Linear Circuit Analysis - DeCarlo-Lin 3e

'
f-. '
LINEAR
CIRCUITS
TIME DOMAIN, PHASOR, AND LAPLACE

TRANSrORM APPROACHES
THIRD
EDITION
Raymond A. DeCarlo
Purdue University
Pen-Min Lin
Purdue University
Kendall Hunt
p u b l i s h i n g
c o m p a n y
n
o
Cover image (^^Jîkiaui
^ Used under license from Shutterstock, Inc.
Kendall Hunft
p u b l i s h i n g
c o m p a n y
www.kendallhunt.cpm
Send all inquiries to:
4050 Westmark Drive
Dubuque, lA 52004-1840
Copyright 2001, 2009 Raymond A. DeCarlo and Pen-Min Lin
Copyright 1995 Prentice-Hall, Inc.
ISBN 978-0-7575-6499-4
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise,
without the prior written permission of the copyright owner.
r^
Printed in the United States of America

10 9 8 7 6 5 4 3
O
TABLE OF CO N TEN TS
Preface......................................................................................................................................................................vii
Chapter 1 Charge, Current, Voltage and Ohms Law ............................................................................ 1
Chapter 2 Kirchhoffs Current & Voltage Laws and Series-Parallel Resistive C ircu its..............51
Chapter 3 Nodal and Loop Analyses....................................................................................................... 107
Chapter 4 T he Operational Amplifier..................................................................................................... 155
Chapter 5 * Linearity, Superposition, and Source Transform ation................................................... 191
Chapter 6 Thevenin, Norton, and Maximum Power Transfer Theorems.................................... 227
Chapter 7 Inductors and C apacitors....................................................................................................... 269
Chapter 8 First Order RL and RC Circuits...........................................................................................321
Chapter 9 Second Order Linear Circuits................................................................................................379
Chapter 10 Sinusoidal Steady State Analysis by Phasor Methods .................................................431
Chapter 11 Sinusoidal State State Power Calculations.......................................................................499
Chapter 12 Laplace Transform Analysis L Basics................................................................................. 543
Chapter 13 Laplace Transform Analysis II: Circuit Applications................................................... 603
Chapter 14 Laplace Transform Analysis III; Transfer Function Applications.............................683
Chapter 15 * Time Domain Circuit Response Computations: The Convolution M ethod...... 763
Chapter 16 Band-Pass Circuits and Resonance....................................................................................811
Chapter 17 * Magnetically Coupled Circuits and Transformers........................................................ 883
Chapter 18 Tw o-Ports...................................................................................................................................959
Chapter 19 Principles o f Basic Filtering ............................................................................................. 1031
Chapter 20 Brief Introduction to Fourier Series .............................................................................. 1085
In d ex................................................................................................................................................................... 1119
o
n
p ,
0
o
n
o
0
o
n
o
o
0
0
0
n
O
n
^
o
PREFACE
For the last several decades, EE/ECE departments o f US universities have typically required two
semesters o f linear circuits during the sophomore year for EE majors and one semester for other
engineering majors. Over the same time period discrete time system concepts and computer engi
neering principles have become required fare for EE undergraduates. Thus we continue to use
Laplace transforms as a vehicle for understanding basic concepts such as impedance, admittance,
fdtering, and magnetic circuits. Further, software programs such as PSpice, MATLAB and its tool
boxes, Mathematica, Maple, and a host o f other tools have streamlined the computational drudg
ery o f engineering analysis and design. MATLAB remains a working tool in this 3'''^ edition o f
Linear Circuits.
In addition to a continuing extensive use o f MATLAB, we have removed much o f the more com
plex material from the book and rewritten much o f the remaining book in an attempt to make the
text and the examples more illustrative and accessible. More importantly, many o f the more diffi
cult homework exercises have been replaced with more routine problems often with numerical
answers or checks.
Our hope is that we have made the text more readable and understandable by todays engineering
undergraduates.
Charge, Current, Voltage

and Ohms Law
CHAPTER O U TLIN E
1.
2.
3.
4.
5.
6.
7.
8.
Role and Importance o f Circuits in Engineering

Charge and Current
Voltage
Circuit Elements
Voltage, Current, Power, Energy, Relationships
Ideal Voltage and Current Sources
Resistance, Ohms Law, and Power (a Reprise)
V-I Characteristics o f Ideal Resistors, Constant Voltage, and
Constant Current Sources
Summary
Terms and Concepts
Problems
CHAPTER O B jEC TIV ES

1.
Introduce and investigate three basic electrical quantities: charge, current, and voltage,
and the conventions for their reference directions.
2.
3.
Define a two-terminal circuit element.

Define and investigate power and energy conversion in electric circuits, and demonstrate
4.
that these quantities are conserved.

Define independent and dependent voltage and current sources that act as energy or sig
5.
6.
7.
8.
nal generators in a circuit.

Define Ohms law, v{t) = R i{t), for a resistor with resistance R.
Investigate power dissipation in a resistor.
Classify memoryless circuit elements by dieir terminal voltage-current relationships.
Explain the difference between a device and its circuit model.
ch ap ter 1 Charge, Current, Voltage and O hm s Law
1. ROLE AND IM PORTANCE OF CIRCUITS IN ENGINEERING

Are you curious about how fuses blow? About the meaning o f different wattages on Hght bulbs?
About the heating elements in an oven? And how is the presence o f your car sensed at a stoplight?
Circuit theory, the focus o f this text, provides answers to all these questions.
W hen you learn basic circuit theory, you learn how to harness the power o f electricity, as is done,
for example, in
an electric motor that runs the compressor in an air conditioner or the pump in a dish
washer;
a microwave oven;
a radio, TV, or stereo;
an iPod;
a car heater.
In this text, we define and analyze common circuit elements and describe their interaction. Our
aim is to create a modular framework for analyzing circuit behavior, while simultaneously devel
oping a set o f tools essential for circuit design. These skills are, o f course, crucial to every electri
cal engineer. But they also have broad applicability in other fields. For instance, disciplines such
as bioengineering and mechanical engineering have similar patterns o f analysis and often utilize
circuit analogies.
W H A T IS A C IR C U IT ?
A circuit is an energy or signal/information processor. Each circuit consists o f interconnections o f
simple circuit elements, or devices. Each circuit element can, in turn, be thought o f as an ener
gy or signal/information processor. For example, a circuit element called a source produces a
voltage or a current signal. This signal may serve as a power source for the circuit, or it may rep
resent information. Information in the form o f voltage or current signals can be processed by the
circuit to produce new signals or new/different information. In a radio transmitter, electricity
powers the circuits that convert pictures, voices, or music (that is, information) into electromag
netic energy. This energy then radi
ates into the atmosphere or into
space from a transmitting antenna.
A satellite in space can pick up this
electromagnetic energy and trans
mit it to locations all over the
world. Similarly, a T V reception
antenna or a satellite dish can pick
up and direct this energy to a T V
set. T h e T V contains circuits
(Figure 1.1) that reconvert the
information within the received
signal back into pictures with
sound.
FIG U RE 1.1 Cathode ray tube with surrounding circuitry for
converting electrical signals into pictures.
Chapter 1 Charge, Current, Voltage and O hm s Law
2. CH A RGE AND CU RREN T

CH A R G E
Charge is an electrical property o f matter. Matter consists o f atoms. Roughly speaking, an atom
contains a nucleus that is made up o f positively charged protons and neutrons (which have no
charge). T he nucleus is surrounded by a cloud o f negatively charged electrons. Th e accumulated
charge on 6.2415 x 10^ electrons equals -1 coulomb (C). Thus, the charge on an electron is
-1 .6 0 2 1 7 6 X 10-19 C.
Particles with opposite charges attract each other, whereas those with similar charges repel. The
force o f attraction or repulsion between two charged bodies is inversely proportional to the square
o f the distance between them, assuming the dimensions o f the bodies are very small compared
with the distance o f separation. Two equally charged particles 1 meter (m) apart in free space have
charges o f 1 C each if they repel each other with a force o f 10^ c^ Newtons (N), where c = 3 x
10^ m/s is the speed o f light, by definition. The force is attractive if the particles have opposite
charges. Notationally, Q will denote a fixed charge, and q or q{t), a time-varying charge.
Exercise. How many electrons have a combined charge o f -5 3 .4 0 6 x 10
C?
AN SW ER; 333,3 9 1 ,5 9 7
Exercise. Sketch the time-dependent charge profile q{t) = 3 (l-^ ^ 0 C, ? > 0, present on a metal
plate. M ATLAB is a good tool for such sketches.
A conductor refers to a material in which electrons can move to neighboring atoms with relative
ease. Metals, carbon, and acids are common conductors. Copper wire is probably the most com
mon conductor. An ideal conductor offers zero resistance to electron movement. Wires are
assumed to be ideal conductors, unless otherwise indicated.
Insulators oppose electron movement. Common insulators include dry air, dry wood, ceramic,
glass, and plastic. An ideal insulator offers infinite opposition to electron movement.
C U R R EN T
Current refers to the net flow o f charge across any cross section o f a conductor. T he net move
ment o f 1 coulomb (1 C) o f charge through a cross section o f a conductor in 1 second (1 sec)
produces an electric current o f 1 ampere (1 A). The ampere is the basic unit o f electric current
and equals 1 C/s.
The direction o f current flow is taken by convention as opposite to the direction o f electron flow,
as illustrated in Figure 1.2. This is because early in the history o f electricity, scientists erroneously
believed that current was the movement o f only positive charges, as illustrated in Figure 1.3. In
metallic conductors, current consists solely o f the movement o f electrons. However, as our under
standing o f device physics advanced, scientists learned that in ionized gases, in electrolytic solu
c h ap ter 1 Charge, Current, Voltage and O hm s Law
tions, and in some semiconductor materials, movement o f positive charges constitutes part or all
o f the total current flow.
One Ampere
of Current "
One
;
;
Cloud o f \
se co n d ^ ....... |---- 6.24x10 1
later
i
;
k electrons
Boundary
FIG U RE 1.2 A cloud o f negative charge moves past a cross section of an ideal conductor from right
to left. By convention, the positive current direction is taken as left to right.
One Ampere
of Current
One
Coulom b
One
of positive
'second
later
charge
Boundary
FIGURE 1.3 In the late nineteenth cenmry, current was thought to be the movement of a positive charge
past a cross section of a conduaor, giving rise to the conventional reference direction of positive current flow.
Both Figures 1.2 and 1.3 depict a current o f 1 A flowing from left to right. In circuit analysis, we
do not distinguish between these two cases: each is represented symbolically, as in Figure 1.4(a).
The arrowhead serves as a reference for determining the true direction o f the current. A positive
value o f current means the current flows in the same direction as the arrow. A current o f negative
value implies flow is in the opposite direction o f the arrow. For example, in both Figures 1.4a and
b, a current o f 1 A flows from left to right.
1A
-1A
>
<
(a)
(b)
FIG U RE 1.4 1 A of current flows from left to right through a general circuit element.
In Figure 1.4, the current is constant. The wall socket in a typical home is a source o f alternating
current, which changes its sign periodically, as we will describe shortly. In addition, a current direc
tion may not be known a priori. These situations require the notion o f a negative current.
E X A M P L E 1.1.
Figure 1.5 shows a slab o f material in which the following is true:
1. Positive charge carriers move from left to right at the rate o f 0.2 C/s.
2. Negative charge carriers move from right to left at the rate o f 0.48 C/s.
Given these conditions,
a) Find
and /^;
b)
Describe the charge movement on the wire at the boundaries A and B.

B
Connecting
o
0
Connecting
wire
wire
Sem iconductor iVlaterial
F IG U R E 1.5 Material through which positive and negative charges move.

S o lu tio n
a)
The current from left to right, due to the movement o f the positive charges, is 0.2 A. The
current from left to right, due to the movement o f the negative charges, is 0.48 A.
Therefore, /^, the total current from left to right, is 0.2 + 0.48 = 0.68 A. Since ly is the
current from right to left, its value is then -0 .6 8 A.
b)
T he wire is a metallic conductor in which only electrons move. Therefore, at boundaries

A and B, negative charges (carried by electrons) move from right to left at the rate o f 0.68
C/s.
Exercise. In Example 1.1, suppose positive-charge carriers move from right to left at the rate o f 0.5
C/s, and negative carriers move from left to right at the rate o f 0.4 C/s. Find
and
AN SW ER: /, = - 0 .9 A; ^ = 0.9 A
If a net charge
crosses a boundary in a short time frame o f At (in seconds), then the approxi
mate current flow is

/=
Aq
At
( 1 . 1)
where I, in this case, is a constant. The instantaneous (time-dependent) current flow is the limit
ing case o f Equation 1.1, i.e.,
dq{t)
dt
( 1. 2)
Here q{t) is the amount o f charge that has crossed the boundary in the time interval [tQ, t] . The
equivalent integral counterpart o f Equation 1.2 is
q{t) = J i{r)dr
(1.3)
E X A M P L E 1.2
The charge crossing a boundary in a wire is given in Figure 1.6(a) for ? > 0. Plot the current i{t)
through the wire.
(a)
(b)
FIG U RE 1.6 (a) Charge crossing a hypothetical boundary; (b) current flow
associated with the charge plot o f (a).
S o lutio n
As per Equation 1.2, the current is the time derivative o f q{t). The slopes o f the straight-Une seg
ments o f q{f} in Figure 1.6(a) determine the piecewise constant current plotted in Figure 1.6(b).
l-cos(co?)
Exercise. The charge crossing a boundary in a wire varies as q[t) = ---------------- C, for t >Q.
Compute the current flow.
A N SW ER: sin(cof) A, for f > 0
Exercise. Repeat the preceding exercise if q{t) = 5e

A N SW ER:
C, for t > 0 .
A, for f > 0
E X A M PLE 1.3
Find q{t), the charge transported through a cross section o f a conductor over [0, f], and also the total
charge Q transported, if the current dirough the conductor is given by die waveform o f Figure 1.7(a).
-l-*-t(se c)
FIG U RE 1.7 (a) Square-wave current signal; (b) q{t) equal to the integral of i{t) given in (a).
S o lutio n
From Equation 1.3, for t>Q,

q{t)=p{T)clT
Thus, q{t) is the running area under the i{t) versus t curve. Since i{t) is piecevv'ise constant, the
integral is piecewise linear because the area either increases or decreases linearly with time, as
shown in Figure 1.7(b). Since q{t) is constant for ^ > 3, the total charge transported is Q = q{5) =
3 C.
Exercise. If the current flow through a cross section o f conductor is i{t) = cos(120jtf) A for ? > 0
and 0 otherwise, find q{t) for t>Qi.
AN SW ER: q{t)
120jt
C for r > 0
Exercise. Suppose the current through a cross section o f conductor is given in Figure 1.8. Find
q{t) for t > 0 .
FIGURE 1.8
AN SW ER; q(t) =
C for 0 <
1; q{t) = IC for r > I
T Y P ES OF C U R R EN T
There are two very important current types: direct current (do) and alternating current (ac).
Constant current (i.e., dqldt = / is constant) is called direct current, which is illustrated graphi
cally m Figure 1.9(a). Figure 1.9(b) shows an alternating current, generally meaning a sinusoidal
waveform, i.e., current o f the form y4sin(w? + ()>), where A is the peak magnitude, co is the angu
lar frequency, and (|) is the phase angle o f the sine wave. W ith alternating current, the instanta
neous value o f the waveform changes periodically through negative and positive values, i.e., the
ch a p ter 1 Charge, Current, Voltage and O h m s Law
direction o f the current flow changes regularly as indicated by the + and - values in Figure 1.9(b).
Household current is ac.
Lastly, Figure 1.9(c) shows a current that is neither dc nor ac, but that nevertheless will appear in
later circuit analyses. There are many other types o f waveforms. Interestingly, currents inside com
puters, C D players, TV s, and other entertainment devices are typically neither dc nor ac.
i(t) (A)
t(sec)
-H ----------------------1 -
-I-----
3
(a)
F IG U R E 1.9 (a) Direct current, or dc; i{t) = Iq\ (b) alternating current, or ac;
i{t) = 1 2 0-^ sin (1 2 0 ?) A; (c) neither ac nor dc.
10
Because the value o f an ac waveform changes with time, ac is measured in different ways. Suppose
the instantaneous value o f the current at time t is A!sin(ci)i- + (j>). The term peak value refers to K
in K sin(co? + (j)). The peak-to-peak value is 2K. Another measure o f the alternating current,
indicative o f its heating effect, is the root mean square (rms), or effective value. The rms or effec
tive value is related to the peak value by the formula
rms =
X peak-value = Q .lO llK
(i.4)
A derivation o f Equation 1.4 with an explanation o f its meaning will be given in Chapter 11.
A special instrument called an ammeter measures current. Some ammeters read the peak value,
whereas some others read the rms value. One type o f ammeter, based on the interaction between
the current and a permanent magnet, reads the average value o f a current. From calculus, Fave!
the average value o f any function y(^), over the time interval [0, 7] is given by
(1.5)
For a general ac waveform, the average value is zero. However, ac signals are often rectified, i.e.,
converted to their absolute values, in power-supply circuits. For such circuits, the average value o f
the rectified signal is important. From Equation 1.5, the average value o f the absolute value o f an
ac waveform over one complete cycle with T = 2jt/co, is
K ^
2.K
Average Value = ^\s,m{wt)\dt = ----- J
0
2K -cos{(ot)
T
(O
0.5T
sin(cot)clt
^ 0
2,K
= 0.636K
jt
( 1.6)
i.e., 0 .636 X peak value.
Exercise. Suppose i{t) - 169.7 sin(50jtr) A. Find the peak value, the peak-to-peak value, the rms
value o f i{t), and the average value o f
AN SW ER: 169.7, 339.4, 120, and 107.93 A, respectively
3. VO LTAG E
W hat causes current to flow? An analogous question might be. W hat causes water to flow in a pipe
or a hose? W ithout pressure from either a pump or gravity, water in a pipe is still. Pressure from
a water tower, a pressured bug sprayer tank, or a pump on a fire truck will force water flow In
electrical circuits, the pressure that forces electrons to flow, i.e., produces a current in a wire or
a device, is called voltage. Strictly speaking, water flows from a point o f higher pressure say,
p o in ts to a point o f lower pressure say, point 5 along a pipe. Between the two points and
B, there is said to be a pressure drop. In electrical circuits, a voltage drop from point A to point B
11
along a conductor will force current to flow from point A to point B; there is said to be a voltage
drop from point A to point B in such cases.
Gravity forces the water to flow from a higher elevation to a lower elevation. An analogous phe
nomenon occurs in an electric field, as illustrated in Figure 1.10(a). Figure 1.10(a) shows two con
ducting plates separated by a vacuum. O n the top plate is a fixed amount o f positive static charge.
On the bottom plate is an equal amount o f negative static charge. Suppose a small positive charge
were placed between the plates. This small charge would experience a force directed toward the
negatively charged bottom plate. Part o f the force is due to repulsion by the positive charges on
the top plate, and part is due to the attraction by the negative charges on the bottom plate. This
repulsion and attraction marks the presence o f an electric field produced by the opposite sets o f
static charges on the plates.
The electric field indicated in Figure 1.10 sets up an electric pressure or voltage drop from the
top plate to the bottom plate, which forces positive charges to flow downhill in the way that
water flows from a water tower to your faucet. Unlike water flow, negative charges are forced
uphill from the negatively charged bottom plate to the positively charged top plate. As men
tioned in the previous section, this constitutes a net current flow caused by the bilateral flow o f
positive and negative charges. The point is that current flow is induced by an electric pressure
called a voltage drop.

A 0
Positive
negative
charge, q
Electric Field
Force on
Electric Field
charge q
charge
Negative
B charge,-q
B
0
Force on

(a)

(b)
FIGURE 1.10 (a) Positive charge in a (uniform) electric field; (b) negative charge in a uniform elearic field.
As mentioned, in Figure 1.10, the positive charge ^ at ^ tends to move toward B. We say, quali
tatively, that point A in the electric field is at a higher potential than point B. Equivalently, point
5 is at a lower potential than point A. An analogy is now evident: a positive charge in an electric
field falls from a higher potential point to a lower potential point, just as a ball falls from a high
er elevation to a lower elevation in a gravitational field.
Note, however, that if we turn the whole setup o f Figure 1.10(a) upside down, the positive charge
q still moves from point A to point B, an upward spatial movement. Similarly, if a negative charge
- q is placed at B, as in Figure 1.10(b), then the negative charge experiences an upward-pulling
force, moving from the lower potential, point B, to the higher potential, point A.
n
12
Chapter 1 Charge, Current, Voltage and Ohms Law
-----------------------------------------------------------------------------------------------------------------------
Again, consider Figure 1. 10(a). As the charge q moves from point ^ toward B, it picks up veloci
ty and gains kinetic energy. Just before q hits the bottom plate, the kinetic energy gained equals
the (constant) force acting on q multiplied by the distance traveled in the direction o f the force. The
kinetic energy is proportional to q and to
the distance traveled. Therefore,
energy converted = kinetic energy gained oc q
The missing proportionality constant in this relationship is defined as the potential difference or
voltage between A and B, The term voltage is synonymous with potential difference.
Mathematically,
,
. ,
voltage = potential difference =
energy converted
magnitude of charge
( 1.8)
The standard unit for measuring potential difference or voltage is the volt (V). According to
Equation 1.8, i f 1 joule {]) o f energy is convertedfrom one form to another when moving 1 C o f charge
from point K to point B, then the potential difference, or voltage, between A and B w i VTIn equation
form, with standard units of V, J, and C, we have
1V = 1 ^
(1.9)
The use of terms such as elevation diflFerence, energy converted, potential difference, or
voltage implies that they all have positive values. If the word difference is changed to drop
(or to rise), then potential drop and elevation drop have either positive or negative values, as the
case may be. The following four statements illustrate this point in the context of Figure 1.10:
The voltage between (or across) A a n d 5 is 2 V.
The voltage between (or across) B and A is 2 V.
' The voltage drop from A to B is 2W.
The voltage drop from B to A Is - 2 V.
This discussion describes the phenomena of voltage. Voltage causes current flow. But what pro
duces voltage or electric pressure? Voltage can be generated by chemical action, as in batteries. In
a battery, chemical action causes an excess of positive charge to reside at a terminal marked with
a plus sign and an equal amount of negative charge to reside at a terminal marked with a negative
sign. When a device such as a headlight is connected between the terminals, the voltage causes a
current to flow through the headlight, heating up the tiny wire and making it Ught up. Another
source of voltage/current is an electric generator in which mechanical energy used to rotate the
shaft of the generator is converted to electrical energy using properties of electro-magnetic fields.
All types of circuit analysis require knowledge of the potential difference between two points, say
^
^
A and B, and specifically whether point A or point 5 is at a higher potential. To this end, we speak
of the voltage drop from point A to point B, conveniently denoted by a double-subscript, as Vj^.
If the value of
is positive, then point ^ is at a higher potential than point B. On the other
hand, if
is negative, then point 5 is at a higher potential than point A. Since
stands for
the voltage drop from point B to point A,
The double-subscript convention is one o f three methods commonly used to unambiguously specify
a voltage drop. Using this convention requires labeling all points o f interest with letters or integers so
that
KiO ^12
^13
sense. A second, more-common convention uses + and - markings
on two points, together with a variable or numerical labeling o f the voltage drop from the point
marked + to the point marked - . Figure 1.11 illustrates this second convention, where Vq denotes the
voltage drop from A (marked +) to B (marked - ) . If Vq is positive, then ^4 is at a higher potential than
B. O n the other hand, if Vq is negative, then 5 is at a higher potential than A. The value o f Vq, togeth
er with the markings + and
stipulates which terminal is at a higher potential; neither alone can do
this. For a general circuit element, the (+, ) markings that is, the reference directions can be
assigned arbitrarily. A third method for specifying a voltage drop, using a single subscript, will be dis
cussed in Chapter 2.
B
-I-
FIGURE 1.11 The + and - markings establish a reference direction for voltage drop. For accuracy,
always place the (+, - ) markings reasonably close to the circuit element to avoid uncertainty.
The following example illustrates the use o f the double subscript and the (+, - ) markings for des
ignating voltage drops.
E X A M P L E 1.4
Figure 1.12 shows a circuit consisting o f four general circuit elements, with voltage drops as indi
cated. Suppose we know that
= 4 V, and
= 9 V. Find the values o f
V^q and
CD-
-I-
3V
FIG U RE 1.12 Arbitrary circuit elements for exploring the use of (+, - ) for specifying a voltage drop.
14
S o lutio n
T he meaiiing o f the double subscript notation and the (+, - ) markings for a voltage imply that
'DA
^ 5 C = 3I V
^CZ> = - ^ Z )C = -(-2 ) = 2 V
Exercise. In Figure 1.12, find
and Vp.^-
A N S W E R :- 3 V ; - 2 V
Exercise. T he convention o f the (+, - ) markings is commonly used as described. Figure 1.13 shows
an old 12-V automobile battery whose (+, - ) markings cannot be seen because o f the corrosion o f
the terminals. A digital voltmeter (DVM ) is connected across the terminals, as shown. The display
reads -1 2 V. Figure out the (+, - ) marking o f the battery terminals.
A N SW ER: left terminal,
right terminal, +
DVM
12V
battery
FIG U RE 1.13 Digital voltmeter connected to a 12-V (car) battery whose plus and
minus markings have corroded away.
One final note: As with current, there are different types o f voltages dc voltage, ac voltage, and
general voltage waveforms. Figure 1.9, with the vertical axis relabeled as v{t), illustrates different
voltage types.
4. C IR CU IT ELEM EN TS
Circuits consist o f interconnections o f circuit elements. The most basic circuit element has two
terminals, and is called a two-terminal circuit element, as illustrated in Figure 1.14. A circuit eie-
Chapter 1 Charge, Current, Voltage and O h m s Law
15
ment called a source provides either voltage, current, or both. The battery is a very common
source, providing nearly constant voltage and the usually small current needed to operate small
electronic devices. Car batteries, for example, are typically 12 volts and can produce large currents
during starting. The wall outlet in a home can be thought o f as a 110 -volt ac source. Figure 1.14(a)
shows a (battery) voltage
across a general undefined circuit element. A current z(r) flows
through the element. Recall from our earlier intuitive discussion that voltage is analogous to water
pressure: pressure causes water to flow through pipes; voltage causes current to flow through cir
cuit elements. Total water into a pipe equals total water out o f the pipe. Analogously, the current
entering a two-terminal device must, by definition, equal the current leaving the two-terminal device.
Current
FIGURE 1.14 (a) General circuit element (connected to a battery) as an energy

or signal processor: v(i) is the voltage developed across the circuit element, and z'(r)
is the current flowing through the circuit element; (b) practical example of a
general circuit element (car headlight) connected to a car battery.
The circuit element o f Figure 1.14(a) has a specific labeling: the current i(f) flows from the plus
terminal to the minus terminal through the circuit element. Such a labeling o f the voltage-current
reference directions is called the passive sign convention. In contrast, the current iij) flows from
the minus terminal to the plus terminal through the battery; this labeling is conventional for
sources but not for non-source circuit elements.
In addition to sources, there are other common two-terminal circuit elements:
The resistor
The capacitor
The inductor
For a resistor, the amount o f current flow depends on a property called resistance; the smaller the
resistance, the larger the current flow for a fixed voltage across the resistor. A small-diameter pipe
offers more resistance to water flow than a large-diameter pipe. Similarly, different types o f con
ductors offer different resistances to current flow. A conductor that is designed to have a specific
resistance is called a resistor. If the device is an ideal resistor, then v(f) = Ri{i), where i? is a con
stant o f resistance. More on this shortly.
The circuit elements called the capacitor and the inductor will be described later in the text. Also,
future chapters will describe the operational amplifier and the transformer that are circuit elements
having more than two terminals.
16
5. VO LTAG E, CURRENT, POW ER, ENERGY, RELATIONSHIPS

The relationship between voltage across and current through a two-terminal element determines
whether power (and, thus, energy) is delivered or absorbed. The heating element in an electric
oven can be thought o f as a resistor. The heating element absorbs electric energy and converts it
into heat energy that cooks, among other things, turkey dinners.
In Figure 1.14(a), a battery is connected to a circuit element. Figure 1.14(b) concretely illustrates
this with a 12-V car battery connected to a headlight. W ith reference to Figures 1.14(a) and
1.14(b), suppose v{t) = 12 V, and i{t) = 5 A: 5 A o f current flows through the headlight. The head
light converts electrical energy into heat and light. Power (in watts) is the rate at which the ener
gy is converted. At each instant o f time, the electrical power delivered to (absorbed by) the head
light is pit) = v[t)i{t) - 12 X 5 = 60 watts. Similarly, at each instant o f time, the battery can be
viewed as delivering 60 watts o f power to the headlight. Inside the battery, the stored potential
energy o f the chemicals and metals undergoes a chemical reaction that produces the electrical
potential difference and the current flow to the headlight: chemical energy is converted into elec
trical energy that is converted into light and heat.
Figure 1.15 depicts a more general scenario: a circuit element is connected to its surrounding cir
cuit at points A and B. (One, o f course, could imagine that the remainder o f circuit is a battery,
and circuit element 1 is a headlight.) Suppose there is a constant voltage drop from A to B, denot
ed by
Also assume that a constant current
flows from terminal A to terminal B through
circuit element 1, as shown.
FIG U RE 1.15 A general circuit in which a two-element circuit element is extracted

and labeled according to the passive sign convention.
For discussion purposes, assume
> 0 and
> 0. During a time interval o f T s, (V^g x T) C o f
charge moves through circuit element 1 from A to B. In falling from a higher potential, point A, to
a lower potential, point B, the charge loses electric potential energy. The lost potential energy is con
verted within element 1 into some other form o f energy heat or light being two o f several possibil
ities. According to Equation 1.8, the amount o f energy converted {absorbed by the element) is
y. T) >Q. The power absorbedhj element 1 is, by definition, the rate at which it converts or absorbs
energy. This rate equals
â b (âb ^ T)
^Vab I a b > 0 .
T
17
Exercise. In Figure 1.15, the current ÂB - 5 niA, and
= 400 V. W hat is the energy absorbed

by circuit element 1 in one minute? W hat is the power absorbed by circuit element 1?
AN SW ER: W = 120 J; P = 2 watts
W ith respect to Figure 1.15, for constant (direct) voltages and currents, we arrive at a very simple
relationship:
P\-V ab I ab
where
0 -1 0 )
is the power (in W ) absorbed by the circuit element. Consequently, the energy, W , (in
J), absorbed during the time interval Tis
W^=P\xT
(1-11)
Now, let us reconsider Figure 1.15. One can think o f-/ ^ g as flowing from A w B through the
remainder o f the circuit. In this case,
-1 ^ ^ ^ ^ < 0 . This means that the remain
der o f the circuit absorbs negative power or equivalently delivers |^ 5 (
| = â ÂB
circuit
element 1. As such, the remainder o f the circuit is said to generate electric energy. By definition,
the electric power generated by the remainder o f the circuit is the rate at which it generates elec
tric energy. From Equation 1.8, this rate equals
--------- ---------- - ^ ab âb
Observe that the rate at which the remainder o f the circuit generates power precisely equals the
rate at which circuit element 1 absorbs power. This equality is called the principle o f conserva
tion o f power: total power generated equals total power absorbed. Equivalently, the sum o f the
powers absorbed by all the circuit elements must add to zero,
Exercise. In Figure 1.15, -Pg =
+ Pq = y^gl^B
ÂB^^ÂB^ ~
watts, i.e., the remainder o f the circuit absorbs - 1 0 watts o f
power. How much power does circuit element 1 absorb?

A N SW ER: 10 watts
In general, whenever a two-terminal general circuit element is labeled according to the passive sign
convention, as in Figure 1.15, then P =
whereas
P = V^b âb ^
> 0 means the element absorbs (positive) power,
absorbs negative power or delivers (positive) power to whatever it
is connected. As a general convention, non-source circuit elements are labeled according to the
passive sign convention. Usually, sources are labeled with the current leaving the terminal labeled
with +. For such labeling o f sources, if the product o f the source voltage and the current leaving
the + terminal is positive, then the source is delivering power to the network.
Chapter 1 Charge, Current, Vohage and O hm s Law
RULE FOR C A LC U LA TIN G A B SO R B ED PO W ER

The power absorbed by any circuit element (Figure 1.16) with terminals labeled A and B is
equal to the voltage drop from A m B multiplied by the current through the element from A
to 5,
V AB
FIGURE 1.16
Exercise. Compute the power absorbed by each o f the elements in Figure 1.17.
-1A
_____________
-2A
_____________
2A
>
< Z3
10V
10V
10V
(a)
(b)
(c)
FIGURE 1.17
AN SW ER: (a) 10 W; (b) - 2 0 W; (c) 20 W
As mentioned, power is the rate o f change o f work per unit o f time. T he ability to determine the
power absorbed by each circuit element is highly important because using a circuit element or
some device beyond its power-handling capability could damage the device, cause a fire, or result
in a serious disaster. This is why households use circuit breakers to make sure electrical wiring is
not overloaded.
Exercise. In Figure 1.18, a car heater is attached to a 12-volt D C voltage source. How much power
can the car heater absorb before the 20 -amp fuse blows.
20 Amp Fuse
FIGURE 1.18 Car heater connected to a 12-volt car battery through a 20-amp fuse.
A N SW ER: 240 watts
19
As mentioned earlier, the calculated value o f absorbed power P may be negative. If the absorbed
power P is negative, then the circuit element actually generates power or, equivalently, delivers
power to the remainder o f the circuit. In any circuit, some elements will have positive absorbed
powers, whereas some others will have negative absorbed powers. If one adds up the absorbed
powers o f ALL elements, the sum is zero! This is a universal property called conservation o f
power.
PRIN C IPLE OF CO N SERVA TIO N OF PO W ER

The sum o f the powers absorbed by all elements in a circuit is zero at any instant o f time.
Equivalently, the sum o f the absorbed powers equals the sum o f the generated powers at each
instant o f time.
The 2"*^ edition o f this text contains a rigorous proof o f this principle. For the present, we will
simply use it to solve various problems. The following example will help clarify the sign conven
tions and illustrate the principle o f conservation o f power.
E X A M P L E 1.5
Light bulbs come in all sorts o f shapes, sizes, and wattages. W a t t l e measures the power consumed
by a bulb. Typical wattages include 15, 25, 40, 60, 75, and 100 W. Power consumptions differ
because the current required to light a higher-wattage (and brighter) bulb is larger for a fixed out
let voltage: a higher-wattage bulb converts more electric energy into light energy. In Figure 1.19,
the source delivers 215 watts o f power. W hat is the wattage o f the unlabeled bulb?
7?
100V
watts
watts
watts
FIG U RE 1.19. Three bulbs connected to a 100-V battery.

S o lutio n
From conservation o f power, the total power delivered by the battery equals the total power
absorbed by all the bulbs. Therefore, the power absorbed by the unknown bulb is
215 - 4 0 - 100 = 75 watts
Exercise. Determine the current / leaving the battery in Example 1.5.

AN SW ER: 2.15 amps
20
EXA M PLE 1.6

An electroplating apparatus uses electrical current to coat materials with metals such as copper or
silver. In Figure 1.20, suppose a 2 2 0 -V electrical source supplies 10 A dc to the electroplating
apparatus.
10A
Electroplating
Apparatus
FIGURE 1.20 Electrical source operating an electroplating apparatus.

a)
b)
W hat is the power consumed by the apparatus?

If electric energy costs 10 cents per kilowatt-hour (kW h), what will it cost to operate the
apparatus for a single 12 -h day?
S o lutio n
Step 1. From Equation 1.10, the power consumed is

/> = 220
10 = 2 2 0 0 W, or 2.2 kW
Step 2. According to Equation 1.11, the energy consumed per 12-h period is
2.2
12 = 26.4 kWh
Step 3. Therefore, the cost to operate is

26 .4
.01 = $ 2.64 / day
Exercise. Suppose the electroplating apparatus o f Example 1.6 draws 12 A D C at the same volt
age. W hat is the cost o f operation for a single 12-h day? W hat is the cost o f operating for a 20
workday month?
AN SW ER: $3,168; $63.36
E X A M PLE 1.7
Each box in the circuit o f Figure 1.21 is a two-terminal element. Compute the power absorbed by
each circuit element. W hich elements are delivering power? Verify the conservation o f power prin
ciple for this circuit.
Chapter 1 Charge, Current, Voltage and O hm s
21
FIG U RE 1.21 Circuit containing several general circuit elements.

S o lutio n
Step 1. Compute power absorbed by each element. Using either Equation 1.10 or the power con
sumption rule, the power absorbed by each element is
a)
For element 1 P i = 4 X 1 = 4 W
b)
For element 2 P l = 8 x 2 = 1 6 W
c)
For element 3 ^ 3 = 10 X 1 = 10 W
d)
For element 4
e)
For element 5 P 5 = 2
For element 6 Pe = 1 0 X ( - 2 ) = - 2 0 W
14
x
(-1)=-14W
2 = 4W
Step 2 . Verify conservation o f power. Since P 4 and Pg are negative, element 4 delivers 14 W, and
element 6 delivers 20 W o f power. T he remaining four elements absorb power. Observe that the
sum o f the six absorbed powers, 4 + 16 + 10 - 14 + 4 - 2 0 = 0, as expected from the principle o f
conservation o f power. Equivalently, the total positive generated power, (14 + 20) = 34 W, equals
the total positive absorbed power, (4 + 16 + 10 + 4) = 34 W.
Exercise. In Figure 1.22, find the powers absorbed by elements 1, 2, and 3.
FIG U RE 1.22
AN SW ER: 8 W, 20 W, - 2 8 W; element 3 equivalently delivers 28 W
22
Exercise. In Figure 1.22, suppose the current 2 A were changed to - 4 A. W hat is the new power
absorbed by element 3?
A N SW ER: 56 watts
If the power absorbed by a circuit element is positive, the exact nature o f the element determines the
type o f energy conversion that takes place. For example, a circuit element called a resistor (to be dis
cussed shortly) converts electric energy into heat. If the circuit element is a battery that is being
charged, then electric energy is converted into chemical energy within the battery. If the circuit ele
ment is a dc motor turning a fan, then electrical energy is converted into mechanical energy.
N O N -D C PO W ER A N D EN ER G Y C A LC U LA TIO N S
Consider Figure 1.23, where i{t) is an arbitrary time-varying current entering a general two-ter
minal circuit element, and v{t) is the time-varying voltage across the element. Because voltage and
current are functions o f time, the power p{t) = v{t)i{t) is also a function o f time. For any specific
value o f ^ = ?j, the value p{t^) indicates the power absorbed by the element at that particular
time hence, the terminology instantaneous power for p{t).
i(t)
Circuit Elem ent

Absorbing Power
p(t)
FIGURE 1.23 Calculation of absorbed power for time-varying voltages and currents for circuit ele
ments labeled with the passive sign convention; here, power is p{t) = v{t)i{t).
Equation 1.12 extends Equation 1.10 in the obvious way.
p{t) = v{t)i{t)
( 1. 12)
i.e., the instantaneous (absorbed) power p{t), in W, is the product o f the voltage v{t), in V, and
the current i{t), in A, with labeling according to the passive sign convention. This product also
makes sense from a dimensional point o f view;
,
volts X amps =
joules
coulombs
joules
; - x
=
coulomb
second
second
Knowing the power p{t) absorbed by a circuit element as a function o f t allows one to compute
the energy W{tQ, t) absorbed by the element during the time interval [^q, t > Iq], W[tQ, t) (J) is the
integral o f p{t) (W) with respect to t over [?q, t], i.e..
23
W(to,t)^ r p ir ) d r
where
the lower limit o f the integral, could possibly be -oo. For the dc case, p{t) = P (a con
stant). From Equation 1.13,
t
t
W(tQ,t) = f p ( r ) d T = P f d r = P(t-tQ) = P x T
where T = t - t^, as given in Equation 1.11. If, in Equation 1.13, tg = -oo, then W (-co, t) becomes
a function only o f t which, for convenience, is denoted by
t
W{t)= f p ( r ) d r
L
(1.14)
W{t) = W{00, t), in joules, represents the total energy absorbed by the circuit element from the
beginning o f time to the present time rwhen p{t) is in watts.
Exercise, a) Suppose the power absorbed by a circuit element over [0,oo) is p{i) =
W (0, oo). b) Now suppose the absorbed power o f the circuit element is
p{t) =
j >0
watts. Find
for t > 0 .
A N SW ER: 4 J; (4+t) J
Since energy is the integral o f power, power is the rate o f change (derivative) o f energy.
Differentiating both sides o f Equation 1.14 yields the expected equation for instantaneous power.
dW(t)
v m o = P ( o = ^
(1.15a)
or, equivalently, for t > (q,
Exercise. Suppose that for t > 0 , the work done by an electronic device satisfies W{t) = 10(1
J- If
the voltage supplied by the device is 10 V, then for t > 0, find the power and current supplied by the
device, assuming standard labeling, i.e., the passive sign convention.
AN SW ER: p{t) = \0e- watts; i(f) = e' A
24
EXA M PLE 1.8

In the circuit o f Figure 1.23, the current i{t) and vokage v{t) have the waveforms graphed in Figure
1.24. Sicetch p{t), the instantaneous power absorbed by the circuit element, and then sketch W(0,
t), the energy absorbed over the interval [0 , i\.
FIGURE 1.24 (a) Current and (b) voltage profdes with respect to t for circuit o f Figure 1.23.
S o lutio n
A simple graphical multiplication o f Figures 1.24(a) and (b) yields the sketch o f the curves in
instantaneous power shown in Figure 1.25(a). From Equation 1.13 with
= 0, we have, for 0 <
t< %
,2
p(T)dr - J c/t =
0
and for t> 5,

t
W ( 0 , t ) - J p(r)dT = J p{x)dT + J p(r)dT - 5 + J ' d T - 5 + ( t - 5 ) = l

0
Figure 1.25(b) presents the resulting graph.
25
(b)
FIGURE 1.25 (a) Profile of the instantaneous power p{t) = v{t)i{t) for the current and voltage wave
forms of Figure 1.24; (b) associated profde of energy versus time.
6. IDEAL VO LTAG E AND CU RREN T SO URCES

Two-terminal circuit elements may be classified according to their terminal voltage-current rela
tionships. The goal o f this section is to define ideal voltage and current sources via their termi
nal voltage-current relationships.
The wall socket o f a typical home represents a practical voltage source. After flipping the switch
on an appliance plugged into a wall socket, a current flows through the internal circuitry o f the
appliance, which, for a vacuum cleaner or dishwasher, converts electrical energy into mechanical
energy. For modest amounts o f current draw (below the fuse setting), the voltage nearly maintains
its nominal pattern o f 120 / 2 sin(120 lit) = 169.7 sin(120 nt) V. This practical situation is ide
26
alized in circuit analysis by the ideal voltage source symbol shown in Figure 1.26(a), a circle with
a reference inside. The symbol is more commonly referred to as independent voltage source.
FIGURE 1.26 Equivalent representations of ideal voltage source attached to a hypothetical circuit.
The waveform or signal v{t) in Figure 1.26 represents the voltage produced by the source at each
time t. The plus and minus (+, ), on the source define a reference polarity. T he reference polari
ty is a labeling or reference frame for standardized voltage measurement. T he reference polarity
does not mean that v(t) is positive. Rather, the reference polarity (+, - ) means that the voltage drop
from + to - is v{t), whatever its value/sign. Finally, the voltage source is ideal because it maintains
the given voltage v{t), regardless o f the current drawn from the source by the attached circuit.
voltage (V)
V,
1(A)
(b)
FIGURE 1.27 (a) Ideal battery representation of ideal voltage source; (b) v-i characteristic of ideal battery.
Figure 1.2 7 (a) shows a source symbol for an ideal battery. The voltage drop from the long-dash
side to the short-dash side is Vg, with Vjj > 0. In commercial products, the terminal marked with
a + sign corresponds to the long-dash side o f Figure 1.27(a). An ideal battery produces a constant
voltage under all operating conditions, i.e., regardless o f current drawn from an attached circuit
or circuit element, as indicated by the v-i characteristic o f Figure 1.27(b). Real batteries are not
ideal but approximate the ideal case over a manufacturer-specified range o f current requirements.
Practical sources (i.e., non-ideal); voltage sources, such as commercial dc and ac generators; and
real batteries deviate from the ideal in many respects. One important respect is that the terminal
voltage depends on the current delivered by the source. The most common generators convert
mechanical energy into electrical energy, while batteries convert chemical energy into electrical
27
energy. There are two general battery categories: nonrechargeable and rechargeable. A discussion
o f the dramatically advancing battery technology is beyond the scope o f this text.
Besides batteries and ideal voltage sources, devices called ideal or independent current sources
maintain fixed current waveforms into a circuit, as illustrated in Figure 1.28. T he symbol o f an
ideal current source is a circle with an arrow inside, indicating a reference current direction. An
ideal current source produces and maintains the current i{t) under all operating conditions. O f
course, the current i{t) flowing from the source can be a constant (dc), sinusoidal (ac), or any other
time-varying function.
FIGURE 1.28 Equivalent ideal current sources whose current i{t) is maintained
under all operating conditions o f the circuit.
In nature, lightning is an example o f an approximately ideal current source. W hen lightning
strikes a lightning rod, the path to the ground is almost a short circuit, and very little voltage is
developed between the top o f the rod and the ground. However, if lightning strikes a tree, the path
o f the current to the ground is impeded by the trunk o f the tree. A large voltage then develops
from the top o f the tree to the ground.
Independent sources have conventional labeling, as shown in Figure 1.29, which is different from
that o f the passive sign convention. Here the source delivers power if p{t) = v{t)i{t) > 0 and would
absorb power i f p{t) =
< 0. A complicated circuit called a battery charger can deliver ener
gy to a drained car battery. T he car battery, although usually a source delivering power, exempli
fies a source absorbing power from the charger.
FIG U RE 1.29 Common voltage and current source labeling.
28
Another type o f ideal source is a dependent source. A dependent source or a controlled source
produces a current or voltage that depends on a current through or voltage across some other ele
ment in the circuit. Such sources model real-world devices that are used in real circuits. In the text,
the symbol for a dependent source is a diamond. If a appears inside the diamond, it is a depend
ent voltage source, as illustrated in Figure 1.30. If an arrow appears inside the diamond, it is a
dependent current source, as illustrated in Figure 1.31. In Figure 1.30, the voltage across the dia
mond-shaped source, v{t), depends either on a current, labeled
through some other circuit
device, or on the voltage
across it. If the voltage across the source depends on the voltage v^,
i.e., v{t) = p
then the source is called a voltage-controlled voltage source (VCVS). If the volt
age across the source depends on the current z^, i.e., v{t) =
then the source is called a cur
rent-controlled voltage source (CCVS).
FIGURE 1.30 The right element is a voltage-controlled voltage source (VCVS) if v{t) =
(p is here
dimensionless), or a current-controlled voltage source (CCVS) if v(t) = rî (r^ here has units of ohm).
Exercise. The voltage across a particular circuit element is

element is
= 5 V, and the current through the
0.5 A, using the standard labeling.
a)
If a V CV S (Figure 1.30) with p = 0.4 were associated with the controlled-source branch,
fmd vit).
b)
If a CCV S (Figure 1.30) with
= 3 2 were associated with the controlled branch, fmd
v{t).
ANSW ER: a) 2 V; b) 1.5 V
There is dual terminology for dependent current sources. The configuration o f Figure 1.31 shows
a voltage-controlled current source (VCCS), i.e., i{t) = g^v^, or a current-controlled current
source (CCCS), for which i{t) =
r~\
29
i(t) =
or
Pi
Q V
or
Pi:
-o
FIG U RE 1.31 The right element is a voltage-controlled current source (VCCS) if i{t) =
(g^ has
units o f siemens) or a current-controlled current source (CCCS) if i{t) = |3/^ ((3 is dimensionless).
Source voltages or currents are called excitations, inputs, or input signals. A constant voltage will nor
mally be denoted by an uppercase letter, such as V, Vq,
cally be denoted by /, /g, /p
V^, and so on. A constant current will typi
and so on. The units are volts, amperes, and so on. Smaller and larger
quantities are expressed by the use o f prefixes, as defined in Standard Engineering Notation Table 1.1.
Exercise. The voltage across a particular circuit element is

ment is
= 5, and the current through the ele
= 0.5 A using the standard labeling.
a)
If a VCCS (Figure 1.31) with ^^ = 0.1 S were associated with the controlled-source branch, find
b)
If a CCCS (Figure 1.31) with P = 0.5 were associated with the controlled-source branch, find i{i).
i{i).
AN SW ER; a) 0.5 A; b) 0.25 A
TABLE 1.1. Engineering Notation for Large and Small Quantities
Name
Prefix
Value
femto
10-15
pico
10-12
nano
10-9
micro
10-6
milli
10-3
kilo
103
mega
lO^
gga
109
tera
1012
30
7. RESISTANCE, O HM 'S LAW, AND POW ER (A REPRISE)

Different materials allow electrons to move from atom to atom with different levels o f ease.
Suppose the same dc voltage is applied to two conductors, one carbon and one copper, o f the same
size and shape. Two different currents will flow. T he current flow depends on a property o f the
conductor called resistance: the smaller the resistance, the larger the current flow for a fixed volt
age. The idea is similar to water flow through different-diameter pipes (analogous to electrical con
ductors): for a given pressure, a larger-diameter pipe allows a larger volume o f water to flow and,
therefore, has a smaller resistance than a pipe with, say, half the diameter.
A conductor designed to have a specific resistance is called a resistor. Hence, a resistor is a device
that impedes current flow. Just as dams impede water flow and provide flood control for rivers,
resistors provide a means to control current flow in a circuit. Further, resistors are a good approx
imate model to a wide assortment o f electric devices such as light bulbs and heating elements in
ovens. Figure 1.32(a) shows the standard symbol for a resistor, where the voltage and current ref
erence directions are marked in accordance with xhie.passive sign convention. Figure 1.32(b) pic
tures a resistor connected to an ideal battery.
I
+
R
V
(a)
FIG U RE 1.32 (a) Symbol for a resistor with reference voltage polarity and current direction
consistent with the passive sign convention; (b) resistor connected to an ideal battery.
In 1827, Ohm observed that for a connection like that o f Figure 1.32(b), the direct current
through the conductor/resistor is proportional to the voltage across the conductor/resistor, i.e., I
= V. Inserting a proportionality constant, one can write
or, equivalently,
1 = V GV
^
(1 .1 6a)
V = R1
The proportionality constant R is the resistance o f the conductor in ohms. The resistance R meas
ures the degree to which the device impedes current flow. For conductors/resistors, the ohm (Q)
is the basic unit o f resistance. A two-terminal device has a 1-Q resistance i f a 1-V excitation causes
1-A o f current to flow. In Equation 1.16(a), the proportionality constant is the reciprocal o f R, i.e.,
G = HR, which is called the conductance o f the device. T he unit for conductance according to
the International System o f Units (SI) system is the siemen, S. In the United States, the older term
for the unit o f conductance is the mho ^5, that is, ohm spelled backward, which is still widely
used. In this text, we try to adhere to the SI system. If a device or wire has zero resistance {R = 0)
or infinite conductance {G = t), it is termed a short circuit. On the other hand, if a device or
wire has infinite resistance (zero conductance), it is called an open circuit. Technically speaking,
31
a resistor means a real physical device, with resistance being the essential property o f the device.
In most o f the literature on electronic circuits, resistor and resistance are used synonymously, and
we will continue this practice.
O H M 'S LAW
Ohms law, as observed for constant voltages and currents, is given by Equation 1.16(b), with
its equivalent form in Equation 1.16(a). However, it is true for all time-dependent waveforms
exciting a linear resistor. Thus, we can generalize Equation 1.16 as
v (0 = W
or
i{t) = v{t) = Gv{t)
(1.17b)
according to Figure 1.33, whose voltage-current labeling is consistent with the passive sign con
vention.
i(t)
AO
+
^
v(t)
----- OB
-
FIG U R E 1.33
If either the voltage or the current direction is reversed, but not both, then Ohms law becomes
v(t) = -Ri{t). As an aid in writing the correct v-i relationship for a resistor. Ohms law is stated
here in words:
For a resistor connected between terminals A and B, the voltage drop from A to B is equal to the
resistance multiplied by the current flowing from A to B through the resistor.
Exercise. Find the resistance R for each o f the resistor configurations in Figure 1.34.
AN SW ER: (a) 12 Q ; (b) 3 Q; (c) 6 Q
-1A
+
R
12V
(a)
4A
-
R
12V
-2A
- -
(b)
R
12V
(c)
FIG U RE 1.34
Once the voltage and the current associated with a resistor are known, the power absorbed by the
resistor is easily calculated. Assuming the passive sign convention, then combining Equation 1.12 for
32
power and Ohms law (Equation 1.17), the instantaneous absorbed power is
9
p{t) = v{t)i{t) = i' ^ { t ) R ^ ^ ^

R
(1.18a)
which for the dc case reduces to
Exercise. Find the power absorbed by each o f the resistors in Figure 1.35.
80
+
12V
(a)
4A
-
R
100
(b)
90
-
12V
(c)
FIGURE 1.35
AN SW ER: (a) 18 W; (b) 160 W; (c) 16 W
Equations 1.18(a) and (b) bring out a very important property; a resistor always absorbs power,
dissipating it as heat. Intuitively speaking, electrons that flow through the resistor collide with
other particles along the way. The process resembles the action in a pinball game: the pinball suecessively collides with various pegs as it rolls from a higher to a lower elevation. W ith each colli
sion, part o f the electrons kinetic energy is converted into heat as the voltage pressure continues
to reaccelerate the electron.
Electrical energy that is converted to heat or used to overcome friction is usually called a loss. Such
losses are termed /-squared-i? {f-R) losses because o f the form o f Equation 1.18. On the other
hand, a stoves heating element purposely converts to heat as much electric energy as possible, in
which case, the P-R loss is desirable. This heating effect also proves useful as the basis for the oper
ation o f fuses. A fuse is a short piece o f inexpensive conductor with a very low resistance and a
predetermined current-carrying capacity. When inserted in a circuit, it carries the current o f the
equipment or appliances it must protect. W hen the current rises above the fuse rating, the gener
ated heat melts the conducting metal inside the fuse, opening the circuit and preventing damage
to the more-expensive appliance. Oversized fuses or solid-wire jumpers circumvent safe fuse oper
ation by permitting unsafe operation at overload currents, with consequent electrical damage to
the appliance that may cause overheating and fire.
Resistance o f a conductor depends on the material and its geometrical structure. For a specific
temperature, R is proportional to the length I o f a conductor and inversely proportional to its
cross-sectional area A,
R= p^
(1.19)
33
where the proportionaUty constant p is the resistivity in ohm-meters (Q m). T he resistivity o f

copper at 2 0 C is 1.7 x 10^ Q m. Table 1.2 lists the relative resistivities o f various materials with
respect to copper.
Table 1.2 Resistivities of Various Materials Relative to Copper.
Silver
0.94
Chromium
1.8
Tin
6.7
Copper
1.00
Zinc
3.4
Carbon
2 .4 X 10^
Gold
1.4
Nickel
5.1
Aluminum
1.6
*The resistivity of copper at 20 degrees C is 1.7 x 10"^ Qxm.

EXA M PLE 1.9
Sixteen-gauge (16 AWG) copper wire has a resistance o f 4 .0 9 4 Q for every 1,000 feet o f wire. Find
the resistance o f 100 feet o f 16 AWG aluminum wire and 100 feet o f 16 AWG nickel wire. Then
find the voltage across each wire and the power absorbed (given off as heat) by each wire if a 10A direct current flows through 100 feet o f each wire.
S o lutio n
The resistivities o f aluminum and nickel wire relative to copper are 1.6 and 5.1, respectively.
Hence, 100 feet o f aluminum/nickel wire has a resistance o f
(aluminum) 1.6 x 0.4094 = 0.655 Q
(nickel) 5.1 X 0.4094 = 2.088 Q
Given a 10-A current flowing through 100 feet o f copper, aluminum, and nickel wire, Ohms law
implies
(copper) V = /?/ = 0 .4 0 9 4 x 10 = 4 .0 9 4 V
(aluminum) V = RI = 0 .655 x 10 = 6.55 V
(nickel) V = RI = 2 .088 x 10 - 2 0.88 V
Finally, from Equation 1.18(b), the absorbed power given off as heat is
(copper) P = V I ^ R I - = 0 .4 0 9 4 x 100 = 4 0 .9 4 W
(aluminum) p = VI = RI^^ 0 .655 x 100 = 6 5.5 W
(nickel) P = VI = RI~ = 2 .088 x 100 = 208.8 W
Notice that every 100 feet o f 16 AWG aluminum wire would absorb 65.5 - 4 0.9 = 2 4 .6 W more
power than copper. And nickel wire absorbs even more power:
^ ^
208.8
4 0.94
times more power than copper per unit length. This absorbed power, given off as heat, is why
nickel wire is used for heating elements in toasters and ovens.
34
Chapter 1 Charge, Current, Vokage and O hm s Law
Exercise, (a) If a constant current o f 10 A flows through 1,000 feet o f (16 AWG) copper wire, how
many watts o f heat are generated by the wire?
(b) If the wire o f part (a) were changed to (16 AWG) aluminum, how many watts o f heat would
be generated?
AN SW ER; (a) 409.4 watts; (b) 65 5 .0 4 watts
Temperature also affects resistance. For example, light bulbs have a cold resistance and a hot
resistance o f more importance during lighting. For most metallic conductors, resistance increases
with increasing temperature except carbon, which has a decrease in resistance as temperature
rises. Since resistors absorb power dissipated as heat, they should have adequate physical dimen
sions to better radiate the heat or there must be some external cooling to prevent overheating.
EXA M PLE 1.10
T he hot resistance o f a light bulb is 120 Q. Find the current through and the power absorbed by
the bulb if it is connected across a constant 90-V source, as illustrated in Figure 1.36.
-O 90V
90V
R =120Q
-O -
FIG U RE 1.36 Light bulb and equivalent resistive circuit model.

S o lutio n
Step 1. From Ohms law. Equation 1.16(a),
V
90
/ = - = ----- = 0.75 A
R
120
Step 2 . By Equation 1.18(b), the power absorbed by the lamp is

P = 0.752
120 = 67.5 W
Step 3 . C/?eck conservation o f power. T he power delivered by the source is 90 x 0.75 = 67.5 W.
Therefore, the power delivered by the source equals the power absorbed by the resistor. This ver
ifies conservation o f power for the circuit.
Exercise. In Example 1.10, suppose the battery voltage is cut in half to 60 V. W hat is the power
absorbed by the lamp? W hat is the power delivered by the battery? Repeat with the battery volt
age changed to 120 V.
AN SW ER; 30 watts; 120 watts
35
The following example illustrates power consumption for a parallel connection o f light bulbs.
EXA M PLE 1.11
Figure 1.37 shows four automobile halogen Hght bulbs connected in parallel across a 12-V bat
tery. Find the following;
(a) The effective hot resistance o f each bulb
(b) T he total power delivered by the source
(c) After 700 hours o f operation, the current supplied by the source drops to 11.417 A.
Discover which light bulb has burned out.
ijt )
27 watts
35 watts
50 watts
60 watts
12V
F IG U R E 1.37 Parallel connection of light bulbs.

S o lutio n
(a) From Equation 1.18(b), P - VÎR,
12^
Rxiw -
27
144
^50W =
50
= 5.33D
= 2 .88 Q
144
^ 3 5 W = ^ - 4 .1 1 4 Q
144
- ' 60
= 2.4 Q
(b) The power delivered by the source equals the sum o f the powers consumed by each bulb,
which is 172 W.
(c) Since the current supplied by the source has dropped to 11.417 A, then the power delivered
by the source drops to P;urcenew ~
^ 11.417 = 137 watts, which is 35 watts less than the ear
lier-delivered power o f 172 watts. Hence, the 35-watt bulb has gone dark.
Exercise. Repeat Example 1.11 (a) with the battery voltage changed to 48 V and a new set o f light
bulbs whose operating voltage is 48 V.
AN SW ER;
85.333 Q; R^c,^= 65.83 Q; R^q^ = 4 6 .0 8 Q; R(^^^= 38.4 Q..
E X A M PLE 1.12
W hen connected to a 120-volt source, halogen light bulb number 1 uses 40 watts o f power. W hen
similarly connected, halogen light bulb 2 uses 60 watts o f power.
(a)
(b)
Find the hot resistance o f each bulb.

If the two bulbs are connected in a series, as in Figure 1.38 and placed across the 120-V
source, find the power absorbed by each bulb and the power delivered by the source,
assuming the hot resistances computed in part (a) do not change.
36
(c)
Find the voltage
and V2 across each bulb.
b u ib l
120V
120 V ^ = i-
bulb 2
(a)
(b)
FIGURE 1.38 Series connection o f two light bulbs and equivalent resistive circuit model.
S o lu tio n
Step 1. Find the hot resistances. The hot resistances o f each bulb are given by
120^
Vt
40
bulb\
= 360 and R 2 =
120 "
= 240
Step 2. Find the current through each bulb, the power absorbed by each bulb, and the power delivered
by the source. The circuit o f Figure 1.38(a) has the equivalent representation in terms o f resistanc
es in Figure 1.38(b). By definition, in a nvo-terminal circuit element, the current entering each
resistor equals the current leaving. Therefore, the current through each resistor in the series con
nection is the same, and is denoted /. So the new power dissipated by each bulb/resistor is
^l,new
Snd P2^new ~ ^2^
To calculate these values, we need to know I. By conservation o f power, the power delivered by
the source is the sum o f the absorbed powers, i.e.,
^ sou rce ~
X I P\^new
^ 1,new ~
+ -^2^
Hence, dividing through by /,

120 = RÎ + R 2I = (Ri + R 2)I = 600/
Therefore,
;==o.2A
600
Hence,
= 360 X 0 .2 ^ = 14.4 W ,
= 240 x 0.2^ = 9.6 W,
= 24 W
37
Step 3. Find voltages across each bulb. From Ohms law,

Vl =7?,7 = 72 V and V2 = ^ 2 ^ = 48
Although involved, the solution o f this problem uses the definition o f a two-terminal circuit ele
ment and conservation o f power to arrive at the result in a roundabout way. In Chapter 2, we can
more directly arrive at the answers by using Kirchhoffs voltage and current laws.
A potential problem with series connections o f light bulbs is circuit failure. If one bulb burns out,
i.e., the filament in the bulb open-circuits, then all other lights are extinguished. Parallel circuits
continue to operate in the presence o f open-circuit failures and are easier to fix: only the unlit bulb
must be replaced.
8. V-l CH A RA CTERISTICS OF IDEAL RESISTOR, CO N STA N T

VO LTAG E, AN D CO N STA N T CU RREN T SO URCES
The ideal (linear) resistor is a device
that satisfies Ohms law. Ohms law is a
relationship between the current
through the linear resistor and the volt
age across it. A graph o f this relation
ship is known as the v-i characteristic
o f the resistor. The ideal resistor stud
ied in this chapter has the v-i charac
v(V )
teristic given in Figure 1.39. The slope

o f the line in the v-i plane is the value
o f the resistance.
Recall that an ideal voltage source
maintains a given voltage, irrespective
o f the current demands o f the attached
circuit. For constant-voltage sources, as
shown in Figure 1.40(a), this property is
depicted graphically by a constant hori
FIGURE 1.39 Linear resistor characteristic in which

voltage is the constant times the current through
the resistor.
zontal line (slope equals 0 ) in the v-i plane

(Figure 1.40(b)). This means that the internal resistance o f an ideal voltage source is zero.
Further, if
= 0, the voltage source looks like a short circuit because the current flow, generated
by the remaining circuit, will induce no voltage across the source. For now, we must be content
with this brief discussion. Chapter 2 will reiterate and expand on these ideas.
38
+
V ( + )
V-
/ Circuit
-I
(a)
(b)
FIG U RE 1.40 (a) Constant source
attached to circuit; (b) v-i characteristic
is a constant horizontal line in the v-i plane.

Analogously, an ideal current source maintains the given current, irrespective o f the voltage
requirements o f the attached circuit. For constant-current sources, as in Figure 1.41(a), this prop
erty is depicted by a constant vertical line (infinite slope) in the v-i plane (Figure 1.41(b)). This
means that an ideal current source has infinite internal resistance. Further, if
= 0, the current
source looks like an open circuit because no current will flow, regardless o f any voltage generated
by the rest o f the circuit. Again, we must be content with this brief discussion until Chapter 2 reit
erates and expands on the ideas.
+
Y
/ Circuit /
(a)
FIG U RE 1.41 (a) Constant source
(b)
attached to circuit;
(b) v-i characteristic is a constant vertical line in the v-i plane.
9. SUM M ARY
Building on a simplified physics o f charge (coulombs), electric fields, and charge movement, this
chapter set forth the notions o f current, i{t) or / for dc, and voltage, v{t) or V for constant volt
ages. A rigorous treatment would require field theory and quantum electronics. More specifically,
the notions o f current, current direction, voltage, and voltage polarity, a two-terminal circuit ele
ment (the current entering equals the current leaving), the passive sign convention, power con
sumption [pit) = v{t)i{t) assuming the passive sign convention], and dissipated energy (the inte
gral o f power) were all defined. In general, we can say that every circuit element does one o f the
following:
Absorbs energy
Stores energy
Delivers energy, or
Converts energy from one form to another
39
T he chapter subsequently introduced ideal independent and dependent voltage and current
sources: the voltage-controlled voltage source (VCVS), the current-controlled voltage source
(CC VS), the voltage-controlled current source (VCCS), and the current-controlled current source
(C C C S). A dependent source produces a voltage or current proportional to a voltage across or a
current through some other element o f the circuit. The various types o f dependent sources are
summarized in Table 1.4.
TABLE 1.4 Summary of the Four Possible Dependent Sources.
VCVS
(Voltage-Controlled
Voltage Source, p is dimensionless)
ccvs
(Current-Controlled
Current Source,
is in
ohms)
V CC S
(Voltage-Controlled
Voltage Source,
is in S)
CCCS
(Current-Controlled
Current Source, P is
dimensionless)
-I-
40
The chapter keynoted a special two-terminal element, called a resistor, whose terminal voltage and
current satisfied Ohms law, v(t) = Ri{t), where v{t) is the voltage in volts, R is the resistance in
ohms, and i{t) is the current in amperes. The resistor, as defined in this chapter, is a passive ele
ment, meaning that it always absorbs power,/>{;) = v{t)i{t) = i?-{t)IR = Rp-{t) > 0 since R>Q. This
absorbed power is dissipated as heat. Hence, the (passive) resistor models the heating elements in
a stove or toaster oven quite well. In addition, the resistor models the hot resistance o f a light bulb.
Throughout the text, the resistor will often represent a fixed electrical load. In a later chapter, we
will discover that it is possible to construct a device with a negative resistance, R<Q, which can
generate power. However, such a device is rather complex to build and requires such things as the
operational amplifier covered in Chapter 4.
The various quantities defined and used throughout the chapter have various units. The quanti
ties and their units are summarized as follows:
TABLE 1.5 Summary of Units
Charge
Current
Coulomb
Ampere (A)
(C)
Voltage
Resistance
Conductance
Volt (V)
Ohm (Q)
S (Siemens)
mhof3
Power
w a tt
= volt X
am p
Energy
Joule (J)
Throughout this chapter, a number o f examples illustrated the various concepts that were intro
duced. Some simple resisrive circuits were analyzed. To analyze more complex circuits, one needs
Kirchhoffs voltage and current laws, which specify how circuit elements interact in a complex cir
cuit. These basic laws o f circuit theory are set forth in the next chapter.
10. TERM S AND C O N C EPTS

Alternating current: a sinusoidally time-varying current signal having the form A'sin(co?+(j)).
Battery: a device that converts chemical energy into electrical energy, and maintains approxi
mately a constant voltage between its terminals.
Charge: an electric property o f matter, measured in Coulombs. Like charges repel, and unlike
charges attract each other. Each electron carries the smallest known indivisible amount o f
charge equal to - 1.6 x 10 '^ Coulomb.
Conductance: reciprocal o f resistance, with siemens (S) (or formerly, mhos) as its unit.
Conductor: a material, usually a metal, in which electrons can move to neighboring atoms with
relative ease.
Conservation o f power (energy): the sum o f powers generated by a group o f circuit elements is
equal to the sum o f powers absorbed by the remaining circuit elements.
Current: the movement o f charges constitutes an electric current. Current is measured in
Amperes. One Ampere means movement o f charges through a surface at the rate o f 1
Coulomb per second.
Current source: a device that generates electrical current.
Dependent (controlled) current source: a current source whose output current depends on the
voltage or current o f some other element in the circuit.
41
'w '
Dependent (controlled) voltage source: a voltage source whose output voltage depends on the
voltage or current of some other element in the circuit.
Direct current: a current constant with time.

Ideal conductor: offers zero resistance to electron movement.
Ideal insulator: offers infinite resistance to electron movement.
Independent (ideal) current source: an ideal device that delivers current as a prescribed function
of time, e.g., {2 cos(/) + 12}A, no matter what circuit element is connected across its ter
minals.
Independent (ideal) voltage source: an ideal device whose terminal voltage is a prescribed func
tion of time, e.g., {2 cos{t) + 12}V, no matter what current goes through the device.
Instantaneous power: the value of p{t) =
at a particular time instant.
Insulator: a material that opposes easy electron movement.
Mho: historical unit of conductance equal to the reciprocal of an ohm.
Ohm: unit of resistance. One ohm equals the ratio of IV to lA.
Ohms law: for a linear conductor, the current through the conductor at any time t is proportional
to the voltage across the conductor at the same time.
Open circuit: connection of infinite resistance or zero conductance.

Passive sign convention: voltage and current reference directions, indicated by +, - , and an arrow,
which conform to that shown in Figure 1.15.
Peak-to-peak value: equals 2 K 'm K sin(co^ + (()) of the ac waveform.

Peak value: refers to K m K sin(cor + (|)) of the ac waveform.
Power: rate of change of work per unit of time.
Resistance: for a resistor, v{t) a i{t). The proportionality constant R is called the resistance, i.e.,
v{i) = Ri{t). Resistance is measured in ohms: 1 ohm means the voltage is 1 V when the
current is 1 A.
Resistivity: the resistance of a conductor is proportional to its length and inversely proportional
to its cross-sectional area. The proportionality constant p is called the resistivity of the
material. The resistivity of copper at 2 0 ^C is 1.7 x 10~^ ohm-meters.
Resistor: physical device that obeys Ohms law. There are commercially available nonlinear resis
tors that do not obey Ohms law. Resistors convert electric energy into heat.
Root mean square (rms) or eflfective value: measure of ac current, which is related to the peak
value by the formula rms = 0.7071if, where K sin(o)^ + (|)) is the ac waveform.
Short circuit: connection of zero resistance or infinite conductance.
Siemens: unit of conductance (formerly, mho) or inverse ohms.
v-i characteristic: graphical or functional representation of a memoryless circuit element.
Voltage (potential difference): positive charge, without obstruction, will move from a higher
potential point to a lower potential point, accompanied by a conversion of energy.
Voltage is measured in volts; 1 volt between two points A and B means that the energy
converted when moving 1 Coulomb of charge between A and B is 1 joule.
V oltê source: device that generates an electric voltage or potential difference.

Wattage: measure of power consumption.
42
PROBLEMS
AN SW ER: (a) -0 .1 2 1 3 C; (b) 121,3 A; (c) 3.75

X
10^'; (d) 1 - exp(-5^) for ? > 0 and 0 for ^ <
0 , from left to right; (e) line segments joining
C H A R G E A N D C U R R EN T
PRO BLEM S
(0,0), (0,1), (2,1), (2,-1), (5,-1), (5,1), (6,1),

(6,2), (7,2), (7,-2), (8,-2).
1.Consider the diagram o f Figure P I.la .

(a)
(b)
(c)
(d)
Determine the charge on 7 .5 7 3 X
10 ^^ electrons.
If this number o f electrons moves uni
formly from the left end o f a wire to
the right in 1 ms (milli second), what
current flows through the wire?
How many electrons must pass a given
point in 1 minute to produce a current
o f 10 Amperes?
If the charge profile across the cross-sec
tion o f a conductor from left: to riglit is
given by q{t) = t+ 0.2e'5^- 0.2 C for t >
0 and zero for ? < 0 , plot the profile o f the
current that flows across the botmdary. In
2. For the following questions, draw diagrams

whenever necessary.
(a)
(b)
Determine the charge on 6 .023 X lO '^

electrons.
If this number o f electrons moves uni
formly from the left end o f a wire to
the right in 1 ms (milli second), what
current flows through the wire?
(c)
(d)
How many electrons must pass a given

point in 1 minute to produce a current
o f 5 Amperes?
The charge profile residing in a vol
ume V = 10 cm^ is given by q(i) = t +
0.5 sin( 7i:^) C for t> 0 and zero for t <
what direction would the current flow?
0. Plot the current that flows across

the boundary o f the volume for 0 < ?
< 2 sec. In what direction would the
current flow at ? = 1 second? Explain.
3. Reconsider Figure 1.5 in the text in which
changed to ijyf) and
is
is changed to z'^(?). Suppose
(i)
Positive charge carriers move from left
(ii)
Negative charge carriers move from
to right at the rate o f 2cos(10z-) C/s

right to left at the rate o f 6 cos( 10 ?)
C/s
Repeat part (d) for the charge wave
(a)
Find ij^t) and i^(?) as functions o f time.
form (in coulombs) sketched in Figure

P .l.lb .
(b)
Describe the charge movement on the

wire at the boundaries A and B.
4. (a)
Suppose the charge transported across

the cross section o f a conductor for t >
0 is q(t) = e'' sin( 12071?) C. Find the
current, z(r), t> 0 ,flowing in the con
ductor.
(b)
The charge crossing a boundary in a

wire is given in Figure P I.4 for t > 0.
Plot the current i(t) through the wire.
See Example 1.2.
43
AN SW ER: Q = integral o f current. Hence, Q =

0.4 - 0.2 = 0.2.
7. (a)
(b)
Figure P 1.4 Charge crossing a hypothetical

boundary.
(c)
The current in an ideal conductor is
5. (a)
(b)
given by i{t) = 5 - 3e'^^ - 2e^^ A for t

> 0. Determine the charge transferred,
q{t), as a function o f time for t > 0 .
Repeat part (a) for the current plot
sketched in Figure P I .5.
The current in an ideal conductor is

given by i{t) = 2 - cos(2?) A for t
> 0 and 0 for t < 0. Determine the
charge transferred, q{i), as a function
o f time for ? > 0 .
Now suppose the charge transferred
across some surface for ^ > 0 from left
to right is q{i) = 2 - cos(2?) C.
Find the current i(t) through the sur
face for ^ > 0 from left to right.
Repeat part (a) for the current plot
sketched in Figure P I .7. Again, the
current is zero for t< 0.
i(t) (A)
2--
-t (secs)
1- -
Figure P I.7
-
2- -
CHECK: (c) For ,6 > ? > 3 , g (0 = j - 4 t + 12
Figure P I.5
6 . A plot o f the current flowing past point A is

shown on the graph o f Figure P I . 6 . Find the
net positive charge transferred in the direction
o f the current arrow during the interval 0 < ^<
8 . Find i(t) when the charge transported across

a surface cutting a conductor is shown in Figure
P I.8.
6 sec, in Coulombs.
i (amps)
0.1
-t (sec)
10
-0.1--
t-*-t (sec)
Figure P I .6
Figure P I .8
44
9. (a)
Find the average value o f the voltage,
12.(a)
2k
T = -
(b)
(0
Hint: See Equation 1.6.

(b)
W hich elements in Figure P I. 12 are

labeled with the passive sign conven
tion'
v{t) = K cos(cof) over one period,
Find the average value o f the absolute
In the circuit o f Figure P I. 12, volt

ages, currents, and powers o f some ele
ments have been measured and indi
cated in the diagram.
value o f the voltage, v(f) = K cos(cot)

over one period, T = 2u/cl).
5A
+ 3V -
2V -
+ 5V -
-cz>
V O LTA G E, CURREN T, POW ER,

EN ER G Y
10. In Figure P I. 10, suppose we know that
4A
+
4V
Vab = 8 V and Vad = 18 V. Find the values of
il'
7A
3A
0 '
Figure P I. 12
(i)
If element A generates 28
power, find Va (ii) Find the power absorbed by
ment B.
(iii) I f element C generates 6
power, find Vq
(iv) I f element D absorbs 2 7
power, find //).
6 V
- 4 V
W
ele
W
W
(v) If element E absorbs 4 W power,

Figure P I. 10
11. (a)
(b)
2A
W hich o f the three elements in Figure

P I. 11 is labeled with the passive sign
convention?
find I e (vi) Find the power absorbed by ele

ment F.
13. Consider the circuit o f Figure P I. 13.
Find the absorbed powers for each cir

cuit element in Figure P 1.11.
____________
2A
(a)
___________
20 V
lO V
(i)
(ii)
__ .
Find the power absorbed by the circuit

elements 1 and 2 .
(b) Show that the algebraic sum o f the

absorbed powers is zero. Be careful o f
sign.
4A.
3A
2A
C _ 3
20 V
(iii)
Figure PI. 11
A N SW ERS (b): (i) -40 W; (ii) 20 W; (iii) 60 W
6V
1A
3A
lOV
circuit
element 1
circuit
6V element 2
16v Q ) 2 A
lOV
Figure P I. 13
45
14. (a)
Determine the power absorbed by
+ 80V -
-25A
each o f the circuit elements in Figures

(b)
P 1.14a below.
Show that the algebraic sum o f the
absorbed powers is zero. Be careftil o f sign.
(c)
Repeat parts (a) and (b) for die circuit

of Figure P l.l4 b , where v^{t) = 3 for t>Q, Vjit) = 1 +
^^for t>Q, z'j(?)
for t>Q, and ijit) =
=2 +
for t
> 0.
9A
Figure P I. 15
14A|
circuit
15V
element 1
4A
5A ( ^ ^ 2 0 V
lOA
+ 10V-
circuit
element 2
5V
lOV
- 5V +
5A
(a)
100(1 - e * ) mA for t > 0.

(a)
How much energy does the element A
absorb for the interval [0 , t] ?
(b)
I f element B is a 5 Q resistor, deter
circuit
element 2
-v,(t) +
2A
3.334, Rfj = 20 in ohms.
16. In the circuit shown in Figure P I. 16, i(t) =
Figure P I. 14a
30
CHECK: Re
(c)
mine the power absorbed at time t,

and the energy absorbed for the inter
val [0 , t],
W hat is the energy delivered by the
source over the interval [0 , i\?
4V
(b)
Figure P 1.14b
15. In the circuit o f Figure P I. 15, there are
three independent sources and five ordinary
resistors.
(a)
Determine which o f the circuit ele
(b)
ments are sources and which are resis

tors.
Determine the value o f the resistance
for each o f the resistors.
25V
ti
Figure P I. 16
C H E C K : (b) 125 W, 1 2 5 tJ
17. Suppose energy cost in Indiana is 10 cents
per kwh.
(a)
How much does it cost to run a 100watt T V set 8 hours per day for 30
days?
(b)
How many 100-watt light bulbs run
for 6 hours a day are needed to use
$9.00 o f energy every 30 days?
A N SW ER: (a) 8 cents per day; $2.40 per
month; (b) 5 bulbs
46
Chapter 1 Charge, Current, Vohage and O hm s Law
RESISTA N CE
18. Using Equation 1.19 and Table 1.2 , find
the resistance o f a nickel ribbon having these
dimensions:
length:
width:
thickness:
19. (a)
(b)
(c)
40 m
1.5 cm
0.1 cm
Compute the resistance o f800 feet o f 14gauge copper wire (2.575 Q /1000 ft).
Repeat (a) for 200 feet o f 14-gauge
nickel wire.
If one end o f the copper wire is soldered
to one end o f the nickel wire, find the
total resistance o f the 1000 feet o f wire.
Can you justify your answer?
20. The resistance o f a conductor is function of

the temperature T (in C ). Over a range o f
temperature that is not too distant from 2 0 C ,
the relationship between R{ T) and T is linear
Figure P I.21
22. For the circuit o f Figure P 1.22,
= 1Q , and
the input current to the circuit o f is i(^t) =

400sin(207i^) mA for ? > 0 and zero for t <Q.
(a) Com pute
the
instantaneous
power delivered by the source.

Using a graphing program, graph
the power delivered as a function
o f time for 0 < ? < 0.5 s.
(b) Now compute and graph an expres
sion for the energy dissipated in the
resistor as a function o f for 0 < ? <
0.5 s.
and can be expressed as /?(7) = i?(20)[l + a{T-
20 )] where a is called the temperature coeffi

cient o f the conducting material. For copper a
= 0.0039 per C . If the resistance o f a coil of
wire is 21 Ohms at -1 0 C, what is the resist
ance when the wire is operating at 10 C?
AN SW ER: 22.85 Q
A VERA G E VALU E, PO W ER, A N D

EN ER G Y C A LC U LA TIO N S
2 1 . The current through a 500 2 resistor is
given in Figure P I .21 where /^ = 6 mA.
(a)
(b)
How much total charge is transferred

over the time interval ^ = 0 to ^ = 2 sec
onds?
How much total energy must a source
deliver over the time interval t = 0 X.Ot
= 2 seconds?
(c)
If /(?) in Figure P I.21 is periodic, with

period equal to 2 seconds, i.e., the indi
cated waveform is replicated every two
seconds, find the average power absorbed
by the resistor. Use intuitive reasoning.
Figure P I.22
23. The switch S in Figure P I .23 is assumed to
be ideal, i.e., it behaves as a short circuit when
closed, and as an open circuit when open.
Suppose the switch is repeatedly closed for 1 ms
and opened for 1ms.
(a) W hat is the average value o f z(^)?
(b) W hat is the average power delivered
by the source?
Kt)>'
Figure P I.23
A N SW ERS: 0.25 mA. 1.25m W
47
24. Repeat problem 23 when the switch is

repeatedly closed for 3 ms and opened for 1 ms.
28. In Figure P1.28, F q = 125 V.
(a)
Suppose bulb A and bulb B each use

100 watts o f power. Find /q and the
CH ECK:
= 1.875 m W
hot resistance o f each bulb.

(b)
Suppose bulb A uses 40 watts o f power
25. In Figure P1.25, Vq = 10 V, and the switch S
and bulb B 60 watts o f power. Find /q
alternately stays at position A for 4 ms and at
and the hot resistance o f each bulb.
position B for 1 ms. Find the average value o f i(tj.

5kn
Figure P I.25
C H E C K : Average current is 1.8 mA.
A PPLIC A TIO N S OF O H M 'S LAW
ANSW ER: (b) /g = 0.8 A,
26 (a)
93.75 Q
W hat is the safe maximum current o f
a 0.25 W, 2 77 k fi metal film resistor

used in a radio receiver?
(b) W hat is the safe maximum current o f
a 1 W, 130 2 resistor?
(c) W hat is the safe maximum current o f
a 2 kW, 2 2 resistor used in an electric
power station?
A N SW ERS: (a) 0.95 mA, (b) 87.7 niA. (c)
3 1.6 A
= 62.5 Q ,
29. T he power delivered by the source in the

circuit o f Figure P I .29 is 750 watts.
(a)
If /^ = 5 A, determine the value o f R.
(b)
Suppose now that R = 11 Q. Find

Hint: W hat is the power consumed in
each resistor as a function o f /^?
27. In Figure P I .27, Kq = 120 V.

(a)
T he power absorbed by the bulb in

the circuit shown in Figure P I.2 7 a is
Figure P I.29
60 watts. Find the value o f the hot

resistance o f the bulb.
(b)
The power delivered by the source to

the parallel connection o f two identi
cal bulbs in Figure 1.27b is 150 watts.
Find the hot resistance o f each bulb.
AN SW ER: (a) 6 Q, (b) 4.33 A

30. Consider the circuit o f Figure P I.30. The
power consumed by each resistor is known to be
^2Q ^
watts, PjQ = 48 watts,

= 64 watts,
PjQ = 3075.2 watts, and
= 1944 watts.
(a)
(b)
(a)
(b)
Figure P I.27
AN SW ER: (b) 192 D.
For each resistor, determine the indi

cated voltage or current.
Determine the total power delivered
by the two sources.
48
(ii) Determine the total power deliv

ered by the battery.
20
+ V, -
(iii) Determine the current,

+ V,-
3n
(iv) Assuming that each bulb behaves

as a resistor, determine the hot
40
''" 6
60 <
>50
resistance o f each bulb.

(b)
Figure P I . 3 0
C H E C K : (a)
blow the 15-amp fuse in the circuit o f
= 28 V, Vg = 108 V, (b) 5523.2
Figure P I.3 2 b .
(c)
31. The power absorbed by the resistor R in

the circuit o f Figure P I.31 is 100 watts and
(a)
Find the value o f R.
(b)
Find the value o f the current flow

ing through R and determine its
direction as per the passive sign
convention.
Find the power absorbed by the 20
Q resistor.
Repeat (b) for C C bulbs.

Ignition
Switch
Ko = 2 0 V
(d)
Determine the number o f AA bulbs in

parallel that would be required to
watts
(c)
deliv
ered by the battery.
12V
Find the power delivered by the

source and the value o f /q.
Figure PI.32a
12V
Figure PI.32b
200
Figure P1.31.
C H E C K : /q = 6 A.
32. (a)
Consider the circuit o f Figure P I.32a,

which shows three lamps, AA, BB,
and C C in a parallel circuit. This is a
simplified example o f a light circuit on
a car, in your house, or possibly on a
Christmas tree. Halogen bulb AA uses
35 watts when lit, the Halogen Xenon
bulb BB uses 36 watts when lit, and
the incandescent bulb C C bulb uses
25 watts when lit.
(i) Determine the current through
each bulb.
AN SW ERS: (a) (i) 2 .9 167, 3, 2.08; (ii) 6.2 A;

(iii) 96; (iv) 4 .1 1 ,4 , 5.76; (b) n > 6 ; (c) n > 8
33. An automobile battery has a terminal volt
age o f approximately 12 V when the engine is
not running and the starter motor is not
engaged. A car with such a battery is parked at
a picnic. For music, the car stereo is playing,
using 240 watts, and some o f the lights are on
using 120 watts. W ith this load, the battery will
supply approximately 3 M J o f energy before it
will have insufficient stored energy to start the
car.
(a) W hat power does the battery supply
(b)
to the load?
W hat current does the battery supply
to the load?
49
(c)
Approximately how long can the car

remain parked with the stereo and
lights on and still start the car?
C H E C K : (c) 2.31 hours

34. In Figure P I .34, Vq = 24 V, i?, = 4 Q, the
Figure PI.35b
unknown resistance, Rj, consumes 20 watts o f

power. Find
and Rj- (How many possible
D EP EN D EN T SO U RC E PRO BLEM S
36. Consider the circuit in Figure P I .36.
solutions are there?)
(a)
-A /V ^
R,
If 1/ = 6 V, find
and the power in
watts absorbed by the load R^. W hat is

the power delivered by each source?
I'.
(b)
If the power absorbed by the load

resistor is Pj^ = 80 watts, then find V^,
/( and I/.
Figure P I.34
C H EC K : 1 A, 20 Q or ?????
R, = 20n
35. In Figure P I .35, Vq = 48 V.

(a)
Determine the value(s) o f the current

in the D C resistive circuit o f Figure
P I.3 5 a
(b)
(c)
given
that
the
unknown
Figure P I.36
devices absorb the powers indicated.

W hat value o f Vq results in a unique
C H EC K : (a)
solution for I J
(b) 1/ = 4 V
= 60 V,
= 3.6 watts;
If the circuit is modified as shown in

Figure P I.35b , determine the rwo new
values o f the current 1^.
37. For the circuit in Figure P I .37, determine

and
in terms o f /-^,
, R2 and p.
AN SW ER: (a) 10, 2 A; (b) 35.77 V
Figure P1.37
38. Consider the circuit o f Figure P I .38.
(a)
Determine an expression for
F igure P I . 3 5 a
and
the voltage gain
V
G y = ^
Vin terms o f R^, R2 , a , and V-^^.
(b)
If R.ÂQ. and a = 0.8, determine the
50
(c)
value of J?2 so that the voltage gain Gy
41. For the circuit o f Figure P 1.41, suppose
= 4.
Given your answer to (b), determine
the power gain, which is the ratio of
the power delivered to
divided by
= 1 0 V.
(a)
(b)
the power delivered by the source.
Find the output voltage and output

current.
Find the voltage gain
Gv =
Gp =
ôut
, and the power gain
^
Pin
(c)
(d)
Figure P I.38
Find the power delivered by each

source.
Suppose the power absorbed by the 2
k Q output resistor were 80 watts.
Find the power delivered by the input
source,
and the voltage
39. For the circuit of Figure P I .39, suppose
- 100 mA,
50 Q,
10 Q, and
100 Q.
(a)
Find the output voltage and out
(b)
put current.
Find the current gain,
G/ =
0.1V.
0.21
v,^
Rb
, the v oltê gain

Figure P I.41
Gy =
, and the power gain

C H EC K : (b)
(c)
10
G p = ^
P-
42. For the circuit of Figure P 1.42, suppose
Findthepowerabsorbedbyeachresistor.
^ 3=
10i?i. Find the resistor values

so that Gy = ^
= 1000
Vin
2001
1000
2kO
2kn
Figure P I.39
40. In problem 39, suppose R^= 1 k Q, 7?2=
C ty
10 Q, and R^= 20 Q and P ^ = 80 watts. Find

^2.
Figure P I.42
CHECK; 4 . 1 mA
C H EC K : i?,= 5 k Q
C H A P
KirchhofFs Current & Voltage
Laws and Series-Parallel Resistive Circuits
A CAR HEATER FAN SPEED -CO N TRO L APPLICATION
One use o f resistors in electronic circuits is to control current flow, just as dams control water flow
along rivers. Ohms law, V = RI, gauges the ability o f resistors to control this current flow: for a
fixed voltage, high values o f resistance lead to small currents, whereas low values o f resistance lead
to higher currents. This property underlies the adjustment o f the blower (fan) speed for ventila
tion in a typical car, as represented in the following diagram.
In this diagram, three resistors are connected in series, and their connecting points are attached to
a switch. As we will learn in this chapter, the resistance o f a series connection is the sum o f the
resistances. So with the switch in the low position, the 12-V car battery sees three resistors in series
with the motor. Th e series connection o f three resistors represents a large resistance and heavily
restricts the current through the motor. W ith less current, there is less power, and the fan motor
speed is slow. W hen the switch moves to the Med-1 position, a resistor is bypassed, producing less
52
Chapter 2 K irchhofFs Current & Voltage Laws and Series-Parallel Resistive Circuits
resistance in the series circuit and allowing more current to flow. More current flow increases the
fan motor speed. Each successive switch position removes resistance from the circuit, and the fan
motor speed increases accordingly.
Analysis o f such practical circuits builds on the principles set forth in this chapter.
CHAPTER O U TLIN E
1.
2.
3.
4.
5.
6.
7.
8.
9.
Introduction and Terminology: Parallel, Series, Node, Branch, and so on

KirchhofFs Current Law
Kirchhoffs Voltage Law
Series Resistances and Voltage Division
Parallel Resistances and Current Division
Series-Parallel Interconnections
Dependent Sources Revisited
Model for a Non-ideal Battery
Non-ideal Sources
Summary
Terms and Concepts
Problems
CHAPTER OBJECTIVES
Define and utilize Kirchhoffs current law (KCL), which governs the distribution o f cur
rents into or out o f a node.
Define and utilize Kirchhoffs voltage law (KVL), which governs the distribution o f volt
ages in a circuit.
Introduce series and parallel resistive circuits.
Develop a voltage division formula that specifies how voltages distribute across series con
nections o f resistors.
Develop a current division formula that specifies how currents distribute through a par
allel connection o f resistors.
'
Show that a series connection o f resistors has an equivalent resistance equal to the sum o f
the resistances in the series connection.
Show that a parallel connection o f resistors has an equivalent conductance equal to the
sum o f the conductances in the parallel connection.
Explore the calculation o f the equivalent resistance/conductance o f a series-parallel con
nection o f resistances, i.e., a circuit having a mixed connection o f series and parallel con
nections o f resistors.
Explore the calculation o f voltages, currents, and power in a series-parallel connection o f
resistances.
Revisit the notion o f a dependent source and use a V C C S to model an amplifier circuit.
Describe a practical battery source and look at a general practical source model.
Chapter 2 K irchh offs Current & Voltage Laws and Series-Parallel Resistive Circuits
53
1. IN TRO D U CTIO N AND TER M IN O LO G Y: PARALLEL, SERIES,

N OD E, BRANCH, AND SO ON
The circuits studied in Chapter 1 were interconnections
o f resistors and sources that were two-terminal circuit
elements. This chapter sets forth K irchhoffs voltage
law (KVL) and Kirchhoffs current law (KCL). These
laws govern the voltage relationships and the current
relationships, respectively, o f interconnections o f twoterminal circuit elements.
Some new terminology underpins the statements o f
KVL and KCL. Figure 2.2a shows a series circuit con
sisting o f a sequential connection o f two-terminal cir
cuit elements (resistors) end to end. The common con
FIG U RE 2.1
nection point between any elements is called a node. In

general, a node is the connection point o f one or more circuit elements. Figure 2.2b shows a par
allel circuit, in which the top terminals and the bottom terminals o f each resistor are wired
together. The common connection point o f the top terminals is a node, as is the common con
nection point o f the bottom terminals.
An important property o f the series connection o f Figure 2.2a is that all the rwo-terminal elements
carry the same current, in this case
because the input current for each two-terminal element
must equal the exit current. Similarly, in a parallel connection, such as Figure 2.2b, the same volt
age, in this case, Vj^, appears across every circuit element.
+
node
d-
0
-1-
node 1
Vr
-
node 2
(a)
(b)
FIG U RE 2.2 (a) Series connection of resistors with the property that each resistor carries
the same current; (b) parallel connection of resistors with the property that
the same voltage appears across each resistor.
Sources interconnected with circuit elements produce currents through the elements and voltages
across the elements. For example, a voltage source connected across Figure 2.2a would generate a
current 2^ and the voltages
through v^. K irchhoffs voltage law (KVL) governs the distribu
tion o f voltages around loops o f circuit elements, as shown in Figure 2.2a. Similarly, a current
source connected across the circuit o f Figure 2.2b would produce the voltage
and the currents
Z] through
K irchhoffs current law (KCL) governs the flow o f currents into and out o f a com-
54
Chapter 2 KirchhofPs Current & Voltage Laws and Series-Parallel Resistive Circuits
mon connection point or node, as in the top and bottom connections o f Figure 2.2b. This chap
ter sets forth precise statements o f these laws and illustrates their application.
A proper statement o f KVL and KCL requires the additional notion o f branch. A branch o f a cir
cuit is a generic name for a two-terminal circuit element and is denoted by a line segment, as in
Figure 2.3. T he endpoints o f a branch (the terminals o f the circuit element) are called nodes, as
in Figure 2.3a. Ordinarily, however, node means a common connection point o f two or more cir
cuit elements (branches), as shown in Figure 2.3b.
node A
(a)
(b)
FIG U RE 2.3 (a) Single branch representing a circuit element with terminals labeled as
nodes A and B; (b) interconnection of branches (circuit elements) with common
connection points labeled as nodes A through D.
The voltage polarity and current direction for the branches in Figures 2.2 and 2.3 are labeled in accor
dance with the passive sign convention; the arrowhead on a branch denotes the reference current direc
tion, which is from plus to minus. Recall that the + to - does not mean that the voltage is always posi
tive if measured from the plus-sign to the minus-sign. In general, reference directions can be assigned
arbitrarily. The conventional assignment o f voltage polarity and current direction to voltage and current
soiurces is given in Figure 2.4, which is different from the passive sign convention. Note that with these
conventional assignments, the (instantaneous) power delivered by a source is/^/^) =
power absorbed by a source is
Circuit
Circuit
' 'J O
(a)
th^
= -pjeff)-
(b)
FIG U RE 2.4 Conventional labeling o f (a) voltage, and (b) current sources.
55
2. KIRCH HOFF'S CU RREN T LAW

Imagine a number o f branches connected at a common point, as at node A o f Figure 2.3b. The
current through each branch has a reference direction indicated by an arrow. If the arrow points
toward the node, the reference direction o f the current is entering the node; if the arrow points
away from the node, the reference direction o f the current is leaving. If a current is referenced as
leaving a node, then the negative o f the current enters the node, and conversely.
KIR C H H O FF'S C U R R EN T LAW (K C L )

S tatem ent 1: The algebraic sum o f the currents entering a node is zero for every instant o f
time.
S tatem ent 2 : Equivalently, the algebraic sum o f the currents leaving a node is zero for every
instant o f time.
The two statements o f KCL are equivalent because the negative o f the sum o f the currents enter
ing a node corresponds to the sum o f the currents leaving the node. Further, from physics we know
that charge is neither created nor destroyed. Thus, the charge transported into the node must equal
the charge leaving the node because charge cannot accumulate at a node. KCL expresses the con
servation o f charge law in terms o f branch currents. Moreover, KCL specifies how branch currents
interact at a node, regardless o f the type o f element connected to the node.
Referring to Figure 2.3b, KCL at nodey4 requires that i^{t) +
- iîi) = 0 for all t. KCL at node
Finally, KCL at node D requires that 25(f) =
B requires that
E X A M P L E 2.1
For the node shown in Figure 2.5, find
FIG U RE 2.5 Connection o f five circuit elements at a single node.
56
S o lutio n
By KCL, the sum o f the currents entering the node must be zero. Hence, the current z^(^) =
9cos(2r) - 3cos(2?) - cos(2r) - 2cos(2z) = 3cos(2^) A.
Exercise. 1. Suppose the current through the voltage source in Figure 2.5 is changed to - 2 cos(2?).
Find
AN SW ER: - 4 cos(2^) A.
2. Three branches connect at a node. All branch currents have reference directions leaving the
node. If /j = /2 = 2 A, then find ly
A N SW ER; - 4 A
Two implications o f KCL are o f immediate interest. First, as a general rule, KCL forbids the series
connection o f current sources. Figure 2.6a shows an invalid connection o f two arbitrary current
sources i,(?) and iî), where z,(?)
i^t). It is invalid because KCL requires that i^{t) = i2 {t). On
the other hand, a parallel connection o f two current sources can be combined to form an equiva
lent source, as in Figure 2.6b, where
= i^{t) + i2 {t).
(a)
-O
(b)
FIG U RE 2.6 (a) Invalid connection of two arbitrary current sources

when Z](z)
ijii)- Avoid this violation o f KCL; (b) equivalent representation o f a parallel connection
o f rwo current sources in which
= i]{t) +
A second immediate consequence o f KCL is that a current source supplying zero current [i{t) = 0
in Figure 2.7] is equivalent to an open circuit because the current through an open circuit is zero.
An open circuit has infinite resistance, or zero conductance. This means that a current source has
infinite internal resistance. From another angle, a constant current source is represented by a ver
tical line in the iv plane (see Figure 1.4 lb ). The slope o f the vertical line, which is infinite, deter
mines the internal resistance o f the source.
57
-O
+
v(t)
-O
FIG U RE 2.7 Ideal current source with i{t) = 0 is an open circuit.
A typical application o f KCL is given in the following example.
E X A M P L E 2.2
In the parallel resistive circuit o f Figure 2.8, the voltage across each resistor is 6 cos(z) V. Find the
current through each resistor and the current,
supplied by the voltage source.
'm
6cos(t) V
IQ
<20
<3Q
FIG U RE 2.8 Parallel resistive circuit for Example 2.2.
S o lu tio n
By Ohm s law,
/^ l ( 0 = 6 c o s ( 0 A
6 c o s (/)
'R l W -------- = 3 c o s ( 0
, 3 , - 5 i 2 5 W = 2 c o s ( ,) A
By KCL,
iinif) = ' l ( 0 + ' 2 (0 +
= 6 cos(r) + 3cos(/) + 2cos(/) = llco s(/ ) A
Exercise. 1. In Figure 2.8, suppose the source voltage is changed to a constant, labeled
Iin in terms o f V-^.
AN SW ER:
Find
58
Chapter 2 KirchhofFs Current & Voltage Laws and Series-Parallel Resistive Circuits
2. Suppose the source voltage in the circuit o f Figure 2.8 were changed to - 1 2 cos(2?) V. Find
AN SW ER: - 2 2 cos(2?) A
Kirchhoffs current law holds for closed curves or surfaces, called Gaussian curves or surfaces. A
Gaussian curve or surface is a closed curve (such as a circle in a plane) or a closed surface (such as
a sphere or ellipsoid in three dimensions). A Gaussian curve or surface has a well-defined inside and
outside. Figure 2.9 illustrates the idea o f a Gaussian curve for three (planar) situations.
.............
Two
Terminal
Circuit
Elem ent /
i,(t)
i,(t) -A '/'' ............
(a)
(b)
FIGURE 2.9 Illustrations of Gaussian curves:

(a) enclosure of a two-terminal element; (b) enclosure o f a three-terminal device, such as a transistor;
(c) enclosure of a three-node interconnection with an arbitrary circuit.
For the two-terminal circuit element o f Figure 2.9a, KCL for Gaussian curves implies that i^{t) =
i2 {t), which is precisely the definition o f a two-terminal circuit element. For the three-terminal
device o f Figure 2.9b, KCL for Gaussian curves implies that i^{t) = i^{t) + i2 {t). Finally, for Figure
2.9c, i^ i^ + if^ = 0. From these illustrations, one might imagine that the use o f Gaussian surfaces
might simplify or provide a short cut to certain branch current computations. T he general state
ment o f KCL for Gaussian surfaces is next followed by an example that demonstrates its use for
computing branch currents.
59
K C L FOR G A U SSIA N CU R V ES O R SURFACES

The algebraic sum o f the currents leaving (or entering) a Gaussian curve (or surface) is zero
for every instant o f time.
E X A M P L E 2 .3
This example shows how the use o f a Gaussian curve or surface can sometimes simplify a calcula
tion. Figure 2.10 portrays a complicated circuit whose branch currents and voltages are not solv
able by methods learned so far. Our objective is to find the current
without having to solve a
set o f complex circuit equations.
FIG U RE 2.10 Circuit for Example 2.3, showing a Gaussian surface to compute
directly.
S o lutio n
Using KCL for the indicated Gaussian curve, - 1 .1 5 + / ^ - 0.3 + 0.95 = 0. Equivalently, /^ = 1.15
+ 0.3 - 0.95 = 0.5 A.
In the next chapter, circuits such as the one in Figure 2 .10 are analyzed using a technique called
nodal analysis.
Exercise. 1. Draw a Gaussian surface on the circuit in Figure 2.10 that is different from the sur
face given but still allows one to compute /^.
AN SW ER; One choice is a circle enclosing the bottom node.
60
2. Draw an appropriate Gaussian curve to find / in the graphical circuit representation in Figure
2 . 11.
AN SW ER: 2 A
FIG U RE 2.11 Graph representation of a circuit.
3. KIRCHHOFF'S VO LTAG E LAW

Kirchhoffs voltage law (KVL) specifies how voltages distribute across the elements o f a circuit.
Before conveying four equivalent versions o f KVL, we first set forth several necessary background
concepts. The first is the notion o f a closed path. In a circuit, a closed path is a connection o f
two-terminal elements that ends and begins at the same node and which traverses each node in
the connection only once. Figure 2.12 illustrates several closed paths. One closed path is A-B-CD-E-A, i.e., it begins at node A, moves to node B, drops to node C, moves through element 4 to
node D, down through element 6 to reference node E, and back through the voltage source to A.
A second closed path is A-B-C-E-A, and a third is B-D-C-B.
A second concept pertinent to our KVL statements is that o f a node voltage with respect to a ref
erence. A node voltage o f a circuit is the voltage drop from a given node to a reference node. The
reference node is usually indicated on the circuit or is taken as ground. The circuit o f Figure 2.12
has branches labeled 1 through 6 and nodes labeled A through E, with node E taken as the refer
Vg, Vq and v^. The voltage

denotes
the voltage drop from node A to node E-,
denotes the voltage drop from node D to node E, and
similarly for the remaining node voltages. Node E, being the reference node, has zero as its node
ence node. The associated node voltages are denoted by
voltage.
61
V.
BD
FIG U RE 2.12 Circuit diagram illustrating (i) three closed paths {A-B-C-D-E-A)\
(ii) the concept of node voltages with respect to a given reference node E,
(iii) the concept of branch voltages
and Vj^-,
and
The concept o f a closed path and the concept o f a node voltage allow us to state our first two ver
sions o f Kirchhoffs voltage law.
KIRCH HOFF'S V O LTA G E LAW (K V L)

Kirchhoffs voltage law can be stated in different ways. Following are two equivalent state
ments o f the law.
Statem ent 1: The algebraic sum o f the voltage drops around any closed path is zero at every
instant o f time.
Statem ent 2 : For any pair o f nodes j and k, the voltage drop Vjj^ from node j to node k is
given by
at every instant o f time, where Vj is the voltage at node j with respect to the reference and
is the voltage at node ^ with respect to reference. Herey and k stand for arbitrary node indices.
For example, in Figure 2.12, j, k can be any o f the nodes A, B, C, D, or E.
62
Referring back to Figure 2.12, for the closed
A-B-C-D-E-A, statement 1 o f KVL requires that
ÂB ^ ^BC ^CD + ^DE ÊA ^ O'

for Figure 2.12, from statement 2 o f KVL, the branch
voltages
= v^ - Vg and Vqj^ =
^D- Hence,
= -v^. Thus, by knowing the node
voltages o f a circuit, one can easily compute the branch voltages.
Exercise. 1. Find Vy^g, VgQ and V^^-for the circuit o f Figure 2.1 3 in which we have introduced
the ground symbol at node E to identify the reference node.
AN SW ERS:
= - 3 V, Vg(^=2\ V,
= 18 V
2. Again, with reference to Figure 2.13, find the node voltages V^, Vg, Vq and
AN SW ERS: 2 V, 5 V, - 1 6 V, - 6 V
3. In Figure 2.13, suppose the branch labeled 6 V is now labeled - 1 2 V. Find
AN SW ER: - 3 V
FIG U RE 2.13
A third concept needed for two further equivalent statements o f KVL is that o f a closed node
sequence. A closed node sequence is a finite sequence o f nodes that begins and ends at the same
node. A closed node sequence generalizes the notion o f a closed path. Finally, we define the notion
o f a connected circuit. In a connected circuit, each node can be reached from any other node by
some path through the circuit elements. Figures 2 .12 and 2.14 show connected circuits. However,
in Figure 2.14, the sequence o f nodes A-B-C-D-E-A is a closed node sequence but not a closed
path because there is no circuit element between nodes B and C.
+ 2.5V -
+ 10V D
+
V,
F IG U R E 2 .1 4 Simple dependent source circuit for illustrating the concepts o f a connected circuit and a closed node sequence.
63
This brings us to our last two equivalent statements o f KVL.
KIRCH HOFF'S V O LTA G E LAW (K V L)

Following are two additional equivalent statements o f KVL.
Statem ent 3 ; For connected circuits and any node sequence, say A -D -B-... -G-P, the volt
age drop
ÂP = ÂD + ^DB + - +
GP
at every instant o f time.

S tatem ent 4 : For connected circuits, the algebraic sum o f the node-to-node voltages for any
closed node sequence is zero for every instant o f time.
Referring back to Figure 2.12, statement 3 o f KVL impHes that
Vab +
Referring to Figure 2.14, for the closed node sequence E-A-B-E, V-^ = 10 =
+ Vg = 2.5 + Vg and Vg = 7.5 V. Now, consider the closed node sequence E-C-D-E. For this
sequence,
^CD
Equivalently, 30 = 10 + Vj^, in which case
= 20 V. Finally,
consider the closed node sequence, E-B-C-E, which is not a closed path because there is no cir
cuit element between nodes B and C. Nevertheless, by statement 4 o f KVL, - Vg+ Vb + ^ c ~ ^
or equivalently that
= Vg -
= 7.5 - 30 = - 22.5 V.
Exercise. 1. In Figure 2.14, suppose

= 20 V, V^g = 5 V, and
and Vj^.
AN SW ER: Kg = 15 V,
60 V, V^g = 45 V and
= 40 V
10 V and Vj^ = - 3 V. Find v^ q
2. (a) In the circuit o f Figure 2.15, suppose

(b) Suppose Vg = 1 2 0
cos( 1 2 0 tc?),
Vg^ =
18
= 20 V. Find Vg, Vq
cos( 1 2 0 ti?)
and
3 2 cos(1207t^). Find
at ? =
0 .5 s.
(c) Find
when v^ = 1 0 0 V,
= - 1 0 V and
25V.
SC RA M BLED A N SW ER: 85 V, - 1 3 V, - 7 0 V
FIG U RE 2.15 Circuit with nodes labeled A through E. Node E is taken as the reference node.
64
Chapter 2 K irchhoff s Current & Voltage Laws and Series-Parallel Resistive Circuits
Two further implications o f the KVL are o f immediate interest. First, as a general rule, KVL for
bids the parallel connection o f two voltage sources say,
and V2 (^) for which Vj(?)
as illustrated in Figure 2.16a. O n the other hand, two voltage sources in series can be combined
to form a single source, as illustrated in Figure 2.16b, where
{t) = v-^{t) + V2 {t).
FIGURE 2.16 (a) An improper connecdon of voltage sources when v^{t) ^ i>2 (i);
(b) an equivalent representation of two voltage sources connected in series in which
= Vj{i) +
Second, a voltage source supplying 0 V is equivalent to a short circuit, as illustrated in Figure

2.17. Also, the internal resistance o f a voltage source is zero. One can see this by referring to the
fact that in the iv plane, an ideal dc voltage source is represented by a horizontal line, as was illus
trated in Figure 1.40, The slope o f the line is zero and represents the resistance o f the source. These
ideas are dual to those expressed for current sources earlier.
-O
-o
+
ov
ov
FIG U RE 2.17 A 0 V voltage source is equivalent to a short circuit.

Finally, note that all four KVL statements can be justified using the definition and the notation
for voltage drop presented in Chapter 1. The justification is more readily comprehended via the
analogy o f the gravitational field, also developed in that section. Also, observe that KVL holds for
all closed node sequences, independent o f the device represented by each branch o f the connect
ed circuit. The distribution o f voltages around closed paths can be viewed as a special case o f this
general statement.
4. SERIES RESISTANCES AND VO LTAG E DIVISION

During holidays, one often sees strings o f lights hanging between poles or trees. Sometimes these
strings consist o f a series connection o f light bulbs. Each light bulb contains a filament, a coil of
65
wire, that gives o ff an intense light when hot. In a circuits perspective, the filament acts as a resis
tor and has an equivalent hot resistance. The series connection o f bulbs can be modeled by a series
connection o f resistors, with each resistor paired with a specific bulb. Computing the voltage
across each light (a very important type o f calculation) would then be equivalent to finding the
voltage across each o f the resistors in the equivalent circuit model. It is quite common to model
electrical loads, such as a light, by resistors.
E X A M P L E 2 .4
Figure 2.18a shows a voltage source v-JJ) connected to three resistors in series. The objectives o f
this example are to compute
the voltages Vj{t),j = 1, 2, 3, across each resistor, and
the
equivalent resistance seen by the voltage source.
ijt )
(a)
(b)
FIGURE 2.18 (a) Three series resistors connected across a voltage source. By the
definition o f a two-terminal resistor or by the KCL, the current through each resistor is
(b) equivalent resistance
= R^ Rj + R^ seen by the source, i.e., v-^{t) = Reqii{t)-
So l u t io n
Step 1. Express the voltage across each resistor in terms o f the input current. For the circuit o f Figure
2.18a, the current through each resistor is i^JJ) by KCL. From Ohms law, the voltage across each
resistor is
fo t j = 1, 2, 3.
Step 2. Express v-J^t) in terms ofi-JJ), solvefor i-^<^t), and then compute an expression for v^t) in terms
o f the Rj and v-J^t). By KVL, the source voltage equals the sum o f the resistor voltages, i.e..
(2 . 1)
where we have substituted Rjii{t) = Vj{t). Dividing Equation 2.1 by (i?j + Rj + -^3) yields
Since v-{t) = RjiiJyt) for j = 1, 2, 3,
66
V j{t) =
= --- ---------5
,2 2^)
Equation 2.2a is a voltê division formula for a three-resistor series circuit. This formula imphes
that if a resistance R. is small relative to the other resistances in the series circuit, then only a small
portion o f the source voltage develops across it. O n the other hand, if a resistance R. is large rela
tive to the other resistances, then a larger portion o f the source voltage will develop across it. One
concludes that the voltage distributes around a loop o f resistors in proportion to the value o f each
resistance. The proportion is simply the ratio o f the branch resistance R- to the total series resist
ance.
Step 3. Compute the equivalent resistance R^^ seen by the voltage source. The equivalent resistance
seen by the voltage source for a resistive circuit is implicidy defined by Ohm s law, i.e., Viî) =
êqhri^^nonzero currents, the equivalent resistance is defined as
Figure 2.18b illustrates the idea o f the equivalence. By Equation 2.1, v^î) = R^,îj{t) = (7?^ +
+
implies that the equivalent resistance is R^^ = R-^ + R^ + Ry This means that from the
perspective o f the voltage source, the series connection o f resistors is equivalent to a single resistor
o f value equal to the sum o f the resistances. A formal discussion o f equivalent resistance and its
generalization (the Thevenin resistance) is taken up in Chapter 6 .
Exercise. In Figure 2.18, suppose R^ = 5R^ and R^ = 2Ry Find R

AN SW ER: R^^ = GR^,
^
^rid
and
Example 2.4 suggests some generalizations. Consider Figure 2.19. The first is that the equivalent
resistance R^^ seen by the source is the sum o f the resistors. This means that resistances in series add,
i.e., resistors in series can be combined into a single resistor whose resistance is the sum o f the indi
vidual resistances.
Req = R \ + R 2+ "' + Rn
Further, since vi^t) = Rjii{t), a general voltage division formula can be derived as
Ri
R\ +
+ ' + Rn
(2 .2 b)
fory = 1, ... , n.
67
FIG U RE 2.19 Series circuit o f n resistors driven by a voltage source.
Exercise. In Figure 2.19, suppose each resistor has value Rq. Find the equivalent resistance seen by
the source and the voltage across each resistor in terms o f the source voltages.
A N SW ER: nR^^, v jt)ln
EXA M PLE 2.5

Find the equivalent resistance seen by the source and the voltages
2.20. W hat is the power dissipated in the 14-Q resistor if
and Vj for the circuit o f Figure

= 2 V?
FIG U R E 2.20 Series circuit containing a dependent voltage source.
So l u t io n
From the preceding discussion, R^^ is defined by Ohms law, i.e., v-J^t) = Rgî-Js)Step 1. Express v-^ in terms o f the remaining branch voltages. From KVL,
în = Vi + V2 + 2vj = 3vi + V2
(2.3)
68
Step 2 . Express the branch voltages in terms o f
and substitute into Equation 2.3. To express
and ^2 in
terms of i^^, observe that i^^ is the current through each resistor (KCL or definition o f two-terminal circuit
element) and use Ohms law:
= 2i-^ and Vj = 14/^^. Subsrimting into Equation 2.3 yields

^ in =
Therefore,
20// = R e q iin
= 20
Notice that the dependent source increases the resistance o f the two series resistors by 4 Q.
Dependent sources can increase or decrease the resistance o f the circuit. W ith dependent sources,
it is even possible to make the equivalent resistance negative.
Step 3. Find the power absorbed by the 14-Q. resistor. To find the power absorbed by the l4-2 resis
tor when
= 2 V, first compute i^^ via Ohm s law; i^^ = v J R
=
= 0 .0 1 x 1 4 = 0.14 W.
= 2/20 = 0.1 A. It follows that P
Exercise. Suppose the dependent source in the circuit o f Figure 2 .20 has its value changed to 2[v-^
+ V2 ). Find R^q.
AN SW ER: 48 Q
5. PARALLEL RESISTANCES AND CU RREN T DIVISION

Many o f the electrical outlets in the average home are connected in parallel. W hen too many
appliances are connected to the same outlet or set o f oudets on the same fused circuit, a fuse will
blow or a circuit breaker will open. Although each appliance uses only a portion o f the maximum
allowable current for the (fused) circuit, together, the total current exceeds the allowable limit.
Because o f this common occurrence, an engineering student ought to know how current distrib
utes through a parallel connection o f loads (resistors).
To keep the analysis simple, consider a set o f three parallel resistors driven by a current source.
E X A M P L E 2 .6
Figure 2.21a shows a circuit o f three parallel resistors driven by a current source. Our objectives are to
find expressions for v-Jyf), ij^t) in terms o f the input current i-JJ) and the circuit conductances (the recip
rocal o f the circuit resistances) and the equivalent resistance seen by the current source.
-o
ijt )
-I-
- f-
v jt )
ijt )
v Jt)
-O
(a)
(b)
FIG U RE 2.21 (a) Three parallel resistors driven by a current source;

(b) equivalent resistive circuit as seen from source.
69
So l u t io n
Step 1. Find expressions for ij{t) in terms o f v-J^t). The variable that Hnks the branch current ij{t) to
the input current i^J^t) is the voltage
which by KVL appears across each resistor. Since v-J^t)
appears across each o f the resistors, Ohms law implies that each resistor current is
ij{t) = ^
^
(2.4)
where Gj = HRj is the conductance in siemens andy = 1, 2, 3.

Step 2. Compute v^î) in terms ofi-J^t). Applying KCL to the top node o f the circuit yields
ii{t) = /j(?) + z'2 (?) + i^{t)

Using Equation 2.4 to substitute for each ij{t) and then solving for v-^ yields
(j\ + Cj2
R]
Rn
êq
^3
R^
(2.5)
Step 3. Compute ijyi) in terms ofi-J^t). To obtain a relationship between i-J,t) and ij{t), substitute
Equation 2.5 into Equation 2.4 to obtain
1
G,
+
+
Rl R2
R3
G, ,
G 1 + G 2 + G3
(2 .6)
Equation 2.6 is called a current division formula. It says that currents distribute through the
branches o f a parallel resistive circuit in proportion to the conductance o f the particular branch
G. relative to the total conductance o f the circuit G^^ = G^ + G2 + Gy The greater the conduc
tance, i.e., the smaller the resistance, the larger the proportion o f current flow through the associ
ated branch.
Step 4. Compute the equivalent resistance
= G^ v-^, defines G^^ or, equivalently,
seen by the source. As in Example 2.5, Ohms law, i-^

From Equation 2.5, G^^ = G j + G2 + G3 is the equiva
lent conductance o f the parallel circuit, and the equivalent resistance is
R.
1
+
R,
R2
The idea is illustrated in Figure 2.21b.
R3
G 1 + G2 + G 3
G ,,
70
Exercise. In Figure 2.21a, suppose

through
if
= 1Oe~ A.
= 1 Q,
= 0-5
= 0.5 O.. Find the current
and
A N SW ER: z, (?) = 2eÂ
T he Example 2.6 suggests a very important property. Since
hn ={G\ + G2 + G 3 )v, = G Vj
in addition to implying that G
= G^ + G2 + G^ is the equivalent conductance seen by the source
o f the parallel circuit, one can further interpret this to mean that conductances in parallel add to
form equivalent conductances. This parallels the property that resistors in series add to form equiv
alent resistances. O n the other hand, resistances in parallel do not add, and conductances in series do
not add. We can conclude that from the perspective o f the source, the parallel circuit o f Figure
2.21a has the equivalent representations given in Figure 2.21b.
These ideas generalize to n resistors in parallel, as illustrated in Figure 2.22. In particular, the
equivalent resistance R
o f the parallel set o f resistors in Figure 2 .22 is
Ren = - 1
R,
R2
R.,
(2.7)
êq =
+ *^2 + +
equivalent conductance. Further, the current through each
branch satisfies the general current division formula
R,
G/
1 1
+ ---- -!- +
R
^1
^2
rj A
A-j-+ rG
G
+. r
G2 +
G, .
~ Gr^ q
(2 .8 )
o
-Iijt )
( f
V,
(t)
-o
FIG U RE 2.22 Parallel connection o f n resistors driven by current source.
Exercise. Consider Figure 2.22. Suppose ten 10 Q resistors are in parallel. Find
71
and the cur
rent through each resistor.

AN SW ER:
\ Q. and each current is Q.\i-J^t)
E X A M P L E 2 .7
Consider the circuit o f Figure 2.23 exhibiting a current source driving two parallel resistors. Show
that
Ry
-O
FIG U RE 2.23 Two resistors in parallel driven by a current source.
So l u t io n
Step 1. Find the equivalent resistance seen by the current source. From Equation 2.7, with = 2, it
follows that
R^R2
Ri ^ Ri
This formula, called the product over sum rule, is quite useful in many calculations.
Step 2. Find i^{t) and iît)- From Equation 2.8, with w = 2, it follows that
R^
Gi
X 7X
Ry
and
/2 (0 =
Gi 4- G 2
Ri
,__ Ri + R'^
R2
72
Exercise. In Figure 2.23, suppose
= 12 A and
Find Rj so that
= 10
= 4 A.
AN SW ER: /?, = 20 Q
E X A M P L E 2 .8
the current ijit) through i?2> and the
For the circuit o f Figure 2.24, find the input voltage

instantaneous power absorbed by Rj when
5e
^ 0
t< 0
y r i,(t)
^2
ijt )
U i3 (t)
SG 3
l|i,( t )
<G ,
> = 0 .0 5 U > = 0 .1 5 U > = 0 .0 2 u V = 0 . 0 3 0

-O FIG U RE 2.2 4 Parallel connection of four resistors.
So l u t io n
Step 1. Compute the equivalent conductance and equivalent resistance o f the circuit. Since conduc
tances in parallel add,
^ e q - G j + G 2 + G 3 + G 4 = 0.25 S
and
Step 2 . Compute v^JJ). From Ohms law, the voltage across the current source is
2 0 e~ 'V t^ 0
0
t< 0
Step 3. Compute the current i2 (t). Using the current division formula o f Equation 2.8 yields
3e~'A t > 0
Cjeq
0.25
r<0
73
Chapter 2 K irchhoffs Current & Voltage Laws and Series-Parallel Resistive Circuits
Step 4. Compute the power absorbed by Rj- To compute the power absorbed by
P 2 (0 = v,(O x i2 {t) = 1'2 (0 r ^2 =
for t >0,
Exercises. 1. For the circuit o f Figure 2.24, find i/^(t) and the power absorbed by i?4 .
A N SW ER: 0.6^-^ A, 12
2. In the circuit o f Figure 2.24, suppose each conductance is doubled and i-J^t) = 100 mA. Find
R , V- (?), and the power absorbed by the new Gy

AN SW ER; 2 Q, 200 mV, and 1.6 mW
6. SERIES-PARALLEL IN TERCO N N ECTIO N S

The last two sections covered series and parallel resistive networks. Suppose we take a series circuit
and connect it in parallel with another series circuit; this is a parallel connection o f two series cir
cuits. Alternately, we could take two parallel circuits and connect them in series. This would result
in a series connection o f parallel circuits. We could also put a series connection o f two parallel sub
circuits in parallel with a replica o f itself or some other series or parallel circuit. Many other inter
connections are possible. Arbitrary series and parallel connections o f such subcircuits are called
series-parallel circuits. This section explores the calculation o f the equivalent resistance o f seriesparallel circuits by repeated use o f formulas for series and parallel resistance computation. Related
voltage and current computation is also explored. Example 2.11 presents a practical application o f
series-parallel concepts.
EXA M PLE 2 .9
Find the equivalent resistance, R^^, and the voltage across the source,
the voltages
V2 , the
power absorbed by the 6 kQ resistor, and the power delivered by the source for the circuit o f
Figure 2.25, when
= 20 mA.
FIGURE 2.25 Series-parallel resistive circuit.
74
So l u t io n
Step 1. Compute
first compute R^^y and
To compute
8 x 4.8
Kq\ = ------=
1+ 4 .8
38.4
= 3 kQ
12.e
and
^^^2
1 ^ 6
12
---- 1---- 1--- ------12
The resistance in parallel with the 2k2 resistor is, say,
Reqh ~ 1000 + R^qX + R^q2 ~ 6

Finally,
R, = IkQ. /!R,. =
2 +6
= 1.5 kQ
Step 2. Compute V-^ . From Ohms law.
= Reqhn = 20 X 1.5 = 30 V
Step 3. Compute Vj and V^- By voltage division.
'
Req-i
= >5 V and Vj = - ^ v ; = ?
Step 4. Compute the power absorbed by the 6
resistor.
^ (V 2 )^ ^ 1 0 0
6000
6000
1
60
Step 5. Compute the power delivered by the source.
^source ~ înhn 30 x 0 .0 2 = 0 .6 W
Exercise. 1. W hat is the current through the 2 k2 resistor firom top to bottom?
AN SW ER: 50 mA
2 . In Example 2.9, suppose the resistance o f each resistor is doubled. Find the new R^ and the
AN SW ER: 3 k n , 1.2 watts
This example points out a very interesting fact: finding the equivalent resistance o f a series-paral
lel connection o f resistors requires only two types o f arithmetic operations no matter the network
complexity: adding two numbers and taking the reciprocal o f a number. A hand calculator easily
executes both operations. Such is not the case with a non-series-parallel network. To find the
equivalent resistance o f a non-series-parallel network, one usually must write simultaneous equa
tions and evaluate determinants, a topic detailed in Chapter 3 .
75
It is then important to recognize when a problem belongs to the series-parallel category in order to take
advantage o f the simple arithmetic operations. In the previous series-parallel examples, one and only
one independent source was specified on the circuit diagram. This is part o f the definition o f a seriesparallel network. The independent source must be indicated, or, equivalendy, the pair o f input termi
nals to which the source is connected must be specified. The specification o f the input terminals deter
mines whether or not a network is series-parallel. The following example illustrates the effect of differ
ent input terminal designations on the computation o f equivalent resistance.
E X A M P L E 2 .1 0
For the circuit o f Figure 2.26a, determine whether or not the network is series-parallel as seen
from each o f the following terminal pairs:
1.
2.
C a s e l;( A , B)
Case 2; (A, C)
3.
Case 3: (C, D)
If the answer is affirmative, give an expression and compute the numerical value for the equivalent
resistance, using the notation // (double slash) for combining resistances in parallel, i.e.,
and
are
parallel, and RÎIR2 llR^ means
is in parallel with
means
vvhich is in parallel with Ry
(a)
>R,
A<
R1
D'
(0
FIG U RE 2.26 (a) From terminals (C, D) the network is not series-parallel. However, from terminals
{A, B) the network is a series-parallel one. (b) Redrawing of the network of (a) as seen from terminals
(A, Q; the resulting network is series-parallel.(c) Non-series-parallel network seen from (C, D).
76
So l u t io n
Case 1. Find equivalent resistance seen at (A, B). W ith an independent source connected to nodes
A and B, the source sees a series-parallel network. By inspection o f Figure 2.26a, the equivalent
resistance is
i?eq = ^ i //[(^2 +
+ ^ 5)] = 20//[(4 + 6)//(2 + 8 )] = 4
^
I
Case 2. Find equivalent resistance seen at (A, Q. W ith (A, Q as the input terminal pair, the net
work is again series-parallel. This is made apparent by redrawing the network, as shown in Figure
2.26b, from which
i?eq =
+ [(i ?4 +
= 4//{6 + [(2 + 8)//20]} = 3.0 4 Q
Case 3. Find equivalent resistance seen at (C, U). W ith (C, D) as the input terminal pair, the net
work is not series-parallel, as can be garnered from Figure 2.26c. T he calculation o f
for this
case requires methods to be discussed in Chapter 3 and is omitted.
Exercise. 1. In Figure 2.26b, suppose

AN SW ER: 3.11 Q
is changed to 40 Q. Find
In electrical engineering laboratories, a student often uses a meter to measure voltages associated
with a piece o f electronic equipment. In older laboratories, or when using an inexpensive meter,
the voltage reading will sometimes differ from what the student calculated or expected to meas
ure. Typically, this results from the loading effect o f the meter. Using the concept o f series-paral
lel resistances, the following example explores the phenomenon o f loading.
E X A M P L E 2 .il
Suppose the circuit in Figure 2.27a is part o f a laboratory experiment to verify voltage division. In
this experiment, you calculate the expected voltage Vq and then measure the circuit voltage using
an inexpensive voltmeter.
(a)
Calculate the expected voltage Vq in Figure 2.27a.
(b)
A voltmeter with a 1-kQ/V sensitivity is used to measure
V q.You
use a 0 -10-V range. In
this range, the meter is represented by a 10-kQ resistance, i.e., 10 kD = full-scale reading
meter sensitivity = 10 V x 1 kQ/V. W hat voltage will the meter read?
(c)
A better-quality voltmeter with a 2 0 - k H / V sensitivity is used to measure the same volt-
^0
^ 0 -10-V scale. This better-quality meter is represented by a 2 0 0 -k tl resist
ance. W hat new voltage will the meter read?
77
" h
10 kQ
r-O
+
15V^
10 kO
20 ko
f-O
-1-
IS v X
>10 kO
20 kO
lOkQ
r-O-----1-
20 kQ
-o
FIG U RE 2.27 Three circuits for exploring the effect of loading on a circuit: (a) circuit for validating
voltage division; (b) circuit o f (a) with an attached voltmeter having an internal resistance o f 10 k2;
(c) circuit o f (a) with an attached voltmeter having an internal resistance of 200 k^2.
So l u t io n
(a) Voltage division on the circuit o f Figure 2.27a yields
20
Vo =
(b)
-15 = 10 V
O n the 0-10-V range, the voltmeter internal resistance between the probes is 10 kD, as
stated. This represents a 1G-Id2 load connected in parallel with the 20-kQ resistance, as
shown in Figure 2.27b. The voltage Kg will now change because the 15-V source no
longer sees 10 ld2 in series with 20 kO. Rather, the source sees 10 kQ in series with 6 .67
k ii = 20 kX2//10 kQ. By voltage division,
6 .67
-15 = 6 V
Vo = 10 -H6 .67
This is a 4 0 % deviation from the true answer, V q = 10 V, as calculated in part (a).
(c)
O n the 0 -10-V range with the better voltmeter, the internal resistance between the probes
is 200 kD. As before, this represents a 200-kQ load connected in parallel with the 2 0 -k 0
resistance, as shown in Figure 2.27c. 20 kQ//200 klQ = 18.18 kO. By voltage division,
this yields
18.18
Vb =
-15 = 9 .6 7 7 V
10 + 18.18
This 3.23% deviation is within a reasonable tolerance o f the precise answer o f 10 V.

Example 2.11 demonstrates the effect o f loading due to a measuring instrument, emphasizing the
importance o f choosing a good voltmeter with adequate sensitivity. Although modern-day volt
meters typically have sensitivities better than 20 kQ/V, a meter with a sensitivity o f 1 kQ./Y is used
in the example to dramatize the effect o f loading.
Exercise. 1. Repeat Example 2.11 if the 20-kQ resistance is changed to 40 kO.

AN.SWER: 12 V, 6.667 V, 1 1.538 V
78
2. Th e circuit o f Figure 2.28 shows a voltage divider whose voltage Kq is to be measured by a volt
meter having an internal resistance o f 80 kO. Find Kq without the meter attached, and then find
the value o f Vq measured by the meter.
20 V
FIG U RE 2.28 Voltage divider circuit.

AN SW ER: 15 V, 13.71 V
7. D EPEN D EN T SO U RCES REVISITED

Chapter 1 introduced the notion o f a dependent or controlled source whose voltage or current
depends on the voltage or current in another branch o f the network, i.e., each source has a con
trolling voltage or current and an output voltage or current. Figure 2.29 depicts the four types o f
controlled sources designated by a diamond containing either a or an arrow:
1.
2.
Voltage-controlled voltage source
(VCVS)
Voltage-controlled current source
(VCCS)
3.
Current-controlled voltage source (CCVS)
4.
Current-controlled current source (CCCS)
An arrow inside the diamond indicates a controlled current source having the reference current
direction given by the arrow. A inside the diamond specifies a controlled voltage source, with
the reference voltage polarity given by the sign. A parameter value completes the specification
o f a linear controlled source. In Figure 2.2 9 the (constant) parameters are fx, g^, r^, and |3. These
symbols are common to many electronic circuit texts and have useful physical interpretations to
practicing engineers and technicians. For consistency, a ^^-type controlled source is a V C C S and
a jO,-type source is a V C V S, and so on.
79
O
+
a ................... o
(a) VCVS or |j-type
(b) VCVS or g -type
O
Pi,
>r
6-
(c )C C V S o rr -type
<;
o - ......... .........o
(d) CCCS or P-type
F IG U R E 2.29 Designations for the various controlled sources.

In practical controllecl sources, the controUing voltage (t>j in Figure 2.29a and b) or current (z'j in
Figure 2.29c and d) is ordinarily associated with a particular circuit element, but not always. For
generality, the controlling voltage
in Figure 2.29a and b is shown across a pair o f nodes. Also,
in Figure 2.29c and d, the controlling current Zj is shown to flow through a short circuit. (Strictly
speaking, neither an open circuit nor a short circuit is a circuit element.) In a real circuit, the cur
rent may be flowing through an actual circuit element, such as a resistor or even a source.
In Figure 2.29b, once the controlling voltage v-^ is known, the right-hand source behaves as an
independent current source o f value
Since the unit for
is volt, the unit for^^ is amperes per volt, or siemens. Since
is amperes and the unit for

has units o f conductance, and the
controlling and controlled variables belong to two different network branches,
is called a trans
fer conductance, or transconductance. The other controlled sources have a similar interpretation.
The parameter
has the unit o f resistance, ohms, and is called a transfer resistance. The param
eter |i is dimensionless because the controlling voltage

has units o f volts and the output vari
able
must have units o f volts. Similarly, the parameter (3 is dimensionless. The units and asso
ciation are set forth in Table 2.1.
80
Table 2.1 Units and Association.

Type
vcvs
VCCS
ccvs
cccs
Unit
Appellation
dimensionless
Voltage gain
Parameter
siemens
Sm
Transfer
conductance
ohm
Transfer resistance
dimensionless
Current gain
Figure 2.29 portrays each controlled source as a four-terminal device. In practical circuits, the
great majority o f controlled sources have one terminal or node in common, making them threeterminal devices. The dashed lines joining the two bottom nodes in Figure 2 .29 suggest this quite
common configuration.
The controlled sources as defined in Figure 2.29 have linear v-i relationships. Controlled sources
may also have a nonlinear v-i relationship. In such cases, the element will be called a nonlinear
controlled source. This text deals only with linear controlled sources.
The next few examples describe some o f the unique features o f controlled sources.
Exercise. Find v^,
and the power delivered by each source in Figure 2.30.
FIG U RE 2.30
AN SW ER: 4 V, 0.05 A, 1.6 W, 0.05 W
E X A M P L E 2 .1 2
This example analyzes the circuit o f Figure 2.31. The independent voltage source in series with
the 3-Q. resistor represents a practical source discussed at greater length later in this chapter. The
circuit within the box o f Figure 2.30 approximates a simplified amplifier circuit by a V C C S. The
8 -Q resistor is considered a load and could, for example, model a loudspeaker. Two important
quantities o f an amplifier circuit are voltage gain and power gain, which are computed here along
with various other quantities.
(a)
(b)
Find the equivalent resistance

Compute / .
seen by the independent voltage source.
(c)
Compute /out'
(d)
Compute
(e)
Compute the voltage gain
(f)
Compute the power
(g)
Compute the power delivered by the dependent current source.
(h)
Compute
81
delivered to the amplifier.
the power absorbed by the 8-D. resistor.
Compute the power gain, PgJP;-
(i)
FIG U RE 2.31 Practical source (ideal independent voltage source in series with a resistor)
driving a simplified VCCS approximation of an amplifier circuit loaded by an 8 -Q resistor.
So l u t io n
(a)
Since resistances in series add,
(b)
(c)
By Ohms law, / =
= 0.8 mA.
To compute
one must first compute K j. Here one can use Ohms law directly, since
= 3 + 47 = 50 Q.
we know /-^, or one can use voltage division. Doing the calculation by voltage division,
V = 4 0 X 10'^ = 3 7 .6 X 10'^ V
50
Using this value o f
and current division on the right half o f the circuit yields

0.125
79.8 X 3 7.6 X 1 0 '^ = 2 A
0.125 + 0 .0 6 2 5
(d)
follows by Ohms law

K,.= 2
(e)
8 = 16V
The voltage gain with respect to the input signal is
=
= 400
0.0 4
82
(f)
By Equation 1.18, the power delivered to the amplifier circuit is
p.^ =
(g)
= 47
= 47 X 0.82 >< io-<5 = 30.08
The power delivered by the dependent current source is

^VCCS = "ôut ^ 79 .8 K j = 16(79.8 x 0.0376) = 48.01 W
(h)
(i)
Pout is simply the product o f voltage and current delivered to the load
The resulting amplifier power gain is the ratio o f the power absorbed by the 8 -Q load to
the power delivered to the amplifier, P-^,
Pin
= 1.064 X 10^
3 0 ,0 8
Exercise. Suppose the 8-f2 load resistor in Figure 2.30 is changed to 16 Q. Compute
and the power gain.
AN SW ER: 1.5 A, 24 V, 1.197 X 10<^
The analysis in Example 2.12 required only KCL, KVL, and simple voltage divider and/or cur
rent divider formulas. More complicated linear circuits necessitate a more systematic approach. To
see this need, add a resistor between the top o f the 47-Q resistor and the top o f the dependent
current source in Figure 2.31. The methods o f solution used in the example immediately break
down because the circuit is no longer series-parallel; hence, one cannot use voltage division to
compute V j. Chapter 3 will explain more systematic methods called nodal and loop analysis.
Unlike a passive element such as a resistor, which always dissipates power as heat, a controlled
source may generate power as computed in part (g) o f Example 2 . 12, or may dissipate power in
other cases. Since a controlled source has the potential o f generating power, it is called an active
element.
In Example 2.12, the practical voltage source delivers 30.08 pW o f power to the circuit, which is
easy to accept because the source could have been a small battery On the other hand, the con
trolled source generates 48 W. This seems a litde puzzling. Where does the power come from?
W hy not purchase a controlled source at a local electronics store and use it to power, say, a lamp?
Here it is important to recognize that a controlled source is not a stand-alone component picked
o ff the shelf like a resistor. A controlled source is usually constructed from one or more semicon
ductor devices and requires a dc power supply for its operation. The power delivered by the con
trolled source actually comes from the power supply. Here, we use the controlled source to math
ematically model an amplifier and facilitate analysis o f the circuit.
83
W ith simple series-parallel connections o f resistors, the equivalent resistance is always positive.
When controlled sources are present, a strange result may happen, as illustrated in the next exam
ple.
E X A M P L E 2 .1 3
= 2.
for the circuit o f Figure 2.32 when (a) p = 0.5 and (b) p
O^
FIG U RE 2.32 Calculation of
for a circuit with controlled source for two values o f p.
So l u t io n
W ith p unspecified, we can apply KVL to the single loop, noting that
Consequently, (1 - p)V^ =
= V^. Here,
and
i-x
For p = 0.5, R^q = 2R, which means that the dependent source acts like a resistor o f R Q.. In this
case, it absorbs power. O n the other hand, for p = 2, R,^ = -R , a negative equivalent resistance. In
this case, the dependent source acts like a -2R-Q. resistor and, in fact, delivers power to the inde
pendent source. An important conclusion can be drawn from this example: in the study o f linear
circuit analysis, controlled sources allow the possibility o f negative resistances. Since a negative
resistance generates power, it is also an active element.
Exercise. In Figure 2.32, find the values o f p so that R^q - 0 .5 R and R^q - 2R.
A N S W E R :-!, 0.5
84
Exercise. For Figure 2.33, find
for the following three values of^^: 0.5 mS, 1 mS, and 2 mS.
FIG U RE 2.33
AN SW ER: 2 kl, open circuit, -1 kQ
8. M O D EL FOR A NON -IDEAL BATTERY

The ideal battery o f Figure 1.30, repeated in Figure 2.34a, delivers a constant voltage regardless
o f the current drawn by a load. T he i-v plane characteristic is a horizontal line through V^, as
shown in Figure 1.40b and repeated in Figure 2.34b. Ideal batteries do not exist in the real world.
The terminal voltage always depends on the supplied current. A more accurate representation o f
a practical battery, but by no means a fully realistic one, is an ideal battery in series with a resist
ance, say, R^, as shown in Figure 2.34c. R^ is termed an internal resistance, which crudely models
the effects o f chemical action and electrodes inside the battery.
>
Vs
(a)
(b)
FIG U RE 2.34 (a) Ideal battery; (b) i-v battery char

acteristic; (c) battery model with internal resistance to
crudely approximate effects of chemical action and
presence o f electrodes; (d) nickel-cadmium battery.
(c)
85
E X A M P L E 2 .1 4
This example shows the effect o f the internal resistance o f a battery on the terminal voltage.
Suppose a nickel-cadmium battery has an open circuit terminal voltage o f 6 volts. W hen con
nected across a 2-Q. resistor, the voltage drops to 5.97 V. Find the internal resistance o f the bat
tery.
So l u t io n
Figure 2.35 illustrates the situation. Here, the dashed box represents the battery model with inter
nal resistance R^. In Figure 2.35a, no load is connected to the battery. Hence, no current flows
through the internal resistance, in which case, the terminal battery voltage is 6 V.
(a)
(b)
FIG U RE 2.35 Battery model with internal resistance; (a) open circuited (Is = 0); and
(b) connected to a 2 -0 load.
Figure 2.35b shows the battery connected to the 2-Q resistive load. The measured voltage is 5.97
V. By KVL, the voltage across the internal resistance,
law, the current through
is
is Vj^ = 6 - 5-97 = 0.03 V. From Ohms
= (5.97/2) = 2 .985 A. Again, by Ohms law.

0.03
2.985
= 0 .0 1 0 0 5 Q
Exercise. In Example 2.14, suppose the internal resistance is known to be R^ = 0.005 Q and
although the load resistance is unknown, the load current is 4 A. W hat is the voltage across the
load resistance, and what is the load resistance?
AN SW ER; 5.98 V and 1.495 ^
9. NON -IDEAL SO URCES

Ideal voltage sources have zero internal resistance. Real voltage sources, such as batteries, have an
internal resistance. The value o f this resistance may change with the current load. There may also
86
be other effects. However, for our purposes, a more realistic model o f a voltage source contains a
series internal resistance, as illustrated in Figure 2.36a.
L .(t)
(b)
FIG U RE 2.36 (a) A non-ideal voltage source as an ideal voltage source with an internal series resist
ance; (b) a non-ideal current source as an ideal current source with a parallel internal resistance.
Ideal current sources have infinite internal resistance. Real current sources have a finite, typically
large, internal resistance. Figure 2.36b depicts a more realistic current source model where the
internal resistance is in parallel with the ideal current source.
In the case o f constant voltage and current sources, ideal and non-ideal source models have a
graphical interpretation. The i-v (current-voltage) characteristic o f an ideal constant voltage
source {v^{t) = 1^) is a horizontal straight line. This means that the voltage supplied by the source
is fixed for all possible current loads. An ideal constant current source (z^(z) =
has a vertical
straight line characteristic, which means that the current is constant for all possible voltages across
the source. Figure 2.3 7 illustrates these relationships graphically.
Vout
(a)
FIG U RE 2.37 v-i characteristics o f (a) an ideal constant voltage source, and
(b) an ideal constant current source.
87
The non-ideal case is quite different. Because o f the internal resistance a non-ideal constant volt
age source i-v characteristic satisfies the linear relationship
ôut
^sôut
Ks'
and for a non-ideal constant current source in which
(2 . 10)
= l/R^,
hut ^s'ôul
( 2 . 11 )
Equations 2.10 and 2.11 are illustrated by the graphs in Figure 2.38 when v^{t) =
ideal voltage source and i^{t) =
for the non
for the non-ideal current source. For a voltage source, if the value
o f R; is very small in comparison with potential load resistances, as ordinarily expected, then the
hne in Figure 2.38a approximates a horizontal line, the ideal case. O n the other hand, for a cur
rent source, the line in Figure 2.38b approximates a vertical line whenever
is much much larg
er than a potential load resistance. This would then approximate the ideal current source case.
FIGURE 2.38 v-i characteristics of (a) non-ideal constant voltage source, and (b) non-ideal current source.
In a similar way, non-ideal dependent voltage sources are a connection o f an ideal dependent
source with a series resistance. A non-ideal dependent current source is a connection o f an ideal
dependent current source with a parallel resistance.
E X A M PLE 2 .1 6 Figure 2.39 shows the measured voltages o f a dc power supply found in an old
laboratory. Assuming a non-ideal model o f Figure 2.38a, find
the power supply.
and the internal resistance R^ of
88
Chapter 2 K irch h offs Current & Voltage Laws and Series-Parallel Resistive Circuits
Vout (V)
' '
FIG U RE 2.39 Graph of measured voltages and currents for a dc power supply.
So l u t io n
= -R^
From Equation 2.10, we know that
+ V^. From the graph, when
= 0,
10 V = I/. Further, R^ = - (9.8 - 10)/(0.5 - 0.0) = 0.4 Q.
SUM M ARY
This chapter has presented the essential building blocks o f linear lumped circuit theory, beginning
with the two fundamental laws for interconnected circuit elements: KVL and KCL. KVL states
that for lumped circuits, the algebraic sum o f the voltages around any closed node sequence o f a
circuit is zero. Similarly, KCL says that for lumped circuits, the algebraic sum o f the currents enter
ing (or leaving) a node is zero.
These laws in conjunction with Ohms law allowed us to develop a voltage division and a current
division formula. The voltage division formula applies to series-resistive circuits driven by a volt
age source. The voltage developed across each resistor was found to be proportional to the resist
ance o f the particular element relative to the equivalent resistance seen by the source. For exam
ple, in a two-resistor series circuit, Rj in series with
we found that
Vi =
7^1 + /?2
T he current division formula applies to parallel-resistive circuits driven by a current source. Here,
the current through each resistor with conductance Gi was found to be proportional to G/ divid
ed by the equivalent conductance seen by the source. Since conductance is the reciprocal o f resist
ance, the idea can also be expressed in terms o f the resistances o f the circuit. For example, in a tworesistor parallel circuit, Rj is parallel with /?2>
/i =
G,
G 1 + G2
R, +Rn
In deriving the voltage division formula, we learned that the resistances o f a series connection o f
resistors may be added together to obtain an equivalent resistance, prompting the phrase resistors
in series add. Analogously, the derivation o f the current division formula for parallel circuits led
Chapter 2 Kirchhofifs Current & Voltage Laws and Series-Parallel Resistive Circuits
US
89
to conclude that a parallel connection o f resistors has an equivalent conductance equal to the
sum o f conductances. This is sometimes expressed in terms o f resistances as the inverse o f the sum
o f reciprocal, i.e., for n resistors in parallel,
p .
7^1
which leads to the very special formula for two resistors in parallel,
often referred to as the product over sum rule.

Dependent sources, first introduced in Chapter 1, were re-examined in greater detail. Some prac
tical points were described.
All o f the above ideas were applied to the analysis o f series-parallel networks that are interconnec
tions o f series and parallel groupings o f resistors. Our analysis showed us how to compute the
equivalent resistance o f series-parallel circuits. An example was given that described the applica
tion o f these ideas to voltage measurement. This was followed by a discussion o f battery models
and battery usage. Finally, battery modeling ideas were used to describe non-ideal source models.

Branch: a two-terminal circuit element denoted by a line segment.
Closed node sequence: a finite sequence o f nodes that begins and ends with the same node.
Closed path: a connection o f devices or branches through a sequence o f nodes so that the con
nection ends on the node where it began and traverses each node in the connection only
once.
Connected circuit: one for which any node can be reached from any other node by some path
through the circuit elements.
Current division: the current in a branch o f a parallel-resistive circuit is equal to the input cur
rent times the conductance o f the particular resistor, Gj, divided by the total parallel con
ductance o f the circuit,
= G^ + ... + G^.
Dependent (controlled) current source: a current source whose output current depends on the
Dependent (controlled) voltage source: a voltage source whose output voltage depends on the
Gaussian surface: a closed curve in the plane or a closed surface in three dimensions. A Gaussian
surface has a well-defined inside and outside.
Kirchhoff'* current law (KCL): the algebraic sum o f the currents entering a node o f a circuit
consisting o f lumped elements is zero for every instant o f time. In general, for lumped
circuits, the algebraic sum o f the currents entering (leaving) a Gaussian surface is zero at
every instant o f time.
90
Chapter 2 KirchhoflFs Current & Volt:e Laws and Series-Parallel Resistive Circuits
KirchhoF$ voltage law (KVL); for lumped circuits, the algebraic sum of the voltage drops
around any closed path in a network is zero at every instant of time. In general, for
lumped connected circuits, the algebraic sum of the node-to-node voltages for any closed
node sequence is zero for every instant of time.
Node: the common connection point between each element; in general, a node is a connection
point of one or more circuit elements.
Node voltage: the voltage drop from a given node to the reference node.
Parallel circuit: a side-by-side connection of two-terminal circuit elements whose top terminals
are wired together and whose bottom terminals are wired together.
Series circuit: a sequential connection o f two terminal circuit elements, end-to-end.

Voltage division: each resistor voltage in a series connection is a fraction o f the input voltage equal
to the ratio of the branch resistance to the total series resistance.
// (double-slash): notation for combining resistances in parallel, i.e., RÛR2 means

and i ?2 are
in parallel, and RÎIR2 llRj^ means R^ is in parallel with
which is in parallel with Ry
r\
r\ .
r\
o
n
n
n
91
Chapter 2 K irchh off s Current & Voltage Laws and Series-Parallel Resistive Circuits
PROBLEMS
KIR C H H O FF'S C U R R EN T LAW
1. (a)
Find the value o f /j for each o f the

node connections in Figure P2.1a and
P2.1b given that 1^ = 2 A,
Figure P2.3
1^ = 3 A, and
(b)
= 4 A.
Repeat part (a) when l 2 = I^ = 1^ = 2 A.
A N SW ER: ( b ) 4 A
4. (a)
Find the value o f /j in the circuit o f

Figure P2.4.
(b)
Find the value o f
in the circuit o f
Figure P 2.4 by a single application o f

KCL.
Figure P2.1
A N SW ERS: (b) 0, 2 A
2.
In the circuit o f Figure P2.2, each shaded
box is a general circuit element.

= 20 mA, 1-^2 = ^0 mA,
(a)
Suppose
(b)
I ini = 100 mA> and /-^4 = 0.05 A.

Apply KCL to find /j, Ij, ly and 1^.
Repeat part (a) when
= hni
=
Figure P2.4
A N SW ER: (a) 6 A
KIR C H H O FF'S V O LTA G E LAW
100 mA.
5. (a)
Consider the circuit o f Figure P2.5a

where each branch represents a circuit
element. Find Vj and 1^2-
(b)
Find Kj and
unspecified
for Figure P2.5b. Each

branch
represents
an
unknown circuit element.

.1 0 0
40V
lOOV
Figure P2.2
+ V, -
3. (a)
For the circuit o f Figure P2.3a,
V,
+
VI
Reference Node
5A
Reference Node
(a)
find the value o f the current /j using
lO O v C ^
+
40V
AN SW ERS: (b) (scrambled) 200, -300,

-200, -300 mA
(b)
Figure P2.5
only a single application o f KCL.

(Hint: Construct a Gaussian surface.)
(b)
(c)
Repeat part (a) for 1^

True-False:
can be uniquely deter
mined as in part (a) and part (b).
6 . (a)
For the circuit o f Figure P2.6, deter

mine the voltages f j, ... , ^4 and the
power absorbed by each resistor.
92
(b)
Now determine the node voltages V^,
Vg, Vq and
with respect to the
reference
indicated
node
by
the
ground symbol.
(c)
Compute
50 V
SC RA M BLED AN SW ERS: (a) 2 V, - 2 V
K C L A N D KVL
9. (a)
Consider the circuit o f Figure P2.9a. Use

KCL and KVL to find the voltage across
each current source from the arrow head
Figure P2.6
to the arrow tail and the current through
A N SW ERS: (a) 180 V, 50 V, -110 V, 10 V
each voltage source from minus to plus.

Finally, find the power delivered by each
7. (a)
Find the values o f the voltages Vp Kj,
source and verify conservation o f power,
and Vj in the circuit graph o f Figure
(b)
P2.7, where each branch represents a
For the circuit o f Figure P2.9b, find

the voltages
and V^.
circuit element.
(b)
Now determine the node voltages
K l ^B
respect
to the reference node indicated by
in
iUV
6^
4A
30V
3A
10
the ground symbol.
(c)
Compute
and
(a)
(b)
Figure P2.9
AN SW ERS: ( b ) - 1 7 V 14 V
10.
For the circuit in Figure P 2.10, calculate
the power delivered by each o f the eight inde

pendent sources. Verify the principle o f con
servation o f power.
Figure P2.7
SC RA M BLED AN SW ERS: (a) -65 V, 15 V, 15
V; (b) 35 V, - 4 5 V, 10 V, - 5 V
8 . (a)
Use KVL to determine the voltages
(b)
1/ and V; in the circuit o f Figure P2.8.

Now compute V^g.
Figure P2.10
A N SW ERS: - 4 , - 9 , - 3 6 . 35, 10, 0 , 10, - 6 W
93
11.
(b)
Four circuit elements and a dependent
Supposing that /j = 4 A, /, = 2 A, and

and the power
a = 0.25, determine
delivered to Rj^.
voltage source are shown in the circuit o f Figure

P 2 .11. The current through and the voltage
across each element are identified on the dia
gram. However, one and only one voltage
(or current) value is labeled incorrectly. Mark
the incorrect voltage (or current) on the circuit
diagram and give the correct value for this volt
age (or current).
Figure P2.14
A N SW ERS: (b) 80 Q, 20 W
15.
Consider the circuit in Figure P2.15 in

= 1 A and R^ = 84 Q. Find the value
which
o f R^ for each o f the following cases:
25V
(a)
The power delivered by the source is
(b)
T he power absorbed by
(c)
watts.
The power absorbed by R^ is 13.44
13.44 watts.
Figure P 2 .ll
Find the currents and voltages /^, V^, 1^,
12.
and
is 13.44
watts.
in the circuit o f Figure P2.12.
Figure P2.15
SC RA M BLED A N SW ERS: 21, 16, 336, 56
Figure P2.12
KCL, KVL, A N D O H M 'S LAW

13.(a)
16 . Consider the circuit o f Figure P2.16.

(a)
(b)
If /? = 5 Q, find Vjf
Find the value o f R when
For the circuit o f figure P2.13, sup

= 875 mA and Rj_ = 80Q .
pose
Find V-^, /j, /j, and ly

(b)
= 40 V.
Now suppose that

= 7 A and -^2 ~ ^
A. Find V-^, ly Rj^, and the power
120V
4on
delivered to the load Rj^.
50V
20V
Figure P2.16
+
V.
AN SW ER: (b) 25 Q
200
,4on
Figure P2.13
C H E C K : (b) P^= 160 watts
I4.(a)
For the circuit o f Figure P2.13, deter

mine
in terms o f /j, a and Rj^.
17. For the circuit o f Figure 2.17, find

(a)
the voltage Vj and the power absorbed
by the 10 Q resistor;
(b)
(c)
the voltage V2 ,
the power delivered by each source.
94
R
>J .
10 k n ,
6kO
Figure P 2.17
AN SW ER: (c) 12.5 watts and - 7 .5 watts
'+
V,
(a)
Figure P2.20
18. Find the power absorbed by the unknown

circuit element x and the voltage
in the cir
cuit o f Figure P2.18.
21.
500
50 V
6x 10"
ANSWERS: (b) Vi =
3V
'in
4 8 V 4 kQ
i 6xl 0' -i-/?-i-6axl0
The circuit o f in Figure P2.21 is a blower
motor control for a typical car heater. In this circuit,
0.8 A
<----
resistors are used to control the current through a

motor, thereby controlling the fan speed.
Figure P2.18
19. (a)
(b)
Ignition
Switch
JUA
Find the current 7/j and the voltage

Kuf
circuit o f Figure P 2 .19.
If a resistor o f 7? Q is placed across
the output terminals, determine the
current
and the voltage
and
12V
Chassis .
Ground
the power delivered by the 10 V

14V
44 V
'r
3R
10V
(a)
W ith the switch in the Lo position,

the current supplied by the battery is
source.
2.5 A. The voltage drops across the
Figure P2.19
C H EC K : (b) K^,= 4 V
20. (a)
^BC= 1-5 V, Vc/j = 0.625 V, and Vq

= 3 .125 V. Consider the motor as rep
In Figure P2.20a, Vj = 32 V and the
resented by a load resistance.
power delivered by the source is 80
(i)
mW. Compute
(b)
resistors and motor are Vjg = 6.75 V,
, V-^, and R.
In Figure P2.20b a dependent voltage

source has been added to the circuit o f
Figure P2.20a. Suppose
Determine
= 40 V.
in terms o f a and R. If
= 0.8 mA and a = 5, find K, and R.

(c)
Determine the value o f each resist

ance and the value o f the equivalent
resistance representing the motor.
(ii) Determine the power dissipated

in each resistor and the power
used by the motor.
(iii) Determine the relative efficiency o f
For each circuit o f Figure P 2.20,
the circuit, which is the ratio o f the
determine the resistance seen by the

voltage source,
power used by the motor to the

power delivered by the battery.
95
(b)
W ith the switch in Med-1 position,
delivered is 1250 W. How many possible medi
determine:
um wattages are there and what are they?
(i)
C H EC K : 10 ohms, 40 ohms
T h e voltage drop across each

resistor.
(ii) The current delivered by the battery.

(iii) The relative efficiency of the circuit.
(c)
(a)
Suppose /? = 20 Q, find the power
(b)
Suppose the power delivered by the
delivered by the current source.
Repeat part (b) with the switch in

position Med-2.
(d)
24. Consider the circuit o f Figure P2.24.
The switch is in the high position. A
current source is 120 watts. Find the
winding in the motor shorts out. The
value o f R.
fuse blows. W hat is the largest equiva

lent resistance o f the motor that will
cause the fuse to blow?
A N SW ERS:
(a)
(i)
Rj^g = 2.7 Q.,
=0.6 0., RcD
(t)
^^^lOOV
= 0 .2 5 Q ,
1.25 a
Figure P2.24
= 16.875 W Pbc = 3.75 W

1.5625 W P , , , , = 7.8125
(ii)
C H E C K : (b) 8 </?< 15
W
(b)
(iii) 26%
(i)K^5=0> ^s c = 3 -4 3 V ,
1-43
25. Given that 4 W is absorbed by the 100-Q

=
= 7.14 V
resistor, find V} and the power delivered by the

voltage source in the circuit o f Figure P2.25.
(ii) 5.71 A
(iii) 59.5%
(c)
(i)
=0.
150 Q
2o on
Vcn = 2V,
Vmotor = 10 V
300n'
lo o n
(ii) B A
(d)
(iii) 83.3%
0.4 Q
22. Suppose one has two resistors /?j = 20 Q

and i ('2 = 20 Q that can be conected to a source,
= 100 V. By connecting the resistors to the
Figure P2.25
26. In the circuit o f Figure P2.26, suppose V2 =
60 V. Find
/^, and the power delivered by
the source.
source in different ways, what are the different

i8 o n
wattages that can be delivered by the source?
6on'
The different types o f connections represent

what might occur in an electric space heater
having a low, medium, and high setting.
'' 6
4o n
9o n
1 800
C H EC K : There are three possible connections

with medium using 500 watts.
23. In Problem 22, find the values o f R^ and
so that the lowest wattage delivered by the
100 V source is 200 W and the highest wattage
Figure P2.26
SCRA M BLED AN SW ERS: 3 360, 840, 4
96
Chapter 2 Kirchhofifs Current & Voltage Laws and Series-Parallel Resistive Circuits
27. Find the power delivered by each independ

ent source and the power absorbed by each resis
tor in the circuit o f figure P2.27. (Check; Total of
delivered power = total o f absorbed power.)
0.7 A
lOon
500
0.8 A
20V
C H EC K : (a) 45 <
(c) 8 0 < i? ^ < 125
200
< 65; (b) -85 < 1^2 < '6 5 ;
Figure P2.27
SC RA M BLED A N SW ERS: 59.5, 9, 8, 49.
0.5, 45, 18
(a)
28.
(b)
Write an equation for
in terms o f a
and 4
For the circuit o f Figure P2.28 with the
indicated currents and voltages, find

(a)
(b)
(c)
If 1/ = 40 V and a = 0.5, find the value

o f the current
(c)
Currents /j through
How much power is delivered by the
Voltages Vj through
independent
Power delivered by each independent
power is delivered by the dependent
source?
How
much
current source
source? Verify the principle o f conser

vation o f power for this circuit.
5 mA
500 0
2000
Figure P2.30
C H EC K : 4 = 0.05 A
Figure P2.28
C H E C K : ^2 = 12 V, /^ = 1 mA,
= 60m W
29. For the circuit o f Figure P2.29, find
V O LTA G E A N D C U R R EN T
D IVISIO N
31. Consider the circuit o f Figure P2.31 in
(a)
Voltage drop V j, and
which
(b)
Voltage drop V2
V2 for each o f the following cases:
(c)
The value o f the unspecified resist
(a)
ance, R
= 30 V and
= 20 V. Find
and
Switch 5j is closed and switch S2 is

open.
(b)
Switch
is open and switch S2 is
closed.
(c)
Switch
closed.
is closed and switch S2 is
97
34. For the circuit o f Figure P2.33,
(a)
= 120 V
= 120 Q. Find the value o f
Suppose
that is necessary to achieve V-^ = 90

V. Compute
(b)
Find the values o f R^ and Rj that are

necessary to achieve
Figure P2.31
= 100 V and
1/2 = 80 V.
32. Construct a series voltage divider circuit
wiiose total resistance is 2400 Q as illustrated
in Figure 2.32.
(a)
Suppose Vj = 0 . 7 5 and Vj = 0 .2 5 V^.

Find the values o f R-^,
(b)
^s'
Suppose V"j = 0 . 8 and V2 = 0.5V^.

Find the values o f
(c)
6on
R2 , and R^.
Suppose K, = 0.81/ and Kj = 0.5

Find the values o f 7?,, R2 , and R^.
Figure P2.34
C H EC K ; (b) 60 Q, 240 Q
35 . Figure P2.35 shows a Wheatstone bridge

circuit that is commonly used in a variety o f
measurement equipment. The bridge circuit is
said to be balanced if R J i j = Ri,R^- In this case,
the voltage
= 0 for any voltage V-^^.
(a) Use voltage division to compute the
voltages
Figure P2.32
and V^. Check:
AN SW ER: (c) R^=R^ = 960 Q, R^ = 4 80 Q
Rc + Rci
33. For the circuit o f Figure P 2.33, suppose
= 48 V.
(a)
Find
(b)
about
with the switch in position A,
i.e., the switch is open.

(b)
Find
with the switch in position B,
= 0, then what must be true

and VJ. Show that
=0
if and only if R^R^ = Rh^cSuppose that RJi^ = Rf^R^ and a 0.5D. resistor is connected across the
If
(c)
i.e., the switch is closed.
terminals. Then
6Q
(d)
find the current
through the 0.5-^^ resistor.

Suppose
= 18 V,
= 3 Q,
= 2 Q, an d R ^ = 2^ - Find
Figure P2.33
C H E C K : (b) 25 <
< 30 V
AN SW ER: (d) 3 V
=6
98
36. Find V^, V , and
for the circuit o f Figure
P2.36 when 1/^ = 50 V and V^ 2 = 25 V.

> ;i,
lokn
60kn
Figure P2.39
lokn
C H EC K :
= 6.4 watts
40. Find /p /2>l y V-^, and the power delivered

by the source in the circuit o f Figure P2.40
when I-^ = 120 mA.
Figure P2.36
SC RA M BLED A N S W E R S :- 5 , 10, 15
37. Consider the circuit o f Figure P 2.37 in
-o
which /.^ = 0.1 A, Gj = 2.5 mS,

(a)
i',
W ith the switch in position A, find

"^^d the power
delivered by the source.
(b)
4kn
'
Kq' h
6kn
9kn
Repeat part (a) if the switch is in posi

tion B.
J 8 kO
-O Figure P2.40
C H E C K : V.^ = 360 V
41. Find /[ and I 2 for the circuit o f figure P2.41
when
= 10 mA, 1^^ = ^
and
= 14 mA.
Figure P2.37
SC RA M BLED AN SW ERS: (a) 50, 0 . 5 ,
5, 50, 20
' 7
38. In the circuit o f Figure P2.38, it is required

that /j = 0.81/^-^. Find R (in Q ),
in terms
o f / , and V- in terms o f / .
r '0f
>
Y
30mS>
Figure P2.41
42. For the circuit o f Figure P2.42, find the cur
rents /p /2, ly and
-O V
'
when
= 300 mA.
j"
3oon> 6 o o n /l
I >i 2on
-o Figure P2.38
i 2on >4on
CH ECK:
= 40/, and
39. In the circuit o f Figure

P 2.39,'
o
in
= 10 mA.
Find /p /j, V and the power delivered by the

dependent source.
Figure P2.42
SC RA M BLED AN SW ERS: - 5 0 , 40, 80 (in
mA)
N.
99
Rpo AND RELATED CALCULATIONS

OF SERIES-PARALLEL CIRCUITS
43.
3000
For each o f the circuits o f Figure P 2.43,
find the value o f
3000
5000,
and the power delivered
" V
if a 10-V source were connected.
5000 .
1.5kO<
J 7500
i6kn
lkO<
Ik0<
Figure P2.45
(b)
SC RA M BLED AN SW ERS: (b) 29.63, 675,
0.1185
46.
Find
for each o f the circuits in Figure
P2.46. Notice that the circuit o f (b) is a modifica

tion o f (a) and that o f (c) a modification o f (b).
2kn
15kn
Figure P2.43
A N SW ERS: 0.5R , 5 kQ, 2.6 kQ
44. Find the value o f R-^ for each o f the
circuits o f Figure P2.44.
O-
Figure P2.46
SC RA M BLED A N SW ERS: 60 kn ,
Figure 2.44 (a)
22.5 kO, 135 k n
1.5R
47. Find R^^ in the circuit o f Figure P2.47

(a)
W hen the switch is open
(b)
W hen the switch is closed
Figure 2.44 (b)

45. For each o f the circuits in Figure P2.45,
compute the equivalent resistance R seen by
the source, the input current
delivered by each source, and
the power
when
80 V.
C H E C K : Answers are the same.
100
48. This is a conceptual problem and requires

8000
no calculations for the answer. Consider cir
>5000
cuits 1 and 2 o f Figure P2.48. All resistors are
2kO
greater than or equal to 1 Q.. We wish to deterO-
mine the relationship between R^^^ and R^ ^ 2

the presence o f the finite positive R-Q. resistor
(e)
Figure P2.49
between points a and b. W hich o f the following
SCRAM BLED ANSWERS: 4 0 0 ,7 0 0 , 500, 1500
statements is true?
> K ql
(a)
(b)

(a)
Kill = êq2
(d)
There
is no
general
relationship
between R , and R^^jrelation

ship depends on the value o f R.
(b)
Explain your reasoning.
Suppose
= 320 V,
= 256 V, R^ =
R^ = 800 n . Find R,
and the
resulting R^^.
Suppose
= 320 V, V^= 192 V, R^ =
400 Q, R^ = 800 Q. Find R,
and
the resulting R
C ircuit 1
Figure P2.50
SC RA M BLED AN SW ERS: 500, 1000, 1600,
400, 170.67, 128
C ircuit 2
5 L For the circuit of Figure P2.51:

(a)
Calculate R^^q the equivalent resist
ance seen at terminals A and C, which
would be the reading on an ohmmeter
if the two probes were connected to A
Figure P2.48
and C, respectively.
49. For each o f the circuits o f Figure P2.49,
find the value o f R that makes
= 1000 Q,.
(b)
Calculate R^q the equivalent resist

ance seen at terminals B and C, which
would be the reading on an ohmmeter
O-
if the two probes were connected to B

and C, respectively.
5000
>3kO
O-
7500
O-
be
mulas? State your reasons without per

(b)
1 .2 kn
forming any calculations.
52Sn
o-
(c)
Can the equivalent resistance
calculated using the series-parallel for
(a)
O-
(c)
(d)
(a)
101
For Figure P2.53a, how many bulbs can

be put in parallel before the 15 A fiise
blows? Given the maximum number
that can be put in parallel without blow
(b)
ing the fuse, find R and V^.

In Figure 2.53b, bulbs BB and C C are
24 watt and 36 watt, respectively, at
approximately 12 volts. Find the inter
52. Some physical problems have models that
nal resistances o f each bulb. How
are infinite ladders o f resistors, as illustrated in
many C C bulbs can be present before
Figure P2.52.
(a)
the 15 amp fuse blows? Given this

at the
number o f C C bulbs, find R^^ and
terminals a-b in figure P2.52a. (Hint:

Since the resistive network is infinite,
the equivalent resistance seen at termi
nals a-b is the same as the equivalent
resistance to the right o f terminals c-d\
this means that the network to
the right o f c-d can be replaced
by what???) Evaluate if
=1Q
and
= 100 Q. This type of
problem is useful for represent
ing series and shunt conductance
(b)
in transmission lines.
(b)
Find
Figure P2.53
at terminals a-b for the ladder
network o f Figure P2.52b.
C H E C K ; (a) n = 16; (b) n = 4
- 0 --------------->
>^1
54. In the circuit o f Figure P 2.54,

V and
= 40 Q. The switch
= 330
closes at ? = 5
s, S2 closes at f = 10 s, 5^ closes at ? = 15 sec,

bO -
and 54 closes at ^ = 20 sec. Plot
and cal
culate RAi) for 0 < /< 25 s.
Figure P2.52
N U M ERICA L AN SW ERS: 10.512, 14.177
Figure P2.54
53.
Consider the circuit o f Figure 2.53a in
which
= 0.5 ^. Suppose each AA-bulb rep
resents a 12-watt fluorescent bulb at approxi
mately 12 volts, having an internal resistance o f
12 a

(a)
Find
max[
and the average
(b)
value o f Zj(^).
Find ijit), max[ i 2 {t)], and the average
value o f i2 {t).
102
Chapter 2 K irch h offs Current & Voltage Laws and Series-Parallel Resistive Circuits
i,(t)
58. W ith the car engine turned off, you have
9kn
3kO <
been listening to the car radio. While the radio
6kn<
is on, you turn the ignition to start the engine.
3cost(2t)V
2kn
You noticed a momentary silence o f the radio.

The following circuit analysis explains this
12V
effect quantitatively. Assume that with the car

Figure P2.55
engine not running, the 12-V car battery is rep
56. Consider the circuits o f Figure P2.56. In
The load due to the car radio is represented by
resented by the model shown in Figure P2.58.

Figure 2.56,
(a)
Find
(b)
= 120 sin(377?) V and

ou f
out'
an equivalent resistance o f 240 Q. The starter
=5
and the instantaneous
motor draws 150 A o f current when the igni
power absorbed by 30 D, resistor.
tion is turned on and before the engine starts.
If
Find
is replaced by a current source,

= 120 sin(377?) mA,
up, find
pointing
and the instanta
neous power absorbed by 30 Q, resis

tor. Does
affect the current through
at the moment when the ignition
switch is turned on. Compare this to the volt

age
before the ignition switch is turned on.
W hy do you think the radio goes silent

momentarily?
the other resistors in the circuit?
ignition
model for
car battery
with engine
not running
Chasis
ground
Figure P2.58
C H EC K : /*3oq = 37.97sin^(377r) watts

57. The circuit o f Figure P 2.57 shows a simple
scheme to determine Rq, the internal resistance
o f the battery model. The loading effect due to
the digital voltmeter may be neglected (consid
er that the meter is represented by an infinite
resistance). W ith the switch open, the meter
reads 12 V. W ith the switch (briefly) closed, the
reading drops to 11.96 V. Find the value o f R^.
CH ECK:
= 900 watts
59. The volume o f a car radio is not much

affected by the on/off state o f the headlights.
The following circuit analysis explains this phe
nomenon quantitatively. Assume that with the
car engine running, the 12-V car battery is rep
resented by the model shown in Figure P2.59.
Notice that the effective voltage o f the
car battery increases due to the effect
model for a
12V battery
o f the alternator while the engine is

running. The load due to the car radio
15Q
Figure P2.57
AN SW ER: 0.0376 Q
(a)
represented by an equivalent resist

ance o f 240 2. At 12 V dc, each head
light consumes 35 W on low beam and
65 W on high beam.
Find the equivalent resistance o f each
headlight on low beam.
103
(b)
v(V)
o f each headlight on high

beam.
(c)
Find
(d)
are turned off.

Find
when the low beams
when the headlights
are turned on.

(e)
Find
when the high beams
are turned on.

(f)
v (V)
60
40 -
How much power does each

high beam consume given your
20 -
answer to part (e)? W hy is this
^--------- 1-----------
value different from 65 watts?

(g)
0.5
1 V3
>i(A)
i(A)
(b)
How much power must the

Figure P2.61
battery deliver to overcome its internal

losses and operate the high beams and
radio.
C H E C K ; (a) 4.11 Q ; (f) 195.58 watts
DEPENDENT SOURCE PROBLEMS

62. In the circuit o f Figure P 2.62, determine
so that the power delivered to
the 5-kQ load resistor is
lOOPy^, where P^^ is the
mstantaneous
power
con
sumed by the 8-kQ resistor.

Equivalently,
is the power
delivered by the non-ideal

voltage source.
Figure P2.59
60. A 50-cell lead acid storage battery has an
open-circuit voltage o f 102 V and a total
internal resistance o f 0.2 Q,.
I f the battery delivers 40 A to a load
(a)
Load
resistor, what is the terminal voltage?

(b)
W hat is the terminal voltage when the

battery is being charged at a 50 A rate?
(c)
Figure P2.62
AN SW ER: 6.25 mS
W hat is the power delivered by the

charger in part (b)? How much o f the
power is lost in the battery as heat?
63. Find the equivalent conductance G

and
then the equivalent resistance R seen by the
SCRAMBLED ANSWERS: 500, 112, 5600, 94
current source 1^ in the circuit o f Figure P2.63
61. A non-ideal constant voltage source, an

ordinary resistor, and a non-ideal constant cur
when
in terms o f the literals R\,
rent source have the v-i characteristics given in

Figure P 2.61. Determine the values o f the
source voltage or current, the value o f the
source internal resistance, and, finally, the value
o f the resistance for Figure P2.61.
= 1 kQ, R2 =
gm = 0-2 mS.
and g^. Evaluate
104
MATLAB PROBLEMS
(a) Find the output voltage,
the output
current (what is its direction), and the

power absorbed by the load (8-Q resis
tor) for the circuit o f Figure P2.66.
C H EC K :
Figure P2.63
= 10 kO.

write a node equation that allows
you to find
in terms o f
Figure 2.66
Then find
(b)
Using MATLAB or equivalent, com

pute and PLO T with appropriate labels
the power absorbed by the load, denot
ed by R^, as Rj^ varies from 8 to 64 Q in
increments o f 1 Q. Also plot the current,
again using MATLAB, as a function o f
R^. At what value o f
is the absorbed
power a maximum? Knowing this is
Figure P2.64
important, for example, when matching
C H EC K : 0.75 S.
loudspeaker resistances to the output

65. In the circuit o f Figure P2.65 r^ =
resistance o f your stereo. For this prob
12.5 kQ and^^ = 12.5 mS:
lem, you should use MATLAB. You will
(a)
Compute the output voltage and out
need to turn in an original printout (no
put current in terms o f

(b)
Compute the voltage gain, Gy=
copies permitted) o f your code and

plots.
l/-^.
Hint: Begin your program
2kn
SmV,
8kQ'
2kn<
with the commands listed
5kO
below. ?? indicates that you
r I,
should
insert
the
proper
number or formula.
8kO<
Figure P2.65
RL = 8:1:64;
% This command generates an array o f
numbers for RL beginning at 8 and

ending at 64 in increments o f 1. I f you
do not end it with a semicolon, it will
list every entry o f the array.
V2 = ??
% This value should be precomputed
I L = ??;
PL = RL .* I L . ^ 2 ;
% Note that because IL and RL are
arrays o f numbers . ^ means to square

each number in the array IL and . *
means to multiply each number in IL by
the corresponding number in the array
for RL.
Beginning your MATLAB solution:

% Define element values
R l= 15; R2= 4; R3= 9; R4= 2; R5=8; R6=18;
% To fin d Req start from right side.
Ra= R4 + R5;
Ga= 1/Ra;
plot(RL, PL)
Gb = Ga + 1/Rl;
grid
% Plot IL in mA
Rb = 1/Gb;
% Continue these additions and reciprocals until
plot(RL, IL *]0 0 0 )
obtaining Req.
grid
% typing grid adds a grid to your
% To fin d Vout requires repeated use o f voltage

and current division formulas.
plot. Always add a grid.

% You can put both plots on the
same graph as follows:
plot(RL,IL*WOO,RL,PL)
% The motivated student might
investigate using the hold
command instead.
Geq = 1/Req;
IRc = 20*Gc/Geq;
V Rb = IRc*Rb;
% Now write down the MATLAB expression for
finding Vout.
AN SW ERS: (a) 3 Q; (b) 24 V
67. The analysis o f series-parallel circuits with

numerical element values can be done with
only two types arithmetic operations: adding
two numbers and taking the reciprocal o f a
number. As such, MATLAB is an extremely
convenient tool for finding the equivalent
resistances and the voltages and currents
throughout a series-parallel circuit. This prob
lem illustrates such a use o f MATLAB.
For the circuit o f Figure P 2.67
(a)
Find R^^_
(b)
Find Vouf
R^= 180
--R =20
R =90
20A
105
R,= 150
R =40
Figure P2.67
R =80
106
68. Use MATLAB to find R-^ and
for the
circuit o f Figure P2.68. Turn in your MATLAB

code with your answers. Hint: Label the equiv
alent seen at each node to facilitate computa-
on of
K uf
-Cr
200 mA
---
2kO
-- ---------1,2kO
3kO
^3,2kn
(D
IkO
2.2kO
3kO
1.6kn
-
Figure P2.68
A N SW ERS: 591.2 Q, 8.869 V
69. Use MATLAB to find R^^,
and /j for
the circuit o f Figure P2.69.
300n
300
2on
ion
eon
i3on
10V
'4
4on
i5on
2oon >
135Q
> V
500*
-OFigure P2.69
70. Use MATLAB to find Ri in the circuit
o f Figure 2P.70.
3on
2on
8000
100
6000
600
1300
1500
100V
4000 <
400
1350
-o
Figure P2.70
A N SW ER: 50.53 II, 133.8 mA
5000
< 500
C H A P
Nodal and Loop Analyses
H ISTO RICAL NOTE

For a network consisting o f resistors and independent voltage sources, one can apply KCL to the
nodes, KVL to the various loops, and Ohms law to the elements to construct a large set o f simul
taneous equations whose solution yields all currents and voltages in the circuit. In theory, this
approach completely solves the basic analysis problem. In practice, this approach proves imprac
tical because large numbers o f equations are required even for a small network. For example, a 6branch, 4-node network, with each node connected to the other nodes through a single element,
leads to a set o f 12 equations in 12 unknowns: 3 equations from KCL, 3 equations from KVL,
and 6 equations from the element v -i relationships. The 12 unknowns are the 6 branch currents
and 6 branch voltages.
Before the advent o f digital computers, engineers solved simultaneous equations manually, possi
bly with the aid o f a slide rule, or some primitive mechanical calculating machines. Any technique
or trick that reduced the number o f equations was highly treasured. In such an environment.
Maxwells mesh analysis technique (1881) received much acclaim and credit. Through the use o f
a fictitious circulating current, called a mesh current, Maxwell was able to greatly reduce the num
ber o f equations. For the above-mentioned network, the number o f equations drops from 12 to 3
equations in the unknown mesh currents.
An alternate KCL-based technique (now called nodal analysis) appeared in literature as early as
1901. The method did not gain momentum until the late 1940s, because most problems in the
early days o f electrical engineering could be solved efficiendy using mesh equations in conjunc
tion with some network theorems. W ith the invention o f multi-element vacuum tubes having
interelectrode capacitances, some compelling reasons to use the node method appeared; primari
ly, the node method accounts for the presence o f capacitances without introducing more equa
tions, and secondly, those vacuum tubes that behave very much like current sources are more eas
ily accommodated with nodal equations. By the late 1950s, almost all circuit texts presented both
the mesh and node methods.
Since the 1960s, many digital computer software programs (SPICE being the most ubiquitous)
have been developed for the simulation o f electronic circuits that otherwise would defy hand cal
108
Chapter 3 * Nodal and Loop Analyses
culation. These software packages use a node equation method over the mesh equation approach.
One o f several reasons is that a node is easily identifiable, whereas a set o f proper meshes is diffi
cult for a computer to recognize.
For resistive networks driven by current sources, writing node equations is straightforward.
Certain difficulties arise in writing node equations for circuits containing independent and
dependent voltage sources. During the 1970s, a modification o f the conventional node method
by a research group at IBM resulted in the modified nodal analysis (MNA) technique. W ith the
M N A method, the formulation o f network equations, even in the presence o f voltage sources and
all types o f dependent sources, becomes very systematic.
This chapter discusses the writing and solution o f equations to find pertinent voltages and cur
rents for linear resistive networks.
CHAPTER O U TLIN E
1.
2.
3.
4.
5.
6.
7.
8.
Introduction, Review, and Terminology

The Concepts o f Nodal and Loop Analysis
Nodal Analysis I: Grounded Voltage Sources
Nodal Analysis II: Floating Voltage Sources
Loop Analysis
Summary
Terms and Concepts
Problems
CH APTER O BJECTIVES
1.
2.
Describe and illustrate the method o f node analysis for the computation o f node voltages
in a circuit. Knowledge o f the node voltages o f a circuit allows one to compute all the
branch voltages and, thus, with knowledge o f the element values, all the branch currents.
Define the notion o f a mesh or loop current and describe and illustrate the method o f
mesh or, more generally, loop analysis for the computation o f loop currents in a circuit.
Knowledge o f all the loop currents o f a circuit allows one to compute all the branch cur
rents. Thus, in conjunction with the knowledge o f the branch element information, one
can compute all the branch voltages.
3.
Formulate the node analysis and loop analysis equations as matrix equations and use
matrix methods in their solution emphasizing the use o f existing software for the gener
al solution.
4.
Describe and illustrate the modified nodal approach to circuit analysis. This method
underlies the general software algorithms available for computer simulation o f circuits.
Chapter 3 Nodal and Loop Analyses
109
1. IN TRO D U CTIO N , REVIEW, AND TER M IN O LO G Y

Chapter 1 introduced basic circuit elements, Ohms law, and power calculations. Chapter 2 intro
duced the important laws o f circuit theory, K Y L and KCL, and investigated series, parallel, and
series-parallel circuits. Recall from Chapter 2 that a node voltage is the voltage drop from a given
node to a reference node. As a brief review, consider Figure 3.1, which portrays a circuit labeled
with nodes A through D having associated node voltages, V^, Vg, Vq V^, and eight branches,
one for the current source and one for each o f the seven conductances,
... , Gj. (Since this
chapter deals almost exclusively with dc, the uppercase notation for voltages and currents is com
monplace.)
FIG U RE 3.1. Diagram of a circuit with labeled node voltages, V^, Vg, Vq V^,
with respect to the given reference node.
KVL states that every branch voltage is the difference o f the node voltages present at the terminals
o f the branch: for circuits in this text and all pairs o f nodes, j and k, the voltage drop
from
n o d ej to node k, is
at every instant o f time, where VJ- is the voltage at node j with respect to the reference and
is
the voltage at node k with respect to reference. Here, j and k stand for arbitrary indices and could
be any o f the nodes. A, B, C, or D , in Figure 3.1. These statements mean that knowledge o f all
node voltages in conjunction with device information paints a rather complete picture o f the cir
cuits behavior. This chapter develops techniques for a systematic construction o f equations that
characterize a circuits behavior.
One last introductory point: Throughout this chapter and in many subsequent chapters, software
programs such as MATLAB facilitate calculations. Constructing sets o f equations that character
110
ize the voltages and currents in a circuit is often a challenge. Solving such sets o f equations with
out the use o f software tools presents a much greater challenge. Yet facilitated by MATLAB or
equivalent, the calculations reduce to a hit o f the return key. MATLAB and the circuit simulation
program called PSpice or Spice (utilized in Chapter 4) are but two o f the many modern and
important software tools available to engineers.
2. TH E C O N C EPTS OF N O D AL AND LOOP ANALYSIS

Nodal analysis is an organized means for computing ALL node voltages o f a circuit. Nodal analy
sis builds around KCL, i.e., at each node o f the circuit, the sum o f the currents leaving (entering)
the node is zero. Each current in the sum enters or leaves a node through a branch. Each branch
current generally depends on the branch conductance, a subset o f the circuit node voltages, and
possibly source values. After substituting this branch information for each current in a nodes KCL
equation, one obtains a nodal equation.
As an example, the nodal equation at node A in Figure 3.1 is /^-^ = /j + /y = G j (V ^- V^) + Gy (Vj
- V^). The nodal equation at node C is -/2 + I^ + 1^^- Ij =
Vj- + G j ( V ^ + Gy ( K ( j- V^) = 0. Writing such an equation at each circuit node (except the reference node) pro
duces a set o f independent equations. O f course, one can substitute a KCL equation at the refer
ence node for any o f the other equations and still obtain an independent set o f nodal equations.
T he solution o f such a set o f nodal equations yields all circuit node voltages. Knowing all node volt
ages permits us to compute all branch voltages. Knowing each branch voltage and each branch con
ductance allows us to compute each branch current using Ohms law. The reference node may be
chosen arbitrarily and can sometimes be chosen to greatly simplify the analysis.
A set o f nodal equations has a matrix representation. The matrix representation permits easy solu
tion for the node voltages using MATLAB or an equivalent software package. A variation o f the
nodal analysis method, termed modified nodal analysis, relies heavily on matrix methods for
constructing and solving the circuit equations. The basic principles o f this widespread analysis
technique are illustrated in Section 4.
Because computer-based circuit analysis packages build on a matrix formulation o f the circuit
equations and because o f the widespread use o f matrices in circuits, systems, and control, we will
stress a matrix formulation o f equations throughout this chapter. The student unfamiliar with
matrix methods might look through a calculus text or a linear algebra text for a good explanation
o f their basic properties and uses.
T he counterpart to nodal analysis is loop analysis. In loop analysis, the counterpart o f a node
voltage is a loop current, which circulates around a closed path in a circuit. A loop or closed path
in a circuit is a contiguous sequence o f branches that begins and ends on the same node and touch
es no other node more than once. For each loop in the circuit, one defines a loop current, as illus
trated in Figure 3.2, that depicts three loops or closed paths having corresponding loop currents
/p Ij, and ly O f course, one can draw other closed paths or loops for this circuit and define other
loop currents.
111
90
FIG U RE 3.2. Simple resistive circuit showing three closed paths (dotted lines) that represent
three loop currents, /j, Ij, and 1^; the branch current
which is a difference
o f the two loop currents through the resistor.
Using a fluid flow analogy, one can think o f loop currents as fluid circulating through closed sec
tions o f pipe. The fluid in different closed paths may share a segment o f pipe. This segment is anal
ogous to a branch o f a circuit on which two or more loop currents are incident. The net current
in the branch is analogous to the net fluid flow. Note that each branch current can be expressed
as a sum o f loop currents with due regard to direction. For example, in Figure 3.2, the branch cur
rent 7^3 = ^\ -
Using loop currents, element resistance values, and source values, it is possible
by KVL and Ohms law to express the sum o f the voltages around each loop in terms o f the loop
currents. For example, the first loop, labeled
in Figure 3.2, has the loop equation
^ = 9 / i + 3(/i -/ 2 )+ 6 (/ ^ -/ 3 )
We will explore this concept more thoroughly in Section 5. Here we see that loop analysis builds
on KVL, whereas node analysis builds on KCL.
3. N O D AL ANALYSIS I: G RO U N D ED VO LTAG E SO URCES

As mentioned earlier, nodal analysis is a technique for finding all node voltages in a circuit. W ith
knowledge o f all the node voltages and all the element values, one can compute all branch volt
ages and currents, and thus the power absorbed or delivered by each branch. This section describes
nodal analysis for circuits containing dependent and independent current sources, resistances, and
independent voltage sources that are grounded to the reference node (see Figure 3.3). Floating
independent or dependent voltage sources (those not directly connected to the reference node) are
covered in Section 4.
For the class o f circuits discussed in this section, it is possible to write a nodal (KCL) equation at
each node not connected to a voltage source. A node connected to a voltage source grounded to
the reference node has a node voltage equal to the source voltage. The other node voltages must
112
be computed from the set o f nodal equations. Each nodal equation will sum the currents leaving
a node. Each current in the sum will be expressed in terms o f dependent or independent current
sources or branch conductances and node voltages. The set o f these equations will have a solution
that yields all the pertinent node voltages o f the circuit. Examples 3.1 and 3.2 illustrate the basic
techniques o f nodal analysis.
EX A M P L E 3 .1 .
The circuit o f Figure 3.3a contains an independent voltage source, an independent current source,
and five resistances whose conductances in S are G j through Gy The nodes other than the refer
ence are labeled with the node voltages V^,
and V^, which respectively denote nodes a, b, and
c. T he analysis o f this circuit illustrates the process o f nodal analysis to find the node voltages V^,
y,,andK =
FIG U RE 3.3A. Resistive circuit for Example 3.1. Note that node voltage
is specified by the voltage source.
So l u t io n .
Step 1. Consider node c. A voltage source ties node c to the reference node. Hence, the node volt
age
is fixed at V-^, i.e.,
Because
it is not necessary to apply KCL to this node
unless the current through the voltage source is required, for example, when determining the
Step 2. Sum the currents leaving node a. From KCL, the sum o f the currents leaving node a is
zero. As per the partial circuit in Figure 3.3b, this requires that
Grouping the coefficients o f
and
tion yields our first nodal equation
and moving the source values to the right side o f the equa
113
G 5J 'V -aV .in')
(3.1)
Step 3. Sum the currents leaving node b. Applying KCL to node b, reproduced in Figure 3.3c,
yields the equation
G 2 ^y b- y a) ^G ,V b^ G, {V ,- VJ = Q
After regrouping terms, one obtains our second nodal equation:
G4(v ,- v
G 3V,
G .(V ,-V )
FIG U RE 3.3C
-G ^ V ^ + {G ^ + G ,+ G ,)V ,= G,V.
(3.2)
Step 4 . Write set o f nodal equations in matrix form. Equations 3.1 and 3.2 in matrix form are
Gi + G
2+
G5
-G 2
-G 2
G 2 -I- G 3 -I- G 4
(3.3)
Matrix equations organize relevant data into a unified framework. Because many calculators do
matrix arithmetic, because o f the widespread availability o f matrix software packages such as
MATLAB, and because equation solution techniques in circuits, systems, and control heavily uti
lize matrix methods, the matrix equation formulation has widespread and critical importance.
Step 5 . Solve the matrix equation 3.3: For this part, suppose that the conductance values in S are
Gj = 0.2, G2 = 0.2, G3 = 0.3, G 4 = 0.1, G 5 = 0.4, that = 2.8 A, and that
= 24 V. After sub
stitution, equation 3.3 simpUfies to
114
0.8
-0 .2 - K,'
- 0 .2
0 .6
12.4'
(3.4)
2.4
Solving using the inverse matrix method leads to the node voltages (in volts):
0.8
- 0 .2
- 0 .2 '
0.6
-1
T 2 .4 -
2.4
0 .6
0.4 4 0.2
0 .2 ' 12.4'
18-
0.8
10
2.4
Alternately, one could have solved equation 3.4 via MATLAB, its equivalent, or the age-old hand
method o f adding and subtracting equations. For example, in MATLAB
M =[0.8 -0.2;-0.2 0.6];
b= [12.4 2.4]';
>>NodeV = M\b
NodeV =
1.8000e+01
l.OOOOe+01
% O R EQU IVA LEN TLY
NodeV = inv(M )*b
NodeV =
18
10
Step 6. Compute
The branch voltage V^ = V^ -
18 - 10 = 8 V.
Exercises. 1. Utilize the solution o f Example 3.1 to compute the current leaving and the power
delivered by the independent voltage source.
AN SW ER: 3.8 A and 91.2 watts
2. Referring to Figure 3.3a and the values set forth in Step 5 o f Example 3.1, suppose the value o f
is cut in half, the value o f V-^ is 24 V, and the value o f each o f the conductances is also cut in
half W hat are the new values o f the node voltages?
AN SW ER: All node voltages are the same.
3. By what single factor must the values o f
and V-^ be
multiplied so that the node voltages are doubled?

AN SW ER: 2
L
4. Construct a node equation for
A N SW ER: (G, + G ,)V ^ ^ 1 -
in Figure 3.4.
F IG U R E 3.4.
115
EX A M PLE 3.2.
Consider the circuit o f Figure 3.5a. Similar to Example 3.1, the objective is to find the node volt
ages V^,Vf^, and
. However, in the circuit o f Figure 3.5a, an independent current source has
replaced the independent voltage source o f Figure 3.3a. This change unfreezes the constraint on
the value o f
present in the circuit Figure 3.3a. There will result three nodal equations in the
three unknowns
, and V^.
0.4 U
FIG U RE 3.5A. Circuit containing two independent current sources and

three unknown node voltages
, and V^.
So l u t io n .
Step 1. Sum currents leaving node a. This step is the same as Step 2 o f Example 3.1. By inspec
tion o f node a,
0 . 2 + 0.2(K^ - V^) + 0.4(V^ - V J - 2 = 0
which upon regrouping terms yields
0.8V ^ -0.2V ^ -0.4V ^ = 2
(3.5)
Step 2. Sum currents leaving node b. This step is the same as Step 3 o f Example 3.1. Again, by
inspection,
0.2
- VJ + 0.3
+ 0.1 (V^ - V;) = 0
Simplification yields
-0 .2 7 ^ + 0.6V ^ ^ -0.1K ^ = 0
(3.6)
Step 3. Sum currents leaving node c. Because a current source
o. 4 ( v - v :
drives node c, the similarity to example 3.1 ends, and we must

write a third node equation. Summing the currents leaving
node c, as shown in Figure 3.5b, yields
0 . 4 ( 1 / - K J + 0.1 ( K ^ - K ^ ) - 1 = 0
Upon simplification, we have
-0.4V ^^-0.1K ^ + 0 .5 V ;= 1
(3.7)
116
Step 4. Write equations 3 .5 -3 .7 as a matrix equation and solve. The matrix form o f our nodal
equations 3 .5 -3 .7 is
0.8
- 0.2
- 0 . 4 ' v;'
- 0.2
0.6
- 0.1
- 0 .4
- 0 .1
0.5
'T
(3.8a)
= 0
Solving equation 3.8 using MATLAB or equivalent, using a calculator that does matrix operations,
or solving via some form o f row reduction, one obtains the solution (in volts)
0.8
= - 0.2
Vc
- 0 .4
- 0.2
- 0 .4 '
0.6
- 0.1
-1
'2 .9
1.4
2.6
8.4'
- 0.1
0 = 1.4
2.4
1.6
0 = 4 .4
0.5
2.6
1.6
4 .4
(3.8b)
9.6
Specifically, in MATLAB
M = [0.8 -0.2 -0.4;-0.2 0.6 -0.1;
-0.4 - 0.1 0.5];
>>b = [2 0 1]';
NodeV = M\b
NodeV =
8.4000e+00
4.4000e+00
9.6000e+00
Exercises. 1. Suppose the values o f the current sources in Figure 3.5a are doubled. W hat are the
new values o f the node voltages? Hint: Consider the effect on equation 3.8.
AN SW ER: All node voltages are doubled.
2 . Suppose the conductances in the circuit o f Figure 3.5a are cut in half, i.e., the resistances are
doubled. W hat are the new node voltages?
A N SW ER: Node voltages are doubled.
3. Suppose the conductances in the circuit o f Figure 3.5a are cut in half W hat happens to the
magnitudes o f the branch currents? Hint: Express the branch current in terms o f the branch con
ductance and its terminal node voltages.
AN SW ER: The magnitudes o f the branch currents
stay the same.
I
4. Find two node equations characterizing the cir
cuit o f Figure 3.6.
AN SW ER: (G j + G^) -V~G^^Vy =
and
FIG U RE 3.6
117
The matrices in equations 3.3, 3.4, and 3.8a are symmetric. A symmetric matrix, say A, is one
= A; this means that ifA = [a-^ is an n x n matrix whose i-j

entry is a-, then A is symmetric if a-j = a^j^. In words, the off-diagonal entries are mirror images o f
whose transpose equals itself, i.e.,
each other. For example.
0.8
- 0.2
- 0 .4 '
A = - 0.2
0.6
- 0.1
- 0 .4
- 0.1
0.5
present in the circuit., as in Examples 3.1 and 3.2, the coefficient matrix o f the node equations (as exemplified in equations 3.3, 3.4, and 3.8) is always sym
metric, provided the equations are written in the natural order.
When only resistances, independent current sources, and grounded independent voltage sources are
present in the circuit, the value o f the entries in the coefficient matrix o f the nodal equations can be
computed by inspection. The 1-1 entry o f the matrix is the sum o f the conductances at node a (or
1); the 2 - 2 entry is the sum o f the conductances at node b (or 2). In general, the i-i entry o f the
coefficient matrix is the sum o f the conductances incident at node i. Further, the 1 -2 entry o f the
matrix is the negative o f the sum o f the conductances between nodes a and b (or between nodes 1
and 2), and the 2 -1 entry has the same value. In Example 3.2, the 1 -2 entry o f - 0 .2 S is the nega
tive o f the sum o f the conductances between nodes a and b; the 1 -3 entry o f - 0 .4 S is the negative
o f the sum o f the conductances between nodes a and c (or between 1 and 3, if the nodes were so
numbered). Thus, whenever the circuit contains no dependent sources, the node equations can be
written by inspection. Further, if independent voltage sources are absent, then the right-hand side o f
the nodal matrix equation can also be written by inspection: the i-th entry is simply the sum o f the
independent source currents injected into the ith node at which KCL is applied.
W hen controlled sources are present in the circuit, the resultant nodal matrix is generally not sym
metric, as illustrated in the following two examples.
EXA M PLE 3.3.

The circuit o f Figure 3.7 represents a small-signal low-frequency equivalent circuit o f an amplifi
er in which the input signal V-^ is amplified at the output,
= V'2. Small-signal means that
the input signal should have a relatively small magnitude so that a LIN EA R circuit will adequately
represent the amplifier. Similarly, low-frequency means that the frequency o f any sinusoidal input
must be relatively low for the (resistive) circuit model o f the amplifier to remain valid.
The amplifier circuit model contains a current-controlled current source (CC CS) and a voltagecontrolled current source (VCCS). These two dependent sources have currents that depend on
other circuit parameters and require some special handling when constructing node equations.
Our objective is to set forth the methodology for writing the node equations when dependent cur
rent sources are part o f the circuit and to compute the magnitude o f the voltage gain, |
V-^ |
= \y 2 |y^n\
Note that the source voltage, V-^, specifies the voltage at the node at the bottom o f Gp hence, a
nodal equation at this node is unnecessary. Nodal equations must be written at the remaining
118
nodes, which are labeled with the voltages Vj, Kj (=
and V^. (Numbering and labeling is
often a matter o f personal preference. In this example, we have chosen 1, 2, and 3 as node labels,
in contrast to the previous two examples, where we used a, b, and c.)
G,
FIG U RE 3.7. An equivalent circuit model o f an amplifier.
So l u t io n .
Step 1. Sum the currents leaving node 1. Summing the current leaving node 1 leads to
(^1 -
+ ^2 (^1 - ^ 3) + ^3 (^1 -
+ P
or, equivalently, after grouping like terms,

(Gi + G2 + G3)Ki - G3 K2 -
(3.9)
Step 2. Substitute fo r i^ in equation 3 .9 and simplify. In equation 3.9, |3 i^ accounts for the effect
o f the C C C S at node 1 and is not given in terms o f the circuit node voltages. To specify this term
in terms o f the circuit node voltages, observe that in Figure 3.7, i^ is the current from node 2 to
node 3 through G^. Hence,
I3t -I3G^{V2 - K3) = I3G^V^ - I3G^V^
(3.10)
Substituting equation 3.10 into 3.9, again grouping like terms, one obtains the first nodal equation,
{G, + G2 + G3) Vi + il3G^ - G3) K2 - (G 2 + PG^) K3 =
(3.11)
Step 3. Sum the currents leaving node 2. By the usual methods,

G,{V^ - V,) +
+ G,{V^ - K3) +
which, after regrouping terms, reduces to

- G 3 K1 + (G 3 + Gg 4. G,)V^ - G4 K3 +
= 0
( 3 . 12)
119
Step 4, Specify
in terms o f node voltages, substitute into equation 3 . 1 2 , an d simplify.
Inspecting the circuit o f Figure 3.7 shows that
is the voltage across
from node 1 to node 3.
Hence,
gm^. = & j y x - y , ) - g n y x - g r r y ,
Substituting
(3-13)
o f equation 3.13 into equation 3.12 leads to our second nodal equation,
^ 3)^ , + (G 3 + G4 + G ,)V ^ -{G , + ^ J K 3 = 0
(3.14)
Step 5. Sum the currents leaving node 3. Applying KCL to node 3 yields,
G 2 ( V ,- V , ) ^ G , i V , - V , ) ^ G , V , - p i ^ - g ^ v ^ = 0
(3.15a)
Using equations 3.10 and 3.13 for , z and g v respectively, we have

0
= G , { V , - V ,) + G , { V , - V ,) + G ^ V ^ - ^ G , { V , - V , ) - g ^ { V , - V , )
Grouping like terms leads to our third equation in the three unknowns V^, V^, and V^:
-(G i + kJ V
x-
+ ^ 4)^2 + (^2 + G4 + I3G^ + G 5 + gJV ^ = 0
(3.15b)
Step 6. Put nodal equations in matrix form. The three nodal equations 3.11, 3.14, and 3.15b
have the matrix form
Gi + G 2 + G 3
Sm - G
PG^ G 3
~ ^ 2
G 3 + Gg + G 4
^2 8 m
~Â -
-G 4 -
G 2 + G 4 + fiG^ + G 5 + g
y,'
G.V^v/
^2 =
^^3
Step 7. Substitute values and solve. Suppose that the various circuit conductances have the fol
lowing values in [xS: G j = 1,000, G j = 2.0, G j = 1.0, G^ = 10, G j = 2 0 ,100, and Gg = 200 .
Suppose further that
= 2.1 V, (3 = 4/1010 and^^ = 21,112 [xS. This allows us to generate the
following M ATLAB code for the solution:

G 1 = 1000e-6;G 2 = 2e-6; G 3 = le - 6 ; G 4 = lOe-6 ;
G5 = 20100e-6; G 6 = 200e-6; Vin = 2 . 1;
beta = 4/1010; gm = 2 1 1 12 e-6 ;
M =[G1+G 2+G 3
gm-G3
-G2-gm
beta*G 4-G 3
G 3+G 6+G 4
-G 4-beta*G 4
b = [G l* V in
NodeV = M\b
NodeV =
2 . 0000 e +00
-l.OOOOe+02
l.OOOOe+00
0]';
-G 2-beta*G 4;
-G4-gm ;
G2+G4+beta*G4+G5+gm ];
120
in which case,
V,-
V^2 = - 1 0 0
V3
Step 8. Compute the voltage gain. The voltage gain o f the amplifier is given by
Kut
V2
-1 0 0
Vin
Vin
2.1
= 4 7 .6 2
Exercises. 1. Suppose V-^ in the circuit o f Figure 3 .7 is doubled. W hat are the new node voltages?
Hint: Consider the matrix equation o f Step 6.
A N SW ER: Node voltages are doubled.
2. Suppose all conductances in the circuit o f Figure 3.7 are cut in half (resistances are doubled)
and (3 is held constant. How must^^ change for the node voltages to remain at their same values?
AN SW ER:
must double.
Realistic problems do not permit hand solutions. For hand solutions, the smallest number o f
equations is generally desired. For matrix solutions using software packages such as MATLAB,
more variables with more equations may often be easier to construct and may often result in more
reliable numerical calculations. This can be illustrated using the equations o f Example 3.3. All the
pertinent basic equations o f the circuit o f Figure 3 .7 can be written down as follows: from equa
tions 3.9 and 3.10 we have
and
However, in contrast to the example, we do not substitute 3.10 into 3.9 to obtain 3.11. Rather,
we just let them be two independent equations. Further, from equations 3.12, 3.13, and 3.15a,
we have
- G 3 K, + (G 3 + G4 + G ,) K2 - G 4 K3 + g^v^ = 0
and
By not substituting for and v^, we avoid unnecessary hand calculation, and if there is an error,
it is easier to find. The resulting equations have the matrix form where i and
now appear as
additional unknowns, easily handled by a software program:
121
G] 4- G 2 + G 3
-G 3
-G 2
13
G4
-G 4
-I
-G 3
G 3 -1- G 4 -HGg
-G 4
8m
-1
-1
-G 2
-G 4
G2 + G 4 + G5
-/?
~Sm
V,-
^2
=
iy
0
0
0
As a general rule, we would reorder the equations so that rows 1, 3, and 5 came first, as they cor
respond to the three nodal equations at Kj, V2 , and
equations for i and
. Then we would write the constraint

. Such a reordering leads to certain symmetry properties discussed earlier.
Exercise. Solve the above matrix equation in MATLAB or equivalent, using the numbers o f
A and - 1 V.
Example 3.3 to verify that = -1 .0 1 x 10^-3 A
Matrix methods as used in the above examples and in the ones to follow necessitate the power o f
a calculator or a software program such as MATLAB for easy solution. Such programs permit a
straightforward calculation o f the required answers and are not prone to arithmetic errors.
The next example illustrates how to write node equations for circuits containing a voltage-con
trolled voltage source (VCVS) grounded to the reference node. The analysis o f CCVSs grounded
to the reference node is similar. The more challenging analysis o f circuits containing floating
dependent or independent voltage sources is taken up in the next section.
EX A M P L E 3 .4 .
The circuit o f Figure 3.8 models a poor operational amplifier circuit' in which the output voltage
1/^^ = V2 approximates
For the analysis, let |i = 70. The adjective poor arises because ^
should have a value much larger than 70.

R3= lO k O
V. =
F IG U R E 3.8. A two-node (amplifier) circuit containing a grounded VCVS with jx = 70.
So l u t io n .
The circuit contains two nodes labeled
and
(equivalently nodes 1 and 2) not constrained
by voltage sources. The goal o f our analysis is to find these node voltages by writing two equations
in these voltages and solving. As is commonly the case, resistances are in ohms and will be con
verted to conductances in S for convenience in writing the node equations.
/~N,
122
Step 1. Compute conductance values in S. Conductances are the reciprocal o f resistances, i.e., G-
= ÎRj- Hence,
= 3.3 3 3 3 3 lQ-5, G^ = 10^, G^ = 0.01, and G^ = lO'^
G j = 2.0 10-3,
Step 2 . Write a node equation at node 1. Summing the currents leaving node 1 yields
Grouping like terms leads to
{G ,^ G ^ ^ G ,)V ,-G ,V ^ = G,V^n

Inserting numerical quantities yields the first node equation
(3.16)
Step 3. Sum currents leaving node
. Summing the currents leaving node 2 yields
^3 (^^2 - ^l) + ^5 ^2 + G4 {I /2 +
Kj) = 0
The dimensionless coefficient |i is placed with the conductance,

obtain
(HG4 -
while grouping like terms to
+ (G 3 + G4+ ^ 5)^2 = 0
Inserting the numerical values produces the second node equation

0.699 9 Vj + 0.0111 ^2 = 0
(3.17)
Step 4. Write equations 3 .1 6 and 3 .1 7 in matrix form and solve. In matrix form
3.3333
-1
-V f
0 .6 9 9 9
O .O Ill
.^2.
2'
0
Using the formula for the inverse o f a 2 x 2 matrix (interchange the diagonal entries, change the
sign on the o ff diagonals, and divide by the determinant), one obtains
Vf
0.0111
0 .0 3 0 1 2 6
.^2.
0.7369
-0 .6 9 9 9
3.3333
-1 .8 9 9 6
In Example 3.4, observe that
^^ 2
= -1 -8 9 9 6 , which approximates - 2 V-^ since V- = 1 V.
Exercises. 1. Write MATLAB code to solve the above example. Check that your code works. Hint:
See Example 3.3.
2. If R2 is changed to 100 k2 in Example 3.4, show that V2 = -1 .9 0 6 3 V.
123
4. N O D AL AN ALYSIS II: FLOATING VO LTAG E SO URCES

A floating voltage source means that neither node o f the source is connected to the reference
node. When a floating dependent or independent voltage source is present with respect to a given
reference node, a direct application o f KCL to either terminal node o f the voltage source is unfruit
ful. There are several ways to handle this situation. One fruitful method is to enclose the source
and its terminal nodes by a Gaussian surface, i.e., a closed curve, to create what is commonly
called a supernode, as illustrated in Figure 3.9. One would then write KCL for the supernode as
is done in a number o f circuit texts. However, there is a conceptually more straightforward
approach, which is often called the modified nodal analysis, or MNA. In MNA, we add an addi
tional current label to each floating voltage source. In Figure 3.9, we have added the current label
l^y. This additional current becomes an unknown in a set o f nodal equations generated by apply
ing KCL to each node. At this point, further explanation is best done by an example, but the con
cept is similar to the discussion following Example 3.3.
E X A M P L E 3.5.
Find the node voltages V^, Vy,
and the unknown current
in the circuit o f Figure 3.9, when
the bottom node is taken as reference.
FIG U RE 3.9. Resistive circuit containing a floating voltage source for the given reference;
generally, the reference node may be chosen arbitrarily.
So l u t io n .
Step 1. Write a node equation at node a. Summing the currents leaving node a yields
8 + 0.15
+ 3 + 0.2 ( K ^ -K .) = 0
After grouping terms appropriately, we have

(0.15 + 0 .2 )K ^ - 0.15^ -^ - 0 .2 V ;+ 8 + 3 = 0
or, equivalently,
0.351^ ^ -0.15K ^ -0.2V ^ ^ = - 1 1
This provides our first equation in four unknowns.
(3.18)
124
Step 2. Write a nodal equation at node b. Here,

=0
- 3 + 0 .1 5 ( K ^ - K ; + 0.05
or equivalently,
- 0 .1 5
(0 .1 5 + 0.05) n - / ^ ^ = 3
Simplifying this expression leads to

- 0 .1 5 K , + 0.2K^-/^^ = 3
(3.19)
Step 3. Write a nodal equation at node c. Here,

^.^ + 0-25 V ^-25 +
{V^- K J = 0
or equivalently.
- 0 .2 K ,+ 0.45K^ + /,, = 25
(3.20)
Step 4. Write the node voltage relationship fo r the terminal nodes o f the floating voltage source,
i.e., between the voltages
and V^. The voltages
Mathematically, this constraint is 1/ -
and
are constrained by the voltage source.
= 440, i.e.,
K ^ = 440
(3.21)
Step 5. Write thefour equations 3.18, 3.19, 3.20, and 3.21 in matrixform and solve. In matrix form,
0.35
- 0 .1 5
- 0 .2
-1 1
- 0 .1 5
0.2
-1
- 0 .2
0.45
25
-1
hb
(3.22)
440
Because o f the extra variable, the equations become too large for hand calculation. Hence, we use
MATLAB as follows:
>>M = [0.35 - 0 .1 5 - 0 .2 0 ;-0 .1 5 0.2 0 - 1 ;
- 0 .2 0 0.45 1 ; 0 - 1 1 0];
b = [-11 3 25 440]';
x = M\b
X =
-9.0000e+ 01
-3 .1 0 0 0 e + 0 2
1.3000e+02
-5.1500 e+ 01
Hence,
= - 90 V,
= - 3 1 0 V,
= 130 V,
= -5 1 .5 A
12 5
In a conventional nodal analysis, all unknowns are node voltages. Here we have the additional
unknown current,
. Because o f this additional unknown current, the method is called a mod
ified nodal analysis.

Also, in this example, node d was taken as the reference node. However, one could just as easily
take node b as the reference node, in which case, the voltage source would not have been floating.
A home problem investigates this choice o f reference node.
Exercise. 1. For Example 3.5, compute the voltages
and
2. For Example 3.5, compute the power absorbed by the 0.15 S resistor.
3. Compute the power delivered by the floating voltage source.
AN SW ERS in random order: 220 V, 22.6 6 kw, 310 V, 440 V, 7260 watts
The next example investigates a circuit having floating independent and dependent voltage
sources. By convention, the reference node o f this circuit, figure 3.10, and all subsequent circuits,
will be the bottom node o f the circuit unless stated otherwise.
E X A M P L E 3.6.
The circuit o f Figure 3.10 contains a floating independent and a floating dependent voltage
source. Find the node voltages V^, Vy, V^, and the unknown currents
and
Then find the
power delivered by the 30 V source and the dependent source.

500
FIG U RE 3.10. Resistive circuit containing a floating dependent voltage source and a floating
independent voltage when node d is chosen as the reference node.
So l u t io n .
Step 1. Sum currents leaving node a. Here,
126
Equivalently,
100
(3.23)
100
Step 2. Sum currents leaving node b. Here,
100
500
Equivalently,
(3.24)
Step 3. Sum currents leaving node c. Here,
1
(3.25)
800
Step 4. Write an equation relating the terminal voltages o f the independent voltage source. Here,
K ,-n = 3 0
(3.26)
Step 5. Write an equation relating the terminal voltages o f the dependent voltage source. Here,
40
K ,- K = 4 0 ., =
Equivalently,
(3.27)
0.6V;, + 0 .4 y ^ - V ^ = 0
Step 6. Write equations 3.23 through 3 .2 7 in matrix form an d solve in MATLAB. Combining
the above equations into a matrix produces
0 .03
-
0.01
-0 .0 1
0.012
-1
0.0 0 1 2 5
-1
âc
30
Icb,
-1
0 .6
0.4
________
-1
2.2'
0
V'i
Again, this matrix equation is too large for hand computation. Hence in MATLAB,
M = [0.03 - 0 .0 1 0 1 0;
-
0.01 0 .012 0 0
0 0 0.00125 - 1 1;
0 - 1 1 0 0;
0.6 0 . 4 - 1 0 0];
1;
127
>.b= [2.2 0 0 30 0]';

x = M\b
X =
l.OOOOe+02
5.0000e+01
8.0000e+01
-3 .0 0 0 0 e -0 1
-4 .0 0 0 0 e -0 1
Hence,
100
50
=
80
âc
- 0 .3
Icb_
- 0 .4
Step 4. Compute the power delivered by the 3 0 Vsource. The power delivered by the 30-V source is
^^./ = - 3 0 / ,^ = 3 0 x 0.4 = 1 2 W
Step 5. Compute the power delivered by the dependent source. The power delivered by the depend
ent source is
Pdel =
- 4 0
^ ( - 0 . 3 ) = 0.12(V^ - V^) = 6 W
100
Exercises. 1. For Example 3.6, compute the voltages
and the power absorbed by the 800
Q resistor.
A N SW ERS in random order: 8 watts, 20 V, - 8 0 V
2. Suppose the two independent voltage source values in Example 3.6 are doubled. W hat are the
new node voltages? W hat are the new branch currents?
A N SW ERS: Node voltages are doubled and branch currents are doubled.
3. Suppose all resistances in the circuit o f Figure 3.10 are doubled and the value o f the parameter
on the dependent source is also doubled. W hat are the new branch currents?
AN SW ER: All branch currents are cut in half.
The above example increases the number o f unknowns beyond the node voltages to include the
two currents through the floating voltage sources. However, we could have included additional
currents to the set o f equations making the dimension even higher. W ith a tool like MATLAB,
this poses no difficulty. However, it does make hand computation a challenge. For example, we
128
= l^y as an additional variable with a corresponding increase in the num
could have included
ber o f equations. By adding addirional unknowns we would simplify the writing o f the individual
node equations but increase the dimension o f the matrix equation. Specifically, the node equation
at a becomes
_
t i
50
IV
50
and the resulting larger matrix equation is

0.02
2.2
0.002
-1
0.00 1 2 5
-1
-1
-1
-4 0
-1
-100
0
âc
30
Ic h
0
0
This completes our discussion o f the standard nodal equation method o f circuit analysis. T he next
section takes up a discussion o f an alternative analysis method entided loop analysis.
5. LOOP ANALYSIS
Loop analysis is a second general analysis technique for computing the voltages and currents in a
circuit. Mesh analysis is a special type o f loop analysis for planar circuits, i.e., circuits that can
be drawn on a plane without branch crossings. For planar circuits, loops can be chosen as mesh
es, as illustrated in Figure 3.2, or as in 3.11 below. Associated with each loop is a loop current.
Loop currents circulate around closed paths (loops) in the circuit. Similarly, for planar circuits,
the term mesh current is used traditionally for loop current. By KVL, the sum o f the voltages
across each branch in a loop is zero. By expressing each o f these branch voltages in terms o f the
designated loop currents, one can write an equation in the loop currents for each designated loop
in the circuit. For branches that are often common to two or more designated loops, the branch
current equals the net flow o f the loop currents incident on the branch. Writing an equation for
each loop produces a set o f equations called loop equations. If sufficient independent loops are
defined, one can solve the loop equations for the loop currents. Once the loop currents are known,
we can easily compute the branch currents and then the branch voltages in the circuit. Then we
can compute any other quantities o f interest, such as power absorbed, power delivered, voltage
gain, etc.
EXA M PLE 3.7.
Consider the planar circuit o f Figure 3.11 with the three specified loops, which are also called
meshes. Denote the loop currents for each loop by /j, 1^, and /j. The objective is to write three
equations in the currents /j,
and /j using KVL and solve these equations for their values. Then
we will compute the power absorbed by the 2 -Q resistor marked with the voltage v. Suppose the
source voltages are
= 4 0 V andV^2 = 20 V.
129
FIG U RE 3.11. Resistive circuit containing only independent voltage sources

for the loop analysis of Example 3.7.
So l u t io n .
Step 1. Write a KVL equation based on loop 1 by summing voltages around this loop. Summing
the voltages around loop 1 using Ohms law and the defined loop currents produces
K., = 40 = /i + 4 (/; - /j) + Va + ih - ^3) = 6 /, - 4 /2 - 7 3 + 20
(3.25a)
Here, observe that the 4 Q resistor is incident on two loops; the net current flowing from top to
bottom, i.e., with respect to the direction o f loop 1, is /j - l 2 - The idea is analogous to a pair o f
distinct water pipes that share a common length. The common length is analogous to the 4 -Q
resistor. The flow rate in each pipe is analogous to the currents /] and I 2 , which in fact, are rates
at which charge flows past a cross sectional area o f the conductor. It follows that the net flow
through the common length o f pipe with respect to the direction o f loop 1 is the difference in the
net flow rates o f pipes 1 and 2, respectively. This is precisely the meaning o f /j - /2 . A similar
explanation can be made for the 1 -Q resistor common to loops 1 and 3 for which the net flow
rate with respect to the direction o f loop 1 is 7j - ly
Simplifying equation 3.25a yields
6/1 - AI^ - 73 = 20
(3.25b)
Step 2. Write a KVL equation based on loop 2 by summing the voltages around this loop. Applying
Ohms law and KVL to loop 2 produces
0 = 4(72 - 7,) + 272 + 2(^2 -
(^.26)
h'
Notice that with respect to the direction of loop 2, the net flow rate through the 4 Q resistor is
Step 3. Finally, write a KVL equation based on bop 3. Stmiming the voltages around loop 3 yields
V^2 = 20 = 2(73 - 72) + 73 + (73 - 7,) = - 7i - 272 + 473
(3.27)
130
Step 4 . Write eqtiations 3.25b, 3.26, an d 3 .2 7 in matrix form an d solve. Writing the above three
loop equations in matrix form yields
'6
-4
-r
-4
-2
-1
-2
20'
(3.28)
20
Solving Equation 3.28 by the matrix inverse method (by a numerical algorithm or by Cramers
rule) yields the loop currents in amps as
h'
-4
-r
-1
0 .35
'2 0 '
= -4
-2
-1
-2
20
= 0 .2 2 5
0 .225
0 .2
20'
0 .2 8 7 5
0 .2
0 .2
0 .4
20
0 .2
11
=
8.5
12
Step 5. Compute the power consumed by the 2 Q resistor. Knowledge o f the loop currents makes
it possible to compute all voltages and currents in the circuit. For our purpose, the voltage
and the power absorbed by the 2 Q resistor is
/ 2 = 24.5 watts.
Exercises. All exercises are for the circuit o f Figure 3.11.

1. Compute the power delivered by the 20 V source.
AN SW ER: 20 watts
2. Compute the power absorbed by the 4 Q resistor.
AN SW ER: 25 watts
3. Suppose the source values are doubled. W hat are the new values o f the loop currents?
AN SW ER: loop currents are doubled
4. Suppose the resistance values are multiplied by 4. W hat are the new loop currents? W hat are
the new node voltages?
AN SW ERS: Loop currents are 0.25 times their original values, and node voltages are unchanged.
Observe that there are no dependent current or voltage sources in the circuit. Similar to the nodal
analysis case, whenever dependent sources are absent and the equations are written in the natural
order, the loop (or mesh) equations are symmetric, as illustrated by the coefficient matrix o f equa
tion 3.28 where, for example, the 1 -2 and 2 -1 entries coincide, as do the 1 -3 and 3 -1 entries,
etc. Also, the value o f all entries can be computed by inspection. The 1-1 entry o f the matrix is
the sum o f the resistances in loop 1; the 2 - 2 entry is the sum o f the resistances in loop 2, etc. In
general, the i- i entry is the sum o f the resistances in loop i. T he 1 -2 entry o f the matrix is 'L{R^
(the large sigma means summation), where each
is a resistance common to both loops 1 and 2.
Use the + sign when both loop currents circulate through
in the same direction, and use the -
131
sign otherwise. Further, if independent current sources are absent, then the right-hand side o f the
loop equations can also be written by inspection. The i-th entry is simply the net voltage o f the
sources in the ith loop that tends to deliver a current in the direction o f the loop current.
Exercises. 1. Use the inspection rules described above to write two mesh equations for the circuit
o f Figure 3.12, when both mesh currents are assigned clockwise direction.
FIG U RE 3.12.
2. Use the inspection rules described above to write two mesh equations for the circuit o f Figure
3.12, when the left mesh current is clockwise and the right mesh current is counterclockwise.
3. Use the inspection rules described above to determine the right-hand side o f the mesh equation
for the circuit o f Figure 3.13.
AN SW ERS: 8, 0, 10
A simplifying reduction to the set o f loop equations occurs if an independent current source coin
cides with a single loop current. The analysis becomes simpler because that loop current is no
longer an unknown; rather it is equal to the value o f the source current if their directions coin
cide, or to the negative value if their directions are opposing. Because the associated loop current
is known, there are fewer loop equations to write and solve. One would apply KVL to such a loop
only if it were necessary to compute the voltage across the independent current source, which
might be necessary for determining the power delivered by the source. Th e entire situation is anal
ogous to an independent voltage source tied between a node and the reference in nodal analysis.
The following example illustrates the details o f this discussion.
132
EXA M PLE 3.8.

The circuit o f Figure 3.1 4 is a modification o f the one o f 3.11 in which (i) a 1 ohm resistor on
the perimeter o f the circuit is replaced by an 8 A independent current source, and (ii) the values
o f the voltage sources are doubled. The currents for each loop are again denoted by /j, I 2 , and ly
Our objective is to find all the loop currents, the voltage V^, and the power delivered by the 8 A
source.
FIG U RE 3.14. A resistive circuit containing an independent current source on the perimeter
o f loop 3 forcing /^ = 8 A.
So l u t io n .
Step 1. So/ve fo r
by inspection. Because
is the only loop current circulating through the
branch containing the independent 8 A current source, /j = 8 A. This phenomena is similar to the
fact that in nodal analysis, the node voltage o f a grounded voltage source is fixed at the voltage
source value.
Step 2. Write a KVL equation fo r loop 1 by summing voltages around this loop. Summing the
voltages around loop 1 using Ohms law and the designated loop currents produces
28 = / j + 4 ( /j - /2) + 12 + ( /j - 8) = 6 /, - 4 /2 + 4
Hence,
6 /, - 4 /2 = 24
(3.29)
Step 3. Write a KVL equation fo r loop 2 by summing the voltages around this loop. Applying KVL
and Ohms law to loop 2 produces
0 = 4 (/2- /i) + 2 /2 + 2 (/2 - 8 ) = - 4 /, + 8 /2 - 16
Equivalently,
- 4 /1 - 8 /2 = 16
(3.30)
133
Step 4 . Write above loop equations in matrix form and solve. The matrix form o f equations 3.29
and 3.30 is
' 6 -A
'i r
24
-4
l2
16
Using the inverse matrix technique to compute the solution, we have

II"
6 -4
I2
-4
-1
24 '
16
32
' 24'
4 6 16
'8
's'
6
Step 5. Compute V^. By KVL,

v; = 2 (/2 - 8) + 12 + (/j - 8) = 8 V
Step 6. Compute power delivered by 8 A source. Observe that the 8 A current source is labeled
according to the passive sign convention, in which case,
Pdel = -
= - (8
8) = - 64 watts
Hence, the source actually absorbs 64 watts.
Exercise. In the circuit o f Figure 3.15, nvo o f the three mesh currents coincide with independent
source currents. By writing and solving just one mesh equation, find /j.
A N SW ER: 3 A
Not only do independent current sources constrain loop currents, but dependent currents sources
do also. This situation is illustrated in Example 3.9.
E X A M PLE 3 .9 .
This example illustrates the writing o f loop equations for a simplified small signal equivalent cir
cuit, Figure 3.16, o f a two-stage amplifier that contains a current-controlled current source
(C C C S) and a current-controlled voltage source (CC VS). This process extends the techniques o f
Examples 3.7 and 3.8 to find some important characteristics o f the amplifier. Specifically, find
(a) the input resistance seen by the source, i.e.,
= v-Jij^,
134
(b) the voltage gain, v
j and
(c) the voltage v across the dependent current source.
FIG U RE 3.16. Small signal equivalent circuit for a two-stage amplifier. Signals in amplifiers are
usually time dependent, so we adopt the lowercase notation for voltages and currents.
So l u t io n .
The circuit o f Figure 3.16 contains three loop or mesh currents. The direction o f the loops is a
user-chosen preference. For convenience, we have chosen mesh current z'2 to be consistent with
the direction o f the arrow in the dependent current source. Because this dependent current source
lies on the perimeter o f the circuit, it constrains the value o f /'2> i-e-) ?2 ^ P ^b-
the control
ling current,
, z'2 = . ih ~ ^ h'
relationship implies that the mesh current o f loop 2
depends directly on the mesh current o f loop 1. This observation allows us to skip constructing a
mesh equation for loop 2. Only equations for loops 1 and 3 are needed, thereby reducing the
number o f simultaneous equations from three (because there are three loops) to two.
Step 1. Apply KVL to loop/mesh 1. Here, by KVL and the observation that
în = h h +
('1 + P ' 1) =
+ (1 + p)
12
= P zp
i\
From this equation, we can immediately compute the input resistance

V
in
in
(3.31)
Step 2. Apply KVL to loop/mesh 3. In this case, observe that z^ = (z^ + z'g) = - (|3 z'j + Zj). By BCVL,
'3 -
K + K ('3 + ' 2) =
h +
(P h + ' 3) + K ('3 +P ' 1) = 0
Combining like terms, it follows that
K +
ph +
'3 = 0
(3.32)
135
Step 3. Write equations 3.31 and 3.32 in matrix form an d solve. The matrix form o f these equa
tions
IS
/?^ + (P + l ) ^ ,
în
'3
(3.33)
V.
Because the solution is desired in terms o f the hteral variables, we solve equation 3.33 using
Cramers rule, which utilizes determinants. In this task, first define
Using the notation A for the determinant, Cramers rule provides the solution for i^ according to
the formula
det
în
R^ + R ^+r
(3.34)
I =
Applying Cramers rule for the solution o f Zj, yields
det
Step 4. Compute
/?^ + (|3 + l)7 ? e
(3.35)
in terms o f v-^ and then the voltage gain vjv^^. As per the circuit o f Figure
3.11 and equations 3.34 and 3.35,
Vo =
V ; = r
ic = 'w(PM + '3) = r,i
After substituting for A, the voltage gain is
---V.
in
P/?
I d
r= y
Step 4 . Compute v. To compute the voltage across the dependent current source, apply KVL to
mesh 2 to obtain
n
= R^ (z'2 + /'i) + R, (2 + ^3) = [P
h +
136
Exercise. Find a simplified loop equation for
in the circuit o f Figure 3.17.
FIG U RE 3.17.
c' '1
To see the importance o f the calculations o f the amplifier circuit o f Example 3.9, suppose two
amplifiers are available for use with a non-ideal voltage source. The non-ideal voltage source is
modeled by an ideal one-volt source in series with a 100 Q source resistance. Suppose amplifier 1
has a voltage gain, vjv^^ = 1 0 and
= 100 k2. Suppose amplifier 2 has a voltage gain o f 100
and R -^ 2 = 5 3. If amplifier 1 is attached to the non-ideal source, then by voltage division,
100,000/(100,000 + 100) = 0.999 V, whereas in the case o f amplifier 2, v -^ 2 = 5/(100 + 5) =

0 .0 4 7 6 V. In the first case, the gain from
to
is 10, yielding
= 9.99 V. In the second case,
the same gain is 100, yielding v^ 2 =
V. One concludes that amplifier 1 is better suited to this
particular application, although it has a lower voltage gain than amplifier 2. Hence, Example 3.9
illustrates the need to know both the voltage gain and the input resistance to determine the out
put voltage in practical applications. Further, using the literal solution to the example allows us to
apply the formulas to different sets o f parameter values without repeating the complete analysis.
In the previous two examples, there were current sources on the perimeter o f the circuit. Such cur
rent sources were incident to only one loop. It often happens that independent and dependent
current sources can be common to two or more loops. When this happens, a situation analogous
to floating voltage sources in nodal analysis occurs. To handle such cases, many texts define some
thing called a supermesh and write a special loop equation for this supermesh. Supermeshes often
confuse the beginner. There is an easier way.
Example 3.10 below illustrates how to write loop equations when current sources are common
to two or more loops. In such cases, we introduce auxiliary voltage variables across current
sources common to two or more loops. The resulting set o f simultaneous equations will contain
not only the loop currents as unknowns, but also the auxiliary voltages as unknowns. Because the
resulting set o f equations contains both loop currents and additional (auxiliary) voltage variables,
the equations are called modified loop equations. T he process o f writing modified loop equa
tions is extremely systematic and straightforward. Further, it allows us to avoid explaining the very
confusing concept o f a supermesh. On the other hand, the presence o f auxiliary voltage variables
increases the number o f unknowns, i.e., the number o f simultaneous equations increases.
Because o f the availability o f software packages such as MATLAB, M ATH EM ATICA, and
MAPLE, this increased dimension is not a hindrance.
137
E X A M P L E 3 .1 0 .
Consider the circuit o f Figure 3-18 in which
= 28 V and 7^2 = 0 .06 A. Note that the inde
pendent current source is common to loops 1 and 3 and a voltage-controlled current source is
common to loops 1 and 2. Find values for the loop currents 7p Ij, l y and the power delivered by
each independent source.
200Q
FIG U RE 3.18. Circuit containing a current source between loops.
So l u t io n .
To begin the solution, we introduce two auxiliary voltage variables Vj and Vj associated with the
current sources common to two (or multiple) loops. The purpose o f these variables is to facilitate
the application o f KVL for constructing the loop equations. This will require that we obtain three
KVL equations, one for each loop, and two constraint equations, one for each current source.
Step 1. Apply KVL to loop 1. By a clear-cut applicanon o f KVL,
28 = 2007] - V] - V2
(3.36)
Step 2. Apply KVL and Ohms law to loop 2. Again applying KVL and Ohms law to loop 2, we
obtain 100 7j + 200 (Jr^ 7^) + V2 = 0. After grouping like terms.
3 0 0 7 , - 20 0 /3 + V2 = 0
(3.37)
Step 3. Apply KVL to loop 3. Applying KVL to loop 3 yields 150 7^ + Vj + 200 (/j - Tj) = 0.
Equivalendy,
- 200/2 + 350/3 + V| = 0
Step 4. Write a constraint equation determined by the independent current source. Here, loops 1
and 3 are incident on the independent current source so that
0 .0 6 = / , -/ 3
(3.39)
138
Step 5. Write a constraint equation determined by the dependent current source. In a straightfor
ward manner, we have
/ , - / , = 0 .0 2 V, = 0 .02 r2 0 0 (/3 -
12
) = 4 7 , - 47,
After simplification,
/i + 3/2 - 4 /3 - 0
(3.40)
Step 6. Write equations 3 .3 6 to 3.40 in matrix form an d solve. The matrix form o f these equations is
0
-1
-r
300
-2 0 0
-2 0 0
350
-1
''I
0 .0 6
-4
V2
28
0
=
(3.41)
Solving equation 3.41 by the matrix inverse method or by an available software package yields the
solution (currents in A and voltages in V) given by equation 3 .42 below:
200
-1
300
-2 0 0
-2 0 0
350
-1
-4
Li
,''2.
-r -1 28
0 300 -200 0 1
0 -200 350 1 0
1 0 -1 0 0
1 3 -4 0 0];
b = [28 0 0 0.06 0 ];
LoopIplus= M\b
Looplplus =
l.OOOOe-01
2,0000e-02
4,0000e-02
-l.OOOOe+01
2.0000e+00
0.1
0 .02
= 0 .0 4
0 .06
-To
which can be obtained using the following MATLAB code:

M = [ 2 0 0 0 0 -1 -1
(3.42)
139
Step 7. Compute the powers delivered by the independent sources. First, the power delivered by the
independent voltage source is
Py-source = 28 /j = 2.8 watts

The power delivered by the independent current source is
Pl-source =
V^ = - 0 .6 watts
This last value indicates that the independent current source actually absorbs power from the cir
cuit.
Exercise. For the circuit o f Figure 3.19, write the modified loop equations having two unknowns
/j and V, following the procedure described in Example 3.10. Solve the equations and find the
power absorbed by the 2 -Q resistor.
4A
AN SW ER: 18 watts
One final point before closing our discussion o f loop analysis. Loops can be chosen in different
ways. Cleverly choosing loops can sometimes simplify the solution o f the associated equations. For
example, by choosing a loop that passes through a current source so that no other loop is com
mon to the source, the loop current is automatically specified by that current source.
6. SUM M ARY
This chapter introduced the technique o f nodal analysis. Nodal analysis is a technique for writing a set
o f equations whose solution yields all node voltages in a circuit. With knowledge o f all the node volt
ages and all the element values, one can compute all branch voltages and currents. As mentioned, when
ever there are no dependent sources present, the coefficient matrix o f the node equations is always sym
metric. Hence, whenever dependent sources are absent, it is possible to write the nodal equation coef
ficient matrix by inspection. Further, if independent voltage sources are absent, then the right-hand side
o f the matrix form o f the nodal equations can also be written by inspection: the entry is simply the sum
o f the independent source currents injected into the node at which KCL is applied. When VCCSs are
present, the steps for writing nodal equations are the same as illustrated in Example 3.3. Generally, in
such cases, the resultant coefficient matrix is not symmetric.
140
W hen floating dependent or independent voltage sources are present with respect to a given ref
erence node, we introduce new current variable through these floating sources as unknowns. The
node equations then incorporate these additional unknown currents, as was illustrated in
Examples 3.5 and 3.6. This method increases the number o f equations but simplifies the con
struction o f the individual equations. W ith a tool like MATLAB to compute solutions, there is no
difficulty, although hand computation may become more difficult. This concept is the basis o f the
modified nodal analysis method used in circuit simulation programs like SPICE.
Loop/mesh analysis, an approach dual to nodal analysis, was introduced in Section 5. Mesh analy
sis is a special case o f loop analysis for planar circuits when the loops are chosen to be the obvious
meshes, similar in geometry to a fish net. In loop analysis, one sums the voltages around a loop or
mesh to zero. Each o f the branch voltages in the loop is expressed as a product o f resistances and
(fictitious) loop currents that circulate through the branch resistance, as illustrated in Figures 3.10,
3.14, and 3.16. The branch current o f the circuit are equal to the net flow o f the loop currents
incident on a particular branch, meaning that each branch current is expressible as a sum o f loop
currents. The desired set o f loop equations is produced by summing the voltages around each loop,
expressing these voltages either as source values or as resistances times loop currents. One solves
the loop equations for the loop currents. Once the loop currents are known, we can then compute
the individual branch currents and then the branch voltages, and thus any other pertinent current,
voltage, or power. Whenever there are no dependent sources present, the coefficient matrix o f the
loop equations is always symmetric. Whenever dependent sources are absent, it is possible to eas
ily write the loop matrix by inspection.
As the size o f an arbitrary circuit grows larger, there are two good reasons for choosing the nodal
method over the loop method; (i) the number o f nodal equations is usually smaller than the num
ber o f loop equations, and (ii) the formulation o f nodal equations for computer solution is easier
than methods based on loop equations. Writing nodal equations is particularly easy if the circuit
contains only resistances, independent current sources, and VCCSs for short, an R
I
g net
work. For an
network, one simply applies KVL to every node (except the reference node)
and obtains a set o f node equations directly. For floating independent or dependent voltage
sources, the task is more complex. Examples 3.5 and 3.6 illustrate cases where, besides the node
voltages, additional unknown auxiliary currents are added. By adding additional auxiliary variables
to the formulation o f the nodal equations, we described the concept behind the modified nodal
analysis (MNA) method. The MNA method retains the simplicity o f the nodal method while
removing its limitations and is the most commonly used method in present-day computer-aided
circuit analysis programs.

Connected circuit: every pair o f nodes in the circuit is joined by some set o f branches.
Cram ers rule: a method for solving a linear matrix equation for the unknowns, one by one,
through the use o f determinants; the method has serious numerical problems when
implemented on a computer, but is often convenient for small, 2 x 2 or 3 x 3, hand cal
culations.
Floating source: neither node o f the source is connected to the reference node.
141
Gaussian surface: a closed curve or a closed surface surrounding two or more nodes.
Linear matrix equation: an equation of the form Ax =b, where A i s z n x n matrix, x is an n-vector of unknowns, and b is an n-vector of constants.
Loop (closed path): a contiguous sequence of branches that begins and ends on the same node
and touches no node more than once.
Loop analysis: an organized method of circuit analysis for computing loop currents in a circuit.
Knowledge of the loop currents allows one to compute the individual element currents
and, consequently, the element voltages.
Loop current: a (fictitious) current circulating around a closed path in a circuit.

Matrix inverse: the inverse, if it exists, of an n n matrix yl, denoted b y ^ ^ satisfies the equation
A A~^ = A~^ ^ = I,, where / is the w x identity matrix; the solution of the linear matrix
equation, ^
is given hy x = A~^b,
Mesh: After drawing a planar graph without branch crossing, the boundary of any region with
finite area is called a mesh. Intuitively, meshes resemble the openings of a fish net.
Mesh analysis: the special case of loop analysis for planar circuits in which the loops are chosen
to be the meshes.
Mesh current: a fictitious current circulating around a mesh in a planar circuit.

Modified nodal analysis: a modification of the basic nodal analysis method in which the
unknowns are the usual nodal voltages plus some naturally occurring auxiliary currents.
Nodal analysis: an organized method of circuit analysis built around KCL for computing all node
voltages of a circuit.
Node voltage: the voltage drop from a given node to a reference node.
Symmetric matrix: a matrix whose transpose is itself I f ^ =
is a x matrix whose i- j entry
is
then A is symmetric if a - = a-.
142
3. For the circuit o f Figure P3.3, suppose

=
1.2 A. Write a single node equation in the volt
PROBLEMS
age V and solve.
SIN G LE N ODE PROBLEM S

1. For the circuit o f Figure P3.1, write a single
node equation in
, Gj > <^3 >
For a fixed K > 0 , R-^ = R , R2 = 2R ,
Compute Kj
^ 2-
= 2R.
Figure P3.3
in terms o f R and V^j

A N SW ER: - 6 V
4 Ksi-
M ULTIPLE N ODE PROBLEM S

4. The purpose o f this problem is to write the
nodal equations direcdy by inspection o f the cir
cuit diagram o f Figure P3.4. Recall that when the
Figure P3.1
network has only independent current sources and

resistors, the nodal equation matrix is symmetric
A N SW ER: K, = l.5V^^
and the entries can be written down by inspection
2 . The battery o f your car has been dealt a sud
as per the discussion following Example 3.2.
den death by the sub-zero North wind and a
Construct the nodal equations in matrix form for

the circuit o f Figure 3.4 by inspection.
faulty alternator. Unable to fight the elements,

you wait a few days hoping for a thaw, which
comes. You replace the alternator. Then, using
your roommates car, you attempt a jump-start.
Nothing happens. You let it sit for a while with
your roommates car running juice into your
battery for 20 minutes. Still, nothing happens.
W hy wont your car start? Consider the circuit
reference node
o f Figure P3.2. Notice that your dead batter
Figure P3.4
is labeled as Vq. Your roommates battery is
A FEW A N SW ERS: 'Fhe 3-3 entry is G^ + G^
labeled 12 V. Each battery has an internal
+ G^ + Gg, and the 2-1 entry is -G y
resistance o f 0.02 Q and the starter, an internal

resistance o f 0.2 Q. The starter motor requires
5. Consider the circuit o f Figure P3.5 in which
50 A to crank the engine. Find the minimum
/^, = 0.5 A and V^ 2 = 40 V. Furtiier, let G, = 5 mS,

G2 = 2.5 mS, G3 = 2.5 mS, and G^ = 12.5 mS.
(a)
By inspection, what is the
value o f voltage Vq needed before the starter

can draw 50 A and work.
value o f
50 A
>0.020
7 0.20
''load <
v
I
Live
Battery
Write a minimum number o f node

equations and put in matrix form.
0.02Q
V
12V
(b)
(c)
Solve the node equations for the voltages

and
using MATLAB or the for
mula for the inverse o f a 2 x 2 matrix:

Dead
Battery
Figure P3.2
Starter
Motor
a b
c d
'd -b'
ad - be
-c a
143
(d)
Find 1 ^ ,7 ^ and
(e)
Find the power delivered by each

source and the power absorbed by
each resistor. Verify the principle of
conservation of power.
8 . The circuit of Figure P3.8 is an experimental
reference node d
Figure P3.5
measurement circuit for determining tempera
In the circuit of Figure P3.6,
ture inside a cavern underneath the Polar ice

6 . (a)
= 8
cap. The cavern is heated by a fissure leading to
V* Further,
= 5 kQ,
R l = R^ = R^ = 20 k2, and R^ = 10
kfl. Find the node voltages,
and
some volcanic activity deep in the earth. The
V^, and also the voltage, V^. Compute
the range -2 5 C to +25C . The nominal tem
the power absorbed by R^ and the
perature of the cavern is 0C. In this type of cir
power delivered by each of the sources.
cuit, the voltage
It is suggested that you write your
the temperature changes. Suppose that
equations in matrix form and solve
V, and in kQ, R-^ = 20, i?2 44, R^ - 20, and
using MATLAB or the formula for a 2
R^ = 12.5. Note that the 4 4
mA, V^2 -
(b)
2 inverse given in problem 5.
resistor
changes its value linearly from 15
IdQ to 65 k n as a fiinction of temperature over
- Vg is a measure of how
= 50
resistor is a
result of manufacturing tolerances that often
Repeat part (a) when all resistances are
permit deviations from a nominal of, say, 40
cut in half
kflt, by as much as 2 0 %. As usual, it is cost ver

sus precision.
(a)
Write a set of nodal equations in the
(b)
Assuming
----------^/S/^-------
variables
and Vq
= 40 kQ at 0C, put
the nodal equations in matrix form

and solve for the node voltages,
and Vq
(c)
Determine the power delivered by the

source.
Figure P3.6
(d)
Use MATLAB to solve for all the node
7. In the circuit of Figure P3.7,

^2
= 4 V, and
- 1 mA. Further, in mS, G| =
voltages as
0.4, G2 = 2, G3 = 3, and
out. Then find the linear equation
analysis to find
= 5. Use nodal
and V^. Then compute the
power delivered by the independent sources

and the power absorbed by G2 .
varies from 15 to 65
ki2 in 1 kfi increments. Do not print

relating
to temperature. Plot Vq
- y c as a W c tio n of temperature, i.e.,
over the range - 2 5 C to +25C . Over
what range of temperatures about 0
degrees would the sensor be reason
ably accurate? W hy and why not?
144
II.
Consider the circuit o f Figure P 3 .l l .
Choose node D as the reference node. This

choice eliminates the floating voltage source
and hence the nodal equations can be written
without the need o f a so-called supernode. Let
= 0.0 8 S, Gj = 0.08 S,
0.0 2 S, G 5 = 0 .0 2 S,
= 0.3 A, and
= 0.01 S, G^ =
= 0.3 A, 7^2 = 0 .2 A, 7^3
= 50 V. Write and solve a set
o f nodal equations for the voltages
Figure P3.8
and Vg = Vg0 . Then compute the powers

9. In the circuit o f Figure P3.9, ail resistances
are 1 ItQ, except
= 500 Q. Suppose
delivered by each o f the sources.
=
G,
100 V and 1 ^ 2 = 0-3 A. Compute all tiie node

voltages o f the circuit. You may want to use
MATLAB or a calculator that inverts matrices
G,
to compute the answer. Compute the power
'
delivered by the independent sources.
Figure P 3 .1 1
= 20 kQ, T?3 = 20 kQ,
= 5 mA. Find
Figure P3.9
1,7,0.6,7
10. In the circuit o f Figure P3.10,
= 30 V,
= 0.6 A. Use nodal analysis
on the circuit below, as indicated:

(a)
Figure P3.12
Write a nodal equation at node A.
(b)
Write a nodal equation at node B.
(c)
Write a third nodal equation at node C.
(d)
Solve the 3 equations in 3 unknowns
13. Consider Figure P3.13.

(a)
M ATLAB, or using some other soft

ware program to obtain all the node
this, let G-=\ I R -.

(b)
W ith K ^ = 150V ,7?^ = lk Q ,7?, = 5 k Q ,
(c)
= 15
, Vg the power delivered by
, and the power absorbed by R^ .
Compute I 2 , the current through Rj
^2 = 10 kQ, T?3 = 10 kQ, and
voltages. Show ALL work/procedures.
mS, find
Find the power delivered by the inde

pendent voltage source.
ion
100
Write the nodal equations and place in

matrix form prior to solving. In doing
by hand, with your calculator, using
(e)
and the power delivered by
the dependent current source.
ANSWERS IN RANDOM O R D E R
7^2 = 1-2 A, and
= 5 kQ,
= 0.55 x 103, 7-
from left to right.
ion V.
ion.
Reference node
Figure P 3 .1 0
Figure P 3 .13
145
14. Consider the circuit of Figure P 3.14

(a)
17. Use nodal analysis to find the voltages
Write two node equations in terms of

the literal variables in Figure P 3.14
(c)
(e)
in the circuit of Figure P 3.17.
= 20 Q, i?2 = 10 Q,
= 10 Q, and I^=
= 4 Q,
A. Note that in
and put in matrix form.
0.1 S,
Solve the node equations for the volt
solving this problem, you are to generate three
and Vq when
= 0.1 A, 7^2 =
(nodal) equations in which the unknowns are
0-2 A,
= 7 mS;
= 2 mS,
=
500 Q ; i?2 = 333.33 2; and /?3 = 1 fl.
Determine Kq.
you could eliminate the equation
ages
(d)
, and
Suppose
for
but this problem is to illustrate that
such elimination is not necessary. Finally, deter
Determine the power delivered by
mine the equivalent resistance seen by the inde
each source. (Be careful of sign.)
pendent current source.
. 9.iVa
'4
Figure P3.14
Figure P3.17
15. Consider the circuit of Figure P3.15 in which

= 20 V and
(a)
= 0 ; node voltage C is
Write the two nodal equations in

terms of the literal variables.
(b)
Suppose 11 = 6 and the following in S

are given:
= 0.5, G2
= 4, 6*5 = 1. Solve for
and
Check: 20 and 10 volts.

(c)
Find
and then find the equivalent
resistance seen by the independent

voltage source.
(d)
"
18. Consider the circuit of Figure P3.18. By

choosing node C as the reference node, we elim
inate a floating voltage source. Write an appro
priate set of nodal equations, with node C as the
reference node. Solve the nodal equations, speci
fy the voltages V^q Vbo V^Q and V
and the
power delivered by the sources. Finally, find the

equivalent resistance seen by the current source.
Leti?i = 9 k Q ,7 ? 2 = l S k i 2
= 6 kQ,
kQ,
= 3000 Q, and = 20 mA.
= 9

pendent source and the dependent
source.
(e)
W hat is the power absorbed by the

output resistor?
-------- --------------
Figure P3.18. By choosing node C as the ref

erence node, it is possible to simplify the con
struction of the node equations.
19. Consider the circuit of Figure P 3.19 in
Figure P3.15
16. Redo problem 15 with
\
J
= 60V .
0.25 S and
which
60 ^2.
- 20 Q, /?2 ~ 20 Q,
= 2 0 Q,
= 30 Q, R^ =
= 12 V, and
146
FLOATING VO LTAG E SO URCE

PROBLEM S
0.6 A. The point o f this problem is to illustrate

how a good choice o f reference node may sim
plify the calculation o f node voltages, whereas a
poor choice may lead to a complicated formu
22. For example 3.5 suppose all resistance values
lation o f the node equations.
are doubled, the floaring voltage source remains

Choose a reference node so that there
the same at 440 V, and all current sources are
are no floating voltage sources. Write
scaled down to one-half o f their original values.
three equations in the unknown volt
(a)
Compute all node voltages and the
ages. Solve for the node voltages.
current
C H EC K : i^ = 2 A k and iy = - 3 0 A.
(b)
Compute the voltages
and
(b)
(c)
Compute the power absorbed by the
each source.
0.075 S resistor.
(c) Determine the power absorbed by
(d)
Compute the power delivered by the
each resistor.
floating voltage source.
(d) Verify conservation o f power using the
(a)
results o f parts (b) and (c).
23. For the circuit o f Figure 3.10 in Example 3.6,

suppose the 110 V source is changed to 200 V and
the 50 Q resistor is changed to 500 2. Find the
node voltages
rents
and
Vy,
and the unknown cur
Then find the powers delivered by
the 30 V source and the dependent source.

which
= 2 0 0 V, V^2 = 5 0 V, R^ = 5 0 Q, R.^
= 20 Q, R^ = 50 Q, and R^ = 40 Q.
(a)
Identify the floating voltage source and
(b)
W rite
add a current label through the source.

20. The nodal equations for the circuit in
Figure P3.20 are
0.03
0.09
modified
nodal
equations,
which include both node voltages and

- 0 .0 r Vf
0.04
unknown currents through any float
'A-.'
ing voltage sources.
.^0.
Compute the values o f R^, Rj ,
(c)
,a n d .
Solve the equations for the node voltages Kg and Vq and the current
through the 50 V source. C H ECK :
Vg = 50 V and
Figure P3.20
= 100 V.
(d)
Find the power consumed by R^ .
(e)

each o f the sources.
21. Consider the circuit o f Figure P3.15, which

has nodal equations {R^
0 .0 0 8
0 .0 1 9
- 0 .0 0 5 '
-0 .0 0 1
0 .0 0 5 5
- 0 .0 0 2
-0 .0 0 5
- 0 .0 0 2
0 .1 0 7
Compute the values of /?j,
0) given by
0
=
0
O .n/
0
=
0
G A.
>-^3 >-^4 -^5 ^nd ,u.
Figure P 3 .2 4
147
27. T he modified nodal equations for the cir
which l/j = 2 50 V,
cuit o f Figure P 3.27 are
: 20 Q,
(a)
(b)
= 5 0 Q,
= 5 0 V,
= 50 Q ,
= 40 Q , and R^ = 10 Q.
0.004
-0 .0 0 1
-0 .0 0 2
add a current label through the source.
-0 .0 0 1
0.001
-1
W rite
-0 .0 0 2
0.004
-1
-1
Identify the floating voltage source and

modified
nodal
equations,
which include both node voltages and
0 T 4 '
O'
yB
0
^CB
unknown currents through any float

(c)
ing voltage sources.
Compute all four resistor values and (3. Hint:
Solve the equations for the node volt-
Find all the conductances first and then convert
ages Vg and Vq and the current
to resistances.
through the 50 V source. C H EC K :

5 0 V a n d V^= 100 V.
(d)
(e)
Find the power consumed by R^ .

Rc
Figure P3.27
= 0.02
= 10 Q.,
=

S, ^2 = 0.025 S, Gg = 0.2 S,
0.4 A, and V^ 2 ~ 12 V. Use nodal analysis to

find all node voltages, the current , the power
absorbed by
and the power delivered by
the two sources.

Figure P3.25
26. Consider the circuit o f Figure P3.26. R^ =
10 Q ,
= 100 Q, R^ = 100 Q,
= 50 Q ,
= 100 V, K^2 = 60 V, V^3 = 100 V,

= 14
A. Label appropriate currents
and I j^q
through the floating voltage sources.
(a)
Write the modified nodal equations

for the three unknown node voltages
and two unknown currents.
(b)
Solve for the five unknowns (in MATLAB).
(c)
Find the power delivered by each o f
(a)
Determine Vq
(b)
Label the current /ABUsing V^, Vg,

and 4 as
unknowns, write a 4 x 4 matrix set o f
(c)
nodal equations.
(d)
Solve the nodal equations for V^, Vg,
ÂB and 4 .
(e)
(0
Determine the power absorbed by G2 .

Determine the power delivered by all
sources.
AN SW ERS (D) IN R A N D O M O R D E R : 12,
0.24, 9.6, 0.16
the sources.
29. Repeat Problem 28, except this time write

Reference node
Figure P 3 .2 6
only three nodal equations in the variables V^,
Vg, and l^g. Notice that you must express
in
148
terms o f
and the appropriate conductance.
One can even reduce the number o f equations
AN SW ERS (R A N D O M IZ ED ): 250, 325, 1 2 5 ,7 5 ,2 5 0 , 25
to two using the so-called supernode approach,
32 Consider the circuit o f Figure P3.32, where
which is the subject o f other texts.
= 4 kQ,
= 100 Q,
mS,fj. = 4 S,
= 1
= 4/3 kQ,
= 0.75
= 160 V, and
= 40 mA.
= 300
(a)
Specify V^.
2 A,
= 2 A, Vj2 ~
nodal analysis
to find all node voltages, the current
, the
(b)
Write modified nodal equations.
= 20 Q,
= 20 Q, G4 = 0.09 S,
(c)
nodes B and C, and the power delivered by the
(d)
independent sources as follows:

the sources.
(a)
Determine Vq
(b)
Write a set o f modified nodal equa
(e)
tions that contain extra current vari

ables including
Solve the modified nodal equations in

MATLAB.
power absorbed by the 20 Q resistor between
and
Compute the power absorbed by each

resistor.
(f)
Verify conservation o f power.
(c)
Solve your nodal equations for the
(d)

independent sources.
unknowns. C H EC K : Vg= 180 V.
(e)
Compute the power absorbed by the

20
resistor between nodes B and C.

Figure P 3.32
33. Consider the circuit in Figure P 3.33 in
which V-^ = 60 V, G j = 0.1 S, G2 = 0.1 S,
= 0.3 S, G4 = 0.4 S, G 5 = 0.1 S, Gg = 0.1 S, Gj

= 7/480 S, ^ = 3, and (3 = 2 .
(a)
Figure P3.30
using
31. Use nodal analysis on the circuit o f Figure

P3.31 as indicated.
all resistors are 10 Q.
(a)
W rite
= 100 V,
modified
1^2
nodal
= 1 A, and
equations
including the extra variable IgQ

(b)
(c)
Write the modified nodal equations
(b)
(c)
Vg Vq and /^âs unknowns.

MATLAB.
Find
the equivalent
resistance seen by the independent

voltage source.

MATLAB.
Find the power delivered by each of
the sources.
Reference node
Figure P3.31
Figure P3.33
AN SW ERS T O (C) IN RA N D O M O R D E R :
40, -2 5 ,3 8 .7 5 ;/ ?
= 12 Q .
149
SIN G LE LOOP-EQUATION
PROBLEM S
= 50 Q, and
= 0.5. If
= 400 Q,
= 50 V, find 4 ,
the power delivered by the independent and
37. Consider the circuit o f Figure P3.37

(a) Suppose
sources.
R,
dependent voltage sources, and the equivalent

resistance,
= 2 0 0 Q, R^ = 300 Q, R2 = 500
Q, /jj = 750 mA and I ^ 2 = 100 ^lAh

the power delivered by each o f the independent
--------- ----------- ---
, seen by the independent source.
Figure P3.37
C H EC K : /j = 100 mA.
(b)
C H EC K S: R^^ = 2 50 Q, and
3 5 . In the circuit o f Figure P 3.35,

= 5 0 Q, and
= 50 V, find
= 8 watts.
= 4 00 2,
= 50 Q, and jj. = 0.5. If

, the power delivered by the
independent and dependent voltage sources,

and the equivalent resistance, R^^ , seen by the
independent source.
Now suppose /^j = 4 0 0 mA and 1 ^ 2 =

100 mA and the loop equation for /j
written in the standard way directly
yields 2000/|= 60. Find R-^ and R2 if
R^ = 600 fil. Note: If the equations are
not written in the standard way, the
solution is not unique. For example,
multiplying both sides o f the above
equation by 0.5 yields a different
answer in which R-^ = 140 2 , as
opposed to the correct answer o f R-^ =
400 Q.

/j2 = 100 mA, 7^3 = 2 0 0 mA, and
mA. Find
Figure P3.35
CH ECKS: R^^ = 200 Q, and
= 56 V,
= 100 mA
and the power delivered by each
independent source.
12.5 watts.
= 200 V
and 1 ^ 2 = 20 mA. Find

Then find the power
delivered by each o f the independent sources.
Finally, find the power absorbed by each resistor
and verify conservation o f power for this circuit.
Figure P3.38
Figure P3.36
C H EC K : 7, = 4 mA.
C H EC K : Sum o f powers delivered by the

sources is 15.68 watts.
150
with internal resistances R

(a)
(b)
(c)
Suppose
= 2 50 Q,
= 5 0 0 Q,
20 Q and R2 = 80
Q (faulty connection) respectively connected in
= 100 V, and /S = 0.5. Use loop analy
parallel to supply power to a load o f 7?^ = 80 Q.
sis to find /j and R^^ .
Compute the power absorbed by the load R^
Compute the power dehvered by each
and the power delivered by each independent
source and absorbed by each resistor.
source. W hich battery supplies more current to
Verify conservation o f power.
R^ and hence more power to the load? How
Compute R^^ as a function o f R^,
much power is wasted by the internal resistanc
and 13 . Suppose
5 00 Q , plot R
s 13 s
= 250 Q and
as a function o f
,0
es o f the battery?
Pi,
0 '
C H EC K : P.s\
3.15 watts and P ,2 = 1 .8 watts.
Figure P3.39
C H EC K : Power absorbed by resistors is 15
watts and R^^ > 4 50 Q .
42. Reconsider the circuit o f Problem 3.41,

redrawn with different loop currents in Figure
3.42a and 3.42b. Th e point o f this problem is
40. (a) For the circuit o f Figure P3.40, R^ = 1

kQ,
= 5 kQ, /?3 = 4 kQ,
= 100
mA and g = i xlO '^ S. Find /j and

by writing two equations in the
two unknowns /j and
. The first
to verify that different sets o f independent loop

equations produce the same element currents
and branch voltages.
(a)
/j and
equation is the usual loop equation
(b)
(c)
3.42a, and then find the voltage across
and the second determines the rela

tionship o f /j and
Write the new loop equations and find
and the power consumed by R^ .

(b)
Write the new loop equations and find
Given your answer to (a), find the
/j and I 2 for the circuit o f Figure
equivalent resistance, R^^, seen by the

independent source.
3.42b, and then find the voltage across

and the power consumed by R^ .

dependent source.
Figure 3.42
Figure P3.40
43. The matrix loop equation o f the circuit o f

Figure P3.43 is
M ULTIPLE LOOP PROBLEM S

41. The circuit o f Figure P3.41 represents two
non-ideal batteries
= 21 V and
= 24 V
150
-4 0
-1 0 0 '
-4 0
140
= -2 0
-1 0 0
150
20
h'
TOO
151
Find the value of each resistance and each

source in the circuit.
(b)
the current in the locomotive motor and
(c)
repeat parts (a) and (b) when the loco
the power absorbed by the locomotive

motive is 1/3 distant from either station
CHECKS: V^2 = ^0 V,
= 40 Q.
44. The mesh equations for the circuit of

Figure P 3.44 are
40
-8 0
- 1 0 \h'
-3 0
130
-5 0
-1 0
-5 0
70
[*3.
n
=
-V2
0
Figure P3.45
Find
*
46. Reconsider Problem 3.45. Let
= 590 V
and R^ = 1.296 Q. This time, suppose there are

two locomotives on the track. One is 1/3 dis
tant from the East side station, and the other is
1/3 distant from the West side station.
(a)
Determine the resistance R in Figure
P 3.46.
(b)
Using the indicated currents, write a

set of three mesh equations and solve
(c)
45. Figure P 3.45a shows an electric locomotive
(d)
propelled by a dc motor. The locomotive pulls

a train of 12 cars. The motor behaves
like a 590 V battery in series with a
for
, ^2
H
Determine the two motor currents.
each of the 660 V sources.
---------------------- -------- ------------- 1--------- -----------------R

__ I
^
R
A
1.296 Q resistor. Suppose the train is

midway between stations, West side 660 V
and East side, where 66 0 V dc
sources provide electricity. The

resistance of the rails affects the cur
rent received by the locomotive. The
equivalent circuit diagram is given by Figure
P3.45b, where R
660VVf
R
R
Figure P3.46
0.15 Q. Using m e ^ analy4 7 . Reconsider the Problem 3.5 and the circuit
sis find
(a)
l=
the currents
and I 2
of Figure P3.5. Draw two loop currents and
152
solve for these currents. Then compute the
node voltages
which
= 40 V and V^ 2 = 20 V. Write a set o f
three loop equations by inspection. Refer to
n\
Example 3.7 and the discussion following the
> and
48. In Figure P 3.48, let

= 6 kQ,
= 9 kQ,
= 9 kQ,
= 18 kQ,
=
example. Solve the loop equations using matrix
and /2.
software program. Compute the voltage v. Note
= 3000 Q, and
methods via your calculator or an appropriate
20 mA.
(a)
Write two mesh equations in
that /| and I 2 should have values identical with
Put in matrix form and solve.
those in example 3.7. Finally, find the power
(b)
Specify the voltages 1^^,
(c)
^BO ^CD
K iD'
delivered by each o f the sources.
IQ
the sources.

which 1/j = 250 V, V^ 2 = 50, V,
= 50 Q,
= 20 Q,
= 50 Q,
= 40 Q, and
= 10 Q.
(a)
51 . Consider the circuit o f Figure P3.51.

(a)
Write three standard loop equations
matrix form in terms o f the literal
and put in matrix form.

(b)
Write two mesh equations and put in
parameters.
Solve the equations for the loop cur
(b)
rents and determine the node voltages
Solve the mesh equations for the

unknown currents assuming
and
Q,
= 40 Q,
A, and
= 20 Q,
V.
= 100
= 80 Q,
(c)
Find the power consumed by
(d)
(c)
Find
(d)
r^
and V^.
pendent sources.
(e)

dependent source.
r )
n
Figure P 3 .4 9
Figure P 3.51
153
52. Repeat Problem 51 when 7?, = 100 3, ^2 =

40 Q,
= 80 Q,
= 60 2,
= 1 A, and
K2 =
V-

(a)
Write two mesh equations in terms o f

the literal parameter values.
(b)
Solve the mesh equations assuming

= 100 Q, ^2 = 40 Q,
= 60 f i,
80 Q,
1^2
= 80 Q,
= 1 A, and V^j =
40 V.
(c)
Compute
(d)
Find the power delivered by all the
and
sources in the circuit.
CHECK: /, = 0.1 A, K- = 50 V, and V = 20 V.

(a)
Write the modified loop equations (using

the indicated loops) in matrix form.
(b)
Figure P3.53
(c)
If I/, = 200 V, 7^2 = 0.3 A, /?! = 7^3 =

100 2, T?2 = 400 2, and /i = 0.5, find
7j , 72 , and V^ 2
W hat is the power delivered by the
three sources?
54 . Repeat Problem 53 when

= 40 Q,
30 3,
0.25 A, and
= 40 Q, 7?2 =
= 20 2,
= 10 2, 7^2 =
= 60 V.
>mv,-
M O D IFIED LOOP ANALYSIS

PROBLEM S
55 . Consider the circuit o f Figure P3.55. The
objective o f this example is to illustrate a
numerical approach to loop analysis where the
number o f variables to be found is quite large,

(a) Write the modified loop equations in
matrix form.
but the equations are quite easy to write and do

(b)
not require multiple substitutions.

(a)
A, 7?, = 200 2, 7?2 = 400 Q,
If the loop equation matrix is o f the
Q,
form below, compute the undeter

(c)
1
1 + ??'
??
-1
R3
????
-1
(b)
If
y s -R ih n
A'
=
0
0
= 500 2, 7^2 = 100 Q,
Q, 7?4 = 100 2,
= 100
= 500 Q, and TJj = 1150 Q,
compute 7j, 72 , 73 , and
mined entries.
'7?2 + ? ?
If 1/, = 210 V, K^2 = 150 V, 7^3 = 0.1
= 400
= 150 V, / = 0.5
A, find the three unknowns.

154
C H EC K : /j = 0.4 A and
= 50 V.
58. Consider the circuit of Figure P3.58.

(a)
Write the modified loop equations in

matrix form in terms of the literal values.
(b)
If
= 4 0 0 ^ ^ ^ 2 = 2 0 0 V,i?i = 3 0 Q ,
= 20 Q,
= 270 Q,
= 80 Q, T?5
= 140 Q, and compute /p /2 , 73, and

(c)
Compute the power delivered by all

sources.
(d)
Compute the power absorbed by each

resistor and verify conservation of power.
Ri
R3
R.
-VESA
's <-----N/N^
Rs
Figure P3.58
CHECKS: V, = 265 V and
watts.
= -3 9 7 .5
C H A P
'
"
L in D lif ie r
The Operational Amplifier
Amplification o f voice allows announcers at sports events to convey their comments on the playby-play action to the crowd. At concerts, high-powered amplifiers project a singers voice and the
instrumental music into a crowded auditorium. Electronic amplifiers make this possible. One o f
the simplest and most common amplifiers is the operational amplifier, the subject o f this chapter.
The word operational, though, suggests a purpose beyond simple amplification. Often one must
sum signals to produce a new signal, or take the difference o f two signals. Sometimes one must
decide whether one dc signal is larger than another. The operational amplifier is operational pre
cisely because it can be configured to do these things and many other tasks, as we will see later in
the text.
CH APTER O BjEC TIV ES

1.
Introduce the notion o f an ideal operational amplifier, called an op amp.
2.
3.
4.
Describe and analyze basic op amp circuits.

Describe and illustrate a simple method for designing a general summing amplifier.
Describe and illustrate the phenomenon o f saturation in op amp circuits and describe cir
cuits that utilize saturation for their operation.
SECTIO N HEADIN GS
1.
2.
3.
4.
5.
Introduction
The Idealized Operational Amplifier: Definition and Circuit Analysis
The Design of General Summing Amplifiers
Saturation and the Active Region of the Op Amp
Summary
Terms and Concepts
Problems
156
Chapter 4 T h e Operational Amplifier
1. IN TRO D U CTIO N
Chapters 1 and 2 defined and discussed independent and dependent voltage and current sources.
Chapter 3 investigated the nodal and loop analysis o f resistive circuits containing such sources.
Ofi:en, dependent sources supply energy and power to a circuit, making them so-called active ele
ments. O n the other hand, resistors are passive elements because they only absorb energy. Circuit
models o f real amplifiers (see Examples 3.3 and 3.4 with associated Figures 3 .7 and 3.8, respec
tively) contain controlled sources that underlie their analysis and performance evaluation. Indeed,
the VCVS is the core component o f the operational amplifier (op amp), the main focus o f this
chapter. Thus, the op amp is an active circuit element whose analysis is done with the techniques
o f Chapters 1 through 3.
A real op amp is a semiconductor device consisting o f nearly two dozen transistors and a dozen
resistors sealed in a package from which a small number o f terminals protrude, as shown in Figure
4.1(a). Despite its apparent internal complexity, advances in integrated circuit manufacturing
technology have made the op amp only slightly more expensive than a single discrete transistor.
Its simplicity, utility, reliability, and low cost have made the op amp an essential basic building
block in communication, control, and the instrumentation circuits that can be found in all under
graduate EE laboratories.
Top View
Balance 1 [
TO-5
Dual-in-line
Inverting
input
Noninverting ^ [
input
E- 4 [
3 6 Output
H 5 Balance
(a)
Inverting
E+-
input
V-
Output
Reference
node
Noninverting
E-
input
(c)
(d)
FIG U RE 4.1 (a) Typical op amp packages; (b) typical terminal arrangement of an op amp package;
(c) dual power supply notation; (d) essential terminals for circuit analysis.
Figure 4.1 (b) shows a typical arrangement o f terminals for a dual-in-line op amp package. Th e ter
minal markings and the symbol shown in Figure 4.1(b) do not appear on the actual device, but
Chapter 4 T h e Operational AmpUfier
157
are included here for reference. In Figure 4.1(b), the terminal labeled N C (no connection) is not
used. The E+ and E- terminals (Figure 4.1(b)) are connected to a dual power supply, illustrated in
Figure 4.1(c), where
typically ranges between 3 V and 15 V, depending on the application;
adequate voltage is required for proper operation. The three terminals in Figure 4.1(b) marked
inverting input, non-inverting input, and output interact with a surrounding circuit, and
correspond to V,
and Vq in Figure 4.1(d). The two terminals labeled balance or ojf-set have
importance only when the op am is part o f a larger circuit: resistors o f appropriate values are con
nected to these terminals to make sure the output voltage is zero when the input voltage is zero.
This balancing process is best discussed in a laboratory session.
This chapter sketches the basic properties o f the op amp: just enough to understand some o f the
interesting applications. The ideal op amp model and the saturation model are described. Using
these models and the principles o f analysis covered in Chapters 1 through 3, we then analyze the
behavior o f some widely used op amp configurations. These application examples hint at the
importance o f the op amp and furnish motivation for the study o f electronic circuits.
Several o f the examples include a SPIC E simulanon o f the circuit being analyzed. SPIC E is a
sophisticated circuit simulation program. Behind the user-interface, SPIC E uses complex models
o f the real operational amplifier. Our purpose in using SPIC E simulation is to verify or test the
theoretical analysis set forth in the examples. W hat we show is that the simplified theoretical
analysis provides a very good approximation to the actual circuit behavior represented in the
SPIC E simulation results. Industrial circuit designers often use SPIC E to visualize the expected
behavior o f very complex circuits. Later chapters cover some o f the more complex op amp appli
cations.
2. TH E ID EALIZED O PERATIO N AL AM PLIFIER

This section analyzes resistive circuits containing an operational amplifier. Figure 4 .2 explicitly
shows an op amp embedded in a surrounding resistive circuit.
FIG U RE 4.2
One possibility for analyzing op amp circuits is to represent the op amp by one o f the simplified
models shown in Figure 4.3 that do not account for saturation effects. The first model o f Figure
4.3(a) consists o f an input resistor,
an output resistor,
and a V CV S with finite gain A.
O f practical import is the idealization o f this model (Figure 4.3(a)) to the one o f Figure 4.3(b) by
158
(1) letting R-^ become infinite, setting up an open circuit condition at the input terminals; (2) let
ting
become zero, making the output voltage o f the op amp equal to that o f the V C V S; and
(3) letting the gain A approach infinity. These conditions are idealizations because (1) with R-^
infinite, there is no loading to a circuit attached to the input; (2) with R^^^ = 0, the full output
voltage appears across any circuit connected to the output; and {5) A ^
leads to a simplifica
tion o f the associated analysis. These conditions, stated below as equation 4.1, define the so-called
ideal operational amplifier;
-> 00 (infinite gain)
R-^
Rout
(4.1a)
00 (infinite input resistance)
(4.1b)
0 (zero output resistance)
(4.1c)
Rest of Circuit
Rout
+
+
V
A(v - V
(a)
FIG U RE 4.3
To see how this idealization simplifies op amp circuit analysis, consider an equivalent set o f conditions for the ideal op amp, called the virtual short circuit model:
(4.2a)
(4.2b)
v^ = v_
(4.2c)
From Figure 4.3(b), the conditions that = 0 and i_ = Q follow directly from the open circuit con
dition at the input terminals. The condition that
= v_ (hence, the term virtual short circuit)
will be discussed later, but occurs because A ^ co, forcing {v^ v j
0.
159
The recommended way to analyze circuits containing op amps is to replace any ideal op amp by the
model o f Figure 4.3{b), the virtual short circuit model o f equation 4.2. Th e following examples illus
trate the use o f the virtual short circuit model.
EX A M P L E 4 .1 . This example investigates the inverting am plifier o f Figure 4.4, which is used in
a wide range o f commercial circuits. The objective is to compute
Rjs and
in terms o f
Rf
FIG U RE 4 .4 Inverting amplifier, assuming an ideal op amp in which V, = ---- V: .
So l u t io n
Step 1. Compute
and v_. Since the + terminal is grounded,
= 0. From the virtual short prop
erty o f the ideal op amp, v_= v^ = Q.

Step 2. Compute i^-^. Since v_ = 0, the voltage across R-^ is v-^. From Ohms law,
v,
R
Step 3. Compute iy. Again, since v_ = 0, the voltage across Rf is
From Ohms law.
if = R
Step 4. Relate the currents i-^ and ip and substitute the results o f Steps 2 and 3. From KCL, i^^ i_
+ ir= 0. From the properties o f the ideal op amp, i_ = 0, in which case, ir= -i-^. This imphes that
R*
V
Hence, the voltage gain relationship o f the inverting op am circuit is
Rf
V ou t= --^ ^ in
Rin
(4.3)
Equation 4.3 shows that the input and output voltages are always o f opposite polarity, hence the
name inverting am plifier. One also observes that by choosing proper values for Rjr and R^ a volt
age gain o f any magnitude is possible, in theory. In practice, other factors limit the range o f obtain
able gains.
160
Exercises. 1. Find
lOOkO
25kQ
-s/ W '
50 mV
FIG U RE 4.5 Inverting amplifier.

2. Find
lookn
FIG U RE 4.6 Inverting amplifier with additional resistor.

A N SW ER F O R BO T H :
= - 200 mV.
A few remarks are in order. Op amp configurations in which one o f the input terminals is ground
ed, as is the non-inverting terminal in Figure 4.4, are said to operate in the single-ended mode. The
input terminal can be grounded directly or through a resistor, as in Exercise 2 above. Also, since
v_ =
or v_-v^ = 0, the terminals are virtually short circuited even though there is no hard-wired
direct connection between them. This condition is called a virtual short circuit. Further, if one
o f the terminals is grounded, then the other terminal is said to be virtually grounded, as is the
case in Figures 4.4, 4.5, and 4.6. Specifically, in Figure 4.4, there is a virtual ground at the invert
ing input terminal.
The next example continues the investigation o f the ideal inverting amplifier for the two-input,
single-output op amp circuit o f Figure 4.7. The solution again makes use o f the virtual ground
and virtual short circuit properties o f the ideal op amp.
161
E X A M P L E 4 .2 . For the circuit o f Figure 4.7 , our objective is to compute

and the two input voltages
in terms o f R^, ^2>
and v^2 Rf
+
V
FIG U RE 4.7 Inverting (ideal) amplifier with rwo inputs for which
Rf
Rj
v,=-^v,|V ,2.
So l u t io n . A s in Example 4.1, and by the same reasoning described there, v_ = v^ = 0.

Step 1. Compute ij and
Since v_ = v^ = 0, the voltage across
is
and the voltage across Rj
is v^2 - From Ohms law,
R^
Step 2. Compute ij-. Again, since v_ = 0, the voltage across Rj- is
and from Ohms law.
'/ = R
/
Step 3. Relate the currents /j, ^2
and then substitute the results o f Steps 1 and 2. From KCL,
H i_ + ijr = ^- From the properties o f the ideal op amp, i_ = 0, in which case, ijr = -(/j +
This implies that
Hence,
Rf
Rj
(4.4)
162
Exercise. In Figure 4.7, suppose Rjr= 100 IcD. Find

AN SW ER:
= 25 kD and
and ^2 so that
= 50 kD.
E X A M P L E 4 .3 . This example analyzes the non-inverting operational am plifier circuit o f Figure

4.8. As in Examples 4.1 and 4.2, the objective is to compute
in terms o f R^, R2 , and v-^. We
show that
''out
V.
FIG U R E 4.8 A non-inverting op amp circuit.
So l u t io n
Step 1. Compute
and v_. Since the + terminal is connected to the input voltage source,
= v-^.
From the virtual short property o f the ideal op amp, v_ = v^ = Vj^.

Step 2. Compute zj . Since v_ =
the voltage across the resistor
is
Observe that the current,
Zj, has reference direction different from the passive sign convention. Hence, from Ohms law,
h=Step 3. Compute ip Again, since v_ = v-^, the voltage across R^-is
lf =
From O hm s law,
Vout - Vin
Step 4 . Relate the currents z'j and ip and substitute the results o f Steps 1 and 2. From KCL, z'j - z_ +
Zyr= 0. From the ideal op amp property o f equation 4.2, z_ = 0, forcing z^= -Zj. This implies that
Rf
/?,
163
Chapter 4 T h e Operational AmpHfier
Hence, the input-output voltage relationship
Rf\
(4.5)
1+ ^
R i)
/
R \
From equation 4.5, the voltage gain is greater than 1, i.e., i 1 + - i - i > 1, and
I
}
have the same polarity; the circuit is naturally called a non-inverting amplifier.
Exercise. For the non-inverting amplifier o f Figure 4.8, find
and
and
always
so that the gain is 2, and
when v-^ = 5 V, the power absorbed by R^ is 5 mW.

i,n
AN SW ER: Rf = 5\<Q. and /?,^ = 5 kQ
EXA M PLE 4 .4 . This example analyzes the ideal general di^Ference amplifier circuit o f Figure
4.9. We show that
Kf
^s 2 - - r ^ s \
R\
'" " ' I
In a basic difference amplifier, the output is the difference o f two input voltages. For the gener
al difference amplifier o f this example, the output is a difference o f the scaled input voltages,
for appropriate positive
and a^.
So l u t io n . From the ideal op amp property o f equations 4.2, v_ =
and no current enters the
inverting and non-inverting op amp input terminals.

Step 1. Write a node equation at the non-inverting input terminal o f the op amp. Summing the cur
rents leaving the + node o f the op amp yields
G 2 { v + - v , 2 ) + GgV+=Q
164
Solving for
leads to
(4.6a)
Step 2. Write a node equation at the inverting input terminal o f the op amp. Recall v_ = v^. T he sum
o f the currents leaving the - node satisfies
G i ( v+ - v, ] ) + G ^ ( v+ - v,) = 0
Thus,
Cl
ôut =
1+
(4.6b)
Step 3. Combine Steps 1 and 2. Substituting equation 4.6a into 4.6b yields
(4.7a)
G t + Go
or, in terms o f resistances,
[R g + R2 j
(4.7b)
^s2
Equations 4 .7 have the desired form:
appropriate positive constants <?2
a ^, which can be obtained by proper choices o f the resistors.
Two special cases o f Example 4 .4 are o f practical importance. First, if
and R^ = /?2> then
equation 4.7b reduces to the classical difference amplifier equation.
out
= K {v , 2 -V si)
with K = I and for an arbitrary K > 0 , Rj- = KR^ and R^ = KR2 fits the bi
bill.
Exercises. 1. In Figure 4.9, if
= 7?2 = 5 kQ and K = 2 , find R p and R^.
ANSW ER: R j - = R^= \0 kQ

2. Using the circuit o f Figure 4.9, design a difference amplifier so that
= 4(t^^2 ^ji) ^^d the
feedback resistance R^ = 20 kQ.

A N SW ER: R, = R , ^ 5 kQ and R^ = 20 kQ
3. In Exercise 1, suppose R^ and /?yare scaled by a positive constant A'j, i.e., R^^^ = K^R^y and R2
and R are scaled by a positive constant K^. Determine the new input-output relationship.
o
A N SW ER:
w ith
the sam e as in Exercise 1
165
The point o f Exercise 3 is that the group
, Rj] can be independently scaled by
and the group
{Rj, R^ independently scaled by K 2 without affecting the gain o f equation 4.7b.

E X A M P L E 4 .5 . This example analyzes a special case o f the non-inverting amplifier called the
buffer or isolation amplifier, shown in Figure 4.10, where
= v-^. W hen connected between
two circuits, the buffer amplifier prevents one circuit from having a loading effect on the other.
FIG U RE 4.10 The buffer or isolation amplifier for which
So l u t io n . From the connection shown in Figure 4 .1 0 , v-^ =

ties o f the ideal op amp,
= v_, in which case
and
=
= v_. From the proper
= v-^.
Exercise. Compute the power delivered by the source in Figure 4 .1 0 and the power delivered to
the load R^.
^
AN SW ER: 0 and
R,
The circuit o f Figure 4.10 is called an isolation or buffer amplifier, because no current is drawn
from the source
maintaining
However, the op amp does supply current (and power) directly to the load by
under the condition that
put current rating. Since Vgf{t) =
not exceed the manufacturers maximum out
the circuit is also called a voltage follower.
Figure 4.11 shows a SPIC E simulation that verifies the behavior arrived at in Example 4.5. Here
a dc voltage sweep, Q <v-^< 12 V, was input to a highly accurate SP IC E model o f a Burr Brown
741 connected to 10 V power supply. Observe in Figure 4.11 that the output follows the input
up to the 10-volt value, after which, the output remains at 10 V despite increased input values.
This non-ideal phenomenon, called saturation, is due to the power supply voltage level and is dis
cussed in Section 4.
166
buffer -DC Transfer-2
(V)
+0 .0006+000
+2 .000
+4.000
+6.000
V2
+8.000
+10.000
+12.000
FIG U RE 4.11 Spice simulation of voltage-follower circuit.
Exercise. Find
for the circuit o f Figure 4.12, the power supplied by the source
power supplied to the 12 kQ load.

AN SW ERS:
=v,.0.
1 2 x 10-
FIG U RE 4.12 Isolation of load from source using buffer amplifier.
and the
167
3. TH E DESIGN OF G EN ERAL SUM M IN G AMPLIFIERS^

Often data acquisition equipment and active fdters require multi-input single-output amplifiers
having a more general summing characteristic, such as
ôut -
+ 2 'a 2 ) +
(4.8)
where the constants ay>0 and |3>0. The inverting and non-inverting amplifier configurations
(Examples 4.1, 4.2, and 4.3), as well as the difference amplifier configuration o f Example 4.4, are
special cases o f equation 4.8. W ith a little cleverness, it is possible to design by inspection an op
amp circuit whose input-output characteristic is precisely equation 4.8. The op amp circuit o f
Figure 4.13 having the four inputs
V^2 > ^hv ^bl accomplishes this. The circuit looks ordi
nary except for the presence o f one additional conductance, AG, incident on the inverting termi
nal o f the op amp. T he dashed lines in Figure 4.13 are present because this conductance may or
may not be needed. Computation o f the values o f AG and
are explained in design Step 2,
below.
b2 0 s / \ / V
G
FIG U RE 4.13 A general op amp circuit that realizes equation 4.8.
Design Choices for the General Summing Circuit o f Figure 4.13

The first two design steps constitute a preliminary or prototype design, meaning that the feedback
resistor is normalized to 1 Q , or equivalently, 1 S. After completing the prototype design, an engi
neer would scale the resistances to more practical values without changing the gain characteristics.
T he scaling procedure is explained in Step 3.
D esign Step 1. Prototype design. Set G ^ = l S , G , , = a , S ,G ,2 = 2 S, G^, = P i S , and CJ^2 = P2
S. For the design to remain simple, the total conductance incident on the inverting terminal must
equal the total conductance incident on the non-inverting terminal. This is achieved by proper
choice o f AG and/or G . T he proper choices are given in Step 2.
168
Design Step 2. Prototype design continued: Computation ofG^ andhr tS.G so that the total conduc
tance incident at the inverting terminal o f the op amp equals the total conductance incident at the non
inverting terminal.
To achieve this equality, recall that in design Step 1, Cy= 1 S,
S, and
S. Define
a numerical quantity
^ = (1 + a i +
) - (^1 +
The sign o f 6 leads to two cases:

Case 1: If 8 > 0, then set (7 = 8 and AG = 0.
.5
Case 2. If 8 < 0, set G to some value, for example, G = 1 S and AG = |8| + G .

o
&
&
Design Step 3. Scaling to achieve practical element values. Multiply all the resistances (divide all
conductances) incident at the inverting input terminal o f the op amp by a constant
Similarly,
multiply all resistances (divide all conductances) incident at the non-inverting terminal o f the op
amp by
It is permissible to choose
but this is not necessary.
EXA M PLE 4 .6 . Design an op amp circuit having the input-output relationship

>ô/ = - 7 V a l - 3 v ^ 2 + 2 v i , + 4 v^2
(4.9)
So l u t io n
Step 1. Prototype design. Using Figure 4.13, choose Gy- = 1 S, G^j = 7 S, G^2 = 3 S, G^, = 2 S, and
G^2 = 4 S .
Step 2. Equalization o f total conductances at inverting and non-inverting terminals. Since 8 = (1 +
7 + 3) - (2 + 4) = 5 > 0, set AG = 0 and G^ = 8 = 5 S. The circuit in Figure 4.14(a) exemplifies
the prototype design.
Step 3. Scaling. To have practical element values, let us choose
to a design with resistances Rjr= 100 kQ, 7?^, = 14.28 kQ,
25 kD and R = 20 kQ.
= 10^. This scaling leads

= 3 3.33 kQ,
= 50 kQ, R^^ =
169
14.28 kQ
V3, o------
lOOkO
25 kQ
(a)
(b)
FIGURE 4.14. (a) Prototype design of equation 4.9; (b) final design
after scaling with
= lO^.
EXA M PLE 4 .7 . Design an op amp circuit to have the input-output relationship:

I'ow/ = - 2 v i - 4 v^2 + 7 v/,| + 5 v^2
So l u t io n
Step 1. Prototype design. Again, using Figure 4.13, choose
(4.10)
1S,G =2S,G ,2 = 4S,G^,=7
S, and G^2 = 5 S.
Step 2. Equalization o f total conductances at inverting and non-inverting terminals. 8 = (1 + 2 + 4)
- (7 + 5) = - 5 <
0; set C = 1 S, AG = |8| + G = 5 + 1 = 6 S. This prototype design is given in
A
o
Figure 4.15(a).
Step 3. Scaling. To have practical element values, let us again choose
= 10^. This scal
ing leads to a design with resistances Rjr = 100 kQ, R^j = 50 k fl, R ^2 = 25 kQ,
= 14.28 kQ,
R^j = 20 kQ,
- 100 kQ, AR = 16.67 kQ. The final design is set forth in Figure 4.15(b).
16.67kn
50 kQ
V3, o------v X / X . -
lOOkQ
25 kQ
'â2 O------- s/\v^v^
14.28 kQ
lOOkQ
'^b2 O--20 kQ
(a)
(b)
FIGURE 4.15 (a) Prototype design of equation 4.10; (b) final design after scaling with
= 10^.
170
Exercise. 1. Obtain an alternative design for Example 4 .7 such that
= 0, implying the saving
o f one resistor.
AN SW ER: In prototype design, AG = 5 mho.
2. Design a difference amplifier so that
- v^-^, with
= 10 kD.
AN SW ER: See Figure 4.16.

10 kQ
At this point, the reader may wonder how this simple procedure is derived. The derivation o f this
procedure is beyond the scope o f the Hght edition^ Th e interested reader is directed to the 2nd
edition o f this text.
Exercise. 1. Find
AN SW ER:
the G- for the circuit in Figure 4.17(a).
in terms o f
_
' out -
G,
G,
Gj
G3
,v2
2. Find t',,
in terms o f for the circuit in Figure 4.17(b).
AN SW ER:
- 7z>2 +
6K U
(a)
FIG U RE 4 .17
171
4. SATURATION AND TH E A CTIVE REGION OF TH E OP AMP

In the previous sections, we assumed the op am functioned ideally:
= v_ and
= i_=Q. For the
inverting amplifier o f Example 4.1, this led to the very simple gain formula,
'ôut
_
Rh,
Thus, as the input voltage increases, the output voltage increases proportionately. For real circuits,
this proportional relationship holds only when
< V^^^for some value o f
that is associat
ed with the power supply voltage. Intuitively speaking, an op amp cannot generate an output volt
age beyond that o f its power supply voltage, typically less than or equal to 15 V. W hen the V^^limit is reached, further increases in the magnitude o f v-^ produce no change in the value o f
This behavior is called saturation.
To explain this saturation behavior, we refer to Figure 4 .1 8 . In Figure 4.18,/'(v^ - v j represents
a nonlinear controlled voltage source, as opposed to the linear relationship A(v^ - p_), shown in
Figure 4.4(a). However, because the op amp functions more or less linearly until reaching its sat
uration limits, we can approximate f(v^ - v_) by the three-segment piecewise linear relationship
shown in Figure 4.19(a), wherein the saturation effects are captured by segments II and III. One
observes that when
, the voltage/(f^ - z 'J clamps at
the voltage/(z'^ - v j clamps a

saturation occurs are
. If
and when
As observed, the critical threshold voltages o f t^ât which

= 15 V and A= 10^, the critical threshold voltages are 0.15
mV; if A is infinite, as in Figure 4.19(b), then saturation occurs when \v^ > 0.
FIG U RE 4.18 Practical op amp model with a nonlinear controlled voltage source.
The linear r e l a t i o n s h i p , - v^=A{v^ - v J , holds for segment 1 in Figure 4.19(a), which is said
to be the linear region or active region o f the op amp, denoted by
172
Typical values o f finite A range from 10'^ to 10*^. The active region is the ordinary region o f oper
ation. In the active region, the op amp provides a very high (open loop) voltage gain A, the slope
o f segment I. The phrase open loop gain means that there is no connection through a wire, a
resistor, or some other device back to the input terminals.
Models o f the three operating regions o f the op amp are summarized in Table 4.1.
A f(v^-v)
V
Positive
Saturation
/
Active /
Region
= V -V
Negative
Saturation
-V
(b)
FIG U RE 4.19 A piecewise linear (three-segment) curve for the op amp that specifies the active
and positive/negative saturation regions of operation: (a) finite gain A, and (b) (ideal) infinite gain A.
TABLE 4.1 Operating Regions o f the Op Amp with Associated Models
C urve Seg m ent
N ame
of
R e g io n D efining E quations
Vcl =
Active
f(Vd)
A
and
*^sar
II
Positive
saturation
Vsa,
A
and
sat
III
Negative
saturation
<and
I dealized C ircuit M od el
Chapter 4 T h e Operational Amphfier
173
The use o f a three-segment curve in Figure 4 .1 9 is different from the techniques o f earlier chap
ters. The operating point,
determines the proper segment to be used for analysis. If the
input is small, one reasonably assumes the operation is in the active region, segment I. However,
when the input magnitude is large, one must guess and check to determine the appropriate oper
ating region. For example, should the guess be incorrect, then the model for one o f the other
regions must be used and the analysis repeated until a valid solution (and operating region) is
obtained. The following example illustrates the approach.
E X A M P L E 4 .8 . The purpose o f this example is to illustrate that an op amp may operate in any
o f three regions and also to illustrate that the determination o f the region o f operation using the
guess and check method. Recall the inverting am plifier o f Figure 4.5. Suppose
= 50 kQ,
V; (b)
= 10 kQ and
= 4 V; and (c)
= 15 V. Find
and
is infinite,
for the following three cases: (a)
= 0.5
= - 5 V. Finally, verify the theoretical analysis using SPICE.
So l u t io n
(a) Assume the op amp operates in its active region. From equation 4.3 in Example 4.1, the out
put voltage is
vout
= - ^ v , m, = - ^JQx 0 . 5 = - 2 . 5 V
Since |-2.5| <
= 15 V the op amp operates in its active region; the answers
= - 2 .5 V and
v^ = 0 are valid.
(b) W ith v-^ = 4 V, assuming operadon in the active region,
However, since |-20| >

ôut
^
50
= -----= -------------- x 4 = - 2 0 V.
10
Rl
= 15 V, the op amp does not operate in its active region. Therefore,
invalid, but does suggest operation in the negative saturation region. The negative
saturation model o f Table 4.1 yields the circuit o f Figure 4 .20 in which
= - 1 5 V.
FIG U RE 4.2 0 Op amp operating in negative saturation region.

By writing and solving a single node equation at the inverting input terminal designated by the
minus sign in Figure 4.20, we obtain Vj = - 0 .8 3 V.
(c) With v-^ = - 5 V, assuming operation in the active region,
R'
50
= 25 V
This result suggests that the op amp is really operating in the positive saturation region. Using the
positive saturation model o f Table 4.1, Figure 4.21 shows the proper circuit configuration with
174
ôut ~ 15 V. As in the previous case, by writing and solving a single node equation at the invert
= 1.667 V. In this case and in case
ing terminal designated by the minus sign in Figure 4 .2 1 ,

(b) above,
0, as we were not in the active region o f operation, and it was necessary to change
the guessed region o f operation to obtain valid results.
FIG U RE 4.21 Op amp operating in positive saturation region.

A SPIC E simulation was used to validate the theoretical analysis^. A D C sweep, - 4 <
< 4 V,
is an adequate input to demonstrate the saturation effects. In the SPIC E simulation, an accurate
model for a 741 op amp manufactured by Burr Brown was used. T he resulting dc transfer curve
is shown below in Figure 4.22.
Lin/Decarlo E xI-D C Transfer-4
Output voltage
(V)
-4.000
- 1.000
+ 0 .0006+000
Vin
+
1.000
+3.000
+4.000
V(IVM)
From this curve, one can see that the op amp saturates for input voltages v-^ such that
> 3,
and the op amp operates in its linear region whenever \v-^ < 3. As hoped, the simplified three-seg
ment model in Table 4.1 yields very good results in all regions o f operation relative to the realis
tic SPIC E simulation.
O ne can conclude from the above example that for the purpose o f faithfully amplifying an input
signal, the input should not be so large as to drive the op amp into saturation. Driving an op amp
into saturation distorts the output signal relative to the input. O n the other hand, for some spe
cial applications, such as the com parator, saturation is precisely the property to be utilized. Figure
17 5
4.23 shows two comparator circuits. A com parator circuit compares tlie input voltage v-^ with a
reference voltage Vygjr{or some multiple o f
Only two different output voltages are produced,
and the other for v-^ <
one for
EXA M PLE 4 .9 . For the com parator circuits shown in Figure 4.23, each op amp has infinite gain
and a saturation voltage
= 1 5 V"^:
v -^ relationship for the comparator o f Figure 4.23(a).
(a)
Find the
(b)
Repeat part (a) for Figure 4.23(b).

Note that in both circuits, there is no connection between the output and inverting input
terminals, a departure from all the previous circuit configurations. Because o f this, for
^ v_, and
almost all voltages,
=
Vin O------s / W 2 0 kQ
''re f O --------
= -2 0 V
80 kn
1 .
1 .
(a)
(b)
FIG U RE 4 .23 Two comparator circuits that are used to determine when an input
voltage is above or below a reference voltage.
S
o l u t io n
(a) For
> - 5 V, the voltage v^=
v_ = -5 - v-^ < 0. Referring to Figure 4.19(b),
ôut ^ ~^sat = - 1 5 V. Similarly, for v-^ < - 5 V, the voltage v^ = v^ - v_ = -5 - v-^ > 0, and hence
ôut = ^sat = 15 V.
(b) By the fact that no current flows into the input terminals o f Figure 4.23(b), using nodal analy
sis, we have that
V
- V
20x10^
80x10^
m which case,
- 4 + 0.8v,
For
> - 5 V, the voltage v^ =
ration curve o f Figure 4.19(b),
- v_ = 0 - {-A +
Here, referring again to the satu
= - 1 5 V. Similarly, when v-^ < 5 V, the voltage
v^ = v ^ -v _ = Q -{-A + 0.8V.J > 0; hence,
= 15 V.
176
To verify this analysis, the circuit o f Figure 4.23(b) was simulated in B2Spice using a Burr Brown
741 op amp model. T he results o f the simulation are given in Figure 4.24. The theoretical analy
sis based on the simplified models o f Table 4 .1 shows a very good match with the more realistic
SPIC E simulation results.
example4.8-DC Transfer-6
Output voltage
(V )
-1 0 0
+2.000
+3.000
V(IVM)
FIG U RE 4.2 4 B2Spice simulation o f the comparator circuit of Figure 4.23(b).
Exercise. For the circuit o f Figure 4.25, suppose
= 12 V, find the range o f v^2 for which the op
amp is in positive saturation. Then find the range o f v^ 2 fof negative saturation.
AN SW ER:
when
< -^ V, and
=-
when
V , C3------- v N / \ - -
75 kO
''s2 o-----s /s y \/-
-o
+
25 kQ
1 .
FIG U RE 4.25
177
5. SUM M ARY
This chapter has introduced the operational amplifier and a number o f practical circuits that uti
lize this new device. These circuits include the inverting and non-inverting amplifiers, the buffer
amplifier, the difference amplifier, and the general summing amplifier. W ith regard to the gener
al summing amplifier, a simple design algorithm is described and exemplified. The analysis o f
these circuits builds on the definition o f an ideal op amp, meaning that, when properly config
ured, no current enters the input terminals and the voltage across the input terminals is zero; these
properties are referred to as the virtual short circuit model o f the op amp, i.e., the ideal op amp
has infinite input resistance, zero output resistance, and an infinite internal gain, A. (See equations
4.1 and 4.2.) Practically speaking, the gain A, is not infinite, but ranges between lO'^ and 10*^.
After exploring properties o f the ideal op amp, we discussed the phenomena o f output voltage sat
uration. By introducing output saturation, the ideal model o f the op amp gives way to a more real
istic one, characterized by three regions o f operation, each having its own ideal model, as set
forth in Table 4.1. In practical design and applications, output saturation is either to be avoided
or utilized to some advantage, as in the case o f the comparator circuit studied in Example 4.8. For
a faithful amplification o f an input signal, saturation is to be avoided.

Active element: A circuit element that requires an outside power supply for proper operation and
has the capability o f delivering net power to a circuit such as is the case for an op amp or
negative resistance.
Buffer: A circuit designed to prevent the loading effect in a multistage amplifier. It isolates two
successive amplifier stages. Characteristics o f an ideal buffer are infinite input impedance,
zero output impedance, and constant voltage gain.
Com parator: an op amp circuit that compares the input voltage
(or some multiple o f
with a reference voltage
only two different output voltages are produced, one for
<
v-^, and the other for v-^ < v-^.

Difference amplifier: given two inputs,
and
a difference amplifier produces the output
ôut ^
appropriate constant k, often taken as 1.
General summing amplifier: an op amp circuit having the input-output relationship
constant a - and p ..
Ideal op amp: An operational amplifier with infinite input resistance and infinite open-loop gain.
-
+ ^ n ^ a r) +
+ - + ^ m ^ bn )
p o s itiy c
Inverting amplifier: An operational amplifier connected to provide a negative voltage gain at dc.
Linear active region: In the op amp output vs. input transfer characteristic, the region where the
curve is essentially a straight line through the origin is called the linear active region.
Non-inverting amplifier: An operational amplifier connected to provide a positive voltage gain
at dc.
Open-loop gain: The ratio o f the output voltage (loaded, but without any feedback connection)
to the voltage across the two input terminals o f an op amp. The slope, p, o f the straight
line in the active region o f an op amp is the open loop gain under no load condition.
When a load
is present, the open loop gain is reduced to
the output resistance o f the op amp.
+ R^, where R^ is
178
Chapter 4 The Operational Amplifier
Operational amplifier (abbreviated op amp): A multi-stage amplifier with very high voltage
gain (exceeding 10"^) used as a single circuit element.
Passive elements: a circuit element that cannot deliver net power to a circuit such as a resistor.
Saturation regions: In the op amp output vs. input transfer characteristic, the region where the
curve is essentially a horizontal line is called the saturation region. There are two such
regions: one for positive input voltage, and the other for negative input voltage.
SPICE: Acronym for Simulation Program with Integrated Circuit Emphasis. It is a very sophisti
cated software tool for simulating electronic circuit behavior.
Virtual ground: When an ideal op amp has one of its input terminals grounded, and is operat
ing in the active region, then the other input terminal is also held at the ground potential because of the virtual short effect (see below). Such a condition is called a virtual
ground (in contrast to a physical ground).
Virtual short circuit: When an ideal op amp is operating in the active region, the voltage across
the two input terminals is zero, even though the two terminals are not hard-wired togeth
er. Such a condition is called a virtual short circuit (in contrast to a physical short circuit).
Voltage follower: A voltage-controlled voltage source with gain equal to 1, often utilized to sep
arate stages of amplification in a multi-stage amplifier device.
^The circuit proposed in this section is a modification of one proposed in W. J. Kerwin, L. P. Huesman, and R.
W. Newcomb, State-Variable Synthesis for Insensitive Integrated Circuit Transfer Functions, IEEE Jr.
of Solid State Circuits, Vol. SC-2, pp. 87-92, Sept. 1967. The modification consists of an additional
resistor, which greatly simplifies the design calculations and was published by P. M. Lin as Simple
Design Procedure for a General Summer, Electron. Eng., vol. 57, no. 708, pp. 37-38, Dec. 1985.
'
^ See Linear Circuit Analysis by DeCarlo and Lin, 2nd edition, New York: Oxford University Press, 2002.
^ Any of the SPICE or PSPICE software programs available by a variety of vendors will suffice to obtain the indi
cated curve.
^ An op amp and a comparator as seen in a parts catalog are essentially the same, except that the comparator device
has a modified output stage that makes it compatible with digital circuits.
179
Problems
ANALYSIS USING IDEAL OP
AM P M O D EL
1. Consider the inverting amplifier circuit o f
Figure P 4 .1, in which
= 4 V.
(a)
If
= 2 kD, find i ?2
dehvered to
(b)
that the power
= 100
Figure 4.3
is 4 W.
Now suppose T(2 = 12 kQ. Find
so
that the power delivered to Rj^ = 2 Id l

is 450 mW. Then find the power con
sumed in i?, and Rj-
4. In the circuits o f Figure P4.4, a source is rep

resented by an ideal voltage source, v-J^t) = 4 V,
in series with R^ = 10 Q resistor. The loadin
both cases is
(a)
= 40 Q.
W ith the load connected directly to

the source, as shown in Figure P4.4(a),
find the load voltage, the load current,
the source current, and the power
delivered to the load.
(b)
Figure P 4.1
Check: (b) 1500 <R^< 2000, and
As in Figure 4.4(b), a buffer amplifier

separates the source and the load.
= 10 m W
Again, find the load voltage, the load

current, the source current, and the
power delivered by the op amp to the
2. Consider the non-inverting circuit o f Figure

P 4.2, in which v- = 4 V.
(a)
If R^ = 2 kQ, find 7?2 so that the

power delivered to
= 100 Q is 4 W.
(b)
Now suppose i ?2 = 13 kO. Find R-^ so

that the power delivered to
load.
+
Rs
= 2 kH
' l >r
is 450 mW. Then find the power con-
sumed in R-^ and i?2(a)
(b)
Figure P4.4
SCRAMBLED ANSWERS: (a) 0.256, 3.2, 0.08
5. Figure P4.5 contains three circuits that
explore loading and the elimination o f loading
effects using either a dependent source or an
equivalent buffering op amp circuit.
(a)
For the circuit o f Figure P4.5(a), com
pute
and
in terms o f v^. Observe
Figure P4.2
Check: (b) 1500 <
< 3000
3. For the circuit o f Figure P4.3, find the volt
age gains,
G,
andG , =
^ Vin
in terms o f the literal resistor values.
that the 80-Q -240 Q resistor combina

tion loads down the 3200 resistor.
(b)
For the circuit o f Figure P4.5(b), com

pute
and
in terms o f v^. Notice
that
is different from the answer
V,
180
computed in part (a) because the 80-
A N SW ER: (a) R, = 5 kQ; (b) 6 .76 m W
2-240 Q resistor combination is iso

lated from the 320-Q. resistor.
(c)
For the circuit o f Figiire P4.5(c), again

compute v-^ and
7. (a)
in terms o f v^. Your
answers should be the same as those in

part (b). The buffering op amp circuit
again isolates the 80-l-240 Q. resistor
comination from the 320 Q resistor.
<,v,
320 n<
Repeat part (a) for the circuit o f Figure

P4.7(b).
(c)
If for Figure 4.7(b),

3 kn ,
12 kD,
= 4 kQ, Rj = 1 k2,
=
= 1.5
V, and v^ 2 = 2 V, find the power deliv

ered to the load R^= 100 Q.
-o
+
V r + ^
as a function
and the R^.

(b)
80 n
80 n
For the op amp circuit o f Figure

P4.7(a), find
AN SW ER: (c) 0.04 watts
240
(a)
80 n
son
-o
+
240 n
320 n<
(b)
80 0
son
-v \ ^ -
-o
+
240 n
320 n
Figure P4.5
AN SW ER: (a)-V /
3
0 5V
6. In the circuit below, R^= 10 kQ.

(a)
Find R^ and R^ so that
(b)
Given correct answers to part (a), sup
= -2t^^j 8. (a)
">â-
Find the power delivered to the load if
= 20. If
= 600 mV.
= 0.6, find the power deliv
ered to the 8 -Q load.

(b)
Now suppose R^^'^ kQ, and find R^
so that
(c)
out
20
Finally, suppose Rj =
and find
their common value so that

F igu re P 4 .6
out
= 6 kQ, and find R^ so that
pose a 1 kO resistor is attached as a load.
= 200 mV and
For the circuit o f Figure P 4.8, suppose
out
20
181
A N SW ER: (a) 40 kl; (b) /?, = 15 k2

2kn
-O-
11. Consider the circuit shown in Figure

P 4 .l l.
-0 -,
sn
and v^2 -
(a)
Find
(b)
If
= 250 m V and v^ 2 - 500 mV,
find the power delivered to the 1 kQ
in terms o f
load resistance.
Figure P4.8
A N SW ERS (in random order): 14 kO, 8 kf2,
10 k n , 18
lOkn
lokn
lokn
5kn
2kQ
9. In the circuit below, Rr= 12 ki2 and ^3 > 1 k2.

(a)
Find
and ^2 so that
1 kn<
= -lOr^^j -
'
20^,2(b)
Given correct answers to part (a), find

the power delivered to the load if
- 2 0 0 m V and
Figure P 4.11
C H EC K : (b) Pj^ = 12.25 m W
12.
(a)
For
the
circuit
o f Figure
P4.12a, the input voltage
= 2 V and
the input voltage v^2 = 3 V. Find

and
the power delivered to the 1 kI2 load
1 kn
resistor.
(b)
Repeat part (a), when
= 4 V and
the input voltage v^ 2 = 2 V.

C H EC K : (b)
Figure P4.9
= 0.1 watt
(c)
Reconsider part (b). Find the mini

mum value o f R so that the maximum
amount o f power consumed in either
R-ohm input resistors is 2 mW.

(a)
Find the value o f R =

= R2
^ât
the power delivered to R^= 1.25 kO is
0.5 watt when
(b)
3R
= 1 V.
Suppose 6R^ = i ?2
v^ = 2Y . Find
and R2 so that the power delivered
to
= 1.25 k il is 2 watts.
Figure P4.12
13. (a)
For the circuit o f Figure P4.13(a), the

input voltage
= 1 V, and the input
voltage v^2 ~ 500 mV. Find /?, in terms
Figure P 4.10
o f R so that
= 10 V.
182
(b)

P 4.13(b), given that
= 2.5 V.
12R
r e
R,
-V S/V
2R
5R
-O
+
3R _ d
(a)
Figure P 4 .15
C H EC K : (a)
12R
2R
rO
5R
16. For the circuit o f Figure P 4.16, find

-o
+
terms o f
in
V, v^2
2.5R
(b)
Figure P4.13
AN SW ERS; (a) 6R-, (b) 3R
14. For the circuit in Figure P4.14, the input
voltages are
= 2 V,
= '1-5 V, and
=2
V.
(a)
Find
(b)
If /? = 10 kQ, find the power delivered
17. For the circuit o f Figure P 4.17, find
by each o f the operational amplifiers.

4R
2R
0.5 R
+ Av^2 -
1.5R
-o+
0.75R
0 - .
in terms o f R so that
and
Figure P4.14
C H EC K :
i = 0.9 mW,
= 0-20667
watts
15. For the circuit o f Figure P4.15, the input volt
age
= 5 V, and the input voltage
(a)
(b)
If
= 8 R and Rj - Ry find
If R^ = 8R and R^ = ARy find
(c)
If /?2 =
that
find
= lOV.
in terms o f R so
CH ECK:
Figure 4.17
= 8R
18. For the circuit o f Figure P 4.18, find R^ and

i?2> and Rj in terms o f R so that
= 8f^j +
10v^2~2^s3- Hint: Consider Problem 17 first.
183
Time in s
AN SW ERS:
= AR, R^ =
(0
R, = O m
Figure P4.19
NON -IDEAL OP
AM P-SATURATIO N EFFECTS
19. The op amp in Figure 4.19(a) has
V,
= 4 ld2, and /?2 = 20 kQ.

(a)
Plot the
versus
for
20. Repeat Problem 19 for the op amp circuit of

Figure P4.20 when
= 4 ld2, and R2 = 20 kfl.
=15
-O+
given in
Figure 4.19(b).
for 0 < /< 6 s for v(J) in
(b)
Plot
(c)
Verify your analysis in part (a) using
Figure 4.19(c).
Figure P4.20
SPICE. Assume that the op amp is a

type 741 whose model should be
available within your SPIC E program.
21. For the circuit o f Figure P4.21, each amplifiersaturates

at
= 15 V.
sat
(a)
Suppose the input voltage
= 5 Y
and the input voltage v^2 = 2.5 V.
R,
Find
-o (b)
and
If
= -2 -5 V and
= 15 V, find
so that no amplifier saturates.
(a)
4R
1.5R
2R
-O
Figure P4.21
AN SW ER: (a) 15 V; (b)
= 3.75 V
22. For the circuits o f Figure P 4.22, suppose R^
(b)
= 40 k n , and R^ = 120 Q.
(a)
For the circuit o f Figure P4.22(a),

compute
the power delivered by
the source, and power delivered to the
load R^ in terms o f v-^.
184
(b)
For the circuit o f Figure P4.22(b),

compute
(b)
compute

load
(c)

in terms o f v-^^.
load
Discuss the differences in your solu
(c)
in terms o f
tions to (a) and (b). Specifically, dis
cuss the effect o f using a voltage fol
lower to isolate portions o f the circuit.
(a)
(a)
(b)
Figure P4.24
25. For the circuits o f Figure P4.25, suppose /?, =

20 Q, and /?2 = 160 Q, R^ = 40 Q., and R^ = 120
a.
(a)

compute

Figure P4.22
23. In the circuits o f Figure P4.22, all resist
ances are 100 Q, and
= 1 V.
(a)
(b)
the source current again when all

resistances are 100 Q.
I f a buffer amplifier separates the
source and the load, as in Figure

compute
For the circuit o f Figure P4.22(a), find

the load voltage, the load current, and
(b)
load Rj^ in terms o f v-^.
load R^ in terms o f v-^.

(c)

P4.22(b), find the source current, the

load voltage, the load current, and the
current supplied by the op amp.
AN SW ERS: (a) 0.5V, 5 mA, 5 mA; (b) 0 A, 1
V, 10 mA, 10 mA
24. For the circuits o f Figure P4.24, suppose
= 40 Q , and Rj = Rl = 120 Q.
(a)

compute
load Rj^ in terms o f v-^.
Figure P4.25
AN SW ERS: (a)
3.7037 x l O 'V .
185
= 0.003i^y; (b) Pl =
26. Figure P 4.26 contains three circuits that

explore loading and the elimination o f loading
effects using either a dependent source or an
equivalent buffering op amp circuit.
(a)
27. Two non-ideal voltage sources are each rep

resented by a connection o f a (grounded) inde
pendent voltage source and a series resistor.
Denote the parameters o f each connection by
{Vs^,
^st)' Design an op amp cir
cuit such that the output voltage with respect

to ground is

compute I'j and
Observe that the
8-Q -24 Q. resistor combination loads

down the 3 2 -Q resistor.
(b)
for all values o f

and R^2 tie greater
than or equal to 100 kS2 so that only small
compute
is
amounts o f current are drawn from the buffer
different from the answer computed
amplifiers. Note that the general difference
in part (a) because the 8-Q -24 Q resis
amplifier circuit o f the chapter will not work
tor combination is isolated from the
here because o f the presence o f the resistances
and
Notice that
3 2 -Q resistor.
(c)
and R^2 - To achieve such a design, it is nec
For the circuit o f Figure P4.26(c),
essary to isolate the (practical) sources from the
again compute Pj and
Your
difference amplifier inputs using buffer ampli
answers should be the same as those in
fiers, as shown in Figure P 4.27. Explore your
part (b). The buffering op amp circuit
design for various values o f
again isolates the 8-Q -24 O, resistor
SPICE. Do the SPIC E simulations verify that
combination from the 32 Q resistor.
the output is independent o f the values o f
and R^2 using
Figure P4.27
8Q
80
-o
+
circuit o f Figure P 4.28, suppose the op amp has

24 n
32 0 <
(c)
Figure P4.26
AN SW ERS: (a) 0.6665 V^, 0.5 V^; (b) and (c)
0.8 V^, 0.6
28. Following Example 4.9, for the comparator

infinite gain and a saturation voltage
=15
V. Find the
versus
relationship and plot
as a function o f v-^. Verify your analysis using
SPICE. Assume that the op amp is a type 741
whose model should be available within your
SPIC E program.
186
Check your design using SPICE. Hint: How
80 kO
can the circuit o f Figure P4.29 be modified to

achieve the correct polarities?
/?] _
A N SW ER:
32. Find the
comparator
Figure P4.28
29. (a)
Find the
1
1.5
versus v-^ relationship for the

circuit
of
Figure
P 4.32.
Specifically, show that when
versus
relationship for
Vin >
, then
and when
the comparator circuit o f Figure

P4.29. Specifically, show that when
Vin <
R,
and when v-^ > ??, then
then
V
= - V^sa f
'ôut
(b)
în<-^^ref , then
Now suppose
^ref ~
characteristic if
= R2 = 100 kQ and
ôut versus v-^
= 15 V. Verify
Figure P4.32
your analysis using SPICE. Assume

that the op amp is a type 741 whose
model should be available within your
SP IC E program.
G EN ERA L SUM M IN G
AM PLIFIER (ID EA L OP AM P
M O D EL)
3 3 .(a)
Assuming the op amp in Figure P4.33

is ideal, derive the relationship
Figure P4.29
AN SW ERS: (b)
= 15 V if
< 2 V, and
(b)
Rf
Rf
R,
Ro
Suppose
so that
25 kO. Find
, i?2>
is the negative o f the
average o f
, V2 , and Vy
30. Using a 1.5 V battery, an op amp with

= 10 V, and some resistors, design a comparator
circuit such that
and
= 10 V when
< 1 V,
= - 1 0 V when v-^ > 1 V. Check your
design using SPIC E. Use part (a) o f Problem 29

as a guide.
31. Using a 1.5-V battery, an op amp with
= 10 V, and some resistors, design a comparator
circuit such that
= 10 V when v-^ >
1 V, and
= - 1 0 V when v-^ < I V.
V l-
Figure P4.33
AN SW ER: (b) R. = 75 kO
187
34. Using the topology o f Figure 4.13, design

an op amp circuit to have the input-output

relationship
relationship
+H i
- 3^.2 +
Two different designs are to be produced for
+H i
comparison and selection:

(a)
Design 1: Rjr= 100 kQ.
(b)
1:
. = 50 kQ.
iqq
Design 2: Rj^= 50 kQ. Specify all final
values in terms o f Q..

C H ECK :
= 25 kO, and R^2fiil ^ 33.33 k2
^
39. Generalizing the topology o f Figure 4.13,
design an op amp circuit to have the input-output relationship

ôut ~â\ ~ '^'âl ~

I .he final circm, Rf - 40 k a
" ' 'P
ôut ~ ~^â\ ~ '^âl
^^h\
^'^bl

40. Generalizing the topology o f Figure 4.13,

design an op amp circuit to have the input-output relationship
out
(a)
(b)
Design 1: Rr= 100 kl.

Design 2: R^= 50 k tl. Specify all final
values in terms o f Q.
= -4v^j + 2z;^j +
n
/ r^
^/=
VARIABLE GAIN AM PLIFIERS

4 1
x h e circuit o f Figure P4.41 is a modificanon o f the basic non-inverting amplifier. In the
relationship
modification, a potentiometer R^ is connected

between the output terminal and Rq, with the
ôut ^
~ '^'âl ^
^'^bl
Your design must have Rj-= 10 kX2 in the final
sliding contact between points A and B, as

shown. Show that as the sliding contact o f the
potentiometer is moved between positions A
circuit, and all other resistors should be within
and B, the range o f voltage gain achievable is
the range 2 kQ to 20 kQ.
ôut

relationship
ôut =
- ">âl + ^^b\ +
hi
Your design must have all resistors, including
Rjr, in the range 5 kQ to 25 kQ.
Figure P4.41 Variable gain non-inverting amplifier.
188
42. The circuit o f Figure P4.42 is a simple

modification o f the basic inverting amplifier
circuit in which a potentiometer is connected
to the feedback resistor Rp as shown. Show that
the range o f gains achievable by this circuit is
h .
SIM ULATION OF
C O N TR O LLED SO URCES
USING OP AM PS
44. Design an op amp circuit to simulate the
grounded VCVS in Figure P 4.44 when p > 1.
Hint: Consider the non-inverting amplifier o f
Example 4.3.
Figure P4.44 Grounded VCVS.

grounded V CV S in Figure P4.44 for any p > 0.
Hint: Try a voltage follower in cascade with two
inverting op amp circuits.
Figure P4.42
43. T he circuit o f Figure P4.43 is another mod
ification o f the basic inverting amplifier to
obtain a variable gain amplifier. Show that as
the sliding contact o f the potentiometer is
moved between the two extreme positions, the
range o f achievable voltage gain is
R
U a = 1, H
where
:-a -
H -
Rq
Rf
46. Reconsider the design o f Problem 45 so

that only two op amps are used. In this case,
one still needs the voltage follower. Why? Hint:
Consider using a voltage divider followed by a
non-inverting amplifier circuit.
R,
Rp
R0n // /Ii\f
Rf
Hint: Apply KCL to the non-inverting input

terminal, and make use o f the virtual ground
property o f an ideal op amp.

grounded VCVS in Figure P 4.44 when p < 0.
Hint: Consider an inverting amplifier configu
ration in conjunction with a buffer amplifier.
48. For the circuit o f Figure P4.48, show that
the load current
equals V/R^, which is inde
pendent o f the load resistance R^. Hence, this

op amp circuit converts a grounded voltage
source into a floating current source. (This is
sometimes called a voltage-to-current con
verter.)
Figure P4.48 Op amp circuit simulating a

floating current source.
49. In Problem 4.48, since 7^ depends on Vand
only, the load need not be a resistor. For
example, Rj^ may be replaced by an LED (lightemitting diode), as shown in Figure P4.49.
Then by turning the knob o f the 10-kf2 poten
tiometer, one can control the brightness o f the
LED. The current through the load is supplied
by the op amp. The potentiometer, which con
trols the brightness o f the LED , uses a low-voltage part o f the circuit. Find the magnitude o f
the LED current if the potentiometer is set at
(a)
= 5 k^ and (b) 7?, = 8 k tl.
A N SW ERS: 1.32 niA, 2.1 niA

10 kn
50. This problem is a variation o f Problem 4.49

in which the load current flows in the opposite
direction. For the circuit o f Figure P 4.50, show
that the load current
equals v^JR^, which is
independent o f the load resistance R^. Hence,

this op amp circuit converts a grounded voltage
source into a floating current source in which
the current enters the op amp output terminal.
(This is sometimes called a voltage-to-current
converter.)
Figure P4.50 Op amp circuit simulating a float

ing current source.
189
Linearity, Superposition, and

Source Transformation
H ISTO RICAL NOTE
In the mid-nineteenth century, before the introduction o f the alternating current (ac), electricity
was available mainly as direct current (dc). This time period saw the evolution o f basic laws for
the analysis o f electrical circuits composed o f dc voltage sources and resistors: Ohms law, KVL,
and KCL. Application o f these laws to the analysis o f circuits led to the development o f the mesh
and nodal techniques requiring the solution o f simultaneous equations. Before the computer age,
manual solution o f a (large) set o f equations was very difficult. To circumvent this difficulty,
researchers developed a number o f network theorems that (i) simplified the aforementioned man
ual analysis, (ii) reduced the need for repeated solution o f the same set o f equations, and (iii) pro
vided insight into the behavior o f circuits. These network theorems remain useful even in the pres
ent day o f high-powered computing.
CH APTER O BjEC TIV ES

1.
2.
3.
Introduce and apply the property o f linearity.

State and explore the two consequences o f linearity called superposition and
proportionality to simplify response computation.
Use superposition and proportionality to simplify manual analysis and to gain better
4.
insight into circuit behavior.

Introduce and apply the source transformation theorem to again simplify manual analysis.
SECTIO N HEADIN GS
1.
2.
3.
4.
5.
6.
Introduction
Linearity
Linearity Revisited: Superposition and Proportionality
Source Transformations
Equivalent Networks
Summary
192
Chapter 5 * Linearity, Superposition, and Source Transformation
7.
8.
Terms and Concepts

Problems
Chapter 3 covered nodal and loop/mesh analyses. Node voltage or loop current calculation pro
ceeds by constructing a set o f simultaneous node or loop equations and solving them by hand, by
MATLAB, or with some equivalent software package. Few o f us will attempt a paper-and-pencil
solution o f four equations in four unknowns. Yet, MATLAB, Mathematica, or some other com
putational software program, can easily and reliably crunch numbers, relieving us o f tedious hand
calculations. Nevertheless, manual analysis in some form remains important for a deeper under
standing or insight into a circuits behavior, as well as a way to check the validity o f a program
output.
Experience teaches us that manual analysis is ordinarily practical only for small circuits.
Fortunately, the network theorems studied in this chapter and the next can often reduce seemingly
complex circuits to simpler ones amenable to manual analysis. They also provide shortcuts for
computing outputs and allow us to obtain deeper insights into a circuits behavior.
This chapter talks about linearity and superposition, which are motivated by the following ques
tions: What is the effect on the circuit output (voltage or current) o f a single independent voltage
source, say
acting alone. Acting alone means that the independent source, Vj^, has a nonzero
value, while all other independent sources are set to zero. A deactivated voltage source acts as short
circuit (see Chapter 2), and a deactivated current source acts as an open circuit (again, see Chapter
2). Is there a shortcut to computing the response if Vj^ is doubled in value?
To answer the above questions and others, our discussion begins with the important property o f
linearity. Linearity relates the values o f independent sources to a circuit output with a very com
pact equation. This equation defines the effect o f any independent source on a circuit output.
After studying linearity, we discuss two special consequences called superposition and propor
tionality. Each o f these concepts helps reduce manual computation o f responses, and each provides
insight into circuit behavior. Next, we state the source transformation theorem and show how this
method can reduce a complex circuit to a more simple form. Finally, we set forth the notion o f an
equivalent two-terminal network and then outline a proof o f the source transformation theorem.
193
2. LIN EARITY
This section investigates the circuit property o f linearity, which we introduce with a motivating
example.
E X A M P L E 5 .1 . For the circuit o f Figure 5.1, find the outputs
current /^j, and the source voltage V^2 -
and Vg in terms o f the source
will derive the relationships Vg = 40/^j + ^ 1^2 and
jgQ .sz
I.
60 0
+
V
120Q
FIG U RE 5.1. Resistive circuit driven by current and voltage sources.
So l u t io n
Step 1. Find the voltage Vg. A node equation at the top o f the current source is
Vb , V B - V s l ^ r
120
60
Solving for Vg yields
= 4 0 7 ,1 + - V ,2
Here, Vg appears as a constant times /^j, plus another constant times K^2>^ so-called linear com
bination.
Step 2. Find the current I
From Step 1, we know Vg. The current
40 / , i + - V , 2 - V ,2
Similar to Step 1, the output current
linear combination.
satisfies
180 '
is a constant times /^j plus another constant times V^2>^
194
Chapter 5 ' Linearity, Superposition, and Source Transformation
Exercise. 1. In Example 5.1, suppose the 60 Q resistor is changed to 120 Q. Find the outputs
and Vg in terms o f the sources, /^j and V^2 AN SW ER: Vg = 60/ ,, + 0 .5 V ,2 and
= 0 .5 / ,, -
2. For the circuit o f Figure 5.2, find Vg in terms o f
and V^2-
--- ---------- ---- --------120 Q +

60 Q
> 120Q
(j
V..
FIG U RE 5.2 Resistive circuit for Exercise 2.

A N SW ER:
= 0.251/, + 0 .5 V^2
In the above example and exercises, the desired output voltage or current was a so-called linear
combination o f the independent source values. This is, in fact, a quite general phenomena, as indi
cated by the linearity theorem below.
LIN EA R ITY TH EO R EM
For all practical linear resistive circuits, as per Figure 5.3, any output voltage,
or any cur
rent, ig, can be related linearly to the independent source values, as in the following equa
tions:
= î^s\ + - +
^nfsm
(5.1a)
= l Kl + - +
+ + ^mhm
(5.1b)
or
where the a - and
are properly dimensioned constants.
Chapter 5 Linearity, Superposition, and Source Transformation
195
Linear Circuit
containing
no
independent
sources.
FIG U RE 5.3. A linear circuit driven by n independent voltage sources and m independent current
sources with outputs of
and
A rigorous proof o f the linearity theorem entails solving a set o f modified nodal or loop equations
using matrix algebra and is beyond the scope o f this text.
EXAM PLE 5.2. For the circuit o f Figure 5.4, our objective in this example is to express
ear combination o f /^j, Iq, and
lin
as per equation 5.1a. In doing this, we review nodal analysis.

g .v .
)v .
FIG U RE 5.4
So l u t io n
Step 1. Write nodal equation at A. For node A,
(5 .2 )
196
Step 2. Write nodal equation at output node. At the output node,
or equivalently,
(5.3)
Step 3. Write equations 5.2 and 5.3 in matrix form. In matrix form, the nodal equations are
G ^ -g^
G
8m
2+
/.I
(5.4)
J s l + G ^ V ,,
Vou t,
Step 4 . Solve equation 5.4. Solving equation 5.4 for
and
yields
-1
-V
Vou t,
'G l- g m
8m
G 2
h i
G2 + G 3
{G :-gJ(G 2+ G ,)
-8 m
s\
G, - I
fs2+G3V,3
It follows that
Vout--
8m
(G i-g ,)(G 2 + G 3 )
hl+-
Gi-g.
(G ,-g ,)(G
jGl-8m)G3
2+
G3)
(G i-g ,)(G
2+
-V.s3
3)
(5 3 )
as set forth in equation 5.1(a).
Exercise. 1. In equation 5.5, suppose G j = 1 S, G2 = 2 S, G3 = 3 S, and^^ = 5 S. Find the numer

ical expression for
A N SW ER:
= 0.25/,, + 0.27,2 + 0-6 ^^,3
2. Suppose the dependent current source in Figure 5.4 is changed from^^V^

the new expression for
if G j = 1 S, G2 = 2 S, G3 = 3 S, and
Smôuf Compute
= 5 S.
AN SW ER: V ;,= 0.17,2+ 0 .31/3

3. Suppose the dependent current source in Figure 5.4 is changed from
Compute the new expression for
AN SW ER:
if G j = 1 S, G2 = 2 S, G3 = 3 S, and
to
+ ^0 ^)-
= 5 S.
^ ^.'i + ^ h i + J ^v3
E X A M P L E 5 .3 . A linear resistive circuit has two inputs

and z,2 with output
as shown in
Figure 5.5. Rows 1 and 2 o f Table 5.1 list the results o f two sets o f measurements taken in a lab
oratory. The measurements are taken in a practical way by first setting the value o f the current
197
ch a p te r 5 " Linearity, Superposition, and Source Transformation
source to zero, i.e., i^ 2 = 0
exciting
with a dc power supply set to 5 V; then the voltage
source is removed and replaced by a short circuit using a jumper cable, i.e.,
rent source is excited by a power supply producing a constant current o f
(a)
Derive the linear relationship
(b)
Find
when
= 0, and the cur

= 0-2 A.
+ 50/^2 using the data in Table 5.1.
= 10 V and ^'.2 = 0.5 A, i.e., complete the third row o f Table 5.1.
FIG U RE 5.5. Linear resistive circuit driven by two sources.

TABLE 5.1. Two Sets of Measurements o f a Linear Circuit in which
One is Allowed to Set Each Source Value to 0
i^ 2 (amperes)
Vout (volts)
0.2
10
10
0.5
(volts)
So l u t io n
From the linearity equation 5.1(a),
ôut =
+hhl
for appropriate ttj and ^2 - From the data in rows 1 and 2 o f Table 5.1,
4 = a , x 5 + P2><0 = 5a, ^
U j = 0.8
and
10 = a , X 0 + p2 X 0.2 = O.2 P 2 => P 2 = 50
in which case,
+ 5 0 i,2
So from row 3 o f Table 5.1, if i',] = 10 V and z'^2 = 0.5 A, we have that
0 .8 X 10 + 5 0 X 0 .5 = 33 V.
(5.7)
198
Exercise. 1. For Example 5.4, suppose
= 50 V and i^ 2 ~ 0-4 A. Find
AN SW ER: 60 V
2. Suppose the data in row 1, column 3, o f Table 5.1 is changed to 10 V. Find y^^^when
V and
= 50
= 0.4 A.
AN SW ER: 120 V
Comparing the development o f equation 5.7 in Example 5.3 with equation 5.1 suggests that the
coefficients Qj and P 2 can be defined as ratios:
1 =
''out
and 132=
I,-,v2
Example 5.3 and these equations suggest the algorithm for finding the coefficients in equation 5.1
by setting all inputs to zero except the input associated with the desired coefficient. This approach
is sometimes impractical. It is not always possible to set an independent source voltage or current
source to zero: imagine turning off a generator for downtown Manhattan to obtain a coefficient.
T he following example illustrates an alternate approach.
E X A M P L E 5 .4 . Consider Figure 5.6, which has two inputs
and
output
Table 5.2
lists measurement data taken in a laboratory. Row 1 ofTable 5.2 lists the nominal operating con
ditions o f the circuit. Rows 2 and 3 illustrate measurements in which one source has its value only
slightly changed (although the change may be arbitrary) while keeping the other source value the
same. From the linearity theorem, we know
find
to complete row 4 ofTable 5.2.
+ ^2^s2- Compute
FIG U RE 5.6. Linear resistive circuit driven by two sources.
and P j, and then
199
TABLE 5.2. Two sets o f measurements of a linear circuit.
i^ 2 (amperes)
hut (amps)
0.25
-1
5+0.1
0.25
- 1 .0 3
0.25+ 0.05
- 0 .9
15
0.5
????
(volts)
So l u t io n
From rows 1 and 2 o f Table 5.2,
- 1 = ttj X 5 + P2 X 0.25
(5.8a)
- 1 .0 3 = ttj x (5 + 0.1) + P 2 X 0.25
(5.8b)
and
Subtracting equation 5.8(a) from equation 5.8(b), we have

- 0 .0 3 = ttj X 0.1 => ttj = - 0 .3
Similarly, from row 3, we have that
- 0 .9 = ttj X 5 + P 2 X (0.25 + 0.05)
(5.8c)
Again, subtracting equation 5.8(a) from equation 5.8(c), we have

0.1 = p 2 x 0 .0 5 ^
Pa = 2
Equation 5.1 for the given data has the linear form
(5.9)
Hence, for row 4 o f Table 5.2, we have that
2,, = - 0 . 3 x 1 5 + 2 x 0 . 5 = - 3 .5 V
200
Exercise. Find the unknown value in Table 5.3 using linearity.

TABLE 5.3. Two Sets of Measurements of a Linear Circuit
(volts)
i^2 (niA)
20
100
15
22
100
15.9
20
110
15.6
28
80
???
AN SW ER: 17.4 A
As a final comment on linearity, we note that by simply using the data o f rows 1 and 2 o f Table
5.2, one can solve for the coefficients by solving simultaneous equations. Specifically, using the
data o f rows 1 and 2 ofTable 5.2, we have the following matrix equation
5
0 .2 5 ' i
5.1
0.25
.^1.
-1
-1 .0 3
whose solution yields the proper coefficients o f equation 5.9.

5
5.1
.A .
0 .2 5 '
0.25
-1
-1
- 1 .0 3
"
= -4 0
'0 .2 5
-5 .1
- 0 .2 5 ' -1
5
-0.3'
-1 .0 3
Exercise. Find the unknown entry in Table 5.4 after finding a , and (3j in the equation
+ Pl^.2TABLE 5.4. Two Sets of Measurements of a Linear Circuit
(volts)
AN SW ER:
ia (itiA)
ôut
10
100
15
20
100
20
30
150
???
= 0.5z^,i + 100/^2
201
3. LIN EARITY REVISITED: SUPERPO SITION AND

PRO PO RTIO N A LITY
T he linearity principle o f equation 5.1 has the more simple form
y=
( 5 . 10)
+ ... + au
Here, j denotes an output, whether it be current or voltage, and each
denotes a source input,
whether it be voltage or current. A special consequence o f the linearity principle is the superpo
sition property. Equation 5.10 says that the total response jy is the sum o f the responses '"aju".
Each a-u is the response o f the circuit to u- acting alone, i.e., when all other independent
sources are set to zero. Although implied by linearity, this property is so important that we single
it out.
THE SUPERPOSITION PROPERTY

For almost all linear resistive circuits containing more than one independent source, any out
put (voltage or current) in the circuit may be calculated by adding together the contributions
due to each independent source acting alone with the remaining independent sources deacti
vated, i.e., their source values are set to zero.
EXA M PLE 5 .5 . A linear resistive circuit has two inputs
and
with output
as shown in
Figure 5.7, where

= 2 Q,, Rj = 2.5 2, and R^ = 10 Q. Find
by the principle o f superposi
tion. Then, compute the power absorbed by the 10
resistor. We show that
=
0.5 V^l + 0.4V^2 where
is the contribution o f the source
acting alone for k = 1, 2.
FIG U RE 5.7 Linear resistive circuit driven by two voltage sources;
Rj = 2 Q , R 2 = 2.5
So l u t io n
Step 1. Fini/ the contribution to
and R^ = 10 Q.
due only to V^j. Denote this contribution by
W ith V^2
= 0, the equivalent circuit is shown in Figure 5.8(a). Here, the 2.5 Q and 10 Q, resistors are in par
allel, yielding an equivalent resistance o f 2 = 2.5 x 10/12.5 f i. By voltage division,
2 +2
202
+
Voût
FIG U RE 5.8 (a) Circuit equivalent to Figure 5.7 when

(b) circuit equivalent to Figure 5.7 when
Step 2. Find the contribution to
due to
= 0;
= 0.
Denote this contribution by
W ith
= 0,
the equivalent circuit is shown in Figure 5.8(b). Here, the 2 Q and 10 O resistors are in parallel,
yielding an equivalent resistance o f 5/3 = 2 x 10/12 Q. By voltage division,
2.5 + 3
Step 3. Compute
by superposition. Using superposition,

Vout = ]/o u t ^+
^ out
= 0 5 K , ^+ 0 4K .,
Step 4 . Compute the power absorbed by the R^= 10 Q resistor.
Pr3 =
(youtT
= 0 . 1(0.5
+ 0.4V,2 f = 0 . 1(o .2 5 v /i + 0 .2 V ,iy ,2 + 0 . 16V / 2 )
Note that the total power, Pj^^, is not the sum o f the powers due to each source acting alone
because o f the presence o f the cross product term. Hence, in general, superposition does not apply
to the calculation o f power.
For dc circuit analysis, the principle o f superposition does NOT apply to power calculations.
Exercise. Reconsider Figure 5.7 in which

principle o f superposition.
AN SW ER;
0.5K^, + 0.25
= 2 Q , 7?2 = 4 Q, and R^ = A Q.. Find
by the
203
The next example adds a controlled source to the circuit o f Figure 5.7 and repeats the superposi
tion analysis.
E X A M P L E 5 .6 . For the circuit o f Figure 5.9, suppose i?, = 2 Q,

S. Using superposition, find
in terms o f
and v^2 show that
= 0 . 5 z^^j + 0 .4 t/^2 where
=5
-^3 = 10
= 0-2
superposition theorem to
is the contribution from the source
acting alone for k = 1, 2 .
q V.A
FIG U RE 5.9 Circuit containing a dependent source for illustrating

the principle o f superposition.
So l u t io n
Step 1. Compute the contribution due only to
Setting v^ 2 ~ ^ leads to the circuit o f Figure 5.10,
where we note that
Applying KCL to the top node yields

0 - 5 ( - L - " . l ) + (0-2 + 0 . 1 + 0 .2 ) .i ,= 0
Therefore,
d r = 0 - 5 ^ .i
Step 2. Compute the contributions due only to
o f Figure 5.11, where this time,
= ôut~
Setting
= 0 in Figure 5.9 leads to the circuit
204
As in Step 1, we apply KCL to the top node to obtain

(0.5 + 0.1 + 0 .2 ).2 ^ ,- 0.2.^2 + 0 - 2 K i - -. 2) = 0
Therefore,
Step 3. Using superposition, add up the contributions due to each independent source acting
alone.
ôut =
Exercise. Repeat Example 5.6 with

AN SW ER:
+ 0-4 ^',2
= /?2 = ^3 = ^ ^
Sm ~
(5.11)
S.
= 0.25i',i + 0.5t^,2
The above examples used voltage division and superposition to compute an output voltage due to
two independent voltage sources.
E X A M PLE 5 .7 . This example illustrates the principle o f superposition for the three-input op
amp circuit o f Figure 5. 12 . Show that
is the contribution o f
acting alone for k = 1 ,2 , 3.
FIG U RE 5.12 Three-input op amp circuit.
+ 2 . 5 K^2 +
205
So l u t io n
due only to
Denote this output by
W ith V^ 2 = ^ 3
= 0, the circuit o f Figure 5.12 reduces to that o f Figure 5.13(a). The properties o f an ideal op amp
ensure that i^ = 0, making
= -0.5i?z^ = 0. Thus, v_ = v^=Q implies

Vi r = - K . = - 4 V ,
2R
(a)
FIG U RE 5.13
due only to V^2 - W ith
shown in Figure 5.13(b) where we denote the output as
= 0, the equivalent circuit is

From op amp properties and volt
age division,
^ -V ,2 = 0 .5 V ,2
R+R
Hence, from Example 4.3,
due only to
as that o f Figure 5.13(b) with V^ 2 replaced by
T he equivalent circuit in this case is the same

Therefore, the output due to source
acting
alone is
Step 4 . Sum up contributions due to each source. By the principle o f superposition.
Vout =
out
out
Exercise. 1. For Example 5.7, suppose
out
= -^4 sK\ , + 2 .5 Ks2t + 2 .5

K.
^ si
^3 ^
(5 . 1 2 )
ôuf
AN SW ER: 2 V
2. Now suppose
A N SW ER: - 1 2 V
= 8 V,
= ^s3 =
saturates at |
= 12 V; compute
206
The above examples have generated the linearity formula, equation 5.1, using superposition, i.e., the
response o f a circuit is the sum o f the responses due to each source acting alone. The technique is equiv
alent to that described in Example 5.3. However, superposition alone is not equivalent to linearity.
Linearity is equivalent to the properties o f superposirion AND proportionality, which is now stated.
T H E P R O P O R TIO N A LITY PRO PERTY

For almost all linear resistive circuits, when any one o f the independent sources is acting
alone, say j, with output 7 , then y =
for some constant a^. Proportionality says that if
is multiplied by a constant K, then the output is multiplied by K, i.e.,
However, for dc analysis, the proportionality property does N O T apply for power calculations.
T he proportionality property is easily illustrated by equation 5.12 o f Example 5.7:
- V'.., V I , * V i ,
. K,
If
and
* 2-5^2 ^ 2-5^3
. V . 0, then V ^ . - 4 ( V ) . - 4 K ) -
Exercises. For certain nonlinear circuits, the principle o f superposition may be satisfied, but pro
portionality not satisfied, or vice versa. This exercise explores these distinctions.
1.
sV ^^ôw that the principle o f
If a circuit has input-output relationship
superposition is satisfied, but proportionality is not satisfied.

2.
If a circuit has input-output relationship
= a,Wjj +
^ôw that the
principle o f proportionality is satisfied, but superposition is not satisfied.
A very interesting and significant application o f the proportionality property occurs in the analy
sis o f a resistive ladder network. A resistive ladder netw ork is one having the patterned structure
shown in Figure 5.14, where each box represents a resistor.
-H
v.Q
FIG U RE 5 .14 A ladder network.

A typical analysis problem follows:
Given
and all resistances in Figure 5.14, find all node voltages. O ne can, o f course, solve the
problem by writing and solving a set o f mesh equations or node equations. A simple trick using
the proportionality property allows us to solve arbitrarily long ladder networks without simulta
207
neous equations, as follows: assume
= 1 V. We can sequentially compute currents and voltages
= \ V. Suppose we call this
in a backwards fashion to obtain the required source value to yield

voltage
Define K =
to be the proportionality constant, where is the actual source
voltage. Then the correct output voltage is

E X A M P L E 5 .8 . Find all the voltages V^-, i = 1, ..., 6 in the resistive ladder network o f Figure 5.15.
-----------
^
L
"
R =100
'4 W R4 = 6 Q
R =5Q
1.
tS
V =50V
-------- '
^
V.
V,
R3 = 5 0
R j= 1 0 Q
t .
V,
R, = 1 0 Q
FIG U RE 5.15 A simple resistive ladder network.
So l u t io n
Assume Vj = 1 V. Repeatedly apply Ohms law, KCL, and KVL as follows: (Q, V and A are used
throughout):
(Ohms law)
/2 = / i =
V2 = /?2 / 2 =
V3 =
/3
/2
1 0
(KCL)
0 .1
0 .1
+ V2 =
(Ohms law)
(KVL)
(Ohms law)
= ^ = 0.4
^3
+ / 3 = 0 .1 + 0.4 = 0.5
V4 = R4 / 4 =
Vg =
\/3
X 0.5 = 3
(Ohms law)
(KVL)
+ y4 = 5
(Ohms law)
7, = "^ = 0 .5
^5
^6 = 7 4 + 7 5 = 0.5 + 0.5 =
V6 = 7?67e=5
y, = y5 + y6 = io
(KCL)
(KCL)
(Ohms law)
(KVL)
We conclude that if Vj = 1 V, the source voltage must be

= 10 V. But the actual source voltage is
50
50 V. Define K =
= = 5 . By the proportionality property, if
= 50 V, then
- I^x 1 =
5 V Similarly,
- 5 V,
= W V,
= 15 V, and V5 = 25 V.
208
In the solution given above, we have separated the expressions into calculation blocks to empha
size the repetitive pattern. For example, the expressions in block #3 are simply obtained from
block #2 by increasing all subscripts by 2. When the ladder network has more elements, the
sequence o f expressions contains more blocks, each o f which entails two additions and two mul
tiplications. This method then allows us to straightforwardly solve ladder networks o f any size
without writing or solving simultaneous equations.
Exercise. In Example 5.7, change all resistances to 2 Q and find V^. Would it make any differ
ence in the voltage Kj if all the resistors were changed to R ohms?
50
AN SW ER: Vi = = 3.85 V, and no difference.
'
13
4. SO U RCE TRAN SFO RM ATIO N S

The words source transformation refer to the conversion o f a voltage source in series with an Rohm resistor to a current source in parallel with an R-ohm resistor, and/or vice versa. This section
explains the details o f such transformations and how they can simplify analysis. But first we must
recall from Chapter 2 that voltage sources in series add together (such as batteries added to a flash
light) and that current sources in parallel combine into an equivalent single current source. This
is illustrated for multiple voltage sources in series in Figure 5.16. Similarly, Figure 5.17 shows how
multiple current sources combine into a single source.
4.5 V
FIG U RE 5.16 (a) Three voltage sources in series; (b) equivalent single voltage source.
s,eq
FIG U RE 5.17 (a) Three independent current sources in parallel;

(b) equivalent single source circuit.
209
SOURCE TRANSFORM ATION THEOREM FOR INDEPEN DEN T SOURCES

A 2-terminal network consisting o f a series connection o f an independent voltage source
and a nonzero finite resistance R is equivalent to a 2-terminaI network consisting o f an
independent current source, /^ = VJR'm parallel with R.
Conversely, a 2-terminal network consisting o f a parallel connection o f an independent cur
rent source
and a nonzero finite resistance R, is equivalent to a 2-terminai network con
sisting o f an independent voltage source,
= RIy, in series with R. T he reference directions
for voltages and currents are as shown in Figure 5.18.
V =RL
FIG U RE 5.18 Illustration of source transformation theorem for independent sources.

A justification for the source transformation theorem will be given in the next section. Practically
speaking, it can save significant computational effort. For example, in the circuit o f Figure 5.19 in
Example 5.9 below, a solution approach using mesh analysis requires writing and solving three
simultaneous equations. Nodal analysis at A and B requires writing and solving two simultaneous
equations. Applying the source transformation theorem is a third avenue that avoids all simultane
ous equations.
E X A M P L E 5 .9 . Find
in Figure 5.19 by repeated applications o f the source transformation
theorem. Then find the power absorbed by the 4 kD resistor.
5 kO
6kQ
50 V
FIG U RE 5.19 Circuit for Example 5.9.
210
So l u t io n
Step 1. Substitute all series V^- R combinations by their parallel
- R equivalents, where in each
case, /^ = . Applying the source transformation theorem four times results in Figure 5.20.
10 mA
10 mA
FIG U RE 5.20 Circuit equivalent to that o f Figure 5.19 by source transformation theorem.
Step 2. Combine the parallel resistances and the parallel current sources.
To the left o f point A are two independent current sources and two resistors, all in parallel.
Similarly, to the right o f B are two current sources and two resistors in parallel.
Combining current sources and resistors to the left o f A results in a single current source o f 5 mA
directed upward and an equivalent resistance o f 4 kD. To the right o f B, the current sources can
cel each other out, and the equivalent resistance is 2 kXl. T he resulting simplified circuit is shown
in Figure 5.21.
5 mA
10 mA
FIG U RE 5.21 Simplification o f the circuit in Figure 5.20.

Step 3. Apply the source transformation theorem a second time to each o f the
allel I ; - R pairs become series V^- R pairs, as illustrated in Figure 5.22.
4kO
--A
20 V
4kQ
B
20 V
FIG U R E 5 .2 2 Further simplification of Figure 5.21.
- R pairs. These par
Step 4 . Find
and P^j^- From Ohms law,

ÂB -
Thus,
211
20 + 20
4+4+2
= 4 mA
= 4 0 0 0 X (0 .0 0 4 )2 ^ <54
Exercises. 1. For the circuit o f Figure 5.23(a), /^j = 50 mA and

R combination to a series V^- R combination, where
= 500 Q. Convert the parallel
= ? and
= ? in Figure 5.23(b).
AN SW ER: 25 V, 500 Q
f^series
-o Circuit
V.
Circuit
r +
B
O
B
-O
(b)
(a)
FIG U RE 5.23
2. For the circuit o f Figure 5.24(a), /^j = 50 mA and R^ = 500 Q , while I ^ 2 = 1 5 0 mA and R^ =
300 Q . Convert the two parallel 1^R combination to a single series V^R combination, where
= ? and
= ? in Figure 5.24(b).
AN SW ER: - 2 0 V, 800 Q
A
ho-
-o Circuit
Circuit
V
B
-O -
B
-O
(b)
(a)
FIG U RE 5.24
3. Consider the circuit in Figure 5.25(a). Using a source transformation and resistance combina
tions, determine the values o f
and Rp^^^ in Figure 5.25(b).
212
100Q
100Q
FIGURE 5.25
AN SW ER: 150 LX 50 niA
5. EQ U IVALEN T N ETW ORKS

The source transformation theorem above is based on the notion o f equivalent networks, as is the
material o f the next chapter. So we now explore a precise understanding o f equivalent 2-tenninal net
works. Figure 5.26 illustrates four 2-terminal networks, all enclosed by dashed boxes, labeled N j,
N 2, Ng, and N^. Their characteristic is that there are only two accessible nodes for connection to other
circuits. Note however that any controlling voltage or current must be contained within the dashedline box. Such dashed-line boxes are often omitted to avoid cluttering in circuit diagrams.
FIGURE. 5.26 Examples of 2-terminal networks, i.e., networks in which only two terminals are
available for connection to other networks.
Observe that networks N j and N 2 in Figure 5.26(a) and (b) have the same terminal characteris
tics: at the terminals o f N j, the v - i characteristic is
v = 2i+ 10
213
At the terminals o f N 2, the characteristic is

v = 2( + 10
i= - - 5
2
The two equations are identical. We then say that a pair o f 2-terminal networks are equivalent if
they have the same terminal characteristics. Therefore,
and N 2 are equivalent.
Now, observe that networks
and
are also equivalent to A^j and A^2- To see this, note that
for N y
V
v-1 5
6/ = 3 v - 3 0
v = 2i + 10
And for N^, first observe that i = \0i^ from KCL, in which case
10 + 20/'a ^
= O.lz; further, from KVL,
v = 2 i + 10
as was to be shown.
Because equivalent 2-terminal networks have the same terminal v i characteristic, if one network
is interchanged with its equivalent, all currents and voltages outside the box remain the same as
illustrated in Figure 5.27; i.e., all voltages and currents in the rest o f the circuit are the same as
before.
A
o
N
N.1
V
J
Rest of
Circuit
-O B
FIG U RE 5.27 The networks denoted N^, i = 1,2, are equivalent when the v-i values at the terminals
are identical; logically then, all voltages and currents inside the Rest o f Circuit remain the same.
These examples allow us to justify the source transformation theorem as follows. Both 2-terminal
networks in Figure 5.18 have the same v - i relationship: v = Ri +
v = Ri + Rlj, = Ri +
and
.
Therefore, the two networks o f Figure 5.18 are equivalent, and the source transformation is a valid
analysis technique.
214
6. SUM M ARY
This chapter covers the notions o f Unearity, superposition, proportionaHty, and source transfor
mations. Linearity states that for any hnear resistive circuit, any output voltage or current, denot
ed as y, is related to the independent sources by the formula j =
+ ... +
are the voltage and current values o f the independent sources, and
ate constants. O nce values for the
where
through
through
are appropri
are known, one can compute the output for any (new) set o f
input values without having to resolve the circuit equations, a tremendous savings in time and
effort. A special consequence o f linearity is the widely used principle o f superposition.
Superposition means that in any linear resistive circuit containing more than one independent
source, any output (voltage or current) can be calculated by adding together the contributions due
to each independent source acting alone with the remaining independent source values set to zero.
Practically speaking, this is the customary path to computing the coefficients, a^, in the linearity
formula.
Proportionality, another consequence o f linearity, means that if a single input is scaled by a constant,
with the other inputs set to zero, then the output is scaled by the same constant. This property led
to a clever technique for analyzing ladder networks without writing simultaneous equations.
Since
power
is
proportional
to
the
square
o f a voltage
or
current, P = =
for dc resistive circuits, the principle o f linearity and its consequences, superposition and propor
tionality, D O N O T APPLY for power calculations.
Using the notion o f an equivalent 2-terminal network, the chapter set forth the theorems on
source transformations for source-resistor combinations: a 2-terminal network consisting o f a
series connection o f an independent voltage source
and a nonzero finite resistance R is equiv
alent to a 2-terminal network consisting o f an independent current source,
= VÎR, in parallel
with R, as illustrated in Figure 5.18. These transformations, applied multiple times to a circuit,
often simplify the analysis o f a circuit.
215

2-term inal network: an interconnection o f circuit elements inside a box having only 2 accessible
terminals for connection to other networks. T he concept is extendible to n-terminal net
works.
Equivalent 2-term inal networks: two 2-terminal networks having the same terminal voltage-cur
rent relationship. I f two 2-terminal networks
and N 2 are equivalent, then one can be
substituted for the other without affecting the voltages and currents in any attached net
work.
and Uj, each acting alone, be y-^ and y 2 . U2 are applied simultaneously, the response is =
Linearity property: let the responses due to inputs

W hen the scaled inputs
and
0 2
+ 2^2- Linearity implies both superposition and proportionality, and vice versa.
Linear resistive element: a 2-terminal circuit element whose terminal voltage and current rela
tionships is described by Ohms law.
Linear resistive circuit/network: a network consisting o f linear resistive elements, independent
voltage and current sources, op amps, and controlled sources.
Proportionality property: when an input to a linear resistive network is acting alone, multiply
ing the input by a constant, K, implies that the response is multiplied by K.
Source transformation: a 2-terminal network consisting o f an independent voltage source in
series with a resistance is equivalent to another 2-terminal network consisting o f an inde
pendent current source in parallel with a resistance o f the same value.
Superposition property: when a number o f inputs are applied to a linear resistive network simul
taneously, the response is the sum o f the responses due to each input acting alone.
216
4 . Consider the circuit shown in Figure P5.4.
Problems
(a)
Find the coefficients a , and (3j in the

linear relationship
LIN EARITY
(b)
and |3j in the
linear relationship v^ 2 =

= 5 2 and
= 20 Q.
(a) Using linearity,
= a j V^j +
Find the coefficients
^l^sT
rnay be expressed
Compute
ttj and
(b)
If v^{t) = 10 cos(10?) V and I 2 = 2 A,
(c)
find v^Jt).
Redo part (a), but this time express
ttj and P 2 in terms o f the hterals
G] = and G t = .
Figure P5.4
(a)
Find the linear relationship between
(b)
If
(c)
(Challenge) W hat is the effect o f dou
and the input sources V^j and l^j= 20 V and ^ = 0.5 A, find 1/
bling all resistance values on the coef

ficients o f the linear relationship
'^(t)
found in part (a)?
Figure P5.1
A N SW ER: (b) 8 cos(lO^) - 8 V
2. For the circuit o f Figure P5-2,
(a)
(b)
find Vg in terms o f
and Gy and
*^2
find Ig in terms o f
and Gy
V^2 > î> ^ 2
CH ECK:
= 0 .2 5 ^ 1 +????/,2
6. For the circuit o f Figure P5.6, find the linear

relationship between
and the independent
sources. Hint: Write a single loop equation.
Figure P5.2
Figure P5.6
(a)
find Ig in terms o f 7^,, /^2>^1 >-^2

Ry and
(b)
find Vg in terms o f /^,, /^2>

+
. ^2>
7. Consider the circuit shown in Figure P 5.7 in
^3-
which /?, = 80 Q. R2 = 20 Q, R^ = 80 Q. and

r^ = 20.
(a)
ôia
the two independent sources.
Hint: Write two loop equations.
-t"
(b)
Figure P5.3
If
= 20 V and 7^2 = 0.125 A, com
pute the power delivered by the
dependent source.
217
G,
<+ i;
I .Q
>R,
'
Figure P5.7
C H EC K : (a)
Figure P5.9
????? I/, + 16/^2 39 m W >
10. A linear resistive circuit has two independent

sources, as shown in Figure P5.10. If
with v^2 ^t) = 10cos(2 ?) V, then
8. Consider the circuit o f Figure P5.8. in which
V. On the other hand, if
Ry = 18 Q.
=9
and R^= 18 D.
with v^^{) = 0, then
(a)
^3 = 18
Ra = 36 Q,
it) and the four independent sources.

out
(b)
= 20 cos(2 i)
= 10cos(2^) mA
= 2 cos(2z-) V. Find the
linear relationship between
and
= 0
and the inputs,
Now compute
when
= 20cos(2/) mA and v^2 (^) = 20 V.
(Challenge) If each o f the resistances

is doubled, what is the new linear rela
tionship. (Reason your way to the
answer without having to resolve the
circuit. Hints: Investigate the effect o f
Linear
resistive circuit
with
dependent sources
changing the resistance in Ohms law

for fixed current. Investigate the effect
o f equal changes in all resistances on a
V________________^
voltage divider formula.)
Figure P5.10
11. Again, consider the configuration o f Figure
= 0 with v^2 ^t) = 10 V, then
P5.10. If
= 55 V. On the other hand, if z^j(r) = 4cos(2^) A

with v^2 ^t) = 0, then
= -2cos(2?) V.
(a)
If z^j(?) = 2cos(2z-) A and v^2 ^t) = -
(b)
If /^j(z-) = -4cos(5^) A and v^2 ^t) =
10cos(2z-) V, find
20cos(5^) V, find
C H EC K : (b) v^Jt) = 108cos(5^) V
Figure P5.8
CH ECK:
27
- - Vi + ?? *4 -
12. Consider again Figure P5.10. Suppose the

+ ?? V2
measured data are as follows: (i)

when
9. For the circuit o f Figure P5.9, express

a linear combination o f /^j, /^2>
equation 5.1(a). Assume
= 0.4 S, ^2 = G j =
0.05 S, and^^ = 0.1 S.
= 15 V
= 2 A and v^ 2 = 10 V, and (ii)
=10
V when
= 3 A and v^ 2 = 5 V.
(a)
Determine the linear relationship
(b)
Find
AN SW ER: 7.5 V
when z^j = 1 A and
218
13. Consider the linear network o f Figure
(a)
P 5.13, which contains, at most, resistors and

linear dependent sources. Measurement data is
given in Table P5.13.
Compute the coeiTicients o f a linear

relationship among the output and
three inputs.
(b)
(a)
Find the linear relationship
(b)
Find the power consumed by the 10

resistor when
= 20 V and
If 7,1 = - 1 A, 1/2 = 4 0 V, and I/3 = 10

V, find the power absorbed by R^.
AN SW ER: 16 W
= 500 mA
Resistive Circuit
witli
Dependent
Sources
V. . -
Figure P 5.I3
Figure P5.15
Table P5.13
I',, (volts)
z'^2 (amperes)
0.4
-1
16. Again consider Figure P5.15. Suppose the

data measurements are given in Table P5.16.
10
Table P5.16
Km
14. Reconsider Figure P5.13 Two separate dc
/j, (mA) ^.2 (V)
measurements are taken. In the first experi

ment,
= 7 V and hi - 3 A, yielding
=1
A. In the second experiment,
= 9 V and z'^2
= 1 A, yielding
= 3 A.
(a)
Find the coefficients o f the linear rela
(b)
tionship
+ p 2i,2 Given the equation found in part (a),
compute
when
= 5 A.
AN SW ER: (b) 90 watts
= 15 V and z^2
15 . The box in the circuit o f Figure P 5.15 con
Case 1
30
-1
11.5
Case 2
40
-1
13
Case 3
30
2.2
-1
11.6
Case 4
30
- 0 .9
11.9
Case 5
40
10
(a)
ments.
Case 1
50
-2
Case 2
Case 3
^^.3 (V)
1
(b)
Find the power consumed by Rj^ for
the data in Case 5.

A N SW ER: 25 watts
Table P5.15
(mA) 1^.2 (V)
Find the coefficients in the linear rela

tionship
+ a2^^2 +
without any matrix inversions.
tains resistors and dependent sources. 7?^ = 100

Q. Table P5.15 contains various data measure
(V)
17. Again consider Figure P 5.15. Suppose the
-1 3
data measurements are given in Table P5.17.
219
Table P5.17
h\
Table P5.18
^ 2 (V)
^",3 (V)
Kut (V)
i,4 (mA)
-out (V)
Case 1
30
-1
11.5
Case 2
-2 0
27
10
Case 3
-1 0
-3
-1 4
Case 4
40
10
10
???
>
(a)
Find the coefficients in the linear rela

tionship
using a matrix inversion.
(b)
Find the power consumed by 7?^ for
the data in Case 4.

A N SW ER: 102.01 watts
SUPERPO SITIO N AND

PRO PO RTIO N ALITY
R2 = 50 Q, 'sX = 12 V, and
(a)
1^2
using
Find
= 200 Q,
mA.
superposition.
18. The linear resistive circuit o f Figure P5.18 has
Specifically, first find
due to
four independent sources. Three o f these sources
acting alone, and
to z,2 act-
have ftxed values. Only one,
is adjustable. In
a laboratory, the data set forth in rows 1 and 2 o f
out
ing alone.
AN SW ER: 2.4 V, 2.4 V, 4.8 V
Table P5.18 were taken. Complete the last two

rows o f Table P5.18 using linearity and the data
(b)
Find
and i^ 2
from the first two rows. For the data in row 3,

find the power delivered by the current source z^.
Hint: To solve this problem, recall from the lin
in terms o f the literals
compute the spe
cific numerical relationship.

(c)
If
= 10
determine
earity equation 5.1
, R^-,
12 V and j2 = ^ x 60 mA,
using the proportional
ity theorem by first computing
ôut =
+ ^2^.2 +
\'s\
due to the modified

and
'ih-i
acting alone,
due to the modified z'^2 act
ing alone.
We have used the fact here that the term (a,v^
+ ^2^j2
constant because the associat
ed source values are constant. Thus,
+
^2î2
^^
some K. Hence, one can
use the data from the first two rows o f Table
P5.18 to solve for
and K.
Figure P5.19
which
= 20 2 , R2 = 60 Q, R^ = 20 Q.
(a)
Linear resistive
network with
dependent
sources
tionship
+ a 2 i, 2 + ^3^,3 by
superposition. Specifically, first find
a
due to z/^j acting alone, due to z^2

acting alone, and
ing alone.
Figure P5.18
g^f due to z^g act
220
(b)
(c)
Repeat part (a), but express your answers
Using superposition, find

acting alone, and then find
Find
power delivered to
v^2 acting alone. What is
V,
(d)
(a)
in terms o f the literals R-,i= 1, 2, 3.
= 2 A, and
and
when
(b)
due to
Redo part (a) using the literals G/ =
Ri
= 4 A.
Repeat part (c) when

is tripled, and
and the
= 100
due to
is doubled, i^2
is halved.
1
AN SW ER: (a)
5
2
7
= -''.v i- '' m = - ',v2
AN SW ER: (c) 100 V, 500 watts
Figure P5.22
Figure P5.20
23. For the circuit o f Figure P5.23, suppose R^ =
21. In the circuit shown in Figure P5.21, R-^ =

180 Q, R^ = 360 a , T?3 = 90 Q, an R^ = 720 Q.
(a)

tionship
+
fi by superposition.
Repeat part (a), but express your
0 -2 1
(b)
answers in terms o f the literals G- =
20 Cl, /?2 = 50 Q,
(a)
= 100 Q and
= 0.02 S.
Using superposition, find
due to
acting alone, and then find

due to v^ 2 acting alone. W hat is
A N SW ER:
= 0.5j',| + 0.9i^,,
(b)
Redo part (a) using the literals G,- =
l lR- ,i = 1 , 2 , 3 , 4.
(c)
Find
and the power absorbed by
R^ when
= 100 V and v^ 2 = 50 V.
= 60 V;
= 5 watts
AN SW ER:
(d)
Repeat part (c) when
= 0.5 x 100 V
and v^ 2 = 10 x 2 V.
Figure P5.21
Figure P5.23
24. For the circuit o f Figure P 5.24, find the
contribution to
from each independent
source acting alone, and then compute

'., .,.> U
- 0
power absorbed by the 900 Q resistor.

AN SW ER: 38 V and 1.6 watts
22. For the circuit o f Figure P5.22, suppose R^

= 20 Q., /?2 = 50 Q and R^ = 100 Q.
by
the principle o f superposition. Finally, find the
221
i8on
225 n
20V
Figure P5.27
0.1 A
28. (a)
^900 0
MATLAB program.
Figure P5.24
25. For the circuit shown in Figure P5.25,
160 V. Find
(b)
If it is known that
(c)
Find the equivalent resistance seen by
Then find the intermediate
node voltages. Hint: Assume
For the circuit shown in Figure P 5.28,

If /j = 1 A, find /j by writing a
= 200 mA, find /j.
the current source.
= 1V
and use proportionality, as per Example

5.11.
C H E C K : Answer is an integer.
+
Figure P5.28
LIN EARITY AND OP AM P

CIRCU ITS
26. For the circuit shown in Figure P5.26,
64 mA. Find
Hint: Assume
29. Consider the circuit in Figure P5.29.

(a)
= 1 A and
then use proportionality.

1n
1n
1n
1n
III
2n
1n
2n
2n
2n
Find the contribution to
due only
to
(b) Find the contribution to
due only
'^^2o>î (c) Find
by superposition.
r2n
Figure 5.26
2 7 .(a)
For the circuit shown in Figure P5.27,

If Vj = 1 V, find V; by writing a
MATLAB program to solve the prob
lem, given that R-^ = 10 Q ,
= 10
= 5 Q, ^4 = 6 Q ,
= 10 Q ,
Figure P5.29
=
(b)
5 Q,
= 20 Q, and
= 5 Q..
If it is known that
= 175 volts, find
(c)
^1Find the equivalent resistance seen by
(d)
the voltage source.

Suppose
is changed from 1 to 10
Q in steps o f 0.25 O.. Obtain a plot o f
vs
by modifying the MATLAB
code o f (a). Assume Vj = 1 V.

(a)
due only
(b)
to I/j.
(c)
to
Find
by superposition.
due only
222
Figure P5.30
Figure P5.32
31. (a) For the circuit of Figure P 5.31, find
Vût^ the voltage due to each source

acting alone in terms of the literal values,
(b)
Find
(c)
Now suppose that

= 2Rq,
= 3i?o
= 4 i^ , Rr= URq, and
= 100 Q,
33. (a)
in terms of the literals.

(b)
If the input voltages are

= - 2 .5 V and
Vout.
Suppose each voltage source has value 2

V: (i) Find the power absorbed by the
load to each source acting alone, and
For the circuit in Figure P 5.33, find

the linear relationship between
(c)
= 5 V, V^2
= 2 V, determine
If the voltages are all halved, what is

the new
(ii) the actual power delivered to the

load when all sources are active.
r^
34. (a)
For the circuit in Figure
P 5.34, find the linear relationship
Figure P5.31
Kd-
between V^^^ând
(a)
due only
(b)
to Ki
due only
to
(c)
Find
(d)
If
= ^3 = 0 . 5 ^ = 5 kQ
and V;i = 2V;2 = 4 V, find the
(b)
If the input voltages are
V^2 = 0-5 V and

Vour
(c)
= 0.25 V,
= 2 V, determine
If the voltages are all halved, what is

the new
by superposition.
power delivered to the 1
load.
n
Figure P 5 .3 4

(a)
Find the contribution to K^^^due only
(b)
(c)
to 1/2.
Find
(d)
223
(a)
= Q.25R^= R2 = Ry
due only
(b)
to
due only
(c)
to
Find
which
due only
by superposition.
by superposition.
Find the power delivered to Rj^ when

is acting alone, i.e.,
then find the power delivered to R^
(e)
when V^2 acting alone when Rj =

R and R^ = R.
Find the total power delivered to R^
when R 2 = 4 R and R^ = R.
Figure P5.37
Find the contribution to Vout due only
(a)
to l/i(b)
Figure P5.35
(c)
(b)
= Q.25Rjr.
Find the
contribution to
only to
(c)
only to V^2 Find

by superposition.
(d)
Find the power delivered to Rj^ when
which
(a)
(d)
Find the contribution to Vout due only

to K2Find the contribution to Vout due only
to V^,3.
Find
by superposition.
due
due
is acting alone, i.e., V^2 =

and
then find the power delivered to
(e)
when V^ 2 is acting alone when R ^ ARy

Find the total power delivered to
Figure P5.38
when Rjr=
SO U RCE TRAN SFO RM ATIO N S

39. Use a series o f source transformations to
simphfy the circuit o f Figure P 5.39 into one
consisting o f a single voltage source in series
with a single resistance.
224
200 n
0
Figure P5.39
AN SW ER: 6 V source in series with 12 Q resis
40 0
80
40 n
20 n
son
Figure P5.41
AN SW ER: 13.5 and 9.1125
42. Use source transformations on the circuit o f
tor
Figure P5.42, to compute the value o f

which /,i = 10 mA, 1^2 = 20 V, and I/3 = 80 V.
(a)
need
ed to deliver a current o f I = 0.25 A.
Use a series o f source transformations

to find a single voltage source in series
with a resistance that is in series with
the 9.6 kQ resistor.
(b)
Then find the power absorbed by the

9.6 kQ resistor.
Figure P5.42
A N SW ER: 28 V
Figure P5.40
AN SW ER: (a) 48 V in series with 3.2 kQ; (b)
135 mW
43. For Figure P5.43, use a series o f source

transformations to find the value o f
so that
the power delivered to
is 16 watts.
40 A
41. In the circuit o f Figure 5.41, 1/j = 240 V

and 1^2 = 0-25 A.
(a)
Use a series o f source transformations
to reduce the circuit o f figure P5.41 to
sn
2A
<
50<
= 1n
a current source in parallel with a sin

gle resistor in parallel with the 20
resistor across which
V ^ appears.
(b)
Find
Find the power dissipated in the 20 Q
resistor.
(c)
I f both sources have their values

increased by a factor o f two, compute
the new value o f
Can you do this
by inspection? Explain.
is n
Figure P5.43
44. Apply source transformations to the circuit
shown in Figure P 5.44. Then write two nodal
equations to find
and V2 .
Figure P5.44 Source transformations simplify

writing node equations.
AN SW ER: 25 V, 20 V
45. Apply source transformations to the circuit
o f Figure P5.45. Then write two nodal equa
tions to find K| and Vj-
Figure P5.45
AN SW ER: 2.8 V, - 0 .4 V
225
Thevenin, Norton, and

Maximum Power Transfer Theorems
H ISTO RICAL NOTE
In the early days o f electricity, engineers wanted to know how much voltage or current could be
delivered to a load, such as a set o f street lamps, through a complex transmission network. Before
the days o f computer-aided circuit simulation, simplification o f complex circuits allowed engi
neers to analyze these very complex circuits manually. In 1883, a French telegraph engineer, M.
L. Thevenin, first stated that a complex (passive) network could be replaced by an equivalent cir
cuit consisting o f an independent voltage source in series with a resistor. Although stated only for
passive networks, the idea o f a Thevenin equivalent evolved to include active networks. Its wide
spread use has simplified the homework o f students for many years now and probably will con
tinue to do so for many years to come.
A more recent but quite similar idea is the Norton equivalent circuit consisting o f an independ
ent current source in parallel with a resistance. At the time o f E. L. Norton (a scientist with Bell
Laboratories), the invention o f vacuum tubes made independent current sources a realistic possi
bility. Many electronic circuits were modeled with independent and dependent current sources.
The appearance o f Nortons equivalent circuit was a natural outcome o f advances in technology.
CHAPTER OBJECTIVES
1.
2.
3.
4.
Define and construct the Thevenin and Norton equivalent circuits for passive networks.
Define and construct the Thevenin and Norton equivalent circuits for active networks
containing dependent sources or op amps.
Illustrate several different techniques for constructing the Thevenin and Norton equiva
lent circuits.
Investigate maximum power transfer to a load using Thevenin or Norton equivalents.
228
Chapter 6 Thevenin, N orton, and M axim um Power Transfer Theorem s
SECTIO N H EADIN GS
1. Introduction
2. Thevenin and N orton Equivalent Circuits
3. A General Approach to Finding Thevenin
6 . Thevenin and N orton Equivalent Circuits
7. M aximum Power Transfer Theorem
8 . Summary
9. Terms and Concepts
10. Problems
for Linear Passive Networks

and N orton Equivalents
for Active Networks
for Op Amp Circuits
from Measured Data
Practicing electrical engineers often want to know the power absorbed by one particular load. The
load may be a large machine in a factory or a lighting network in the electrical engineering build
ing. Simple resistances often represent such loads. Usually the load varies over time in which dif
ferent resistances are used at different times to represent the load. W hat is the effect o f this load
variation on the absorbed power and on the current drawn by the load? To simplify analysis, the
rest o f the linear network (exclusive o f the load) is replaced by a simple equivalent circuit consist
ing o f just one resistance and one independent source.
For our purposes, a (resistive) load is a two-terminal network defined in Chapter 1, meaning that
the current entering one o f the terminals equals the current leaving the other. More generally, a
two-terminal network is any circuit for which there are only two terminals available for connec
tion to other networks. (See Figure 6.1.) T he important question for our work in this chapter is:
How does one characterize a two-terminal networks As is shown in Figure 6.1(a), there is a voltage
v{t) across the terminals and a current i(t) entering one terminal and leaving the other. T he rela
tionship between the voltage v{t) and the current i{t) characterizes the two-terminal network. For
example, if v{t) = Ri{t), we would recognize the terminal network as an equivalent resistance R.
Or, if v{t) = Ri{t) + Vq, we might recognized this equation as that o f a resistance in series with a
voltage source. In fact, this equation could be represented as graph, e.g. Figure 6.1(b).
This leads to our next question: When are two 2-terminal networks equivalent'^ As developed in
Chapter 5, two 2 -terminal networks are said to be equivalent when their terminal v-i characteris
tics are the same. O f particular interest for this chapter is an equivalent network consisting o f a
voltage source in series with a resistance, called the Thevenin equivalent network, and a current
source in parallel with a resistance, called a Norton equivalent network. Figure 6.1c shows a
Thevenin equivalent for a linear resistive circuit.
Chapter 6 Thevenin, N orton, and M axim um Power Transfer Theorems
N-
229
i(t)
2-terminal
Linear
o
+
Resistive
Network
V(t)
FIG U RE 6.1. (a) a 2-terminal linear network with terminal voltage v{t) and current i{t)\
(b) graphical representation of the equation v{t) =
+ Vq,
(c) Thevenin equivalent network
having the same terminal v{t) and i{t) relationship as (b).
This chapter investigates the replacement o f a network N by its Thevenin equivalent or its Norton
equivalent. The first section describes the Thevenin and Norton equivalent theorems for passive net
works, those containing only independent sources and resistors. Following that, we generalize the
statements to include active networks. However, because op amps have peculiar properties, Thevenin
and Norton equivalents o f circuits with op amps are explored exclusively in Section 4. Following
this, in Section 5, we describe how to obtain a Thevenin or Norton equivalent from measured data
without having to know anything about the internal circuit structure. This is particularly useful
when one has equipment such as a power supply but no schematic diagram o f the internal circuit
ry. Unfortunately, not all linear devices have a well-defined Thevenin or Norton equivalent. The
homework exercises illustrate a few cases. Section 6 explores the problem o f maximum power trans
fer to a load in the context o f the Thevenin equivalent circuit, which ends the chapter.
2. TH EVEN IN AND NORTON EQ UIVALEN T CIRCU ITS FOR

LINEAR PASSIVE N ETW ORKS
Our first objective is to develop and illustrate the celebrated Thevenin theorem for passive net
works. Then we will state and illustrate Nortons theorem, dual to Thevenins theorem.
To develop Thevenins theorem, consider Figure 6.2(a) consisting o f two 2-terminal networks, N
and Nj^ , joined at A and B. Only resistors and independent sources make up N, while
con
230
tains arbitrary even nonlinear elements. Suppose
undergoes various changes as part o f an
experiment, while N, complicated in its own right, remains unchanged. To simplify repeated cal
culations, N is replaced by its Thevenin equivalent, as illustrated in Figure 6.2(b). The more sim
ple Thevenin equivalent consists o f a single voltage source,
in series with a single resistance.
Rthf
------ o-----+
Resistances
and
independent
Sources
Arbitrary
Networi<
r-\
(a)
i, = 0
-A
NResistances
and
independent
Sources
-o
-I-
-o
NReslstances
with independent
Sources
-o
R..
Deactivated
(c)
(d)
FIG U RE 6.2 (a) Network TVattached to an arbitrary network load, N^,

(b) N replaced by its so-called Thevenin equivalent,
(c) circuit for computing
still attached to 7V^;
(d) circuit for computing
in which all independent sources inside N are deactivated.

This brings us to a formal statement o f Thevenin s theorem for passive networks.
231
TH EVEN IN 'S TH EO REM FOR PASSIVE N ETW ORKS

Given an arbitrary 2-terminal linear network, N, consisting o f resistances and independent
sources, then, for almost all such N, there exists an equivalent 2-terminal network consisting
o f a resistance,
in series with an independent voltage source,
Th e voltage,
called the open-circuit voltage, is what appears across the 2 terminals o f N. R^j^, called the
Thevenin equivalent resistance, is the equivalent resistance o f N when all independent sources
are deactivated. Figure 6.2(c) shows the appropriate polarity for v^J^t),
In the above theorem, for almost all means there are exceptions. For example, an independent
current source does not have a Thevenin equivalent. More generally, any two-terminal network
characterized by i{t) = constant does not have a Thevenin equivalent. This leads us to suggest that
there ought to be an equivalent current source formulation o f an equivalent network. From
Chapter 5, the source transformation theorem tells us that the Thevenin equivalent o f Figure
6.2(b) when
0 is equivalent to a current source in parallel with R^j^, as in Figure 6.3(b).
Figure 6.3 leads us to a formal statement o f the so-called Norton theorem.

NResistances
and
Independent
Sources
(c)
FIG U RE 6.3 (a) Arbitrary 2-terminal linear network o f resistors and independent sources;
(b) Norton equivalent circuit; (c) circuit for computing
with
computed, as per Figure 6.2(d).
N O RTO N 'S TH EO REM FOR PASSIVE NETW ORKS

Given an arbitrary 2-terminal linear network, N, consisting o f resistances and independent
sources, then for almost all such N, there exists an equivalent 2-terminal network consisting
o f a resistance, R^^, in parallel with an independent current source, z'^^(r). Th e current,
called the short circuit current, is what flows through a short circuit o f the 2 terminals o f N,
as per Figure 6.3(c). R^^ as before, is the Thevenin equivalent resistance o f N computed
when all independent sources are deactivated.
A single voltage source does not have a Norton equivalent, and as mentioned a single current
source does not have a Thevenin equivalent. Both Thevenin and Norton equivalents exist for a 2terminal linear circuit when R^^
0 and is finite. W hen both the Thevenin and Norton equiva
lents exist for the same network, the source transformation theorem and Ohms law imply that
thhc'kt)
(6.1a)
232
Chapter 6 Thevenin, Norton, and M axim um Power Transfer Theorem s
4^ 0, then
and when
r,
th -
(6.1b)
This formula turns out to be useful in calculating
for a variety o f circuits, especially op amp
circuits.
E X A M P L E 6 .1 . For the circuit o f Figure 6.4, using literals, find the open circuit voltage,
short circuit current,
200 Q ,
and the Thevenin equivalent resistance,
the
Then, if R^ = 50 Q,
= 100 V, and i^ 2 = 2 A, construct the Thevenin and Norton equivalent circuits.
-O
-O
40 n
160V
40 0
4A
-o
-O
(c)
(b)
FIG U RE 6.4. (a) Resistive 2-terminal network; (b) Thevenin equivalent; (c) Norton equivalent.
S
o l u t io n
Step 1. Find
Using superposition, we have by voltage division and Ohms law,

/?2
^1^2
~ R\+R2
^1 + ^2
Substituting the given values into this formula yields

= 0.8
100 + 4 0 x 2 = 160 V
Step 2. Find i^^. As per Figure 6.4, with terminals A and B shorted ^together, all the current from
flows through the short circuit. From superposition, isc = 's2'^----- Substituting numbers
into this formula yields
i^^ = 2 + 0.02
100 = 4 A
233
Step 3. Find
Replacing
by a short circuit and z'^2 t>y an open circuit implies that

^
^ 40
R, +R2
Step 4 . Determine the Thevenin and Norton equivalent circuits. The Thevenin equivalent circuit fol
lows from Steps 1 and 3 and is illustrated in Figure 6.4(b). The Norton equivalent circuit follows
from Steps 2 and 3 and is illustrated in Figure 6.4(c). We also note that
R ,, =
V =155 =
he
40 Q
as expected.
It is important to note that for many circuits, especially when the deactivated circuit is a seriesparallel connection o f resistances, one can obtain the Thevenin equivalent by a series o f source
transformations.
Exercises. 1. Redo Example 6.1 using a series o f source transformations.

2. In Example 6.1, suppose
= 100 Q,
= 400 Q, v^^ = 100 V, and i^ 2 = 2 A. Find
v^^, and i^^.
AN SW ER; 80 Q, 240 V, 3 A
Among the three quantities, R^j^, v^^, and
if two have been calculated, then the remaining one
follows easily from Equation 6.1. In some cases, the choice o f which two to find first either
increases or decreases the amount o f calculation. The following exercises illustrate this point.
Exercises. 1. For the circuit o f Figure 6.5, R^^= 200 Q, R2 = 50 O., R^= 10 Q,
= 50 V. Find R^f^, i^^ and v^^ in this order.
= 100 V, and v^2
A N SW ERS: 8 Q, 1.5 A, 12 V
2. For the circuit o f Figure 6.5 with the same values as in Exercise 1, find v^^, i^^, and
this order.
AN SW ER: Same as in 1, but v^^^. is harder to find.
3. For the circuit o f Figure 6.5, find the Thevenin equivalent circuit using a series o f source trans
formations.
The next example illustrates the computation o f the Thevenin and Norton equivalent circuits
using loop analysis.
Chapter 6 Thevenin, Norton, and M axim um Power Transfer Theorem s
234
E X A M PLE 6 .2 . Find the Thevenin and Norton equivalent circuits seen at the terminals A-B for
the circuit depicted in Figure 6.6, where
= 100 V and
= 3.2 A. We show that
= 4 0 0 Q,
= 200 V, and
= 0.5 A.
1500
500 Q
FIGURE 6.6 Two-source circuit for Example 6.2 with loop currents shown;
So l u t io n
Step 1. Compute
= 100 V and i^ 2 = 3.2 A.
To compute R^^^, we set all source values to zero. Each voltage source becomes
a short, and each current source becomes an open. This leads to the circuit o f Figure 6.7. Here, we
have a 500 Q in series with 100 Q, yielding 600 Q. Since this 600 Q resistance is in parallel with
400 2, the resulting equivalent resistance is 240 Q. Hence,
500 0
= (150 + 240 + 10) = 400 Q..
1500
FIG U RE 6.7 The circuit of Figure 6.6 with all independent sources deactivated.
Step 2. Compute an expression for
A-B is N O T present. Hence,
Because we are computing
the short across the terminals
= 0 and no current flows through the 150 Q resistor. This means
its voltage drop is zero. (One ofi:en says that the 150 Q resistor is dangling.) Thus, from KVL we
have
= 'ôozj + ioz;2
(6 .2)
Chapter 6 Thevenin, Norton, and Maximum Power Transfer Theorems
23^
Step 3. Compute iy The only unknown in Equation 6.2 is /j, since
~ 3-2 A. Hence, around
loop 1,
^s\ =
+ 100/^2= 1000/jand
in which case,
/| = 0.0 0 ly^j + 0.1/^2

Thus, from Equation 6.2,
= 400(0.001
+ 0.1/^2) + lO/^, = 0.4r/^, + 50/^2 =
+ 160 = 200 V
Step 4 . Construct the Thevenin and Norton equivalent ciraiits. Equation 6.3 with
(6.3)
= 400 Q yields
the Thevenin equivalent o f Figure 6.8(a). Further, from the source transformation theorem,
(6.4)
Equation 6.4 leads to the Norton equivalent circuit o f Figure 6.8(b).

400 Q
FIGURE 6.8 (a) Thevenin equivalent o f circuit of Figure 6.6;

(b) Norton equivalent o f Figure 6.6.
Step 5. Compute i^^ directly so as to verify the above calailation. This step is merely given to illus
trate the direct calculation o f i^^ and is unnecessary at this point to the solution o f the problem.
Referring again to Figure 6.6 and assuming that the short acro.ss A-B is present, then /2 = i^^.
Hence, around loop 1,
^S\
= 500/, + 400 (/j - / J
+ 100(/j - /^2)
in which case.
v^\ + 100/^2 = 1000/j - 400/^^ = 420 V
Around loop 2 we have
560/;^-400/, = 10/^2 = 32 V
In matrix form, the pertinent equations are
1000
-4 0 0 '
-4 0 0
560
h
Jsc
420'
32
2M^
Chapter 6 Thevcnin, Norton, and Maximum Power Transfer Theorems
Thus,
' i\
1000
-4 0 0 '
j.sc
-4 0 0
560
-1
420
0.62'
32
0.5
Consequently, / = 0.5 A as was found earlier using the easier method o f
Vo,.
th
Exercises. 1. Suppose all source values in the circuit o f Figure 6.6 are doubled. What is the new
v j Does
change?
A N S W E R : /, =-)()() V. no
2. Suppose all resistances in the circuit o f Figure 6.6 are multiplied by 4 and the independent cur
rent source is changed to 0.6 A. Find
and
Hint: For
in equation 6.3, the value 5 0
is in ohms, so if the resistances are multiplied by four, what is the new value?
AN SW ER: r
= 160 V. A',
// 4 x 400
1600 Q. and /SC

. = 0.1 A
3. A 400 L resistor is connected in series with terminal A o f the circuit o f Figure 6.6. Find the
I V
ANSWLR:
im
V, A',,,
=K) ti. ,nd
- .2S A
4. A 400 Q resistor is connected across terminals A and B o f the circuit o f Figure 6.6. Find the
V
" I V
ANSV('-UR:
O.SA, /(, titJi - 2 0 0 U . a n d r oC- lOl) V
In the above two examples, deactivation o f all independent sources led to a series-parallel network.
Calculation o f
wâs then straightfor\vard. In fact, we can state a corollary to Thevenin and
Nortons theorems.
CO RO LLA RY TO TH EVEN IN AND N ORTO N'S TH EO REM S

FOR PASSIVE N ETW ORKS
When a network contains no independent sources,
= 0 , and the Thevenin or
Norton equivalent consists o f a single resistance R^f^. For a series-parallel net%vork, R^f^ can be
computed by straightforward resistance combinations.
3. A GEN ERAL APPROACH TO FINDING THEVEN IN AND

NORTON EQUIVALENTS
Consider Figure 6.10(a) where we have a network N connected to the remainder o f a larger cir
cuit. Our goal is to replace the net\vork N by its Thevenin equivalent, as shown in Figure 6.10(b).
Chapter 6 The\enin, Norton, and Maximum Power Transfer Theorems
23'
The terminal v-i characteristics o f the network N and its Thevenin equivalent must be the same.
Consider that the v-i characteristic at A-B o f the Thevenin equivalent o f N is
(6.5)
while the Norton equivalent o f N as per Figure 6.10(c) has the v-i relationship
'a
- 1
^th
(6 .6 )
- ^sc -
These relationships tell us that if we have a linear net\vork and assume there is a voltage
across
its terminals and a current /^j entering the network, as shown in Figure 6.10(a), then obtaining an
equation o f the form
(6.7)
or o f the form
(6.8)
Vj n
allows us to match the coefficients o f equations 6.7 and 6.5 to determine

the coefficients o f equations 6.8 and 6.6 to determine ^th = - and
and
or to march
This sometimes proves
^th
an easier approach for non-simple circuits, as the next two examples illustrate.
Remaining
Network
^
A
------oLinear
Network
A8
- o .......+
B
(a)
Remaining
Remaining
Network
(c)
I'lG U R M 6.10 (a) Nervvork N attached to an unknown network;
(b) theTheveinin equivalent of N attached to the unknown nervvork; (c) the
Norton equivalent of N attached to the unknown network.
23.S
E X A M PLE 6 .3 . This example revisits Example 6.2 using the new approach. Again, we find the
Thevenin and Norton equivalent circuits seen at the terminals A-B for the circuit depicted in
Figure 6. I I , where
= 100 V and
= 3.2 A. Our goal is to find the v-i characteristic at the ter
minals A-B.
FIGURE 6.11 Two-source circuit for Example 6.2 with loop currents
shown;
= 100 V and ip = 3.2 A.
So l u t io n
Step 1, Consider i^ loop. Around the loop for i^, we have
ÂB ~ 560/^ + 400/j + 10/^2 = 560/^ + 400/j + 32
(6.9)
Step 2. Consider loop 1. From Example 6.2, around loop 1 we have,

= 100/^2 = 420 = 1000/, + 400;;^ V
Thus,
=
42 0 - 400/^
(6 . 10)
1000
Step 3. Substitute. Substituting equation 6.10 into 6.9 yields

4 2 0 - 400/4
+ 32 = 400/.A + 200 V
V.
o = 560/.A + 4 0 0 ----------------^
AU
jQQQ
Step 4. Match coefficients. Matching coefficients o f equations 6.11 and 6.5 implies that
R,h = 4 0 0 Q,\-
= 200 V and /,, = ^
= 0.5 A.
th
(6.11)
Chapter 6 Thcvenin, Norton, and Maximum Power Transfer Theorems
ly )
EXAM PLE 6.4. For the circuit o f Figure 6. ] 2, find theThevenin equivalent o f the 2-terminaJ Network
N defined by the dashed line box. We show that
= 9.6 V,
= 4.4 Q., and = 2.1818 A.
FIGURE 6 . 12 A current source is is attached to N for computing
and
So l u t io n
Our objective is to compute the relationship o f the form o f equation 6.7 using Nodal analysis and
then match coefficients with equation 6.5 to obtain R^j^ and
Assume /^, = 2 A and
~^
Step 1. Write nodal equations. For writing the equations o f this circuit, the reader might first
review Example 3.2. Alternately, using the inspection method, the matrix nodal equations are
- 0 .2
0.8
- 0 .4 ^1
- 0 .2
0.6
-0.1
- 0 .4
-0 .1
0.5
Step 2. Solve equation 6.12 fo r
h\
' 1
\'2
\>2
( 6. 12)
js l + U .
^'ab
using Crammer's rule. First, we note that

0.8
det(A/) = det
- 0 .2
0.2
0.6
- 0 .4
-0 .1
- 0 .4 '
-
0.1
= 0.:
0.5
From Crammers rule,

0.8
- 0 .2
det - 0 .2
0 .6
h\
0
- 0 .4
-0 .1
s2 + U
âb -
d et(M )
which from the properties o f determinants becomes

0.8
- 0 .2
O'
det - 0 .2
0.6
- 0 .4
-0 .1
d et(M )
4 . 4/4
0.8
- 0 .2
0.8
- 0 .2
O'
det - 0 .2
0.6
det - 0 .2
0.6
- 0 .4
-0 .1
- 0 .4
-0 .1
. /. -1------"
d et(M )
+ 2 . 6/^1 + 4 .4 /^ 2 = 4.4/,^ + 9 .6
/ . 4- -----
d et(M )
( 6 .1 3 )
240
Equation 6.13 shows that
is calculated finding four determinants numerically using MATLAB
or equivalent.
Step 3. Match coefficients o f equations 6.13 and 6.5. Matching coefficients o f equation 6.13 with
equation 6.5, we obtain
V,
= 4.4 a and
2.1818 A
Exercises. 1. If the independent current sources in the circuit o f Figure 6.11 arc set to zero, find
the Thevenin equivalent circuit.
AN.SW'j-^R: The I hcvcnin ci|uiv.ilcnt .onsist.s oi a single rc.sisn)r.
2. Find
= 4.4 12.
when i^^= 10 A and i^2 = 5 A.
A N S W I- R ; -48 \
3. A 4.4 d resistor is connected in series with terminal A o f the circuit o f Figure 6.12. Find the
V
ANSW'l-.R:
. ).(, \.
= 8.8 12. and
=1
4. A 4.4 Q resistor is connected across terminals A and B o f the circuit o f Figure 6.12. Find the
ANS\V1;R:
- 2 .1SIS A.
= 2.2 f l and
M .S V
At this point, we end our development in this section with an example that shows how to com
pute a Thevenin equivalent from measured, e.g., in a laboratory setting where there is a power sup
ply with an adjustable voltage.
EX A M PLE 6 .5 . Consider Figure 6.13, which show^s the Thevenin equivalent o f an unknow'n network N attached to a variable voltage,
power supply, which also shows the current delivered
to the unknown network N, i.e.,

Two measurements o f the unknown network N are taken, and
the data is displayed in Table 6.1. Find the Thevenin equivalent o f N.
N
r
R..
Variable
Voltage
Power
Supply
oc
B
MCURI-. 6.13 Thevenin equivalent of an unknown network N
connected to a variable voltage power supply.
2/f
TABLE 6.1
/Î (mA)
-
(V)
10
24
20
40
o l u t io n
Substituting the measured data in Table 6.1 into equation 6.5
yields
24 = 0.01,;,+ V
from row 1 o f Table 6.1, and
40 = 0 . 0 2 ,^ * V
from row 2 o f Table 6.1. In matrix form,
0.01
0.02
^th
24
40
Solving produces
^r/i
'V>(Hence, R,,, = 1600
1
-1 24
r -I 24
= -100
-0.02 0.01 40
40
0.02 1
0.01
and
1600'
8
= 8 V.
Thus, one can use the technique o f Example 6.5 to determine Thevenin equivalent circuits in the
laboratory.
4. TH EVEN IN AND NORTON EQ U IVALEN T CIRCU ITS FOR

A CTIVE N ETW ORKS
Constructing Thevenin and Norton equivalents for active networks, those containing dependent
sources and op amps, presents us with some unique challenges. Except with one extra condition,
Thevenin and Nortons theorems and their corollary are valid for active networks. Because active
networks contain dependent sources, the extra condition is that all controlling voltages or currents
be within the 2-terminal network whose Thevenin/Norton equivalent are being sought.
Chapter 6 *Thevenin, Norton, and Maximum Power Transfer Theorems
242
TH EVEN IN AND N ORTO N'S TH EO REM S FOR A CTIVE

N ETW O RKS
For almost every 2-terminal linear network, N, as in Figure 6.14(a), consisting o f resistances,
independent sources, and dependent sources whose controlling voltages and currents are con
tained within N , there is an equivalent 2-terminal network consisting o f either (i) a resist
ance,
in series with an independent voltage source,
(Figure 6.14(b)), or (ii) a resistance,
called the Thevenin equivalent
in parallel with an independent current source, /y^(^),
called the N orton equivalent (Figure 6.14(c)). In most cases, both the Thevenin and Norton
equivalent circuits exist. Computation o f y^îs characterized by Figure 6.14(a), computation
of
by Figure 6.14(d), and computation o f b y Figure 6.14(e).
/^Neq
NResistances,
independent
and
R..
dependent
sources
(a)
(b)
N-
(c)
NResistances,
Independent
sources
deactivated
independent
and
(d)
(e)
-o
dependent
sources
M GURE 6.14 (a) Arbitrary linear nerwork N; (b) Thevenin equivalent of N;

(c) Norton equivalent of N; (d) N with independent sources deactivated for calculating R^j',
(e) N with short circuited terminals for calculating
As in the previous section, a corollary to Thevenin and Nortons theorems is that if the nerwork
N has no internal independent sources, then the Thevenin and Norton equivalent circuit consists
o f a single resistance /?^yr. However, in contrast to passive networks, R^f^ can be negative. As a first
example illustrating the above theorems, we consider an active nervvork containing no internal
EXA M PLE 6.6 . Find the Thevenin equivalent circuit for the 2-terminal network (marked by
dashed line box) in Figure 6.15(a) using the method o f Section 3. (The dependent source acts as
a voltage amplifier.)
B
O(b)
FIGURE 6.15. (a) circuit with terminal voltage i/^ând input current /^;
(b)
/?, = /?,//
SO L u n o s
Step 1. Since there are no independent internal sources, the Thevenin equivalent consists o f a single
resistance,
i.e., v^^ = i^^= 0.
Step 2. Write a nodal equation. Writing a single node equation we have
Step 3. Match coefficients with equation 6.6. Matching coefficients implies that
= ((7j + (p +
1)G^) in which case,

^2
R^R
j A2
{ 1
J_ \
R2+(M + ^)R\
d 4.
R^
We recognize equation 6.14 as the parallel combination o f the resistance

To illustrate a typical calculation, suppose p = 199, /?j = 100
(6.14)
'(A^ + 1)
and _.^2
and Rj = 4 kf2.
Then
R.u = 500 / /20 = 19.23 a 20 =
iJ + \
Exercises. 1. For the above example, suppose p = 99, R^ = 500 Q, and /?2 = 1
A N SW ER: 10 Q
P*rid R^jy
244
2. For the circuit o f Figure 6.16, find the Thevenin equivalent resistance by obtaining
of
in terms
'AB-
FIG U RF 6.16 A circuit having no independent sources, in which case
0 and
the Thevenin equivalent consists only of a single resistance,
îh -
('jth
E X A M PLE 6 .7 . Find the Thevenin and Norton equivalent circuits seen at the terminals A-B in
Figure 6.17 when = 50 mA. Our computations will proceed using loop analysis to find the ter
minal v-i characteristic A-B.
5001
MCiURF 6.17 Arbitrary network for finding Thevenin equivalent.

S
o l u t io n
We first note that \.=
and
= 0.05 A.
Step 1. Write a set o f loop equations for the circuit o f Figure 6.17.
For loop 1, vvc obtain
0 = 1000(/j + i^) + 1000(/, - i^) - m i ^
which simplifies to
- 5 = 2000/, - 1500/;^
For loop A, we have
Chapter 6 * Thevcnin, Norton, and Maximum Power Transfer Theorems
Jn
Step 2. Write the loop equations in matrix form and solve. Writing the loop equations in matrix
form yields
2000
- 1 5 0 0 ' 'i\
-1 0 0 0
-50'
1100
Solving for i^ using for example Crammers rule produces

2000
del
IA=
det
-5 0 '
det
-1000
[2 0 0 0
-1 0 0 0
-1 5 0 0
-= -5 0 -
2000
-1 0 0 0
det
AB
700x10-
2000
()
-1 0 0 0
700x10-
1100
50
2
------- 1* AR-----700
700
(6.15)
or equivalently,
(6.16)
'/IS = 350/;^ + 25
Step 3. Match coejficietits o f equation 6.15 with equation 6.6 or equation 6.16 with equation 6.5 to
obtain
50
1
1
= = A, G,i, = =
S. R, = 350 Q, and v,,,. = 25 V
700
14
700
350
Exercises. I. In Example 6.7, if /j= 5 n-L\, find

I
140
A.
350
S.
R^f^, and
Hint: Use proportionalir)^
= 350 a a n d i ^ = 2.5 V
2. In Example 6.7, if /^= 5 mA and the 100 Q resistor is replaced by a short circuit, find
and
ANSWKR:
Hint: We have removed the dangling resistor in this case.

= 2.S \'. K,i, = 250 Q.G,,, =
S, and
= - ^ = 0 .0 1 A
3. Find the Norton equivalent at the terminals A-B o f the circuit o f P'igure 6.18 when mA.
AN SW ER:
= 200 LI and
= 0.1 25 A
8 0 0 I.
200 0
--- O
<
I
<
^ 800 0
800 Q
----- (
B
FIG U R H 6 .1 8 Modification o f the circuit o f Figure 6 .1 7 .
2-i6
5. TH EVEN IN AND NORTON EQ UIVALEN T CIRCU ITS FOR OP

AM P CIRCU ITS
Op am circuits arc active circuits. However, because the op amp is a device with special proper
ties, such as the virtual short circuit in the ideal case and such as output saturation in the non
ideal case, their discussion warrants special consideration. Our discussion begins with a Thevenin
equivalent o f a non-inverting amplifier with a dangling resistor at the output terminal.
EXA M PLE 6 ,8 . Find the Thevenin equivalent seen at the terminals A-B for the op amp circuit
o f Figure 6.19.
S o lution
Step 1. Find
2 0 -h 5
' Cfi Step 2. Find v^g. By inspection.
Step 3. Match coefficients with equation 6.5. Matching coefficients we obsen'e that
R.i. = 16 Q. V r =
, and
=
16
Our next example illustrates how to construct a negative resistance using an ideal op amp.
Chapter 6 *Thevenin, Norton, and Maximum Power Transfer Theorems
EXA M PLE 6 .9 . Find the Thevenin equivalent seen at the terminals A-B for the (ideal) op amp
circuit o f Figure 6.20.
lO k O
S o lution
By V-division and the properties o f the op amp,
VcB -
5 ,
Thus, computing /^j we have

.
ÂB - Vc
10x10^
lOxlO-"'
Matching coefficients with equation 6.6 we have

-4
S, R, = - 2 .5 kQ, and VV;,. = 0
10x10^
Exercise. For the circuit o f Figure 6.21, find the Thevenin equivalent circuit at A-B.
AN SW ER:
= 0 and R^,, = -R
Chapter 6 Thevenin, Norton, and Maximum Power Transfer 1 hcorems
:-i
Our third example constructs a Thevenin equivalent of the standard inverting op amp configura
tion with a terminal resistance. However, we will consider both the ideal and non-ideal cases.
EXA M PLE 6 .1 0 . Find the Thevenin equivalent seen at the terminals A-B for the op amp circuit
o f figure 6.22 when
(a) w'hen the op amp is assumed ideal, and
(b)
when the op amp has a saturation voltage,
= 15 V.
S o lution
Step 1. Fiuci the Theveniti equivalent seen at the terminals A-B assuming an ideal op amp. I'h e prop
erties of an ideal op amp imply that
On the other hand,
with set to zero,
ed into node A. Hence,
= 0 for all possible currents,

inject
-5V^ implies
= 0 and
= -5K^. The
= 0. In flict,
Thevenin equivalent seen at the terminals A-B consists only o f a voltage source o f value
=5
for the ideal op amp case.

Step 2. Find the Thevenin equivaloit seen at the termitials A-B assuming an op amp ivith output sat
uration. When the non-ideal op amp operates in its linear region, the Thevenin equivalent by Step
I is a voltage source having value v^^^ = - 5 V^. When, |-5 V} >
= 15 V , or equivalently, when
I
> 3 V, then the op amp saturates at 15 V. Specifically, when K > 3 V, then v^^ = - \5V and
when K, < - 3 V, then v^^ = 15 V. The Thevenin equivalent for an op amp with output saturation
is summarized in Figure 6.23, where v^^. takes on three separate values depending on the region o f
operation o f the amplifier.
1 5 V fo rV ^ < -3 V
- 5 V f o r |V j< 3 V
1 5 V fo rv / > 3 V
FIGURE 6.23 Thevenin equivalent at output terminals of an inverting amplifier (Figure 6.22)
with non-ideal op amp.
Chapter 6 Thcvcnin, Norton, and Maximum Power Transfer Theorems
249
This cnd-s our investigation of I hevenin equivalents o f op amp circuits. There arc many more
interesting examples that are beyond the scope o f this text.
7. M AXIM UM POW ER TRANSFER THEOREM

Figure 6.24 shows the Thevenin equivalent o f a network N connected to a variable load desig
nated Rj. 1'he load voltage, Vj, the load current, ij, and the power,
delivered to the load arc
all functions o f Rj. The main objective o f this section is to show that for fixed R^j^, maximum
power is transferred to the load when R^ = R^f^. We illustrate this assertion with an example that
shows the power delivered to Rj as a function o f Rj. Throughout this section, it is assumed that
all resistances are non-nesative.
FIG URE 6.24 Thcvcnin equivalent o f network N connectcd to a variable load, Rj.
E XA M PLE 6 .1 1 . For the circuit o f Figure 6.24, suppose that R^j^ = 20 Q and
= 20 V, Plot the
power delivered to the load R^ as a function o f

SO L U T IO N
The power delivered to the load Rj is
L
(6.17)
Plugging in the known values yields
IL =
^ (2 0 )'
To obtain the plot w'e use the following MA'FLAB code, resulting in the plot o f Figure 6.25.
voc = 20; Rth = 20;
RL = 0 :0.25:100;
PL = RL .* voc^2 ./ ((RL + Rth) .^2);
>>plot(RL,PL)
grid
Chapter 6 Thevenin, Norton, and Maximum Power Transfer 1 heorems
250
RL in Ohms
FIG URE 6.25 Plot of power delivered to the load in Figure 6.24 as a function of Rj.
= R^j^ = 20
From the curve, maximum power is transferred at
In a neighborhood o f R^ =
20 ^2, the curve remains fairly flat. At R^ = 40 Q. and Rj = \0 Q, the curve shows that about 88%
o f maximum power is transferred.
This experimentally observed fact, that maximum power transfer occurs when R^ =
plays an
important role when matching speaker resistances to the output resistance o f a stereo ampli
fier or when trying to get as much power as possible out o f an antenna and into a receiver.
M AXIM UM POW ER TRAN SFER THEO REM

Let a two-terminal linear network, N, represented by its Thevenin equivalent, as in Figure
6.24, be connected to a variable load, R^. For fixed R^j^, maximum instantaneous power is
transferred to the load when
=,/,
and the maximum instantaneous power is given by
rL.max
In the dc case, the instantaneous power is a constant for all t.
A verification o f the maximum power transfer theorem proceeds using differential calculus. From
equation 6.17, the power absorbed by the load is
2S1
R
Pl =
L____
VV? o c
[^L + ^th)
Following the standard procedure o f calculus for determining a maximum/minimum, we com
pute the derivative o f w i t h respect to Rj, set to zero, and solve for R^.
dPL ^ d
cJRl
(IRl
Rr
-9 .
oc
Rl v I c
= y2
/
\3
from which R^ = R^^^ and R^ = ^ are the only possible solutions. But, if R^ = oc, then
= 0.
Hence, because equation 6.17 is positive for /?^ > 0, /?^ = R^j^ produces maximum power,/>^, deliv
ered to the load.
Further, substituting
= R^f^ into equation 6.1 7 yields
Rfll
P L ,m a x
^n/(6.18)
This completes the verification o f the maximum pow'er transfer theorem.

E X A M PLE 6 .1 2 . Consider the circuit o f Figure 6.26a. Find (i) the value o f 7?^ for maximum
power transfer and (ii) the corresponding
.......N ................................
600 O
<
V -/
300 0
Thevenin
equivalent
(b)
FIGURE 6.26 (a) A network N connected to a load /?/-, (b) Thevenin equivalent of N connected to
2^2
SO L U T IO N
urrent
the independent voltage source becomes a short and the independent cur,
Step 1. To compute
= 200
source becomes an open. Finding the equivalent: resistance seen at the terminals produces
Hence, maximum power is transferred when
Step 2.
= 200
may be computed by any o f the methods discussed throi4ghoi4t this chapter. For example, by
repeated source transformations, the network N reduces to its Thevenin equivalent shown in Figure
6.26(b) with
Plugging
16 V. In fact, this approach would have found R^j^ and

16 V and
at the same time.
= R^^ = 200 Q into equation 6.18 yields
PU
vi
(1 6 )
4R,th
800
= 320 mW
Exercise. Suppose the 400 Q resistor in Figure 6.26(a) is changed to 100 Q. Find
R^j^, and pj
A N S W l'R S : 24 V. ISO 12, 0.96 wat t s
EXA M PLE 6 .1 3 . This example shows that the Thevenin equivalent cannot be used to calculate
power consumption within the network N it represents. For this demonstration, consider the net
work N given in Figure 6.27(a) with its Thevenin equivalent given in Figure 6.27(b). Compute
the power loss within the actual network N and within its Thevenin equivalent. We show that
these are different.
N.
-O1n
1A
20
3V
--Thevenin equivalent -(a)
(b)
FIGURE 6.27 (a) A network N; (b) Thevenin equivalent of (a).

S O L U T IO N
W ithin the network N, the power loss is
;
watts
PN, actual = 2 x 2 + 2 x 1 - = 10 w
Within
the power loss is
Pm
- 1 X 1- -
Chapter 6 Thevenin, Norton, and iMaximum Power Transfer Theorems
2^3
This means that theTIicvenin equivalent is not, in general, representative o f pov/er relationships
within the network, i.e., the losses that are dissipated as heat, for example.
When a network N is a voltage source in series with a resistance R^, and hence is its own Thevenin
equivalent, one may ask about maximum power transfer when
fixed, assuming
is variable and the load Rf is
is also fixed. The following example is an experiment for investigating this sit
uation.
E X A M P L E 6 .1 3 . For the circuit o f Figure 6.28, suppose /?^ = 20 Q and
delivered to the load as a function o f R^ along with the power loss,
FIGURE 6.28 A network N in which R^ can be adjusted with
= 20 V. Plot the power

in R^.
and /?^ fixed.
SO L U T IO N
The power delivered to the load R^ is
R,
(R
+ Rs )
0
vr =
20 X 400
(20+ / ?,)
To obtain the plots, we use the following MATLAB code, resulting in the plot o f Figure 6.29.
vs = 20; RL = 20;
..Rs = 0:.25:50;
>>PL = RL .* vsA2 ./ ((RL + Rs) .^2);
plot(Rs,PL)
grid
hold
Ploss = Rs .* vs'^2 ./ ((RL + Rs) .'^2);
plot(Rs,Ploss,b)
Chapter 6 Thcvcnin, Norton, and Maximum Power 1 ransfer Theorems
25-1
FIGURE 6.29 Plot of power delivered to load as a function of
for circuit o f Figure 6.28.
According to Figure 6.29, the maximum power o f 20 warts is delivered when

if R^< R^ (the usual case), then mi n i m i z i n g m a x i m i z e s
Ploss
= 0. Observe that
However, if R^ > Rj, minimizing
niaximizep^.
The proof for the maximum power transfer theorem given earlier considers Rj as the independ
ent variable and sets
dRi
to zero, standard practice in calculus. There is, however, an alternate
approach whose derivation is simpler mathematically, but is more meaningful for applications in
the sense that the load can be a general 2-terminal linear network, N^, instead o f a single resistor.
For this alternate derivation, refer to Figure 6.30. We ask the question. What v-i characteristic
should the load network
have so that maximum power is transferredfrom N to N J
FIGURE 6.30 The Thcvcnin equivalent of a network N connected to a loading ncrvvork N^.
To find the value o f v, we note that the power
= y i= y l^ - V
R,th
To find the value o f v that maximizes
transferred from N lo Nj^ is
we differentiate
with respect to v and set the result to zero:
lIPL
dv
ôc R,i,
Solving for v yields
V = 0.5v
(6.19a)
at which value
(6.19b)
ih
th
which are the conditions on i/and i for maximum power transfer to

resistor
If
consists o f a single
it has the v-i characteristic o f equations 6.19. Then from O hm s law,
^ L - ~ - îh
I
At V= 0.5v , the corresponding maximum power is
( 6 .20 )
PLmax = V X /= -
4/?,th
EX A M PLE 6 .1 4 . In the circuit o f Figure 6 .31,
and R^f^ = 2 0.. Find the value o f that
maximizes power transfer to the network N^.
S o lution
According to equation 6.19(a), maximum power transfer occurs when v =

equation 6.19(b), i =
= 0.5 A. Thus,
^th
V= \ = 0 .2 / +
= 0.\ + Vj
=o
Vf = 0 .9
= 1 V and from
2S 6
Chapter 6 Thcvcniii, Norton, and Maximutn Power Transfer'Ihcorcnis
Exercises. 1 If the 0.2 2 resistor is changed 4 Q, find the value o f
that maximizes power trans
fer to the network N^.

ANSWHR; -1 V
2. If the 0.2 Q resistor is changed 2 Q, find the value o f Vj that maximizes power transfer to the
network N^.
AN SW ER: 0 V
3. IF the 0.2 Q. resistor is variable and Vj = 0.5 V, find the new' value of the 0.2
resistor that
maximizes power transfer to the Nj.

AN SW ER: 1 Q
8. SUM M ARY
This chapter has set forth a powerful strateg)' for analyzing complex networks by replacing portions o f the nerwork by their simpler Thevenin and Norton equivalents. The Thevenin and
Norton theorems assure us that almost any 2-terminal linear nerwork, no matter the number o f
internal elements, is equivalent to a simple nerwork consisting of an independent source either in
series with or in parallel with a resistance. O f course, an independent current source does not have
a Thevenin equivalent, and an independent voltage source does not have a Norton equivalent.
\4ore generally, there are some circuits that have one but not the other. Further, some circuits have
neither.
'
rhe chapter has illustrated various techniques for constructing the Thevenin and Norton equiv-
alents. For passive networks, the ordinary approach is to find
first by deactivating all internal
independent sources. If the resultant circuit is series-parallel, then
can be found by combin-
ing series and parallel resistances as learned in Chapter 2. If the resultant nerwork is not seriesparallel, then one should use the main technique set forth in this chapter, which is to find the vi characteristic o f the terminals. This technique is valid for all circuit t\'pes.
With the ideas o f a Thevenin and Norton equivalent circuit, we then investigated the problem of
transferring power to a load. When R^f^ is fixed, maximum power is transferred when R^ is adjust
ed to be R^j^. If
is adjustable and R^ is fixed, then maximum power is transferred when R^j^ =
0. It is important to imderstand that a practical dc voltage source (such as a battery in an auto

mobile) is designed to provide nearly constant output voltage for the intended load current.
Accordingly, it has a rather small source resistance R^. Any attempt to transfer the maximum
power from such a source continuously will overload the source and may cause damage to its
internal structure. For example, in a lead acid battery, the plates may warp or the solution bub
ble. Hence, maximum power transfer is not o f critical importance for power transmission net
works, whereas for communication networks, maximum power transfer is important.
Chapter 6 Thcvcnin. Norton, anil iMaximum Power Transfer Theorems
25'
9. TERM S AND C O N CEPTS

2-term inaI network: an interconnection o f circuit elements inside a box having only 2 accessible
terminals for connection to other nervvorks.
D eactivating an independent current source: replacing the source by an open circuit.
D eactivating an independent voltage source: replacing the source by a short circuit.
Equivalent n-term inal networks: t\vo n-terminal networks having the same terminal voltagecurrent relationships. Alternately, two n-terminal networks N j and N-, are equivalent
when substituting one for the other in every possible network N; the voltages and cur
rents in N are unaffected,
the current through a short circuit placed across the output terminals of a 2-terminal network.
M aximum Power Theorem : let an adjustable load resistor
be connected to the Thevenin
equivalent o f a 2-terminal linear network. Maximum power is absorbed by the resistor

when Rj = Rî^.
N ortons equivalent circuit: any 2-terminal net\vork consisting o f independent sources and lin
ear resistive elements is equivalent to an independent current source in parallel with a
resistance.
R(h (Thevenins equivalent resistance): the resistance that appears in the 7'hevenin equivalent
circuit o f a 2-terminal linear network. It is also the equivalent resistance of the 2-terminal net\vork w'hen all internal independent sources are deactivated.
Thevenins equivalent circuit: any 2-terminal network consisting o f independent sources and lin
ear resistive elements is equivalent to an independent voltage source in series with a resist
V
ance.
: the open circuit voltage o f a 2-terminal network N when no load is connected.
' For a generalization o f this condition to the case where the controlling voltage or current is out
side o f N, see the article by Peter Aronheim entitled Frequenc)' Domain Methods in The Circuits
and Filters Handbook, BocaRaton, FI.: C R C Press, 1995, pp. 682-691.
258
Prob ems
= 12 V,
= 60 Q, /?2 = 60 Q, and R^ = 40 Q.
0.4 A,
(a) Find the Thevenin and Norton equiva
TH EVEN IN /N O R TO N FOR
PASSIVE CIRCU ITS
lents seen at the terminals A-B.

(b) If a load resistor o f 90 Q is connected to AB, find the power absorbed by this resistor.
1. For the circuit o f Figure P6.1, find Rî^,

and
in terms o f the literals. Hint: Consider
(c) Repeat (b) for a 30 Q resistor. Which resis

tor, 30 Q. or 90
absorbs die most power?
using C/ = .
R.
Figure P 6.1
Figure P6.4
2. Find the Thevenin and Norton equivalent

circuits seen at the terminals A-B for the circuit
depicted in Figure P6.2.
5. In the circuit o f Figure P6.5, R^ = 2 kH, Rj
= 8 k n , Ri = 6 kQ,
= 60 V, i^ 2 - ^9 mA and
= 5 mA. Find the Thevenin and Norton

equivalents o f the circuit in the dashed box.
Then find ij and the power absorbed by R^.
5kQ
-o-r
Figure P6.2
-O -v
3. For the circuit o f Figure P6.3, /?, = 3 kH, Rj

= 6 kD,
= 30 V, and /p = 10 mA.
(a) Find the Norton and Thevenin equiva

lents.
(b) Suppose a variable load resistor
attached
across
M ATLAB
A-B.
Plot
is
using
or equivalent the power
Figure P6.5
C H E C K : 200 V, 10 kQ
R.
^(l)
' 6
Figure P 6 .6
Figure P6.3
ANSWF.R: /?;/; =2 kil, isr = 20 mA, // . = 40 V
= 48
V. Find the Norton and Thevenin equivalents.
absorbed by R^ when \00 < R^< 4 kH.
- 6
= 18 kH,
/?2 = 9 kQ, ^3 = 3 kD, R^ = 6 kQ, and
CH ECK:
= 2 mA
Chapter 6 Thcvcnin, Norton, and Maximum Power Transfer Theorems
2S9
7. Find the Thevcnin equivalent seen at A-B o f
(c) Find the value o f Rj for maximum
the circuit o f Figure P6.7, where R, = 18 kH, R-,
power transfer and the resultant power
= 9 kD, /?3 = 3 kD. /?4 = 6 kD.
= 3.6 kQ.
= 32 k n .
= 48 V, and s2 = 8 mA. Hint: Use

the result o f Problem 6 to find the Thevenin
equivalent of the network between C and D.
delivered to the load.

(d) If the value o f
is doubled, what is the
power delivered to the load under the

condition o f maximum power transfer?
R,
' 6
8R
6R
Figure P6.9
C H ECK S: (a) 900 Q, 30 V; (c) 250 mW, (d) 1
Figure P6.7
C H EC K : R.th = 10 kD
10. Find the Thevenin equivalent seen at A-B

o f the circuit o f Figure P6.10. Hint: For this
t)'pe o f problem, the more natural solution
8. (a) Find the Thevenin equivalents for the
technique is source transformations. Why?
circuit o f Figure P6.8 in terms o f the lit

erals
and i^2 (b) If the A-B is terminated in a 15 kl^ load,
and
= 30 V, and 1^2 =
power delivered to the load.
(c) What is the proper resistance across the
terminals A-B for maximum power
transfer and what is the resultant power
Figure P6.10
delivered to the load?

ANSW'ER:
= 2R,
V()
= +
11. Find the Thevenin equivalent o f the circuit

o f Figure P 6 .ll enclosed in the dashed-line
box. Then compute
and the power absorbed
by the 2 kQ resistor. Hint: W hat resistances are

extraneous to the solution?
Figure P6.8
A N SW FR: R^f, = 10 k ii,
= 204 V
9. Consider the circuit Figure P6.9, in

which v^= 120 V and R = 300 Q.
(a) Find the Thevenin equivalent circuit
to the left o f the terminals A-B.
(b) For Rj^ = 300 Q, 600 iX and 1200 Q.,
find die power absorbed by R^. Does
the use o f a Thevenin equivalent reduce the
effort needed to obtain these answers?
Figure P 6 .11
A N S W E R : 5 2 V. 2 4 k X 2 mA
Chapter 6 Thev enin, Norton, and Maximum Power Transfer Theorems
260
12.(a) Find theThevenin equivalent circuit for

the 2-terminal non-series-parallel net
work shown in Figure P6.12. Use the
1 kn
2kO
general method.
(b) If a load resistance
is connected to
terminals A-B, use MATLAB to calcu

late and plot the power absorbed by the
Figure P6.14
load for 1/ = 30 V, and 10 < /?^ < 200 ^2

in 5
steps. At what value o f
is max
15. (a)
imum power achieved?
= 3 kH for the circuit
Find a so that
o f Figure P6.15.
(b)
100 o
100 Q
Repeat part (a) so that
R^^^ = -1 kQ.
Hint: Do problem 14 first, and then
100 Q
modify the Thevenin resistance appro
200 Q
priately.
C H EC K : (b) a = 4000 Q
i.
Figure P 6.12
ANSW FR: For (a),
------O
= 1(H) LI
1 kO
1 kO
2kn
TH EVEN IN /N O R TO N FOR
A CTIVE CIRCU ITS
13.(a) Find the value o f
so that the
Figure P6.15
Thevenin equivalent resistance o f the

16.(a)
circuit shown in Figure P 6 .13 is 5

(b) Repeat part (a) for the case when
- 2 5 0 Q. C H EC K : (b) 6.25 mS
Find a so that R^^^ = 5 kQ for the cir

cuit of Figure P6.16.
(b)
Repeat part (a) so that
200 n
1 kn
2k O
800 n
A N SW FR:
= -1 kl^ .
Figure P 6.13
= 1000 uS
Figure P 6 .16
17. For the circuit shown in Figure P6.17, find
the Norton equivalent circuit.
14.(a) Find a so that G,/, =

S for the circuit
^ ih
o f Figure P6.14.
(b) Repeat part (a) so that G,/, =
200 n
= -0 .0 0 1
A
7/l
18kQ
Figure P 6 .17
ANSWFR: - 6 0 0
/ = 0
261
18, Use loop analysis to compute the Thevenin

equivalent for the circuit shown in Figure
2 1 .(a) Find the Thevenin and Norton equiva

lent circuits for the network shown in
Figure P6.21, assuming that k = 0,025 S
P6.18, What is the Norton equivalent?
and
= 20 V,
(b) For what value o f k is the open circuit

voltage zero. For this value o f k, deter
100 0
-----1----- -
loon
mine R^fj.
'300 0
' 0.01
0>
800 n
ANSW I-R;
Figure P 6,18
= 0,
= 250 12
19.(a) Find the Norton and Thevenin equiva

lents o f the circuit o f Figure P 6 ,19,
(b) If a load resistor
output
is attached across the
terminals,
plot
the
AN SW ER:
Figure P6.21
and
= 60
= 18 V
power
absorbed by the load for 1 <^Rj < 24 Q

usingMATLAB or equivalent. For what
22. For the circuit shown in Figure P6.22, b =

- 0 .0 2 S and a = 25
Find the Norton equiv
value o f
alent circuit.
does the load absorb maxi
mum power? Determine the power

delivered to the load at maximum power
transfer.
i.j
50 0
V. <v 50 O
'
50 0
bv
Figure P 6.19
A X 'S W I- .R :/'DC=6/<. / w' = U Mf, P,//
Figure P6.22
W SW I'R; R^,, = 100 12.
= 50/,
20. Find the Thevenin equivalent o f the circuit
in Figure P6.20 where
P6.23.
(a) Find the Thevenin equivalent.
= 0.2 A.
(b) If a load resistor R^ is connected across

400 Q
' 0
1 kO
'400 0
200i
terminals A-B, determine R^^ for maxi

mum power transfer and determine the
maximum power delivered to R^.
(c) If a resistor /?, were added in series with
terminal A o f figure P6.23, what is the
Figure P6.20
AN SW FR: r^. = 60 \. A,;, = SOO LI
Thevenin equivalent resistance o f the

augmented network.
2(>:
ANSW ER: (a) R,i, = R^
' 6
Figure P6.23
27. Find the Thevenin and Norton equivalents of
2 4 .(a) Find the Thevenin equivalent for the

network shown in Figure P6.24.
(b) If the values of each source are cut in
the circuit o f Figure P6.27 when
= 20 mA.
AA
half, what is the new

ANSWER: (a) 1.6 kLl - 260 V: (b) / = - 130 V
400 n
2000i
io o v (^ ^
0.1 A
ANSWI-:R:
= 12.S V,
= 650 Q
Figure P6.24
25. Find the Norton equivalent for the circuit
shown in Figure P6.25 when = 30 niA,
=
0.04 S, /?j = 100 Q and
= 400 n .
28. Find the Thevenin equivalent seen at A-B

for the circuit o f Figure P6.28 when /^= 10 mA,
= 24 0 0 Q and /?2 = HOOO Q, R^ = 5600 Q,
R^ = 1500 n , R^ = 1000 Q, and
= 0.25 x
10-^ S.
C H EC K :
Figure P6.25
0.09 A
26. Consider the circuit o f Figure P6.26, where

= 32 V, /?, = 80 Q, /?2 = 240 Q,
and
(a) Replace the circuit to the left o f nodes A

and B with its Thevenin equivalent.
(h) Given your answer to (a), assume that Rf
= 150 n , and find and the power con
C H EC K : R^,^ = iQ Q.
Figure P6.28
= 8 mA
= 60 Q,
= 2.
sumed by Rj.
CH ECK:
OP AM P PROBLEM S
2 9 .(a) Find the Thevenin equivalent seen at the
terminals A-B for the op amp circuit o f
Figure P6.29.
(b) What is the value o f a load resistor /?^
263
C H EC K :
= 5000 Q
3 2 .(a) Find
the
attached across the terminals A-B for

maximum power transfer. What is the
power absorbed by this Rj}
Thevenin
and
Norton
Equivalent circuits o f the op amp con

figuration o f Figure P6.32 seen at A-B.
(b) Repeat (a) for the terminals C-B.
Figure P6.29
C H EC K : /?,/,=/?3
30. (a) Find the Thevenin and Norton equiva
lents seen at the terminals A-B for the op amp
circuit o f Figure P6.30.
Figure P6.32
AN SW ER: ( b ) = A*,,
=0
3 3 .(a) Find
the
Thevenin
and
Norton
Equivalent circuits o f the op amp con

figuration o f Figure P6.33 seen at A-B.
(b) Determine the value o f a load resistor R^
connected across the terminals A-B for
maximum power transfer. If
= 4 V
and ryp = 5 V, determine the maximum

power transferred to this R^.
C H E C K : (b)
1 ( /?! + /?2
rA
20 kQ
= 0.9 watts
50 kO
3 1.(a) Find the Thevenin equivalent to the

right o f the terminals A-B for the (ideal)
op amp circuit o f Figure P6.31.
(b) If the practical source indicated in the
figure is attached to A-B, find the current
in terms o f
15 kn
34. Find the Thevenin equivalent seen at the

terminals A-C for the op amp circuit o f Figure
P 6.34 when the op amp has output saturation,
>5 V.
Chapter 6 Thcvcniii, Norton, and Maximiun Power Transfer Theorems
2 64
N
Linear resistive
network with
dependent sources
and fixed
independent
sources
Power
Supply
Figure P6.36
Tabic P6.36
3 5 .(a) Find the Thevenin equivalent seen at the
terminals A-C for the op amp circuit of
/^l (mA)
Figure P6.35 when the op amp has out

put saturation,
12
= 12 V.
(b) Find the Thevenin equivalent seen to the

left o f the terminals B-C and the maxi
C H E C K : ( b )P ,,,= 2 m W
mum power that will be absorbed by the

24 k n resistor for all variations in V..
37. Repeat Problem 36 with the data given in

Rible P6.37.
Table P6.37
r j (mA)
Figure P6.35
C H EC K :
/W =
144
2800
10
54
40
66
C H E C K : (b)
.5625 W
= 6 mW
THEVEN IN AND NORTON

EQUIVALENTS FROM
M EASURED DATA
36. In a laboratory, the data set forth in rows I
and 2 o f Table P6.36 were taken.
(a) Compute the Thevenin and Norton
equivalents o f N.
38. The data listed in Table P6.38 was taken for

the network N o f Figure P6.38.
(a) Fill in the values for the third column o f
Table P6.38 and find the Thevenin and
Norton equivalents o f the linear resistive
nervvork N.
(b) To what resistance should
be changed
to achieve maximum power transfer?

What is P..
(b) After the power supply is removed, what

resistance, Rj, should be connected
Table P6.38
across A-B for maximum power transfer?
(niA )
What i s />?
2
10
10
j>
Chapter 6 lhc\xnin, Norton, and Maximum Power Transfer Theorems
26S
41. This problem is the first of t%vo problems
that outline a laboratory measurement proce

dure for finding the 'Fhevenin equivalent o f a
Linear resistive
network with
dependent sources
and fixed
independent
sources
linear resistive 2-terminal nerwork. For this

problem, consider Figure P6.41 in which the
circuit under test contains no independent
sources. I'he experimental apparatus includes a
C H EC K :
Figure P6.38
= 10.667 mW
resistance decade box, denoted R, a dc volt

meter with internal resistance
and a signal
generator having known internal resistance, R_.

39. Repeat Problem 38 using the data in Tible
P6.39.
To begin the procedure, one sets R = R^ = 0 and

adjusts the dc level,
o f the signal generator
to obtain a reasonable meter reading, say

Table P6.39
R, (2)
q, where the subscript 1 indicates our first

(niA)
meter reading. (For an analog meter, the read

ing should be almost full scale.) Leaving the sig
200
>
nal generator set at this value o f V^, increase R
1200
>
until the meter reading drops to

Record this value o f R as R^.
^''o-
(a) Suppose R^^^ = x and R^ = 0. Show that

CH ECK:
31.25 m\V
,/, = 2(b) Now suppose R^^^ = x> and R^ ^ 0. Show
40. The data listed in Table P6.40 was taken for

the network N o f Figure P6.40 with a volt
meter (VM ) whose internal resistance is 10
M n . Fill in the values for the third column o f
Table 6.3 6 and find theThevenin equivalent o f
the linear resistive network N.
that/^,/^=y?2(c) F'inally, suppose R^^ând Râre nonzero and

finite. Show that
and then solve
^ th
for R.th-
= R j - R^
^m
Table P6.40
(pA)
R, (M Q)
2
0.4
>
10
>
N
Linear resistive
network with
dependent sources
and fixed
independent
sources
42. 'Fhis problem is the second of rwo problems

that outline a laboratory measurement procedure
for finding the The\'enin equivalent of a linear
resistive 2-terminal net\vork. For this problem,
consider the new configuration o f Figure P6.42 in
which the circuit under test contains independent
Figure P6.40
=4 V
sources and has a non-zero
'Fhe experimental
apparatus includes a resistance decade box, denot

ed R, and a dc voltmeter with internal resistance
Chapter 6 Thcvenin, Norton, and Maximum Power Transfer Theorems
266
All devices are connected in parallel. Because

?= 0, a sign;il generator is not needed, as in
Problem 41. l b begin the procedure, open circuit
AN SW ER: 4 WLl and 20 V

45. The linear circuit shown in Figure P6.45 is
= co,
found experimentally to have the voltage and
and set the scale on the voltmeter to obtain a rea
current relationship shown. Find its Thevenin

Norton equivalent.
the decade box, or equivalently set /? =

sonable meter reading, say
= q, where the
subscript 1 indicates our first meter reading.

(For an analog meter, the reading should be
almost full scale.) Next, reconnect the decade and
decrease R until the meter reading drops to V^p =
0.5
Record this value o f R as /?2.

(a) Suppose R^^^ = co. Show that R^f^ = R-,.
(b) Suppose R^^^ is nonzero and finite. Show
R,R
that
= /? 2 înd then solve for
R ,l,+ R ,
R^!^. Then show that
1+
Rm/
'0
ANSWT.R: 0.5 Q, 4 A
46. Repeat Problem 45 for the measurement
curve shown in Figure P6,46, Then determine
the value of a load resistor for maximum power
transfer and compute
.
Figure P6.42
43. The Thevenin equivalent o f a linear resistive
network containing no independent sources is to
be found experimentally using the method o f
Problem 41. fhe voltmeter has an internal resist
ance R^^^= 20 k n . The dc signal generator has an
internal resistance, R^ = 2 kI2. The following
measurements are taken: (i) with R = 0,
is
adjusted until the voltmeter reads 4 V; (ii) keeping
fixed, the decade box is adjusted until the
voltmeter reads 2 V. For this voltage, the decade

box shows R = 6 kf2. Find R^/^.
AN SW FR: 5 kLl
47. The i-v curve o f the network N in Figure

P6.47a is measured in a laboratory, and is
approximated by the straight-line segments
shown in Figure P6.47b. The meter readings
are shown in Table P6.47.
Table P6.47
44. The Thevenin equivalent of a linear resistive

network containing independent sources is to be
found experimentally by the procedure o f
Problem 42. The voltmeter has an input resistance
R^j = 1 M ti. The following me;isurements are

taken: (i) when R is opcn-circuited, the voltmeter
reads 4 V, and (ii) when R is decreased to 800 kl^,
the voltmeter reads 2 V. Find R^jând
0.2 V
0.1 mA
0.7 V
10.1 mA
(a) Find the Thevenin equivalent for the

range 0 < i < 0.1 mA.
(b) Find the Thevenin equivalent for the
range 0.1 < / < 10.1 mA.
Chapter 6 rhcvcnin, Norton, and Maximum Power Transfer Theorems
(c) If R = 500 Q, Vît) = 50 sintdOOO t) mV,

100 mV, find i{t). Hint: Use a
and
dated voltage
267
and the power delivered to
the load.
suitable Thevenin equivalent for N.

(d) If R = 50
and
v^{t) = 200 sintdOOO t) mV,
= 500 mV, find /(/).
R
scale)
' 6
Figure P6.50
(a)
Figure P6.47
ANSWl-.R: (a)
= 0.
= 2 kl2; (h)
51 (a) For the circuits o f Figure P6.51, find the

load resistance R^ needed for maximum
0.195V, R,/, = 5 o 'h ; (c) 0.04 + 0.02 sin (1000
power transfer, the associated voltage

and the power delivered to the load.
/) mA: (d) .^.05 + 2 sindOOO t) mA
(b) If the load resistance is constrained as 5

k n < R^ < 10 kD, repeat part (a).
M AXIM UM POW ER TRANSFER
(c) If the load resistance is constrained as 15
48. For the circuit o f Figure P6.48, /?, = 160 Q.,
R-, = 480
and
k n < R^ < 20 k n , repeat part (a).
= 80 V. Find the value o f
for maximum power transfer and
e
2 mA
12 kO
8kn
24 V
6kfi
Figure P6.48
C H EC K : 7.5 watts
Figure P 6.51
49. For the circuit o f Figure P6.49, R^ = 900 Q,
R j= 180
/?3 = 50 Q,
21 V. Find the value o f
= 60 mA, and v^ 2 =
for maximum power

(a) Find the value o f
for maximum
power transfer to the three-resistor load.
transfer and
(b) Find the power delivered to each load

resistor, i.e., to R^, R J2 , and /?^/3.
40 V
Figure P6.49
lon
50. For the circuits o f Figure P6.50,
= 10 V
and v^ 2 - ^5 V. Find the load resistance R^^

needed for maximum power transfer, the asso-
Figure P 6 .5 2
Chapter 6 I'hevcnin, Norton, and Maximum Power transfer Theorems
268
53. For the circuits o f Figure P6.53, /?j = 200

= 1000 Q, /?3 = 400 n ,
>^ v.,(V)
= 8 mS, and /;, =
0.4 A. Find the load resistance
80 -
needed for
<
maximum power transfer, the associated voltage,
--- 0 +
a
and the power delivered to the load.
o b
40 -
(a)
>
1
0.2
1
1
0.4
i(A)
^
r
(b)
Figure P6.56
57. (a)
Figure P6.53
For the circuit o f Figure P6.57 com

pute, (i) the value o f R which leads to
maximum power transfer to the load,
54. T he circuits o f Figure P6.54 have the load
(ii) the voltage across the load, and
resistor
(iii) the power absorbed by the load.
connected in different ways. For
each circuit, (i) compute the value o f
which
Hint: In MATLAB, the roots o f a
leads to maximum power transfer, (ii) the volt
quadratic, aQ x- + a, x + a2> are given
age across the load, and (iii) the power

absorbed by the load. Which configuration
absorbs more power?
by roots([aO al a2]).
(b)
To verify the results o f (a) write a

MATLAB program to calculate and
plot the power absorbed by the load as
son
R varies from 0 Q to 400
15Q
30 V
15
in 2 Q
increments.
^ (b
Load
(a)
R.
30 Q
(b
+ V
15Q
.s v Q
30 V
Figure P6.57
(b)
C H EC K : (a)
Figure P6.54
10 watts
55. Suppose the polarity o f the 15-V-source in

Problem 54 is reversed. Repeat Problem 54 and
determine which configuration transfers more
power to the load.
58. The i-v relationship o f certain type o f LED

(light emitting diode) in its operating range o f
1-7 V -3 V is represented by a 2 V voltage-source
in series with a 50 H resistance. The load con
sists o f a network o f n such diodes connected in
parallel. The source network is represented by a
5 V voltage-source in series with a 50 ^2 resist
ance. Assume that the power delivered to each
diode is totally converted into light. Determine
56. T he linear resistive circuit o f Figure

P6.56(a) is found experimentally to have the
voltage-current relationship plotted in Figure
P6.56(b). Find the maximum power that can
be absorbed by placing a load resistor across ter
minals a-b?
how' many LEDs should be connected in paral

lel for maximum brightness. W hat is power dis
sipated by each diode?
ANSW ER: M - 5,
= 2S mW
C H A P
Inductors and Caoacitors
CAPACITIVE SM O O TH IN G IN POW ER SUPPLIES

Every non-portable personal computer contains a power supply that converts the sinusoidal volt
age o f the ordinary household outlet to a regulated dc voltage. Regulated means that the output
voltage stays within very tight limits o f its nominal value (e.g., 12 0.1 V) over a wide range o f
power requirements. Engineers design power supply circuits with regulators that produce voltages
with a small oscillation because to generate a truly dc voltage is practically impossible.
REGULATION
RECTIFICATION
O
O
SMOOTHING
This process o f converting ac to dc has three stages: First, the ac waveform is rectified into its
absolute value. Then a smoothing operation takes place that reduces the variation in the voltage
to a reasonable but still unacceptable level. This first level o f smoothing is nccessar) becausc the
voltage regulator is a precision subcircuit that requires a fairly constant voltage for its proper oper
270
Chapter 7 Inductors and Capacitors
ation. The partially smoothed waveform is fed into a voltage regulator, which limits the voltage
oscillation between critical levels even when the load drawn by any connected device (e.g., your
computer) varies in the course o f its operation.
As mentioned, the rectified sine wave is smoothed before entering the voltage regulator. A crude
smoothing can be accomplished with a capacitor, a device studied in this chapter. Intuitively,
capacitors resist voltage changes and are designed to steady the voltage at a constant level. In this
chapter, we will study the capacitor and investigate a simplified smoothing operation for a power
supply.
CH APTER O BJECTIVES
1.
Define the notion o f inductance and introduce the inductor, whose terminal voltage is
proportional to the time derivative o f the current through it.
2.
3.
Investigate the ability o f an inductor to store energy and the computation o f the equiva
lent inductance o f series-parallel connections.
Define the notion o f capacitance and introduce the capacitor, whose current is propor
tional to the time derivative o f its terminal voltage.
4.
Investigate the ability o f a capacitor to store energy and the computation o f the equiva
lent capacitance o f series-parallel connections.
5.
Define and illustrate the principle o f conservation o f charge.
CHAPTER O U TLIN E
1.
2.
Introduction
The Inductor
3.
The Capacitor
4.
5.
6.
Series and Parallel Inductors and Capacitors

Smoothing Property o f a Capacitor in a Power Supply
Summary
7.
8.
Terms and Concepts

Problems
This chapter introduces two new circuit elements, the linear inductor and the linear capacitor,
hereafter referred to as an inductor and a capacitor. The inductor, shown in Figure 7.3, is a device
whose voltage is proportional to the time rate o f change o f its current with a constant o f propor
tionality I , called the inductance o f the device, i.e.
as set forth in equation 7.1. The unit o f the inductance Z., is the henry, denoted by H.
Macroscopically, inductance measures the magnitude o f the voltage induced by a change in the
current through the inductor.
The capacitor, shown in Figure 7.15, is a device whose current is proportional to the time rate o f
change o f its voltage, i.e.,
ic(0= C
dvcO)
(it
as set forth in equation 7.5. Here, the constant o f proportionality, C, is the capacitance o f the
device with unit farad, denoted by F. Capacitance measures the devices ability to produce a cur
rent from changes in the voltage across it.
By adding the inductor and the capacitor to the previously studied devices (the resistor, inde
pendent and dependent sources, etc.), one discovers an entire panorama o f possible circuit
responses, to be explored in the next four chapters. Together, these devices allow one to design
radios, transmitters, televisions, stereos, tape decks, and other electronic equipment. In this chap
ter, our goal is to understand the basic operation o f inductors and capacitors.
2. TH E IN DU CTO R
SomePhysics
In Figure 7.1, a changing current flowing from point A to point B through an ideal conductor
induces a voltage
between points A and B according to Faradays law. Joseph Henr}' inde
pendently observed the same phenomenon at about 1831. The induced voltage,
to be proportional to the rate o f change o f current, i.e.,
was found
= .
dt
FIG URE 7.1 A time-varying current flowing through an ideal conductor.

The following experiment illustrates the idea. Suppose the conductor in Figure 7.1 is 6 feet o f #22
copper with resistance 16.5 Q/1,000 ft. The 6-foot length has a resistance o f about 0.1 U. Using
a current generator, we apply a pair of ramp currents (shown in Figure 7.2a) to the conductor, as
per Figure 7.2b. The measured responses are shown in Figure 7.2c and, as expected, satisfy Ohms
law.
Chapter 7 * Inductors and Capacitors
(a)
(b)
(c)
(d)
(e)
FIGURK 7.2 (a) Ramp currcnt inputs to iincoilcd and coilcd wire, (b) Six feet of #22 wire attached
to a current generator, (c) Voltage responses to ramp current inputs of uncoiled wire.
(d) Six feet o f #22 wire coiled into 45 turns 1 long and 1 in diameter.
(e) Voltage responses to ramp current inputs o f coiled wire.
273
Now suppose the wire is coiled into a qrlinder 1 in diameter and 1 long, as in Figure 7.2d. Apply
the same ramp currents o f Figure 7.2a to the coiled wire. This time, the measured responses are as
shown in Figure 7.2e. These responses have the same shape as those of Figure 7.2c, except for the
offsets o f 30 mV and 60 mV, respectively. These offiet voltages are proportional to the derivatives
of the input currents, i.e.,
Offset -
for k = 1 , 2 , where L is the proportionality constant, called the inductance o f the coil. Since the
derivative o f i s lO'^ A/sec, and the derivative o f
the coil can be computed as
3x10
Offset
0.03
= j: = ^
10^
ini^t)
is 2 x lO'^ A/sec, the inductance
0.06
of
,
.
hennes
2x10^*
dt
As mentioned earlier, the heniy, equal to 1 volt-sec/amp and abbreviated H, is the unit of induc
tance. Also, from the above experiment, one concludes that the inductance o f a cylindrical coil o f
wire is much greater than the inductance of a straight piece of wire, which in the above experi
ment was not measurable by our apparatus.
The physics of the preceding interaction is governed by Maxwells equations, which describe the
interaction between electric and magnetic fields. A time-varying current flow through a wire creates a time-varying magnetic field around the wire. The magnetic field in turn sets up a time-varying electric field, i.e., an electric potential or voltage. One can verify the presence o f this magnetic field by bringing a compass close to a wire carrying a current. The magnetic field surrounding
the wire will cause the compass needle to deflect. Physically speaking, a changing current causes a
change in the storage o f energy in the magnetic field surrounding the conductor. The energy trans
ferred to the magnetic field requires work and, hence, power. Because power is the product of volt
age and current, it follows that there is an induced voltage between the ends of the conductor.
W hat is even more interesting is that if a second wire is immersed in the changing magnetic field
of the first wire, a voltage will be induced between the ends of the second wire. A proper (mathematical) explanation of this phenomenon is left to a fields course. For our purposes, three fects
are important: (1) energy storage occurs, (2) the induced voltage is proportional to the derivative
o f the current, and (3) the constant of proportionality is called the inductance of the coil and is
denoted by L.
As mentioned, a straight wire has a very small inductance, whereas a cylindrical coil o f the same
length o f wire has a much greater inductance. This inductance can be increased many times over,
possibly several thousand times, simply by putting an iron bar in the center o f a cylindrical coil.
Alas, the calculation o f inductance is the proper subject of more advanced texts, e.g., on field the
ory or transmission line theory. Nevertheless, there are empirical formulas for estimating the
inductance of a single-layer air-core coil as described in the homework exercises.
274
BASIC DEFIN ITIO N AND EXAM PLES

D EFIN ITIO N OF TH E LINEAR IN D U CTO R
The linear inductor, symbolized by a coiled wire as shown in Figure 7.3, is a two-terminal
energy storage device whose voltage is proportional to the derivative o f the current passing
through it. The constant o f proportionality, denoted by Z,, has the unit o f H enry (H), equal
to 1 volt-sec/amp. L is said to be the inductance o f the coil. Th e specific voltage-current rela
tionship o f the linear inductor is given by
(7.1)
dt
i,(t)
h/Y Y V
+ V jt) FIG U RE 7.3 The inductor and its differential voltage-current relationship
as per the passive sign convention.
EXA M PLE 7.1

Compute Vi{t) for the inductor circuit o f Figure 7.4 when ij{t) = e'~.
0.5H
FIG URE 7.4 A 0.5 H inductor driven by a current source.
S o lution
From equation 7.1, direct differentiation o f the inductor current /^(/) leads to
y^(t) = 0 . 5 - " '
dt
^ = 0 .5 (-2 f)e-''
Exercises. 1. In Example 7.1, suppose ijit) = 0.5sin(20r + 7t/3) A. Compute v^{t).

AN SW FR: 5 cos(20t + tt/3) V.
2. In Example 7.1, suppose ii{t) = (1 AN SW ER: ///(/) = lOOe--'*'V.
V for / > 0 and 0 otherwise. Find
t> 0 .
The differential equation 7.1 has a dual integral relationship. Safely supposing that at / =
inductor had not yet been manufactured, one can take
= 0, in which case
the
275
(7.2)
L-fh
The time
represents an initial time tiiat is o f interest or significance, e.g., the rime when a switch
is thrown or a source excitation is activated. The quantity
specifies the initial current flowing through the inductor at ^q. This quantity,
sums up the
entire past history o f the voltage excitation across the inductor. Because o f this, the inductor is said
to have memory.
EXA M PLE 7 .2
For the circuit o f Figure 7.5a, determine /^(O) and ij{t) for / > 0 when Vj{t) =
V as plotted in
Figure 7.5b.
\(t)
ijt)
,(t,
L = 0.5H
(b)
(a)
Inductor Current (A)
FICJURE 7.5 (a) Simple inductor driven by a voltage source, (b) Source waveform Vj{t).
(c) Resulting inductor current
1~ G
S olution
A direct application o f equation 7.2 leads to

I
T,
= z j l .
1
i / o " " ' ' = z + i
It follows that
il^(0) = = 2 A and //^(0 = 2 - e '

-
The graph o f ij\t) for all t is given in Figure 7.5c.
Exercises. 1. In Example 7.2, compute an expression for i^{t) for / < 0.

A N S V V K R : i j U ) - 2<-'' A tor t < 0.
2. Repeat Example 7.2 with L = H and with v,{t) = cos(27ir) V for r > -0 .2 5 sec and zero
I
.
4jt
otherwise.
ANSWl'.R: //(()) = 2 A, /y(/) = 2 + .sinUni) A for r > 0.
E X A M PLE 7 .3
Consider the circuit o f Figure 7.6a with voltage excitation v^{t) shown in Figure 7.6b. Find the
inductor current /^(r) for f > 0, assuming that /^(O) = 0.
i,(t)
,(t )
L = 0.5H
(a)
FIGURE 7.6 (a) Voltage source driving inductor, (h) Square wave excitation !(/).
S o lution
It is necessary to apply equation 7.2 to each interval, [0, 1], [1, 2], ... , [;;, n + 1], .... For this we
need to first specify the initial conditions for each interval.
Step 1. Compute i^\). From equation 7.2,
Chapter 7 Inductors atul Capacitors
Step 2. Cotnpute ii{2).

// (2 ) = /^(0) + f \'i^{T)dT = X Net Area = 0
Step 3. Compute the iuitial condition for the interval [n,u + 1] for u even. Again from equation 7.2,
with t = n and n even, we liave
1
Hence i^{n) = 0 for ail even values o f n.

Step 4. Compute the initial condition for the interval [n,n + 1] for n odd. From equation 7.2, with
t = n and n odd, we have, utilizing steps 1 and 3,

I
/I-1
+j
/j
f
ri-1
y i(r)d T = J
v7 (tV /t = 2
n-l
since n - 1 is even.
Step 5. Compute ij^t) over [n,n + 1] ivith n even. If n is even, then the value o f the inductor cur
rent over the interval [;/,;/ + 1] is
ilSt) = iL{n) + ^ cIt = i i{ n) + l { t - n) = 2{t - n) A
l^Jn
Observe that iît) = 2 t - In A is the equation of a straight line having slope +2 and^-intercept -In .
Step 6. Compute i^(t) over [n,n + 1] with n odd. If n is odd, then for the inter\'al [n,n + 1], the
inductor current is
/^(/) = //(/2) + ( dT = i i i n ) - 2{t - n ) = 2 - 2{t - n) A
[^J It
Here, i^{t) = 2 + 2n - 2t is the equation o f a straight line, with slope - 2 and )'-intercept 2+2.
Step 7. Piece segfnentsfivm steps 5 and 6 together. Thus the segments computed in steps 5 and 6 inter
cept the /-axis at the same points. Figure 7.7 sketches the resulting triangular response for /> 0.
FIGURE 7.7 Triangular shape o f inductor current for the square wave voltage excitation of
Figure 7.6b applied to the circuit of Figure 7.6a.
27H
Exercises. (All time is in seconds.) 1. Again consider the circuit o f Figure 7.6a. Compute iyr (?) for
(i) 0 < r < 1, (ii) 1 < t < 3 , and (iii) 3 <t For the waveform o f Figure 7.8a, assuming i^(0) = 0.
2. Again consider the circuit o f Figure 7.6a. Compute i^(^) for (i) 0 < r < 1, (ii) 1 < ^ < 3, (iii) 3 <
( < 4, and (iv) 4 < /, for the waveform o f Figure 7.8b, assuming i^(0) = 0.
F IG U R U 7.8 Voltage excitations for Exercises 1 and 2.
It is important to recognize that the square wave voltage input o f Figure 7.6b is discontinuous but
the current waveform o f Figure 7.7 is continuous. Integration (computation o f area) is a smooth
ing operation: it smoothes simple discontinuities. This means that the inductor current is a con
tinuous function o f t, even for discontinuous inductor voltages, provided that the voltages are
bounded. A voltage or current is bounded if the absolute value o f the excitation remains smaller
than some fixed finite constant for all time. Thus, equation 7.2 leads to the continuity property
o f the inductor: if the voltage Vf{t) across an inductor is bounded over the time interval /] < t <
tj, then the current through the inductor is continuous for
< t < tj. In particular, if
then /^(ô) =
^
The notation
and + on /q is used to dis
tinguish the moments immediately before and after /q- For example, in Figure 7.9, t = 2 shows a
discontinuity o f
The value o f *^^(2) is 1 and the value o f
is - 1 . The value Vj{2*) can
be seen as the limiting value o f z^^(r) when approaching r - 2 from the right, whereas Vf{2~) can
be seen as the limiting value o f v^{t) when approaching t
2 from the left.
279
\ {2 )
(U
Lu
T3
C
rtJ
01
CT>
TO
*->
o
>
u
O
c
Time (seconds)
F IG U R E 7.9 A possible discontinuous voltage v^{t) appearing across an inductor of 1 H,
and the resulting continuous inductor current.
PotverandEnergy
Rccall that the instantaneous power absorbed by a devicc is the product o f the voltage across and
the current through the device assuming the passive sign convention. For an inductor,
dt
Plit)=\'L{t)ilU) =
where
w atts.
is in volts, i^{t) in amps, and L in henries.
Since energy (absorbed or delivered) is the integral o f the instantaneous power over a given time
interval, it follows that the net energy
stored' over the inter\'al [/q, /,] in the magnetic
field around the inductor is

f'o V
(h
/
(7.3)
/(/,)-//^(/q ) joules.
for L in henries and

in amps. From equation 7.3, whenever the current waveform is bounded,
the net energy stored in the inductor over the interval [/q, rj] depends only on the value o f the
inductor current at times r, and /q, i.e., on //(^j) and //(/q) respectively. This means that the
stored energy is independent o f the particular current waveform between
and
If the current waveform is periodic, i.e., if ij{t) =
+ T) for some constant 7'> 0, then over any
time interval o f length T, the net stored energ)' in the inductor is zero because
= /^(/-q + 7)
Cliapter 7 Iiuluctors and Capacitors
280
forces equation 7.3 ro zero. To further illustrate this propert)', consider Figure 7.10a, which shows
a 0.1 H inductor driven by a periodic current /^(^) = sin(27tr) V. This current signal has a funda
mental period T = I, i.e., the smallest 7 'over which the signal repeats itselh From equation 7.3,
VV^(OJ) =
pi iOdt = I L/7(l)
L/7(0) = 0
However, we can interpret this result in terms o f the waveform o f pi{t). First note that the volt
age across the inductor in Figure 7.10a is Vjit) = 0.27Tcos(27if) V. Hence, the instantaneous power
is pf{t) =
= 0.2jtcos(2Tt^)sin(27tr) watts, as plotted in Figure 7.10b. Observe the shaded
regions o f Figure 7.10b in which the area under the power curve has equal parts ot positive and
negative area. This means that all the energ)' stored by the inductor over the part o f the cycle o f
positive power is delivered back to the circuit over the portion of the cycle when the power is neg
ative. Fhis is true for all periodic signals over any period. Because no energ)' is dissipated, and
because energy is only stored and returned to the circuit, the (ideal) inductor is said to be a loss
less device.
IlW
v jt)
sln(27T) A
0.1 H
= 0.2n cos(2n) V
(a)
P lW
(b)
FIGUllE 7.10 (a) Inductor excited by periodic current, (b) Plot of the power absorbed by the inductor.
It is convenient to define the instantaneous stored energy in an inductor as
(7 .4 )
281
for all t. Equation 7.4 can be viewed as a special case o f equation 7.3 in which r,) = -oo and /^(-oo)
= 0. Thus, equation 7.4 can be interpreted as the change in stored energ)' in the inductor over the
inten'al (^x>, t].
E XA M PLE 7 .4
Find the instantaneous energy stored in each inductor o f the circuit o f Figure 7.11 a for the source
waveform given in Figure 7.1 lb. In Figure 7.1 lb , note that ij^t) = 0 for r < 0.
FIGURE 7.11 (a) Series inductors excitcd by a source current, (b) Graph o f the source current.
S olution
From KCL, i^{t) =
for all t. Since i^{t) = 2r A for 0 < r < 1 and i^{t) = 2 A for /> 1, equa
tion 7.3 or 7.4 immediately yields the instantaneous stored energies (in J) as plotted in Figure 7.12:
0^/<l
1
4r
0:sr<l
1s
> t
> t
(a)
(b)
FIGURE 7.12 (a) Encrg)' (in J) stored in inductor Z.,. (b) Energy (in J) stored in inductor Ly
282
Exercises. 1. For the circuit o f Figure 7.1 la, find analytic expressions for the instantaneous stored
energ)' for the current excitation in Figure 7.13a for r > 0.
2. Repeat Exercise 1 for Figure 7.13b.
FIG U RF 7.13 Current excitation for Exercise 2.

ANSWUPvS:
1.
0 .2 5 r
UV/) =
{) ^ / < 2
t-
()s/<l
0 .2 5 (9 - 6/+ / )
i s / < 3 and U '2(0 =

3s I
0 s /< 2
2s/
4/
0 s /< 1
(9-fv + r )
I s / <3
3</
EXA M PLE 7.5

For the circuit o f Figure 7.14 in which v^{t) = cos(t) V for r > 0 and 0 otherwise, find the input
current i^{t) for r > 0 and the energy stored in each o f the inductors for the intervals [0, t] for 0 <
t < 1 and [0, t] for 1 < t.
i.(t)
L^ = 1H
FIGURE 7.14 Parallel inductive circuit with switch in which v^{t) = cos(t) V for /> 0 and 0 otherwise.
283
S olution
Step 1. Since no voltage is applied to either inductor for / < 0, /,(0) = 0. Further, no voltage
appears across the second inductor until r > 1. Hence, /^(l) = 0.
Step 2. Equation 7.2 implies that, for 0 < r < 1,
( s ( 0 = /i(/) =
L|
v ,(T )r/ T =
-jL|
/ ^ c o s (t)Jt =
sin(/) A
Step 3. At ^ = 1, the switch closes. T he r\vo inductors are then in parallel, and the source voltage
appears across each. Hence, by equation 7.2,
co s (t
W t = sin( I ) + sin(/) - sin( I ) = sin(/) A
Also, equation 7.2 applied to L-, implies
i2{t)= /2(l)+ Jj^
cos(T)r/T = sin(/) - s i n ( l ) A
From the KCL, the input current ij^t) = /,(/) + ijit) = 2sin(/) - sin(I) A for / > 1.
Step 4. Compute the energy stored in the inductors over the interval [0, t]. From equation 7.3, it fol
lows that for 0 < r < 1,
t) - 0.5 sin^(r) joules, whereas
->((), t) = 0.
Step 5. Compute the energy stored in the inductors over the interval [0, f] for 1 < t. Again from equa
tion 7.3, for 1 < t,
sin^(l)] joules.
t) = 0.5 sin^(^) joules and
t) = 0.5[sin~(r) - 2 sin (l) sin(/) +
Exercise. Repeat the calculations o f Example 7.5 for

= 2 sin(r) V for / > 0 and 0 otherwise.
A N SW ERS: For 0 < / < 1. U'} ,(0, f) = [2 - 2 cos(r)] J, whereas W) ,(0, t) = 0; for 1 < U'^^,(0.
/) = [2 - 2 cos(/)]- J and U'} ,(0, ;) = [ 1.0806 - 2 cos(/)]- J.
28-
3. THE CAPACITO R
DefinitiojisandProperties
D EFIN ITIO N OF T H E C A PA CITO R
Like the inductor, the capacitor, denoted by Figure 7.15a, is an energy storage device.
Physically, one can think o f a capacitor as two metal plates separated by some insulating mate
rial (called a dielectric) such as air, as illustrated in Figure 7.15b. Placing a voltage across the
plates o f the capacitor will cause positive charge to accumulate on the top plate and an equal
amount o f negative charge on the bottom plate. This generates an electric field between the
plates that stores energy. Hence, for a capacitor.
(7.5)
clt
d\
where q{() is the accumulated charge on the top plate, which is proportional to the voltage
V({t) across the plates; thus q{i) = Cv^t), with proportionality constant C denoting capaci
tance and having the unit o f Farad (F). One Farad equals 1 amp-sec/volt. The capacitance C
is a measure o f the capacitors potential to store energ)' in an electric field.
ic(t) >r
^
V ,(t)
(a)
+ -h
-i- + 4-
4- + + -I--1- +
A- + + + + + +.
(b)
FIG URE 7.15 (a) The symbol for the capacitor with conventional voltage and current direc
tions. (b) Illustration o f electric field between plates of a parallel-plate capacitor.
Modern-day capacitors take on all sorts o f shapes and sizes and materials. In keeping with craditio n , the parallel-platc concept remains the customar)' perspective. Calculating the capacitance o f
t^vo arbitrarily shaped conducting surfaces separated by a dielectric is, in general, ver)- difficult
Fortunately, the ordinary capacitor o f a practical circuit is o f the parallel-plate variety, with the
plates separated by a thin dielectric. The two plates are often rolled into a tubula, ruu, .,>1
complete structure is sealed.
EXA M PLE 7.6
For the capacitor circuit o f Figure 7 . 16a, compute i^t) when v j t ) = r5""'sin(1000r) V for / > 0.
28S
ic(t)
v Jt )
2mF
(a)
Time in milli-seconds
(b)
FIG URE 7.16 (a) A 2 mF capacitor connectcd to a voltage source,

(b) Plots of capacitor voltage and current waveforms.
S o lution
A direct application o f equation 7.5 yields
i^{t) =
sin(1000/) +
dt
Exercises. 1. In Figure 7.16, suppose

=e
(preferably in MATLAB) i^^t) for 0 < r < 0.5 sec.
cos(l()00/) A
V for r > 0. Compute
for r > 0. Sketch
ANS\V1-:R: - 0 . 0 5 , A.
2. Repeat Exercise 1 with
= e~-^^ cos( 100/) V for f > 0 but plot over die time interval [0, 0.15 sec].
A.\S\V1-:K: -.- -^ q 0 .0 5 COS. KJOr) r 0.2 sin(lOOr)] A.
The differential relationship o f equation 7.5 has the equivalent integral form
' t < '> =
^ f J c W d T + I j ; " ic W d T
(7.6)
C '^0
where
is in volts, /^r) is in amps, and C is in farads, and where we have taken
=0
because the capacitor was not manufactured at t = -oo. The time /q represents an initial time o f
interest or significance, e.g., the time when the capacitor is first used in a circuit. The quantit)'
286
specifies the initial voltage across the capacitor at ^q. This initial voltage,
sums up the entire
past history o f the current excitation into the capacitor. Because o f this, the capacitor, like the
inductor, is said to have memory.
EXA M PLE 7 .7
Suppose a current source with sawtooth waveform
shown in Figure 7.17h, drives a relaxed
0.5 F capacitor (zero initial voltage) as in the circuit o f Figure 7.17a. Compute and plot the volt
age across the capacitor.
i.(t) A/v^(t)V
(a)
(b)
FIG U RE 7.17 (a) Current source driving a capacitor.

(b) Sawtooth current waveform and voltage response of a 0.5 F capacitor.
S o l u t io n
The input waveform is periodic in that it repeats itself every 2 sec. Therefore, the solution will pro
ceed on a segment-by-segment basis.
Step 1. Consider the interval 0 < f < 2. For this interval ij^t) = {It - 2) A. With
= 0, it fol
lows from equation 7.6 that

v^(/) =
(2 t - 2)ilT = 2 ( r - 2/) V for 0 < t <2
Step 2. Consider the interval 2 < t <A. Observe that at / = 2,
= 0; hence, the capacitor volt
age over the interval 2 < /< 4 is simply a right-shifted version o f the voltage over the first inter
val. Right-shifting is achieved by replacing t with t - 2 . In other words,
v^{t) = 2[{t - 2)2 - 2{t - 2)] V for 2 < r < 4

Step 3. Consider the general interval 2k < t < 2{k + 1). For interval 2k < t < 2{k + 1),
v^t) = 2{{t - 2k)^ - 2{t - 2k)], /^= 0 , 1, 2 ,
...
Lastly, obser\ê that the voltage across the capacitor, as illustrated in Figure 7.17b, is continuous
despite the discontinuity o f the capacitor current. Again, this follows because the capacitor volt
age is the integral (a smoothing operation) o f the capacitor current supplied by the source.
287
Exercise. Consider the capacitor circuit o f Figure 7.18. Suppose the current source is i^{t) = e~ A
for /> 0 and
= 1 V. Compute the capacitor voltage
the resistor voltage
and the
voltage vj^t) across the current source for r > 0.

-H V(t) -
+
I (t) = e-'u(t) 0
20
M O
0.5 F
v,(t)
FIGURE 7.18 Scries RC circuit driven by a currcnt source for accompanying exercise.
AN SW ERS: v^^t) = 3 -
for t> 0,
= lc~' for t > 0. and, by K\'l., v^{t) = 3 V for t > 0.
It is important to emphasize that the sawtooth current input depicted in Figure 7.17b is a dis
continuous function, but the associated voltage waveform is continuous because integration
(equation 7.6) is a smoothing operation. This means that the capacitor voltage is a continuous
function o f t even for discontinuous capacitor currents, provided they are bounded. This obser
vation leads to the continuity property o f the capacitor: if the current i(^t) through a capacitor
is bounded over the time interv'al
< ^ < ^2> then the voltage across the capacitor is continuous
for fj < r < tj. In particular, for bounded currents, if fj <
< tj, then
= V(- (tQ"^), even when
At the macroscopic level, there appear to be some exceptions to the continuit}' propert}' o f the
capacitor voltage, e.g., when two charged capacitors or one charged and one uncharged capacitor
are instantaneously connected in parallel. In such cases, KVL takes precedence and will force an
instantaneous equality in the capacitor voltages, subject to the principle o f conservation o f
charge, to be discussed shortly. Another example is w'hen capacitors and some independent volt
age sources form a loop. When any o f the voltage sources has an instantaneous jump, so will the
other capacitor voltages. Upon closer examination, however, we see that there is really no excep
tion to the stated continuity rule: it can be shown that in all o f the cases where the capacitor volt
age jumps instantaneously, an impulse current flows in the circuit. Physically, an impulse cur
rent is one that is ver)' large (infinite from an ideal viewpoint) and o f very short duration. The cur
rent is not bounded, and consequently, the capacitor voltage may jump instantaneously. This
jump does not violate the rule, which presumes that the currcnt is bounded.
Relatio7ishipofChargetoCapacitorVoltageandCurrent
We have defined the capacitance o f a two-terminal device strictly from its terminal voltage-current
relationship the differential equation 7.5 and the integral equation 7.6, which is now repeated:
v 'c ( 0 = V c(fo) + ^ f ^
C 'M)
Physically speaking, the integral o f i(^t) over [/q, t\ represents the amount o f charge passing
through the top wire in Figure 7.19 over [rQ, r].
288
ic(t)
/ ''+ + +q + + /
/+ + +
^
/ ++++++
A- + + + + + +y
-q
FIGURE 7.19 Capacitor cxcitcd by a currciu.
Bccausc o f the insulating' material (the dielectric), this charge cannot pass through to the other
plate. Instead, a charge o f +q{t) is stored on the top plate, as shown in Figure 7.19. By KCL, if
i(it) flows into the top plate, then

of
must flow into the bottom plate. This causes a charge
to be deposited on the bottom plate. The positive and negative charges on these two
plates, separated by the dielectric, produce a voltage drop
from the top plate to the bottom
plate. For a linear capacitor, the only t)'pe studied in this text, the value o f V(^t) is proportional
to the charge
The proportionalit)' constant is the capacitance o f the device. Specifically,
qit) = C\U)
(7.7)
where q{t) is in coulombs, Cis in farads, and tî^) is in volts. Thus, equation 7.6 has the following phys
ical interpretation: the first term,
is the capacitor voltage at /q; the integral in the second term,
ic(T )d r.
represents the additional charge transferred to the capacitor during the interval [r,j, /]. Dividing
this integral by Cgives the additional voltage attained by the capacitor during [^q, ^]. Therefore,
the sum o f these rwo terms, i.e., equation 7.6, is the voltage o f the capacitor at r. Since q(/) =
it follows direcdy that
=
(7.8)
(It
cl!
ThePj'hicipleofConservationofCharge
It is important in terms of modern trends in circuit applications to further investigate the rela
tionship o f charge to capacitor voltages and currents. The principle o f conservation o f charge
requires that the total charge tramferred into a junction {or out o f a junctiori) be zero.~ This is a direct
consequence of KCL. To exemplify, consider the junction o f four capacitors shown in Figure 7.20.
289
v,(t)
+ V ,(t) -
i,(t)
i3(t)
- V3(t) +
v,(t)
+
l4(t)
FIGURE 7.20 Junction of four capacitors.

By KCL
/,(/) + ijit) + i^{t) + i^{t) = 0
Since the integral o f current with respect to time is charge, the integral o f this equation over
( - 00, t] is
/ ^ (m
'2
+'3
M^
=^/1 ) +^/2(0 +qj, (/) +f/4(/) =0
(7.9)
where qj^{t) is the charge transferred to capacitor k. By equation 7.6, at ever}' instant of time,
qi{t) = C-v.{t)
(7.10)
which defines the relationship between transported charge, capacitance, and the voltage across the
capacitor. Hence, from equations 7.9 and 7.10, at every instant of time,
C^vît) + C2 V2 U) +
+ C^v^{t) = 0
This simple equation relates voltages, capacitances, and charge transport. The following example
provides an application o f these ideas.
EXA M PLE 7 .8
This example shows that under idealized conditions, capacitor voltages can change instanta
neously. Consider the circuit o f Figure 7.21 , in which ^q(0~) = 1 V and
and V(^{t) for f > 0.
= 0 V. Find
290
t=0
C l = 1F
C2
Cl
C2 = 1F
FIGURE 7.21 Two parallel capacitors connectcd by a switch.
S o lution
At t = 0 , the charge stored on C, is C, ^ ^ (0 ) and that o f C j is

requires that
For t > 0, KVL
Therefore, after the switch is closed at r = 0, some charge must be
transferred between the capacitors to equalize the voltages. According to the principle o f conser
vation o f charge, the total charge before and after the transfer is the same. Thus, conservation o f
charge requires that
F,quivalently, <7i(0'^) - <7i(0 ) + qîO*) - ^2(0 ) = 0- From equation 7.10,

C,
Since ^ ^ (0 ) = 1 V, V(^{0 ) = 0, and from K \T

C,
= V(y{0*), it follows that
v - c , ( 0 ^ ) - l l + C 2 [ i c i ( 0 ^ ) - 0
Hence, (Cj + C ,)y Q (0 ^) = I implies that
= Vq^{0*) = 0.5 V.
Exercises. 1. In Example 7.8, make C, = 0.75 F and C2 = 0.25 F, and compute ^/^(O'^).
AN SW ER: /Y-,(0") = 0.75 V.
2. In Example 7.8, sufipose ^q ( 0 " ) =10 V and Vqj,{Q~) = - 8 V. Also let C, = 0.75 F and C-, = 0.25
F. Compute
ANSW ER: /.v.,(0^) = 5.5 V.
Example 7.8 is illustrative o f a charge transport that is germane to switched capacitor circuits,
which are o f fundamental importance in the industrial world.
EnergyStorageinaCapacitor
As with all devices, the energy stored or utilized in a capacitor is the integral o f the power absorbed
by the capacitor. The net energ}' entering the capacitor over the interval [/q, /J is
Jff,
Pci'^)dT = f'' Vc(T)/c(T)i/T
2 )1
chciT )]
= c r '' (
(It
Vcih)
Oo)
dT =
(7.11)
= -C
for C in farads,
in volts, and energy in joules (J). From equation 7. 11, the change in energy
stored in the capacitor over the inter\'al [rQ, r j depends only on the values o f the capacitor volt
ages at times /q and
i.e., on
and v^t^. This means that the change in stored energy is
independent o f the particular voltage v/aveform between

odic, i.e., if V(\t) =
and r,. If the voltage waveform is peri
+ T) for some r > 0, then over any time interval [t, / + 7], the change in
the stored energy in the capacitor is zero because
+ 7) =
forces equation 7.11
to zero. Analogous to the inductor, for all periodic voltages, the capacitor stores energy and then
returns it to the circuit and is thus called a lossless device.
As with the inductor, it is convenient to define the instantaneous stored energy in a capacitor as
Wc{t) = l^Cvc{t)
(7.12)
which is really the integral o f power over the interval (x , /], assuming that all voltages and cur
rents are zero at r = - x .
E X A M PLE 7 .9
Consider the circuit o f Figure 7.22, in which
= 0. It is known that
for f > 0, the source current is i^{t) and the voltage across the capacitor is
V for ^ > 0. Compute (i) the energ)', in joules, stored in the capacitor for
/ > 0, (ii)
and (iii)
FIGURE 7.22 Parallel RC circuit.

S o lution
(i) Since
= 0, from equation 7.11 (or 7.12),
VV^^(0,/) = lc v J ( / ) = 8 C / ?
\-e
RC
(ii) To find the capacitor current, recall

= C ^ ^ ^ = 4
<li
RC
) - 4/? \ - e
2 )2
(iii) To find
we first compute
flowing fi-om top to bottom:
\ -e
Thus
1- e-
+ 4^
E X A M PLE 7 .1 0
= 4 A
ijt )

For the circuit o f Figure 7.23a, it is known that the voltage across
the capacitor is
= 20sin(2f + rr/6) V for r > 0. Compute and
plot the instantaneous power absorbed by the capacitor and the

energy stored by the capacitor during the time interval [0, f].
+
N
5mF
k
-
T im e t in seconds
(b)
F IG U R E
7.23 (a) Capacitor with known voltage v^^t) connectcd to a network N.

and the net cncrg)',
t), stored over the interval lO, ^].
(b) Plot of power,
S o lution
Step 1. Compute
From equation 7.5, for / > 0

,-^(,) = C ^ ^
dt
= 0 .2 co s 2t + -
6}
Step 2. Computep(\t). By direct multiplication and a standard trig identit}',

/
k\
/
k\
{A
watts
P c(t)= V(-(/)/(-(/)= 20 sin 2 / + - X 0. 2 cos 2/ + - = 2 sin 4t + 6j
\
3!
l
6j
Step 3. Compute
t). From equation 7.11 with
= 20 sin(7r/6) = 10 V, we obtain
\V^(0.0 = 0.5 C \c {t) - 0.5 C v c (0 ) = sin-
2t + - - 0 . 2 5 J
6}
Plots o f pf4f) i^nd V\^^0, t) arc given in Figure 7.23b. Notice that WTÔ, /) can be negative,
because Wôo, 0) = 0.25 joules, meaning that at r = 0, there is an initial stored energy that can
be returned to the circuit at a later time . Figure 7.23b substantiates this.
4. SERIES AND PARALLEL IN D UCTO RS AND CAPACITO RS
Sei'iesInductors
Just as resistors in series combine to form an equivalent resistance, inductors in series combine to
form an equivalent inductance. As it turns out, series inductances combine in the same way as
series resistances.
E X A M PLE 7 .1 1 .
Compute the equivalent inductance o f the series connection o f three inductors illustrated in
Figure 7.24. Then find the voltages
as a fraction o f the applied voltage
Leq
+
V ..
+
Leq
Vl2
+
V .,
Q-
o+
êq
o -
(a)
FICIJRH
Leq
(b)
(a) Scries connection of three inductors, (b) Equivalent inductance.
29-4
S o lution
First we must answer the question o f what it means to be an equivalent inductance. Earlier, we
defined the inductor in terms o f its terminal voltage-current relationship. Two 2-terminal induc
tor circuits have the same inductance if each circuit has the same terminal voltage-current rela
tionship as defined in equation 7.1.
Step 1. The voltage labeled
il
appears across the series connection, and, by KCL, the current

= l y
flows through each o f the inductors, i.e.,
equivalent inductance,
is defined by the relationship

=
in terms o f Z j, L-,, and Ly
O ur goal is to express
Step 2. Find
(7.13)
dt
in terms ofi^^^. To obtain such an expression, observe that, by KVL,

^Leq = ^/.l + ^12 +
Since each inductor satisfies the v-i relationship
dt
it follow's that
^'Uq - (^1 + ^2 + ^ 3 )
dt
Hence, the series inductors o f figure 7.24a can be replaced by a single inductor with inductance
= -^-1 + ^2 +
Finally, since
= Lj
dt
= Lj ^
dt
and
_
dt
= { L, +L^ +
~
'
dt
, it follows that
Lj
(L| 4- Zy-> -l- L-^)
which is analogous to the voltage divider formula for resistances.
Exercises. 1. If, in Example 7. 11, Z,, = 2 mH,

AN SW ER:
= 8 mH.
2. Find
in terms o f
3
ANSW ER: '/.:=
= 5 mH, and
= 1 mH, find
295
Extension o f the formulas in the above example to n inductors is fairly clear, and we state the
results without rigorous proof: the formula for series inductances is
(7.14a)
and the formula for voltage division o f series inductances is
^'IJ =
(7.14b)
L\ + Lo + ... + Lfj
InductorsinParallel
The same basic question as with inductors in series arises with a parallel connection o f inductors:
what is the equivalent inductance? Rather than derive the general formula, let us consider the case
o f three inductors in parallel, as illustrated in Figure 7.25a.
E X A M PLE 7 .1 2
For this example our goal is to show that the equivalent inductance o f the circuit o f Figure 7.25a
is given by the reciprocal o f the sum-of-reciprocals formula,
- ~\
i
T
+ +
(7.15)
Ly
We then show a formula for current division.

Leq
Leq
O +
L3
L2,
êq
Leq
L,
L,
o-
o (a)
(b)
FIG URE 7.25 (a) Parallel connection of three inductors, (b) Equivalent inductance.
So l u t io n
Once again, equation 7.13 defines the relationship for the equivalent inductance:
Û-q
The goal is to construct
in terms o f Z ,,
êq
and
in a way that satisfies equation 7.13. This
will produce equation 7.15.

Step 1. Write KCL for the parallel connection shown in Figure 7.25a. Here, by KCL,
U q = 'Z.1 +
+ 'L 3
2%
DifFerentiating both sides with rcspect co time yields
(Hl\ ^
dt
dt
ill
dt
di,
Step 2. Find Ul in terms o f a n d L^,. From equation 7.1, For each inductor
dt
dt
Li
~ î\ ~ ^L1 ~ ^L5
Substituting into the result o f step 1 and noting that
di,
dt
( \
Ln
1
n
+ +
Li-q
L]
L t^)
L'S
This has the form o f equation 7.13, which implies equation 7.15, i.e.,
Le, =
+ ---- +
Z/j
L~i L,'^
lb generate a currcnt division formula we first note that
= ^/1 =
1 '
1 '
= J y i J c i T ) d T = J \ ' , ^ { T ) d T and
^ _-TT
'
= / v^,^(tV/t
^ _nr.
Thus
(/)
Exercises. 1. If, in Flxample 7.12, Z.j = 2.5 mH, Z., = 5 mFl, and
ANSWq-R:
= 1 mH, find
= 0.625 mH.
The above arguments easily generalize. Suppose there are u inductors, /,,, Z-,, ... , Z.,^, connected in
parallel. Then the equivalent inductance is given by the reciprocal o f the sum-of-reciprocals formula.
(7.16a)
and the current division formula.
Chapter 7 * Inductors aiul Capacitors
ii.jn
~i-------- [
(7.16b)
Exercise. For two inductors Zj and L-, in parallel, show chat the equivalent inductancc satisfies the
formula
(7.17)
Series-ParallelCombinations
This subsection examines series-parallel connections o f inductors. This allows us to use the for
mulas developed above in an iterative way.
EXA M PLE 7 .1 3
Find the equivalent inductance,
o f the circuit o f Figure 7.26.
S o l u t io n
Step 1. In the circuit o f Figure 7.26, several inductors are enclosed by an ellipse. Let
denote
the equivalent inductance o f this combination. Observe that the series inductance o f the 5/6 H
and 0.5 H inductors equals 4/3 FI. This inductance is in parallel with a 1 H and a 4 Fi induc
tance. Hence,
, =
I
1 I
3= 1
!----- !---1
Step 2. The equivalent circuit at this point is given by Figure 7.27. This figure consists o f a series
combination o f a 1.5 FI and a 0.5 FI inductor connectcd in parallel with a 6 H inductor. It fol
lows that
29S
-------------+ 0 .5 + 1.5 6
= -= 1 .5
Exercise. In Example 7.13, suppose the 5/6 H and 0.5 H inductors are both changed to 0.4 H
inductors. Find L o f the circuit.
AN SW ER: 1.443 H.
CapacitorsinSeines
Capacitors in series have capacitances that combine according to the same formula for combining
resistances or inductances in parallel. Similarly, capacitances in parallel combine in the same way
that resistances or inductances in series combine. This means that the equivalent capacitance o f a
parallel combination o f capacitors is the sum o f the individual capacitances, and the equivalent
capacitance o f a series combination o f capacitances satisfies the reciprocal o f the sum-of-reciprocals rule. These ideas are illustrated in the examples to follow.
EXAM PLE 7 .1 4
Compute the equivalent capacitance,
o f the series connection o f capacitors in Figure 7.28a.
2)9
o-
Cl
'C2
o+
C3
eq
Q-
(a)
(b)
FIGURI^ 7.28 (a) Series combination of three capacitors, (b) Equivalent capacitance,
So l u t io n
The equivalent capacitance denoted in Figure 7.28b is defined implicitly by the current-voltage
terminal conditions according to equation 7.5, i.e.,
; - r
dt
Our goal is to express this same terminal v-i relationship in terms o f the capacitances, C j, C 2, and
Cy After this we set forth a formula for voltage division.

Step 1. Set forth the i-v relationship fo r each capacitor. For each capacitor, k = 1, 2, 3,
Ck - Q
But, by KCL, i^ =
dt
Hence,
dt
C,
Step 2. Apply K V L From KVL,
Differentiating this expression with respect to time and using the result o f step 1 yields
dv/^
d v f^ ]
d v 'c '')
d v (--i
\ .
^ = + + = \ + + \ir
dt
dt
dt
dt
l,C|
Cj
c J
3D0
Step 3. Compute
. From the result o f step 2, solve for Iq to obtain
\
iQ -
[C l
dt
H--------h --
Cl
dt
C^j
It follows that
Q ,=
C,
C3
To set forth a formula for voltage division, we first note that
= /q =
1 '
1 '
'
''CA-(0 = J ick^T)dr = J ic{T)dT and C,^ Vci 0 = f ic(r)dT
^ -rr-
^ rri
Thus
''a - ( 0 = f ic(r)dT = -
v c(f) = -^------- j------- p v c (0

C,
C2
Exercises. 1. In Example 7.14, suppose Cj = 5 pF, Cj = 20 |.iF, and
C3
= 16 pF. Compute
ANSW ER: 3.2 uE

2. Find
in terms of
.ANSWER: /V;: = 0 .1 6 /y ;
Generalizing the result o f Example 7.14, we may say that capacitors in series satisfy the reciprocal
o f the sum-of-reciprocals rule. Thus, for n capacitors C j, C2, ... , C^, connected in series, the
equivalent capacitance is
1
- 1 -------------- r
+ + ... +
C,
G
C
(7.18a)
and the general voltage division formula is

1
' c ( 0
------- i---------
+ --- +...
C,
Co
C.,
( 7 .1 8 b )
3 01
Exercise. Show that if two capacitors Cj and C-y are connected in series, then
(7.19)
c ,+ c .
CapacitorsinParallel
If rvvo capacitors are connected in parallel as in Figure 7.29a, there results an equivalent capaci
tance
= Cj + C2 and a simple current division formula to be derived.

I.
a
+
'c
Cl
o
+
'C2
eq
a-
a(a)
(b)
FIGURE 7.29 (a) Parallel combination of two capacitors, (b) Equivalent capacitance,
Since the voltage
appears across each capacitor, and since /^= /q +
by KCL it follows that
Hence,
êq -
^2
One surmises from the above example that, in general, capacitors in parallel have capacitances that
add. And, indeed, this is the case: if there are n capacitors C j, C2, ... ,
in parallel, the equiva
lent capacitance is
C
= C, + C2 + ... + C
(7.20a)
Exercise. Show that in the above derivation

Ck .
Cf,
Ceq
Q + Q
and that for n capacitors in parallel.

'CA- -
Q
C] + C j + + C
'c
( 7 .2 0 b )
302
Series-ParallelCombinations
Wc round out our discussion o f capacitance by considering a simple series-parallel interconnection.
EX A M PLE 7 .1 5
Consider the circuit o f Figure 7.30. Compute the equivalent capacitance,
0.45 mF
0.6 mF
FIG U RE 7.30 Series-parallel combination o f capacitors.

S
o l u t io n
Step 1. Combine series capacitances. Observe that the rwo series capacitances o f 0.5 mF and 0.5
mF combine to make a 0.25 mF capacitance.
Step 2. Combine parallel capacitances. First, as a result o f step 1, the three capacitances, 0.3 mF,
0.25 mF, and 0.45 mF, add to an equivalent capacitance o f 1 mF. Further, the two parallel capac
itances, 0.3 mF and 0.6 mF, at the bottom o f the circuit, add to make a 0.9 mF capacitance. The
new equivalent circuit is shown in Figure 7.31.
a-
II mF
1.125 mF
0.9 mF
a-
FIGURE 7.31 Circuit equivalent to that o f Figure 7.30.

Step 3. Combine series capacitances. From equation 7.18,
^ eq-
1 1
-------- + - + ----1.125
1 0.9
Exercise. Suppose the two 0.5 mF capacitors in Figure 7.30 are changed to 2.5 mF capacitors.
Find the new
A N SW ER: 0.4 mF.
303
5. SM O O TH IN G PROPERTY OF A CAPACITO R IN A POW ER

SUPPLY
As mentioned in the chapter opener, a power supply converts a sinusoidal input voltage to an
almost constant dc output voltage. Sucii devices are present in televisions, transistor radios, stere
os, computers, and a host o f other household electronic gadgets. Producing a truly constant dc
voltage from a sinusoidal source is virtually impossible, so engineers design special circuits called
voltage regulators that generate a voltage with only a small variation between set limits for a given
range o f variation in load. The voltage regulator is a precision device whose input must be fairly
smooth for proper operation. A capacitor can provide a crude, inexpensive sm oothing function
that is often sufficient for the task. This section explores the design o f a capacitive smoothing cir
cuit. In practice, such a circuit is used only for low-power applications.
i.(t)
FIGURE 7.32 Simple power supply with capacitive smoothing for low- power applications.
Consider, for example, the circuit shown in Figure 7.32. The four (ideal) diodes are arranged in a
configuration called a fidl-wave bridge rectifier circuit. An ideal diode allows current to pass only
in the direction o f the arrow. The diode configuration ensures that i^{t) remains positive, regard
less o f the sign o f the source current. Specifically, the diodes ensure that /j(f) =
Using the
integral relationship (equation 7.6) o f the capacitor voltage and current, it follows that
V cit)= yc(h)) + ^ f
ici-^)ch = VcUo) + ^ f
[M t)|-/o(t)]^/t
(7.21)
Because o f the difference |/j(r)| - iît) inside the integrand o f the integral, i^{t) tends to increase
the capacitor voltage, whereas i^^{t) tends to decrease the capacitor voltage. Further, because the
diodes are assumed ideal, it follows that
v^t) > I v^{t)
(7.22)
To see this, suppose the opposite were true; i.e., suppose

One o f the diodes would
then have a positive voltage across it in the direction o f the arrow. The diode is said to be forivarcl
biased. But this is impossible, because an ideal diode behaves like a short circuit when forward
biased. The consequence is that V(4,t) will be 12 V whenever |?>'^(^)| is 12 V. This occurs every 1/120
o f a second. Thus, the rectifier output will recharge the capacitor every 1/120 o f a second. Between
charging times, the current, i(){t), will tend to discharge the capacitor and diminish its voltage.
Chapter 7 Inductors aiui Capacitors
304
The design problem for the capacitive smoothing circuit is to select a value for C that guarantees
that v^t) is sufficiently smooth to ensure proper operation of the voltage regulator. Here, suffi
ciently smooth means that the maximum and minimum voltages differ by less than a prescribed
amount. To be specific, suppose that i>(\t) must remain between 8 V and 12 V. Recall that i^{t)
tends to increase the capacitor voltage, while
tends to decrease it. The design requires select
ing a value for C to ensure that i^{t) can keep up with
so that the capacitor voltage remains
fairly constant. The value for /(,(^) is obtained from the specification sheet o f the voltage regulator.
Suppose this value is a constant 1 A. It remains to select C so as to ensure that V({t) remains above
8 V between charging times. From equation 7.21, it is necessary that
CJio
Now we need consider only values for t between 0 and 1/120, because the capacitor will recharge
and the process will repeat itself every 1/120 o f a second. Thus, because i^{t) will only increase the
capacitor voltage, to ensure that v^{t) remains above 8 V, it is sufficient to require that
With i^{t) = 1 and
= 12, it follows that

1 A X 120 sec
= 2.083 mF
4 V
A 2,100 |.iF capacitor satisfies this requirement. A method for computing the capacitor voltage
waveform is described in Chapter 22 o f o f 2"^ edition. However, using SPIC E or one o f the other
available circuit simulation programs, one can generate a plot o f the time-varying capacitor volt
age produced by this circuit, as shown in Figure 7.33. In the figure, it is seen that the capacitor
voltage varies between 12 and 9.02 V, which is smaller than the allowed variation o f (12 - 8) V.
Two factors contribute to this conservative design: (1) we used C = 2,100 uF instead o f the cal
culated value, C = 2,083 uF, and (2) the increase in the capacitor voltage due the charging current
is is not included in the calculation.
volts
FIGURE 7.33 Time-varying capacitor voltage generated by the circuit in

Figure 7.32 when C = 2,100 |.iK
The preceding brief introduction made several simplifying assumptions to clarify the basic use o f
a capacitor as a smoothing or filtering device. Practical power supply design is a challenging field.
A complete design would need to consider many other issues, some o f which are the nonzero
resistance o f the source, the non-ideal nature o f the diodes, the current-handling abilit)' o f the
components, protection o f the components from high-voltage transients, and heat-sinking o f the
components.
6. SUM M ARY
This chapter has introduced the notions o f a capacitor and an inductor, each o f which is a lossless
energ}' storage device whose voltage and current satisfy a differential equation. The inductor has a
voltage proportional to the derivative o f the current through it; the constant o f proportionalit}^ is
the inductance L. T he capacitor has a current proportional to the derivative o f the voltage across
it; the constant o f proportionalit}' is the capacitance C. It is interesting to observe that the roles of
voltage and current in the capacitor are the reverse o f their roles in the inductor. Because o f this
reversal, the capacitor and the inductor are said to be dual devices.
That the (ideal) inductor and the (ideal) capacitor are lossless energy storage devices means that
they can store energ)- and deliver it back to the circuit, but they can never dissipate energ)^ as does
a resistor. The inductor stores energy in a surrounding magnetic field, whereas the capacitor stores
energy in an electric field between its conducting surfaces. Unlike energ)' in a resistor, the energy
stored in an inductor over an interval [r^, fj] is dependent only on the inductance L and the val
ues o f the inductor current //(/^()) and //(/^j). Likewise, the energ)' stored in a capacitor over an
interval [r^, /,] is dependent only on the capacitance C and the values o f the capacitor voltage
and
Both the inductor and the capacitor have memor)'. The inductor has memory because at a partic
ular time Tq, the inductor current depends on the past histor}' o f the voltage across the inductor.
The capacitor has a voltage at, say, time
that depends on the past current excitation to the capac
itor. The concept o f memory stems from the fact that the inductor current is proportional to the
integral o f the voltage across the inductor and the capacitor voltage is proportional to the integral
o f the current through the capacitor. This integral relationship gives rise to the important proper
ties o f the continuity of the inductor current and the continuit)' o f the capacitor voltage under
bounded excitations.
rhe dual capacitor and inductor relationships are set forth in Table 7.1.
Finally, we investigated the smoothing action o f a capacitor in a power supply.
306
lABLE 7.1. Summary of the Dual Relationships ot the Capacitor and Inductor
ic )
icit) = C
d\'c )
dt
V[{t) = L
dilit)
dt
//(0 =//(ô)+ t /^ V;(T)^/T
LJ k>
7. TERM S AND CO N CEPTS

Bounded voltage or current; voltage or current signal whose absolute value remains below some
fixed finite constant for all time.
Capacitance o f a pair o f conductors: a propert)' o f conductors separated by a dielectric that per
mits the storage o f electrically separated charge when a potential difference exists
between the conductors. Capacitance is measured in stored charge per unit o f potential
difference between the conductors.
Capacitor (linear): a two-terminal device whose current is proportional to the time derivative o f
the voltage across it.
C oil: another name for an inductor.
Conservation-of-charge principle: principle that the total charge transferred into a junction (or
out of a junction) is /.ero.
C ontinuity property o f the capacitor: property such that if the current i(\t) through a capaci
tor is bounded over the time interval
< t < t-,, then the voltage across the capacitor is

< tj, then t^(^tQ~) =
when
continuous for /, < r < tj. In particular, if fj <
C ontinuity property o f the inductor: propert}' such that if the voltage

is bounded over the time interval r, < r <
continuous for r, < r <
In particular, if r, <
across an inductor
then the current through the inductor is

then /^(/o~) =
when
Coulom b: quantit)' o f charge that, in 1 second, passes through any cross section o f a conductor
maintaining a constant 1 A current flow.
Dielectric: an insulating material often used between two conducting surfaces to form a capacitor.
Farad: a me;Lsure o f capacitance in which a charge o f 1 coulomb produces a 1 V potential difference.
Faradays law o f induction: law' asserting that, for a coil of wire sufficiently distant from any mag
netic material, such as iron, the voltage induced across the coil by a time-varying current
is proportional to the time derivative o f the current; the constant of proportionality,
307
denoted Z, is die inductance o f the coil. Faradays law is usually stated in terms of flux
and flux linkages, which are discussed in physics texts.
H enry: the unit o f inductance; equal to 1 V-sec/amp.
Inductance: property of a conductor and its local environment (a coil with an air core or iron
core) that relates the time derivative o f a current through the conductor to an induced
voltage across the ends o f the conductor.
Inductor (linear): a two-terminal device whose voltage is proportional to the time derivative of
the current through it.
Instantaneous powen p(t) = */(/)/(<), in watts when v{t) is in volts and i{t) in amps.
Lossless device: device in which energy can only be stored and retrieved and never dissipated.
Lossy device: a device, such as a resistor (with positive R), that dissipates energy as some form o f
heat or as work.
Maxwells equations: a set o f mathematical equations governing the properties o f electric and
magnetic Beids and their interaction.
M emory: property o f a device whose voltage or current at a particular time depends on the past
operational history o f the device; e.g., the current through an inductor at time /q depends
on the history o f the voltage excitation across the inductor for t< /q.
Unbounded voltage or current: a voltage or current whose value approaches infinity as it nears
some instant o f time, possibly r = oo.
Voltage r^ u la to r: circuit that produces a voltage having only a small variation between set lim
its for a given range o f load variation from a fairly smooth input signal.
^The word stored emphasizes that the energy in the inductor is not dissipated as heat and can be recovered by
the circuit, whereas the word absorbed is used to mean that the energy cannot be returned to the circuit. In a
resistor, energy absorbed is dissipated as heat.
^ More generally, conservation of charge says that the total charge transferred into a Gaussian sur&ce (or out of a
Gaussian surface) is zero.
31)8
Chapter 7 Induccors and Capacitors
Problems
4. (a) For i^{t) = 10sin(2000r) mA in Figure

P7.4, calculate and sketch
for 0 <
t < \5 ms assuming both inductor cur
TH E IN D U CTO R AND ITS

PROPERTIES
(b) What is the instantaneous power deliv
1. If the length o f a single-layer air coil is
(c) Compute and sketch the energy stored in
rents are zero at f = 0.

ered by the dependent source?
greater than or equal to 0.4 times its diameter,
the 2 niH inductor for 0 < f < 15 msec.
then its inductance is approximately given by

the formula
L =
lOv ft)
o f tiirns)~
4 X
18 {dkimeter)+ 40 (lengths)
length o f the coil are in meters. A 2 cm diam
Figure P7.4
eter coil has 48 turns wound at 12 turns/cm.

Compute the approximate value o f the induc
tance.
C H EC K : 18 pH < I < 2 0 pH.
2mH
0.2 mH
where L is in henries, and the diameter and
5. For Vsii) sketched in Figure P7.5a, compute

and sketch
P7.5b. What is the instantaneous power deliv
2. (a) Find and plot for 0 < r < 5 sec the

V (t) (mV)
inductor voltage Vj{t) for the circuit o f

Figure P7.2a driven by the current
waveform of Figure P7.2b.

1
1
(b) Find and plot the instantaneous stored

energ)^
t
I
(c) Find and plot the stored energy U^(l,r)
-1
1
6
as a function o f time for 5 > / > 1.

-1 -
L(A)
-2 L(t)
t(s)
(a)
Figure P7.2
-2
(a)
, 0.8 mH
(b)
(t)
3. Repeat Problem 2 for:

(a)
and
(b) the waveform sketched in Figure P7.3.
6 ijt)
v,{t)
0.6 mH
0.75 mH
1.5 mH
i.JA )
(b)
Figure P7.5
6. (a) Find and plot for 0 < r < 6 sec the induc
Figure P 7.3
tor current /^(f) for the circuit o f Figure
3 0 ')
(a) the voltage waveform sketched in Figure
P7.6a driven by the voltage waveform of
P7.8b, and
Figure P7 .6 b.
(b) the voltage waveform sketched in Figure
(b) Find and plot the instantaneous stored

energy.
P7.8c.
(c) Find and plot the stored energy

as a function o f time for 5 > ^ > 1.
(d) Find and plot t/j (/)
i,(t)
vJt)
(a)
(b)
Figure P7.6
7. Repeat Problem 6 for
(a)
u(t), and
(b) the voltage waveform in Figure P7 .7 .
Figure P7.8
9. Consider the circuit in Figure P7.9 in which
Z,
= 0 .2 H , Z2 = 0 .5 H , and
100sin(0.257t/) mV for / > 0 and zero other

wise.
(a) Find the current /^(/) for r > 0.
(b) Compute the energy stored in each
inductor over the intervals 0 < r < 2 sec
and 2 <
8 . For the circuit in Figure P7.8a, suppose Z, =
0.8 H and L2 = 0.2 H. Compute and plot the
waveforms / j (/) and /^J f ) for
310
v,(t)
% iQ
ic(t)
Figure P7.12
Figure P7.9
13. In Figure P7.13a, the capacitors C, = 4 mF
10. The circuit o f Figure P 7 .10 has two induc

tors, Z., = 20 mH and Lj = 50 mH, in parallel.
The input is v^{t) = 200cos(5007rr) mV for t >
and C j = 12 mF are driven by the voltage

specified in Figure P7.13b. Plot /q(/),
and iA i).
0 and zero otherwise. The switch between the

two inductors moves down at r = 4 ms.
Compute the currents
and
for 0 <
r < 4 ms and 4 ms < /. Also find the energy

stored in each inductor as a function o f t for the
same time intervals.
(a)
Figure P 7.10
THE CAPACITOR AND ITS

PROPERTIES
11.(a) Suppose that a 20 pF capacii
charged to 100 V. Find the charge that
(b)
resides on each plate o f the capacitor.
Figure P7.13
(b) If the same charge (as in part (a)) resides

on a 5 pF capacitor, what is the voltage
14. T he C = 2 pF capacitor o f Figure P 7 .l4 a
across the capacitor?

(c) What is the voltage required to store 50
pC on a 2 pF capacitor?
(d) Find the energy required to charge a 20
has current ;^ r) shown in Figure P 7 .l4 b . If

= 4 V, compute
at ^ = 1, 2, 3, 4 ms.
Now compute the energy stored in the capaci
tor over the intervals, [0, 2 ms], [2 ms, 3 ms],
pF capacitor to 100 V.
and [0, 4 ms].

12.(a) The C = 2 pF capacitor o f Figure P7.12
has a terminal voltage o f
= 100[1 +
cos(lOOOTtr)] V. Find the current i(^t)

through the capacitor.
(b) Now suppose the voltage is v^^t) =
10sin(2000r) V and Iq = 10cos(2000r)
mA. Find the capacitance C.
31
Chapter 7 Induaors and Capacitors
15. Suppose
as specified for all time in
Figure P7.15a, excites the circuit of Figure

P7.15b, in which Cj = 0.2 pF and C2 = 0.1 pF.
(a) Plot
for 0 < ^< 8 msec.
and
(b) Compute and plot the energy stored in

the 0.2 pF and 0.1 pF capacitors for
0 < r < 8 . Hint: use MATLAB to plot the
answers.
(c) Find
and
as t
00.
O '
w
O '
Figure P7.16
(b)
0
Figure P7.15
16. For the circuit in Figure P7.16a, C = 0.25

mF. Compute and plot the waveforms of the
voltage,
given
as sketched in Figures
P7.16B and c.
17.(a) Consider the circuit sketched in Figure

P 7.17 in which Cj = 20 pF and C2 = 0.1
mF. Suppose v^{t) = 5sin(2000f) V for t
> 0 and suppose
= 10 V. Find
for r > 0. Is the output voltage

independent of the initial voltage on Q ?
Why?
0
(b) W hat is the instantaneous power deliv
(c) Find the energy stored in Cj over the

interval [0 , t].
i jt )
(a)
0
0
0
Figure P 7 .1 7
312
18. Repeat Problem 17 when v^{t) =

V for / > 0 and 0 otherwise.
19. Reconsider the circuit o f Figure P 7 .17.
Suppose, however, that
is given by the plot
in Figure P 7 .19.
(a) Find and sketch
Figure P7.21
for 0 < / < 6
msec.
(b) What is the instantaneous power deliv
22. In the circuit o f Figure P7.22 v^{t) = 25 V

and the

(c) Compute and plot the energ)' stored in
C,.
= 100 mF capacitor is uncharged at
f = 0 . Compute V({t) for 0 < r < 2 sec and 2 sec

< t when C, = 4 0 0 mF.
t = 2s
Figure P7.22
23 . Fhe circuit ol Figure P7.23 has two capac

itors in parallel, C| = 30 mF, C-, = 50 mF. Fhe
20. Repeat Problem 17 for the waveform of
input current is iî) = 360f~'^^ mA for r > 0
Figure P7.20.
and 0 otherwise. Suppose each capacitor is

imcharged at /= 0 . The switch between the t%vo
capacitors opens at r = 2 msec.
(a) Find the voltage,
for 0 < r < 2 ms
and 2 ms < t.
(b) Compute the energy stored in each
capacitor as a function o f t for the same
time intervals.
Figure P7.20
(c) Com pute the current through each

capacitor over each time interval.
21. For the circuit o f Figure P7.21, suppose C,

= 0.6 ml', Ct = 1.2 mF,
t = 2 ms
= 0.4 mF, Q = 1.6
mF, ijj) = 120sin(100r) niA for / > 0 and 0 for

r < 0.
(a) Find the equivalent capacitances
the series combination and
for
for the
parallel combination.
Figure P7.23
(b) Find and sketch

(c) What is the instantaneous power deliv
(d) What is the instantaneous energ)' stored
in C4?
M IX ED C A PA CITO R AN D
IN D U C TO R PRO BLEM S
24. Consider the circuit o f Figure P7.24,
2.5 H, Cj = 1 mF, which is excited by the cur-
313
rent waveform
= 200fr ^' mA for ^ > 0

as a function o f i^{t) and
(a) Compute
and 0 otherwise.
(a) Compute and sketch I'lit),
the capacitances Cj and C j.
and
in terms o f i^{t) and the
(b) Now find
circuit parameter values.
(b) Compute and sketch the energy stored

in the inductor for r > 0.
(c) Compute and sketch the energy stored
L.
in the capacitor for t> 0.
Figure P7.27
28. For the circuit o f Figure P7.28, compute
and
Figure P7.24
as a function of
and the circuit parameters.

25. In the circuit o f P'igure P7.25, suppose Z., =
0.25 H, Cj = 2.5 mF, ij^t) = 20sin(400/) mA for
r > 0 and 0 otherwise. Ail initial conditions are
zero at r = 0.
(a) Find Vjit).
+
" 0
c:
\|>
-5
L ,2
L,
(b) Find V(^t).

(c) Find the instantaneous stored cnerg)' in
the capacitor.
Figure P7.28
SERIES-PA RALLEL IN D U C TO R S
29. In the circuit o f Figure P7.29, all inductors
are initially relaxed at /^= 0 and /.j = 6 mH, L-,
Figure P7.25
26.(a) In the circuit o f Figure P7.26, (i = 10, C,
= 38.5 mH,
= 22 mH. A voltage
200re~' mV is applied for r > 0. Find,
Vjj(t), and
Challenge: Find
= 20 pF, C , = 80 mF, Z., = Z , = 20 mH

are
initially
uncharged.
If
vj^t) =
10/sin(20^) V for r > 0 and 0 otherwise,

(a) Find i^{t) for r > 0.
(b) Now find
for / > 0.
(c) Compute the energ)- stored in the 20

mH inductor for r > 0.
i.(t)
C H E C K S: 20 mH and \AOte~' mV.
Figure P 7 .2 6
314

Suppose Z., = 3 mH, L j= \2 mH,
(mA)
= 36
= 120cos(1000r) mA.
mH, and
(a) Find L
and
(b) Find
(c) Plot the instantaneous power deliv
ered by the source for 0 < r < 14
msec.
/
L.
-200
--
Figure P7.32
(!)
33. In Figure P7.33, Z-j = 5 mH, L-, = 20 mH,

= 20 mH,
= 80 mH, and ij^t) =
lOsin(lOOOr) mA for r> 0 and 0 otherwise.

(a) With the switch in position C, find the
Figure P7.30
C H EC K :
12 mH, - 1 .4 4 sin(lOOOr) V, 90
equivalent inductance,
Vj^, and
(b) Repeat part (a) with the switch in posi
cos(lOOOr) mA.
tion D.
31. For the circuit o f Figure P7.31, Z., = 260
mH,
= 26 mH, L-^ = 39 mH, and
10^"^ tiiA.
(a) Find
(b) Compute
(c) Com pute
stored in
and i[^2 and

the
instantaneous
energy
as a function o f t.
C H EC K S: 52 mH. 0.2
0.8 i j t ) ,
-0.104^6" ' V, Vj2 i^) =

34. Consider the circuit o f Figure P7.34a with
voltage source excitation given in Figure P7.34b.
Let the inductor values be those given in
Problem 33. Suppose the switch is in position C.
Note that each inductor is relaxed at r = 0.
(a) Find Z^^^.
(b) Compute and sketch
for r > 0.
(c) Find the instantaneous (total) energ)'

stored in the set o f four inductors as a
function o f time.
Figure P7.31
32. Repeat Problem 31 for the waveform of
Figure P7.32.
(d) Compute and sketch //2(0CH ECK: L
= 20 mH; /.(1) = /.(3) = 0.8 A

while /y,/2) =
=0
31
ijt )
____rvY V
v X
L.
L,
1-6
L,
i(t)
f r r \ ____ T Y Y \
(a)
i-N/YYA___ TYYV
v(t)(V)
>k
16'
16-
L.
(b)
rOA___ TYYV
Figure P7.34
35. Find
L,
(b)
for the circuit o f Figure P7.35, (a)
when the s\vitch is open, and (b) when the

switch is closed. The unit o f L is henries.
Figure P7.36
SCRA M BLED ANSWERS: 0.1, 0.08, 0.6 (in 11)
36. Find
P 7.36, where Z., = 5 niH,

= 20 mH,
niH, I 4 = 150 mH, Z.5 = 50 mH,
mH, L j = 120 mH,
= 40
= 180
= 35 mH. Notice that
the circuit o f (b) is a modification o f (a) and

that o f (c) is a modification o f (b). Connections
can create interesting behaviors.
37 . Three 60 mH inductors are available for

interconnection. List all equivalent inductances
obtainable over all possible interconnections o f
these elements.
C H EC K : There should be seven different val
ues.
316
38. Find L
P7.38.
4 mH
AN.SWER: C//I
for
. all / values.
40. Like Problem 39, this is a conceptual prob
lem and requires no calculations for the answer.
lOmH
1 mH
Consider circuits 1 and 2 o f Figure P7.40. All
5 mH'
wish to determine the relationship bet\veen
inductors are 1 H except the one labeled L. We

3mH'
36 mH
and L^^-, in the presence o f the finite posi
Bo(a)
tive inductor o f L henries between points a and

b. W hich of the following statements is true?
/YYV
7mH
/
2.4 mH
(h)
< /v,2-
1.2 mH
(d) There is no general relationship between
êq\
0.6 mH
relationship depends
on the value o f L
(b)
Figure P7.38
ANSWHR: (a) 13 m il: (b) 2 in 11
39. This is a conceptual problem
and requires no calculations for
the answer. Consider circuits 1
and 2 o f Figure P7.39. All induc
tors are 1 H except the one labeled
L. We wish to determine the rela

tionship between
and
presence
o f the finite positive inductor o f L henries

between points a and b. Which of the following
statements is true?
êq\ ^ êql(b)
< K ,l(d) There is no general relationship between
and
Any relationship depends
on the value o f L.
Circuit 1
Circuit 2
Figure P7.40
ANSW ER: /.
SERIES-PARALLEL CAPACITORS
41. (a) Find the indicated equivalent capacitance
for the circuit o f Figure P 7.4la where C,
= 4 pF, C 2 = 3 pF, C3 = 2 pF, Q = 4 pE
Then find vj^t) when ii^t) = lOcos(lO'^r)

mA for r > 0 and 0 otherwise.
(b) Repeat for Figure P 7 .4 lb in which C, =
60 pF, C , = 18 pF, C 3 = 18 pF,
Q = 36 pF, and C 5 = 10.8 pF.
Then find vj^t) when i^{t) =
1 Osin( 1O^f) mA for r > 0 and 0

otherwise.
31'
Chapter 7 Induaors and Capacitors
-C,-
Vw/'
C. c,
<!>
^ 1
Figure P7.43
44. For the circuit of Figure P 7.44, Cj = 8 mF,
C2 = 6 mF, C3 = 12 mF, and
240sin(200r) mA for ^ > 0 and 0 otherwise.

(a) Find C,^.
(b) Find v- {i). Note: All capacitors are ini
(b)
tially uncharged. Why?
Figure P7.41
CHECKS: 6 pF, 66 ^F
If
if if
if
c,
c.
4 2 .(a) Find the indicated equivalent capaci

tance for the circuit o f Figure P7.42a
assuming Cj = 48 pF, Cj = 16 |jF,
c,
20 )jF, Q = 80 pF, and C5 = 8 pF.

(b) Repeat for Figure P7.42b assuming Cj =
Figure P7.44
3 pF, C2 = 6 pF, C3 = 3.6 pF, Q = 6 pF,

C5 = 4.5 pF, Q = 48 pF,
= 48 pF, Cg
= 24 pF, Cg = 24 pF.
4 5 . Three 12 pF capacitors are available for

interconnection. List all equivalent capaci
tances obtainable over all possible interconnec
tions of these capacitors.
oc
C -L
c.
C,
(a)
4 6 . This is a conceptual problem and requires

no calculations for the answer. Consider cir
cuits 1 and 2 o f Figure P7.46. All capacitors are
1 F except the one that is labeled C. We wish to
determine the relationship between S ' and
êql
presence o f the finite positive C F
capacitor between points a and b. Which of the

following statements is true?
(b)
Figure P7.42
^ êq2 '
(W
êq\ - êql43. Find
for the circuit of Figure P 7.43, (a)
when the switch is open, and (b) when the
(d) There is no general relationship between

and
Any relationship depends
switch is closed, assuming that Cj = C4 = 12 pF,
on the value of C.
C2 = C5 = 40 pF, C3 = Cg = 2 0 pF.
(c) Repeat parts (a) and (b) for Cj = 12 pF, Cj
= 40 pF, C3 = 20 pF, C4 = 4 0 pF, C5 = 20 pF,

Cg = 100 pF.
318
ages at f = 0 are zero, find all three voltages,
ICjit), i = 1,2,3, for r > 0. Then find the instan

taneous stored energy at / = 0.05 sec.
v jt)
c,
V,
Figure P7.48
C H EC K ;
= 6(1 -
49. In the circuit o f Figure P7.49, suppose Cj

= 3 pF, C3 = 0.5 pF, C2 = 1.5 pF, and
10sin(400r) mV for f > 0 and 0 otherwise.

(a) Find v^^{t) and v^ît) for ^ > 0.
(b) Find an expression for the energy stored
in C] and C , over the interval [0 , /].
Circuit 2
Figure P7.46
ANSWMR: C ; <
v jt)
47. In Figures P7.47a and b, the charge on Cp

and C3 is Q = 72 X
C. hi Figure
P7.47a, the voltages on C ,, C , and
Figure P7.49
are 2 V,
3 V, and 4 V, respectively; and in Figure 7.47b
Vq = A \ while the charges on Cp C , and
50. In the circuit o f Figure P7.50, suppose C,
are Qj = 48 x 10^ C, Q 2 = ^0 ^ ^0^ C, and

Q 3 = 72 X
C, respectively.
= 5 mF, Cj = 20 mF,
(a) Find
for the circuit o f Figure P7.47a.
(b) Find
for the circuit o f Figure P7.47b.
= 4 mF, Q = 80 mF,
= 10 0 e~^^ V for f > 0 and 0 otherwise.

(a) Find
and
for r > 0 .
(b) Compute the energy stored in the C-, over

the interval [0 , t].
(b)
(a)
ANSWHK: (a)
Figure P7.47
= 8 mF
48. In the circuit o f Figure P7.48, C| = 6 mF,

C2 = 12 mF, C3 = 36 mF, v J t ) = 20(1 V for t> 0 and 0 otherwise. If all capacitor volt-
v,(t)
Figure P7.50
51 . In the circuit o f Figure P7.51, suppose Cj

= 4 mF, C-) = 80 mF, C3 = 20 mF, and
1OO^*^^ mA for t > 0 and 0 otherwise.
(a) Find /q(^) and

for t > i).
(c) Compute the energy stored in C-, over
the interval [0 , t].
31')
the green left turn signal. The interesting variation

is that the v-i inductor relationship is different for
time-var\'ing inductances:
WYiLit)
v^(/) = -
at
The following highly simplified circuit illus
Figure P7.51
trates the principle o f operation, although the
M ISCELLAN EO U S
52. Find and sketch
configuration and values may not be what are

for 0 < r < 4 sec for
actually used. Consider the circuit o f Figure
the circuit o f Figure P7.52a, assuming all
P7.54a consisting o f an inductor driven by a
capacitors are initially at rest for the excitation
current source. When the car with its steel
o f Figure P7.52b. Are any o f the capacitors
frame moves over the coil o f wire, the induc
redundant as far as
tance o f the coil changes from I , to some larg-
is concerned?
er vâlue
0.5 F
0.2
+
pkv,
y+\
__
^ . 4 f^ . 2
-_ y
f
3L^ as illustrated in Figure
P7.54b, where the time
depends on the
speed at which the car is slowing down and
i/y
0.4 F
1-^4
2 T f)
(a)
Plot vît) for t >0 assuming ij{t) =
V ,(t)
a con
stant value, and that the front edge o f the car

begins to cross the first edge o f the coil at t = 0 .
Explain how this voltage signal might be u.sed
to control the traffic light.
Figure P7.52
0.5H
(a)
C H EC K : v,,,(/) = :^ v ,(/ )
53. Using the circuit given in Figure 7.32,
select a capacitor value to filter the voltage
for a regulator requiring 14 V <
Use
< 20 V.
= 20 cos(2007tf) and /^(f) = 2 A.
C H E C K : C > 1.667 mR
54. When driving a car into a left-hand turn
lane, one often sees a large circular or hexago
nal cut in the concrete. Embedded in these
cuts is a coil o f wire. When your car (contain
ing a large percentage o f iron) passes over this
coil, its inductance changes. This change o f
Time (in sec)
inductance can be used as a sensor to activate a
(b)
circuit that stops oncoming traffic and lights
Figure P7.54
L(t)
First Order RL and RC Circuits
When watching a manufocturing process, a visitor

might see a pair of robotic arms assemble an engine or
machine a block o f metal with perfectly timed
maneuvers. Timing is a critical aspect of a manufac
turing process. In T V transmitters there is a signal
called the raster, which is critical to the generation of
the screen image. In an oscilloscope a timing signal
called a horizontal sweep acts as a time base, which
allows one to view measured input signals as a func
tion o f time. All these applications utilize a signal hav
ing sawtooth shape and called a linear voltage sweep.
The linear voltage sweep is nicknamed the sawtooth,
rhis sawtooth is pictured here together with an
approximating exponential curv'e for comparison.
Linear Sweep or Sawtooth Waveform
Exponential Approximation
Ideally, the sawtooth voltage increases linearly with time until reaching a threshold where it imme
diately drops to zero, which reinitiates the process. The threshold voltage corresponds to a fixed
unit o f time. The linear voltage increase then acts as an electronic second hand, ticking o ff the
322
Chapter 8 First Order RL and RC Circuits
smaller units o f time. In practice, the linear increase in voltage is approximated by the linear part
o f an exponential response o f an RC circuit. ^X1ên the voltage across the capacitor reaches a cer
tain threshold, an electronic switch changes the equivalent circuit seen by the capacitor, allowing
the capacitor to discharge ver)^ quickly, i.e., the capacitor voltage drops to zero almost instanta
neously. Once the voltage nears zero, the electronic switch reinstates the earlier circuit structure,
causing the capacitor to charge up again. The process repeats itself indefinitely.
CHAPTER O BJECTIVES
1.
Explore the use o f a constant-coefficient first-order linear differential equation as a model
2.
for first-order RL and RC circuits.

Derive from the differennal equation model, the exponential response form (voltage or
4.
current) o f first-order RL and RC circuits without sources and with constant excitations.
Interpret the solution form o f the differential equation model in terms o f the circuit time
constant and the initial and final values o f the capacitor voltage or inductor current.
Develop techniques to handle s\vitching and piecewise constant excitations within first-
5.
order RL and RC circuits.

Investigate waveform generation and RC op amp circuits.
3.
SECTIO N HEADIN GS
1.
2.
Introduction
Some Mathematical Preliminaries
3.
Source-Free or 2^ro-Input Response
4. D C or Step Response o f First-Order Circuits

5. Superposition and Linearity
6 Response Classifications
7. Further Points o f Analysis and Theory
8 . First-Order RC Op Amp Circuits
9. Summary
11.
Problems
Our study prior to Chapter 7 focused exclusively on resistive circuits. Recall that all nodal equa
tions and loop equations for resistive circuits lead to (algebraic) matrix equations whose solution
yields node voltages and loop currents, respectively. Chapter 7 then introduced the capacitor and
the inductor. Interconnections o f sources, resistors, capacitors, and inductors lead to new and fas
cinating circuit behaviors. How? Inductors and capacitors have differential or integral voltage-cur
rent relationships. Interconnecting resistors and capacitors or resistors and inductors leads to cir
cuits that must satisfy both algebraic (KVL, KCL, and Ohms law) and differential or integral rela
tionships for L and C values. When only one inductor or one capacitor is present along with resis
tors and sources, these relationships lead to first-order RL and RC circuits. When the sources are
dc, such circuits have vohages and currents o f the form A + Be~^ for constants A, B, and X. The
main purpose o f this chapter is to develop techniques for computing the exponential responses o f
first-order RC and RL circuits driven by dc sources. A simple example serves to explain some of
these points.
In the series RC circuit o f Figure 8.1, suppose an initial voltage
is present on the capacitor,
where 0~ designates the instant immediately before zero. Often vve distinguish among 0, 0, and
O'*' when switching occurs or when discontinuities o f excitation functions occur at r = 0.
v(t)
v,(t)
FIG URE 8.1 Series/?Ccircuit.

A loop equation for the series RC circuit leads to
vp) = Ri(ir) +
Since iciO = C
(8.1)
equation 8.1 becomes

v,(/) =
at
+Vc(/)
Dividing through by RC yields the constant-coefficient first-order linear diflferential equation

( 8 .2 )
dt
RC
RC
subject to the initial condition Vf^Qr). This equation says that the derivative o f the capacitor volt
age plus MRC times the capacitor voltage equals MRC times the source voltage. The equation
enforces constraints on the capacitor voltage, its derivative, and the source voltage, and is differ
ent from the algebraic node or loop equations studied earlier. The terminology first-order differ
ential equation applies because only the first derivative appears. Equation 8.2 is linear because it
comes from a linear circuit. Our goal is to find capacitor voltage waveforms that satisfy the con
straints imposed by the differential equation 8.2.
Exercise. For the circuit o f Figure 8.1, show that the capacitor current i({t) satisfies a differential
equation o f the form
cti.it)
dt
1 .
RC
1 dv,{t)
dt
32 4
Our scope in this ciiaptcr is limited to circuits containing one inductor or one capacitor equiv
alently, first-order RL or RC circuits. W ithin this category we further constrain our investigation
to circuits with no sources but nonzero initial conditions, circuits driven by constant (dc) sources,
circuits driven by piecewise constant sources, and circuits containing switches. First-order circuits
driven by arbitrary source excitations are covered in later chapters using the Laplace transform
method.
2. SOM E M ATHEM ATICAL PRELIM INARIES

Ver)' often our interest is in source excitations such as v^{t) = 2e~^ V for /> 0 and 0 otherwise. To
conveniently represent such time-restricted waveforms, we define a signal called the unit step
function, denoted by u{t), as
(/) =
1
0
/ < ()
The unit step function is a universally used function and will appear many times in the remain
der o f this text. MATLAB code for specifying the step function is
function f = ustep(t)
t = t + le-12;
f = (sign(t)+l)*0.5;
With the unit step so defined, v^{t) - 2e~"'u(t) V, and both relations are plotted in Figure 8.2.
F IG U R E 8 .2 Unit step function and v^{t) = 2 e
V.
32S
Chapter 8 * First Order RI. and liC Circuits
Further, if v<^t) = l e
for t>t^ and 0 for r < /q, then v^{t) = lOf
r^) would be the prop
er representation because the shifted unit step function, //(/- /q), means
Plots o f v^{t) = 2e~û{t - ^q) and u{t - t^) are given in Figure 8.3 for
= 0.5.
FIGURH 8 . 3 . Plots of u{t - 0.5) and v^{t) = 2e ^'u{t - 0.5).
Exercise. Plot //(/) and (/q t). Hint: For what values o f t are the functions zero and for what
values are they 1?
A working model ot a physical system underlies an engineers ability to methodically anai)'ze,
design, or modify its behavior. Linear circuits are physical systems that have differential equation
models. The RL and RC circuits investigated in this chapter have differential equation models o f
the form
dx{t)
dt
= > ..V (0 + /(/ ).
-Y(/(,) = .Vo
(8.3a)
or, equivalently.
dxU)
dt
(8.3b)
valid for t> /q, where a-(/q) =

is the initial condition on the differential equations 8.3. T he term
J{t) denotes a forcing function. Usually, y(r) is a linear function o f the input excitations to the cir
cuit.
326
Chapter 8 First Order RI. and RC Circuits
Before proceeding, it is appropriate to explore the intuitive nature o f a differential equation.

Equations 8.3 are first-order constant-coefficient linear differential equations. They are first order
because o f the presence o f only the first derivative o f some unknown function x{t). For example,
in equation 8.3a the derivative o f x{t) equals a constant X times x{t) plus a known forcing func
tion y(r), w h e r e in c o r p o r a te s the effect o f all the circuit excitations. Rigorously speaking, lin
ear means that under the assumption o f zero initial conditions, if the pairs o f voltage waveforms
(/j(f), A'j(r)) and (fjit), X2 it)) each satisf)' equations 8.3, then for any scalars a^ and a^, the pair
{a/ît) + ajfjU),
+ ajXjU))
also satisfies equations 8.3.

The parameter X, denotes a riaturalfrequency o f the circuit. Natural frequencies are natural modes
o f oscillation such as, for example, in the ringing o f a bell. For physical objects natural frequen
cies are called natural modes o f vibration. All physical objects have a vibrational motion even
though it may be imperceptible. Knowledge o f these modes is important for the safety and reliabilit)' o f large buildings and bridges. For example, the Tacoma Narrows Bridge had natural modes
o f vibration that the wind excited. Undulations in the wind intensit}' resonated with the natural
vibrations of the bridge, causing a swaying motion to increase without bound until the bridge col
lapsed. In circuits, the natural modes o f oscillation are reflected in the shapes o f the voltage and
current w'aveforms the circuit produces. A more thorough and mathematical discussion o f the
notion o f natural frequency will take place in the next chapter, when we study second-order {RLQ
circuits.
Let us return to the goal of finding a solution to the differential equations 8.3. The solution to
equations 8.3 (a derivation will appear shortly) for t >tQ has the form
( c ^ ' - " /( x U k
Jk)
(8.4)
This means that the expression on the right-hand side o f the equal sign (1) satisfies the differen
tial equations 8.3 [its derivative equals K times itself plus/r)], and (2) it satisfies the correct ini
tial condition, xit^) = x^^. A simple example illustrates this point.
E XA M PLE 8.1.
Compute and verify the solution o f equation 8.3a using equation 8.4.
SO L U T IO N
Suppose in equation 8.3a., J{t) = u{t - 1), a shifted unit step function, X = - 1 ,
10, in which case
cit
From equation 8.4, for ^ > 1,

,v(/) =
\)(k=
+ I
= 1, and at(1) =
32'
Chapter 8 First Order Rl. and RC Circuits
To verify that [9^
+ 1] does indeed satisfy the differential equation, observe that for ? > 1,
dx{t) _ d
dt
+ I = -,v(/) + 1
= -9 e
~~dt
1] = 10, which is the mandatory initial condition. Thus, x{t) = 9^
Further, at ^ = 1, [9f
+ 1 is a valid solution for r > 1.

Example 8.1 spells out the application o f the solution (equation 8.4) to the differential equation
8.3a. It also verifies that the computed solution satisfies the differential equation and the proper
initial condition. Although not shown, equation 8.4 also satisfies equation 8.3b. A formal deriva
tion o f the solution o f equation 8.4 requires the use o f the integrating factor method, the sub
ject o f a differential equations course. Briefly, the first step o f this method entails multiplying both
sides o f equation 8.3a or 8.3b by a so-called integrating factor e~^. For equation 8.3b, this results
in
dx(t)
_>j
------ h e x i l ) = e
dt
f/,\
f{ t )
(8.5)
By the product rule for differentiation, the sum on the left equals
d_
dt
e""x{t)
in which case equation 8.5 becomes

( 8 .6)
dt
One can integrate both sides o f equation
f -dr
J'o
8 .6
from
Tq
to t as follows:
dr = e
(8.7)
JI q
Bringing the term e'Ôx{tQ) to the right-hand side o f equation 8.7 and multiplying through by
results in the solution to the differential equation 8.3a or 8.3b, given by equation 8.4. This
completes the derivation o f the very powerful formula o f equation 8.4.
There are four points to remember about the preceding discussion: (1) circuits have behaviors
modeled by differential equations such as equations 8.3; (2) the solution to a first-order differen
tial equation is a waveform (also called a signal or response) satisfying equation 8.4; (3) the for
mula o f equation 8.4 works for all continuous and piecewise continuous time functionsy(/); and
(4) a solution to a differential equation means that the waveform satisfies the given differential
equation with the proper initial condition.
328
Chapter 8 Hirst Order RL and RC Circuits
Exercise. Show that the hinction .v(/) = (1 0. ôr r > 0, is a solution to the difFerential equation
^
dt
= -.v(/) + //(/) with initial condition a-(0) = 0 by showing that x{t) satisfies the difFerential
equation and has the proper initial condition at r= 0.
3. SOURCE-FREE OR ZERO-IN PUT RESPONSE

Figure 8.4 depicts the most basic (undriven or source-free) RL or RC circuit: a parallel connection of
a resistor with an inductor or a capacitor without a source. In these circuits, one assumes the pres
ence of an initial inductor current or initial capacitor voltage. The complication introduced by a volt
age or current source is taken up later. Once the source-free or zero-input behavior is understood,
one can understand more easily the responses resulting from constant source excitations.
i,(t)
>r
+
^
yf
v^(t)
R ^
ijt )
\ fi)
^R(t) <
-
S v jt )
(a)
(b)
FIGURK 8.4
Our first goal is to derive differential equation models for the RL and RC circuits o f Figures 8.4a
and 8.4b, respectively. We do this in parallel.
(1) At the top node o f Figure 8.4a, KCL implies
(1) Similarly, for Figure 8.4b, KVL implies
= v^t)
ifiU) = -iiit)
(2) However,
(2) However,
Vi it)
.d\'c{t)
n it ) R icit) RCdt
L d ii U)
(3) Making the obvious substitution and
(3) Making the obvious substitution and
multiplying by R/L yields the differential
dividing by RC yields the differential
equation model
equation model
dii it)
with //(/()) a given initial condition.
dV( {t) _
(8.8a)
dr
with
RC
v'c(/)
a given initial condition.
(8-81 )
Both differential equation models have the same general form,
^
dt
= hc{t) = - - m
(8.9)
i.e., the derivative of x(t) is a constant, X = -1/x, times itself. Applying equation 8.4 to equation
8.9 implies that both equations 8 .8a and 8 .8 b have solutions given by
where x is a special constant called the time constant o f the circuit. Equation 8.10 means that the
responses for
W
(q o f the undriven JiL and /?C circuits are, respectively, given by
1^(0 =e
-^ '-'0 )
vc(0 = e
- ^ ' - 0)
(8.11)
vc(/q)
where the time constant o f the RL circuit is T = and the time constant o f the RC circuit is x
= RC.
The time constant of the circuit is the time it takes for the source-free circuit response to drop to
e~^ = 0.368 o f its initial value. Roughly speaking, the response value must drop to a little over onethird of its initial value. This is a good rule of thumb for approximate calculations involving decaying exponentials.
The mathematics that underlie the solution to the differential equation 8.9 given in equation 8.10
is nothing more than elementary calculus. To see this, consider the exponential solution form
(8. 12)
where K 'ls an arbitrary constant. The fiinaion
has the property that its derivative is----- e~' ^
This is precisely what equation 8.9 requires. Therefore equation 8.12 satisfies the differ-
ential equation 8.9 and is said to be a solution. To completely specify x{t) it only remains to iden
tify the proper value of K from the initial condition. Evaluating x(r) at ^
yields
Mt) = /Cf-Vr
in which case
Substituting this value o f i n t o equation 8.12 produces the solution given in equation 8.10,
which is adapted to specific RL and R C circuits in equations 8 . 11. Figure 8.5 plots equation 8.12
for arbitrary K and x > 0 . This plot proves instructive for understanding how the response decays
as a function o f the time constant.
330
Chapter 8 First O rder RL and RC Circuits
Time
FIGURI: 8.5 Plot o f equation 8.12. For f = x, one time constant,
decays to 0.368
of its maximum value.

In summary, the circuits o f Figure 8.4 motivate the development o f the rudimentary machinery
for constructing solutions to undriven RL and RC circuits. For more general circuits, those con
taining multiple resistors and dependent sources, it is necessary to use the Thevcnin equivalent
resistance seen by the inductor or capacitor in placc o f the R in equation 8.11. Figure 8.6 illus
trates this idea.
Linear
\
>f
Resistive
Circuit
i,(t)
\
Linear
Resistive
Circuit
L
p
No
Sources
-N V jt)
No
Sources
i,(t)
+
v,(t)
F I G U R I : 8 .6 R ep la c em en t o f resistive part o f c irc u it by its T h e v e n in eq u iv alen t.
331
These facts imply that the general formulas for computing the responses o f undriven RL and RC
circuits have the structures
(8.13)
VrU) = e
The difference between equations 8.11 and 8.13 is that in equations 8.13 R^f^ is the Thevenin
equivalent resistance seen by the inductor or capacitor.
EXA M PLE 8.2

For the circuit o f Figure 8.7, find i^{t) and v^(t) for r > 0 given that
S closes at r = 0.4 sec. Then compute the energy dissipated in the 5
val [0.4, co).
= 10 A and the switch

resistor over the time inter
t = 0.4sec
S
20Q
5Q
+
v,(t)
8H
FIGURE 8.7 Parallel RL circuit containing a switch.

S
o l u t io n
Step 1. With switch S open, compute the response for 0 < /< 0.4 sec. From the continuity property
o f the inductor current, /^(O'*) = ^^(0) = 10 A. Using equation 8.13,
'1 ( 0 = c
We note that //(0.4) = 3.679 A.

Step 2. With switch S closed, compute the response for t > 0.4 sec. For this time interval the Thevenin
equivalent resistance seen by the inductor is
= 20||5 = 4 Q, i.e., the equivalent resistance o f a par
allel 20 Q and 5 ^ combination. According to equation 8.13, the response for t>tQ = 0.4 sec is
,- ,( 0 = /
Step 3. Write the complete response as a single expression using step fitnctions:
i^{t) = 10^>-2-5^[(^) - (/- 0.4)] +
0.4)A
(8.14)
Chapter 8 * First Order RL and R C Circuits
332
Step 4. Plot the complete response. To plot this using MATLAB, we use the following m-file along
with the code given earlier for the unit step function:
t = 0:0.005:1.4;
iL= 10*exp(-2.5*t) .* (ustep(t).* ustep(t - 0.4)) + 3.679*exp(-0.5*(t-0.4)) .* ustep(t - 0.4);
plot(t,iL)
grid
Using this code, Figure 8.8 illustrates the complete response, showing the two different time con
stants. The 0.4 sec time constant has a much faster rate o f decay than the lengthy 2 sec time con
stant.
a
E
<
Time (seconds)
FIGURE 8.8 Sketch of response i^{t) for Example 8.2.
Step 5. Compute i^^(r). It is a simple matter now to compute v^^t) since

vi^t) =
In particular, t'^(0'*') = -200 V. Hence for 0 < t< 0.4
.IL l
For t> 0.4, however, the circuit structure changes and

= 14.716 V. Thus,
yi(f) = e
(0 .4 "') =
in which case i'^(0.4'^) = 4 x 3 .679
Step 6. Compute the energy dissipated in the 5 O. resistor over the interval [0.4, oo). The power
absorbed by the 5
resistor for 0.4 < / is

vi ( 0
5
= 43.31
The energ)' dissipated over [0.4, oo) is given by

IVjn ( 0 .4 ,cc) =
PfaU/)A/ = 4 3 .3 12 j;^ ^
= 4 3 .3 12 J
Exercises. I. Plot vît) using the above m-file, ustep, and the appropriate .code.
2. Repeat the calculations o f Example 8.2 with the 8 H inductor changed to 8 mH and a switch
closing time o f 0.4 ms.
A N SW ER: ij{t) = l{)r-^'^"'//(f)//(0.4 x lO"-^ - r) +
- 0.4 x Ur-^) A
E X A M PLE 8.3
Find V(^t) for r > 0 for the circuit o f Figure 8.9 given that y^^Ô) = 9 V.
So l u t io n
Because there is a switch that changes position at r = 1 sec, there are two time intervals to consider.
Step 1. Compute the response forO < t < 1. Over this time interv\il, the equivalent circuit is a par
allel /?C circuit, as shown in Figure 8.10a.
334
0.1 F
0.1 F
80
3Q
V jt)
+
V jt)
(a)
(b)
FIG URE 8.10 Equivalent circuits for Figure 8.9: (a) 0 :s r < 1 and (b) 1 s t.
) = 9 V. Therefore from equation 8.11,
By the continuity o f the capacitor voltage,

1
------- 1
v c(t)= e
Vc(O^) =9^- 5^V
Step 2. Compute the response for r > 1. Figure 8.10b depicts the pertinent equivalent circuit.
Observe that
= 2.58 V and
=3
Again by equation 8.11, for r >

r-I
-St-to)
V c (0 = f
= 1,
v .c (^ ) = 2 .5 8 e
Step 3. Use step functions to specify the complete response. By using the shifted unit step function,
the two expressions obtained previously can be combined into a single expression:
t-\
V cir)=9e~^-'^'[u{t)-u{f-\)] + 2.5Se
- \) V
Step 4. Obtain a plot o f the response. Using MATLAB and code similar to that used in Example
8.2, the plot in Figure 8 . 1 1 w'as obtained. Here the part o f the response with the 0.3 sec time con
stant shows a greater rate o f decay than the longer 0.8 sec time constant.
Time (seconds)
F IG U R E 8 .1 1 Response,
for the circuit o f Figure 8.9 .
Chapter 8 First Order RI, and RC Circuits
Exercises. 1. Show that ti{t) - u{t - 1) =
- t)-
2. Suppose that in Example 8.3 the switch moves to the 4.5 ^ resistor at r = 0.5 sec instead oF 1
sec. Compute the vakie V(\t) at f = 1.2 sec.
ANSW ER: 0.4671 V
For all o f these examples x > 0 and the response is a decaying exponential. Intuitively, the response
decays because the resistor dissipates as heat the energy initially stored in the inductor or capaci
tor. One o f the homework exercises will ask the student to show that the total energ)' dissipated
in the resistor from
to oo equals the decrease in energ)^ initially stored in the inductor or capac
itor at ^Q. When controlled sources are present,
may be negative, in which case x < 0. Here the
negative resistance supplies energ)' to the circuit and the source-free response will grow exponen
tially. This is illustrated in the next example.
E X A M PLE 8 .4
Find Vf^t) for the circuit o f Figure 8.12, assuming that^^^^ = 0.75 S and i^(;(0) = 10 V.
------ o -----q V
0.25F
4Q
+
v,(t)
<
0 -----0.25F
-2 Q
v,(t)
-o-
-o-
FIGURE 8.12 Parallel /?Ccircuit with dependent currcnt sourcc.
So l u t io n
It is straightforward to show that theThevenin equivalent seen by the capacitor is a negative resist
ance,
- -2 Q, as shown in Figure 8.12b. Again, by equation 8.11,

v cit) = e
vc{0^)=\0e^'u{t) V
Because o f the negative resistance, this response grows exponentially, as shown in Figure 8.13. A
circuit having a response that increases without bound is said to be unstable. Practically speaking,
an unstable circuit will destroy itself or exhibit a nonlinear phenomenon that clamps the voltage
at a finite value, as in the case o f saturation in an op amp.
336
Tim e (seconds)
FIGURE 8.13 Plot of unbounded voltage response due to presence of negative resistance.
Circuits with such responses arc said to be unstable.
Exercises. 1. For Example 8.3 show that

icU ) = e
2. in Example 8.3,
f > 0.
ANSWI-.l^S: 8 2.
ic{0-^) = 5e-'ii{t)A
= 0.125 S. Find the equivalent resistance seen by rhc capacitor and

V
3. Show that in general, for / >
the form o f the capacitor current is similar to the voltage form.
Hint: Apply the capacitor v-i relationship to equation 8.11.
4. DC OR STEP RESPONSE OF FIRST-ORDER CIRCU ITS

The circuits o f the previous section had no source excitations. This section takes up the calcula
tion o f voltage and current responses when constant-voltage or constant-current sources are pres
ent. It is instructive to start with the basic series RL and RC circuits as shown in Figure 8.14.
Linear
Resistive
Circuit
with
i^Ct)
Linear
Resistive
Circuit
with
Constant
Sources
Constant
Sources
(a)
(b)
\{t)
FIGURE 8.14 (a) Driven first-order RL circuit, (b) Driven first-order RC circuit, (c) Thevenin
equivalent representation oF (a), (d) Thevenin equivalent representation of (b).
Given these basic circuit representations and initial conditions at Tq, what is the structure o f a dif
ferential equation mode! that governs their voltage and current behavior for t>tQ' The first objec
tive is to derive the differential equation models characterizing each circuits voltage and current
responses. It is convenient to use ij{t) as the desired response for constructing the differential
equation for the series RL circuit (Figure 8 .l4 c ), whereas for the series RC circuit (Figure 8 .l4 d ),
is the more convenient variable.
(1) The circuit mode! for the inductor is
vl O )= L
(i) The circuit mode! for the capacitor is
dv(^(t)
dilit)
ic(t)= C
dt
dt
(ii) By KCL and Ohms law,
(ii) By KVL and O hm s law,
ic (t) -
(iii) Substituting for ;^(^) leads to the

differential equation model
d iijt)
dt
R.
(iii) Substituting for /^r) leads to the differentia!

equation model
dvc(r)
+
(8-> 5a)
ôc R,ill
dt
_
-^ ^ 'C (0 + -
v o c (8 .1 5 b )
338
initial condition V(\t^ ) =
with initial condition //(ô") =

is constant (not impulsive).
since
is constant (not impulsive).
Exercise. Construct differential equation models for the parallel RL and RC circuits o f Figure
8 . 15 . Note that these circuits are Norton equivalents o f those in Figure 8.14a and Figure 8.14b.
Again choose /^(r) as the response for the RL circuit and v^^t) as the response for the RC circuit.
(constant)
(constant)
FIG URE 8.15 Driven RL and RC parallel circuits.
dll (/)
.\NSWI-RS: V di
V/(
I.
//(/)+
L
and ^ - = ------^ ''r ( n +

d!
R,i,C ^
C
A simple application o f basic circuit principles has led to the two differential equation models of
equations 8.15. The next important question is: What do these t%vo differential equation models
tell us about the behavior o f each circuit? Equivalently, how do we find a solution to the equa
tions? Observe that both differential equations 8.15 have the same structure:'
dx(t)
dt
(8.16a)
for RL circuits and x = R^j^C for RC circuits, and F= v J L for RL
where the time constant T =
circuits and F = vJiRî^j for the RC case. This equation is valid for t>
Equation 8.4, rewritten
here with/^y) = F, presents the general formula for solving the differential equation 8.16a:
) + ( e^^^'-f^Fdcj
A-(/)=
where A =
(8.16b)
'h)
is a natural frequenc)' o f the circuit, and where we have emphasized the use o f
the initial condition at
Note that as long as x{t) is a capacitor voltage or inductor current, the
initial condition is continuous, i.e., x{t^p = x(fQ+), because F h a . constant (non-impulsive) forc
ing function. A straightforward evaluation o f the integral o f equation 8.16b yields
t
x(/) = e
^ \v(fo)+ F e
( I-In
f e
Jif)
dq = e
^ 'x (t^ )+ F T
Some rearranging o f terms in equation 8.16c produces the desired formula
- e
(8.16c)
339
(t-tn
x(
0=
f t
+ U 4 )-
ft
(8.17)
= -r
which is valid for t > /q. After some interpretation, this formula will serve as a basis for comput
ing the response to RL and RC circuits driven by constant sources. A homework exercise will ask
for a different and direct derivation of this formula.
At this point it is helpful to interpret the quantity i r in equation 8.17. For RL circuits, when x(r)
= /*(/), equation 8.15a implies that
v J L , x = UR^f^ and hence F l = vJR^f^ =
For RC cir
cuits when ;c(/) = V({t), equation 8.15b implies that F = vJR ^ C , x = Rf^,C, and hence Fz =
This interpretation is valid for both positive and negative values of x. If x > 0, then
t-tQ y
jr(oo) = lim x(t) = lim
/*00
t~ * 0 0
Fx + \ x {t^ )-F x y
isc =
= Ft =
for RL case
Rih
(8.18)
for R C case
This means that for the RL case, /^(oo) =
= vJR^f^ and for the RC case, V({<x>) =
O f course,
is computed by replacing the inductor with a short circuit, and

is computed by replacing
the capacitor with an open circuit. See Chapter 6 for details. Mathematically, any constant, such
as x(/) = constant, that satisfies a differential equation is called an equilibrium state of that dif
o
ferential equation. Since the constant x(/) = F i satisfies the differential equation 8.16, / r is an
equilibrium state of the differential equation 8.16.
Whenever X > 0, equation 8.18 implies that the formula (equation 8.17) for the solution o f equa
tion 8.1 6 given constant or dc excitation becomes
.
a: ( 0
_ ^-'o
(8.19a)
= j :(>) + U ( ^ ^ ) - J c( oo) e
and when x(^) = /(/),

(t-to )
iL (0 = / z .( ) + k ( 4 ) - ' L ( )
o
and when x(/) = V({t)y
(8.19b)
t-t
vc(0 = vc(00) + Vc(ô )-^ c ()
K/.C
(8.19c)
Note that x > 0 is true whenever R^j^ > 0, C > 0, and Z, > 0, i.e., the circuit is said to be passive.
This allows us to state a nice physical interpretation of equation 8.19a:
elapsed time
w
w
's . ;
x { t )= [ F in a l value^ + iÎnitial v a lu e ]-[F in a l value\)e ttê constant

Graphically, equation 8.19a is depicted in Figure 8 .1 6 for xipo) >
x (/ q ).
340
Elapsed Time
I'iG U RE 8.16 Graphical interpretation o f equation 8.19a for the case xico) > x{tQ).
Exercise. Redo the curve o f Figure 8.16 for the case x{<x)) <
The initial value
computed from initial conditions and possibly the value o f the source
excitation, or it can be computed from past excitations up to

trate the use o f equation 8.19.
Several examples will now illus
EXA M PLE 8.5

For the circuit o f Figure 8.17, suppose a 10 V unit step excitation is applied at r = 1 when it is
found that the inductor current is
= 1 A. The 10 V excitation is represented mathematically
as
= 1 0 (/ - 1) V for r> 1. Find ij{t) for r > I.
R = 5Q
v^(t)
1^(1-)=1 A
F IG U R L 8 .1 7 Driven series R L circuit for Example 8.5 with /^(1) = 1 A.
.Vtl
S o l u t io n
Step 1. Determine the circuits differential equation model. Since the circuit o f Figure 8.17 is a driv
en series RL circuit, equation 8 .1 5a implies that the differential equation model o f the circuit valid
for r > 1 is
cl'iAt)
10
w here the tim e c o n sta n t t = 0,4 sec.
Step 2. Determine the form o f the response. Since /^(1 ) =
equation 8.19b implies that
/^(/)=/^(oo)+^/^(r)
Here the presence o f u{t - 1) emphasizes that the response is valid only for / > 1.
Step 3. Compute iî^X)) and set forth the fin al expression for /^(r). Since x = 0.4 > 0, we replace the
inductor in the circuit o f Figure 8.17 with a short circuit to compute i^^. = //(oo) = 2 A. It follows
that
[2 + (1 -
- 1) = (2 -
- 1) A
Step 4 . Plot iît). One cannot presume that the response is zero for r < 1. Hence, using MATLAB
or the equivalent, one can construct the graph o f i^{t) for / > 1 as given in Figure 8.18.
c
0;
u
O
tj
3
o
c.w
Time (sec)
FIGURE 8.18 Plot of /^(r) for Example 8.5.
Step 5. Compute vît). Given the expression for the inductor current in step 3, it follows that for
t> 1,
cliLit)
Vi{t) = Ldt
/-I
u{t -\^ ) = 5 e
* ^ / (r -l'* ')V
342
Exercises. 1. Verify that in Example 8.5 vît) can be obtained without differentiation by
V -V iW 2. In Example 8.5, suppose R is changed to 4 Q. Find iît) at r = 2 sec.

ANSW ER: 1.8647 A
Note that we have used the differential equation 8.16 (or equations 8.15) to obtain the solution
form o f equation 8.19. However, when using equations 8.19, it is not necessary to reconstruct the
differential equation of the RL or RC circuit. Specifically, we need only compute xit^), x{cc), and
the time constant x = LIRî^ or
The method described for computing final values can also be used to find the initial values o f
and i^ at f =
if dc excitations have been applied to the circuit for a long time before t = ?q. The
next example illustrates this technique and extends the preceding discussion.
EXA M PLE 8.6
The source in the circuit o f Figure 8.19 furnishes a 12 V excitation for / < 0 and a 24 V excita
tion for r < 0, denoted by yy^(r) = [\2u{-t) + lAu{t)] V. The switch in the circuit closes at / = 10
sec. First determine the value o f the capacitor voltage at r = 0", which by continuity equals
Next determine
for all ^ > 0.

R = 6 kn
+
v,(t)
FIGURE 8.19 Switched driven
circuit for Example 8.6.
So l u t io n
Step 1. Compute initial capacitor voltage. For r < 0, the 12 V excitation has been applied for a long
time. Therefore, at r = 0 , the capacitor has reached its final value and looks like an open circuit
to the source. Hence the entire source voltage o f 12 V appears across the capacitor at r = 0~, i.e.,
^(40~) =
= 12 V by the continuity property o f the capacitor voltage.
Step 2. Use equation 8.19c to obtain vît) fb r Q < t< 10 sec. Equation 8 .1 9c requires only that we
know V(iQ*) (step 1), x, and Vf^co). For 0 < t< 10 sec, r = R^C= 3 sec. It is important to realize
here that for 0 < r < 10 sec the circuit behaves as if the switch were not present. Hence, the com
putation o f v^(co) proceeds as if no switching would take place at r = 10 sec. Here v^^ = v^ôo) =
24 V. Hence, for 0 < ^ < 10, equation 8.19c implies
.vi3
v'cW = ''c(= o ) + [wc(0^)
= 2 4 + (1 2 -2 4 )c^ - ^ = 2 4 - 1 2 ^
Step 3. Compute the initial condition fo r the interval 10 <
(8.20)
i.e., t'^^lO'^). Plugging into equation
8.20 and using the continuity property o f the capacitor voltage yields
= 2 4 - 12f>-io/3 = 2 3 .57 V
Step 4. Find V(^t) for / > 10. For r > 10, the resistive part o f the circuit can be replaced by its
Thevenin equivalent, which yields Figure 8.20.
Rtn = 2 k O
+
V(t)
FIGURE 8.20 Circuit equivalent to that of Figure 8.19 for / > 10.
Here, equation 8.19c applies again. The value for y^^co), however, is now 8 V and the new time
constant is
= 1 sec. Hence, for ; > 10,

/-lo
v c (/ )= V c ( ) + [v c (IO *)-V (-(= c )]e
= 8 + ( 2 3 . 5 7 - 8 ) t '- * '- '" = 8 + I5 .5 7 e * ' '" V
Step 5. Set forth the complete response using step functions. Using step functions, the response V(\()
for f > 0 is
v ^ {t)= 2 A -\ le
+ S + \5.51e^'
Step 6. Plot V(it). Plotting i^(^t) yields the graph of Figure 8.21.
3-m
o
OJ
O)
Q.
Time (seconds)
FIGUllK 8.21 Capacitor voltage
for r > 0 .
Exercise. Suppose the switch in Example 8.6 opens again at r = 20 sec. Find vît) at r = 25 sec.
ANSW ER: 20.98 \
EXA M PLE 8 .7 The circuit o f Figure 8.22a has a capacitor voltage given by the cur\'e in Figure
8.22b. We note that ^fÔ.l) = 7.057 V. Find, Kq, y^Ô), the time constant r = RC, the exact value
of j/(j(Q.25), and the value o f C i f /? = 100^2.
------ O-
v.(t)= V u (t)
v,(t)
-o(a)
345
Chapter 8 First Order lU. and RC Circuits
01
IB
I
Q.
fO
U
T im e (s)
(b)
FIGURE 8.22 (a) Scries /?Ccircuit, (b) Capacitor voltage,

So
lu t io n
. A simple inspection o f the graph indicates that
V(^{Q) = 2 V. One recalls that

/
) - \(-(0C) e ^
Hencc as / -> oo, v^{t) -> v^co) = 10 V. Since the capacitor looks like an open circuit at / = oo, Vq
=
= 10 V. From the given problem data,

y ^ 0 .1 ) = 7 .0 5 7 = 10
Simplifying yields
-
T =
0.1
/ 1 0 - 7 .0 5 7
= 0.
8
Therefore C = 0.01 F and M 0 .2 5 ) = 9..34.33 V.
When switching occurs frequently, or the excitation changes its constant level frequently, then
hand analysis, as in lixample 8.6, becomes very tedious. For such problems a SPIC E simulation
(or the equivalent) proves useful and saves time. The next example u.ses SPIC E to compute the
waveform o f a simple RC circuit whose input excitation is a square wave. Like the previous exam
ple, the solution is broken down into time intervals such that during each time interval inputs are
constant. Because no switching occurs, the time constants for all time intervals are the same. In
applying equation 8.19c, the quantities that vary from one time interval to the next are the initial
values and final values.
346
EXA M PLE 8.8

The first-order RC circuit o f Figure 8.23a is excited by the 50 Hz square wave input voltage o f
Figure 8.23b given that the capacitor is initially relaxed.
(a)
(b)
Plot
for 0 < r < 60 ms, using SPIC E or equivalent software.
Find the initial value and the final value in equation 8.19 when t is very large, for exam
ple, at the beginning and end o f the interval 1 < t< 1.01 sec. Plot the v^t) wave for this
interval using MATLAB or the equivalent.
V (t) (V)
> t (msec)
10
20
30
(b)
FIG U RE 8.23 (a) Series RC circuit excited by the 50 Hz square wave o f (b).
So
lu t io n
Part (a)
Doing a SPICE or equivalent simulation gives rise to the response curve shown in Figure 8.24, over
which the square wave input is superimposed. Observe that the response V(\t) has an approximate tri
angular shape. What is happening is that from zero to 10 msec, the circuit sees a step and hence the
capacitor voltage rises toward one volt. At 10 n:isec, the square goes to zero for 10 msec. The capacitor
then discharges its stored energ)' through the resistor, causing a decrease in its voltage value. The decrease
does not go to zero, however. So when the square wave again is at 1 volt the capacitor voltage begins to
rise again and achieves a slighdy higher value at f = 30 msec compared to f = 10 msec. In fact, one notices
in Figure 8.24 that the peak and minimum values are increasing slighdy as time increases. Eventually the
peak and minimum values will reach their respective fixed viilues, c;illed steady-state values. To find these
values, a simulation program could require a very lengdiy simulation interv'al, which often proves
impraaical. The steady-state values can be computed analytically as in part (b).
10
20
30
40
50
t(m se c)
FIG URE 8.24 Response of circuit of Figure 8.23a calculated using SPICE. For reference, the input
square wave excitation is superimposed on the plot.
_____________________________________________________
Part (b)
Let Tq = mT, where 7'= 20 msec is the period o f the square wave and m is some large integer. Then,
0.57-
+ 0.571 = 1 + {v^t^) - 1) /
Further, in steady state, V(it^ + T) =
= 1+
- 1]^-
(8.21)
which implies that
0.5T
+ 71 =
+ 0 .5 7 ) ^ ' KC .
+ 0.571^'-' =
(8.22a)
Equivalently, equation 8.22a implies that

+ 0 .5 7 ) = V(itQ)e^
(8.22b)
Substituting equation 8.22b into equation 8.21 yields
= 1 + [i/c(ro)the solution o f which is
^ "
1+e
= 0-2689 V
It follows that
r/J/o + 0 . 5 7 ) = i^(3<ro)e> = 0 .7 3 1 1 V
An examination o f the response in Figure 8.24 shows that the minimum and peak values are
approaching the steady-state values o f 0.26 8 9 V and 0.7311 V, respectively.
Exercise. Based on the response in Figure 8.24, roughly sketch the capacitor current,
At
what time instants is the capacitor current discontinuous?
5. SUPERPO SITION AND LINEARITY

Superposition, a special case o f linearity, helps simplify the analysis o f resistive circuits, as discussed
in Chapter 5. Recall that linear resistive circuits are interconnections o f resistors and sources, both
dependent and independent. Does superposition still apply when capacitors and inductors are
added to the circuit? The answer is yes, provided one properly accounts for initial conditions.
In order to justify the use of superposition for RC and RL circuits, consider that resistors satisfy
O hm s law, a linear algebraic equation. Capacitors satisfy the differential relationship
; - r -----ir
^
dt
r\
348
which is also a linear equation. To see linearity in this i-v relationship, suppose voltages
and
V(^ individually excite a relaxed capacitor producing the respective currents

:
Let
_^ dvc2
, >c2 - c
a - c
be the current induced by a voltage equal to the sum o f
and
i.e.,
C3 = (^-^(^Cl+^C2)
However, the linearity o f the derivative implies the property of superposition;
.
ic=C-
_ dvf
^
+ C = ic i + t 2
dt
dt
By the same arguments, the current due to the input excitation t/Q - ^\^c\
+
On the other hand, suppose two separate currents / q and
^2^Cl
^C3 ^l^Cl
individually excite a relaxed capac
itor C Each produces a voltage given by the integral relationship
^Ci (0 =
(<:) d t , vc2 (0 =
ic2 (T^) ^
By the distributive property of integrals, the combined effect of the input,
+ <?2^C2 would
be a voltage,
vc3 (0 = ^ f_ Ja \ ic \ (T ) + a2ic2('^)] d r
= ai
\ 7 S ' - J c 2 W ^
Thus linearity and, hence, superposition hold.

Arguments analogous to the preceding imply that a relaxed inductor satisfies a linear relationship,
and thus superposition is valid, whether the inductor is excited by currents or by voltages.
The interconnection of linear capacitors and linear inductors with linear resistors and sources sat
isfying KVL and KCL produces linear circuits because KVL and KCL are linear algebraic con
straints on the linear element equations. Hence, the property of linearity is maintained, and as a
consequence superposition holds for the interconnected circuit.
To cap off this discussion we must account for the presence of initial conditions on the capacitors
and induaors o f the circuit. For first-order RC and RL circuits, this need is clearly indicated by
the first term o f equation 8.4. For a general linear circuit, one can view each initial condition as
being set up by an input that shuts off the moment the initial condition is established. Hence the
r-
efFect o f the initial condition can be viewed as the effect o f some input that turns o ff at the time
the initial condition is specified. This means that when using superpositioti on a circuit, one first
looks at the effect o f each independent source on a circuit having no initial conditions. Then one
sets all independent sources to zero and computes the response due to each initial condition with
all other initial conditions set to zero. The sum o f all the responses to each o f the independent
sources plus the individual initial condition responses yields the complete circuit response, by the
principle o f superposition. A rigorous justification o f this principle is given in a later chapter using
the Laplace transform method.
The following example illustrates the application o f these ideas.
E XA M PLE 8.9
The linear circuit o f Figure 8.25 has two source excitations applied at r = 0, as indicated by the
presence o f the step functions. The initial condition on the inductor current is
= -1 A.
Compute the response /^(r) for r > 0 using superposition.
So
lu t io n
Because the circuit is linear, having a linear differential equation, superposition is but one o f sev
eral methods for obtaining the solution. An alternative approach is to find the Thevenin equiva
lent circuit seen by the inductor. As we will see, the superposition approach sometimes has an
advantage over the Thevenin approach.
Superposition must be carefully applied, however. First one computes the response due only to the
initial condition with the sources set to zero. Second, one computes the response due to Vj with
all initial conditions and all other sources set to zero. Third, one computes the response due to /j
with all initial conditions and all other sources set to zero. Finally, one adds these three responses
together to obtain the complete circuit response.
Step 1. Compute the part o f the circuit response due only to the initial condition, with all independ
ent sources set to zero. With both sources set to zero, there results the equivalent circuit given by
Figure 8.26. The Thevenin equivalent resistance is
= 4 Q, resulting from the parallel combi
nation o f
and R->. Figure 8.26 depicts the equivalent undriven RL circuit having response
352
Oliapccr 8 First Order RI. and RC Circuits
An approach based on the Thevcnin equivalent circuit seen by the inductor would allow one to
quickly compute the complete response, but not in a way that identifies the contributions due to
each o f the individual sources. Answêrs to the preceding three questions would have required
repeated solutions to the circuit equations. However, if one keeps the source values in literal form,
then the Thevenin equivalent approach would be as efficient.
6. RESPONSE CLA SSIFICA TIO N S

Having gained some understanding of the form o f the behavior of RL and RC circuits, it is instruc
tive to classify the responses into categories. The zero-input response o f a circuit is the response
to the initial conditions when all the inputs are set to zero. The zero-state response o f a circuit is
the response to a specified input signal or set of input signals given that the initial conditions are
all set to zero. By linearit)', the sum o f the zero-input and zero-state responses is the com plete
response o f the circuit. This categorization is the convention in advanced linear systems and lin
ear control texts.
I'Vequently circuits texts include two other notions o f response, the natural response and the
forced response. However, decomposition o f the complete response into the sum of a natural and
a forced response applies only when the input excitation is (1) dc, (2) real exponential, (3) sinu
soidal, or (4) exponentially modulated sinusoidal. Further, the exponent of the input excitation,
for example, a'm j{t) =
must be different from that appearing in the zero-input response.
Under these conditions it is possible to define the natural and forced responses as follows: (1) the
natural response is that portion of the complete response that has the same exponents as the zeroinput response, and (2) the forced response is that portion of the complete response that has the
same exponent as the input excitation provided the input excitation has exponents different
from that of the zero-input response.
This decomposition is important for rwo reasons. First, it agrees with the classical method o f solv
ing linear ordinary differential equations with constant coefficients where the natural response cor
responds to the com plem entary function and the forced response correspontls to the particular
integral. Students fresh from a course in linear differential equations will feel quite at home with
these concepts. The second reason is that the forced response is easily calculated for dc inputs. For
general systems this type o f decomposition is not used.
7. FURTHER POINTS OF ANALYSIS AND TH EO RY

In deriving equations 8.17 and 8 .1 9a, the quantin' x{t) was thought o f as a capacitor voltage or an
inductor current. It turns out that any voltage or current in an RC or RL first-order linear circuit
with constant input has the form
{I- to)
x{l) = X^,+
~ r~
(8.23)
For T negative or positive equation 8.23 is identical to equation 8.17 with Pi = X^. Further, T is
the circuit time constant and X is that voltage or current of interest computed under the condi-
Chapter 8 First Order Rl. ami RC Circuits
3 S3
tion that the inductor is replaced by a short circuit for the RL case or the capacitor is replaced by
an open circuit for the RC case.
How do we justify the form o f equation 8.23 for all variables? We invoke the linearity theorem o f
Chapter 5 and the source substitution theorem o f Chapter 6 o f 2"^ edition. Suppose in a firstorder RC circuit we have found V(\t). By the source substitution theorem, the capacitor can then
be replaced by a voltage source whose voltage is the computed V(it). This new circuit consists o f
constant independent sources, one independent source o f value
and resistors and depend
ent sources. By linearity, any voltage or current in the circuit has the form
x(r) =
for appropriate ATj and Kj- By equation 8.17,
which implies that
Mr) = K , + ^6^
for appropriate
"
This is the same form as equation 8.23 for proper choices o f
Exercise. Show that, for t > 0,
and
-X^.
^rid
Note that equation 8.23 requires that the initial value be evaluated at ^
instead o f r =
This
is because only the inductor currents and the capacitor voltages are guaranteed to be continuous
from one instant to the next for constant input excitations. The capacitor current and the induc
tor voltage as well as other circuit voltages and currents may not behave continuously.
E XA M PLE 8 .1 0
This example illustrates the application o f equation 8.23. For the circuit o f Figure 8.29,
-18//(-r) +
V. Find i- it) for f > 0.
6kQ
3kQ
2kQ
v jt )
0.5 mP
FIGUFIK 8 .2 9 R C circuit with / (/) as the desired response.
354
SO L U T IO N
Step 1. Compute /y(0^). To obtain /y(0*), we first compute V(\Q~) =
Since for r < 0, - 1 8
V has excited the circuit for a long time, the capacitor looks like an open circuit. By voltage divi
sion,
j = ( - 1 8 ) = - 6 V
Thus at r = 0*, the equivalent circuit is as shown in Figure 8.30.
rigurc K.30 Circuit equivalent to that of Figure 8.29 at t = 0"^.

Application o f superposition to Figure 8.30 shows that Kj = - 1 .5 V. Flence,
9 - ( - 1 .5 )
6x10-^
= 1.75 mA
Step 2. Find the circuit time constant and the equilibrium value o f ijj,t). From Figure 8.29, the
equivalent resistance seen by the capacitor is
= 4 kl^. Hence, the time constant is t = 2 sec.
Further, since t><d,X^ = ij{'^), which is computed when the capacitor is replaced by an open cir
cuit. In this case.
9 x 1 0 -'
= 1 mA
Step 3. Apply equation 8.23. Using equation 8.23, we have, for r > 0,
= 1 + 0.75^'
-0.5/
mA
Exercise. In Flxample 8.10, find /Ô ), /(Ô"^), and i^^t) for r > 0 using equation 8.23 directly.
AN SW ERS: 0, 2.25 mA. 2.25^-** "^ mA
Note that in Example 8 .1 0 we used
instead o f /,(0~) to obtain the correct answer. Some
straightforward arithmetic shows that /y(0) = - 2 mA. Since /y(0'^) = 1.75 niA, the input current
is discontinuous at f = 0, unlike the capacitor voltage, which is continuous at f = 0. This empha
sizes the need to use x{t^) in equation 8 .23.
3S5
In several o f the examples o f sections 3 and 4, the circuits contain switches that operate at pre
scribed time instants. In some electronic circuits, the switch is a semiconductor device whose
on/off state is determined by the value o f a controlling voltage somewhere else in the circuit. If the
controlling voltage is below a certain level, the electronic switch is off; if the voltage moves above
a fixed level, the electronic switch is on. The time it takes for a controlling voltage to rise (or fall)
from one level to another is very important because timing is as critical in electronic circuits as is
scheduling for large organizations. For first-order linear networks with constant excitations, cal
culation o f the time for a voltage or current to rise (or fall) from one level to another is straight
forward because all waveforms are exponential functions, as per equation 8.19. The situation is
illustrated in Figure 8.31.
c
0)
k_
I
D
u
<U
cn
(Z
4->
o
>
FIGURE 8.31 First-order response showing a rise from the voltage or current level A'j to the voltage
or current level A'2, for which the elapsed time is h - ty
In equation 8.19a, let .v(rj) =
and .v(^^) = X , be the two levels o f interest. A straightforward
manipulation o f equation 8.19a leads to the elapsed tim e formula for first-order circuits.
h -ri
[^1 --V(oo)
(8.24)
^2 - -v(x)
E X A M PLE 8.11
This example uses the elapsed time formula o f equation 8,24 for the circuit o f Figure 8.32. The
switch in this circuit is used to produce r^vo different final capacitor voltages. When the switch
is open, the final capacitor voltage is 12 V. When the switch closes, the final capacitor voltage, by
V-division, changes to 4 V. Thus the switch causes the capacitor to charge and discharge repeat
edly. For our purposes we show that the choice o f resistances produces an approximate triangular
waveform.
For the purposes o f this example, suppose the switch in Figure 8.32 is controlled electronically so
that it closes when
SNvitchings.
rises to 9 V and opens when V(^ falls to 5 V. Find and plot v^^t) for several
FIGURE 8.32 Switched driven RC circuit used to generate an approximate triangular waveform.
So
lu t io n
subsequently opening at ^ = tj^ and closing again at t^tc,
Suppose the switch first closes at f =

and so on. For 0 < r <
the time constant T = 3 sec, V(\Q) = 0, and
= 12 V. From equa
tion 8 .1 9c,
t
VcU) = 12 1 - e 3
From the elapsed time formula o f equation 8.24,
- 0 = 3//?
Now, for
19-12^
= 3 X 1.386 = 4 .1 5 9 s
t < tj^, there is a new' time constant x= 1 sec,
= 9, and
= 4. .^gain using
equation 8.19c,
From the elapsed time formula, equation 8.24,

//9-4\
- t =/n
Finally, for the time interval
- /n{5) = 1.61 s
\5-4/
= 5, and t'^co) = 12. Using equation 8 .1 9c,
, x= 3 sec,
JJz Itl
v c(/) = 1 2 - 7 e
and from the elapsed time formula,

^ ///5- 12\
- th =3/n
rh e w aveform o f
v^t)
for 0 < / <
9-12/
^ / ; (1\
\3
= 2.54 s
is plotted in Figure 8 .3 3 .
Chapter 8 1-irst Order RL and RC Circuits
Capacitor Voltage (V)
From the preceding solurion
= 4.16 sec,
= 4.1 6 + 1.61 = 5.77 sec, and t^ = 5.77 + 2.54 = 8.31
sec. If we proceed to calculate the waveform for t >
the waveform begins to repeat itself, as is
evident Irom Figure 8.33. Practically speaking, the first c)'cie o f a periodic, approximately trian
gular waveform occurs in the time intervâl [t^, r j , and the period is
= 8.31 - 4.1 6 = 4.15
sec. Note that the triangular waveform has a frequency

I
/ =
period
2.5 4 + 1.61
4. 15
= 0.241 Hz
The waveform in figure 8.33 is approximately triangular. This is due the fact that two time con
stants, 1 and 3 s, have the same order o f magnitude. If we select the resistances so that the charg
ing time constant is much larger than the discharging time constant, then the capacitance voltage
waveform will look more like a sawtooth waveform. Sawtooth waveforms arc used to drive the
horizontal sweep of the electronic beam in an oscilloscope or a T V picture rube.
8. FIRST-ORDER RC OP AM P CIRCU ITS

RC op amp circuits have some singular characteristics that set them apart from standard passive
RC and RL t)^pes o f circuits. Specifically, because ol the nature o f the operational amplifier, the
time constant o f the circuit will often depend only on some o f the resistances. We present four
important examples to illustrate the behavior o f RC op amp circuits.
3^H
EXA M PLE 8 .1 2
Compute the response
for the ideal op amp circuit o f Figure 8.34.
1.
FIGURE 8.34 Differentiating op amp circuit.
So
lu t io n
Observe that
Chapter 4. Also,
by the virtual short-circuit propert)' o f the ideal op amp, as set forth in

Hence, from these equalities and the definition o f a
capacitor,
= ,,( ,) = - / c (0 = - R C ^
=- R C ^
(8.25)
Since the output is a negative constant (user chosen) times the derivative o f the input, the circuit
is called a differentiator.
Exercise. Suppose y/(^) = cos(250^). Find R for the circuit o f Figure 8.34 so that
= sin(250/)
V a n d C = 1 pF.
AN SW FR: 4 kH
EXA M PLE 8 .1 3
Compute the response
for the ideal op amp circuit o f Figure 8.35 assuming V(^0~) =
= 0.
1.
FIG U R K 8 .3 5 Integrating op amp circuit.
So
lu t io n
= v-^{t)IR by the virtual short circuit property o f the ideal op amp. Also, i^^t)
Observe rhat
3S9
Hence, from these equalities and the integral i>-i relationship o f a capacitor,
(8.26)
Since the output is a negative constant (user chosen) times the integral o f the input, the circuit is
called an integrator.
Exercise. Suppose
= cos(250r) V. Find R for the circuit o f Figure 8.35 so that
sin(250r) V and C = 1 pF.
A N SW ER: 4 kD
EXA M PLE 8 .1 4
This example considers the so-called leaky integrator circuit o f Figure 8.36, which contains an
ideal op amp. The input for all time is v^{t) =
V. Rj represents the leakage resistance o f the
capacitor. Given C and Rj, the resistance /?| is chosen to achieve a dc gain o f 10. The objective is
to compute the response
assuming
= 0 and compare it to a pure integrator having a
gain o f 1. This problem is reconsidered in Chapter 13.

+
v jt )
1
FIG U RE 8.36 Leaky integrator op amp circuit in which v^{t) = -5u{t) V.
So
lu t io n
Because there is only one capacitor, the circuit o f Figure 8.36 is a first-order linear circuit. Because
the inverting terminal o f the op amp is at virtual ground,
and the capacitor sees
an equivalent resistance Rî^ = /?2- Hence x = /?,C = 10 sec > 0, Equation 8.19 implies that
360
(8.27)
Because the voltage source v^{t) = 0 for r < 0,
) = -V(i^ ) = 0. For f > 0, v^{t) = - 5 V. Since
the source voltage is constant, the capacitor looks like an open circuit at r = co having final value
.. ( co) = - A ( . 5 ) = 5() V
Entering numbers into equation 8.27 yields
= 50 + (0
= 5 0 (l
ii{t) V
A plot o f the op amp output voltage appears in Figure 8.37 along with that o f an ideal integrator.
O ne observes that the more realistic leaky integrator circuit approximates an ideal integrator only
for 0 < r < 0 .1 5 t before the error induced by the feedback resistor R-^ becomes noticeable. Such
integrators need to be reinitialized periodically by resetting the capacitor voltage to zero.
OJ
Ol
ra
>
4-'
D
CL
4-1
D
Q.
E
<
Q.
O
Time (s)
FIGURE 8.37 Outpiu voltage of leaky integrator that

approximates an ideal integrator.
So h r we have assumed an ideal op amp. In practice, the output voltage will saturate at a level
determined by the power supply voltage and the specs o f the particular amplifier used. Further,
practical op amps have complex models. To evaluate the preceding analysis, Figure 8.38 shows a
SPIC E simulation using the standard 741 op amp.
361
Chapter 8 * First Order RL and RC Circuits
Leaky lntegrator-Transient-0
Time (s)
Observe rhar the response approximaces the ideal up to about 0.1 5t = 1.5 sec, which corroborates
our analysis using the ideal op amp. Note, however, that the simulation accounts for saturation
present in practical op amps but absent from the ideal.
EXA M PLE 8.1 5

In trying to build an inverting amplifier in a laboratory, a student inadvertently reverses the con
nection o f the two input terminals, which results in the circuit of Figure 8.39a. Assume a practi
cal op amp model with
output o f
= 15 V and a finite gain of/I = 10^. Instead o f seeing the expected
= - 4 V, the student observes a 15 V output. Explain how this 15 V output could
possibly exist.
4kO
Capacitance
(a)
(b)
F IG U R E 8.39 (a) Incorrectly connectcd Inverting amplifier, (b) Circuit model, including a stray
capacitancc C = 1 pF and a finite gain A = lO"^.
362
So l u t io n
C'hapter 4 op amp models contain only resistors and controlled sources. One way to explain the
situation described in this example is to postulate a small stray capacitance, C = 1 pF, across the
input terminals. In fact, this is a more accurate circuit model and is shown in Figure 8.38b. This
means that the response will be o f the form o f equation 8.23.
The first quantit)' to compute is the circuit time constant t =
C. The equivalent resistance
looking to the right of C, is obtained from the circuit o f Figure 8.40.
FIG URE 8.40 Circuit for computing
note the artificial 1 V excitation.
From Figure 8.40 and our knowledge o f constructing Thevenin equivalems,
'
4000
Hence,
= 1 =
/,
1 -1 0 ^
Observe that in the actual circuit,
is in parallel with 1 k il. Hence, the Thevenin equivalent
resistance seen by C is
R^,, = 1000
= lOOOll ( - 2.5) = - 2 .5 0 6 Q.
The time constant o f the first-order circuit is

T = R ^^C =
- 2 .5 0 6
10"'^ sec, or -2 .5 0 6 picosecond (psec)
The negative time constant spells instabilit)'. The complete response may be written directly with
the use o f equation 8.23, where x{t) =
Suppose a very small noise voltage,
To use equation 8.23, we need aÔ"^) =
=E V, appears across C Then
= lO-le V
and
363
To compute the equilibrium output voltage
Kg =
we open-circuit the capacitor and compute
with C open-circuited
/
-2 5
\ j
= 1 X ------ 10' = -2 5 .0 6 3 V
V
1000-2.5/
From equation 8.23, the complete response is
= -2 5 .0 6 3 + (lO'Ê + 2 5 .0 6 3 )e
0.339xl0^r
(8.28)
For any small positive initial capacitor voltage E, equation 8.28 implies that the output would
increase exponentially, had the op amp been ideal. Because this particular real op amp saturates at
15 V, the output more or less instantaneously saturates at 15 V, the phenomenon observed by the
student. Had the initial capacitance voltage been sufficiently negative,
-2 5 .0 6 3
< ------- j ---- V,
10-^
equation 8.28 implies that v
would saturate at - 1 5 V.
9. SUMMARY
This chapter has explored the behavior o f first-order RL and RC circuits (1) without sources for
given ICs, (2) for constant excitations (dc), (3) for piccewise constant excitations, and (4) with
switching under constant excitations. In general, first-order RL and RC circuits have only one
capacitor or one inductor present, although there are special conditions when more than one
inductor or capacitor can be present. Our discussion has presumed only one capacitor or one
inductor is present in the circuit.
Using a first-order constant-coefficient linear differential equation model o f the circuit, the chap
ter sets forth rwo t}'pes o f exponential responses, the source-free response and the response when
constant independent sources are present. The source-free responses for the RL and RC circuits
have the exponential forms
ii(0= e
where
iiito)
vcit)=e
is the initial condition for the inductor and
itor. For an RC circuit, the time constant x=
the initial condition on the capac
where R^^^ is theThevenin equivalent resistance
seen by the capacitor. For an RL circuit, the time constant T= L/Rî^, where R^^^now is theThevenin
equivalent resistance seen by the inductor.
When independent sources are present in the circuit, the response o f a first-order RC or RL cir
cuit has the general form
I-10
x{t) = ,v(3c) + [,y(/o ) -
lie-
36-i
forT > 0. Stated in words, this Formula is

elapsed lime
x{t) =[Final value] +{[Initial value]- [Final value])e

provided the time constant
> 0. When the time constant
<^<>nsvdn{
< 0, then it is necessary to modify
the interpretation as discussed in section 7, with equation 8.23 identifying the form o f any volt
age or current in the circuit:
The time constants of- a circuit can be changed by switching within the circuit. By changing time
constants in a circuit, one can generate different t)'pes o f waveforms such as the triangular wave
form o f Figure 8.32. As mentioned at the beginning o f the chapter, wave shaping is an important
application o f circuit design. When inductors, resistors, and capacitors are present in the same cir
cuit, many other wave shapes can be generated. RLC circuits arc the topic o f the next chapter and
allow even greater freedom in waveform construction.
As a final application o f the concepts o f this chapter, we looked at the leaky integrator op amp cir
cuit. Integrators are present in a host o f signal processing and control applications. Unfortunately,
ideal integrators do not exist in practice. The leaky integrator circuit o f Figure 8.35 provides a rea
sonable model o f an ideal integrator.

Com plete response: sum o f zero-input and zero-state responses.
D ifferential equation o f a circuit: equation in which a weighted sum o f derivatives o f an impor
tant circuit variable (e.g., a voltage or current) is equated to a weighted sum o f derivatives
o f the source excitations to the circuit.
D ifferentiator circuit: op amp circuit whose output is a constant times the derivative o f the
input.
Equilibrium state o f a differential equation: c o n stan t, say x{t) = X^, th at satisfies the differential
eq u atio n in the variable A.*(r).
First-order differential equation o f a circuit: difterential equation whose highest derivative is

first order.
Forced response: that portion o f a complete response that has the same exponent as the input
excitation, provided the input excitation has exponents diflêrent from that o f the zeroinput response, under the condition that the input excitation is either (1) dc, (2) real
exponential, (3) sinusoidal, or (4) exponentially modulated sinusoidal.
Integrating factor method: mathematical technique for finding the solution o f a differential
equation in which multiplication by the integrating factor e~^ on both sides o f the dif
ferential equation leads to a new equation that can be explicitly integrated for a solution.
Integrator circuit: op amp circuit whose output is a constant times the integral o f the input.
Leaky integrator circuit: op amp circuit having a response approximating an ideal integrator, as
described in Example 8.13.
365
Natural iirequency of a circuit: natural mode of oscillation of the circuit. For a first-order circuit having a response proportional to
it is the coefficient X in the exponent.
Natural response: that portion of the complete response that has the same exponents as the zeroinput response.
Passive RLC circuit: circuit consisting of resistors, inductors, and capacitors that can only store
and/or dissipate energy.
Sawtooth waveform: triangular waveform resembling the teeth on a saw blade and typically used
to drive the horizontal sweep of the electronic beam in an oscilloscope or a T V picture
tube.
Source-free response: response of a circuit in which sources are either absent or set to zero.
Step response: response, for ^> 0, of a relaxed single-input circuit to a unit step, i.e., a constant
excitation of unit amplitude.
Stray capacitance: small capacitance always present between a conductor and ground. It usually
can be ignored, but as Example 8.14 shows, it can critically affect the response of a cir
cuit.
Superposition: in linear RC and RL circuits, the complete response is the superposition of the
relaxed circuit responses due to each source with all other sources set to zero, plus the
responses to each initial condition when all other initial conditions are set to zero and all
independent sources are set to zero.
Time constant: in a source-free first-order circuit, the time it takes for the circuit response to drop
to e~^ = 0.368 of its initial value. Roughly speaking, the response value must drop to a
litde over one-third of its initial value or rise to within one-third of its final value. For RL
circuits x = LIR^f^ and for RC circuits x =R^/jC.
Unit step (unction: function denoted (/) whose value is 1 for f > 0 and 0 for ^< 0.
Unstable response: response whose magnitude increases without bound as t increases. The time
constant for first-order circuits is negative for an unstable response.
Zero-input response: response in which all sources are set to zero.
Zero-state response: response to a specified input signal or set of input signals given that the initial conditions are all set to zero.
w
w
^ It happens that all variables in a first-order RL or R C circuit satisfy a differential equation o f the same form. The
interpretation o f the solution is somewhat different fi-om what follows. A detailed explanation o f the general solu
tion is presented in section 7.
366
Problems
i,(t)
UNDRIVEN RESPONSE WITH

GIVEN INITIAL C O N D ITIO N S
1. For the RC circuit o f Figure P8.1, R = 2.5
kiQ and C = 50 |.iF.
(a) If V(^Q) = 10 V, find
Plot your
answer for 0 < / < 5x, where x is the cir

cuit time constant.
(b) If V(\Q) = 10 V, find
with
out differentiating your answer to part

(a). Plot your answer for 0 < f < 5x,
where x is the circuit time constant. At
what time is the energy stored in the
capacitor about 1% o f its initial value?
(c) Compute ;^(r) for ^^0) = 5 V and y<;^0)
= 20 V without doing any further calcu
lations, i.e., by using the principle o f lin
earity.
Figure P8.2
3 . In Figure P8.1, suppose R = 25 kH and
vÔ) = 20 V.
(a) Find C so that vÔ.25) = 2.7 0 6 7 V.
(b) Given your answer to part (a), find
C H EC K ; C = 5
pF
4 . In Figure P8.2, suppose R = 2.5 kl^ and

/^(O) = 20 mA.
(a) Find L so that at /, = 1 msec,
2.7 0 6 7 mA.
(b) Given your answer to part (a), find
C H E C K ; Z.= 1.25 H
5. The response o f an undriven parallel RC cir
cuit is plotted in Figure P8.5. Find the time
ic(t)
constant of the circuit, at least approximately. If
C= 0.25 mF, find R.
. v,(t)
Figure P8.1
2. Consider the RL circuit o f Figure P8.2 in
which R = 50 Q and Z, = 0.1 mH.
(a) If the energy stored in the inductor at t
= 0 is 2 [ij, find /^(O) and
Plot your
answer for 0 < r < 5x, where x is the cir

cuit time constant.
(b) If the energy stored in the inductor at t
= 0 is 2 pj, find Vjit) without differenti
Figure P8.5
ating your answer to part (a). Plot your

answer for
< t < 5x, where x is the cir
cuit time constant.

(c) Repeat part (b) for /j^(0 ) = 50 mA and
/^(O) = 250 mA. Hint; What principle
makes this a straightforward calculation
given your answer to part (a)?
6. In the circuit o f Figure P8.6, suppose /?j = 50
Q., Rj = 200 Q., L = 2 H, /^(O) = 100 mA, and

the switch opens at t = 50 msec.
(a) Find ij {t) for r ^ 0. Plot your answer in
MATLAB for 0 s / ^ 0.1 sec.
(b) Find y^(0) and
t > 0. Plot your
answer in MATLAB for 0 < / < 0.1 sec.
36"
where T is the circuit time constant.

(b) Let a = - 1 1 . Compute the equivalent
resistance seen by the capacitor. Find
y^ f) for
> 0. Plot
for 0 < f < 2 t
where T is the circuit time constant.

(c) Find the range o f
(X
for which the time
constant is positive.
Figure P8.6
R.
7. In the circuit o f Figure P8.7, suppose /?j = 5
= 20 V,
kQ, /?2 = 20 k n , C = 50 pF H,
and the switch opens at r = 0.4 sec.
R,
(a) Find v^^t) for f > 0 . Plot your answer in
0.25 mF
MATLAB for 0 < f < 4 sec.

(b) Find
Figure P8.9
and /^(/), t > 0. Plot your
answer in MATLAB for 0 < r < 4 sec.

t=0.4s
C H E C K : (c) a > - 3
10. In the circuit o f Figure P8.10, /?, = 100 Q,
= 20
P = 2 0 0 ,1 = 0.5 H, and /'^(O) = 250
i{t)
mA. The switch opens at / = 0.03 sec. Find the
,Vc(t)
Thevenin equivalent resistance seen by the

inductor before the switch opens, and then
compute /^(r) and
for r > 0.
Figure P8.7
Consider the circuit o f Figure P8.8.
(a) Find the value o f
and the initial con
dition /^(O) so that /^(0.05 msec) =
R,
9.197 niA and /^(0.15 msec) = 1.2447

mA.
ijt)
(b) Given your answer to parr (a), compute

and plot /^(r) for 0 < r < 5x, where I is
the circuit time constant.
Figure P8.10
C H EC K :
= -2 5 n
ijt )
11. Consider the circuit o f Figure P 8 .ll in
1 kf)
which /?j = 25 ^2,

50 Q, and L = 2.5 H.
Suppose /^(O) = 2 A.
80 mH
(a) With
CX
= 0.1, compute the Thevenin
equivalent resistance seen by the induc
Figure P8.8
tor; then compute /^(f) for f > 0. Plot in
AN SW ER: (a) 600 Q, 25 mA
M A T I^ B for 0 < r < 5 t, where T is the

9. Consider the circuit of Figure P8.9, in which
= 100 Q and /?2 = 50 Q. Let
(a) Let a
= 500 mV.
= 7. Compute the equivalent
resistance seen by the capacitor. Find

i/^r) for f > 0. Plot
for 0 < ? < 5 t
time constant o f the circuit.

(b) W ith a = 0.1, compute
(c) Repeat part (a) for a = 0.02. Determine
the time, say r,, when the inductor has
lost 99% o f its initial stored energy.
Chapter 8 First Order RL and R C Circuits
368
t=4RC
i,(t)
R.
1=0
.4R
\4R
R,
Figure P 8 .1 1
Figure P8.14
12. In Figure P8.12, the current source
has
been applied for a long time before the switch

opens at r = 0. Find /^(O^) and /^(r) for f > 0 in
terms o f /^, R, and L, where is in A, R in
and L in H. Sketch /^(r) for 0 < ^ < 4x where x
is the circuit time constant for f > 0.
15. In Figure P8.15, the current excitation is
= Vîi{-t) V. Find
and
V(^t) for ^ > 0 in terms o f
R, and C, where
R is in Q and C in F. Sketch v^^t) for 0 < f < 3x,
given by
where x is the circuit time constant for f > 0.

t=RC
.4R
Figure P8.15
Figure P8.12
13. In Figure P8.13, the current excitation is
given by
A. Find /^(O^) and iît)
for r > 0 in terms o f /^, R, and L, where R is in

and L in H. Sketch /^(r) for 0 < r < 4x, where
X
4R
is the circuit time constant for r > 0.
16. Repeat Problem 14, except find

i(^t) for /> 0.
and
and
i(^t) for r > 0.
RESPONSE OF DRIVEN
CIRCUITS
18. Consider the RC circuit o f Figure P8.18 in
which R = 10 k n and C = 0.4 niF.
Figure P8.13
(a) If
14. In Figure PS. 14 the voltage source
has been applied for a long time before the

switch opens at r = 0. Find
and t^r) for
r > 0 in terms o f V^, R, and C, where
is in V,
R in Q, and C in F. Sketch V(^t) for 0 < r < 3x,

where x is the circuit time constant for / > 0.
= 0 and
V, find
U(^t). Plot your answer for 0 < r < 4x,

where x is the circuit time constant.
(b) II /^(^-(O) = 10 V and
= 0, find
Plot your answer for 0 < / < 4x, where x
is the circuit time constant.
Now making use o f linearit)' and its
associated properties, compute the indi
cated responses without any further cir
cuit analysis.
(c) If V(^Q) = 10 V and
find V(^t).
= 2i)ti(t) V,
(d) If c<0) = - 2 0 V and i/,.() = -I0 ( ) V,

find V(it).
369
20. In Figure P 8.20,

= 50 Q , 7?2 = 200 Q, C
= 2.5 mF, and the voltage excitation is given by
(e) If i/Ô) = 10 V and v-J^t) = 20(r) V.

find i({t) without differentiating your
answer to part (c). Plot your answer for
0 < r < 4 t , where x is the circuit time
constant.
where
= -1 0 V
and V^2 = 20 V.
(a) Find ^(;^0'*^) and V(it) for / > 0.
(b) Sketch v^^t) for 0 < f < 5x, where x is the
circuit time constant for r > 0.
(c) Identify the zero-input response (f > 0)
and the zero-state response (r > 0) for the
yoj
answer computed in part (a).
'.w O
(d) Now compute
for f > 0 assuming
the switch opens at r = 0.2 5 sec. Plot

your result ForO < f < 0.5 sec.
Figure P8.18
19. Consider the RL circuit o f Figure P 8.19.
Suppose /? = 100 Q, Z = 0.2 H.
(a) If /^(O) = 0 and
= 20(^) V, find
i.
'j (D
Plot your answer for 0 < / < 5t ,

Figure P8.20
(b) If ij{G) = - 5 0 mA and v-J^t) = 0, find

/^(r). Plot your answer for 0 < ^ < 5x,
21. In Figure P 8.21, R^ = 50 Q , i?2 = 200 Q,
Now making use o f linearity and its
= 2 H , and the voltage excitation is given by
Kl = -1 0 V
associated properties, compute the indi

cated responses without any further cir
cuit analysis.
(c) If/^(O) = - 5 0 mA and
find
= 20(r) V,
Plot your answer for 0 < r <
5x, where x is the circuit time constant.

(d) If/^(O) = 25 mA and y.(r) = -1 0 (f) V,
find
Plot your answer for 0 < ^ <
5x, where x is the circuit time constant.

(e) If /^(O) = - 5 0 mA and v-^{t) = 20(/) V,
find Vj{t) without differentiating your
and 1^2 = 20 V.
(a) Find /^(O^) and ijit) for t> 0.
(b) Sketch i^{t) for 0 < r < 4x, where x is the
circuit time constant for t> 0.
(c) Identify the zero-input response (t > 0)
and the zero-state response (^ > 0) for the
answer computed in part (a).
(d) Now compute Vj{t) for ? > 0 assuming
the switch opens at ^ = 0,0 4 sec. Plot
your result for 0 < f < 0 .2 sec.
answer to part (c). Plot your answer for

0 < ^ < 5x, where x is the circuit time
constant.
I'
'<b O
Figure P8.21
Figure PS. 19
22. Consider the RC circuit o f Figure P8.22a in

which
= Vq ((), where Kq = 100 V.
(a) Find
'Vw>
the time constant o f the cir-
370
cuit, Rp and /?2

circuit
response v^^t) is given by Figure P8.22b.
Assume C = 0.25 mE
and C so that the circuit
(b) Find
response v^t) is given by Figure P8.22b.

Assume i?2 = 10
n
+
;vc(t)
'.( s O
r\
(b)
Figure P8.23
(a)
24. In the circuit o f Figure P 8.24, V(^Qr) = 2$

V,
= 50u{t) mA, and
= 25(/) mA
(a) Find the zero-input response, i.e., the

response due only to the initial condi
tion.
(b) Find
for r > 0 due only to
(c) Find V(^t) for ? > 0 due only to

(d) Find the zero-state response.
(e) Find the complete response
for t >
r>
0.
Time in seconds
(f) Suppose the initial condition is doubled
(b)
and each independent source is cut in
Figure P8.22
half Find the new complete response

23. Consider the RL circuit o f Figure P 8.23a in
which
using linearity.
= Vqu(/), where Vq = 100 V.
(a) If Z = 2 H , find
the circuit time
constant, R^, and /?2 so that the circuit
1 kn
response i^(t) is given by Figure P8.23b.

(b) If /?2 = 2 k n , find ij^(0*), R-^, and L so
that the circuit response ij{t) is given by
Figure P8.23b.
Vu(t)
r>
i,(t)
d)
Figure P8.24
25. In Figure P 8.25 Ri = 2 0 0 Q, R2 = 6 0 0 Q,
(a)
T?3 = 650Q , Z = 20 H,
= -1 0 0 (-r ) +
50u{t) V, and
= 5 0 ( ? - 0.5) V. Compute
iît) for ^> 0. Plot your answer using MATLAB
or its equivalent for 0 < r < 8x.
371
directly without differentiating your

answer to part (a).
(c) W hat are the new responses if the value
o f each source is doubled?
Figure P8.25
26. Repeat Problem 25, except compute VjjJ)
for t>Q.
27. In Figure P 8.27
= 2 0 0 Q, /?2 = 6 0 0 Q,
= 8 5 0 a . C = 2.5 mF, v^^{t) = - 5 0 M +
100W V, and
Figure P8.30
= - 5 0 ( f - 5 ) V. Compute
31. The switch in the circuit o f Figure P8.31
V(^t) for t > 0. Plot your answer using
has been open for a long time before it is closed
MATLAB or its equivalent for 0 < ? < 6x.
at t = 0 . Suppose
= 6R, R2 = 30i?, R^ = 20R,
and
= Vqu{T - /). In terms of Vq, R, C,
and T = 6RC,
(a) Find V(^0~) and t/(;;(0'^).
(b) Find the Thevenin equivalent seen by
the capacitance for 0 < ^ < T.
(c) Using the Thevenin equivalent found in
pan (b), find an expression for Vf^t)
Figure P8.27
valid over 0 < t < 7.

28. Repeat Problem 27, except compute the
(d) Find the expressions for V(iT~) and
capacitor current i^t) for t > 0 .

-2Qu{-t) + 20(^) V. Find V(iO~) and v^^t) for

r > 0. Plot V({t) for 0 < ? < 40 msec.
(e) Find the time constant valid for t> T.

(f) Find V(^t) for t> T.
(g) Plot
for 0 < r < 4 7 using MATLAB.
20 msec
2kO
'. w ( D
8kO
=0
1.6 kn
+
Vc(t).
S jjF
20 V
+
.Vc(t)
vjt)
Figure P8.29
Figure P8.31
3 2 . For the circuit of Figure P 8.32, -V^ = - 1 0
which /?j = 300j^, /?2 = 800 Cl, R^ - 600 Q, L

= 4 H,
- -24u{-t) + 24(/) V, and
=
V, Kj = 20 V, R^ = 6 0 0 Q, /?2 = 2 0 0 Cl, and C
24u{t) V.
= 12.5 hF(a) Find yÔ"^).
(a) Compute the response ii{t) for ? > 0.

Plot for 0 < ^ < 4x where x is the circuit
time constant.
(b) Find the inductor voltage v^{t) for r > 0
(b) Using the initial condition computed in

part (a), find
< t< 160 msec.
Plot the result for 0
372
34. The voltage waveform
of Figure
P8.34a drives the circuit o f Figure P8.34b. The

voltage-controlled switch Si closes when the
capacitor voltage goes positive and opens when
the capacitor voltage v^^t) goes negative.
Compute the voltage Vf^t) across the capacitor.
Assume that
has been at - 1 0 V for ?< 0
for a very long time. Hint: Use the elapsed time

formula as needed.
>
20V
v,(t)
5
t(Msec)
2.5
(b)
--10V
Figure P8.32 (a) Pulse driving

RC circuit of part (b).
----------(a)
IMegO
33. For the circuit of Figure P 8 .3 3 ,, -V^ = - 1 0 V,

Kj = 20 V, /?! = 80 Q.7?2 = 20 Q, andZ = 4 H.
(a) Find
(b) Using the initial condition computed in
part (a), find
Plot the result for 0
< r < 160 msec.
(b)
Figure P8.34 (a) Pulse waveform exciting

RC circuit in part (b).
35. Repeat Problem 34, except find i(^t) for t>

80
t(msec)
40
36. Consider the circuit o f Figure P8.36. Suppose
--V.0
(a)
V ({0 ) = 0
and find V(î) for
as follows:
(a) Find the Thevenin equivalent circuit

seen by the capacitor.
(b) Find the complete response V(^t) for t >
0. W hat is y(j(oo)?
- v,(t) +
400
(b)
5kO
C = 0.1 F
IV
V,
Figure P 8 .3 3
Figure P 8 .3 6
lOlv.
373
OP AMP CIRCUITS
41. Figure P 8 .4 la shows an op amp integrator

with positive gain, and Figure P 8 .4 lb shows a
= K
sin(cor)(r) V and all capacitor voltages are zero

at f = 0. Find
and
of/e, C,/C and (0.
differentiator with positive gain for the con

stant K>
(a) For each o f the circuits find a literal
for / > 0 in terms
expression for
in terms of
(b) For /?j = 10 k n and (7 = 0.1 mF, find
Vo,{t) when
= 100sin(20r)//(r) mV
assuming that
= 0 in each case.
v.(t)
R.
O v/W
+
Vjt)
A .
Figure P8.37
KR
38. In the circuit o f Figure PS.38, /? = 10 klT2,

C = 10 jiF,
= 10 sin(50f)(r) mV, and all
capacitor voltages are zero at / = 0. Find

and plot it for 0 < r < 6 sec.
Figure P8.41
M ISCELLAN EOUS
Figure P8.38
42. Although most o f the first-order circuits

= \QOe~~ii{t)
mV, But plot from 0 to 6 seconds.
considered in the text have only one capacitor

or one inductor, it is possible to have a firstorder circuit containing more than one energ}'
40. In the circuit o f Figure P8.40, /?j = 10 kH,
storage element. Consider the situation depict
Rf = 40 kQ, C = 12.5 |.iF, and v^{t) = -lOOw(f)
ed in Figures P8.42a and b. Here
mV.
Vcii^*) are given. The networks A^, and N j are
(a) If V({0~) = 0, compute
for r > 0.
and
equivalent under two conditions:
(b) Repeat part (a) for V(jS^~) = 50 mV.

- v,(t) +
r
C, + Co
Prove this equivalence using the integral rela
tionship o f a capacitor to show that the i-v ter
minal conditions are the same for both
N,.
Figure P 8 .4 0
and
374
C .d - v ,
N1
(a)
(b)
Figure P8.42
43. As mentioned in Problem 42, although
most of the first-order circuits considered in the
text have only one inductor or one capacitor, it
is possible to have a first-order circuit contain
ing more than one energy storage element.
Consider the situation depicted in Figures
P8.43a and b. Here
and
given. The networks
and A/'j are equivalent
under two conditions:
êq -
Prove this equivalence using the int^ral relation

ship of an inductor to show that the i-v terminal
conditions are the same for both A^j and A^2 -
(a)
45. In the circuit of Figure P8.45, Cj = C2 = 2

F and the switch closes at time ^= 0. The initial
conditions on the two capacitors are t/^(0 ) =
4 V and vc2(0-) = 0 V.
(a) For = 0.5 ft, find an expression for the
current /^(f) for ^> 0.
(b) For /? = 0.5 ft, find
for ? > 0 and
for f > 0. Note that
^ 0 for
all r > 0.
(c) Compute the energy stored in each
capacitor at r = 0^. Also compute the
energy stored in each capacitor at r = c.
Finally, compute the decrease in total
energy stored in the capacitors fi-om t =
O'^ to ^= 00.
(d) Compute the energy dissipated in the
0.5 ft resistor fi-om r = 0"^ to r = 00. Verify
that the energy dissipated in the resistor
equals the decrease in total energy stored
in the capacitors fi-om r = 0* to ^= 00.
(e) Does the dissipated energy depend upon
the value of P?. What does R affect?
Verify that conservation of energy holds
for the circuit.
(b)
Figure P8.43
Figure P8.45
44. In the circuit of Figure P8.44, suppose /?j =

50 Q, /?2 = 200 a , q = 0.06 F, Cj = 0.3 F,
yQ(0) = 15 V, and i'qCO") = 5 V. Let
=
40tt(r) V. Use the equivalence set forth in
Problem 42 to compute
for r > 0.
C. -
C, - t
V,
46. Repeat Problem 45 for the circuit of Figure

P8.45 when q = 1 F, Cj = 0.25 F, i/îCO-) = 3
V, and t/f^(0~) = 8 V.
47. In the circuit of Figure P8.47,
=
1 lOtt(-f) + 220u(t) mV, Zj = 110 mH, ^2 = 11
mH, and /? = 10 ft. Compute and plot the
waveforms for inU) and /^(/). Hint: Adapt the
results of Problem 43 to ê case of two paral
lel inductors with initial currents.
Figure P 8 .4 4
r>
375
and K 2 . Find
for appropriate
'JO
and
K2 in terms of K^.
iA n i'
(b) Suppose the input v^{t) = -\2u{-t) +
24(/). Find y(0 ) by inspection of the
circuit, and v(0+) by the principle of

conservation of charge.
Figure P8.47
(c) Use equation 8.1 7 to write down direct

48. Repeat Problem 4 7 for the case where v-J^t)
ly the answer for v{t), f > 0. Had y(0 )
= 220W mV,
been used, would the answer still be cor
= 4 4 mA, and
rect?
11 mA.
51. The solution to the basic RL or RC differ
49. Consider the circuit of Figure P 8.49 in

which Cj = 1 F, Cj = 4 F, v-^ = 10 V, and R = 2
ential equation in this chapter, equation 8.3,
a.
builds on the integral solution o f equation 8.4,

(a) Compute v^Qr) and Vj^{Q*). Hint: How
which is valid for arbitrary y(/). This powerful
does the charge distribute over the two
formula will be studied in a course on differen
capacitors at ^ = 0^?
tial equation theory. Wheny(r) = F, a constant,
(b) Compute Vjfi) for r > 0.
it is possible to develop an alternative deriva

tion o f equation 8 .1 7 using no more than some
basic knowledge of calculus. Since the solution
to the source-free case is the exponential, it is
reasonable to expect (or to try) a solution for
the constant input case o f the form
'> ^ '
( 1)
Figure P8.49
and K2 are two constants to be deter
where
mined. The constant K 2 arises from the con

50. The circuit o f Figure P 8.50 contains two
stant input, suggesting that the response would
capacitors.
intuitively contain a constant term.

(a) Substitute equation 1 into equation
8 .16. You should obtain the result
Pi.
(b) With K2 determined, evaluate x(/) at f =
tQ* to obtain an expression for
. Your
result should be
Figure P8.50
(a) Suppose vjî) = Kû{t) V. Show that for t
W
> 0, the voltage v{i) satisfies the firstorder differential equation
dv
It
= -K
iv + K 2
x Uq
) - F t
(2)
(c) Finally, substitute K2 = Fx and equation

2 into equation 1.
3 "6
(ii) S is at position B when v is greater than
APPLICATIONS
52. An approximate sawtooth waveform can be
60 V and decreasing, and it moves to
produced by charging and discharging a capac
position A when i drops to 1 mA and v
itor with widely different time constants. The
drops to 60 V.
circuit o f Figure P8.52 ilkistrates the idea. Vj{t)

= 20 kQ,
= 10 V,
Assume that at r = 0, switch S is at A,
= 1 kQ, and C= 10 pF.
The switch S is operated as follows: S has been
and z^,^;(0) = 60 V. Find
at position B for a long time, and S is moved to
cycle o f operation (i.e., charging and dis
for one
position A at / = 0 to charge the capacitor.
charging the capacitor), and roughly
When V(~increases to 9 V, switch S is moved to
sketch the waveform. Whau is the fre
position B to discharge the capacitor. When
quency o f the sawtooth waveform?
decreases to 1 V, switch S is moved to position

A to charge the capacitor again. The process
repeats indefinitely.
(a) Compute the waveform o f
for four
switchings.
(b) Plot
Is the name sawtooth wave
form appropriate? What is the frequen

cy in hertz o f the sawtooth waveform?
i (mA)
idealized i-v curve

ofa neon lamp
(not to scale)
/:
slope
= 1m U
/
Figure P8.52
slope
53. The sawtooth waveform is used in T V sets
negative resistance
region
= 0.5 mU
60
(b)
N fc i
v(V)
-----------
90
(0
Figure P8.53
and oscilloscopes to control the horizontal

motion of the electron beam that sweeps across
m is based on a hypothetical
the screen. One method o f generating such a
energy storage system using an inductor and a
waveform is to repeatedly charge a capacitor
solar cell. Consider the circuit o f Figure P S.54.
with a large time constant and then repeatedly
During the day, the solar cell stores energy by
discharge it with a very small time constant.
increasing the current in the inductor. During
The circuit in the shaded box o f Figure P8.53a
the night, the stored energy is used to power
is a crude functional model for the neon bulb
lights and appliances. Energy from the solar cell
in Figure P8.53b (type 5AB, costing about 75
is scored in the inductor during 0 < r < 7j. At /
cents), whose i-v characteristic is shown in
= 0, the beginning o f the storage interval, /^(O)
Figure PS.53c. The switch S in the model oper
= 0. At r =
ates as follows;
storing energy in the solar cell via the source
(i) S is at position A when v is less than 90
the device
is switched from
V and increasing, and it moves to B when
Vsolar <^0 powering a light denoted by Ry Note

that diere is some overlap in the switching
reaches 90 V (the breakdown volt
movement; this is to ensure continuity o f the
age).
inductor current. At r = T^, the TV, represent

ed by Rj, is also turned on.
377
Remark: All answers to parts (a) to (f) should be

in terms o f
/?,. and R^.
photo tim er used for timing the light in pho
(a) Draw a simplified equivalent circuit
tographic enlarger and printing boxes. Briefly,
5 5 . The circuit o f Figure P8.55 is a transistor
with three circuit elements for 0 < ? < T j,
the circuit operates as follows. When the relay
indicating all device values.
contact closes, the lamp is lit. When the contact
(b) Construct an expression for
0 < ^<
opens, the lamp is turned off. The relay has a

4 0 0 0 Q dc resistance and a negligible induc
(c) Draw a simplified equivalent circuit
tance. The pickup current is 2 mA, and the
with two circuit elements for T-^<t<
dropout current is 0.5 mA; i.e., the contact
closes when the relay current increases from
(d) Construct an expression for /^(/), T^^<t
< Tj. You will need a value or expression

for
zero to 2 mA, and it opens when the current

drops below 0.5 mA.
After obtaining your expres
sion, for simplicity, let
denote
To use the timer, switch S2 is closed first.

Switch Sj is normally in the B position. When
(e) Draw a simplified equivalent circuit
it is thrown momentarily to position A, the bat
with two circuit elements for T2 < t,
tery charges the 1000 pF electrolytic capacitor

C to 1.5 V. When Sj is then thrown back to

(f) Construct an expression for
Tj < t.
position B at /^= 0, the capacitor discharges and
You will need an expression for
produces a current iy, which, after amplifica
After obtaining your expression, for sim
tion by the transistor, actuates the relay and
plicity, let ^Tl- denote /^(jT2 )

Remark: For the remaining parts, all
amplified current drops below a point for the
answers are to be given in terms of
relay to open and the lamp is turned off.
K kf
tori
^1>
^ 2> d
This is to prevent the substitution of
Compute
possibly incorrect answers from prior
the middle of its full range (i.e., only 5 kQ is
parts for
used in the circuit).
(g) For each o f the four devices
^ston>
^store
down an expres
sion for the power absorbed at time t.
Call the results
and PRstore
(h) For time t give an expression for the

energy Wj(J) stored in the inductor if
W i( r = 0 ) = 0.
^soljr
-N
-------------- ^
Figure P 8 .5 4
turns on the lamp. At some later instant
the
if the 10 kH potentiometer is set at
378
1.5V
lO kn
E
Potentionmeter
crude transistor
circuit model
Figure P8.55
56. The circuit o f Figure P 8.56 suggests a way of generating a sustained sinusoidal oscillation. All
op amps are assumed to be ideal. Capacitors, C = 0.1 fiF are uncharged at f = 0. The first two op
amps are differentiators and the last is an inverting amplifier.
(a) With switch S at position A and v^{t) = sin(lOOOr) V, find vj^t),
and
for r > 0.
(b) If at a later instant switch S is quickly moved to position B, what would you expect
to be?
Figure P8.56
Second Order Linear Circuits
Warming up snacks in a microwave oven is

a common activity in student dormitories.
It works much faster than a conventional
s B s m
oven: heating a sandwich takes about 30

seconds. How does the microwave oven do
this? While a precise explanation is beyond
the scope o f this text, the basic principle
can be understood through the properties
o f a simple LC circuit.
Recall that two conducting plates separated
by a dielectric (insulating material) form a
capacitor. Suppose some food were placed
between the plates in place o f an ordinary
dielectric. The food itself would act as a dielectric. Ordinar)' food contains a great number o f
water molecules. Each water molecule has a positively charged end and a negatively charge end,
with their orientations totally random for uncharged plates, as illustrated in part (a) o f the figure
below. Applying a sufficiently high dc voltage to the plates sets up an electric field produced by
the charge deposited on the plates. This causes the water molecules to align themselves with the
field as illustrated in parr (b). If the polarity o f the dc voltage is reversed, the molecules will realign
in the opposite direction as illustrated in part (c). If the polarity o f the applied voltage is reversed
repeatedly, then the water molecules will repeatedly flip their orientations. In doing so, the water
molecules encounter considerable friction, resulting in a buildup o f heat, which cooks the food.
Microwave cooking is therefore very different from conventional cooking. Instead of heat coming
from the outside, the heat is generated inside the food itself.
380
Chapter 9 Scconcl Order Linear Circuits
food
food
conducting plate
conducting plate
(b)
(a)
food
0
Reversal of the polarity o f the applied voltage at a low frequency can be easily achieved with the
circuit elements studied in earlier chapters: the resistor, the capacitor, and the inductor. However,
the friction-induced heat production is inefficient at low frequency, l b produce a useful amount
o f heat for cooking purposes, ver)' high frequencies must be used. The t}'pical frequency used in
a microwave oven is 2.45 gigahertz, i.e., the water molecules reverse their orientations 2 x 2.45 x
10^ times per second. At such a high frequency, capacitors and inductors are quite different in
their behavior from their conventional forms. For example, the LC circuit becomes a resonant
cavit} and the connecting wire becomes a waveguide. These microwave components will be
studied in a future field theory course. T he theory studied in this chapter will enable us to under
stand the low-frequenc}' version o f the phenomenon, i.e., how a connected inductor and capaci
tor can produce oscillator)' voltage and current waveforms.
CHAPTER OBJECTIVES
1.
Investigate the voltage-current interactions that occur when an ideal inductor is con
nected to an ideal capacitor with initial stored charge.
2.
Use a second-order differential equation for modeling the series RLC and parallel RLC
circuits.
3.
Learn to solve a second-order differential equation circuit model by first finding the nat
ural frequencies o f the circuit, then looking up the general solution form, and finally
determining the associated arbitrary constants.
Chapter 9 Sccond Order Linear Circuits
3<S I
4.
Define and understand the concepts of underdamped, overdamped, and critically
5.
damped responses.
Investigate and understand the underlying principles o f various oscillator circuits.
SECTION HEADINGS
1.
Introduction
2.
Discharging a Capacitor through an Inductor
3.
Source-Free Second-Order Linear Networks
4.
Second-Order Linear Networks with Constant Inputs
5.
Oscillator Application
6.
Summary
7.
Terms and Concepts
8.
Problems
1. IN TRO DUCTION
The previous chapter developed techniques for computing the responses of first-order linear net
works, either without sources or with dc (constant) sources, having first-order linear differential
equation models. Recall that the source-free response contains only real exponential terms.
This chapter focuses on second-order linear networks having second-order linear differential equa
tion models. Usually, but not always, a second-order net\vork contains t^vo energ)^ storage ele
ments, either {L, Q , (C Q , or {L, L). Second-order circuits have a wide variety o f response wave
forms: exponentials
sinusoids (/l,cos(coy) + y4-,sin(coy)), exponentially damped
sinusoids, and exponentially growing sinusoids, among others. Tables 9.1 and 9.2 catalogue the
various response types. W ith no sources or with constant-value sources, some straightforward
extensions o f the solution methods o f Chapter 8 are sufficient to compute the various responses.
The behavior of second-order circuits is a microcosm o f the behavior o f higher-order circuits and
systems. Many higher-order systems can be broken down into cascades o f second-order systems or
sums o f second-order systems. This suggests that our exploration o f second-order circuits can
build a core knowledge base for understanding the behavior o f higher-order, more complex phys
ical systems.
Many introductory texts discuss only parallel and series RLC circuits, stating separate formulas for
the responses o f each. Our approach seeks a unified treatment. To this end, we formulate a basic
second-order differential equation circuit model. The associated solution techniques become
applicable to any second-order linear nersvork and, for that matter, to second-order mechanical
systems.
An oscillator circuit (section 5) motivates our study o f second-order linear networks. The chapter
contains several other practical examples illustrative o f the wide variety o f second-order circuit
382
Chapter 9 Second Order Linear Circuits
applications. Some advanced applications pertinent to higher-level courses include low-pass, highpass, and bandpass filtering (covered later in the text); dc motor analysis; position control; and
many others. Most important, the concepts presented in this chapter are common to a host o f
engineering problems and disciplines. Hence, the techniques and concepts described here will
prove useful time and time again.
2. DISCHARGIN G A CAPACITOR TH RO UGH AN IN DUCTO R

Chapter 8 showed that the voltage o f an initially charged capacitor in parallel with a resistor
decreases exponentially to zero: the capacitor discharges its stored energy through the resistor.
When an inductor replaces the resistor, as in Figure 9.1a, very different voltage waveforms emerge
for V(-{t) and
In order to construct these new waveforms, we first develop a differential equa
tion model o f the LC circuit.
EX A M PLE 9 .1 . The goal o f this example is to develop a differential equation model o f the cir
cuit in Figure 9.1b. In Figure 9.1a, with the switch S in position A, the voltage source charges the
capacitor to
volts. At / = 0, the switch moves to position B, resulting in the new circuit o f
Figure 9.1 b, valid for f > 0.

A
(a)
(b)
FICiURF 9.1 (a) A voltage source charges a capacitor, (b) An LC second-order linear network in
which the cnerg) stored in the capacitor in part (a) is passed back and forth to the inductor.
S
o l u t io n
Step 1. Write down the terminal i-v relationship for the capacitor and itiductor; then apply KCL and
KVL, respectively. Using the i-v relationships for L and C (see Chapter 7) in conjunction with KCL
and KVL, it follow's that
ih C
_ >c _
dt
and
L_
V
J V
C definition
KCL
di
(9.1a)
(9 .1 b )
L definition
KVL
383
Step 2. Obtain a dijferential equation in the capacitor voltage, V(j^t). For this, first differentiate
equation 9.1a to obtain
1 dii^
dt
C dt
Substituting equation 9.1b into this equation yields

^
= - VC
dt
LC ^
(9.2)
Equation 9.2 is a second-order linear differential equation circuit model o f Figure 9.1b in terms o f
the unknown capacitor voltage, v^t). Equation 9.2 stipulates that the second-order derivative
o f the unknown function, vît), must equal the function itself multiplied by a negative constant,-----
LC
Step 3. Obtain a differential equation in the current, ij{t). An alternative circuit model in i^ is obtained
by first differentiating equation 9.1b and substituting equation 9.1a into the result to produce
Equation 9.3 has precisely the same form as 9.2: the second-order derivative o f the unknown
function, ij{t) , equals the function itself multiplied by a negative constant, -----. This similarity
suggests a similarity o f solutions, which we shall pursue further.
Exercise. Fill in the details o f the derivation o f equation 9.3 from 9.1.
Our next goal is to construct the waveforms V(^t) and ij{t), which are the solutions o f the differ
ential equations 9.2 and 9.3. Although differential equations are not usually part o f the common
background o f students in a beginning course on circuits, the solutions o f equations 9.2 and 9.3
do not demand this background. Some elementary knowledge o f differential calculus is sufficient.
Specifically, recall the differential properties o f the sine and cosine functions:
dt
dt
sin(co/+ 0 ) = (o co s(o )/-t-0 ) and c o s ( o ) / + 0 ) = - a ) s i n ( c o / + 0 )
Differentiating a second time yields
d2
^y2
7 sin((or-f-0) = - 0) sin(co/+ 0 ) and 7 cos(o)/+ 0 ) = - c a cos(co/+ 0 )
dt-
dt-
In both cases, the second derivative equals the function itself multiplied by a negative constant.
This is precisely the propert}' required by equations 9.2 and 9.3. Thus one reasonably assumes that
the solutions o f equations 9.2 and 9.3 have the general forms
V(^t)
and
K cos(ojr
+ 0)
(9 .4 a )
38^
Chapter 9 Sccoiul Order Linear Circuits
/^(r) = /^cos(tor+0)
(9.4b)
7'hese forms are general because the cosine function can be replaced by the sine function with a
proper change in the phase angle. Specifically, we note that A'sin((i)/ + (j)) = A'cos(co/ + (j) - 0.5ti)
= K cos((or + 0) with 0 = ([) - ().5ti. Computing values for (o, A', and 0 specifies the solutions to
equations 9.2 and 9.3.
E X A M PLE 9.2. Find A"and B for the capacitor voltage in equation 9.4a.
So
lu t io n
Step 1. Differentiate equation 9.4a to obtain
dvrit)
(It
= -A c o sin (O )/+ 0 )
/'o S'!
Step 2. Dijferentiate n second time. Differentiating equation 9.5 (the second derivative o f 9.4)
yields
d~Vr
">
-----^ = -AÔ) cos(O)/+0) = -(O i <7
dt~
___
(9.6)
Step 3. Match the coefficients o f equation 9.6 with those o f 9.2 to specify O). Under this matching,
' o1r CO =
(0 ^ = -----
LC
7)
j= =
fo
fiC
^ ^
Equation 9.7 specifies co, the angular frequency o f oscillation, in rad/sec, o f the capacitor voltage.
Step 4. Compute K an d f) in equation 9.4a. These two constants depend on the initial conditions
as follows: when the switch is at position A, the capacitor is charged up to I^q volts and the induc
tor current is zero; immediately after the switch moves to position B, i.e., at r = 0+, the continu-
it>' properry o f the capacitor voltage ensures that y^Ô*) = Vj, and^the continuity property o f the
inductor current ensures that //(O^) = 0. The initial value,
^ , is now calculated from
equation 9.1a as
^/r(--(0 ) i( ^ ( 0 )
------------ = ----------- = ------------- = ()
dt
Evaluating equations 9.4a and 9.5 at / = 0^ , we have
= K c o m = v;,
(9.8a)
and
</v'r(0" )
^
= -A o )s m (e ) = 0
(9 g y
From equation 9.8b, 0 = 0. Consequently from 9.8a, K = V^^. Hence the capacitor voltage, i.e.,
the solution o f the second-order differential equation 9.2, is
Vc(f) = VqCOS
[J
lc
(9.9)
38S
As per equation 9.1a, one can obtain ij{t) directly by differentiating equation 9.9 and multiply
ing by C. However, one could aLso solve equation 9.3 by repeating the above steps to arrive at
the same answer.
Exercise. Assuming that /^(r) = K coslcor + 6), solve for (o, K, and 0 in terms o f the initial condi
tions, and show that
Several very interesting and significant facts about this parallel ZCcircuit and the solution method
are apparent:
(1)
For the source-free LC circuit o f Figure 9.1, the voltage and current responses are sinusoidal
waveforms with an angular frequency equal to _L _ . Since the amplitude o f sinusoidal
oscillations remains constant (i.e., does not
VZc
damp out), the circuit is said to be undamped.

(2)
rhe frequenq^ (o, depends on the values o f L and C only, while the amplitude K and the
phase angle 0 depend on L, C, and the initial values o f the capacitor voltage and inductor
current.
(3)
Although the instantaneous energ}" stored in the capacitor, \V^{t), and the instantaneous
energy stored in the inductor,
both vary with time, their sum is constant. (This is
investigated in a homework exercise.) Physically there is a continuous exchange o f the energ)' stored in the magnetic field o f the inductor and that stored in the electric field of the
capacitor, with no net energy loss. This is analogous to a frictionless hanging mass-spring
system: because o f the absence o f friction, the up-and-down motion of the mass never
stops; in such a mechanical system there is a continuous interchange between potential and
kinetic energy.
Figure 9.1 shows what is, in theory, the simplest circuit that generates sinusoidal waveforms. Such
an electronic circuit is an (idealized) oscillator circuit. Oscillator circuits play an important role
in many communication and instrumentation systems.
3. SOURCE-FREE SECON D-ORDER LINEAR N ETW ORKS

Unlike their ideal counterparts, real capacitors and inductors have resistances. A better under
standing o f a realistic oscillator entails the analysis o f an RLC circuit. This section investigates
source-free RLC circuits having two energ)' storage elements. Our investigation begins with the
development o f the differential equation models o f the series and parallel /^//.'circuits. Both mod
els are special cases o f an undriven general second-order linear diflerential equation. Hence we will
discuss the solution o f a general second-order linear differential equation and adapt the solution
to the series and parallel RLC circuits. We will also illustrate the theory with a second-order cir
cuit that is not a parallel or series RLC.
386
Developmeut o f Dijferential Equation Models for Series/Parallel RLC Ciraiit

The first goal o f this section is to develop differential equation models for series and parallel RLC
circuits as detailed in the following example.
EXA M PLE 9 .3 . For the series and parallel RLC circuits shown in Figure 9.2, develop two sec
ond-order differential equation models (one in
and one in
for each circuit.
+
V
''c = V''l
= V
''r
(b)
FIGURE 9.2 (a) Series RLC circuit, (b) Parallel RLC circuit. Passive sign convention is assumed as usual.
S
o l u t io n
Wc do this in parallel rather than in series.

Step 1. Apply KVL to the series RLC.
Step 1. Apply KCL to the parallel RLC.
Step 2. Choose i^âs circuit variable

and express
v^, and Vq in terms
Step 2. Choose
and express i^,
as circuit variable
and iQ in terms o f
^ fL(ii
1
Ril + L + / ^ ( t V t = 0
dt
C
oo
'
\f^ I f
,
dv/"
+
vrix)dx + C ^ = 0
R
L ^ ^
dt
Differentiate, rearrange terms,
Differentiate, rearrange terms, and
and divide by L to obtain
divide by C to obtain
dîl
Rdi^
1 .
^
^ + ------ ^ + ----- 1, = 0
dr
L dt
LC ^
d^Vf
dt-
d\>r
RC dt
1
H-------= 0
LC
Step 3. Choose Vq as the circuit
Step 3. Choose ij as the circuit
variable and express Vjând

terms o f Vq
variable and express

terms o f i^.
in
Again using the KVL o f step 1,
d ir
Rin -f" L ------- 1= 0
^
dt
^
arid i^ in
Again using the KCL o f step 1,

-H/, + C ^ = 0
dt
38
= C dv^^dt,
Hence, substituting for
L dij/dt,
Hence, substituting for
rearranging, and dividing through by
rearranging, and dividing through by
LC yields
LC yields
R dvr
L dt
d t
llL
dr
LC
RC dt
LC
Each circuit has two second-order differential equation models, one each for
= 0
and
as the
unknown quantity.
Exercise. Show that
and Vj^ satisfy second-order differential equations similar to those
derived in Example 9.3.
Solution o f the General Second-Order Dijferential Equation Model

The final differential equations o f Example 9.3 force the current
or the voltage
to satisfy cer
tain differential constraints. All four (differential) equations have the general form
d^x
dx
~ + h ----- h (.'.V = 0
dtdt
(9 . 10)
for appropriate constants b and c, where .v is either ij or
Equation 9.10 stipulates that the sec
ond derivative o f the function x{t) plus h times the first derivative o f x{t) plus c times x(t) itself
adds to zero at all times, t. Unlike the example o f section 2, where a sinusoidal solution was easi
ly predicted, the present differential equation requires a more careful mathematical analysis. Recall
from elementary calculus that the derivative o f an exponential is an exponential. Thus the first
and second derivatives o f an exponential are proportional to the original exponential. This sug
gests postulating a solution o f the form at(^) = Ke^ where we make no a priori assumptions about
s. If it is truly a solution, it must satisfy equation 9.10. Under what conditions will x{t) = Ke^ sat
isfy equation 9.10?
E X A M PLE 9.4. Determine conditions under which the postulated solution x{t) = Ke^ satisfies
equation 9.10.
S
o l u t io n
Step 1. Substituting Ke^ for ,v(/) in equation 9.10 produces
rfV '
dr
+ bK
<h
+ cK e = Ke" (,v^ + hs + c) = 0
^
(9.11)
Step 2. Interpret equation 9.11. For nontrivial solutions, K\s nonzero. The function e^ is always
different from zero. Hence the quadratic in s on the right side o f equation 9.11 must be zero. This
necessarily constrains $ to be a root of
r + bs + c =0
(9 .1 2 )
3<S8
Chapter 9 * Sccond Order Linear Circuits
Step 3. Solve equation 9.12. From rhe quadratic formula, rhe roots o f equation 9.12 are
.V|, .v. =
-b
- 4c
(9.13)
C O N C LU SIO N : x(/) = Ke^' satisfies equation 9.10 provided s tatces on values given by equation
9.13. Equation 9.10 does not constrain K\ however, the initial conditions will.
Equations such as 9.12 whose solution is given by equation 9.13 are a common characteristic o f
second-order networks. Hence, equation 9.12 is called the characteristic equation o f the secondorder linear circuit. The associated roots, equation 9.13, are called the natural frequencies o f the
circuit. These are the natural or intrinsic frequencies o f the circuit response and are akin to the
natural frequencies o f oscillations o f a pendulum (for small swings) or o f a bouncing ball.
From elementary algebra, a quadratic equation (the above characteristic equation) can have dis
tinct roots or equal roots. Distinct roots can be real or complex. Thus
and Sj can be two dis
tinct real roots, two distinct conjugate complex roots, or two repeated (equal) roots, depending
on whether the discriminant, Ip- - 4r, is greater than, less than, or equal to zero. This trifold
grouping separates the solution o f equation 9.10 into three categories, listed below as cases 1, 2,
and 3:
Case 1. Real and distinct roots, i.e., b~ -A c> 0. If the roots are real and distinct, then for arbitrar}'
and K2 , both
constants
.v( o = .vi (/) =
a:,6^"''
and
xif) = X2it)=K2e-'^-'
satisfy the second-order linear differential equation 9.10, i.e., are solutions to the differential equation. Since equation 9.10 is a
differential equation, by superposition the sum x(r) = x,(^) +
X2 (t) is also a solution, a fact easily verified by direct substitution. Therefore, whenever
s-y, the
most general form o f the solution to equation 9.10 is
xit)=Kê^^' + K2e^^-
The constants
(9.14)
and K-, depend on the initial conditions o f the differential equation, which
depend on the initial capacitor voltages and inductor currents. For example, if a.*(0'^) and a''(0'*')
are known, then from equation 9.14,
.v (0^ )=
A y '*'-h
and
.v-(O^) = ^
(It
= .s'l
1= 0 ^
+ stK')
Chapter 9 * Second Order Linear Circuits
389
These are simultaneous equations solvable for

If
and
and $2 are negative, the response given by equation 9 .1 4 decays to zero for large f and the cir
cuit is said to be overdamped.
Case 2. The roots,
and $2 , o f the characteristic equation are distinct hut complex, i.e., iP- -A c <Q.
Since Jj ^
general form of the solution to equation 9.1 0 is again given by equation
9.14, i.e.,
=
with complex
and
$2
given by
+K2e"^^
_______
-b
. y l^ c -b ^
^2 = Y J ------ ------- =
,
J^ d
(9.15)
y l^ c-b ^
where a = bH and co^ = ------ ^ S i n c e s^ and S2 are conjugates, so are dîând e^2în equation 9.14.
For x(t) to be real, the constants /Cj and A'j in equation 9.1 4 must also be complex conjugates,
i.e.,
Using Eulers formula,

giy = cos
+ j sin y
the two terms in equation 9.14 combine to yield a real time function:
= e
co s(co jf) + jKisinioi^t)] + e
ATj cos((0^/) - yATj sin(cojO
(A"! + A"! )cos(to^f) + (yî + jK i )sin((Ojf)
Thus the solution to equation 9.1 4 with
and S2 complex is given by the (damped) sinusoidal
response
x(r) = e~^^ [A cos((o/) + B sin((0/ )]
where ^4 =
+^
= 2 Re[^j] =
+ A^2 and 5
(9.16)
- y Af] = - 2 ImATJ = /A j - yATj are real con
stants and where Re[ ] denotes the real part and Im[ ] denotes the imaginary part. The solution
expressed in equation 9.16 is completed by specifying ^4 and B. As before, A and B depend on the
initial conditions, x(0^) and xXO"*^) as follows:
jc(0'*) =
[Acos(tOjf) + Bsin((0^r)])^_^^ = A
390
jr'(0'*) =
-a e ^^^y4cos (o)^0 + 5sin (coô) + e
co^y4sin (coÔ + (O^Bcos (coô)
= - o A + (o^B
and
which are easily solved for A and B.
Making use of a standard trigonometric identity, the general solution of equation 9.16 has the
equivalent form
x{t) =
[A cos(coj) + B sin(o)^)] = Ke~^^ cos(o)^ + 0)
(9.17a)
where
K = ylA^ + B^ ,Q = tan"
f-B \
(9.17b)
KA)
and the quadrant of 0 is determined by the signs o f - B and A. In MATLAB, one uses the command
atan2(-BvA) to obtain the angle in the proper quadrant. Note that the response waveforms have
oscillations with angular frequency O)^ These oscillations are bounded by the envelope Ke~^*. If
Re[jj] = - a < 0, the amplitude of the oscillations decays to zero and the response is said to be underdamped. If Re[jj] = - a > 0, the amplitude of the oscillations grows to infinity.
C ased. The roots are real and eqml, i.e.,lP' - 4 c = 0 . When the two roots of the characteristic equa
tion are equal, equation 9.1 4 does not represent the general solution form because if
two terms collapse into a single term. However, the general solution for
the
is
x(t) = {K^+K2t)e^^^
(This is investigated in a homework exercise.) Calculation of
and
in equation 9.18 is
straightforward:
x{0-) =
and
i(0+) =
Substituting the value o f
into x'(O^) yields a simple calculation for Kj.
If^l = i2 is negative, the response decays to zero and is said to be critically damped. Critically
damped defines the boundary between overdamped and underdamped. This means that with a
slight change in circuit parameters, the response would almost always change to either over
damped or underdamped.
T he discussion o f the three cases is summarized in Table 9.1.
391
TABLE 9.1. General Solutions for Source-Free Second-Order Networks
General solution o f the homogeneous differential equation
d~x
dx
dr
dt
^ + h + cx - 0
having characteristic equation
+ r = (j - 5 j)(j -
= 0, where
- h yjb" 4 c
^2=
-----------
Case 1. Real and distinct roots, i.e., ^ - 4 c > 0:

x {t)=
+ Kie^-
where
40^) =
and .v (0") = s^K^ + s^K^,
Case 2. The roots, i| = - a + p i^ an d ^2 =

~j^d>
tion are distinct but complex, i.e., tr - Ac < Q:
characteristic equa
x{t) = e~^ [A cos(o)/) + B sin(o)/)] = Ke~^' cos(to/ + 0)

where
x(0^) = A . x'(0+) = - 0 .4 + CO/
and
+
, e = tan"^
Case 3. The roots are real and equal, i.e.,
(-B \
A
= Sj and Ir - Ac
xU) = (K^ + K2t)e^'

where
x{0*) =
and a-'(O^) = s^K^ + Kj
Figure 9.3 displays the various response forms described above for the case where Re[xj] and Re[j-,]
are negative or zero. Because o f their similarity, it is not possible to distinguish between the over
damped and the critically damped responses by merely looking at the waveforms. Both types o f
response may have at most one zero-crossing.
392
Chapter 9 Second Order I.incar Circuits
> t
FIGURE 9.3 (-'tncric waveforms corresponding to the four cases o f damping: (a) undamped (sinu
soidal) response, (b) undcrdamped (exponentially decaying oscillatory) response, (c) overdamped
(exponentially decaying) response, and (d) critically damped (exponentially decaying) response.
The terms undamped, underdamped, overdamped, and critically damped stem from an
intuitive notion o f damping. The sourcc-frce response o f an undamped second-order linear sys
tem, whether electrical or mechanical, has an oscillatory response (waveform) o f constant ampli
tude. Damping, due to system elements that consume cnerg\', means a monotonic decrease in the
amplitude o f oscillation. In electrical circuits, resistances produce the damping effect. In mechan
ical systems, friction causes damping. When the amount o f damping is just enough to prevent
oscillation, the system is critically damped. Less damping corresponds to the underdamped case,
where oscillation is present but eventually dies out. A greater amount o f damping corresponds to
the overdamped case, where the waveform is non-oscillatory, and a ver\ small perturbation o f any
circuit parameter will not cause oscillations to occur.
In summary, once the roots o f the characteristic equation are found and the expression for the
general solution selected from the above cases or Table 9.1, it remains to find the constants A', and
3^)3
K j (or A and 5 ) from the initial conditions on the circuit. In the above development,
and Kj
(or A and B) are given in terms of a:(0^) and x'(O^). Since jf(/) represents either a capacitor voltage
or an inductor current, its value at f = 0"^ is usually given, or can be determined from the past his
tory o f the circuit. (See Example 9.5.) The value o f x'(O^), on the other hand, is often unknown
and must be calculated. If >:(/) =
then the capacitor v-i relationship implies that
xX0*) = v'c{0*) = ^ ^ .
^
If x{t) =
then the v-i relationship of an induaor implies that
The problem then reduces to finding an unknown capacitor current,
, or an unknown
inductor voltage, y^(O^).

To find
or y^(O^), we construct an auxiliary resistive circuit valid at r = 0^. Since the initial
values,
and /^(O^), are known, we replace (each) capacitor in the original circuit by an inde
pendent voltage source o f value
and (each) inductor in the original circuit by an inde
pendent current source o f value /(0'*^). Here the current /(;^0^) retains its original direction and
the voltage ^^(0^) retains its original polarity. After the replacements, the (new) circuit is resistive.
Values for
and y^(O^) follow by applying any of the standard methods of resistive circuit
analysis learned earlier. This allows us to specify x(0^) and x'(O^) in terms of the initial conditions
on the circuit. Two equations in the two unknowns A'j and K2 (or A and B) result. Example 9.5
and, in particular. Figure 9.4c illustrate this procedure.
Response Calculation o f Source-Free Parallel and Series RLC Circuits

Before any additional circuit examples, let us summarize the solution procedure.
Procedurefor Solving Second-Order RLC Circuits

Step 1. Determine the differential equation model of the circuit.
Step 2 . From the differential equation model, construct the characteristic equation and
find its roots using the quadratic root formula.
Step 3. From the nature of the roots (real distinct, real equal, or complex), determine the
general form of the solution from Table 9.1; the solution form will contain two unknown
parameters.
Step 4. Find the two unknown parameters using the initial conditions on the circuit.
The following example illustrates these calculations for the three cases described in Table 9.1.
39-'i
E XA M PLE 9 .5 . In the circuit o f Figure 9.4a, the 1 pF capacitor is assumed to be ideal, and the
inductor is modeled by a 10 mH ideal inductor in series with a 20 Q resistor to account for the
resistance o f the coiled wire. Suppose the switch S in Figure 9.4a has been in position A for a long
time. The capacitor becomes charged to 10 V. Then the switch moves to position B at f = 0. Find
and plot V(\t) for r > 0 for the following three cases; (1)
= 405
(2) /?2 = 0, (3)
180 ti.
Each o f these cases produces a different response type.
:L=10mH
lO O
+
10V
C=1
R=200
Practical Inductor
mF
(a)
i,(0*) = 0
L=10mH
+
C=1
mF
(b)
FKJURH 9.4 (a) Discharge of a capacitor through a practical inductor in series with a resistance /?,.
(b) Kquivalent circuit for t > 0. (c) Equivalent circuit at / = 0^ for calculating
in which the
inductor has been replaced by an independent current source of value //(O^) and the capacitor by an
independent voltage source of value
SoL u rioN
From the problem statements,
= 10 V. When the switch moves to position B,
- 10 V by continuit)' o f the capacitor voltage; the circuit now becomes a series RLC, for t
> 0, with / ?= / ?!+ R-y as shown in Figure 9.4b. The first step in the calculation o f the circuit
response is to find a second-order differential equation in the unknown
From Example 9.3,
for the series RLC,
R dvf'
*- + -------^ +
dt~
L dt
\r = 0
LC ^
(9.19)
Since L and C are known, the series RLC characteristic equation is

o R
1
s~ H .y H------ = .v + ( 2 0 + /?-,)10^y-r-1 0 ^ = 0
L
LC
With this framework, we can separately investigate each o f the three cases.
( 9 .2 0 )
Case I: /?2 = 405 Q or /? = 425 O.

Step 1. Find the characteristic equation and the generalform o f the response using Table 9.1. If R-, =
405
the characteristic equation o f 9.20 is
+ 42,500x + 10^ = 0. Solving for the roots by the
quadratic formula yields

2 = - 2 1 ,2 5 0 18,750 = -2 5 0 0 , - 4 0 ,0 0 0 sec ^
Real distinct roots imply an overdamped response o f the form
v c(0 =
+ K2e^~ =
(9.21)
valid for r > 0.

Step 2. Fijid
and K-,. Evaluating at f = 0"^ implies

z;^0") = 10 = /Tj + /f2
(9.22a)
Differentiating equation 9.21 implies
vL(0+) = -^^^^ = -2.5xlO^/r, - 4 0 x \ 0 ^ K 2

C
From the circuit o f Figure 9.4c, v ^ (0^ ) =
(9.22b)
^ = 0 , where //^(O'*') = i[{0~)
by the continuity o f the inductor current. Solving equations 9.22a and b after substituting the
above values yields
AT, = 10.667 and K-, = - 0 .6 6 7
Step 3. Set forth the solution for V(^t). For / > 0,
v^t) = 10.667^-2500^-0.667e-^0'V
This function is plotted in Figure 9.5.
Exercise. You may verify' the above answer with the Student Edition o f MATLAB (version 4.0 or
later) by typing the command: y = dsolve(D 2y+42500*D y + leS^y = 0,y(0) = 10, Dy(0) = O).
Case 2: /?, = 0 or /? = 20 Q
Step 1. If /?2 = 0> then from equation 9.20, the characteristic equation is
+ 2,000^ + 10^ = 0.
Since tr - Ac = -3 9 6 ,0 0 0 ,0 0 0 < 0, the roots are complex. From the quadratic root formula,
= - 1 0 0 0 + y9950 = - a + yo)^ and Sj = - 1 0 0 0 -y '9950 = - a
From Table 9.1, the underdamped response form is
v^t) =
[A cos(w/) + B sin(co/)] = ^"1000/
cos(9950f) + B sin(9950f)]
(9.23)
396
Step 2. Fiiid A and B. Ir remains to determine A and B in equation 9.23. From equation 9.23
and its derivative,
i / ^ 0 ^ ) = 1 0 = /I
(9 .2 4 a )
and
= -oA + ( .) / = -1000/1 + 9950/y
(9.24b)
^ = - = 0. Substituting into equations 9.24 and
As in case 1, V(- '(O"^) = ^

solving yields
C
A = 10 and
= = 1.005
CO,/
Step 3. Set forth the solution for v^{t). For t > 0,
v^t) = ^-'OOOrjio cos(9950?) + 1.005 sin(9950^)] = 10.05^'-^^^' cos(9950^ + 5.7) V

This waveform is also plotted in Figure 9.5.
Exercise. You may verify the above answer with the Student Edition oF MATLAB (version 4.0 or
later) by typing the command: y = dsolve(D 2y+2000*D y + le8*y = 0,y(0) = 10, Dy(0) = O).
Case 3: ^2 = 180
Step 1. If
=180
or R= 200 Q
then the characteristic equation from 9.20 is
+ 2 0 ,0 0 0 j + 10^ = 0, whose
roots are
= -lO '*, implying a critically damped response. From Table 9.1, the general criti
cally damped response form is, for r > 0,
v^(/) = (/r, + K2t)e^^' = (A", + K2t)e~^^'^
(9-25)
Step 2. Find /Tj and Kj. From equation 9.25, its derivative, and the known initial conditions
from cases 1 and 2,
vîO^) = 10 = A",
(9.26a)
and
^yO ") =
+ K, = -10^ A', + A'2 = 0
(9.26b)
Solving equations 9.26 yields

A', = 10 and Kj = --^lA", = 10^
Step 3. Set forth the sohition for vî). For /^> 0,
v'c(n = ( l0 + 1 0 ^ ) e " '^ '
The waveforms o f V(fJ) for the three cases (underdamped, critically damped, and overdamped)
are plotted in Figure 9.5.
39'
FIGURU 9.5 Waveforms of vj,t) in Example 9.5 for three different degrees of damping. Critical
damping represents the boundary between the overdamped condition and the oscillatory behavior of
underdamping.
Exercise. Verify the answer calculated in Example 9.5 using the Student Edition o f MATLAB and
the desolve command.
On a practical note, commercially available resistors come in standard values each with an associ
ated tolerance. Tolerances vary from 1% (precision resistor) to as much as 20% . Further,
because o f heating action over a long period o f time, resistance values change. Given the above
example, in which the type o f response depends on the resistance, one can imagine the care need
ed in the design o f such circuits: without consideration o f precision and long-term heating effects,
a desired critically damped response could easily become oscillator)'.
Not all second-order circuits arc RLC. Some are only RC but with two capacitors and some are
RL with two inductors. Passive RC or RL circuits cannot have an oscillatory response. The proof
o f this assertion can be found in texts on passive network synthesis. However, with controlled
sources a second-order RC or RL circuit can have an oscillatory response that is not characeristic
o f a first-order circuit, but o f a second- or higher-order circuit. The example below illustrates the
analysis o f a second-order RC circuit containing controlled sources that has an oscillator)^
response.
398
EX A M PLE 9 .6 . This example illustrates the analysis o f the second-order RC circuit shown in
Figure 9.6. The objective is to find
= 10 V and
= 0.
'Cl
1 kn
I
for / > 0 given the initial conditions V(^{0)
and
mF
FIGURE 9.6 Sccond-order RC circuit with controlled sources that has an oscillatory response.
S o l u t io n
Step 1. Write a dijferential equation in V(^^{t). From the properties o f a capacitor and KCL at the
left node,
di
Multiplying through by 10*^ yields
d\
^^ 10
10^ v^2 ~ 1 0 'vf'i
dt
(9.27a)
We expect a second-order differential equation, so differentiating a second time yields

_ 1()5
, Q3
dt
dt~
To obtain a differential equation in
(9.27b)
dt
equation 9.27b must be eliminated. This
requires another relationship between tQand
. At the right node,
106 C l _
_ -0 .
dt
or equivalently,
dt
= -1 0 ^ Vc,
(9.28)
Substituting this expression into equation 9.27b produces
dt
*2
After rearranging terms, we obtain a second-order differential equation in
'2
dt^
dt
(9.29)
Step 2. Determine the characteristic equation, its roots, and the form o f the response. The differential
equation 9.29 has characteristic equation
+ 1 0 = 0
From the quadratic formula, the complex roots are
= - 5 0 0 + 799,998.75 = - a y W
From Table 9.1, complex roots imply an underdamped response o f the form
[A cos(co/) + B sin(co/)] = ^" 500' [A cos(99,998.75r) + B sin(99,998.75r)] (9.30)

Step 3. Find A and B. \x. t = 0, z^qCO) = \() = A. Also, from equation 9.27a and the initial con
ditions,
^^^'Cl(Q) ^ i0 - \ ;^ .,(0 ) - iq 3 v ^ ,(0 ) = - 10^
dt
(9.31)
Differentiating equation 9.30, evaluating at r = 0, and equating the result with equation 9.31
produces
dl
"
in which case B = 5.0001 x 10^.

Step 4. Set forth the final form o f V(-^{t). The final form o f the response is
v^{t) = ^'-500^ [10 cos(99,998.75^) + 5.0001 x 10"2 sin(99,998.75r)] V

= 1 0 ^ 5 0 0 'c o s (9 9 ,9 9 8 .7 5 r- 0.2865) V
Step 5. Plot the response. A plot o f the (underdamped) response is given in Figure 9.7.
Exercise. Construct a parallel RLC circuit to have the same second-order difFerential equation
model as 9.29. Note that there is no unique solution.
C H EC K : /?C= 10-^ and ZC = \0~^^
It is important to observe here that the design o f Example 9.6 achieves a second-order RLC
response without the use of an inductor, which is important for integrated circuit technolog)'.
4. SECON D-ORDER LINEAR NETW ORKS WITH CON STANT

INPUTS
The preceding section studied source-free second-order linear networks. When independent
sources are present, such as in the circuit o f Example 9.7 below, the network (differential) equa
tions are similar to the source-free case except for an additional term that accounts for the effect
o f the input:
^
clf-
+ / ; ^ + c-.v = /(/)
dt
(9.32)
wherey{r) is a scaled sum o f the inputs and/or their first-order derivatives. Ordinarily one might
expecty(f) to be the value o f the input. A homework problem illustrates that j{t) can depend not
only on the source input, but also on the derivatives o f the input. For general circuits, those not
reducible to parallel or series RLC circuits, constructing equation 9.32 can be a challenge. Further,
the solution o f 9.32 for arbitrary inputs and initial conditions is no less challenging but is best
obtained via the Laplace transform method, which is a topic studied in a second circuits course.
However, when the input excitations are constant,/r) = F, the solution to 9.32 is a straightfor
ward modification o f the source-free solution, as explained in the remainder o f this section.
Since the expressions o f Table 9.1 satisfy the homogeneous diflferential equation 9.10, the gen
eral solution to equation 9.32 follows by adding a constant
to each o f the solution forms given
in Table 9.1. Specifically, the general solution o f the driven differential equation
tl~x
dx
- + h ^ cx = F
(9 . 33 )
xit) = .V,//) + Xj:
(9.34)
dr
dt
where-v^j(f) is the solution to the homogeneous equation 9.10 (equivalently, equation 9.33 with F
= 0). Recall that the form o f x j t ) is determined by the roots of the characteristic equation r + bs
+ r = (i - ^])(j - ^2) = 0, given by the quadratic formula
h yjh~ - 4 c
i |-) = -------------9
9
*0 1
Chapter 9 * Scconcl Order Linear Circuits
To verify that the structure o f equation 9.34 is a solution to 9.33 and to compute the value o^Xp sub
stitute the structure given by equation 9.34 into 9.33. Since
satisfies the homogeneous equation
9.10, it contributes zero to the left-hand side. What remains is cXp= F. Therefore,
= which
is independent o f the roots o f the characteristic equation. However, if Re[j|] and Re[^2] < 0, then
tends to zero for large t. Hence x{t) tends to Xp for large t. Consequently Xp is termed the
final value o f the response.
Because o f the trifold structure o f
as summarized in Tible 9.1, the solution form o f equation
9.34 once again breaks dow'n into three distinct cases. We summarize this trifold structure for the
constant-input case in Table 9.2.
TABLE 9.2 General Solutions for Constant-Source Second-Order Networks
General solution o f the driven differential equation
-z- + h + cx = F
dr
dt
having characteristic equation p- + bs + c = (j - Jj)(^ - ^2) = 0 with roots
h yjb^ - 4 c
-7 = ---- --------------
2
2
Case 1. Real and distinct roots, i.e.,lP' 4 r > 0:
x {t)=
+ Xp
F
with Xp = . Further,
c
x(0'^) =
+ K2 + A'yând x'(0'^) = s^K^ + s^K^
Case 2. The roots,

= - a + y'co^y and S2 = - o
o f the characteristic equation
are distinct but complex, i.e.,
- 4 c <Q. The general solution form is
x(r) = e~^^ [A cos(oj/) + B sin (to/)] + Xp = Ke~^^ cos(to/ + 0) + Xp
where again Xp = , with
x{0^) = A + X p x(0^) = - g A +
K
V
Case 3. The roots are real and equal, i.e., s, = s, and Ir -A c= Q . The solution
form is
.v(/) = (/ r ,+ ^ 2 0 ^ '
where Xp = , and
x{Q^) = K^+Xp
and .v W "-) = .v ,^ i-h AT2
402
The interpretation Xp = F/c. is a mathematical one. When the differential equation describes a lin
ear circuit with constant inputs, there is a physical interpretation o f Xp and a circuit theoretic
= Xp
method for computing its value, even without writing the differential equation. Since
= a constant satisfies the differential equation 9.33, it is also a constant solution to the circuit.
Hence Xp\s either a constant capacitor voltage or a constant inductor current. If a capacitor volt
age is constant, its current is zero; this is interpreted as an open circuit. Similarly, if an inductor
current is constant, its voltage is zero; this is interpreted as a short circuit. Therefore, Xp is an
appropriate (capacitor) voltage or (inductor) current obtained when the capacitor (or capacitors)
are open-circuited and the inductor (or inductors) arc short-circuited. The value o f Xp can be
obtained by analyzing the resistive network resulting when all capacitors are open-circuited and
all inductors are short-circuited. Recall that if Re[^j] and Re[^2] < 0, then x{t) tends to the constant
value Xp Physically speaking, then,
equals either
or /^(o))when Re[;,] and Re[^2] < 0.
Once the proper general solution structure is ascertained from Table 9.2 and the constant Xp is
found, the parameters
(or A and B) are computed by the same methods used in the
and
source-free case. The following example illustrates the procedure for a parallel RLC circuit.
= u{t) A, excites the parallel /^ZCcircuit o f Figure 9.8,
E XA M PLE 9.7 . A step current input,
whose initial conditions satisfy /)(0) = 0 and y^Ô) = 0. This simply means that the current source
turns on with a value of 1 amp at r = 0 and maintains this constant current excitation for all time.
The objective is to find the inductor current,
for / > 0, for three values o f R: (i) R = 500 Q,
(ii) R = 25 n , and (iii) R = 20 Q..
I'IGURE 9.8 Parallel RLC circuit cxcited by a step current input.
So l u t io n
Because the circuit is a parallel RLC, the characteristic equation is
. r + - L , + - L = ,2 + 1 0 l,v + 4 x l 0 ' " = 0
RC
LC
(9.35)
For all positive values o f R, the roots of the circuits characteristic equation have negative real parts.
Thus for large t or ideally at t = oo, the inductor looks like a short circuit and the capacitor like
an open circuit. Hence, for all cases o f this example, Xp = i[{^) = 1 A; note that y^co) = 0 because
the inductor looks like a short at t = <x>.
Case I. For R = 500 Q, the characteristic equation 9.35 reduces to r + 2 0 ,0 0 0 j + 4

From the quadratic formula, the roots are
2 = - 1 .0
lO'^ 7 I .9975
105 = - a yco^
10*^^ = 0.
'103
which indicates an underdamped response o f the form (Table 9.2)
[A cos(coy) + B sin(o)^)] + Xp= Ke~^ cos(to^ + 0) +
/^(r) =
= A + Xp = A + 1, then A = -\. Further,
Since 0 =
elicit)
0=
dt
^
= -0/4 + (HjB
1 = 0^
From derivative o f
expression for ij^t)
From physical circuit
This implies that

fi = = - 5 .0 0 6 3 x 1 0 "
(0,y
Hence for / > 0,
/^(r) = ^.-10.000^ [ cos(1.9975 x IQ5f) + 5.0063 x 10"2 sin(1.9975 x lO^r)] + 1
= 1 .0 0 13^>-*^^^'cos( 1.9975 x 105^+ 2.866') + 1 A
Case 2. For R = 25 O., the characteristic equation 9.35 reduces to

From the quadratic formula, the roots are
+ 4 x 10^5 + 4 x lO' = 0.
j, 2 = - 2 . 0 x 105
indicating a critically damped response o f the form (Table 9.2)
iL^t) = (K ^ + K 2t)e '' + Xp

Since 0 = /^(O^) = A^, + Xp =
Q
+ 1, then
= - 1 . Further,
_ vc(0'*') _ vz.(Q~^) _ d ilit )
dt
= SiKi + K2 = 2xW ^ -hK2

1=0*
This implies that K y - - I y. 10^ and for r > 0,

-2xio-\
Case 3 . For R = 20 Q, the characteristic equation 9.35 is

quadratic formula, the roots are
= - 1.0
10^ and S2 ~ ~
specifying an overdamped response o f the form (Table 9.2)
+ 1A
+ 5 x 1O^j + 4 x 10^ = 0. From the
40 4
+ Xp. Furtiier,
Evaluating this response at f = 0+ yields
dt
t=(f
1
Equivalently, -K^
~ Solving these two equations yields K\ = ---- and K-, = .
Therefore the actual response for r > 0 is
^
^
Figure 9.9 displays a graph o f the response for each o f the three cases.
FIGURE 9.9 Underdamped, critically damped, and overdamped

response curves for the parallel RLC circuitof Example 9.7.
Exercises. 1. Show that for t > 0 , the differential equation for the circuit o f Example 9.7 with R
= 500 Q is
clr
+ 2 x l 0 ' ^
dt
+ 4xl0',',.<n = 4 x l 0 JO
'
2. Use MATLAB s dsolve command to verify the solution obtained for case 1 in Example 9.7.
In a linear circuit or system, the response to a step input often indicates the quality o f the system
performance. The problem o f measuring a batter}' voltage using a voltmeter is illustrative o f this
indicator. Here the battery dc voltage is the input and the output is the meter pointer position.
Connecting the meter probes to the batter)'^ terminals amounts to applying a step input to the
voltmeter circuit that drives a second-order mechanical system consisting o f a spring and mass
with friction. Naturally, one would like the pointer to settle on the proper voltage reading quick
ly. If the mechanical system is underdamped, then the pointer oscillates (undesirably) for a short
time before resting at its final position. On the other hand, if the mechanical system is over
damped, the pointer will not oscillate but may take a long time to reach its final resting point.
This also is undesirable. A near critically damped response is the most desirable one: the pointer
will come to rest at the proper voltage as quickly as possible without being oscillatory, and small
changes in the mechanical system will not make it oscillatory.
In the next example, we reverse the process o f analysis and ask what the original circuit parame
ters are given a plot o f the response that might have been taken in a laboratory.
E X A M PLE 9 .8 . Consider the circuit o f Figure 9.10, which shows the response,
of a
(relaxed) series RLC circuit to the voltage input i/y^(r) = 10(r) V. In laboratory, you have meas
ured the capacitor voltage values (approximately). If the response has the form vît) = Ke~^^
cos(ojy + 0) + vYp find
a , 0, K, and the values o f R and C i f it is known that L = 0.5
H. Your lab instructor has told you that toând a are integers.
(a)
___________,
406
(b)
F IG U R t 9.10 (a) Series RLC\ (b) Response to
= 10 u{t) V.
TABLE 9.3
Tim e (sec)
0.316
0 .5 2 3 6
0.839
1.5708
vcKt) (V)
10
13.509
10
10.432
First crossing
First
Second crossing
Second
o f 10 V
peak
o f 10 V
peak
S olution
Step 1. FindXp. By inspection, the curve is settling out at X p - 10 V.
Step 2. Find
40
Now obscr\'e that the first two crossings o f v^^t) = 10 occur at r = 0 .3 1 6 sec,
0.839 sec (Table 9.3). This means that a full k radians is traversed by tlie cosine over [0.316,
0.839] , which is a half c>de or half period. So the period o f the cosine is 7 = 2(0 .8 3 9 - 0.316)
2 ti
= 1.046 sec, making

CO^/ =
= 6.007 = 6 rad/sec.
Step 3. Find O . From Table 9.3, we know that two successive peaks occur at
= 0.523 sec
and t-y = 1.5708 sec. This means that for ^ = 1 ,2 ,

cos(COj//. + 0 ) + X/r
(9.36)
After some manipulation, equation 9.36 implies
COS(Ci),y/l + 0 )
C0.s((0^/r2 + 0 )
Thus
cos((0 ,//| +Q) V c i t j ) - ^ ^,-a(/2-/,)
Xf,-
cos(co^y/'> + 0 )
(9.37)
Equation 9.37 simplifies because two adjacent positive peaks must be 2 tu radians apart, i.e.,
(co/2
+ ^) =
which means cos(to^^^ + H) = cos((i)y, + 0). It follows that
Xc d o - ^ F
3 .509
(9.38)
Solving leads to a = 2.
Step 4. Find 0 and K. At the first crossing o f 10 V, we have
0 =
0.316
0 . 3 1 6 + B)
Thus 6 X 0.316+ 0 must equal 0.5?! or 1.5tt radians. We also know that since
=0 =K
cos(0) + 10, we must have K cos(0) = - 1 0 . Since A"> 0 (always by convention), the value o f
cos(0) must be negative. This means 0.57t < 0 < 1.5Tt. So therefore, it must be that at the first cross
ing
o f 10 V
D
6 X 0 .3 1 6 + 0 =
or 0 = 2.1864 rad.
Therefore K = ^ = 10.553.
cos(0)
Step 5. Find R and C. We know that the characteristic equation o f the series RLC circuit must be
Therefore
f?
1
02
^
^
s~ H- . V.V +H-------=
2Rss +H = (5' + 2)*' + 6 " = s~ + 4.V + 40
------- = .v + IR
L
LC
C
/? = 2 n and C = 0 .0 5 F
-H)K
Chapter 9 Sccontl Order Linear Circuits
In the previous examples one obsen'e that the characteristic equations are independent o f the
source values. I'his is a general property of linear circuits with constant parameters. Hence when
constructing the characteristic equation we may without loss of generality set independent source
values to zero; i.e., independent voltage sources become short circuits and independent current
sources become open circuits. With this operation, some circuits that appear to be non-series/parallel, become series/parallel. This allows us to easily compute the characteristic equation and then
use Table 9.2 and physical reasoning to obtain the solution without having to construct the dif
ferential equation explicitly. The following example illustrates this procedure for a pseudo-parallel/.series RLC. The example will also illustrate the computation o f initial conditions due to past
excitations and the computation o f the complete response w'hen the input changes its dc level.
EXA M PLE 9.9 . The circuit of Figure 9. l i b is driven by the input o f Figure 9.11 a, i.e., vj^t)
-60//(-r) + Gû{t) + 60//(/ - 1) V. Our goal is to find the response
for / > 0.
FIG URE 9.11 (a) Input cxcitation whose dc level changes at r = 0 and t = 1 second, (b) A pseudoparallel RLC circuit; i.e.,when the voltage source is replaced by a short, the circuit reduces to a paral
lel RLC whose characteristic equation is p- +
RC
s+
= 0.
LC
So l u t io n
Step 1. Analysis at 0. Here the circuit has been excited by a constant - 6 0 V level for a long
time. Therefore at ^ = 0", the capacitor looks like an open circuit and the inductor a short cir
cuit. Because the inductor looks like a short, the entire - 6 0 V appears across the 6 i l resistor,
making
= - 60 V and /^(O") = - 60/6 = 10 A.
Step 2. Analysis at 0^. By the continuity o f the capacitor voltage and the inductor current, the
equivalent circuit at 0"^ is given in Figure 9.12.
409
= -6 0 V
^ (0 *)= i,(0 )
= -10A
FICJURE 9.12 Equivalent circuit for analysis at 0^; the capacitor is replaced by a voltage source
o f value
and the inductor by a current source o f value /^(O*) = //;(0).
From the circuit diagram o f Figure 9.12, v^{0*) = 60 - ( - 60) = 120 V,
= - 60/6 = - 1 0 A, and
iff-) = if^{0*)l5 = 40 A. It follows that

Step 3. Find the characteristic equation and the form ofthe response using Table 9.2. To find the char
acteristic equation, we set the independent voltage source to zero. I'he resulting circuit is a paral
lel RLC with characteristic equation
-)
1
I
T
-)
.v + ----- .V+ ------= .v + 4.V + 4 = (.V + 2 ) = 0
RC
LC
where R = 2 LI is the parallel combination o f 6 Q and 3 i i . The characteristic roots are sî = - 2 ,
which correspond to a critically damped response o f the form (Table 9.2)
V(4,t) = (/f,
Kjt)
exp(j,
t) + Xf:
Step 4. Find constants in the response form for 0 < /< 1. The input is constant for 0 < ^ < 1, but
changes its value to 120 V at r = 1 sec. However, the circuit does not know the input is going to
change, and so its response behaves as if the input were to remain at 60 V for all time: the circuit
cannot anticipate the future, and thus its response over 0 < r < 1 behaves as if no further switch
ing were going to occur. If no further switching were to occur and if the input remained at 60 V,
then in Figure 9.10b for large t the capacitor is an open circuit and the inductor is a short circuit;
hence Xp= 60 V. Under these same conditions we find
and K-y. To find A^j, observe that from

= - 6 0 V. Evaluating the response form o f step 3 yields
+ Xp. Equating
these rwo expressions produces - 6 0 = ^'(;(0^) =
+ X^ which implies thar
= - 1 2 0 . To calcu
late Kj, observe that from step 2, /(-(O^) = 40 A. Since C = 0.125, it follows that
step 2,
c!t
H en ce,
K-,
/={)'
= 320 =
+ A'2 = 240 + ^2
= 8 0 . T h u s, the response o f the circu it for 0 < r < 1 is
V({t)
= (-1 2 0 +
m)e~^ +
60 V
410
Similarly, one can compute, for 0 < r < 1,
il{t) = (-2 0 +
+ 10 A
Step 5. Analysis at t = 1. Although the circuit does not know the input will change at r = 1 sec,
we do and we must prepare for the analysis for f > 1. To do this we must evaluate the initial con
ditions at / = 1 and then use the continuity o f the capacitor voltage and inductor current to
obtain the initial conditions at / =
At r = 1, using step 4 we have V(^\~) =
= 54.59 V
and/^d") = /^ (r) = 10 A.
Step 6. Analysis att= I
This step mimics step 2 for r = P . The capacitor is replaced again by an inde
pendent voltage source and the itiductor by an independent current source as shown in Figure 9.13.
Here v^{V) = 1 2 0 - 54.59 = 65.41 Vand
=-in\ +
+ //(O'*') =
= - 9 .0 9 8 + 21.8 + 10 = 22.71 A.
^ 120 - 54.59 ^
I
>
Step 7. Computation o f the responsefor t> 1. Because the characteristic equation is independent o f
the input excitation, the form o f the response is almost the same as in step 3, except for the
replacement o f t by (/ - 1); this substitution follows by the time invariance (constant parameter
values) o f the circuit. Thus, for r > 1,
v^t) = \K^ + K\{t- 1)1 e x p U ,(r- 1)]

Since the source excitation for r > 1 is 120 volts, by inspection o f Figure 9.1 lb
V. To find /f], 54.59 = ^'c^l"^) = ^p+ A',. This implies
f^ôo) = 120
= -6 5 .4 1 . Finally, to find K-, consider
that
=
dt
w hich makes
1 8 1 .7 = .ViA:, + K2 = 1 3 0 .8 + K 2
/=r
= 5 0 .9 . T h u s, for / > 1,
v^t) = [ - 6 5 .4 1
+ 5 0 . 9 ( r - l)]^>-( ' -
+ 120 V
(9 .3 9 )
Chapter 9 Scconcl Order Linear Circuits
41
Time in seconds
FIGURE 9.14 Complete analytical response of the capacitor voltage for 0 < r < 3 sec.
Exercise. Fill in the details for the computation o f iît) = (-2 0 + 20/)^^ + 10 A for 0 < r < I and
then compute /y(/) for 2.5 > ^ > 1. Also, compute /^(/) for t> 2.5.
Despite the idea illustrated in Example 9.9, many second-order RLC circuits are not reducible to
series or parallel RLC circuits when the independent sources are set to zero. Furthermore, when a
dependent sourcc is present, the circuit is generally not reducible to a series or parallel RLC. In
such cases one ordinarily uses a systematic methodolog)' to compute the circuits differential equa
tion and, subsequently, the characteristic equation. This systematic procedure is described in more
advanced texts and in the second edition o f this text. Nevertheless, for some situations one can use
the earlier method s integro-differential equations, which must be differentiated again to eliminate
the integral. This is illustrated in the example o f the next section.
il2
5. OSCILLATOR APPLICATION
An imporrant difference between first-order and second-order linear networks is the possibilit)' o f
oscillatory responses in the latter. In some applications sinusoidal oscillations are intended
responses, while in other applications oscillations arc undesirable. This section presents an exam
ple o f a Wien bridge oscillator circuit.
The goal is to build a circuit that generates a pure sinusoidal voltage waveform at a specified fre
quency. In theory, as per section 2 of this chapter, this is achievable by discharging a capacitor
through an inductor. In practice, both capacitor and inductor have losses. Losses cause the oscil
lation amplitude to decay eventually to zero. For sustained sinusoidal oscillations, some active
element such as a controlled source or op amp must replenish the lost energy. Note that these
active elements require a dc power supply for their operation. Ultimately the dc power supply
replenishes the power losses due to various resistances in the circuit.
E XA M PLE 9 .1 0 . Figure 9.15 shows a Wien bridge oscillator constructed with an op amp as the
active clement. Find the condition on the circuit parameters R^,
and C for sustained sinusoidal
oscillation, and the frequency' o f oscillation.

r
R.
(b)
FIGURE 9.15 (a) Wien bridge oscillator, (b) Equivalent circuit.
S o l u t io n
From the principles described in Chapter 4, the non-inverting amplifier enclosed in the dashed box of
Figure 9.15a is equivalent to a voltage-controlled voltage source with a gain equal to {2Rj- + RJ}IRr= 3.
(See Chapter 4.) Replacing the dashed lx)x with this equivalent yields the simplified circuit o f Figure
9.15b. Using the simplified circuit, the first task is to derive die differential equation model of the circuit.
Step 1. Write a single-loop equation.
(9 .4 0 a )
413
To eliminate rhe integral, we differentiate again to obtain
dt
(9.40b)
7?1 dt
/?,C
Step 2. Express /^j in terms o f V2 - By inspection o f Figure 9.15b we obser\'e that

,
V2
(9.41a)
^c\ ~ ^ ----- '---dt
/?,
dt^
l<2
and thus, differentiating again,

(9.41b)
dt
Step 3. Substitute equations 9.41 into equation 9.40b. Substituting as indicated yields
dt~
/?2 dt
R^C ^
dt
Rj y
- ^ =0
/?, dt
(9.42a)
Grouping terms and dividing by C produces
d Vf
1 dv->
1 dv-y
- + ----------- ^ + ---------- ^ +
dt~
Ro_C dt
/?,C dt
R^R^C-
R^C dt
which simplifies to
dvo
d-V2
dt~
RjC
Vo
(9.42b)
R\C}
Step 4 . Compute the characteristic equation and determine the conditions for sustained oscillations.
The resulting characteristic equation is
s~ + hs + c = s~ +
\ R ,C
= 0
R ,C )
(9.43)
For sustained sinusoidal oscillations to occur, the roots must be purely imaginary. Thus the coef
ficient o f s must be zero, i.e.,
b=
R2C
/?,c
/e|/?2C
=0
(9.44)
Thus the condition for sustained sinusoidal oscillations reduces to /?, = R-,.
Step 6. Find the frequency o f oscillation. Under the condition R^ = Rj, the roots o f the character
istic equation are
/?,C
We conclude that the frequenq^ o f oscillation (in rad/sec) is
(Oo =
(9 .4 5 )
/?,C
A14
> R^, then ^ > 0 and the unforced response is

an exponentially decreasing sinusoid. On the other hand, if R^ < Rj, then b <0, and the unforced
An examination o f equation 9.44 shows that if
response is an exponentially growing sinusoid. For the oscillations to start, the value of/?| should
be designed to be slightly smaller than Rj- Then the value for b in equation 9.58 will be negative,
producing an exponentially growing sinusoidal response. If all circuit parameters are truly con
stant, the amplitude o f oscillation would theoretically grow to infinit)'. In real oscillator circuits,
such growth is limited to a finite amplitude by saturation effects or nonlinearities that clamp the
response when the voltage swing grows large. The resulting waveform then only approximates a
pure sine wave. The analysis o f this nonlinear effect is beyond the scope o f this book. However,
the next example illustrates the growing oscillation when
< /?2 and also shows the effect o f sat
uration to produce an approximate sinusoidal oscillation.
EXA M PLE 9 .1 1 . The circuit o f Figure 9.16a is a B2 Spice schematic for the Wein bridge oscilla
tor o f Figure 9.15. The op amp is a 741 with
= 0. Observe that
= 10
= 15 V. Suppose that
(0) = 10 m V and
= 9.5 kQ. According to the analysis o f Example 9.10, the out
put voltage labeled IVout should be a growing sinusoid. The output response o f Figure 9.16b
shows this growth and the saturation effects induced by the op amp. The waveform is not a pure
sinusoid due to these saturation effects. Also note that the frequency o f oscillation is approximately
16 Hz, which is consistent with equation 9.45, i.e..
=
2k
C2
R1
(a)
2 k ^R^R2 C
16.3 Hz
15
Example 9.11 Oscillator-Transient-4
Time(s)
(b)
FIGURE 9.16 (a) Schematic diagram of Wein bridge oscillator, (b) Voltage response showing grow
ing oscillation clamped at 15 V due to saturation effects of op amp.
An alternative approach to initiating oscillations and simultaneously limiting amplitude is to use

a temperature-sensitive resistor, R^, with a positive temperature coefFicient. Any incandescent
lamp is an example o f a temperature-sensitive resistor. For small voltages the temperature o f an
incandescent lamp is lower than for larger voltages because the dissipated power is lower. Hence
the lamp temperature (and thus its resistance) increases with increasing voltage. In the case o f our
oscillator, we have a desired output voltage swing. The nominal value o f /?, is designed to be
slightly less than R-y when the output voltage swing is bclow' a pre-specified voltage less than
This causes a growing oscillation. As the voltage swing increases, the temperature o f R^ and thus
its resistance increase. When the resistance o f /?, reaches Rj, the amplitude will settle (stabilize) at
the pre-specified voltage swing, at least theoretically. If /?, happens to increase beyond R^, a decay
ing sinusoid wîll result, decreasing the temperature and hence the resistance o f R^. Should the
amplitude o f oscillation decrease for any reason, /?j will decrease, causing a growing sinusoid.
Although the resistance o f R^ may dither about /?2>
amplitude o f oscillation will nevertheless
restore itself to the equilibrium level. In practice this equilibrium level only approximates the spec
ified value due to imperfections in the circuit parameter values. The resulting waveform is almost
a pure sinusoid.
4 16
6. SUMMARY
'I'his chapter has explored the differential equation modeling and response computation of^ sec
ond-order linear circuits having either no input or constant input excitation. Such second-order
circuits contain at least two dynamic elements, either an LC, CC, or LL combination. Secondorder circuits may also contain active elements such as op amps. In contrast to first-order circuits,
second-order linear circuits allow for the possibilit)^ of damped and undamped sinusoidal oscilla
tions.
Analysis o f second-order linear circuits has two phases. Pha.se 1 entails the formulation o f the sec
ond-order differential equation circuit model. For simple I C parallel RLC, or series RLC, the cir
cuit model can be found by inspection.
Phase 2 o f the development centers on the solution o f the second-order differential equation
model o f the circuit. I'he first step here is to compute the (quadratic) characteristic equation and
then solve for the two roots. The roots o f the characteristic equation determine the t}'pe o f
response. The three t)'pes o f roots for a quadratic real distinct, real identical, and complex
specify the three response types of overdamped, critically damped, and underdamped, respective
ly. These three types o f responses characterize all second-order linear differential equation models,
be they o f electrical circuits, mechanical systems, or electro-mechanical systems.
Since sinusoidal waveforms are germane to many electrical systems, this chapter presented an
oscillator circuit that generates a sinusoidal waveform. O f the many types of oscillator circuits, we
chose one containing an RC circuit built around an op amp, avoiding the use o f an inductor.
41

Characteristic equation: for a linear circuit described by a second-order difTerentia! equation
= j{t), the algebraic equation
,v"(f) +
+ r = 0 is called its characteris
tic equation.
Characteristic roots: roots o f the characteristic equation, also called the natural frequencies o f the
linear circuit.
Critically damped circuit: a second-order linear circuit having characteristic roots that are real
and identical. The source-free response o f such a circuit has a non-oscillatory waveform,
but is on the verge o f becoming oscillator)'.
Damped oscillation frequency: in an underdamped second-order linear circuit, the source-free
respon.se has the form [Kc^^' cos(o)y + ()}. The angular frequency
is the damped o.scil-
lation frequenc)', which is the magnitude o f the imaginary part o f the characteristic roots.
Homogeneous differential equation: a differential equation in which there are no forcing terms.
For example, x"{t) + bx\t) + cx{t) = 0.
Natural frequencies: the characteristic roots.
Oscillator circuit: an electronic circuit designed to produce sinusoidal voltage or current wave
forms.
Overdamped circuit: a second-order linear circuit having a characteristic equation whose charac
teristic roots are real and distinct.
Scaled sum o f waveforms: let
...
be a set of waveforms. A scaled sum of these wave
forms is an expression o f the form /r) =
+ ... +
for real (possibly complex)
scalars ^/j, ... ,

Second-order linear circuit: a circuit whose input-output relationship may be expressed by a second-order differential equation o f the form .v"(/') +
+ cx{t) = /(r).
Source fi-ee: there are no independent sources, or all independent sources have zero values.
Step function: a function equal to zero for r < 0 and equal to 1 for r > 0.
Step response: the response o f a circuit to a step function input when all capacitor voltages and
inductor currents are initially zero.
Undamped circuit: a second-order linear circuit where the characteristic roots are purely imagi
nary and the unforced response is purely sinusoidal.
Underdamped circuit: a second-order linear circuit whose characteristic roots are complex with
nonzero real part.
' T h e notations
and KO arc used interchangeably in the literature to den ote the first derivative ol v{t).
418
PROBLEMS
where all coefficients are real (but not necessar

ily positive).
(a)
Prove that x(r) = 0 at some r = T, 0 <
TH EO RY RELATED
r < 00, only if
1. In section 2, the solution to the undriven LC

circuit is given by v^^t) = K cos(cof + 0) V,
Observe that
and /f, have oppo
site signs.
(b)
Prove that the x(^) vs. t curve has at

most one zero crossing for r > 0.
(c)
cos(tt)0 = - W cos(0)/)
State the necessary and sufficient condi

tions on the coefficients
>
and Sj for the presence o f one zero

and
7 sin((0 / ) = - 0 )
crossing.
sin(coO
dt^
5. The voltage or current in a second-order
imply that A cos(tijr) and B sin(tof) are both

solutions to the differential equation
eral form
x(,) = (A:, + K 2 t)e-'>^'
d^\>C
dr
source-free critically damped circuit has the gen
where all parameters are real, but not necessar
LC
ily positive. Prove that x(/) = 0 at some t = T <
By superposition, then, v^t) = A cos(ojr) + B

sin((or) V. Show that for a given A and B, there
exist K and 0 such that Vf\t) = A cos(tof) + B
6. (a)
underdamped
v^t) =
and
satisfies the differential equation
x"{t) + 2}^'{t) + h~x{t) = 0

and AT, are arbitrar)^ constants,
4. The voltage or current in a second-order

source-free overdamped circuit has the general
[10 cos(9950^) + 1.005
the voltage waveform before the peak
W^{t) is constant.
x{t) = (A'l + K^j)e-^
the
+ 5.7)
How many cycles o f ringing occur in
energy stored in C and L for the circuit o f
3. By direct substitution, show that
and
sin(9950/)] = 10.05^" ^' cos(9950f
2. Find the expressions o f the instantaneous
form
is
response is given by
the two solution forms are equivalent.
where
Consider case 2 o f Example 9.5. The

circuit
sin(cof) = K cos(tor + 0). One concludes that
Figure 9.1b. Show that the sum o f
and K j have opposite signs.
COif and only if
value drops from its largest value o f

10.05 to 10.05/f = 0.3 6 8 X 10.05?
(b)
Suppose the characteristic polynomial

is written as
response
+ 2as +
form
x{t)
+ to^ with
=
cos(w y + 0). Prove that for the under

damped case, the circuit will ring for
N = iy ijiln o) cycles before the ampli

tude decreases to Me o i its initial
value.
7. When a dc voltage o f
volts is applied to a
series LC circuit with no initial stored energy,

the voltage across the capacitor reaches a peak
value o f twice the source voltage. To investigate
x{t) =
419
this phenomenon, consider the circuit o f Figure
UNDRIVEN RLC PROBLEMS
P9.7 where switch S is closcd at r = 0. Assume
9. The switch S in the circuit o f Figure P9.9 has
the inductor current and the capacitor voltage
been closed for a long time and is opened at / =
are zero at / = 0.
0. Express vj,t) and /^(r) for r > 0 in terms o f

the literals R, L, C, and /q. Also compute the
t =0
initial stored energy in the inductor and capac

S ^
itor.
Figure P9.7
Show that for f > 0
Figure P9.9
1
.V l
]
c
',
and
10. In the circuit o f Figure P9.10, suppose
Vi{t) = 10 V, /? = 10 n , C = 0.4 mF, Z = 0.25

H, and the switch opens at /^= 0.
yc(t) = Vo 1 -
COS
(a)
Compute
, //(0~), and
(b)
Compute the energ)^ stored in the
(c)
Using only energy considerations,
Jlc\
inductor and the capacitor at / = 0.
8. The circuit in Figure P9.8 is a dual o f the
compute the maximum value o f
previous problem.
f > 0.
(d)
Find the analytical expression for

and verify the maximum value
of
Figure P9.8
computed in part (c).
i,(t)
,v (t)
v(t) I
The switch S is opened at r = 0. Both the induc

tor current and the capacitor voltage are zero at
r = 0. Show that for / > 0
Figure P9.10
11.
vU)
Consider the circuit o f Figure P 9 .ll in
which i r jf ) = -2 0 u (-r ) V, R = 10 Q, C = 0.4

mF, and L = 0.25 H.
and
' l ( 0 = / o 1 - cos
(a)
Compute
/^(O"), and
(b)
;,( 0 -) .
Compute the energ}' stored in the
inductor and the capacitor at r= 0.
(c)

Plot using MATLAB for 0 < / <
200 msec.
420
(d)

Plot using MATLAB for 0 < r < 200
msec.
i,(t)
v(t)
(a) (t>0)
Figure P 9.11
12. Reconsider the circuit o f Figure P 9 .ll
= 50(-/) V, R = 25
Q, C = 0.8 mF, and Z, = 1 H. Repeat Problem
under the conditions
11.
13. Reconsider the circuit o f Figure P 9 .11 under

the conditions v-^{t) = -5 0 (-r) V, R = 25 O., C
Figure P9.15
= 0.8 mF, and Z, = 2 H. Repeat Problem 11.

16. Figure P9.16 shows an overdamped source14. For the circuit o f Figure P9.14, suppose C
= 0.8 mF and determine L so that the fre
quency o f the sinusoidal response, for r > 0, is
2500 rad/sec. Now find y(;;(0~),
free circuit in which R = 0.4 Q., L = 0.5 H, and

C = 0.5 F.
(a)
/(^0),
If VfÔ) = - 2 V and /^(O) = 2.5 A, find
V(it) for r > 0. Use MATLAB or the
/Ô"^), and Vf^t) for r > 0.
equivalent to plot the v^^t) waveform

and verify that there is no zero cross
t=0/\
L
25 0
_L
100 0
ing.
25 mV
(b)
If vÔ) = 2 V and /^(O) = 2.5 A, find
v^t) for t> 0 . Use MATLAB or equiv

alent to plot the V(^t) waveform and
verify that there is no zero crossing.
Figure P9.14
15. Consider the circuit o f Figure P9.15, in
which
= 2 QV, R^ = 2 Cl, Rq = %Q., R = 2 Q,
Z. = 0.5 H, and C = 62.5 mF.
(a)
Figure P9.15 a shows a source-free

parallel RLC circuit whose past history
Figure P9.16
is depicted by Figure P9.15b where

the switch S has been at position A for
17. In Figure P9.17, the switch S has been at
a long time before moving to position
position A for a long time and is moved to posi
B at r = 0. Find
y^Ô'*^), /^0~),
/Ô"^), and V(^t) for r > 0. Plot V(^t)

(b)
tion B at t = 0. Suppose
0.5 a
using MATLAB for 0 < f < 1.25 sec.

Find i^{t) for / > 0. Plot V(^t) using
(a)
MATLAB forO < r< 1.25 sec.
(b)
= 100 mV, R =
/. = 1 H, and C = 0.01 F.
Find
and
for f > 0.
Plot for 0 < / < 50 sec.

Find i^{t) for r > 0. Plot for 0 < / < 50
sec.
421
20. The voltage across the capacitor for the cir

cuit of Figure P9.20a is given by Figure P9.20b.
Suppose R = 25 k2.
v J t)
(a)
Using the plot, estimate the values of
L and C
(b)
Figure P9.17
O '
18. For the circuit of Figure P 9.18,
current.
Clearly show and explain all steps in your cal
0,5u{-t) A.
(a)
culations. Hint: You might assume a general
I f /? = 20 Q , Z = 1 H, and C = 8 mF,
find and plot V(^t) and
(b)
Now estimate the value o f the initial

capacitor voltage and initial inductor
response form v^^t) = Ae~^^ cos(co^ + 0) V and
for f > 0.
then use the plot to estimate a and O)^; what is
Repeat part (a) for R = 22.5
the relation of a and Ci)^ in the characteristic

polynomial of a parallel RLQ
ijt)|
Figure P 9.18
19. In Figure P 9.19. ^^<0") = 25 V, /^(O-) = 50
mA, /? = 2 kQ, Z = 0.1 H, and C = 0.1 ;/E The
switch closes at f = 0.
(a)
Compute y(;;(0^), ^^(0"^), and /^(O^).
(b)
Compute /^(/),
(c)
Plot i^(t) and Vjit) forO < t< 1 msec. |
(d)
and V({t).
|
|
Find the energy stored in the circuit ^

over the interval [0, 0.2 msec], i.e., in
the capacitor and the inductor over
this interval. Is this energy positive or
negative? Also compute the energy
dissipated in the resistor over this same
(b)
interval. The sum o f the energy dissi

pated in the resistors, the energy
Figure P9.20
stored in the capacitor, and that stored

in the inductor should equal zero.
21. In the circuit of Figure P 9.21, R = AO., C
Why?
= 6.2 5 mF, and

i,(t)
/YYV<
L
for f > 0.
(a)
(b)
t=0
Figure P9.19
Compute the value of L.

Find the value of the initial condi
tions, ^^(0*) and
(c)
'n^
= (20 - \200t)e~^^* mV
Find ijit) for t> 0.
422
i jt )
/ Y W
Figure P9.21
Figure P9.24
22. For the circuit o f Figure P9.22, L = 0.04 H,

C = 2.5 mF, and /^, = 10 LI.
(a)
Find the value o f R (in ohms) that
25. Figure P9.25 shows a critically damped
makes the circuit o f Figure P9.22 crit
source-free circuit.
ically damped.
(b)
0.1 H
Given this value o f R, suppose V(^Q) =

160
mV
and
/^(O)
= -6 0
mA.
Compute V(\t).
(c)
40 0
0.25 mF
Determine the first time at which the
+
sVc(t)
capacitor voltage is zero. Plot your

result using MA'FLAB or the equiva
F'igure P9.25
lent to veri5' your calculation.

(a)
If vÔ) = - 5 V and /^(O) = 1 A, find

;^(f) for r > 0. Determine the differen
,_ r Y Y \ =
tial equation for the circuit. Let^ = ij

and use y = dsolve(D2y + 400*D y +
40e3*y = 0,y(0) = l,D y(0) = -3 5 0 ) in
MATLAB to verify your answer.
Figure P9.22
(b)
If i;JO ) = 5 V and /^(O) = 1 A. find
V(^t) for t> 0. Plot the V(^t) waveform

23. Reconsider the circuit o f Figure P9.22 with
and verify that there is one zero cross
/?///?, = 0.8
ing. Again use the dsolve command in
(a)
and A = 0.04 H.
MATLAB to verif}' your calculations.
Find the value o f C so that the circuit is

underdamped with to^= 100 rad/sec.
(b)
(c)
Suppose
= 160 mV and /^(O) =
- 3 0 mA. Compute
Determine the first time at which the
26. The capacitance voltage o f a source-free

parallel RLC circuit, with R = 2.4 LX has the
form
V(^t) =
capacitor voltage is zero. Plot your

result using MATLAB or its equiva
(a)
lent to verify your calculation.
(b)
cos(8r + B)
Find the values o f L and C.

If y^Ô) = 10 and /^(O) = 0, determine
A and 0.
24. In the circuit o f Figure P9.24, /? = 20 Q and
(c)
If the values o f L and C remain

unchanged, find the value o f R for the
//:.(0 = 500f?"'^' sin (l0 V 3 /)
circuit to be critically damped, and the
mA for r > 0. Find the proper values o f L and
general
C to produce this response. Now find t'^Ô'^),
response
under this condition.
Then determine the source-free
response when V(^0) = 10 and /^(O) =
y'cÔ"^), and vît) for r > 0.
0.
form
of
the
source-free
423
27. Almost 75% of fliilures in circuits, i.e., situa

tions where a circuit dramatioilly fails to perform
as designed, are due to opens and shorts o f indi
vidual circuit elements. Heating, c\'cling a circuit
on and off, etc., cause degradation in the circuit
parameters, resistances, capacitances, inductances,
etc. that often precipitates the short or open situa
tion. For example, the material inside a resistor
might become brittle over a period of time and
finally crumble, leaving a break in the circuit. On
the other hand, the material might congeal or
become dense, decreasing the resistance. In the
problems below you are to determine the length of
time it takes for a circuit to move from an over
damped behavior to an imderdamped behavior
due to changes in the resistor characteristic as a
fiinction of time.
(a)
For the parallel RLC circuit in Figure
P9.27a, suppose R = Rq + exp(/ - 5) H
where f > 0 constitutes time in years.
Determine the time f for which the cir
cuit changes its behavior from over
damped to underdamped.
(b)
For the series RLC circuit of Figure
P9.27b, the resistor satisfies R =
+
exp(r - 5)] n , where again t is time in
years. Here it is presumed that the circuit
is pan o f a larger piece o f electronic appa
ratus, such as a TV, which is used exten
sively over a period o f years. The tiine t'
then is not connected with the response
time o f the circuit. Determine the time
f ' for which the circuit changes its behav
ior from overdamped to underdamped.
DRIVEN SERIES AND

PARALLEL /?/.C CIRCUITS
28. (Initial condition calculation) For the cir
//(O'*'),
cuit shown in Figure P9.28, find

;^ 0 ^ ), and
in two steps:
Sff/> I. Find V(^0~) and /^(O") by open-circuit

ing C and short-circuiting L.
Step 2. Construct a resistive circuit valid at t =
O'* and from this find
29.
and
Consider the circuit o f Figure P9.29 in
which
= -10w(-/) + 20(f) V, 7? = 20 Q,
C = 0.1 niF, and L = 0.25 H.

(a)
Compute
/^(0), and
(b)
Compute the energ)'^ stored in the

inductor and the capacitor at r = 0.
(c)
Find the analytical expressions for the

zero-input, zero-state, and complete
responses for
Identif}' the tran
sient and steady-state responses. Plot
V(^t) using MATLAB over [0, 40

msec].
(d)
Find the analytical expressions for the

zero-input, zero-state, and complete
R=
R = 0.8 Q
responses for i/{t) . Plot ij{t) using

MATLAB over [0, 40 msec].
i,(t)
ijO
L= 1H
R = i s n
C = 1/36
(b)
Figure P9.27 Parallel and series RLC circuits subjccc to resistor degradation over time.
30. Reconsider the circuit o f Figure P9.29 under

the conditions
+ 25u{t) V, /? = 25
Q, C= 0.8 mF, and Z, = 2 H. Repeat Problem 29

but construct plots over [0, 400 ms].
424
31. Reconsider the circuit of Figure P9.29

under tiie conditions
= -50(-r) +25u{t)
V, 7? = 25 ii, C = 0.8 mF, and L = 0.2 mH.
Repeat Problem 29 but construct plots over [0,
40 ms].
32. In Figure P9.32 v-J^t) =-250(-r) +750(f)
mV, ^ =0.5 a , I = 1 H, and C= 0.01 F.
(a) Find /^(O^),
and the zeroinput, zero-state, and complete
responses of vît) for r > 0. Identify
the steady-state and transient parts of
the complete response. Plot in MATLAB for 0 < ^< 50 msec.
(b) Find the zero-input, zero-state, and
complete responses of i^{t) for t > 0.
Plot in M ATLAB for 0 < r < 50 msec.
ijt)
fY Y \ h ^
36. In Figure P9.36,
Figure P9.32
T he switch closes at f =0.
(a)
(b)
(c)
ComputeyÔ*), y^(O^), and /^(O^).

Compute the zero-input, zero-state,
and complete responses of
and v^t). Identify the transient and
steady-state parts of the complete
response.
Plot iît) and Vjit) for 0 < r < 1 msec.
Find the energy stored in the circuit
over the interval [0, 0.2 msec], i.e., in
the capacitor and the inductor over
this interval. Is this energy positive or
negative? Also compute the energy
dissipated in the resistor over this same
interval.
i,(t)
33. Repeat Problem 32 for /? =40 Q and v-JJ)

= -0.5tt(-/) + 2(/) V. Plots in M ATLAB
should be for 0 < f < 800 msec.
34. Repeat Problem 32 for /? = 50 Q and and
v.^{t) =-0.5(-/) - 2u{t) V. Plots in M ATLAB
should be for 0 < ? < I sec.
=-1 0 (-^ ) + 40(/)
mA, /? = 4 k n , Z = 0.1 H, and C = 0.1 pF .
(d)
vJt)
Figure P9.35
fY Y \ k
L
+
,Vc(t)
(t)
Figure P9.36
35. For the circuit of Figure P9.35,
=
-0.5M + 2u{t) A.
37. For the circuit of Figure P9.37, R^= 5 0.^
(a) I f /? = 2 Q, Z = 1 H, and C = 8 mF, = 2 0 a , C = 2.5 mF, L = 0.25 H, and v j t ) =
find and plot the zero-input, zero- 20(^) - 20u(t-7) V, where T = 0.25 sec.
state, and complete responses of v^{t)
(a) Find the zero-input, zero-state, and
and ij\t) for / > 0. Identify the tran
complete responses of v^t) for f > 0.
sient and steady-state parts of the
Plot the complete response for 0 < r <
complete response.
0.25 sec.
(b) Repeat part (a) for R= 22.5 Q.
(b)
li')
Find the zero-input, zero-state, and
complete responses o f
200 n .
for t > 0.
= 50
and Rj =
Plot the complete response for 0 < r <

4 3 . Consider the RLC circuit in Figure P 9.43
0.25 sec.
where /?^, = 60 Q, R^2 =

= 5m F
(a)
Q, I = 4 H, and C
= 100(-f) mA and r^2(^) =

2 0u (-t) V. Find the response,
(b)
for
t> 0. Plot for 0 < f < 1 sec.

/y,(r) = 100u(-t) + 500u(t) mA and
1/^2 ^) = 20u(-t) V. Find the response,
for ^ > 0. Plot for 0 < / < 1 sec.
= 50 Q, /?j = 200
Q, C = 0.05 mF, L = 0.5 H , and v j t ) =
- 5 0 (- r ) + 50k(?) - 50(r - 7) V, where T =
0.08 sec. However, only plot the complete
I V ,(t)
responses for 0 < f < 200 msec.

39. Repeat Problem 37 for R^= 100 Q ,
100 Q , C = 0.25 mF, L = 2.5 H , and v-J^t) =
Figure P9.43
-5 0 tt(-f) + 50(f) - 50(^ - 7) V, where T =

100 msec. However, only plot the complete
44. Repeat Problem 4 3 , except find y^(r), ^> 0.
responses for 0 < ^ < 300 msec.

45. Consider the RLC circuit in Figure P9.43
40. Consider the RLC circuit o f Figure P 9.40 in
where Rî = 20 Q, R^2 = 20 Q, L = 0.4 H, C =
which R^ = 100 Q, /?! = 4 0 0 Q , C = 0.125 mF,
4 mF.
L = 0.2 H, and v.{t) = 50u{t) - 50u U - 7) V,

where T = 0.025 sec, vîQr) - - 2 5 V, and
(a)
/^(0 ) = 10 mA.
(a)
(b)
(r) = -u (-t) A and

Find the response,
= 40(-^) V.
for t> 0 . Plot
for 0 < f < 0.8 sec.
Find the zero-state, zero-input, and

complete responses of V(it) for r > 0.
(b)
(f) =
+ 2u{f) - 2uU - 0.4) A
and v^2 (^) = 4 0 ( -/) V. Find the
Plot for 0 < r < 60 msec.
response,
Find the zero-state, zero-input, and
r < 0.8 sec.
for ^ > 0. Plot for 0 <
complete responses of vît) for f > 0.

Plot for 0 < r < 60 msec.
46. Repeat Problem 4 5 , except find
t> 0 .
47. Consider the RLC circuit in Figure P9.43
where /?j| = 16
R^2 = 32 Q, I = 0.4
H, and C = 4 mF.
w
(a)
'n- ^
f > 0. Plot for 0 < r < 0.3 sec.

(b)
Vw>
i M = - 0 .5 ( - / ) A and v,y{t) =
24(-^) V. Find the response,
for
41. Repeat Problem 40 for R^= 140

= 360 a
and Rj
/^,(r) = -0.5u{-t) + 0.5(f) A and v^2 ^t)

= 24u(-t) V. Find the response,
for f > 0. Plot for 0 < f < 0.3 sec.
426
Chapter 9 * Sccond Order Linear Circuits
r> 0.
49. riic current source with /^|(^) = 5//(-/) mA

and the voltage source
= 10 V in Figure
P9.49 drive the circuit in which /? = 1 k^2, C =

0.5 //F, and Z. = 0.184 H.
(a)
Find
(b)
.,(0 ^ ).
Compute
t'(-(0"),
/^(O"^),
/^(0^),
and
for r > 0. Plot for 0 < r
PSEUDO SERIES AND

PARALLEL /?/.C CIRCUITS
/?! = 4 a
= 4 a /. = 5/12 H, C = 25 mF
(a)
Find the roots o f the characteristic
equation. C H EC K : - 8 , - 1 2
(b)
/j^(0"), î(0*), and
< 5 msec.
(c)
Compute
Compute
for / > 0. Plot for 0 < f
ages and currents, draw the equivalent

circuit valid for t > 0. Check your
for r > 0. Plot for 0 < r
answer for
< 5 msec.
(e)
Compute
for r > 0 . Plot for
for t >
0. Hint: After finding the initial volt
< 5 msec.
(d)
If v.^,{t) = -2Qu{-t) V, find 1/(^0-),
using the dsolve
command in MATLAB. Plot
<t
for
0 < r < 2 sec using MATLAB or the
< 5 msec.
equivalent. Be sure you properly label

your plot.
(0
If
(d)
If y,./r) = -2 0 u (-t) + 20u(r) -2 0 u (t- 1)
= - 2 0 (- f) + 20u(r) V, find
for / > 0.
V, find and plot v^^t) for 0 < r < 2 sec.

Figure P9.49
50. Repeat Problem 49 for the new source cur
rent /jj(r) =
0 < r <_10 ms.
-5 ti(r-
0.005) mA. Plot for
51. The switch in the circuit o f Figure P9.51 is

in position A for a long time and moves to posi
tion B at / = 0. Find the voltage
for t > 0
when L equals (a) 0.625 H, (b) 0.4 H, and (c)

0.2 H.
53. Repeat Problem 52, but find i^{t) and v^{t)

without differentiating
= 4 Cl, Rj = 4 Q,
1 = 0.2 H, a n d C = 0 .2 F .
10V
55. Reconsider the circuit o f Figure P9.52 in

which /?, = 600 Q, /?2 = \ 2 0 n ,L = 2 H, C = 1
Figure P9.51
mF and v j t ) = -72u{-r) + 72u(t) -72u{t - 1).

Find the response
for r > 0 as follows:
(a)
Find
(b)
Find z^f;(0'*') and i^CO"*").
(c)
and
Draw the equivalent circuit valid at O'*"

and find ^'/(O'^) and Z(;;;^0'').
^1/
(d)
(e)
Find die characteristic equation and
60. Reconsider the circuit of Figure P9.57.
natural frequencies o f the circuit.
Suppose /?j = 80 Q, /?2 = 4 0

^ = 2 H, and
C = 0.625 mF with
= -1 5 0 tt(-/) + 150(/)
mA.
Determine the general form o f the

response, ij{t)y valid for 0 < ? < 1.
Determine all coefficients in the gen
(a)
Find V(iO~) and V(^0*).
eral form o f the response.
(b)
Find ii(0~) and /^(O'*).
(g)
Determine the form o f the response,
(c)
Find t//^(0+) and /c(0+).
for f > 1. Find the response.
(d)
(h)
Plot the response
(f)
for / > 0 using
MATLAB or the equivalent.
Find the characteristic equation and

natural frequencies o f the circuit.
(e)
Find the response,
(f)
Find the response,
t> 0.
t> 0.
56. Repeat Problem 55, except find Vf4,t).

6 1 . Consider the circuit of Figure P9.61
which
= 2 2, /?2 = 2 Q, Z = 0 .4 H, and C =
0.1 E
(a)
(b)
Suppose /?j = 80 Q, R2 = 40 Q., L = 2 H, and
C = 0.6 2 5 mF with
= 300u(f) mA, Vq =
50 V, and /?3 = 20 Q.
Find V(iO~) and y^Ô'*).
Find the roots of the characteristic
(a)
equation.
(b)
Find i^(0-) and z^(0+).
(c)
Find v^{0*) and /c (0 *).
(d)
= -2u{-t) A. find t;c<0-),
If
^>
0, Hint: After finding the initial volt
natural frequencies o f the circuit for t

>0.
ages and currents, draw the equivalent

circuit valid for ^ > 0. Check your
(e)
Find the response, v^t), t> 0.
answer for ij{t) using the dsolve
(f)
Find the response, ijXt), t> 0.
command in MATLAB. Plot
for
0 < ^ < 1 sec using MATLAB or equiv

alent. Be sure you properly label your
plot.
(c)
If
= -2u{-t) + 2(?) A, find
for ^ > 0.
(d)
If
= -2u{-t) + lu{t) - l u { t - 1) A,
find and plot
for ^ > 0.
62. The switch in the RLC circuit of Figure
'S->
ijt)
d)
P 9.62 opens at ^ = 0 after having been closed

+
> F
L
for a long time. The purpose o f this problem is

to find the complete response o f the capacitor
voltage, ? > 0. Suppose
Figure P9.57
V^
(b)
(c)
59. Repeat Problem 57 with
W
(t) = 1 A, v^2 (^) = 20
V, C = 4 mF, and I = 0.625 H.

(a)
Using a dc analysis, find the initial
Using a dc analysis, find the final value

o f the capacitor voltage, Vfôo).
= 80 2, R2
2 0 a , L = 10 H , and C = 1/240 E
conditions /^(O") and V(^0~).

Find
and
(d)
\2H
(e)
compute its roots. Given tiie roots,
tion with ij{t) as the unknown. Observe that
write down the general form o f the
the derivative o f
response
hand side.
is present on the right-
Solve for the unknown coefficients in

the response form o f part (d) and write
ANSW'F.R: /"(/) + i\t) + i{t) =
down the exact expression for v^^t)

valid for ^ > 0.
40 fi
40 0
Figure P9.62
GENERAL SECOND-ORDER
CIRCUITS
P9.65.
(a)
roots o f the characteristic equation.
63. In the operational amplifier circuit o f

Figure P9.63 is a second-order circuit. Suppose
Write a second-order differential equa

tion with
the unknown. Find the
(b)
If
V, find V(^t) for r > 0.
= R^ = 50 k n .
(a )
Determine the values o f C, and Cj

that produce a characteristic equation
having natural frequencies at - 5 and
-
(b)
10 .
Adjust the value o f /?j so that for a

step function input voltage, the value
o f the output voltage for large t is for
all practical purposes is 5.
(c)
when v^{t) = 2u{t)
P9.66.
and all capacitor voltages are zero at t
(a)
Compute
=
0.
Write a second-order differential equa

tion with
the unknown. Find the
roots o f the characteristic equation.

Then find V(^t) when
(b)
Repeat part (a), for when

unknown.
Figure P9.63 Cascade of leaky integrator circuits

having a second-ordcr response.
P9.64. Write a second-order differential equa-
= u{t) A.
is the
429
67. The second-order circuit shown in Figure
nearly 1 V in about 5 nsec, it behaves approxi
P 9.67 contains two capacitors.
mately as an open circuit. The 0.1 pF capacitor
(a)
(b)
Find the second-order differential
is then charged up with a time constant of
equation with V(^ as the unknown.
about 0.1 msec. As the larger capacitor is
Give the values o f s.
charged up, the output across the smaller one
If vci (0) = 2 V, and ^ ^ (0 ) = 4 V, find

V(^{t) for t> 0.
decreases toward zero.
0.5 0
2F
0.5 n
0.5 0
2F
Figure P9.67
Figure P9.69
68. In the circuit of Figure P 9.68, the voltagecontrolled voltage source has a gain A > 0 . Find
the ranges of
damped,
(b)
for the circuit to be (a) over

underdamped,
(c)
critically
damped, and (d) undamped.
70. Find the value o f the negative resistance
-R for the circuit shown in Figure P 9.70

required to generate sinusoidal oscillations with
constant amplitude.
-R .
V,F
120
Figure P9.70
69. The second-order circuit shown in Figure

's-/'
O '
P 9.69 is of the overdamped type. Find the step

response, i.e., the expression for Vg{t) for ^ > 0,
when the input is vi{t) = u{t) V, and the capac
itors are initially uncharged. Roughly sketch
the waveform of Vpit). Verify your sketch by
doing a SPICE simulation o f the circuit.
so
Remark: The waveform vît) consists of a very

fest rise toward 1 V, and then a relatively slow
exponential decrease toward 0 V. This can be

explained using the first-order RC circuit prop
erties studied in Chapter 8. During the first few

microseconds, the 0.1 jiF capacitor behaves
almost as a short circuit, and the 1 nF capacitor

is charged with a time constant of about 10
nsec. After the smaller capacitor is charged to
'<N->
o
71. Refer to the Wien bridge oscillator of

Example 9.11. Suppose the op amp has a satu
ration voltage 15 V. C = 1 jiF and /?2 = 500 Cl.
is a temperature-sensitive resistor whose
resistance is a fixnction of the amplitude of the
sinusoidal current passing through
. The fol
lowing relationships are given:

?l(f) =
sin(C0f + 0)
/?, = 500 + 1 0 0 ( 7 ^ -0 .0 1 )
(a)
Find the frequency of oscillation (iHn-
(b)
Find the amplitude of voltage at the

op amp output terminal (with respect
to ground).
(c)
Suppose /?j is a fixed resistance of 4 9 0

t^ci(O) = 100 mV, and V(^{0) = 0 in
the Wein bridge oscillator circuit.
Perform a SPICE simulation. Does
the circuit behave as expected?
430
72. In the Wein bridge oscillator example o f

this chapter, let
= R-, = \ kH, Rr= 10
and
C=0.1pF.
(a)
Determine the frequency o f oscillation

in Hz.
(b)
If
(0) = 5 V and v^{0) = 0 V, find
v^{t) for r > 0.

(c)
Use any circuit
simulation
(e.g.,
SPICE) to verify the waveform o f v^{t)

in part (b).
73. In the Wein bridge oscillator o f Figure
9.15a, /?2 = 1 k n . ^/=10
and C = 0.1 pR
The lamp resistance R^ is a function o f the peak
value o f the sinusoidal current as shown Figure
P9.73.
(a)
Determine the frequency o f oscilla

tion.
(b )
Determine the amplitude o f the sinu

soidal waveform v^{t).
Sinusoidal Steady State

Analysis by Phasor Methods
A HIGH-ACCURACY PRESSURE SENSOR APPLICATION
The control o f high-performance jet engines requires highly accurate pressure measurements, with
errors less than one-tenth o f 1% o f a full-range measurement, over a wide range o f temperatures,
- 6 5 to 200 F. The pressure range may be as low as 20 psia or as high as 650 psia. In jet (turbine)
engine applications, knowing pressure and temperature allows one to compute the mass (volume)
air flow, a critical aspect o f an engines performance. A pressure sensor is also a critical component
in the regulation o f aircraft cabin pressure. Such a sensor is depicted here along with a functional
block diagram o f its operation. A diaphragm consisting o f t\\'o fused quartz plates separated by a
vaciumi has a capacitance that changes as a function o f pressure and temperature. This quart/
capacitive diaphragm is an element in a bridge circuit. It is this bridge circuit, in conjunction with
detailed knowledge o f the characteristics o f a pair o f quartz capacitors over the required operating
range o f pressure and temperature, that enables accurate pressure measurements.
Functional block diagram courtcsy o f AllicdSignal Aerospace Com pany.
432
Chapter 10 Sinusoidal Steady State Analysis by Phasor Methods
Because o f the small capacitances, on the order o f picofarads, associated with the quartz
(diaphragm) capacitor, the bridge circuit is driven by an ac source and is called an ac bridge.
Driving the bridge by an ac source moves its analysis outside the realm of the dc and step response
techniques studied in earlier chapters. New methods o f analysis, such as phasor analysis, are nec
essary. Phasor methods, the primary focus o f this chapter, allow us to analrze capacitive and induc
tive circuits excited by sinusoidal (ac) inputs. In particular, phasor techniques permit us to anal)'ze
an ac bridge circuit. Although the analysis ot the pressure sensor shown here is beyond the scope
o f this text, the chapter will end with a simplified pressure sensor circuit based on the one shown.
CHAPTER OBJECTIVES
1.
Review and elaborate on the basic arithmetic and essential properties o f complex num
bers pertinent to sinusoidal steady-state analysis o f circuits.
2.
3.
4.
Develop two complementary techniques for computing the response o f simple RL, RC,
and RLC circuits excited by sinusoidal inputs and modeled by differential equations.
Define the notion o f a (complex) phasor for representing sinusoidal currents and voltages
in a circuit.
Using the notion ol phasor, introduce the notions of impedance, admittance, and a gen
eralized Ohms law lor two-terminal circuit elements having phasor currents and voltages.
5.
Utilizing the methods o f nodal and loop analysis and the nerwork theorems o f Chapters
5 and 6, analyze passive and op amp circuits by the phasor method.
6.
Introduce the notion o f frequency response for linear circuits, i.e., investigate the behav
ior o f a circuit driven by a sinusoid as its frequency ranges over a given band.
SECTION HEADINGS
1.
2.
3.
Introduction
Brief Review o f Complex Numbers
Naive Technique for Computing the Sinusoidal Steady State
4.
5.
Complex Exponential Forcing Functions in Sinusoidal Steady-State Computation

Phasor Representations of Sinusoidal Signals
6.
7.
8.
9.
10.
11.
12.
Elementary Impedance Concepts: Phasor Relationships for Resistors, Inductors,

and Capacitors
Phasor Impedance and Admittance
Steady-State Circuit Analysis Using Phasors
Introduction to the Notion o f Frequency Response
Nodal Analysis o f a Pressure-Sensing Device
Summary
Terms and Concepts
13.
Problems
Chapter 10 vSinusoicial Steady State Analysis by Phaser Methods
-133
1. INTRO DUCTION
Perhaps you have experienced the bouncing motion of a car with broken shock absorbers or
watched the (mechanical) oscillations o f a sw'inging pendulum. These motions reflect the sinu
soidal and damped sinusoidal oscillations in
circuits with conjugate poles o f the characteris
tic equation, as detailed in Chapter 9. In this chapter we allow sources with sinusoidal forcing
functions (such as such as
cos(ov + 0) or perhaps
sin(d)/ + f))), which almost always result in
sinusoidal responses regardless o f the root locations o f the characteristic equation. A sinusoidal
voltage source models the voltage from the ubiquitous wall outlet.
If one hooks up an oscilloscope to measure a voltage in a linear circuit driven by sources with sinu
soidal values, the voltage may not look sinusoidal at first. However, if the circuit is stable, after a
sufficiently long period o f time the screen o f the scope will trace our a sinusoidal waveform. (Here
stable means that any zero-input response consists o f decaying exponentials or exponentially
decreasing sinusoids.) The eventual sinusoidal behavior is not immediately apparent because at
startup, stable circuits exhibit a transient response. Transient means that the circuit response is
transitioning for example, from an initial voltage or current value to another constant value.
Flickering lights during a thunderstorm illustrate the phenomenon o f transient behavior: light
ning may have struck a transmission line or pole, causing the power system to waver briefly from
its nominal behavior.
Because sinusoidal excitations and sinusoidal responses are so common, their study falls under the
heading o f sinusoidal steady-state (SSS) analysis. Here sinusoidal means that source excita
tions have the form
we take
+ 0) =
cos((or + 0) or K sin((0/ + 0). For consistency with traditional approaches,
cos(tor + 0) as the general input excitation, as shown in Figure 10.1, because
sin(to/
cos(tof + 0 - 7t/2). Steady state mean that all transient behavior o f the stable circuit has
died out, i.e., decayed to zero. Observe that every sinusoidal waveform is periodic with angular
argument (to/ + 0). In terms o f angle, each cycle o f the waveform traverses 2 n radians. In terms of
time, each cycle covers a time interval o f T = 27r/o) seconds, called the period o f the waveform.
The number o f cycles contained in 1 second is called the frequency o f the sinusoidal waveform
and is denoted by / T h e unit for/is the herrz (Hz), meaning cycles per second. The quantity
OJ, which specifies the variation o f the angular argument (tor + 0) in 1 second, is called the angu
lar frequency o f the wâveform. The unit o f (O is radians per second (rad/sec). From these defini
tio n s,/ = 1/7'= ti)/27i and o) = 2 k /
Stable circuits driven by sinusoidal excitations produce sinusoidal voltages and currents, as illus
trated in Figure 10.1. The output excitation in Figure 10.1 has the general form
to distinguish it from the input excitation,
cos((0/ + (j))
cos((0/ + 0). Because o f linearity, the circuit can
change only the magnitude o f the input sinusoid

is changed to K^^) and the phase angle o f the
input sinusoid (0 is changed to (j)) while ensuring that the angular frequenq^ to remains the same.
For nonlinear circuits, to can and usually does change.
Chapter 10 Sinusoidal Steady State Analysis by Phaser Methods
434
Input Excitation: Vjcos(cot + 0)
CO
FlGURl'l 10.1 Graphical illustration ot steady-state sinusoidal linear circuit behavior. (Kând
could just as well be
and
or any combination thereof.) Note that 0 and (}) are often different and
that (0 is the same for both input and output excitations.

In Figure 10.1 the steady-state (voltage) response is
been a current response,
+
cos(a)r + ({)). Alternatively this could have
cos(a)r + (j)). Such waveforms have the equivalent structure A cos(wr)
sin(oj/), deducible from trigonometric identities,

cos({or + (}))=
cos((j)) cos(tor) -
= A cos(a)f) + B sin(tor)
where A =
obtains
cos(({)) and B =
sin(i|)) sin(cor)
( 10 . 1)
sin((})). Conversely, by summing the squares oi'A and B, one
V, =
4 3 ^
n ------ ^
+ B-
(10.2a)
By taking the inverse tangent o f the ratio o f - B and A, one obtains
(10.2b)
In using equation 10.2b it is important to adjust the resulting angle for the proper quadrant o f
the complex plane. Equations 10.1 and 10.2 turn out to be useful in developing a conceptually
simple, although naive, technique for computing the steady-state response using a differential
equation model o f the circuit, as explained in section 3.
Sinusoidal steady-state (SSS) analysis o f circuits draws its importance from several areas. The
analysis o f power systems normally occurs in the steady state where voltages and currents are sinu
soidal. Music is a rhythmic blend o f different notes. Mathematically, a musical (voltage) signal can
be decomposed into a sum o f sinusoidal voltages o f different frequencies. The analysis o f a sound
system typically builds around the steady-state behavior o f the microphone, the amplifier, and the
loudspeakers driven by sinusoidal excitations whose frequency varies from around 40 Hz to 20
kHz. Indeed, almost any form o f speech or music transmission requires an understanding o f
steady-state circuit behavior. There are many other areas o f applicability.
This chapter will introduce three techniques for computing the SSS response. The first two, some
what naive, approaches map out a natural motivation and path to the third, ver)' powerful tech
nique o f phasor analysis. Phasor analysis builds on the arithmetic o f complex numbers and the
basic circuit principles studied thus far. To set the stage for phasor analysis, section 2 reviews the
necessary basics o f complex number arithmetic. O f course, the student is assumed to have stud
ied complex numbers in high school and in prerequisite calculus courses.
2. BRIEF REVIEW OF C O M P LEX NUMBERS

Let
= a + jb be an arbitrary complex number, where ^
. The real number a is the real
part o f 2 j, denoted hy a = Re[z,]. The real number b is the im aginary part o f z^, denoted by b
= Im[ 2 j]. It is simple to verify that
= Re ^1 ^ i L i L
2
and
b = Im -1
where
2y
a - jb is the complex conjugate* o f Zy The magnitude or modulus o f 2 ,, denoted by
|2 ,|, satisfies
|z|P =
+ Ip-
36
The number Z| =
+ jb is said to be represented in rectangular coordinates. Another represen
tation o f 2 p called polar form or polar coordinates, follows from the simple geometry illustrated
in Figure 10.2.
FIG UIIE 10.2 Diagram showing relationship between polar and

rectangular coordinates of a complcx niunber.
In Figure 10.2, the number Zj can be thought o f as a vector o f length p = >/ + b" = |zj|, which
makes an angle 0 = tan"*(^/<z) with the horizontal in the counterclockwise direction. (In comput
ing tzn~^{bln) it is important to adjust the angle (principal part) to be in the proper quadrant o f
the complex plane.) Hence Zj = /? + jb is completely specified by its magnitude p and angle 0 i.e.,
z, = /z +
= p cos(0) + yp sin(0) = p[cos(0) + j sin(0)] = peJ^ = pZ.0
where p Z. 0 is a shorthand notation for
and
t/ = cos(0) + ;s in ( 0 )
(10.3)
is the famous Euler identity. The Euler identity can be demonstrated by writing the Taylor series
for
and recognizing it as the sum of the Taylor series for cos(0) added to j times the Taylor
series for sin(0). Note that the symbol Z. has two meanings, depending on the context o f its use:
(1) L z means angle o f the complex number z, and (2)pZ. 0 means the complex number whose
magnitude is p and whose angle is 0. The properties o f the exponential immediately imply that
(10.4)
Exercises. 1. Compute the polar coordinates o f

AN SW ERS:
2. Let z = 6
/T
= -1 - j and Zj = \ + j-
r ; /4S'
where ti/6 has units o f radians; for example, n rad equals 180. Find the real
and imaginary parts o f z.

ANSW ER; 2 = S. 1962 + /3
43*
3. Show by direct computation that since eJ^\ = cos(0j) + j sin(0j) and eJî = cos( 02) + j sin(f)2),
then
= cos(0,+ 0-,) + j sin(0,+ (),) =
With these simple definitions, the product oF t^vo complex numbers z-^ = a +jb =
and z-y
can be found using rectangular coordinates as
z^zf = {a +jb) {c + jd) = a c - bd +j{bc + ad)

or using polar coordinates as
= p ,p 2 cos(G,+ 62 ) + yp ,p 2 sin( 0 ,+ dj)

which in shorthand notation is
~ P 1P 2
^2^
EXA M PLE 10.1. Suppose z^ = 3 -jA = 5 Z .-5 3 .1 3 and ^2 = 8 + ;6 = 1 0 ^ 3 6 .8 7 . Then

,Z 2 = (24 + 24) + y(18 - 32) = 48 - ; 1 4
Equivalently,
z ,z 2 = 5 ^ - 53.130 X 1 0 ^ 3 6 .8 7 = 50 Z ( - 5 3 .1 3 + 3 6 .8 7 )
=
+36.87)
= 50 cos(16.26) - ; 5 0 sin (l6.26 0 ) = 48 - ; 1 4
Exercise. Let Zj = 2 + j l and z^= - 2 + j 6 .

(a)
Compute the polar form o f 2 , and z ,.
(b)
Compute ZjZ-, in rectangular coordinates.
(c)
Compute ZjZ2 in polar coordinates.
A N SW ERS: 2.8284^'/''^", 6.3246^>-/>"''^- *-^, - 1 6 + ;8 , 17.8885^->''^-^-'''-'
Similarly, in rectangular coordinates the arithmetic for the division o f two complex numbers is
Cl _ a + jb _ {a + jb ){c - jd ) _ {a + jb ){c - jd )
Z2
c + jd
ic + j d ) { c - j d )
{ac + bd) + j ( b c - a d )
~
^2
c- + d
c -+ d ~
-i38
In polar coordinates the calculation is more straightfor\vard;
il- =
=-
Z2
) = -Hi-cos(6, - 0 2 ) + 7 sin(6| - G j)
P2
P2
In our shorthand notation,
Exercise. Let Zj = 2 + j l and Zj =

+76 .
(a) Compute z^tzj in rectangular coordinates.
(b)
Compute z'^lzj in polar coordinates.
A N SW ERS:
(J.i
/ 0 . 4 , 0 . 4 4 7 2 ^ ' / "*^^'
O f particular concern in this chapter are equations involving mixed representations o f complex
numbers. For example, suppose an unknown complex number z =
satisfies the equation
{a + jb) = f + j d
Then dividing through by
+ jb yields
c + jd
a + jh
Since Vis the magnitude o f the complex number on the right-hand side o f the equal sign, it fol
lows that
+
_ V c-+ t/ ~
V ^+/r
Here we have used the fact that a complex number that is the ratio o f two other complex num
bers has a magnitude equal to the ratio o f the magnitudes. To determine the angle 0, one uses the
property that 0 equals the angle o f the complex number in the numerator minus the angle o f the
complex number in the denominator,
-1
0 = Z (c + jd ) - Z {a + jb )= tan'
tan
Exercise. Let z, = 2 - 2j and

ANSW ER: 0.471
- 6 8 .4 7 "
-ta n
= 5.5 + jlA . Find V and 0 when
-I
= 2 ,/z,.
EXA M PLE 10.2.

Suppose v(t) =
^ a n d
-5VeJ^ + j6VeJ^ Find V, (j), and v{t).

S o l u t io n
Factoring Ve-^ out to the left and dividing by (-8 + j6) yields
- 8 + j6
Hence. K= 2, (}) = - 9 8 .1 3 and
v{t) =
j)] = 2 cos(to/- 98.13) V
Sometimes a function v{t) is a complex number for each t, such as if{t) =
and v{t) will
satisfy some specific algebraic or differential equation. W hen this is the case, it is possible to use
the differential equation to find values for V and (j). The next two examples illustrate this strategy.
E X A M PLE 10.3. Suppose the function
v{t)
satisfies the differential equation
dr
+ 2 - + 2v = 10e-'< '
dl
Find the values o f A and (}) if O) is known to be 2 rad/sec.

S o l u t io n
Since the function v{t) must satisfy the differential equation, the first step is to substitute into the
differential equation. Substituting
1^ into the differential equation and taking appropriate
derivatives yields
The
term, which is always nonzero, cancels on both sides, leaving
AeJ^ [2 - co2 + ;2co] =

Since cu = 2, one can equate magnitudes and angles to obtain
2 - ( 0 " + y2 (0
-2 + y4
440
Exercise. Repeat the above example if (o = 3 and

ANSWERS: A = 2. (|) = 13J.4 "
is changed to
' HI.2)
The techniques of circuit analysis in this chapter will often require complex number arithmetic.
The voltages and currents o f practical interest are always real. T he complex arithmetic is a short
cut to computing real voltages and currents. The real quantities are obtained by taking the real
part o f the complex number or complex function. The various manipulations depend on some
general properties related to the real part o f complex numbers.
Property 10.1. Re[Z| + Zj\ = Re[z,] + Rel22j.
This property has a particularly nice application to summing trigonometric waveforms. Let v^{t)
= cos(oj/ + 55) and r-y(t) = 10 sin(d)/ - 3 0 ) = 10 cosCd)/ - 120). Note that a - 9 0 shift con
verts the sine to a cosine. Hence,
?'l(r) +
= cos(o)r + 55) + 10 co s(to f- 120)

= Reld-^^^'^^
+ Ret 10^-^^^'^^
= Rek>("*'"
= Relt-''^^^
by the Euler identity, equation 10.3

by Property 10.1
+ lOf*"/ *' )] by equation 10.4 and then factoring
to the left
= Re{f-^'^'1(0.5736 +/).8192) + (-5 - 78 .66 )]} after conversion to reaangular form

= Rel^>'(-4 .4 2 6 - ; 7 . 8 4 l ) ] = Re[9<'>("^
after simplification and con
version back to polar form

= 9 cos(o)/ - 119.4) after taking the real part.
This sequence o f manipulations shows that the magnitude and phase o f two cosines at the same
frequency O) can be represented by distinct complex numbers. One can then add the complex
numbers and determine the magnitude and angle o f a third cosine equal to the sum o f the origi
nal two cosines. This presents a shortcut for adding two cosines together.
Property 10.2 {proportioiuilityproperty). R elazj] = aRefZ]] for all real scalars a .
Properties I and 2 taken together imply
R e [a , 2 , +
= a jR e [ 2 j] + a,Re[z-,]
which is a linearity property for complex numbers with multiplication by real scalars. 'I'he next
property, which underpins the techniques of this chapter, defines how differentiation can be inter
changed with the operation Re[-].
Property 10.3 {differentiation property). Lex. A = ê^^. Then

= Re - ( / t e - '" ' )
= Re
dt
Exercise. Find Re
(10 +
cit
ANSXXTR: 1118 cosdOO ; + 116.57
Our fourth property tells us the conditions for the equality o f two complex-valued time functions.
Property 10.4. For all possibly comple.x numbers A and B, Re[/lf-^^'^T =
only
for all t if and
A = B.
Taken together, the preceding properties imply a fifth, very important propert)'. Here note that a
complex exponential is sometimes referred to as a complex sinusoid.
Property 10.5. I he sum of any number o f (1) complex exponentials, say AjeJ^'^, or (2) derivatives
o f any order o f complex exponentials o f the same frequency co, or (3) indefinite integrals of any
order o f a complex exponential o f the same frequency O), is a complex exponential o f the same fre
quency (1).
This property is another foundation stone on which the phasor analysis o f this chapter builds.
Table 10.1 summari?,es the properties o f complex numbers.
TABLE 10.1 Summar}' of Properties o f Complex Numbers
cos((jL)r + 0) = A cos(cor) +
B sin(co^)
= V-4- +
, ({) = tan"'
Euler identit)'
eJ^ = cos(0) + j sin(0)
Real part o f sum
Re[z, + z-y] = Re[zj] + Re[z2l
Proportional it)' property
RelfXZ]] = a Re[2 i] for all real scalars a

RelcXjZ, + a ,Z 2 ] = ct]
Linearity property
Differentiation propert)'
Equality propert)'
dt
dt^
+ 2
= Re
Re[y4f-^^^T = Re[5^^'^T for all t if and only i f A = B

Sum o f complex exponentials AêJ^^ or deriva
Single-frequency propert)'
tives, or their indefinite integrals o f any order is a

complex exponential o f the same frequency co
442
3. NAIVE TECH N IQ U E FOR C O M PU TIN G THE SINUSOIDAL

STEADY STATE
Property 10.5 o f the previous section suggests a technique for computing the SSS response o f a
circuit. The technique uses a differential equation model of an RL, RQ or RLC circuit as devel
oped in Chapters 8 and 9. In contrast to the dc sources in those chapters, suppose the source exci
tations have the form
cos(cor+0). In addition we assume that the zero-input response consists
o f (eventually) decaying exponentials or (eventually) exponentially decaying sinusoids to ensure

that there is a valid sinusoidal steady state. Thus the form o f a first-order circuit differential equa
tion model with a sinusoidal excitation is
^/v(/)
cit
+ ax{t) = K^cos(co/+ 0 )
(10.5a)
or in the second-order case,
df
(10.5b)
dt
where x{t) is a desired voltage or current, such as
V(\t).
Property 10.5 guarantees that the sum o f any number o f cosines or derivatives o f any order o f
cosines o f the same frequenc)' OJ is a cosine o f the same frequency O). Hence, the circuit response
x{t) in equations 10.5 has a steady-state cosine form o f frequency O). Further, the scaled sum ofx(f)
and its derivatives on the left-hand side o f each differential equation 10.5 must equal V^cos(o)/+0),
the input excitation. This also implies that the steady-state circuit response, x{t), is a cosine o f the
same frequency as the input, but not necessarily the same magnitude or phase. We conclude that
^(^) =
cos(co/+(j)) = A cos(tor) + B sin(to/). The SSS response is then specified upon finding A
and B. The following example illustrates this calculation.
E X A M PLE 10.4. Let the source excitation to the circuit o f Figure 10.3 be
Compute the SSS response /jr(r).
______________
i^lt)
i,
F I G U R E 10.3 Parallel
RL circuit
for Exam ple 10.4.
= /^cos(o)^).
So
443
lu tio n
Step 1. Determine the differential equation model o f the circuit. From KCL applied to the top node
o f the circuit, i jt ) = i^(t) + //_(/) Sincc the resistor and inductor voltages coincide, the t/-i rela
tionship o f the inductor implies that the inductor current satisfies the difTerential equation
( 10.6)
which has the form o f equation 10.5a.
Step 2. Determine the form o f the response. Since the input is a cosine wave, the SSS response will
have the sinusoidal form
ij{t) = A cos(cor) + B sin(oj/)
(10.7)
Step 3. Substitute the form o f the response {equation 10.7) into the differential equation 10.6.
Inserting equation 10.7 into the differential equation 10.6 and evaluating the derivatives yields
L '
cos(co/) = f/\cos(o)/)+ Z?.sin(o)r)l + /4cos(co/)+ fisin(( 0 /)
(It
= - 0) /\sin(coO + to
RA
RB
I-t
cos((or) + cos(co/) + sin(co/)
Step 4. Group like terms and solve for A and B. Grouping like terms leads to
r
Bco H A ----- 1^ cos(o)r) + B -A ii)

L
L ',
.L
]
sin(coO = 0
(10.8)
To determine the coefficients A and B, we evaluate equation 10.8 at two distinct time instants.
Since equation 10.8 must hold at every instant o f time, it must hold at / = 0; i.e., at /= 0,
=0
or, equivalendy,
(10.9a)
In addition, equation 10.8 must hold for t = 7r/(2to), in which case
-COA + - 5 = 0
( 1 0 .9 b )
444
Solving equations 10.9 simultaneously for A and B yields

/?-/,
= ------f ,
CO/?L/
B=
R~ + L^(0-'
/?- + Lco
Step 5. Determine the steady-state response. Since A and B are known,

.
Il (0 = ------ ^
R~ + L~0)~
cos(o3 0 +
R~ +
In the more common alternative form o f i^{t) =
O^RLL
^
SI n(co /)
cos((or + ({)) as per equation 10.2a,
where
= - tan *
(10.10b)
is adjusted to reflect the proper quadrant o f the complex plane.
This example has illustrated a procedure for finding the SSS response o f a circuit. Step 1 is to sub
stitute an assumed sinusoidal response form, such asy4cos(cor) + Bs\n{a)t), having unspecified con
stants A and B, into the differential equation and evaluate all derivatives. Step 2 is to group like
terms, and step 3 is to compute the constants A and B. After finding A and B, one computes the
magnitude,
and phase (j) o f the cosine
cos(oj^ + (|)) via equations 10.2.
The next section offers an alternative approach. Using complex excitation signals o f the form
yêJ^Mt +0)^
computes
and (j) by a more direct route.
4. C O M P LEX EXPON EN TIAL FORCING FUNCTIONS IN

SIN USOIDAL STEADY-STATE COM PUTATION
Complex exponential forcing functions are simply complex exponential input excitations o f the
properties o f complex numbers in section 2, we can
form
or ^;(o)/ + 0) PfQji-j
replace the input excitation
cos{wt + 0) and the assumed circuit response
cos(to^) + j5sin(co^) with their complex counterparts
and
cos(cor + cj)) = /I
respectively, with
out any penalt)'. To recover the actual real-valued responses, we simply take the real parts o f the
complex quantities. Again this is justified by properties 10.1 through 10.5. This process o f sub
stitution and subsequent taking o f real parts actually simplifies the calculations developed in sec
tion 3, because of the simple differential and multiplicative properties o f the exponential function.
The following example illustrates a more efficient calculation o f the steady-state response using
complex exponentials.
vn
EX A M P L E 10.5. For the series RC circuir o f Figure 10.4 let vît) = K^cos(cof). Compute the
steady-state response
FIG URE 10.4. Series /?Ccircuit for Example 10.5.

Step 1. Construct the dijferential equation o f the circuit. Writing a loop equation and substituting
for i(^t) yields
( 10. 11)
at
Step 2. Substitute complex forms o f the input and response into the differential equation. If v^{t) were
to be equal to the complex exponential
if vj^t) =
then the response would be
= V}cos(cor) (as is the case), then V(^t) =
o f complex numbers. Hence, for the moment, let us set vj^t) =
However,
from the properties

and agree that
appropriate real parts.

Substituting the complex expressions into the circuit differential equation 10.11 yields
After canceling the
terms, factoring V,e^^ out to the left, and dividing through by {jixiRC +
1), we obtain
1 + jOdRC
Step 3. Determine the magtiitude
( 10 . 12)
and the angle (}). Equating magnitudes on both sides o f equa
tion 10.12 yields
Vl + ( o V c '
(10.13a)
and equating angles yields

(j) = - tan '(to/?Q
(10.13b)
Step 4. Determine the steady-state response. Using equations 10.13 the desired response is comput
ed by taking real parts:
4 i6
\'cU) = Re
(10.14)
Ks
cos[co/ -tan ha^RC)]
Vl + (0/e-C^
In deriving the relationship 10.12 from the ciifterential equation 10.11, we utilized a complex exponen
tial ftinction as the circuit input. A complex exponentid input is not a signal that cm be generated in the
laborator)'. Nc-vertheless, it is often used in advanced circuit theor)' to simplift' the derivation of many
important results, as was done in the preceding example. If one does not mind a more lengthy derivation,
then the s;imc result (equations 10.12 through 10.14) am be obtained without the flaitious complex
exjxjnentid excitation. For example, let the voltage source in Figure 10.4 represent a reiil signal source
v^{t) = V^cosim) = R e[K / > T

Then the steady-state response has the form
v^t) =
cosiMt + (j)) =
Substituting these expressions into the differential equation 10.11 yields
/ec ( rc v...e
+ Re
= Re
Making use of properties 10.2 and 10.3, move the position o f the operator Re[] outside the first
term to obtain
Re
+ Re
= Re
Evaluating the derivative and using property 10.2 (linearity) produces
(10.15)
By property 10.4, equation 10.15 holds if and only if
+ <l>) =
This is precisely the equation following equation 10.11 that leads to equations 10.12, 10.13, and
finally 10.14.
As we can see, the use o f complex exponentials does indeed lead to a more direct calculation of
the SSS response. However, this method and the method o f section 3 require a difFerential equa-
44 /
tion model o f the circuit. For circuits with multiple sources, dependent sources, and many inter
connections o f circuit elements, finding the differential equation model is ohen a nontrivial task.
In the next section we eliminate the need to find a differential equation model o f the circuit by
introducing the phasor concept.
5. PHASOR REPRESENTATIONS OF SIN USOIDAL SIGNALS

Recall
is shorthand iov
= /lcos(({)) +yy4sin((})). If the frequency co is known, then the
complex number AL^ completely determines the complex exponential

known, then AL^ completely specifies A cos(ojr +({)) =
In turn, if to is
This means that the com
plex number A/L^ can represent a sinu.soidal function A cos(ojf + (})), whenever (O is known.
Complex number representations that denote sinusoidal signals at a fixed frequency are called
phasors. A phasor vo\iz^c or current will be denoted by a boldface capital letter. A typical voltage
phasor is V =
and a ty'pical current phasor is I =
Por example, the current i{t) =
25cos((Of + 4 5 ) has the phasor representation I = 25Z-45. The voltage v{t) = - 1 5 sin(tor + 3 0 )
= 15 cos(w^ + 120) has the phasor representation V = 15/-120.
As all voltages and currents satisfy KVL and KCL, respectively, one might expect phasor voltages
and currents to do likewise. This is not patendy clear. The following simple example demonstrates
why this is true for KCL.
Consider the circuit node drawn in Figure 10.5.
FIGURE 10.5 Single node having four incident branches.

From KCL it follows that
^4(^) =
- ijit) +
= 10 cos(tor) - 5.043 cos(cor + 7 .5 2 ) + 8 cos(cof- 9 0 )
Using trigonometric identities or property^ 10.1 to combine terms on the right-hand side leads to
iît) =
1 0 c o s (tO f- 6 0 )
4 <S
Chapter 10 Sinusoidal Steady State Analysis by Pliasor Methods
For the corresponding phasors to satisR' KCL, it must follow that
10Z-60O =
= I, - I 2 + I3
The right-hand side o f this equation requires that
I, - I2 + I3 = lOZQO - 5.043Z.7.52O + 8Z.-9()0

= 10 - (5 + >0 .66 ) + ( - ; 8 ) = 5 - 78.66 = 10/1-60 = I 4
Thus the phasors (which have both a real and an imaginar)' part) satisfy KCL. KCL is satisfied
because i^{t) = /j(^) - i-,{t) + /^(r) implies
= Re[10^>n - Re[ 5. 043^>>( ' ^' ^] + Re(8f>i< ^ -^0-)]

= Re[( 10 - 5.043<?^'7-5-" +
(10.16)
for all t. By property 10.4, equation 10.16 holds if and only if
IOZ- 6OO = 10 - 5.043e^7-52 +

In phasor notation this stipulates that
It is the properties o f complex numbers and the fact that an equation is true for all t that guaran
tee that phasors satisfy KCL. Although not general, the argument is sufficient for our present ped
agogical purpose. A similar argument implies that phasor voltages satisf}' KVL, as illustrated by
the following example.
EXA M PLE 10.6. Determine the voltage across the resistor in the circuit of Figure 10.6 using the
phasor concept.
Vj(t) = 19.68 sin(a)t
152.8)
V3(t) = 4 .2 1 5 cos(cot + 71.61 )
+ . --------------
v,(t) = 20co s(cot + 53.13)
F K i U R E 10.6 Resistive circuit w ith three sourccs.
44^)
o l u t io n
Firsr note that 19.68 sinltDr+l 52.8) = 19.68 cos(co/+152.890") = 19.68 cos(cor+62.8). From KVL,
Vfît) = v^{t) - Vjit) + v^{t)

Since voltage phasors must satisfy KVL,
V/^ = V, - V 2 + 3 = 20 z i 5 3 . 1 3 0 - 19.68 z^62.80 + 4.215 ^ 7 1 . 6
Changing to rectangular coordinates and adding yields
= 12 + y i6 - (9 + y i7 .5 ) + 1.33 + j4 = 4.3 3 + jl.5
Equivalently,
= 5Z-30 V, and
= Rc[5r>(" ' " ^0")] = 5 co${o)t + 300) V
Exercise. In Figure 10.6, suppose yj(r) = 10 cos(cor) V, v-^ (/) = 10 co s((o r- 0.5tc) V, and
V3(/)= I 0 V 2 cos(cOf - 0.25ti ) . Find the phasorVând then
ANSWKR:
= 20 - /20.
vrU) = 2 ()V : cos(O)/ - 4 5 " ) \
Given that phasor voltages and currents satisfy K\^L and KCL, respectively, it is possible to devel
op phasor O hm s law-like relationships for resistors, capacitors, and inductors operating in the
SSS. This would allow us to do SSS circuit analysis with techniques similar to resistive dc analy
sis. The next section takes up this thread by introducing the notion of (phasor) impedance.
6. ELEMENTARY IM PEDANCE CO N CEPTS: PHASOR

RELATIONSHIPS FOR RESISTORS, INDUCTORS, AND
CAPACITORS
Ohms law-like relationships do exist for resistors, capacitors, and inductors operating in the SSS.
The constraint, operating in the SSS, suggests that any Ohm s law-like relationship should be
dependent on the sinusoidal frequency.
The first objective o f this section is to derive three Ohms law-like relationships, one each for the
resistor, the capacitor, and the inductor. The relationships each take the form V = Z(/to)I, where V
is a phasor voltage, I is a pha.sor current, and Z(/to) is called the impedance o f the device: Zy^(;tij) for
a resistor,
for a capacitor, and Z/(/co) for an inductor. The fact that the phasor voltage V is a
function Z(/co) times a phasor current I indicates a clear kinship with Ohms law for resistors. Indeed
the unit o f impedance is the ohm because it is the ratio of phasor voltage to phasor current. The
impedance Z(/cij) explicitly shows that the relationship is potentially frequency dependent.
450
The derivation o f these elementary impedance concepts will build on the assumption that all volt
ages and currents are complex sinusoids o f the same frequency represented by complex phasors.
This is permissible because real sinusoids can be recovered from complex sinusoids simply by tak
ing the real part. To this end consider the resistive circuit o f Figure 10.7a.
V (t) =
i,(t) =
j
0j(u)t + 6)
(b)
(a)
FIGURE 10.7 (a) Resistive circuit driven by complcx current,

(b) Equivalent phasor representation of the circuit in (a).
From Ohms law,
vjiit) =
In terms o f the phasors
= RIj^
+ 0)
ej^\ this relationship reduces to
= 1^-^^^ and
= /?
= Zpfjia) \j^
(10.17)
and Z^(yco) = R\s the impedance o f the resistor defined by equa
where
= Rlj^ Z.0. If
tion 10.17. Ideally the resistor impedance is independent o f frequency. Thus

=
cos(u)r + 0) = Kq[
then Vjît) = Rlj^ cos(co/ + 0) =
This phasor
relationship restates Ohms law for complex excitations. The distinctiveness o f phasors comes with
their application to inductors and capacitors.
Now consider the inductor circuits o f Figure 10.8. Assume the circuit o f Figure 10.8a is in the
steady state.
i jt ) =
Remainder
Remalnder
of circuit
of circuit
(a)
-I-
V = jcoL 1^
(b)
FIGURE 10.8 (a) Inductor having complex exponential voltage and current, (b)
Phasor relationship of (a).
/is:
The complex current and voltage associated with the inductor are, respectively, /^(f) =
and vît) =
Substituting these expressions into the defining equation for an inductor
yields
j = jcoz.
Canceling out
on both sides yields
the relationship is
and
( 10. 18)
in which case the inductor Impedance is derived as Z^(/co) = p iL . The inductor impedance clear
ly depends on the value ol the radian frequency CO. Specifically, if U) = 0, then the impedance o f
the inductor is 0, i.e., in SSS the inductor looks like a short circuit to dc excitations. If co = oo, the
impedance is infinite, i.e., in the steady state the inductor looks like an open circuit to signals o f
very high frequency.
Equation 10.18 exhibits a frequency-dependent Ohms law relationship for the inductor. From the
properties o f the product o f two complex numbers, the polar form o f the voltage phasor is
\ l = (yco)I^ = (03/./,) ^ ( 0 + 9 0 )
Hence if
/^(/) =
= /, cos(tor + 0) A
then
p^(t) = Re[/wZ/^f>('^' ^ 0)] = RelcoL/ê^^^"
" -^0)] = coL/^ cos(tof + 0 + 9 0 ) V
From this relationship one sees that the voltage phase leads the current phase bv 9 0 . Equivalently,
one can say that the current lags the voltage by 9 0. This leading and lagging takes on a more con
crete meaning when one views phasors as vectors in the complex plane, as per Figure 10.9, which
shows that the voltage phasor o f the inductor always leads the current phasor by 9 0 .
452
The capacitor has a similar impedance relationship, derived as follows. Assume the circuit o f
Figure 10.1 Oa is in the steady state.
C'
Remainder
of
Remainder
circuit
of
circuit
(a)
J
(b)
FIGURF. 10.10 (a) Capacitor having complex exponential voltage and current.
(b) Phasor relationship o f (a).
The complex current and voltage associated with the capacitor are, respectively, i^^t) =
and V(^t) =
Substituting these expressions into the defining equation for a capacitor
yields
Vce 7(o)r+<!>)
Canceling out
on both sides yields
= I(-e^ and V (-=
this relationship becomes
I^=yo)C V ^
or, equivalently.
- i c - ^ cO ^ ^ )Ic
ycoC
(10.19)
Equation 10.19 defines the capacitor impedance as Z(^j(a) = l/(/a)Q. if co = 0, the impedance
o f the capacitor is infinite in magnitude. This means that in SSS the capacitor looks like an open
circuit to dc signals. On the other hand, if OJ = co, then the capacitor has zero impedance and looks
like a short circuit to large frequencies.
4 S3
Chapter 10 * Sinusoidal Steady State Analysis by Phaser Methods
Looking again at equation 10.19, observe that
Vr =
70) C I c
=-
t:
(oC
- 90)
(10.20)
Equation 10.20 has a vector interpretation in the complex plane, as shown in Figure 10.11
Imaginary axis
F ' l C l J R l i 10, 1 1
Diagram of capacitor voltage and current phasors where the voltage phasor lags the
current phasor by 90.
The diagram o f Figure 1 0 .1 1 indicates that the capacitor voltage lags the capacitor current phasor
by 9 0 or that the capacitor current leads the capacitor voltage by 9 0 , which is the opposite o f
the case for the inductor.
Exercises. 1. For the circuit o f Figure 10.12a, show that
ycoL ^
and that
V/
i r ( t ) = ^COS(Cl)/ + 0 - 90)
(OL
2. For the circuit o f Figure 10,12b, show that

I^ = ya)C V ^
and that
i(^t) =
+ ^ + 90)
iiCt)
ijt )
V ,(t) =
V ,(t) =
\J
\J
0 j((i)t + 0)
g j(( t + 0)
(a)
(b)
I-ICURE 10.12 (a) inductor driven by vohagc source, (b) Capacitor driven by voltage source.
Recall that resistance has a reciprocal counterpart, conductance. Likewise, impedance has a recip
rocal counterpart, admittance. Admittance has units o f siemens, S, as does conductance, fh e
admittance, denoted by l^yto), associated with an impedance, Z(/co), is defined by the inverse rela
tionship
K(./co) =
( 10.21)
ZOCO)
provided Z{j(M) is not equal to zero ever)^vhere. What this means is that the phasor i-v relation
ship o f a resistor, capacitor, and inductor satisfies an equation o f the form I = K(y(o)V. Hence, the
admittances o f the resistor, inductor, and capacitor are respectively given by
>"/eO'w) = - ,
r^(./co) =
./(oL
Kc-(yco) = ycoC
( 10.22)
The impedance and admittance relationships o f the resistor, capacitor, and inductor are summa
rized in table 10.2.
TABLE 10.2 Summary of Impedance and Admittance Relationships for Resistor,
Capacitor, and Inductor
Impedance
Admittance
2/?(y(0) - R
K^(;o)) =
yV-(./co) = ycoC
j(oC
JY Y V
Z Lijoi) = j(i)L
j(i)L
In the next section the notion o f impedance is applied to an arbitrar)' two-terminal network. This
generalization will allow us to consider the impedance and admittance o f interconnections o f
capacitors, inductors, resistors, and dependent sources.
7. PHASOR IM PEDANCE AND ADM ITTAN CE

For the resistor, the inductor, and the capacitor, the impedance equals the ratio o f the respective
phasor voltage to the phasor current. Analogously, the impedance o f any t\vo-terminal circuit, as
illustrated in Figure 10.13, is the ratio o f the phasor voltage to the phasor current, i.e..
Z S j^ ) = ^
= R + jX
(10.23a)
1 in
O+
Two
Terminal
Circuit
Z (j(o)orY,(j(o)
FIGURE 10.13 Two-terminal device with phasor voltage \ p h a s o r current 1^^^,
and input impcdance Zy^^(/co).
Because impedance is the ratio o f phasor voltage to phasor current, its unit is the ohm. Inverting
the relationship o f equation 10.23a defines the adm ittance o f a two-terminal device as the ratio
o f phasor current to phasor voltage, i.e..
I;
(10.23b)
Provided Z(yto) ^ 0 for all
cd,
in contrast to a short circuit, then
As an example, the impedance o f an inductor is jwL and its admittance is \/(J(.oL). Historically,
impedance and admittance were first defined as per equation 10.23. However, with the wide
spread use and utility o f the Laplace transform (Chapter 12) in the past several decades, imped
0 6
ance and admittance have become understood as much broader and more useful concepts than
the steady-state presumptions o f equation 10.23, as set forth in Chapter 13.
In general, admittances and impedances are rational functions with real coefficients o f the com
plex variable Ju). At each d) the impedance and the admittance are generally complex numbers.
Since a complex number has a real part and an imaginary part, we can further classify the real and
imaginary parts o f an impedance or an admittance. For an impedance Z(yw) the expression
lm[Z(/co)] = X is called the reactance o f the two-terminal element, while Re[Z(/(o)] = R refers to
its resistance. 1urther, for an admittance Vijto), Im[K(/(o)j = B is called the susceptance of the
two-terminal device whereas Re[)1[yoj)] = G is referred to as the con d u aan ce. These definitions
are summari/,cd in Table 10.3.
I'ABIJ-. 10.3 Summary Definitions of Various Terms
Admittance
Impedance
V/
Y( jo) ) = ^
Z{J(}^) = = R + jX
I in
V/
= G + jB
Resistance
Reactance
Conductance
Susceptance
R = Re[Z(yw)]
X = Im[Z(;to)l
6 = R e [n / o )]
im [K(>j)]
Using equations 10.23, one can compute the equivalent impedance

series, as in Figure 10.14a. Flere
= ^] +
o f two devices in
% Ohms law for impedances, V , = Z|(/to)I, and
V 2 = Zol/w) I ,. But I, = I 2 = I/- Mence,
I.e.,
în (
= ^ = Z, ( ./O)) + Z 2 (./(O)
*iit
(10.24)
This simple derivation has another consequence: given Zy^j(yco) = Z,(/to) + Z 2 (/co) and the fact that
i = 1, 2, a simple substitution yields the voltage division formula,
Z/O )
'
Z,(./(D) + Z2(7(0)
V;.,
(10.25)
Kquations 10.24 and 10.25 are consistent with our early development o f series and parallel resist
ance.
4V
Y,(jco)
V.
V.
V,
oYJjco)
(b)
FIGURE 10.14 (a) Two impcclanccs in series, (b) Two admittances in parallel.
Exercises. 1. Duplicate the derivation ol" equation 10.23 for three impedances in .series.
2. Derive a formula for voltage division when there are three impedances in series.
The admittance o f two devices in parallel, as sketched in Figure 10.14b, satisfies
v,
v,
V,
V,
since
^1 (./ ) = f
V|
Y2 ijc o ) = ^
Vo
we conclude that
(10.26)
Exercises. 1. Duplicate the derivation o f equation 10.26 for three admittances in parallel, i.e.,
show that )^y(/co) = Kj(/o)) +
2. Show that the equivalent impedance o f two devices, Z,(/co) and Z t(/co), in parallel is given by
(10.27)
3. Show that the equivalent admittance o f two devices, Kjlyco) and Y-,{ji.o), in series is given by
458
>U/co)>'2(yw)
K|aco) + K ,0 (o )
4. Show that the admittance o f two capacitors, Cj and C^, in series is yCO
(10.28)
C 1C 2
C\ + C-)
5. Show that the impedance o f two inductors, Z.j and Z.^, in parallel is /O)
Lt + Lo
Now the derivation o f equation 10.26 leads to a current division formula as follows. Since
= Kj(/co) +
and since
for /= 1 ,2 , one immediately obtains the current divi
sion formula,
I; =
'
n-(7(0)
K ,(7co)+ r2(y(o)
I,
(10.29)
Since devices represented by impedances or admittances must satisfy KVL and KCL in terms of their
phasor voltages and currents, and since each device so represented satisfies a generalized Ohms law, i.e.,
V = Z(;o))I
or
I =
it follows that impedances can be 7nauipiilated in the same manner as resistances, and admittances in the
same manner as conductances. The voltage division formula o f equation 10.25 and the current divi
sion formula o f equation 10.27 illustrate this fact. Example 10.7 further clarifies these statements.
Exercises. 1. Derive a current division formula for three admittances in parallel.

2. Find the admittance and then the impedance of each parallel connection in Figure 10.15.
AN SW FRS: Admittances are
3. Compute the equivalent inductance for Figure 10.15a and the equivalent capacitance for the
circuit ol Figure 10.15b.
ANSWI-RS:
L|
L.
4. Find ^ in terms o f
Cj + Ct+C^
L,
for each circuit in Figure 10.15.
I
O-
+
V
(a)
(b)
FIGURE 10 . 15 . (a) Set of three parallel inductors, (b) Set of three parallel capacitors.
4S9
EXA M PLE 10.7. For the circuit o f Figure 10.16, compute the input impedance ^/(/Co) when co
= 500 rad/sec.
FIGURE 10.16 Series-parallel interconnection of different impedances.

S
o l u t io n
As shown in Figure 10.16, Z^-^^(/500) can be seen as the sum o f three impedances,
+ ^2 + Zy
Our approach is to first calculate Zy for each /.

Step 1. Compute Z y Since this is an I C series combination,
A = j 5 0 0 x 0 .0 0 5 -
5 00
0 .0 0 0 4 )
= - j 2 .5 Q
Step 2. Compute Z-) = MYj. From the propert)' that parallel admittances add and series imped
ances add.
Y2 = y'500 X 0.0 0 0 2 -h
10 + ( 1 0 -H7 IO)
= JO. 1 -f 0 .0 4 - jO .02 = 0 .0 4
H en ce,Z 2= 1/^2 = 5 - ; 1 0 a
Step 3. Compute Z 3 = 1/
Here
= j --------------- = o.i-H 70 . 1 - 70.2 = 0 . 1 - 70.1

5 0 0 x 0 .0 1
Hence, Z 3 = 5 + J5
Step 4. Compute Z-^^. Adding the three impedances together yields
Z. = Z j + Z 2 + Z 3 = - ;2 .5 + 5 - 7 IO + 5 +75 = 10 - 77.5 Q
= 12.5 Z - 3 6 .8 7 Q
y0.08
160
Chapter 10 * Sinusoidal Steady State Analysis by Phasor Methods
Calculations performed in this example are most easily done with an advanced calculator or in
M ATLAB.
For example,
in
M ATLAB
the command
for computing
is Z3
l/(sqrt(0.02)*exp(i*pi/4) - j/(50()*0.01)).
EXA M PLE 10.8. Compute the input impedance Zy^^(/a)) o f the ideal op amp circuit o f Figure
10.17.
I,
V,
FIGURE 10.17 Op amp circuit callcd an impedance converter.

S o l u t io n
The trick to solving this problem entails full use o f the ideal op amp properties discussed in
Chapter 4.
Step 1. From the properties o f an ideal op amp, from KVL, and from Ohms law,
V 2 - V , - M 3 = V,
(10.30)
This follows because the voltage across the input terminals of each ideal op amp is zero and no
current enters the + or - terminal o f each ideal op amp. This implies that
Step 2. Using the phasor voltage division formula o f equation 10.25, it follows that
R
=
T " '
R+ j(oC
or, equivalently,
v,=
1+ -
Jc^RC)
Here, o f course, because o f the idealized properties of the op amp, the voltage
the resistor R in the leftmost op amp.
(10.31)
appears across
Step 3. Writing a node equation at the inverting terminal o f the rightmost op amp yields
again by the properties o f an ideal op amp. Simplifying this equation yields

2V, = Vi + V ,
(10.32)
Step 4. Substituting equations 10.30 and 10.31 into equation 10.32 yields
V.
=\. +
+
jCdRC
-/?!
Equivalently,
Z i u ^ )= ^ = m ~ c
I//J
(10.33)
Equation 10.33 suggests that the op amp circuit o f Figure 10.17 can replace a grounded inductor
whose impedance is jii)L with proper choice of R and C, i.e., L = R^C. In integrated circuit tech
nology it is not possible to build a wire-wound inductor. Instead, inductors are simulated by cir
cuits such as that o f Figure 10.17.
The next section continues to develop our skill with and deepen our understanding o f the phasor
technique by computing the steady-state responses o f various circuits.
8. STEADY-STATE CIRCU IT ANALYSIS USING PHASORS

This section presents a series o f examples that illustrate various aspects of the phasor technique.
Our purpose is not only to demonstrate how to compute the SSS, but also to illustrate the pha
sor counterparts o f Thevenin equivalents, nodal analysis, and mesh analysis. Our first example
reconsiders the parallel RL circuit o f Example 10.4, together with the series RC circuit o f Example
10.6. We will demonstrate the superiority o f the phasor technique over the methods presented in
sections 3 and 4.
i6:
E X A M PLE 10.9. Compute the steady-state voltage V(^t) for the circuit o f Figure 10.18 when i^{t)
= cos(lOOr) A.
R=10Q
21,(t)
i ( t ) 0
C=1mF
R = 100
L = 0.1H
FIG U RF 10.18 7?ACcircuit for Example 10.9.

S
o l u t io n
Step 1. Determine I^. Since the phasor
= 1Z.0 A, by current division,

1
1
Z -4 5 "
(10.34)
R
Step 2. Use equation 10.34 and voltage division on the RC part o f the cirniit to compute \ q Using
voltage division and equation 10.34, the capacitor voltage phasor is
ycoC
j(oC
Step 3. Determine V(^t). Converting the phasor
o f equation 10.35 to its corresponding time
function yields
v^t) = co s(1 0 0 ^ - 9 0 ) = sin(100/) V

The next example illustrates voltage division with phasors as well as the basic impedance relationships.
E X A M PLE 10 .1 0 . Consider the circuit in Figure 10.19 where
= 2 H, and vi^t) = 10 cos(2r) V. Find Vfj^t) and ij{t) in steady state.
= 5
C = 0.1 F, /?2 =
463
SO LU TIO N
Step 1. FindZj^^{jl). Y,^cU'^) = +
= 0.2 + 70.2 = 0.2> / 2Z 45" . Hence
ZrcU2) =
Step l.F in dZ j^ iijl).
= 2 .5 V 2 Z - 4 5 " = 2.5 - J2.S
=-+
R
= 0 . 2 5 - ;0 .2 5 = 0 .2 5 V 2 Z - 4 5 " . Hence
j(d L
Zr,.U2) = -
= 2 v /2 Z 4 5 " = 2 + j2
Step 3. F in d Y a t id Vq (t). From volrage division

V
^ -------- Z r c (J2) -------- Y ^ ------- 2 .5 -j2 .:> ------ ^

2 r c 0 '2 ) + Z r i,0 '2 )
2 . 5 - j 2 .5 + 2 + j2
= 7 g 0 9 ^ _ 3 8 .6 6
4.5-J0.5
It follows that V(^{t) = 7.809 cos(2^ - 38.66) V.

Step 4. FindYj^, 1^, a n
d From step 3,
= 10 - 7.809 Z - 38.66 = 1 0 - (6.098 - ;4 .8 7 8 ) = 3 .9 0 2 4 + ;4 .8 7 8

Hence
l,= ^
1 ^ 2 3 ilM iZ ! =
,5 6 ,7 Z - 3 8 .6 6 "
Thus /^(f) = 1 .5 6 l7 c o s (2 f- 38.66) A.
The next example illustrates the computation o f a Thevenin equivalent circuit with the aid o f
nodal analysis. Because impedances may be manipulated in the same manner as resistances and
admittances in the same manner as conductances, the Thevenin theorem, the source transforma
tion theorem (Chapter 5), and node and mesh analysis (Chapter 3) carr)' over directly.
46^
EXA M PLE 10.11

(a) Find cheThevenin equivalent o f the circuit o f Figure 10.20 if to = 4 rad/sec.
(b) Determine the voltage
when a 1.2
load resistor is connected across terminals a and b.
FIGURE 10.20 Z.Ccircuit for Example 10.11.

S o l u t io n
Find theThevenin equivalent circuit, and then using theThevenin equivalent, find
Vj{t).
Step 1. Establish nodal equation. A nodal equation at the left node o f Figure 10.20 in terms o f phasors is given by
I.V = r ^ L + J^CWoc = -J^ L + y ^ o c
,/(oL
Step 2. Determine the relationship between
and
The relationship between
(10.36)
and
as
determined by the dependent source is

V z - V , , = 0 .2 5 [ ;2 V J
Equivalently,
10.37)
V , = (1 Step 3. Substitute equation 10.37 into equatio7i 10.36. Substituting yields
h = m -j)y o c
Solving for
with
= 1Z.0 yields
Vôc =
: I , = (0 .4 - y0.8)I^ = 0 .8 9 4 Z - 6 3 .4 3 V
0.5 + j
Step 4 . Compute the Thevenin equivalent impedance
(10.38)
Consider the circuit o f Figure 10.21,
which is the phasor version o f Figure 10.20 with the output terminals short-circuited. Hence, the
short-circuit current phasor is
1=
4(n
FIGURE 10,21 Phasor version o f Figure 10.20 with short-circuited terminals.

Therefore, from equation 10.38,
Z,/,0'4) = ^ = ( 0 . 4 - y O . S ) Q
.VC
Step 5. Interpret
to generate the Thevenin equivalent circuit. To physically interpret theThevenin
equivalent impedance, consider that
= (0-4 -yO .8 ) = {R^,, + MjAQ Q.

Thus,
= 0.4 Q and C = 0.3125 F. Hence, the desired Thevenin equivalent circuit (valid at O)
= 4 rad/sec) has the form sketched in Figure 10.22.

0 .3 1 25F
-OFIGURE 10.22 Thevenin equivalent of Figure 10.20.

Step 6. Compute
by voltage division. Using voltage division on the circuit o f Figure 10.22,

=
1.2
l.2 + ( 0 .4 - y 0 . 8 )
= (0.6 + y 0 .3 )(0 .8 9 4 Z 6 3 .4 3 )
= 0 . 6 Z - 36.87 V
Converting the load voltage phasor to its corresponding time-domain sinusoid yields
y^^(t) = 0 . 6 cos(4/^ - 3 6 . 8 7 ) V
166
EX A M PLE 10.12. Determine the phasor voltage
and the corresponding time function vj,t)
for the circuit o f Figure 10.23 if co = 100 rad/sec.
j60Q
I'lG U R E 10.23 Phasor domain circuit for Example 10.12. Ail clement values indicate phasor
impedances at 100 rad/sec.
S o l u t io n
To solve this problem, it is convenient to execute a source transformation on the independent cur
rent source and to combine the impedances o f the parallel combination o f the capacitor and
inductor on the right-hand side o f the circuit. After executing these rwo manipulations, one
obtains the new circuit of Figure 10.24.
FIGURE 10.24 Phasor domain equivalent circuit to that of Figure 10.23. All element values indi
cate phasor impedances at 100 rad/sec. I denotes a phasor loop current.
For the circuit o f Figure 10.24, the indicated loop equation is
250Z.-90O = (50 - ; 2 5 ) I - 0.4(501) - ; 1 5 I = (30 - ; 4 0 ) I
Solving for I yields
I = 4 - J 3 = 5 ^ - 3 6 .8 7 A
Consequently,
= 501 = 2 5 0 ^ - 3 6 .8 7 V and vU) = 250 c o s (1 0 0 r- 3 6 .8 7 ) V.
467
9. IN TRO DUCTION TO THE NOTION OF FREQUEN CY

RESPONSE
The frequency response o f a circuit is the graph o f the ratio o f the phasor output to the phasor
input as a function o f frequency, i.e., as the frequency varies over some specified range. Since the
phasor input and the phasor output are complex numbers, the frequency response consists o f twô
plots: (1) a graph o f the magnitude o f the phasor ratio and (2) a graph o f the angle o f the phasor
ratio. Such graphs indicate the magnitude change and the angle change imposed on a sinusoidal
input to produce a steady-state output sinusoid. In steady state, the magnitude o f the output sinu
soid is the product o f the magnitude o f the input sinusoid and the magnitude o f the frequency
response at the frequency o f the input. Similarly, the phase o f the output sinusoid in steady state
is the sum o f the input phase and the frequency response phase at the input frequency. This prop
erty takes on greater importance once one learns that arbitrary input signals can be decomposed
into infinite sums o f sinusoids o f different frequencies, i.e., each signal has a frequency content.
This notion is made precise in a signals and systems course, where one studies Fourier series and
Fourier transforms. The frequency response o f a circuit describes the circuit behavior at each fre
quency component o f the input signal. This permits one to isolate, enhance, or reject certain fre
quency components o f an input signal and thereby isolate, enhance, or reject certain kinds of
information.
EX A M PLE 1 0 .1 3 . Plot the frequency response o f the RC circuit o f Figure 10.25.
-o-
-O
+
0.01 F
10
-o
FIG URE 10.25 RC circuit passing high-frequency content of an input signal.
S o l u t io n
to the input phasor voltage N^
Using voltage division, the ratio o f the output phasor voltage

is given by
___
1
out __________________yO.Ola)
Vi " i + ----- ! _
= m p ))
l + iO .O lc o
/).01(0
where we have designated this ratio as //(/co).

The two universally important frequencies are
O) =
0 and co=
oo.
At these frequencies, H{jO)
0Z .90 and //(; ) = 1^-0. Asymptotically then, the magnitude |//(/ca)|
1 as
CD
oo
and
468
|//(/to)|
0 as to
0. W ith regard to angle, Z.//(/co)
0 as to
oo and Z.//(/to)
9 0 as oj
0. Also, a close scrutiny o f //(/w) indicates that to = 100 rad/sec is also an important frequency.
Here H{j\00) = 0.707^^45. These values give us a prett)' good idea what the magnitude and
phase plots look like. Using a computer program. Figure 10.26a and Figure 10.26b show the exact
magnitude and phase plots. These plots are consistent with our earlier asymptotic analysis.
Frequency (rads/sec)
(a)
(b)
FIG URE 10.26 (a) Magnitude plot of frequenc)' response for Example 10.13.
(b) Phase plot of frequency response.
Do these frequency responses make sense? They should. Going back to the circuit, observe that at
to = 0, the capacitor impedance is infinite. Physically, then, in steady state, the capacitor looks like
an open circuit for dc, i.e., at zero frequenc)'. The magnitude plot bears this out. For frequencies
close
CO
zero, the capacitor approximates an open circuit and, hence, the magnitude remains small.
O n the other hand, for large frequencies, the capacitor has a very small impedance. This means
that most o f the source voltage appears across the output resistor. The gain then approximates 1,
as indicated by the magnitude plot. The frequency response o f the circuit is such that the highfrequency content o f the input signal is passed while the low-frequency content o f the input sig
nal is attenuated. Such circuits are commonly called high-pass circuits.
EXA M PLE 10.14. Investigate the frequency response o f the parallel /^Z,C circuit o f Figure 10.27.
R=10
L=0.04H
C=0.25F
FIGURE 10.27 A parallel RLC circuit having a hand-pass frequency response.

S o l u t io n
The input admittance o f the circuit o f Figure 10.27 is given by
LC
R
RC
. CO
jOiL
C
Inverting to obtain the input impedance yields
w
C
J_ _ ^ 2 ^ ._ o L
LC
RC
Clearly,
y'4co
I00-co-+y4co
Hence the ratio o f the output phasor to the input phasor is simply
Zy^^(yoj). Once again, co = 0 and co = oo are the first two frequencies to look at. Here Zy^,(0) =
0Z .90 and -2^,(oo) = 0Z.-90^. Also at co = 10, the impedance is real, i.e.,
0) = I . These three
points provide a rough idea o f the magnitude and phase response. Two more points are necessary
for a real .sense o f the frequency response. At what frequency or frequencies does the magnitude
drop to 0.707 o f its maximum value or when does the phase angle equal 45? This will occur
when 1100 - co^| = |4co|. This is a quadratic equation. Flowever, because o f the absolute values,
there are rvvo implicit quadratics, co" - 4co - 1 0 0 = 0 and co^ + 4co - 100 = 0. Solving using the
quadratic formula yields co = 8.2, 12.2. Since the magnitude plot is symmetric with respect to
the vertical axis (co = 0 axis), we consider only the positive values o f co. This information provides
a good idea o f the magnitude and phase plots. A computer program was used to generate the fre-
470
quenc)' response plots in Figure 10.28a (magnitude) and Figure 10.28b (phase). The magnitude
plot shows that frequencies satisfying 8,2 < O) < 12.2 are passed with little attenuation. Frequencies
outside this region are attenuated significantly. Such a characteristic is said to be o f the band-pass
type, and the corresponding circuit is a band-pass circuit.
(a)
(b)
F IG U R t 10.28 (a) Magnitude plot o f frequency response for band-pass circuit o f Figure 10.27. (b)
Phase plot of frequency response.
EX A M PLE 10.1 5 . As a final example vve consider the so-called band-reject circuit o f Figure
10.29. A band-reject circuit is the opposite o f a band-pass circuit. A band-reject circuit has a band
o f frequencies that are significantly attenuated while it passes with little to no attenuation those
frequencies outside the band. In this example our goal is to compute the magnitude and phase o f
the frequenc)' response o f the band-reject circuit o f Figure 10.29.
R =10
FIGURE 10.29 Band-reject circuit for Example 10.15.

S o l u t io n
Once again using voltage division, we obtain the phasor ratio
'
LC
t)Ul
-(0
1 0 0 - CD"
_ L _ c o 2 + /o)-5
LC
1 0 0 - 0 ) 2 +^25co
H{jLo) = 1Z.0. Hence, as}^mptotically, |//(/to)| approaches 1 as OJ approach

es 0 and CO. Also at (o = 10, //(/co) = 0 Z .-9 0 while at to = 10+, Hijo)) = 0 Z .-2 7 0 = 0Z.90". For
At to = 0 and
CO = o o ,
this example, to find the frequencies where |//(/to)| drops to l/ V I o f its maximum value o f 1, it
is necessary to equate the magnitudes o f the real and imaginary parts o f the denominator. This
produces two quadratics whose positive roots are to = 3 .5 0 7 8 and OJ = 28 .5 0 7 8 . At these frequen
cies the angles o f //(/w) are 45 and 45, respectively. Fhe computer-generated plots o f Figures
10.30a and 10.30b are, o f course, consistent with these quickly computed values.
472
(a)
(b)
F I G U R t 10.30 (a) Magnitude plot of frcquenc)^ response for band-reject circuit of Figure 10.29. (b)
Phase plot of frequency response.
As wc can see, a wealth o f different kinds of frequency response are obtainable by different inter
connections o f resistors, inductors, and capacitors. Historically, phasor techniques were the essen
tial tool for the analysis and design of such circuits. Nowadays, engineers ordinarily use either
M A T L A B or SPICE to obtain frequency response plots. Two examples follow where we use MATLA B , SPICE, or both to obtain the frequenc) response.
4 '3
E XA M PLE 10.16. Compute the frequency response o f the circuit o f Figure 10.31 using MATLAB and SPICE.
R=10Q
FIG URE 10.31 /^/.Ccircuit for Example 10.16.

S oluti on
This circuit was originally analpêd in Example 10.9. You might want to refer to that example
before proceeding.
SPICE Part. A SPIC E simulation produces the result shown in Figure 10.32.
EX1016 FreqRsp-Small Signal AC-0
+20.000
+40.000
+60.000
+80.000
Frequency (Hz)
+100.000 +120.000 +140.000
FIGURE 10.32 SPICE plot of capacitor voltage for the circuit of Figure 10.31.
MATLAB Part. Although the analysis appears in Example 10.9, we can use MATLAB to more
easily obtain the frequency response. First define Zj(/co) = ytoA and Z-,{p)) = l/y'coC Then from
current division,
= -7 " r 7
R+
Z|(yco)
1 ,0 )
and from voltage division,

2-. (/CO)
R+
Z 2 (y co )
K2( 7(0) +
21^( 700)
474
Assuming a frequency range o f 0 < co < 1000 rad/sec, the following MATLAB code will result in
a suitable magnitude frequency response plot, as shown in Figure 10.33.
L = 0.1;R = 10;C = 0.001;
w = 0:1:1000;
Z1 = j*w*L;
Y2 = jw*C;
>>IL= R./(R+Z1);
VC = 2 1L./(R *Y 2+1);
plot(w/(2*pi),abs(VC),b)
grid
xlabel(Frequency in Hz)
ylabel(Capacitor voltage (V ))
FIGURE 10.33 Magnitude plot of frequenc)^response o f capacitor voltage in the circuit of Figure
10.31. The response is of the low-pass t)'pe.
Now suppose the inductor in the circuit o f Figure 10.31 is replaced by a capacitor C, = 1 mF with
the controlling current changed to /q (0- The frequency response is easily computed with a sin
gle change to the MATLAB code, namely, Z , = 1. ./(j*w *0.001). The resulting plot shows a
band-pass characteristic, as illustrated in Figure 10.34.
47S
F IG U R E 10.34 Magnitude plot of frequency response of capacitor voltage in circuit of Figure 10.31
when inductor is replaced by a 1 mF capacitor. The response is of the band-pass type.
E X A M PLE 10.17. In Chapter 9 we investigated the Wien bridge op amp oscillator circuit,
redrawn in B2 Spice in Figure 10.35.
4" 6
Chapcer 10 Sinusoidal Steady State Analysis by Phaser Methods
Two difTerences are notable: ( !) there is a current sourcc present across the
combination,
and (2)
is now 10 kQ, as opposed to 9.5 kH in Example 9.14, forcing /?, = /?2- This means
that the characteristic equation for the circuit is
1
V*' + / ? i ' + ( = .y~ +
{RlC
R,C)
which indicates a purely sinusoidal oscillation at the frequenc)'
f o = - ^ = -----! = 15.92 Hz
"
2n
2 k R,C
for any initial condition on C ,. In fact one might recall that R^ < Rj causes a growing oscillation
that is limited by the saturation effects o f the op amp.
The current source, set at 1 A, is present in Figure 10.35 so that we can obtain the frequency
response cur\'e shown in Figure 10.36. In Figure 10.36 observ'e that the magnitude response peaks
at/q, as expected from the theoretical analysis. In an actual circuit, the current source would not
be present. Nevertheless, a sustained sinusoidal oscillation will occur because o f the presence o f
noise. W ithout going into the analysis, noise contains an infinite number o f frequency compo
nents, each o f which has a minute magnitude. In particular, noise contains frequency components
around / q that drive the circuit into oscillation. This is precisely what the peak in the frequency
response means: a very small (noise) voltage on Cj will cause a very large-magnitude sinusoid out
put voltage at /q. However, the presence o f nonlinearities such as saturation keep the magnitude
at an acceptable level.
Exi0.17-Small Signal AC-13

+12.000
+13.000
+14.000
+15.000
+16.000
+17.000
Frequency (Hz)
+18.000
F I G U R E 10.36 Frequeno,- response plot o f W ie n bridge oscillator.
+19.000
+20.000
10. NODAL ANALYSIS OF A PRESSURE-SENSING DEVICE

The bridge circuit presented in Figure 10.37, or some variation of this bridge circuit, has been and
continues to be a widely used approach to accurate measurement technology. In this section we
will analyze the ac bridge circuit o f Figure 10.37 as a pressure measurement device. The capaci
tance C t is a diaphragm capacitor consisting o f a hollow cylinder capped on either side by fused
quartz wafers. Bersveen the wafers is a vacuum. The capacitance o f the diaphragm changes with
temperature and pressure. For our analysis we will assume that the temperature is constant and
that the pressure is constant for a time period greater than five times the longest time constant o f
the circuit. This will allow the voltages and currents in the circuit to reach steady state and thus
allow us to use phasor analysis to compute their values.
R =100 0
15cos(20,000nt)V
F1GUR1-' 10.37 Bridge circuit diagram of pressure-sensing device. The capacitance

function of pressure, which causes the voltage
changes as a
to changc as a function of pressure. This is
registered on the attached voltage meter, which has a 1 MQ internal impedance.

As a rule o f thumb, the capacitance C-, Q.llAKAId. This means that the capacitance is inverse
ly proportional to the distance d between the plates and proportional to the area A o f the plates
and to the dielectric constant K o f the material between the plates. Increasing the pressure on the
diaphragm decreases the distance rf'between the wafers, increasing the capacitance. Conversely, a
decrease in pressure will increase the distance between the wafers, thereby decreasing the capaci
tance. As the capacitance changes, the magnitude o f the ac voltage appearing across the voltage
meter will vary accordingly. Hence, two relationships are necessary: (1) the relationship berween
the capacitance C , and the magnitude o f the voltage
- V e and (2) the relationship berween
the pressure applied to the diaphragm and the associated capacitance. Our first task will be to
specify the relationship between the pressure applied to the diaphragm and the resulting capaci
tance. Following this, we will use nodal phasor analysis to determine the magnitude of
and finally, the relationship between pressure and the magnitude V ^ - V e
Pressure is measured in various units. Millimeters o f mercury (mm Hg) is a common standard; 1
mm Hg = 1 torr, and 760 torr = 1 atmosphere (atm), where 1 atm is the pressure o f the earths
atmosphere at sea level, which supports 76 0 mm o f mercur\' in a special measuring tube. Suppose
4 8
it has been found experimentally that the capacitance C j (in pF) varies as a function o f pressure
according to the formuhi
C 2 ( A P ) = Q ) + ^ lo g 10
= 2 6 .5 + 68 log 10
^0
(10.39)
(16i)+A P^
760
A plot o f C2 as a function o f AP is given in Figure 10.38.
QJ
(T3
Q.
U
Change in Pressure
FIG URE 10.3H Plot of capacitancc versus pressure.
Our next task is to develop the relationship between the capacitance of^ the bridge circuit and the
magnitude o f the phasor voltage
and
In our analysis, G, = (/?,) ', Gj =
= (^ 3)"^
= 10~^ S is the conductance o f the meter , According to Figure 10.37, Cj = 20 pF. We
will let Cj range as 0 < C 2 < 40 pF. Finally, co = 2h x 10"^ rad/sec. The following phasor analysis
will be done symbolically so as not to obscure the methodology.
Summing the phasor currents leaving node A leads to the phasor voltage relationship
(G j + G 2 + yujCj)v^ G-jV^ ycoCj
= Gj 15
Similarly, summing the currents leaving node B leads to the relationship

- G 2 V ^ .( G 2 .G 3 .G J V ^ - G ,,V
=0
479
Finally, summing the currents leaving node C produces

- > C ,V ^ .;( o ( C , . q
G JV c-
=0
Writing these three equations in matrix form yields
G] + G 2 +
-/(oC,
-G i
G 2 + G 3 + G ,
-G j
-yco C ,
- G ,
l5 G ,'
V /
=
~^m
G i + ^ ( C | + C 2 )
Vc.
(10.40)
The matrix on the left is said to be a nodal admittance matrix. Its entries can be real or complex,
as indicated. It is nor advisable to solve such a set o f equations by hand over the range o f possible
C2 values. However, using MATLAB one can solve this matrix equation over the range 0 pF < C,
< 40 pF to produce the plot o f Figure 10.39.
c
CO
QJ
01
nj
*->
<U
cn
-o
o
cu
o
D
'E
cn
fO
10
15
20
25
30
35
40
C, in pF
FIGURE 10.39 Plot of the magnitude o f the phasor voltage
V ^ as a function of capacitance.
O f course, one could measure the voltage appearing across the meter, from Figure 10.39 deter
mine the associated value o f Cj, refer to Figure 10.38 for AP, and then determine P = 760 + AP.
This is a long route. To complete our analysis, then, we need to develop the relationship between
pressure and bridge voltage. As we have the relationship between C j and APand the relationship
between C2 and |
|, it is a matter o f using equation 10.39 to derive the value o f
in
equation 10.40. This is best done with a simple MATLAB routine, which yields the plot given in
Figure 10.40.
480
Cliaptcr 10 Sinusoidal Steady State Analysis by Phasor Methods
>
m
>
'o
01
T3
D
'c
ro
300
400
500
600
700
900
800
1000
1100
1200
Pressure in mm Hg
I'lG U RE 10.40 Relationship herween magnitude of bridge output voltage and pressure applied to
diaphragm capacitor C-,.
An actual pressure sensor would, oFcourse, be more complex. For example, there would probably
be a difterential amplifier such as the one shown in Figure 10.41 across the terminals oFthe bridge
circuit, and this would probably drive a peak (ac) detector to determine the maximum value oF
the ac signal appearing at the output oF the dlFFerential amplifier. Fiu ther, the peak value would
probably be read by a digital voltmeter. Nevertheless, our analysis illustrates the basic principles
involved in such a measurement. OF course, one could just as easily use loop analysis to solve the
problem. This is leFt as an exercise in the problems.
kR.
-o V,
F IG U R E 10.41 Difierential amplifier having output voltage
= k { i >2 - /î).
481
Exercise. Prove that
= k{v-, -
for the differential amplifier o f Figure 10.41.
11. SUM M ARY

The two primar)' goals o f this chapter were (1) to de\'elop the phasor technique for the analysis of circuits
having a sinusoidal steady state and (2) to illustrate how this technique leads to die idea of a circuit frequenc)^ response, which characterizes the circuits behavior in response to the frequency content of an input
excitation. In the development, sinusoids were first represented ;is the real part of a complex sinusoid. As
a motivation for the delineation of die phasor method, we showed how the complex sinusoids could be
urilized to compute the sinusoidal steady-state response using difierential equation circuit models. We then
pointed out that a complex (voltage or current) sinusoid is specified by a complex number or pha.sor rep
resenting its magnitude aind phase. After introducing the notions of impedance and admittance for the
capacitor, the inductor, the resistor, and a general two-terminal circuit element, we showed how the pha
sor voltage and phasor current for each such element satisf}" a frequency-dependent Ohms law. I'his
iillowed us to adapt the ;inalysis techniques and network dieorems of Chapters 1 through 6 to the steadystate analj'sis of circuits excited by sinusoidal inputs. For example, diere are voltage division formulas, cur
rent division formulas, source transformations, and The\'enin and Norton theorems all valid for phasor
representations. This permits us to effectively analv/e circuits diat have a src*ady-state response.
The phasor technique opens a door to seeing how circuits behave in response to sinusoids. Given that
input excitations are composed o f different frequenc)' sinusoids, such as a music signal, phasor analy
sis shows why a circuit will behave differently toward the different frequencies present in the input sig
nal. This fact prompts the notion o f a circuits frequency response, which is defined as the ratio o f the
phasor output to the phasor input excitation as a function o f (u in the single-input, single-output case.
The frequency response consists of two plots. The magnitude plot shows the gain magnitude o f the
circuits response to sinusoids o f different frequencies, and the phase plot shows the phase shift the cir
cuit introduces to sinusoids o f different frequencies. The notion o f frequency response will be gener
alized in Chapter 14 afrer we introduce the notion of the Laplace transform.
12. TERM S AND C O N CEPTS

Admittance: o f a two-terminal device, the ratio o f the phasor current into the device to the
1;in
phasor voltage across the device, y< ( /CO) =
in
Band-pass circuit: circuit in which frequencies within a specified band are passed while frequen
cies outside the band are attenuated.
Band-reject circuit: circuit in which one band o f frequencies is significantly attenuated while
those frequencies outside the band are passedes with little to no attenuation.
Com plex exponential forcing function: function o f the form v{t) =
^ = a +yco are complex numbers. A special case (a = 0), f(t) =
out the chapter as a shortcut for sinusoidal steady-state analysis.
, where V =
and
is used through

^
Conductance: real part of a possibly complex admittance.

Current division: in a parallel connection of admittances driven by a current source, the current
through a particular branch is proportional to the ratio of the admittance of the branch
to the total parallel admittance.
Euler identitjr:
= cos(0) + j sin(0).
Frequency: in a sinusoidal function A cos((j)t + d) or B sin((Of + 0), the quantity co is the angular
frequency in radians per second (rad/sec). Equivalendy, A cos(o)f + 6) = A cos(2K/t + 0),
where/is the frequency in hertz (Hz or cycles per second). Note that O) = 2n f.
Frequency response: (of a circuit) graph of the ratio of the phasor output to the phasor input as
a function of frequency. It consists of two parts: (1) a graph of the magnitude of the pha
sor ratio and (2) a graph of the angle of the phasor ratio.
High-pass circuit: circuit with a frequency response such that the high-frequency content of the
input signal is passed while the low-frequency content of the input signal is attenuated.
Imaginary part: the imaginary part of a complex number z = a + Jh for real numbers a and b,
denoted by Im[z], is b.
Impedance: ordinarily complex frequency-dependent Ohms law-like relationship of a two-ter
minal device, defined as Z(/a)) = V/I, where V is the phasor voltage across the device and
I is the phasor current through the device. For the resistor,
= /?; for the capacitor,
= l/(/a)Q; and for the inductor,
= ycoZ.
Magnitude (modulus): the magnitude of a complex number z = a + jb , denoted by |z|, is
Phason complex number representation denoting sinusoidal signals at a fixed frequency. Bold&ce
capital letters denote phasor voltages or currents; a typical voltage phasor is V =
and a typical current phasor is I =
Polar coordinates: representation of a complex number z as p?-^, where p > 0 is the magnimde of z and
0 is the angle z makes with respea to the positive horizontal (real) axis of the complex plane.
Reactance: imaginary part of an impedance.
Real part: real part of a complex number z = a + jb for real numbers a and b, denoted by Re[], is a.
Rectangular coordinates: representation of a complex number z as coordinates in the complex
plane, i.e., zs a + jb for real numbers a and b.
Resistance: real part of a possibly complex impedance.
Sinusoidal steady-state response: response of a circuit to a sinusoidal excitation after all transient
behavior has died out. This definition presumes that the zero-input response of the cir
cuit contains only terms that have an exponential decay.
Stable circuit: circuit such that any zero-input response consists of decaying exponentials or expo
nentially decaying sinusoids.
Susceptance: imaginary part of an admittance.
Voltage division: in a series conneaion of impedances driven by a voltage source, the voltage
appearing across any one of the impedances is proportional to the ratio of the particular
impedance to the total impedance of the connection.
Zero-input response: response of the circuit when all source excitations are set to zero.
In the literature, both z and z* are used to denote the conjugate o f a complex number z. However, in matrix
arithmetic, Z* usually means the conjugate transpose o f the matrix Z. We will sometimes interchange the usage.
In MATLAB, * means multiplication and conj(Z) means conjugated. So there is some ambiguity in the usage.
^
^
483
Problems
SO LUTIO N OF DIFFEREN TIAL
EQUATIONS W ITH CO M PLEX
EXPO N EN TIALS
'
Figure P I0.3
A N SW ER S: 0 .8 sin (2500r -
3 6 .8 6 >). 80
1. Construct the difFerential equation model o f
s in (2 5 0 0 r-3 6 .8 6 )
the series RL circuit o f Figure PI 0.1 in which L

= 0.25 H and R= 100 Q. Then use the method
4. Construct the differential equation model of
o f section 4 to compute the steady-state

response when v, (/) = 20V 2 cos(400/) V.
the parallel /?Z.Ccircuit o f Figure PI 0.4 for R =

100
C = 1 pF, and Z. = 40 mH. Then use
the method o f section 4 to find the steady-state

response when / (f) = 20 cos(2500f) mA.
vJt)
R < v .J t )
Figure P I0.1
AN SW TR: 20 cos(400/ - jr/4) V
Figure P I0.4
AN SW ER: 1.6 cos(2500; + 36.8~) V
2. Find the differential equation model o f the

series RC circuit o f Figure P I0.2 in terms o f
and V(it) assuming that C = 5 pF and R
= 800 Q. Write
as a function o f
and V(^t). Then use the method o f section 4 to

determine the steady-state response when
iV(/) = 20>/2sin(250/) V.
+ v,(t) -
KCL AND KVL W ITH

PHASORS
5. Find the phasor current I and /(/) for each
circuit o f Figure Pi 0.5 when (O = IOOti rad/sec.
(2+j4)A
+
v(t)
V (t)
Figure P i0.2
ANS^XTR: 20 sin(250/ + 0.25k) V
Figure P i0.5
AN SW TRS: (a) 10 cos( 1OOTif - 0 .9 :" )
A. fb)
S.6626 cos( 1OOtt/- 1.798) A

3. Construct the differential equation model o f
the series RLC circuit o f Figure Pi 0.3 in terms
o f /^(r) and
assuming L = 1 0 mH, C = 4
pF, and R = 100 Q. Then use the method o f
section 4 to find the steady-state response /^(t)
when
= 100 sin(2500/) V. Next compute
6. T he circuit o f Figure P i 0.6 operates in the

sinusoidal steady state with the indicated pha
sor currents when i^{r) = 10 cos( 1OOOf) A. Find
the value of the phasor currents and and the
associated /,(r) and
484
j20
(26+j12)V
20
(a)
Figure P i0.6
C H E C K : /,(/) = 25 cos(100r + 0.9273) A
j20
(26+ jl2)V
L
20
7. Suppose that in Figure P I 0.7, v^{t) = 4

cos(o)f) V and ''2^^ ^~ 4-s/2 cos(o)/ 0.25ti ) V.
Find V[{t) = A"cos(ior+ (J)).
+
v^(t)
(b)
e
(26 + jl2)V
Vj(t)
v,(t)
j20
20
(0
Figure P I0.7
ANSW ER; vjU) = 4 cos((./ - )0") \'
Figure P I0.9
8. Use KVL to determine the phasor voltage
10. For the circuit o f Figure P i 0.10, use KCL
in the circuit o f Figure PI 0.8.

8jV
4jV
and KVL to find the phasor voltage V^. and the

phasor current I^,. If the frequency co = 2 0 0 0 ;:
rad/sec, find the associated voltage and current
time functions.
-
(4 + ]4)V
Figure P10.8
(2-j10)A
(2+j4)A
V _y (2+
j
/ Ti
(t)
AN SW ER: - 4 V
(4 + j6) A
(4+j6)V
(5
9. For the circuits of Figures PI0.9a, b, and c,
Figure PIO.IO
compute the indicated phasor currents assuming
ANSW'ERS: V^.= 18.98^17156 V. I. = 2 0 .4 ^
R = 2 0 ., L = A mH, and C = 1 mF. If oj = 500
-1 0 1 .3 '* A
rad/sec, determine the associated time functions.
BASIC IM PED AN CE AND

A D M ITTAN CE CO N CEPTS
11. (a)
A capacitor has an admittance
= y8
mS at OJ = 400 0 rad/sec. Find the
48S
value o f C in pF. Compute the value of

the capacitors impedance at (0 = 500
rad/sec.
An inductor has an impedance Zj =
(b)
j20 Q at (0 = 4000 rad/sec. Find tlie

value o f L in mH. Compute the value
Figure P I0.14
of the inductors admittance at 10,000
A N SW ERS:
rad/sec.
3 0 > A. I,
= 0 .0 2 Z 6 0 > A, 1/ = 0 .0 4 Z = 0 .()2 Z ]5 0 " A. i j t ) = 0.02S3
cosi 11)00/ + 1S>) A

12. In the circuit o f Figure P I0.12,
cos(lOf) V, C = 0.2
(f) = 10
F, Z: = 0.1 H, R= 2.5 Q.
Determine the phasors I j, I 2, and
Find
15 . Consider the circuit o f Figure P i 0.15

where R = 200 H, A = 80 mH, C = 1 pF and
i- p ) =100 sin(2500r) niA. Find the voltage
".a/')-
phasors
r
I''
*
' r
V^,
and
and then compute
"
R
JY Y V
v -
+ V. -
Figure P i0.12
13. The circuit o f Figure PI 0.13 is operating in
the sinusoidal steady state with
= 20
Figure P i0.15
C = 1 mF.
C H E C K : v ;p ) = 28.28 co s (2 5 0 0 r- 135) V or
(a)
v .p ) = 28.28 s in (2 5 0 0 f- 45") V
Suppose /^|(r) = 10 cos(100r + 30'*)

mA and v^Û) = 200 cos(lOOr) mV.
(b)
Find the phasor I^. and then the cur
16. In the circuit o f Figure P IO .I 6 , C = 0.03 S,
rent ip ) .
I = 0.1 H, C = 0.4 mF, /j(r) = 1.2 cos(200r) A
Now let /j,(r) = 10 cos(50/ + 3 0 ) mA
and
and v^2 ^t) = 200 cos(lOOr) mV. Find
(a)
the current
(b)
How does this part
= 40 sin(200^) V.
Find the phasors Ij and
Find the phasor 1^ and the associated
time function ijit).
differ from part (a)?

fi.(t)
v,(t)
Figure P I0.13
14. Find the phasor currents
Figure P I0.16
AN SW ER: (b) 1.2Z-9()'\ 1.2 sin(200r) A
and
and then determine i j p ) for the circuit o f

Figure P I 0.14 in which /? = 1 k li,
C = 1 pF, and
= 0.5 H,
= 20 cos(1000r + 60) V.
17. In the circuit o f Figure P 10.17, /? = 6 f i, Z.

= 80 mH, C = 0.5 mF, Vp{t) = 8 cos(200r) V,
and I, = 0.5Z 90 A. Find the source voltage,
which operates at the same frequency o f
200 rad/sec.
V jt)
Z Jj(o )
Figure P I0.17
ANSWHR: S.831 1 cos(200r +30.96<) V
Figure P 10.20 Parallel Z.C circuit.

ANSWl-.RS: (a) -/1.25 12, (b) 1.6 ml-
18. The circuit o f Figure PI 0.18 operates in the

sinusoidal steady state at a frequency o f cOq =
21. Consider the circuit o f Figure P I0.21.
2000 rad/sec, /?, = 7?^ = 10 Q, V- = 50 V, and
(a)
Find the impedance at (o = 100 rad/sec.
= 2Z. - 53.13 A. Compute the phasor volt
(b)
What happens to the impedance as to
age across Rj and then find the impedance

Now construct a simple series circuit
gets large?
(c)
Ifv/(/) = IOV 2 cosdOOOV, find /.(/).
that represents this impedance at cOq.

ijt )
1 mF
v jt)
lOO
0.1 H
Figure P I0.18
Figure P I0.21
ANSWHR: Z = 2.5 +ylO Q
A N SW FR: (a) S - p U
19. (a)
Find the steady-state response o f the

circuit o f Problem 3 using the phasor
22. For the circuit o f Figure PI 0.22, suppose R
method. Discuss the relative advan
= 100 Q, ^ = 0.5 H, C = 5 îF.

(a)
If
= 0.1 cos(500/) A, find vît).
(b)
Find (O ^ 0 in rad/sec so that the input
tages o f the phasor method.

(b)
Find the steady-state response o f the
admittance is real.
circuit o f Problem 14 using the phasor

method. Discuss the relative advan
tages o f the phasor method.
+
V ,( t)
SERIES-PARALLEL
IM PED A N CE AND
A D M ITTA N CE
CA LCU LA TIO N S
20 . Consider the circuit o f Figure P I 0.20.
(a)
(b)
(c)
Figure P i0.22
If C = 0.01 F, findZ.(/100).
If Zy,^(/100) = 25j O., find the appro
23. Consider the circuit o f Figure P I0.23 in
priate value o f C.
which /?j = 20
/?2 = 10
Z, = 20 mH.
(a)
I f C = 0.3 mF, find K.(;-500).
Using the impedance o f part (b), if
ijU) = 100 cos(100r + 45) mA, find

v^t).
(b)
Find the value o f C that makes the input

admittance real at OJ = 500 rad/sec.
(c)
If C = 0.3
mF and /.(f) = 100
487
zero. Determine the minimum value
cos(500r) niA, find /^](^) and /^-(r)
o flK (/ 0 ))l.
using current division.

/Y Y V
L
(a)
Figure P I0.23
(b)
A N SW ERS: (a) 0.1 + yO.l S, (b) 0.1 mF
Figure P I0.25
24. Consider the circuit o f Figure P I 0.24 in
which C = 1 F. At C0= 2 rad/sec,
= 4 + jl
n.
26.
For a particular two-terminal device,

= 0.002 + y0.002 at O) = 500 rad/sec.
(a)
(b)
(c)
Find the appropriate values o f R and
Construct a parallel RC circuit having this
admittance at (0 = 500 rad/sec. If the circuit is
For 0 < Z < 0.2 H, specify the range o f
excited by a currcnt source with ij^t) = 10
possible reactance values for
cos(500f) niA, find the voltage appearing across

the current source.
If
= 10 cos(2/) V, find Vjî).
A N SW ER : C = 4 uF, A* = 500 Q ..
5
\'^(/) = ;= cos 500/ - 4 5 " ) V
V2
27. For a particular two-terminal device,
'jt ) Q
Z .(;1000) = 2000 + ;2 0 0 0 Q.. Construct a

series RL circuit having this impedance at CO =
1000 rad/sec. If the circuit is excited by a volt
Z,(jw)
age source with v^{t) = 10 cos(1000/) V, find the

current through the resistor.
Figure P I0.24
AN SW ER: Z. = 2 H, /^ = 2 kQ.,
5
/(/) = - = cos 1000/ - 4 5 " ) mA
AN SW ERS: 4 Q, 0.2 H, 0 to x
25. Consider the circuits o f Figure P I 0.25 in
which R = 5 O., L = 32 mH, and C = 5 ^F.
(a)
Find
28. The circuit o f Figure P i 0.28 operates in the
Zy^^(/(0) as a function o f (0. Then com
sinusoidal steady state at the frequency' (0 =
pute the frequency at which Zy^^(/CO) is

purely real, i.e., the reactance is zero.
5000 rad/sec with R =4 Q., L = 0.4 mH, and C

= 0.1 mF. Find Z-^^{j5000) and ^/(y5000).
Determine the minimum value o f
Construct a simple series circuit that is equiva
For the circuit o f Figure P 10.25b, find
construct a simple parallel circuit that is equiv
Ky^;(/0)) as a function o f (0. Then com
alent to this circuit at 0) = 5000 rad/sec. In both
pute the frequency at which
cases specily the element values.
Consider
Figure
P 10.25a.
lent to this circuit at 0) = 5000 rad/sec. Finallv,

(b)
is
purely real, i.e., the susceptance is
488
Chapter 10 * Sinusoidal Stoatly State Analysis by Phasor Mcthoils
(b)
R
Find the value o f to in terms o f R and

C at which the phase angle difference
bet\veen \ a n d
(c)
is 4 5 .
At the w computed in part (b), deter-
Figure IM0.28
-H f
C H EC K : Z.(/5000) =1 + ;2 ^
29. For the circuit o f Figure P i 0.29, let Z.=4

mH, C = 10 |iF, and v^{t) = K^cos(cor) V.
Compute
Find the frequency co (in
Figure P I0.31
rad/sec) at which tlie steady-.state current /^(/) =

0. At this frequenc}', what is the vohage across
32. Consider the circuit o f Figure PI 0.32.
the LC parallel combination?
(a)
If vj^t) = V^^(zos{tlRQ, find the sinusoidal

steady-state response Vf{t) in terms of
V;,, k and C.
(b)
lf/?= 1 0 Hand
10 n/2 cos(10/)
V, find the value o f C so that V(^t) =

10 cos(10r + 0) V.
(c)
For the value o f C found in part (b),

compute the corresponding value ol 0.
Figure P I0.29
AN SW FR: (I) = 5000 rad/sec
SERIES/PARALLEL
IMPEDANCES WITH V/l
DIVISION
^,(t)
30. Consider the circuit o f Figure P I0.30 in

which
/? =
20
Q,
Z.
H,
( 0 = 10>/2 cos(5/) niA.

(a)
Find the input impedance ^,(/w) and

the input admittance
(b)
Figure P I0.32
and
At OJ = 5 rad/sec determine the steady-
ANSW ER: (a) r(^(/) = ^ c o s

v2
RC
-4 5

which y? = 8
(a)
state current ;^(r).
Z, = 8 mFi, and C = 0.125 mF.
Determine the values o f the phasors

I^, and V(^ when
i,(t)
= 2 A and (d =
1000 rad/sec. Specify the correspon

ding time functions.
(b)
Repeat part (a) for o) = 500 rad/sec.
Figure P I0.30
ANSW'F'.KS: (a)
.b; I (I
31. (a)
y20co
5+
/(I)
0.25
. 0.05 - /
(I)
5; - k !4) niA
For the circuit o f Figure P I0.31, find
the ratio
in terms of/^, C, and
CO.
Express the answer in polar form.
Figure P i0.33 Parallel y?/.Ccircuit.

34. In Figure P I0.33, suppose R = 500 Q, I. =
mH,
C
=
0 .1 2 5
i:,Ar) = loV2co.s(co/+ 60^) A.
0 .5
mF,
and
hS9
(a)
Compute the values o f the phasors I^
Using the phasor method, find
I^, and
v j t ) = 50 cos(4000^) mV.
when
= 2 A and OJ =
4000 rad/sec. Specify the correspon
He
ding time functions.

(b)
Repeat part (a) for (O = 8000 rad/sec.
when
100v,(t)
v.(t)
35. In the circuit o f Figure P i 0.35, suppose R
= 2 0 n , L = 0.5 H, and C = 0.625 mF.

(a)
If
= 16 cos(40r) V, find
V(^t) and vît) using phasor voltage
division.
(b)
If
Figure P i 0.38 Iw o coupled /^Ccircuits.
39. Consider the circuit o f Figure P i 0.39 in

which /?j = 200 n , Z = 0.2 H, R, = 200 Q, and
= -32sin(40r) V, find
v^t), and vît) using phasor voltage

division. Hint: avoid repeating the cal
C = 0.05 pF. Use the phasor method to find
i^t) when
= 10 cos(lO'^r) mA.
culations o f part (a); this can be done

by inspection.
Figure P i0.39 Two coupled circuits.

Figure P 10.35 Series /?/,C circuit.

36. Reconsider Figure 10.35 lor /? = 10
Z. =
Find
0.08 H, and C = 0.02 R

(a)
If
which /?! = 500 Q, I = 0.125 H, R ,= m Q,

and C = 5 pF. Suppose Vj^^) = 120 cos(400/) V.
and
= 10 cos{25t) V, find
and vît) using phasor voltage division.

(b)
If za//) = 16 sin(50f) V, find

and
V[ {t)
V({t),
using phasor voltage division.
37. In the circuit o f Figure PI 0.37, /?, = 20 Q.,

i = 2 H ,a n d / ? j = 1 0 a . If|V/V,| = 0.2 at
co= 40 rad/sec, find the necessary value(s) o f C
(in mF).
Figure P i0.40
which 7^1 = 500 QX = 0.125 H, R, = 100 iX
and C = 5 pF. Suppose
''" 6
mA. Find
Figure P I0.37
AN SW TR: 0.625 mF or 0.2083 mF
38. In the circuit o f Figure PI 0.38, /?, = 50 Q,
q
= 1 uF, Rj = 300 Q, and q
= 0.625 pR
and
= 120 cos(400r)
490
AN SW ERS: 2 Z 9 0 >, - 2 sin(400r) A

44. For the circuit o f Figure P i 0.44, find the
and the phasor voltage Y q If
phasor current
the circuit is known to operate at a frequency o f

CO = 1000 rad/sec, find /^(f) and
and the
values o f L and C
2Q
> j2 Q
-j2 0
NETWORK THEOREMS IN
CONJUNCTION WITH V/l
DIVISION.
AN SW ERS; t'fit) = 2 cos(lOOOr) V, i/(/) = cos
(You should consider applying one or more
(1000/-O.S3T) A
Figure P I0.44
network theorems to simplify the solution to

the problems in this section.)
45. In the circuit of Figure Pi 0.45, R=20Q., L = 20

mH, C= 100 pF, and v^.(0 = 2 0V 2 cos(IOOO/)V.
42. In the circuit o f Figure P 10.42, R^ = 60
Compute the value of
in steady state.
/?2 = 40 n , and C = 0.1 mF. Find the phasor Ij

/Y Y V
L
and the corresponding steady-state current /j(/)

when /^.(/) = 5>/2cos(100/) mA. This prob
lem can be solved by direct current division or
by source transformation and impedance con

cepts. Which method is easier?
Figure P i0.45
A N SW ER: 20 co s(1 0 ()0 r- 135) V
i(t)
46.
Consider the linear circuit o f Figure
PI 0.46, which operates at 50 Hz and for which

V, =
+ 1,1^2(a) Find the values o f a and b.
(b)
Figure P i0.42
If v^^{t) = 10 cos(lOOTCf) V and

200 sin(lOO)/^) mA, find vît).
43. In the circuit o f Figure P I 0.43, CO = 400

rad/sec,
= Rj = 2
L = 5 mH, and C =
625 |.iF. Find I^ând the corresponding

steady state. vp)= 12 cos(400^) V
in
j5on
-
J200Q
+ V. -
Figure P i0.46
AN SW ERS (in random order);
Fiaure P I 0.43
- 0 .8 . ./40 Q. - 1 6 cos(IO().-t O V
491
47. The linear circuit o f Figure P I 0.47 is such

that in the steady state, if
can do this with some straightforward
= 10 cos(2007U/)
reasoning without writing any equa
A with v^2 ^t) = 0, then
\t) = 20 cos(2007ir +
45) V. On the other hand if i^^{t) = 0 with
/;2 (/) = 10 cos(2007if + 45) V, then v^{t) = 5
tions.
cos(200)r + 900) V
(a)
Find a linear relationship between
v^2 >and
(b)
If"/^,(/) = 5 cos(2007tr - 45) A and
= 20 cos(2007rr) V, then in the

steady state find v^{t).
(c)
Find
and
Zj
Z - ,.
Develop simple cir
Figure P I0.48 AC Wheatstone bridge circuit.
cuit realizations o f these impedances

valid at OJ = 20071 rad/sec.
THEVENIN AND NORTON

EQUIVALENTS
49. Find theThevenin equivalent for the circuit
o f Figure PI 0.49 when R = AQ. L = 20 niH, C
= 1.25 mF,
= 2Z.45 A and (O = 200 rad/sec.
Be sure to express the open-circuit voltage as a
Figure P I0.47
ANSWHRS: (b) r,(/) = 18.46 cos(200nr -
time function.
22.5") V; (c) Z, = 0.763 + y2.6()5 H Z, = 4 Q
OA
48. The circuit o f Figure P I 0.48 is a general

Wheatstone bridge circuit (the dc version o f
which is described in Problem 35 o f Chapter
2). Here the circuit is used to measure the value
o f the unknown inductance L.
(a)
Suppose
= 0. Show that the steady-
state voltage i',^f(t) = 0 when
UC. Note: In general the condition for

a null voltage, v{t) = 0, in the steady
state is that the products o f the cross
impedances be equal.
(b)
Again suppose
that
= 0. You are given
= 2 sec and that the voltage
source v-^j^t) is a sinusoid with a fre

quency
of
rad/sec.
W ith
the
unknown inductance L inserted in the

circuit as shown, you adjust R[ until
you reach a sinusoidal steady-state
voltage null,
ing value for R^ is 3
(c)
= 0 V. The result
Find the value
ofZ,.
Now suppose R^ ^ 0. Show that the
condition o f part (a) is still valid. You
Figure P10.49 Parallel /?ZC circuit.

A N SW ERS: Z^;, = 4 i l , /-^.(/) = 8 cos(200/ +
45*) V
50 . For the circuit o f Figure P I0.50, let 1 = 1 0

mH, /?
mA.
(a)
=20 a, C = 20
îF, and
l- = 1 0 0 ^ 0
Find the Thevenin equivalent at the

terminals a and b if to = 2000 rad/sec.
(b)
If the circuit is terminated with a load

consisting o f a series connection o f a
20 Q resistor and a 20 mH inductor,
find the sinusoidal steady-state voltage
across the load.
492
OA
Figure P i0.53 Two coupled circuits.

54. The circuit o f Figure P 10.54 operates in the
SSS at cOq.
(a)
20 krad/sec and
= 2Z.0". Find the Thevenin
Find the Thevenin equivalent imped

ance
51. The circuit o f Figure PI 0.51 operates at to =

(b)
Find
if
10 a , ^ = 5 a
equivalent circuit (in the phasor domain) at ter
= 1^^^cos(w^/), R =
and
20 Q.
minals A and B. Use this Thevenin equivalent to

find the magnitude o f the phasor
when the
10 mH and 1 kH series combination load is con

nected to A and B.
i jt )
-jx
Figure P I0.54
55. In the circuit o f Figure PI 0.55, assume co =
100 rad/sec, I = 40 mH, C = 5 mF, R = S Q.,
and a = A Q.. Find the Thevenin impedance
Figure P I0.51
seen at terminals A and B. If

CH ECK : 190 < |V J < 205 V. |V^^| . 0.5| V J
52. For the circuit o f Figure PI 0.52,
Q, Cj = 0.2 pF,
= 20
cos(lOOf) V, find
= 1000
= 500 Q, and C , = 1 \i.
Find the Norton equivalent circuit when v- (t)
= 50 cos(4000f) V.
v.(t)
-OA
R,
Figure P I0.55
OB
GENERAL SSS ANALYSIS

(NODE OR LOOP ANALYSIS)
Figure P I0.52 Two coupled R C circuits.

53. For the circuit o f Figure P I 0.53 R = 2.5
k n . Find the Thevenin equivalent when
56. (a)
Find the phasors
in the cir
cuit o f Figure Pi 0.56 when
= 2 0 V 2 Z 4 5 " V, R = AQ., L = A
mH, and oj = 1000 rad/sec. Specify
the corresponding time functions.
10 cos(4000r) mA.
(b)
Determine the value o f oj for which

the magnitude o f the output voltage
phasor is 20% o f rhe magnitude o f rhe
(0
input voltage.
493
At to = 100 rad/sec, determine the

Thevenin equivalent circuit in which
Z^y^(/'100) is a series combination o f
two circuit elements seen at terminals
/Y Y V
L
A and B.
'6
-OA
2R
2R
;v,(t)
Figure P i0.56
C H E C K S: 5 A, 2o v
-O B
Y.(j)
57. In the circuit o f Figure P I0.57,
= y'30 Q,
Figure P I0.59
= - ; 4 0 Q, V^, = 28 V, V^2 =
^
sinusoidal sources have been operating for a
long time, and Z = 50 - j 4 0 Q. Find V^.
which v(,t) = 20 cos(lOOO^) V,
= 40 Q., Rj =
20 Q.,L = 20 m H .an d C = 7 5 \i.

(a)
Write and solve a nodal equation at

the top node for
6 ''
Then write the
corresponding time-domain expres

sion for v^t).
(b)
Figure P I0.57
z^ = y io a
z^^ = - y i o a
and then write the corre
sponding time-domain expression for
58. For the network o f Figure P i 0.58, a = 20
Calculate
^ = lo a
and
= 20 V. Find the phasor current I^..

Z,
the inductor current

(c)
Find the Flievenin equivalent circuit at

the source frequenc)' relative to terminals
A and B. Draw the Thevenin equivalent
circuit showing the Thevenin impedance
as a series circuit of two elements.
Figure P I0.58
AN SW FR: O.r. + 0.2/
59. Consider the circuit o f Figure P i 0.59 for
which C = 0.8 mF, r.(;100) = 0.01 + ;0 .0 4 S,
and v-^,(r) = 80 cos(lOOf) V.
Find R.
(a)
(b)
(c)
(d)
(e)
-
Find L.
Find
Find
At CO = 100 rad/sec, determine the
Thevenin equivalent circuit phasors
Woe and Z^;^(ylOO) at terminals A and

B.
Figure P I0.60
ANSWKRS: V^. = S - /S
= -/ 0 .:5 . Z./. =
1 0 - ; 10 12
61. The circuit o f Figure P i 0.61 operates at (o
= 2 krad/sec a n d =1. 5 mS with
= 10 0 Z 0
V. Find the Thevenin and Norton equivalent
circuits seen at terminals A and B.
494
determine the asymptotic behavior for large (0

/ 'A
J \
l.s f
and for at least one other frequency without a
0-251
computer or calculator. List these properties in
writing along with your reasoning.
Figure P I0.61
= ;1 0 0 0 Q,
C H EC K :
'
= 25 + ;2 5 V,
= 25 + y25 mA
62. Consider the circuit o f Figure PI 0.62. If i^{t)
Figure P I0.64
= 40 cos(lOOO^) mA, C, = 0.25 1-iF.

= 10 mS,
/?j = 1 kQ, and R-, = 5 kH, find the Thevenin
65. Compute the magnitude and phase o f the
equivalent circuit parameters
frequency response o f the circuit o f Figure
and
P I 0.65 where I = 25 mH and /? = 50 Q. Plot

your response in MATLAB (0 < 0) < 8000
rad/sec) and determine the frequency at which
' '
the magnitude is I/ V 2 o f its maximum value.
R,
Before sketching the responses, determine the

asymptotic behavior for large (0 and for at least
Figure P I0.62
C H EC K :
= 128 - ;3 3 7 .8 8 Cl,
1 4 0 .2 5 Z 9 5 .2 8 V
one other frequency without a computer or cal

culator. List these properties in writing along
with your reasoning.
63. This problem tests whether you can synthe

size ideas from two different parts o f the text.
In the circuit o f Figure P i 0.63, R = 20 Q, L =
(b
1 H, vît) = 50 cos(100r)(r) V (notice the step

function), and /^(O'*') = 1 A. If the response for
t > 0 has the form

i^{t) = A cos(l OOf + (j)) +
then determine the constants A, (j), X, and B.
ijt)
R
v,(t)
Figure P i0.65
66. Inside the black box o f Figure P I0 .6 6 a
there is a two-element circuit composed o f a
resistor o f 10 ^2, capacitors, inductors, or some
combination o f these elements. A variable-fre-
Figure P i0.63
FREQUENCY RESPONSE
64. Compute the magnitude and phase func
tions o f the frequency response o f the circuit o f
Figure P I 0.64 in which L = 4 mH and C = 0.25
mF. Plot your response in MATLAB (0 < CO <
5000 rad/sec). Before sketching the responses,
quency voltage vp) = lOcos(O)r) V is applied to

the box and the voltage v(t) =
cos(cor + 0) is
observed. A plot o f the magnitude o f v{t) with

respect to CO is given in Figure P 10.66b.
(a)
Draw the circuit contained inside the
(b)
box. (There are two solutions.)

Specify the element values.
v,(t) =
10cos(cot)V
Black
Box
v(t)
(a)
(b)
69. Reconsider the pressure-sensing example of
Figure P i0.66
section 10. Specify a set o f mesh currents and

67. Compute the frequency response o f the cir
write a set o f mesh equations that describe the
cuit o f Figure PI 0.67, where R= 100^2, L = 10
circuit. Solve the equations for 1 pF < 2 ^

pF using MATLAB or some other, equivalent
mH, C = 0.1 mF, and
is the output. Use
MATLAB or its equivalent to generate the

magnitude and phase (in degrees)
plots.
Consider 0 < (O < 3000 rad/sec.
software program. Plot the magnitude o f V ^ V âs a function o f Cj- Now construct a plot o f
the magnitude o f V ^ - V ^ a s a function o f pres
sure in mm Hg.
OP AMP CIRCUITS
70. (a)
Figure P I0.67
a single inductor. Let v^{f) be the input excita

tion and ip ) the circuit response. The magni
tude frequency response is given by Figure
P I0 .6 8 b . Draw^ the circuit inside the box and
assign component values if it is known that L =
40 mH.
when
= sin(200r)
mV for the circuit of Figure P10.70a.

(b)
For the circuit o f Figure PI 0.70b, find

C so that when
68. The box labeled V{joi) in Figure P I0 .6 8 a

contains a single resistor, a single capacitor, and
Compute
(c)
= cos(400r) mV,
= sin(400r) mV.
Find the phasor transfer function,
//(yco), and plot the magnitude o f the
frequency response (using iMATL^B
or the equivalent) as a function o f 03 =
Inf, where/is in Hz and to in rad/sec.
496
r>
100 kfi
20 kO
80 kO
o
n
(a)
(b)
Figure P I0.70 Op amp differentiation circuits.

(a)
Compute
when
=
sin(400/) V for the circuit of Figure
P10.71a.
For the circuit of Figure P I 0.7 lb, find
C such that when
=sin(500/) V,
= 5 cos(500/) V. This represents
an integration of the input with gain.
Find the phasor transfer function,
and plot the magnitude of the
frequency response (using M ATLAB
or its equivalent) as a function of O) =
In fi where/is in Hz and O) in rad/sec.
71. (a)
(b)
(c)
n
o
(b)
Figure P I0.72 Leaky integrator circuits.
r s
10|j F
H e-
50 kO
He-
200 kn
+
o
+
vjt)
JL .
(a)
(b)
Figure P I0.71 Op amp integrators.
73. (a) At (0 = 2 X 10^ rad/sec, find the

phasor voltage gain ôuPîn
P
circuit of Figure P 10.73.
(b) Find the phasor transfer function,
//(/to), and plot the magnitude of the
frequency response as a function of O)
= In f, where / is in Hz and (O in
rad/sec using M ATLAB or equivalent
software.
1kn
72. (a) If an 800 Hz sine wave of unit ampli
tude excites the leaky integrator circuit
of Figure PI0.72a, determine the
steady-state output voltage.
(b)
For the circuit of Figure P 10.72a, find
the phasor transfer function, //(/(o),
and plot the magnitude of the fi*equency response (using M ATLAB or
Figure P I0.73
its equivalent) as a function of (O =
271/where/is in Hz and O) in rad/sec.
74. For the circuit of Figure P i0.74, find the
(c) If the input to the circuit of Figure
expression for the phasor transfer function
P I 0.72b is
= cos(2000)r) V,
//(/ cd) =
Assume an ideal operational
determine the values of R and C so
amplifier. Plot the magnitude of the transfer
that v i t ) = 5cos(2000ti^ +135) V.
n
r^
o
O '
^
function as a function o f COusing M ATLAB or
^19'
the equivalent, assuming /? = 10 k fl and C =
0.01 mE
frequency increases to infinity? What

happens as the frequency decreases to
zeroi^
VO
o
o
Figure P I0.74
Figure P I0.76 Ideal op amp circuit.
75. In the circuit of Figure PlO.75 assume the

operational amplifier is ideal and that
= 25
ki2, Cj = 1 ^F, /?2 = 5 k2, and C2 = 0.2 pF.
Compute the gain of the circuit as a function of
(0. Then use M ATLAB or the equivalent to plot
the magnitude and phase of the fi-equency
response as the logarithm of the frequency for 1
< C0< lO'^ rad/sec.
77. Consider the circuit of Figure P I0.77 in

which
= 200 2, Cj = 0.05 |iF, /?2 = 28 IdQ,
and C2 = 0.05 pF.
Use nodal analysis to compute the
ratio
=I
^Hz. Now
use physical reasoning to obtain the
approximate the ratio at/= 1 Hz and
/= 100 kHz.
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
Figure P I0.75
76. For the operational amplifier circuit of
Figure P I0.76, /?j = 5 kQ, C, = 0.02 |iF, /?2 =
5 kfi, and C2 = 0.08 pF.
(a) Write two node equations and solve to
find a relationship between the output
phasor
and the input phasor
at the frequency/= 1000 Hz. Note
that the voltage from the minus termi
nal of the op amp to ground is
which equals the voltage from the plus
terminal to ground, assuming the op
amp is ideal.
(b) Repeat the calculation at/= 100 Hz
and/= 3000 Hz. What happens as the
Figure P I0.77 Op amp circuit having a band

pass type of response.
Note that
is an intermediary variable useful
in the nodal analysis of the circuit.
Sinusoidal State State

Power Calculations
The AM or FM receiver that is often part o f a home stereo system receives signals from radio sta
tions through an attached antenna. The intensity o f these signals or radio waves depends on the
power radiated into the atmosphere by the broadcasting station, the distance between the receiv
ing and transmitting antennas, and the design o f the receiving antenna. The intensity or magni
tude o f the signals picked up by the receiving antenna is very small. The power available from the
antenna and deliverable to the receiver is typically in the microwatt range. Again, this is ver)^ small.
Hence, it is important to have maximum power transfer from the antenna to the receiver input so
that the music signals received can be properly amplified and enjoyed. Since the signals in the
antenna are sinusoidal at very high frequencies, the antenna is represented by a phasor Thevenin
equivalent circuit as is the input circuit o f the receiver. Hence we must understand maximum
power transfer in the context o f sinusoidal steady-state analysis to describe and analyze this prob
lem. An example at the end o f the chapter illustrates some impedance matching techniques to
achieve maximum power transfer from an antenna to a receiver.
CHAPTER OBJECTIVES
1.
2.
Define and investigate the notion o f average power.

Define the notion o f the effective (rms) value o f a periodic voltage or current and its rela
3.
tionship to the average power absorbed by a resistor.

Define the notion o f complex power and its components average, reactive, and appar
ent power and investigate the significance o f each and their relationships.
4.
Introduce the notion o f power factor associated with a load and describe reasons and a
5.
method for improving the power factor.

Prove the maximum power transfer theorem for the sinusoidal steady-state case, and illus
trate its significance for the input stage o f a radio receiver.
500
Chapter 11 Sinusoidal State State Power Calculations
SECTION HEADINGS
1.
2.
3.
4.
Introduction
Instantaneous and Average Powers
Effective Value o f a Signal and Average Power
Com plex Power and Its Com ponents: Average, Reactive, and Apparent Powers
5.
Conservation o f Com plex Power in the Sinusoidal Steady State
6.
Power Factor and Power Factor C orrection
7.
M aximum Power Transfer in the Sinusoidal Steady State
8. Summar}^
10. Problems
1. INTRODUCTION
Chapter 1 defined the concept o f power. The following chapters were primarily devoted to the cal
culation o f voltages and currents. This does not mean that the consideration of power is o f sec
ondary importance. The very opposite is true. A homeowner pays for the energ}' used, not for volt
age and current. The integral o f power over, say, a 30-day period determines the household ener
gy consumed in a month. Hidden in the homeowners cost is an adjustment to cover the power
losses incurred in transmitting energy from the generating station to the home. Thus power con
siderations have a significant impact on everyday life.
A second reason for understanding ac power usage is safety. Each appliance, and its cord that plugs
into the wall outlet, has a maximum safe power-handling capacit)'. Misunderstanding such infor
mation and/or misusing an appliance can lead to equipment breakdown, fire, or some other lifethreatening accident.
Even for electronic equipment in which power consumption is low, such as laptops and handheld
PDAs, power consumption and, thus, battery life are important design factors. Power drainage direct
ly determines the PDAs operating time before the battery needs recharging. In fact, optimizing power
management in laptops and hybrid electric vehicles is an important research area in todays world.
In this chapter we will investigate different notions o f power in ac circuits and discuss their sig
nificance and application. The term ac circuits has a narrow meaning here. It refers to linear cir
cuits having all sinusoidal sources at the same frequency and consideration o f responses only in
steady state. The basic analysis tool is the phasor method o f Chapter 10.
2. INSTANTANEOUS AND AVERAGE POWERS

Figure 11.1 shows an arbitrary two-terminal circuit element isolated from a larger circuit. With
the voltage (in V) and current (in A) having indicated reference directions, the instantaneous
power (in watts) absorbed by the element is given by equation 11.1:
p{t) = v{t)i{t)
(11.1)
501
FIGURE 11.1 Instantaneous power delivered to an arbitrary two-terminal element.

Evaluating battery life or the length o f operation o f your cell phone involves consideration o f a
quantit)' called average power, P (or
for emphasis), defined as the average value o f the instan
taneous power over an interval [T*!, T^. The idea is based on the average value o f a function, say
j{t) , which is defined as

Le =
T-,
T
h - TM r,J
Using this idea we define the average power consumed by a rwo-terminal element as shown in
Figure 11.1 over the interval [T j, Tj] as
Pave(TiJ2) =
T,
piOclt
r .-r ,
( 11. 2)
When the signal is periodic with period T, we speak o f the average power consumed by an ele
ment over the period T as
T
T
(11.3)
It is not necessary that T b e the fundamental period; the evaluation o f the integral is the same for
any integer multiple o f the fundamental period.
EXA M PLE 11.1. Compute the average power absorbed by the resistor R connected to an inde
pendent voltage source as shown in Figure 1 1.2b with the excitation shown in Figure 1 1.2a.
v.(t) f +
F IG U R E 1 1 .2 Triangular voltage waveform
driving resistor R .
S{)2
So
lu t io n
Step 1. Compute the instantaneous power for 0 < f < 7q. Here
IV J
P {t) =
0 < / < 0 .5 ^
o .5 r < / < 7 ;
Step 2. Compute
Using equation 11.3 and observing that the Fundamental period is Tq, we
have
0.57n
/
>ave
p {t)d t =
rj.
TnR
6R
Exercises. 1. Suppose the sawtooth in Figure 11.2a does not drop to zero at r = 0.5 T'q , but rather
continues to increase until reaching ^ = Tq when it drops to zero and repeats. Find the average
power consumed by R.
AN SW ER:
3/^
. K?
2. Show that the average power absorbed by an R Q resistor in parallel with a Vq V dc source is
over any time interval [T'j, 7'^].
^
O f particular importance is the average power consumed by devices in the SSS assuming all exci
tations are at the same frequency, to. Consequently, all voltages and currents are sinusoids at the
sa7nefrequency. To compute the average power absorbed by a circuit element as depicted in Figure
11.1 (assuming a linear circuit), suppose v{t) =
cos(u)t + 9^^ and i(t) =
cos(a)r + 0^) . The
associated instantaneous power is
p ( f ) = v(f ) i { t ) = V, cos(co/ + e^,) X /, cos(cor + 0 ,-)
(11.4)
=
cosO , -
e,-) +
cos( 2 (o/+ e,, + e,.)
Equation 11.4 follows from the trigonometric identit)' cos(x) cos(y) = 0.5 c o s (x - y) + 0.5 cos(a' +
y). Observe that the instantaneous power o f equation 11.4 consists o f a constant term plus anoth
er component varying with time at tivice the input frequency. Figure 11.3 shows typical plots o f
/>(/), v{t), and i{t).
Chapter 11 Sinusoidal State State Power (lalculations
503
0.01
0 .005
0 .0 1 5
0.02
t in secs
FIGURE 11.3 Plots o f tit) = 10 cos(377f) A, v{t) = 2 cos(377/ + 45) V, and p{t).
Using equation 11.3 witii T = 23t/ol), and observing that the integral o f a sinusoid over any peri
od is zero, we obtain the following formula for average power in SSS:
T
_
*Pnvp =
co s(0 , - 0 ; )dt + ^ j c o s ( 2 c o r + 0 , + 0,-)d! =
c o s(0 , - 0 , ) (11 -5)
If the two-terminal element is a resistance R, then v{t) = Ri{t) and 0^^ - 0y = 0 . It follows from
equation 11.5 that for a resistor
p
_ VmI m _ RI^m _ m
âve,R
2
2
2R
If the two-terminal element is an inductance L, then
âveL ~ ^
( 11.6)
= (/(.oZ,)!^ and 0^, - 0^ = 9 0 . Hence,
cos(90) = 0. Similarly, if the two-terminal element is a capacitance C, then
c~ ^
^neans that the average power consumed by
or delivered by a capacitor or au inductor is zero. Even though an ideal capacitor and an ide;il induc
tor neither consume nor generate average power, each may absorb or deliver a large amount ot
= (/coQV^ and 0^ - 0y = - 9 0 . Hence,
instantaneous power during some particular time inten-al.

Before closing this section, we need to investigate the question o f superposition o f average powder.
Is there a principle o f superposition o f average power? If so, when is it valid? When is it not valid?
The following example provides the answers.
504
EXA M PLE 11.2. Consider the circuit o f Figure 11.4, which consists o f a series connection o f two
(sinusoidal) voltage sourccs in parallel with a 1 Q resistor. For this investigation v^{t) =
cos(ojjr + 0|) (having fundamental period /'] = 2:t/tO|) and v-,(f) =
cos(t02^+ B2) (having fun
damental period T-, = 271/(0,). For simplicit)^ we assume that v^{t) and v-,{t) have a common peri
od o f 7 'seconds, i.e., there exist Integers ni and ;/ such that T = uT^ = niT^.
v,(t) Q
1n
< v_(t)
FIGURE 1 1.4 Circuit for investigating superposition o f average power.

So
lu t io n
Compute the average power consumed by the 1 Q resistor. First observe that the power consumed
by the 1 Q resistor with source 1 acting alone, i.e., v-j{t) = 0, is
1
1
0
Also note that the power consumed by the 1 H resistor with source 2 acting alone, i.e., u^{t) = 0,
is
I T
, r ^
0
= v^{t) +
W ith both sources active, linearit)^ (or KVL) implies that
(^ 1 (/)
Pave = - \ V R {t)i,i{t)d t = - \
0
. By equation 11.3,
+ v'2 (r )) dr
vf(r)f/^ + - J v ? ( r V r + - Jv| (t)v2{t)dr

0
T
ôve.l
â\r,2 '
^ Vj (/)V2(/)f/f
0
âve,\
= Pave.\ + f'ave.l + '
ôve.l
COS((Oi /+0| )COS(C02/ +02)<^/^
j [ c o s ( ( ( 0 | + C O 2 )/ + (0 i + 2 ) ) + C 0 s ((0 3 , - ( O 2 ) / + (0 i - 0 2 ) ) ] ^ / ^
0
S()S
When the integral term in this last equation is zero, then
, indicating that
superposition o f average power holds. When this integral term is nonzero, superposition o f aver
age power does not hold. The next question is, under what circumstances is the integral zero and
nonzero.^ There are three cases to consider. Case 1 is when cuj
co^ , which will result in a zero
value o f the integral. In this case, the integral consists o f rwo sinusoids integrated over a common
period T. The integral of a sinusoid over any period is zero. Thus, the integral is zero and super
position o f power holds when cOj ^ O)-,.
Case 2 is when C0 | = co-, but with (Bj - O2) = knl2, k an odd integer. In this case, the integral is
again 0 . This follows because the first term o f the integrand is a sinusoid whose integral is zero
over the period T. The second term o f the integrand is a constant, cos(0j - (),) = cos{kKl2) = 0 ,
also resulting in a zero integral. Hence for case 2, superposition o f power holds.
Finally, we have case 3, for which tOj = co-, but with (6 j - 62 ) ^ kKll, k an odd integer; here
superposition ot power does not hold. The second term of the integrand is a constant, cos(0j - 62 )
0 , resulting in a nonzero integral over the period T. So
P^^^^ j +
For case 3, it is
desirable to use the phasor method o f Chapter 10 to compute the desired voltage and then use
equation 11.5 to compute average power.
Exercises. 1. In Example 11.2, suppose t/j(r) = 3cos(107rr) V and V2 {t) = 4 cos(15n:r + 0.25ti) V.
Compute T, a common period for the two sinusoids, and then compute the average power con
sumed by the 1 Q resistor.
C H EC K ; T = 0.4 sec w-ill work, and
2. In Example 11.2, suppose
age power consumed by the 1
= 12.5 watts
= 3cos(107Tf) V and
= 4 sin(lOJur) V. Compute the aver
resistor.
C H EC K ; P^..=
ave 12.5 watts
3. Now suppose v^{t) = 3cos(10Tcr) V and
=4
c o s
(1 5 7 T /
+ 0.257i) V. Compute the average
power consumed by the 1 H resistor.

C H EC K ; P^^^ = 20.99 watts
Equation 11.6 resembles equation 1.18b for the dc power absorbed by a resistor connected to a
dc source. However, in equation 11.6 the factor 1/2 is present. With the introduction o f a new
concept called the efifective value o f a periodic waveform, the formulas for the average power
absorbed by a resistor can be made the same for dc, sinusoidal, or any other periodic input wave
forms.
S06
3. EFFECTIVE VALUE OF A SIGNAL AND AVERAGE POWER

GENERAL CONSIDERATIONS
From section 2, a resistor o f R ohms excited by a periodic voltage or current absorbs an average
power,
. T he effective value o f any periodic current, i{t), denoted by
is a positive con
stant such that a dc current o f value /^-exciting the resistor causes the same amount o f average
power to be absorbed, i.e.,
The same holds for a resistor excited by a periodic volt
age v{t). Mathematically,
(11.7a)
or
(11.7b)
R
Equation 11.7a suggests that
1
Iq+T'
I f " )
T
Hence, the mathematical definition o f the effective value of a periodic current i{t) is
/m+7
U>ff.R
(11.8a)
and, similarly, the effective value o f a periodic voltage u{t) is

h\+T
ef/M
(11.8b)
In general, the effective value o f any periodic signal y(r) is

In + T
F e jf-
(11.8c)
Observe that the expressions under the radical sign in equations 11.8 constitute the average value
ot the square o f the signal. Hence, the expressions give rise to the alternative name for the effec
tive value, the root-m ean-square (abbreviated rms) value oiJ{t), since
mean value o f the square j{t) over one period.
the square rooto'i the
SO'
Exercises. 1. Show that the average power absorbed by an R Q resistor carrying a periodic current
2. Suppose i{t) = 3cos(2n:/) + 4 cos(47i:f) A flows through a I Cl resistor. Find

ANSWlvR:
= 1 2 .5 watts, and l^,g- =
7?
= 2.5\/2
and
EXA M PLE 11.3. Compute the efFective value o f the periodic voltage waveform sketched in
Figure 11.5.
So
lu t io n
From equation 18.b,

= T
Therefore,
1 1
.2
4J
4J
4J
= 2.3094 V,
Exercise. Repeat the calculation o f Example 11.3 for the case where the values on the vertical axis
o f Figure 11.5 are doubled.
A N SW l'R :
For a sinusoidal signal/r) =
cos(tor + 0) , the effective value can be calculated using the identit}'

cos^ (.v) = 0.5 + 0.5 cos(2.v)
as follows:
f~{t)
Fm
-
= F jC o s ( o j/+ 0 ) =
F~m
c o s (2 o j/+ 2 0 )
S08
Since by assumption to 7^ 0, the average value o f the cosine term is zero. The average value o f the
first (constant) term is itself Hence, by equation 11.8c,
F~
I itt
(11.9)
Eml
72
Thus, for a sinusoidal waveform, the effective or rms value is always 0 .707 times the maximum
value or, equivalently, the ma.ximum value divided by >/2 a basic fact well worth remembering.
The ac voltage and current ratings o f all electrical equipment, as given on the identification plate,
are rms values unless explicitly stated otherwise. For example, the household ac voltage is 110 V,
with a maximum voltage o f 1 lOx-s/^ = 1 5 6 V. A typical appliance such as a coffee maker will
have a 110 V rating, ac, at say, 900 watts. The effective values o f a few other periodic waveforms
are listed in Figure 11.6, with their derivations assigned as exercises.
Feff
=F
dc
sinusoidal
triangular
square
F.= F .
F IG U R E 1 1 . 6 Efleccive values o f some com m on periodic waveforms.
S09
Exercises. 1. Derive the formula for riie cfFectivc value o f a triangular waveform shown in Figure 11.6.
2. Compute the effeccive value of the waveform shown in Figure 11.2a.
AN SW FR:
SINGLE-FREQUENCY ANALYSIS WITH EFFECTIVE VALUES

We recurn now to the case of single-frequency SSS analysis. The average power as per equation 11.5
absorbed by an arbitrary rwo-terminal element mav now be rewritten in terms o f effective values:
cos(0,, - 0y) = - ^
V2 v 2
COS(0^. - 0,-) = yeffêjf COS(0,, - 0, )
For the remainder o f the chapter, all voltage and current phasors will be taken as being effective
values unless the subscript m or
appears, indicating the maximum value. The subscript ^ w ill
be added sometimes for emphasis, however. This practice is widely accepted in the power engi
neering literature. Omitting the subscript eff\v\ equation 11.10 yields
K / c o s (0 ^ -0 .)^ V7cos(0p
where 0^ = 0,^ - 0 y , V= 0.707
( 11.11)
, and / = 0.7071^^ . The angle 0^ is the angle o f the impedance
Z (/co) o f the two-terminal element and is also interpreted as the angle by which the voltage phasor leads the current phasor.
EXA M PLE 11.4. Figure 11.7 shows two t)'pes o f household loads connected in parallel to a 110
V, 60 Hz source,
= 1 1 0 ^ 2 cos(120:tr) V. Lainp 1 and lamp 2 have effective hot resistances
o f 202
(a)
(b)
and 121
respectively. The impedance o f the fluorescent light is Zjfjwi) = 60 + j70
Find the average power consumed by each light.

Find the average power delivered by the source.
Lamp 1
Lamp 2
FIGURK 1 1.7 An example of load current calculation.

So
lu t io n
(a) For lamp 1, Z|(/to) = 202Z.0 Cl. Hence, Ij = Vy^Zj = 0.5446Z .0 A. From equation 11.11,
^\ave ^
z\^ = 110
X 0 .5 4 4 6 cos(0) = 59.9 watts
5 10
Cliapier 11 Sinusoidal State State Power Calculations
This means that lamp 1 is a 60 watt bulb.

Similarly, for lamp 2, Z^C/co) = 121Z.0
Hence, 1, = ^ iJZ j = 0.909 IZ.O^ A. From equation
11. 11,
= 110 X 0.9091 cos(0) = 100 watts
= 56 + y'66 = 86.56Z.49.7 O.. Hence,
Finally, for the fluorescent light,
^ jJZ jj =
1.2 7 Z .-4 9 .7 A. From equation 11.11,
^Jlave =
(b) For this part we first compute
1.27cos(49.7) = 90.4 watts
and then apply equation 11.11 to compute the average power
delivered by the source. Here by KCL,

= I, + I 2 + I 3 = 0 .5 4 4 6 Z 0 " + 0 .9 0 9 1 ^ 0 + 1 .2 7 ^ -4 9 .7
= 2.2759 -y 0 .9 6 9 0 = 2 .4 7 3 6 -^ -2 3 .0 6 A
By equation 11.11, the average power delivered by the source is
Pave = |V/,,l|lm|co.s(0, - 0 , ) =
cos(6 , - 0 ,- ) = 110 X 2 .4 7 3 6 cos(23.06^ ) = 250.35 watts
Observe that the sum o f the individual average powers is 250.3 watts, which equals the power
delivered by the source within the accurac)' o f our calculations, where we have rounded our
answers.
4. COMPLEX POWER AND ITS COMPONENTS: AVERAGE,

REACTIVE, AND APPARENT POWERS
Recall the notion of a phasor. When all source excitations are sinusoidal at the same frequency,
voltages and currents in the SSS can be represented by phasors. Our question here is, can the pha
sor method aid the computation o f power consumption in a circuit? The answer is yes. However,
the formulation will bring out several other concepts o f power associated with the sinusoidal
steady state.
In dc power calculations, the average power consumed by a two-terminal device is the product o f
the voltage and current, assuming the passive sign convention. In SSS, the complex power
absorbed by a two-terminal device, as shown in Figure 11.8, is a complex number defined by the
formula
(1 1 .1 2 )
where I^^ is the com p lex conjugate o f
51 1
eff
o +
Two
Terminal
Veff
Device
o-
FIGURE 11.8 Two-terminal device with phasor voltage and current consistent
witli passive sign convention.
The first useful result o f this definition is that

suppose v(r) =
i(t) =
= Re S = Re
. To see this result,
cos((Of-I-0J.) , which is represented by the phasor
y/l cos((or + 6,-) , which is represented by the phasor
Also suppose
The average power consumed by
a rwo-terminal device excited by this voltage-current pair (Figure 11.8) is given by equation 11.10 as
Now obseiAê that
S = \ //C /f =
j =
h r
in which case
= Re
R e[S]= R e
The curious reader may ask why a conjugate o f the current is used in the definition o f complex
power.
Suppose
Re
one
did
not
have
the
conjugate
o f the
current.
Then
= VeffI(,ffCOs{Q^,+Qj)Î\,y^, i.e., the resulting product would have no physical
meaning. Now because S is a complex number, it has an imaginary part, a magnitude, and an
angle. T he imaginary part o f S defines a quantity called the reactive power absorbed by the twoterminal device in Figure 1.18; i.e., reactivepoioer \s defined as
Q = Im [S]=
J=
sin(0 ,. - 0,) = reactive power
14)
The unit o f reactive power, Q, is VAR, which stands for volt-amp-reactive. It follows immediately
that
where P =
. Also, the magnitude o f S is defined as the apparent power absorbed by the two-
terminal device o f Figure 1.18, i.e.,
________
\^\ = Kffêff = \P ^ JQ\ = 4 p ~ ^Q~ = apparent poNver
^12
The Linir o f apparenr power is VA, short For volt-arnp. The iinerrclationship o f these different pow
ers is illustrated by the right triangle diagram in Figure 11.9, which is often helpful in solving
problems. Observe that the apparent power is always greater than or equal to the average poiuer,
with equality applying to the case o f a purely resistive load.
FIGURE 11.9 Relationships among complex, average, reactive, and apparent powers.
I'h e distinction among these various powers is best understood by computing the powers for some
basic circuit elements. For simplicit)', except when needed or for emphasis, from this point on we
will often drop the subscripts <ând rfw'as given in equations 1 1 .1 3 -1 1 .1 6 .
EXA M PLE 11.5. This example explores the computation o f the various powers for a simple
inductor. Given that 1^0 ) = v2/sin(CO/) in the circuit of Figure 11.10, compute
V^, S^,
Ql, the instantaneous absorbed power />/(/), and the instantaneous stored energ)' Wjit) in terms
o f L, CO, /^, and Vj. After this show that
(i)
max
Ql
(ii)
Wiit) max
Ql
(0
Remainder
of Circuit
FIG U R E. 1 1 . 1 0 Isolation o f an inductor for investigating the concept o f com plex power.
So
SI 3
lu t io n
= -jl^,
= /oZ.1^ = cdZ./^ = V^.
jVJi^ = I\ + yQ^. This implies that
= 0 and (X = V Ji- Further, the instantaneous
absorbed power is Pi ( f ) = v
=
COLy/lIi^cosUor)x J l l i sin(o)/) = V^// sin(2(0/), which
By inspection, and noting that wc again presume effective values,
= VÎ
,a\ ~
is consistent with equation 11.4. It follows immediately that
~ Ql Further,
W' l it ) = i ) . 5 L i l u ) = L I I sin -((o ;) = 0.5 L/; [l - cos(2w/)]

9
[l-c o s (2 c o / )]
[l-c o s (2 o )/ )]
-------------- ^l l -------- z-----------=
2co
2(0
[ l - c o s ( 2 (0 /)]
----------2(0
(f)
Since the bracketed quantity varies between 0 and 2,
\Qi\
0)
, as was to be shown.
EXA M PLE 11.6. This example, like the previous one, investigates the concept o f reactive power,
but in the case o f a capacitor. The calculations will all be dual to those o f Example 11.5. Hence,
given that ^ ^ ( 0 = >/2V(-sin((0/) in the circuit o f Figure 11.11, compute
instantaneous absorbed power
S^^
and the instantaneous stored energ)^
'r
Qq the
terms o f C,
(O ,
Vq and Ir-. After this show that

(i)
P c (0
(ii)
lV c (0
Qc
and
Qc
(0
Remainder
of Circuit
FIGURE 11.11 Isolation of a capacitor for investigating the conccpt of complex power.
So
lu t io n
^C~
By inspection, and noting that we again presume efteaive values,

implies that P r= 0 and Q_(^ =
S^=
absorbed power is /;^(/) =
which
is
consistent
with
Further, the instantaneous
)/(-(/) = (oC>/2\^(7 cos((0/) X V 2 V^7 sin((0/) =

sin(2(0/),
equation
11.4.
It
follows
immediately
that
= | 0 c L Funher,
\V^(i) = 0.5Cv^(/) = CVc sin (03/) = 0.5CVf^ [l - cos(2(0/)]
^ ,,2 [l-c o s (2 (o / )]
(oCVr
------------------ 2w
,, ,
= Vr^c
[l-c o s (2 (o / )]
-----------------2(0
= Qc
[ l - c o s ( 2 (0 /)]
-----------------2(0
^\A
Qc
Since the bracketed quantity varies between 0 and 2, W(-(/)
These quantities,
(0
useful for identify
ing energy storage values in inductors and capacitors in systems where energy is to be recovered
and stored, and for modifying the power factor (to be discussed shortly) in networks with motors.
Energy storage in systems and power management are important research topics in todays world.
In Examples 11.5 and 11.6, one observes that the inductor absorbs reactive power while the capac
itor absorbs negative reactive power or, equivalently, delivers reactive power. This follows from the
definition o f complex power (equation 11.13, i.e., S =
T he structure o f equation 11.13
derives from the convention that whenever the phasor current lags the phasor voltage (as with the
inductor), the device is considered to absorb reactive power, whereas if the current phasor leads
the voltage phasor (as with the capacitor), the devicc is considered to deliver reactive power.
Indeed, the overwhelming majority o f loads (toasters, ovens, hair dryers, motors, transformers,
TV s, etc.) have lagging currents.
When a t%vo-terminal element absorbs an average power
, there is a transformation o f elec
trical energ)' into other forms o f energ}' for example, heat or kinetic energy. In contrast, when a
two-terminal element absorbs reactive power Q, no energy is expended. T he energy transferred
into the two-terminal element is merely stored and later returned to the surrounding network. To
distinguish it from real (expended) power, we use VAR (volt-ampere-reactive) instead o f watt as
the unit for the reactive power Q.
EXA M PLE 11.7. This example investigates the computation o f the various powers defined above
for an /?Ccircuit. Here, consider the circuit o f Figure 11.12, where v,(/) = 100 V 2 c o s (200071:/)
V. Find the complex, average, reactive, and apparent powers absorbed by the load.
100 Q
lO kO
16nF i
FIGURE 11.12 Simple RC circuit for investigating aspects of complex power.

S o lu tio n
S te p 1 .
Compute
Z,ry20007Cj = 100
j -I- 720007T X 16x10
10
-9
Step 2. Compute
Converting
SI^
to a phasor, we have
I,v, = ^
= 10 + j9 .8 5
= 100 V. By O hm s law,
mA
în
Step 3. Cornpute the complex power absorbed by the load. By equation 11.12,
S =
= 100(10 - y9.85)10"-' = 1 - jO .985 VA
Step 4. Given the complex power, the average power is
The reactive power is

Q = I m [ S ] = - 0 .9 8 5 VAR
and the apparent power is
|S| = 1.404 VA
Before doing a more complex example, we will discuss the particulars o f the principle o f conser
vation o f power in the sinusoidal steady state.
5. CONSERVATION OF COMPLEX POWER IN THE

SINUSOIDAL STEADY STATE
Basics and Examples
The basic principle o f power conservation is that instantaneous power is conserved.
GENERAL PRINCIPLE OF CONSERVATION OF POWER

In all circuits, linear or not, instantaneous power is conserved; i.e., the sum o f the absorbed
powers o f all the elements in a circuit is zero. If one thinks o f sources as generating power and
other elements as absorbing power, then we can rephrase this statement as the sum o f gen
erated powers equals the sum o f absorbed powers.
The validity o f this principle follows from KVL and KCL. This principle leads to the particular
fact that complex power is conserved in ac circuits operating in the SSS.
PRINCIPLE OF CONSERVATION OF COMPLEX POWER IN AC CIRCUITS

In ac circuits operating in the SSS, complex power is conserved; i.e., the sum o f the absorbed
complex powers o f all the elements (operating in the steady state) in a circuit is zero.
Consequently, average power is conserved and reactive power is conserved.
Note however, that the conservation principle does not hold for apparent power, i.e., for the mag
nitude o f the complex power. The following example illustrates a basic use o f the conservation law.
S I6
EXA M PLE 1 1.8. This example illustrates the application o f the principle o f conservation o f com
plex power in determining power delivered by a source and the input current to a circuit. We also
show that conser\'ation o f apparent power does not hold. Consider the circuit o f Figure 11.13.
Find the power delivered by the source and the phasor input current,
given that S j = 360 +
;1 6 0 VA, S 2 = 360 - ; 1 2 0 VA, S 3 = 420 + y540 VA, S 4 = 130 + >80 VA, S 5 = 40 - ; 1 0 0 VA.
/
/
lOOV
V- \
S.
\ A
FIGURE 11.13 Bridge circuit where S- represents the complex power absorbed by the element.
S o lu tio n
By the principle o f conservation of power in ac circuits,

= S , + S , + S 3 + S 4 + S 5 = 1310 + j5 6 0 VA
This means that the circuit absorbs 1310 w'atts o f average power; the reactive power is 560 VAR,
and the apparent power is 1425 VA. Notice that the large component o f reactive power makes the
apparent (consumed) power larger than the actual consumed power, P
To compute I , recall that

^so u rce
Hence,
= 'OOI/,, = 1 3 1 0 + p 6 QVA
= 13.1 - J5.6 A.
Exercise. Repeat the above example calculations for S j = 300 + y'400 VA, S 2 = 300 - y 4 0 0 VA, S 3
= 600 + j\ 000 VA, S^ = 60 + y80 VA, S^ = 120 -> 1 6 0 VA. What are the average and reactive pow
ers delivered by the source?
AN SW FR: S^
= 1380 + /920 VA :uid
= 13.8 + / ).2 A. 1380
w ;itts
and 020 VAR
The next example illustrates the computation o f various powers through basic definitions and
application o f the principle o f conservation o f power.
EXAM PLE 11.9. Consider the circuit of Figure 11. 14, which depicts a motor connected to a commer
cial pow'er source. The motor absorbs 50 kW of average power and 37.5 kVAR o f reactive power, and has
a terminal voltage
= 230 V. Find IIj, the complex power delivered by the source, S^, and IVÎ.
R,line = 0 .5 0
FICfURE 11.14 Motor absorbing 50 kW and 37.5 kVAR at a terminal voltage o f 230 V; the value of
^line
exaggerated for pedagogical purposes; electrical code requires that the size of the connecting
wire be large enough that the voltage drop is only a small percentage of the source voltage.
S o l u t io n
Step 1. FiJid the apparent power, |S^^J, absorbed by the motor. Since
S, = ^V-Q = 50t/-37.5kV A
it follows that |S,^J = 62.5 kVA.
Step 2. Find |/J. Here, |S,J =
= 230| lj. Hence, [Ij = 2 7 1 .7 4 A.
Step 3. Compute the line loss.

0-5 X 271.72 = 36 .9 2 kW
Step 4. Compute the complex power delivered by the source. From conservation o f power,
S,
= S, . s,,. = S, *
= 5 0 ;3 7 .5 . 3 6 .92 = 8 6 .9 2 *y 3 7 .5 kVA
Step 5. Compute |Vj.

s.
V. =- *
h
In the above example we choose
s.
= 3 4 8.4 V
I.
large to illustrate the calculations. In practice a line loss of
36.92 kW for a 50 kW motor operation would not be permitted.
6. POWER FACTOR AND POWER FACTOR CORRECTION

In a resistor, average power is dissipated as heat. In a motor, most ol the average consumed power
is converted to mechanical power, say, to run a fan or a pump, with a much smaller portion dis
sipated as heat due to winding resistance and friction. The ratio o f the average power to the appar
ent power is called the power factor, denoted by pf, i.e..
pf=
Average Power
P^ye
Apparent Power
|S|
= cos(0,-0,)
(11.25)
SIS
The right-hand portion o f equation 11.25 follows directly from equation 11.13. Equation 11.25
specifies the power factor as cos(6^, Oy) , i.e., the cosine o f the difference between the angles o f
the voltage phasor V and the current phasor I. Clearly, 0 < pf < 1 . The angle
(11.26)
- 0y) = power factor angle (pfa)
Since cos(x) = cos(-a:), the sign o f (0^^- 0^) is lost when only the pf is given. In order to carry the
relative phase angle information along, the common terminology is p f lagging ov p f leading. A l/igging power factor occurs when the current phasor lags the voltage phasor, i.e., 0 < (0^, - 0y) < 180 .
A leaditjg power factor occurs when the current phasor leads the voltage phasor, i.e., 0 < (0y - 0^j <
180 . Practically all t)^pes o f electrical apparatus have lagging power factors. Some typical power
factor values are listed in Table 11.1.
TABLK 11.1. Power Factors for Common Electrical Apparatus
T
ype o f
L oad
P o w e r F a c t o r ( L a g g in g )
Incandescent lighting
1.0
Fluorescent lighting
0 .5 -0 .9 5
Single-phase induction motor, up to 1 hp
0 .5 5 -0 .7 5 , at rated load
Large three-phase induction motor
0 .9 -0 .9 6 , at rated load
To illustrate the idea o f leading and lagging pf, consider the circuits o f Figure 11.16. Suppose the
circuits operate at a frequency o f 400 Flz or co = 2 5 13.3 rad/sec. For the circuit o f Figure 1 1.6a,
I = (1 -jO.995) 10^ V = 1.41 10^/.44.85 V. Hence, the current phasor lags the voltage pha
sor, i.e., ( 0 J ,- 0y) = 44.8 5 and the pf is cos(44.85) = 0.709 lagging. On the other hand, for the
circuit o f Figure 11.16b, I = (1 + ;2 .5 ) 10-^ V = 2.7 103 Z 6 8 .3 V. Hence, the current phasor
leads the voltage phasor by 68.3", i.e., (0 0^^ = 6 8 .3 and the pf is cos(68.3) = 0.688 leading.
o-
O-
i0 .4 H
IH F
1 kO
1 kQ
O-
(a)
(b)
FIGURE 11.16 (a) A parallel RL circuit illustrating a lagging pf.

(b) A parallel RC circuit illustrating a leading pf
A load with a required average power demand, operating at a fixed voltage with a low pf, say 0.6,
has a relatively high reactive power component. This results in a relatively high apparent power.
SI 9
Since the operating voltage is fixed, the line current needed to drive the load is higher than if the
load operated at a higher pf, say 0.95. Relatively speaking, a higher pf has a lower reactive power
component with correspondingly lower apparent power. Figure 11.9 helps to visualize the rela
tionships. For fixed line voltage, lower apparent power (higher pO means lower line current and
hence lower power loss in the connecting transmission line. In todays world o f energy conserva
tion, it is important to be energy efficient. The following example illustrates how improved pf on
a load can reduce line losses and thus decrease cost o f operation.
EXA M PLE 11.10. This example reconsiders Example 11.9, involving a motor connected to a
commercial power source as illustrated in Figure 11.17. The solution process will emphasize the
basic definition o f pf and the use o f voltage and current phasors. Suppose the motor absorbs 50
kW (about 67 hp) o f average power at a pf o f 0.8 lagging. The terminal voltage,
is 230 V. The
frequency o f operation is 60 Hz or co = 120 j1 . For the first part o f the example the capacitor in
Figure 11.17 is not connected to the motor. In part (c), the capacitor is connected to the motor
to improve the p f This will reduce the magnitude o f the current supplied by the source and hence
reduce the line losses.
(a)
(b)
Find the complex power delivered to the motor.

Find I., V^, and the power delivered by the source, which might represent the power
delivered by the local electric company.
(c)
Correct the power factor o f the combined motor-capacitor load to 0.95 lagging by choos
ing a proper value for C.
(d)
Compute the new power delivered by the source to the combined motor-capacitor load.
R,line =0.5 0
FIG U RE 11.17 Motor absorbing 50 kW and 37.5 VAR at a terminal voltage of 230 V. Again, the
value of
is exaggerated for pedagogical purposes; clectrical code requires that the size of the con
necting wire be large enough that the voltage drop is only a small percentage of the source voltage.
S o l u t io n
(a) Find the complex power delivered to the motor.

Step 1. Use the p f o f 0.8 lagging and the given average power to fin d the apparent power. From the
definition o f pf,
= 50 kW = R e[S] = |S,|cos(e -
) = |S,| x p f = |S,| x 0.8
As such, the apparent pow-er is

50
!s,
0. 8
= 6 2 .5 k V A = \/,/,
( 1 1 .2 7 )
Chapter 1 1 Sinusoidal State State Power Calculations
S20
where V, = |VJ = 230 V and /, =

Step 2. Compute
. Ligging means that current phase lags behind voltage phase, i.e., 0 ^ -
= 0 - 0^ > 0 . Consider the diagram in Figure 11.18, which shows that the current phasor I^ la g s
the voltage phasor, i.e., the current phasor makes an angle o f - 3 6 .8 7 = cos ' (0.8) from the volt
age phasor. Hence,
= 3 6 .8 7 > 0
S m= V IDP m
FIGIJRI- 11.18 Phasor relationship ofV^^,

Step 3. Compute the complex power, S
and
. By definition,
s,,, = | S , k S m =
= 6 2 .2 Z 3 6 .8 7 kVA
= 5 0 + ;3 7 .5 k V A = /^t>Q
(b) Fi7ici
, V., and the power delivered by the source.
Step 1. FindX.. From equation 11.27 and the fact that

62.5 X 10'^
V,
And from Figure 11.18, again since
230
for this part.
= 2 7 1 .7 4 A
, Z.1^ = - 3 6 .8 7 . Hence
= 271.74 Z - 3 6 .8 7 " = 271.74 Z .-0 .6 4 3 5 rad = 217.4 - y l 6 3 A

Step 2. FiudY^. From KVL and Ohms law,
= 0.5 [217.4 - j\ 6 3 ]+ 230 = 3 3 8.7 - y 8 1.5
= 3 8 4 .4 Z - 0.2362 rad = 3 8 4 .4 Z - 13.533'" V
s :i
Step 3. Compute the complex power, , delivered by the source.

S^. = V ,I* = 3 4 8 .4 Z - 13.533" x 2 7 1 .7 4 Z 3 6 .8 7 " VA
= 94.6 6 4 Z 2 3 .3 3 7 kVA = 9 4 .664 LQA07 rad kVA
= (86.918 + ;3 7 .5 ) kVA
Norc that it takes 86.918 kW to run a 50 kW motor. The difference is the loss in the power line.
If we have a way o f reducing the magnitude o f
this line loss will be reduced. In fact, we do,
and this strateg)^ is the goal o f the next part o f the example.
(c) Correct the power factor o f the combined motor-capacitor load to 0.95 biggifig- Since motors are
inductive, a properly chosen capacitor can improve the pf to 0.95 lagging. The new motor con
figuration is that o f Figure 11.17, with the capacitor connected across the motor. The proper value
o f C must be found.
Step 1. What does a p f o f 0.95 lagging require in terms ofcomplex power absorbed by the motor-capac-
itor combination'^
=
Z c o s " l( 0 .9 5 ) = 5 2 .6 3 Z 1 8 .1 9 5 = (50 + 7 1 6.4342) kVA
0.95
(11.28)
Recall that
= (50 + > 37.5) kVA
Step 2. Find a capacitor value to reduce the reactive power. For this step consult Figure 11.19.
FIGURU 11.19 Relationships bcrwecn new and old complex powers.

In Figure 11.19, one observes that
is the same for both the new and old complex powers since
that is what the motor requires for its operation. The reactive powers are different. The new com
plex power with the 0.95 lagging pf has a smaller reactive power component. The capacitor must
be chosen to reduce the old reactive power to this new level. Hence,
jQoU -j^ncw = ;2 1 -0 7 kV.AR = -/reactive power o f capacitor) = -yQ^-
522
Therefore,
jQ c = - j 2 1.07 kVAR = V ,Ic = V, [K c(JO ))V ,]' = -./coC| V,p

It follows that
C = ^
= ^ H lZ iii4 = ,o 5 7 x lO - F
co|V,
12071(230)
(d) Compute the new power delivered by the source.
Step 1. Compute the complex power, denoted
absorbed by the motor-capacitor load. The com-
plex power absorbed by the load is the sum o f the complex power consumed by the motor and
the reactive power of the capacitor, as illustrated in Figure 11.9, i.e.,
S f'= (50 + y i6 .4 3 ) kVA
Step 2. Compute the new
denoted
Since S ^'is the complex power o f the combined motor-
capacitor load,
^ 5 ( W ! M 3 ,0^ = (217 + ; 7 : .43) A

230
J
!
Step 3. Compute . From KVL and Ohms law,

V ;^ = 0 . 5 l f + V, = (3 3 8 .5 - y 3 5 .7 2 ) V
Step 4. Compute the new complex power delivered by the source. By definition,
Sr ' = v f ' ( i ; '" )* = (7 6 - ; i 6 .4 8 ) kVA

Hence, the new average power delivered by the source is 76 kW with pf correction as opposed to
86.9 kW without pf correction. With this pf correction, there is a reduction o f 86.9 - 76 = 10.9
kW o f power loss in the line connecting the source to the load.
Example 11.10 illustrates how adding a parallel capacitor can improve the pf o f a load. The main
motivation for improving the pf was to reduce the power loss in
However, even if
is
negligible, another strong reason exists for improving the load pf. Example 11.11 illustrates how
an improved power factor allows a single generator to run more motors. Example 11.11 will fully
utilize the principle of conservation o f complex power and the two consequences o f equation
11.25.
From equation 11.25 and the fact that S = /^ + jQ, we can express pf directly in terms o f P an d Q
as follows:
. ^
523
with a lagging pf for Q > 0 and a leading pf for Q < 0. Solving for Q from equation 11.29, we
obtain
e=
P .H r - i
P f
(11.30)
with Q > 0 if pf is lagging and Q < 0 if pf is leading.

W ith these formulas we can simplify the process o f power factor correction.
E X A M P L E 11 .1 1 . An industrial plant has a 100 kVA, 230 V generator that supplies power to one
large motor and several identical smaller motors. The resistance o f the connecting line is assumed
negligible in the approximate analysis below'. The large motor, labeled t^-^pe A, draws 50 kW at a
pf o f 0.8 lagging. Each smaller motor, o f type B, draws 5 kW at a p f o f 0.7 lagging. The configu
ration is illustrated in Figure 11.20.
Generator
230 V ^
60 Hz
lOOkVA
Type B Motor
Type A Motor
Type B Motor
FIGURH I 1.20 A generator supplying power to one large motor and several smaller motors.
(a)
(b)
(c)
(d)
Can the generator safely supply power to one large motor and three small motors? What
are the generator current (magnitude) and the power factor o f the combined loads?
Compute the number o f small motors (besides the one large motor) that can be run
simultaneously without exceeding
If the power factor for all motors,
ing appropriate parallel capacitors
(besides the one large motor) can
the generators rating.

large or small, is corrected to 0.9 lagging by connect
(as done in Example 11.10), how many small motors
be run simultaneously without exceeding the genera
tors rating?
Compute the capacitances required in part (c) for the large and the small motors.
S o lu tio n
(a) Compute the reactive power for each motor type. Using equation 11.30, the reactive power for
each type o f motor is given as
- 1
=50
-1
= 37.5 kVA
0.8^
and
Q b = P ,i, / A - >
pfg
= \
-^ 2 -
0.7
= .i . l 0 1 k V A
2 I
By the principle ot conservation o f power, the complex power (in kVA) supplied by the generator
is
V
^ 3(7^^ + ;Q ^ ) = (50 . 15) + ;(3 7 .5 -h 15.303)
= 65 + ;5 2 .8 = 8 3 .7 4 ^ 3 9 .
kVA
By inspection, the apparent power is 83.74 kVA, which is below the generator capacity o f 100
kVA, meaning that the generator can safely operate the large motor and three smaller motors.
The magnitude o f the generator current is 83,740/230 = 364 A. From equation 11.29, the pf o f
the combined loads is
pf =
^
P- +
= 0 .7 7 6 2
\/65 + 5 2 . 8
(b) Compute the number o f stnall motors {besides the one large motor) that can be run simultaneous
ly. When one large type A motor and n smaller t)^pe B motors are connected in parallel, the com
plex power delivered by the generator is
^gen
7 ^
= ^^0 + X 5) + 7(37.5 + ;/ X 5.101) kVA
rhe apparent power is
^ (50 + n x 5 ) ~ + (3 7 .5 + /; x 5 .1 0 1 )" kVA

Since the generator has a capacity ot 100 kVA, then
c
= ( 50+//x 5 ) + (37.5+/; X 5.101)- <
100 =
10^
(11.31)
Replacing the inequalit)' sign in equation 11.31 by an equalit)^ results in the quadratic equation
51.020/r + 882.5750/7 - 6 ,0 9 3 .8 = 0
The resulting zeros are //, = 5.288 and //2 = - 22.58. The largest positive integer satisfying the
inequality 11.31 is // = 5. Thus, at most, five small motors can be run simultaneously with the
large motor without exceeding the generators capacit)'
(c) I f all power factors are corrected to 0.9 lagging, fin d the number o f small motors {besides the one
large motor) that can be run simultaneously. We essentially repeat the calculations o f part (b) with
the new given power factor o f 0.9 lagging:
and
e :r = / '.. A
V PfA
=5o
y - 1 = 2 4 . 2 1 6 kVA
V 0 .9 ^
The complex power (in kVA) supplied by the generator is
= Pa + j Q T + n(Pfl + j Q T ) = (50 + X 5 ) + j(24.216 + /, x 2.416)

The apparent power (in kVA) is
= s j (50 + /2 X 5) + (24.2 \6 + n x 2.4216)
s '" "
ficn
As before, n satisfies the inequaiit)'

(50 + X 5)2 + (24.216 + x 2.4216)2 < 10^
To find n, we compute the largest positive root o f the quadratic equation
30.8641;/2 + 6 l7 .2 8 2 9 n - 6 ,9 1 3 .6 = 0
The roots o f this quadratic are
= 8 and 2 = 28. The largest positive integer that satisfies the
above inequalit}' is n = 8. Thus, eight small motors, as opposed to five in the earlier case, can be
run simultaneously with the large motor without exceeding the generators capacity.
(d) For the large motor, the capacitor must absorb a negative reactive power equal to
Q^.
Equivalently, the capacitor must supply a reactive power equal to

Qw - Q T =
- 24.216) X 1000 = 13284 VAR
(10.32)
From Example 11.6, the reactive power supplied by a capacitor is
I Q-CA I = ^CA^CA =
6 0 Q X 2302 VAR
(10.33)
Equating equations 10.32 and 10.33, we have In x 60C^|2302 = 13284. Solving produces
66 6 .1 6 X 10-<^ E
Similarly, for the smaller motors,
Q-B - Q"b ' = (5.101 - 2.4216) X 1000 = 2679.4 VAR

Also, we have
Qcb\ ~Vcb^Cb\ -
= In
X60Q
X 2302 VAR
Equating these two quantities and solving for C^, w'e obtain Cg = 134.35 x 10^ F. We note that
in the power industry, such capacitors are usually specified only by their kVAR rating, with no
mention o f their actual capacitive value in F.
In the above example the generator capacit}^ was given in terms o f VA, the unit o f apparent power.
The example points out the importance o f reducing reactive power to more fully utilize the power
526
capacity o f the generator. Use o f VA for generator, motor, and transformer capacit)' arises out o f
safety considerations. Most ac machiner)' operates at a specified voltage depending on the insula
tion strength. The size o f the wire and other heat transfer factors determine the maximum allow
able current o f a machine or transformer. Also, the cost and physical size o f most ac equipment
are more closely aligned to the VA rating than to other measures. Hence, the VA rating better
reflects the safe operating capacity o f ac equipment.
Another motivation for improving the power factor is economical. A power company charges a
consumer only for the actual electrical energy used. A meter measures this energy usage in units
o f kWh (kilowatt-hour). As mentioned earlier, most clectrical loads have lagging currents. As
shown in Examples 11.10 and 11.11, for a given required average power, a higher pf means lower
transmission line losses. Also, loads that operate at low pf force power companies to pursue high
er kVA ratings o f the generator equipment. Thus utilities companies encourage consumers to
operate their equipment and appliances at high pfs. Since power companies can supply more
power with the same equipment if the pf is high, they adjust their rates so that energy costs are
less with a high pf and are greater with a low pf.
7. MAXIMUM POWER TRANSFER IN THE SINUSOIDAL

STEADY STATE
Chapter 6 outlined the basics o f maximum power transfer for linear resistive networks. Having
introduced energ)' storage elements L and C, and having studied methods for sinusoidal steady
analysis, it is time to extend the results on maximum power transfer to general linear networks in
the sinusoidal steady state.
MAXIMUM POWER TRANSFER THEOREM FOR AC CIRCUITS

Let a practical ac source be represented by an independent voltage source V^ {voltage phasor in rnis
value) in series with an irnpedance
+ jX^. An adjustable load impedance
=
jX^j
with R l > 0, is connected to the source {Figure 11.21). In steady state, for fixed Z^, Vp and O),
the average power delivered to the load is maximum when Z[^ is the complex conjugate ofZ^, i.e.,
(11.33a)
and
(11.33b)
and the maximum average power is given by

p
^ <e ff
max
(11.33c)
>2.
To derive the conditions o f the maximum power transfer theorem, observe that the current phasor, I, is
(11.34)
1=
{R, + RO + j(X , + XO
Thus the average power delivered to the load is
(11.35)
Here Pave
a hmction o f two real variables
and Xj . To find the conditions for maximum
the partial derivadves ------ and ------ to zero and solve for
set
and X^. Differendating equation 11.35
with respect to R^ yields

V f [ ( R , + R i f + {X, + X 0 - - 2R l (R, + R , , )
dRi
{R^ + Rl)~ +
-)12
(1 1 .36a)
and differentiating with respect to Xj produces
dP
ax,
v "-[-1R l {X,
+ Xl )
= 0
From equation 1 1.36b, the only physically meaningful solution is
(11.36b)
528
which is equation 11.33b. Substituting this result into the numerator of equation 11.36a yields
v3 ^
(R s + Rl T
The only physically meaningful solution here is
^ 0 and
R ,-R ,
which produces equation 11.33a. (Note that this is the condition for maximum power transfer in
purely resistive circuits.) Substituting these results into equation 11.35 produces equation 11.33c,
Ir
4/?,
which verifies the theorem.

The theorem can be established less formally as follows. W ith any existing
connected to the
source, if the total reactance
+X^) is not zero, we can always increase the magnitude of the
current, and hence the power delivered to the load, by tuning out the reactance, i.e., by adjust
ing
to be -X ^. This implies condition 11.33b. Under such a condition, the circuit becomes
resistive, and the maximum power transfer theorem of Chapter 6 may be applied to obtain equations 11.33a and c. The maximum power obtainable with a passive load, given by equation
11.33c, is called the available power of the fixed source.
The conditions for maximum power transfer, as given by equation 11.33, are valid when both
and Xj^ are adjustable. If X^ is fixed and only
is adjustable, then the condition for maximum
power transfer is
=
(11.37)
which is obtained by solving equation 11.36a for fixed X^ and X^ .
r\
If the source is a general two-terminal linear network, then its Thevenin equivalent must be found
before application of the maximum power transfer theorem. If the source is represented by a
Norton equivalent circuit, we can use a source transformation to obtain the Thevenin form and
then apply equations 11.33.
'
As pointed out in Chapter 6, maximum power transfer is not the objective in electric power systems, as the sources usually have very low impedances. On the other hand, it is a very important
factor to be considered in the design of many communication circuits, as illustrated in the following example.
o
S29
EXA M PLE 11 .1 2 . The radio receiver shown in Figure 11.22a is connccted ro an antenna. The
antenna intercepts the electromagnetic waves from a broadcast station operating at 1 MHz. For
circuit analysis purposes, the antenna is represented by the Thevenin equivalent circuit shown in
Figure 11.22b.
(a)
Find the input impedance
o f the receiver if maximum power is to be transferred
(b)
from the antenna to the receiver.

Under the condition o f part (a), find the magnitude of the voltage across the receiver ter
minals, and the average power delivered to the receiver.
receiver input
antenna equivalent circuit
equivalent circuit
(b)
(a)
FIGURE 11.22 Example of maximum power transfer.

S o l u t io n
= 1\ i l and
(a)
From the maximum power transfer theorem, the answers are
(b)
Since the reactances in the circuit have been tuned out, the input current to the receiv
er is simply 14.6/(21 + 21) = 0.348 mA. The input impedance has a magnitude
7.
= 1070.
+ 1070- = 10 7 0 .2 n
Therefore the magnitude o f the voltage across the receiver terminals is 0.348 x 1070.2 = 3 7 2.4
mV (rms). The power transferred from the antenna to the receiver is 0.348^ x 21 = 2.54 uW.
In the preceding discussions of maximum power transfer, we have assumed that the load is
adjustable. In practice the load is often fixed, as for example, in the case o f a loudspeaker having
a4
voice coil. In such cases, one designs coupling networks consisting o f lossless passive com
ponents. These coupling networks transform the fixed load impedance into one whose conjugate
matches the fixed source impedance. This permits maximum power transfer to the load. The fol
lowing example illustrates the principle. A design procedure for some simple coupling networks
will be discussed in the second volume o f this text.
530
EXA M PLE 11.13.

A fixed load resistance /?^ = 100
representing the input resistance o f an amplifier, is connect
ed to the source o f Example 11.12 through a passive coupling network, i.e., a network that does
not generate average power, as shown in Figure 11.23.
(a)
Show that the maximum voltage that can be developed across
(b)
Show that the coupling network shown in Figure 11.23 achieves this maximum voltage
across
is 0 .504 V.
FIGURF, 11.23. Maximum power transfer through a coupling network.

S o l u t io n
(a)
From equation 11.33c, as used in Example 11.12, the available power from the source is
2.54 j.iW. If all o f the power is delivered to R^ , then the voltage,
must be
= ylPmaxf^L = ^ 2 .5 4 X 10^ X 100000 = 0 .5 0 4 V

(b)
The input impedance
o f the coupling network with load must be the conjugate o f
the source impedance. Specifically
în = yw ^ + ------------- r~
j(dC +
R,
Substituting the values OJ = 10^, L = 400.9 x 10^, C = 109.8 x 10
and /? = 100 x 10^ into the
above expression yields

= 21 + ;1 0 7 0 Q
which is indeed the conjugate o f the source impedance. Since
is conjugate-matched to the
source impedance, the maximum power o f 2.54 |aW is transferred to the coupling network. Since
the coupling network consists o f L and C, neither o f which consumes average power, the 2.54 f,iW
power must be transferred out o f the coupling network and into the load resistance. The voltage
across the load resistor, Vj, is given by
= 4 ^
= V 2 .5 4 X 10^ X 100000 = 0 .5 0 4 V
This verifies that the coupling network o f Figure 11.23 enables the largest voltage to appear across
the load resistor.
531
8. SUMMARY
Fundamental to the material in this chapter is the definition o f the effective value (rms value) o f
a periodic voltage or current waveform. For a sine wave, the effective value is the maximum value
divided by -v/2 For a general periodic voltage or current, the effective value is the value o f a dc
waveform that will produce the same amount o f heat as the periodic waveform when applied to
the same resistance. Using the definition o f the effective value o f a waveform, formulas for the
average power absorbed by a linear two-terminal network in ac steady state were set forth and
derived. Recall that for a two-termlnal element with sinusoidal voltage v{t) = ^/2 V'^ cos(co/- 0y) an
current i{t) =
0/), the absorbed average power \s P = ^^^^cos(Oj; - G/), assuming
the passive sign convention. Next w'e presented the definition o f complex power and its compo
nent parts, which include its real part or average power, its imaginary part or reactive power, and
its magnitude or apparent power. Various examples illustrating the calculation o f these powers
were given. Again, for a two-terminal element w'ith sinusoidal voltage v{t) and current i{t) as
above, the reactive power absorbed is defined to be Q = ^ ^ ^ s i n ( 0 y - 0/) VAR (volt-amperereactive). After introducing these different types o f power, we proved the principle o f conserva
tion o f complex power, which implies the consenâtion o f real power and the conservation o f reac
tive power. This was followed by the definition o f power factor, pf, the ratio o f average power to
apparent power, which takes on values between 0 and 1. The need for improving a low power fac
tor and a method for achieving an improved power factor were illustrated with two examples.
The maximum power transfer theorem, first studied in Chapter 6 for the resistive nervvork case,
was taken up again in this chapter for the sinusoidal steady-state case. Here, maximum power
transfer to the load requires that the load impedance be the conjugate o f theThevenin impedance
seen by the load. As pointed out earlier, the theorem has no application in electrical power sys
tems. However, for communication circuits the maximum power transfer theorem is o f extreme
importance. The power that can be extracted from the antenna o f a radio receiver is usually in the
microwatt range, a very small value. It is therefore necessary to get as much power as possible from
the antenna system. Example 11.13 illustrates this principle.
9. TERMS AND CONCEPTS

Apparent power: the apparent power absorbed by a two-terminal element is
assuming the
use o f a passive sign convention . The unit is VA (volt-ampere).

Average power: the average value o f the instantaneous power. For a two-terminal element with
sinusoidal voltage u{t) = yfz V^^cos{o)t + Qi,) and current i{t) = ^
V^jj-cos{o)t + 0/), the
absorbed average power is P = ^ ^ ^ ^ c o s (0 y - 0/), assuming a passive sign convention.

Com plex power: for a two-terminal element absorbing average power P and reactive power Q,
the complex power is defined to be S = P + jQ. The unit o f measuremcnr is VA (voltampere). The magnitude of S is the apparent power.
Conservation o f powers: for any network, the sum o f the instantaneous powers absorbed by all
elements is zero. For any linear network in sinusoidal steady state, the sum o f the aver
age powers, reactive powers, or complex powers absorbed by all elements Is zero. This
property is a consequence of KCL and KVL.
532
Chapter 11 * Sinusoidal State State Power Calculations
Efifiecdve value (nns value): for a sine wave, the effective value is the maximum value divided by
V 2 For a general periodic voltage or current, the effective value is the value of a dc wave
form that will produce the same amount of heat as the periodic waveform when applied
to the same resistance.
Instantaneous power: the power associated with a circuit element as a function of time. The
instantaneous power absorbed by a two-terminal element is p{t) = y(/)/(/), assuming that
a passive sign convention is used.
Maximum power transfer theorem: if a variable load 7.^^ = -^jî is connected to a fixed source
Vy having a source impedance
=7?^+/A^, then the largest average power is transferred
to the load when
is the complex conjugate of Z j , i.e.,
= R^ and AT^ =-X^.
Power factor: the ratio of average power to apparent power. The pf value lies between 0 and 1.
For a passive load, the power factor is said to be lagging when 90 >0j, - 6,->0, and lead
ing when 90 > 0^- 0^, > 0.
Real powen in ac circuits, real power means average power. It is the real pan of the complex
power.
'
Reactive power: for a two-terminal element with sinusoidal voltage v{t) = -y/2 V^cosisat + 0^)
and current i{t) = ^ /^ co s(co f + 0^, the reactive power absorbed, denoted by Q, is
defined to be Q = ^ ^ ^ sin (0 y - 0/), assuming that a passive sign convention is used.
The unit of measurement is VAR (volt-ampere-reactive).
n
o
o
n
n
o
^33
Problems
AV(t)(V)
20
INSTANTANEOUS AND
AVERAGE POWERS
t(sec)
1. For the source current waveform o f Figure

PI 1.1a, which drives the circuit o f Figure
10- -
(a)
PI 1.1b, find the average power consumed by

the 2 Q resistor.
(a)
Figure Pi 1.3 (a) Rectangular waveform, (b)
Triangular waveform.
EFFECTIVE VALUE OF
NONSINUSOIDAL SIGNALS
(b)
Figure P l l .l
AN SW FR. 0.758 watrs
4. (a)
Compute the effective value o f each o f

the periodic signals in Figures Pi 1.4a
and b.
2. Compute the average power delivered to a 1
(b)
kQ resistor by a current o f the form

(a)
1 0co s(10t)m A
(b)
10 |cos(10^)| niA
(c)
10 cos( 1Of) mA
(d)
Plot each o f the instantaneous powers

for 0 < r < 1 sec using MATLAB or its
equivalent.
3. (a)
Compute the average power absorbed

by a 10
resistor whose voltage is
given by each o f the waveforms in
(b)
Using MATLAB, plot the instanta
Figure Pi 1.3.
neous power associated with each
waveform for 0 < f < 3 sec.
For the circuit o f Figure P i 1.4c, find

the power absorbed by Rj if the volt
age source v{t) is given by the wave
form o f Figure PI 1.4a.
(c)
Repeat part (b) for the waveform o f

Figure PI 1.4b.
534
i(t)0
60 O <
= 30 Q
(c)
Figure PI 1.5
6. (a)
Find the effective value o f the source

current plotted in Figure PI 1.1a.
2Q
v(t)
(b)
R = 80
part (a).
Figure P11.4
7. Compute the effective value o f

(a)
Compute the effective value o f each o f
(b)
the periodic signals in Figures P11.5a
(c)
and b.
(b)
resistor in Figure PI 1.1b
using the effective value computed in
(0
5. (a)
Find the average power absorbed by

the 2
v^{t) = 10 + 2 cos(20r)
= 10 cos(2r) + 5 cos(4/)
= 10cos(2r) + 5 cos(4r) +
5 c o s ( 4 f - 4 5 ) V.
For the circuit o f Figure PI 1.5c, find

the power absorbed by
if the cur
rent source i{t) is given by the wave

form o f Figure Pi 1.5a.
(c)
Repeat part (b) for the waveform o f

Figure PI 1.5b.
Ai(t)
AVERAGE POWER
CALCULATIONS IN
SINUSOIDAL STEADY STATE
8. Using equation
44
11.10, find the average
power absorbed by the resistor in the circuit
shown in Figure PI 1.8, where

sin(5r) V . /? = 25 Q, C = 8 mF.
12
15
(a)
t(sec)
-4 -
(b)
Figure PI 1.8
ANSWl-.R: 25 watts
= 50 x
9. In the circuit shown in Figure PI 1.9, i-{t) =
(b)
5 cos(30/) A, /? = 5 Q, and C = 5 niF.

(a)
Find vît).
(b)
Compute the average power delivered

by each source.
(c)
Repeat parts (a) and (b) for R = 50 Q,
Zi = p O Q, and a = 49.
Using equation 11.10, find the instan

taneous and average power absorbed
by the load.
/Y Y V

Load
' 6
aV
Figure 1 11.1 I
ANSW'FRS: (b) in random order: - 9 0 . 100
watts; (c) 4. - 3 .9 watts
Figure P l l .9
AN SW ER: (b) 40 warts
12.
10.
= 120
R = 32 W, Z^ = ;2 0 0 Q.
Z q = y80 ri, and ^ = 4.
Consider the circuit o f Figure PI 1.12,
where
In the circuit o f Figure PI 1,10,
5 0 z i-9 0
/? = 6 Q,
= ;1 2 Q, and
(a)
- ;4 Q.
(a)
Find the phasor current
(b)
and deter
Find the average power delivered by

each source.
mine its magnitude.

(b)
Find the average power (in watts)

absorbed by the resistor.
Using equation 11.10, find the aver

/Y Y V
z. -
age power (in watts) delivered by the

source.
(c)
O nly
R absorbs
average
power.
' 6
III is
Therefore, once
known, the
average power consumed by R is
=
y?|
(d)
bl
Figure PI 1.12
Check your answer to part (b)
using this formula.
ANSWF.RS: (b) in random order: S7.6, 230.4
Repeat parts (a) and (b) when 7? = 30

Q, Z^ = y'50 Q, and Z q = -ylO Q..
watts
13. A coil is modeled by a series connection o f

/Y Y V
L and R. When connected to a 110 V 60 Flz
source, the coil absorbs 300 watts o f average

pow'er. II a 10 Q resistor is connected in series
with the coil and the combination is connected
to a 220 V 60 Hz source, the coil also absorbs
300 watts o f average power. Find L and R.
AN SW FR: R = 0.9901 LI and L = 16.6 mH.
Figure Pi 1.10
A N SW FR; (b) ISO wans
11. For the circuit in Figure P i 1.11,
100.10 V ^ , /? = 5 n , Z^ = ;5 0 a , and ^ = 9.
(a)
Compute the current phasor I^.
14.
Consider
Figure
Pi 1.14,
where
V,(/) = 2 2 0 >/2cos( 120jir) V and R^^-^

=3
The inductance L is adjusted so
that IV
150 V, a n d
watts. Find the magnitude o f
250
the
^36
reactance o f the coil, and the values o f L

and R.
v.(t)
Figure P 11.17
C H EC K : Complex power is 7+y6 VA
Figure P 11.14
18. In the circuit shown in Figure Pi 1.18, to =
ANSW ER: /. = i2.9 niH, R - 14.SS3 Q
= 4 2,

400 watts, and
64 rad/sec,
=
nns with all other values the
= 120/ i60"
= 20 2, /?,
= 4 12, /., = 0.375 H, and L , = (^-125 H. Find

the complex and average powers absorbed by
the load.
same.
Load
C H EC K : L = 7.958 mH, R = 17.794 2

TY"YA--- Q
R.
L.
COMPLEX POWER
CALCULATIONS
16. For the circuit o f Figure PI 1.16, R = 5
= 3 .jA a ,
= -yiO
and
v,(r) = 100v 2 cos(1207i:/) V.
(a)
Find the complex power absorbed by
(b)
Compute the apparent power in VA, the
Figure P 11.18
C H E C K : Complex power is 36 + p 2 VA
the load.
average power (in watts), and the reactive
19. In the circuit shown in Figure Pi 1.19, Z| =
power in VAR delivered to the load.
Zc
{-
+7a, z, =4 +ji2 a, z,^ 2+j i a, v, =
(104 +y50)~V, and

(a)
= (106 + y48) V at 60 Hz.
Find the voltage V , in polar and rec

tangular forms.
C H EC K : , = 1 0 0 + ?
'-0
(b)
Find the complex power absorbed by

each of the three impedances and then
the
Figure P 11.16
power delivered
C H EC K : (b) 300 watts, 400 VAR
sources.
17. In the circuit o f Figure PI 1.17,

v ,(r ) = IOOV2 cos(500/ + 30^) V, /?, = 100 Q,
Z 2 is 100 + ;5 5 0 VA
by the
two
C H EC K : complex power absorbed by
R, = 700 2, and Z. = 1.2 H. Find /^(r), the

complex power, average power, and apparent
power absorbed by the load.
(c)
Using the results o f part (b), verify

conservation o f complex power.
537
(a)
r-
Z.
Z.
power, and the average power deliv

ered by the source.
v .Q
the complex power, the apparent
-I
(b)
Find the source current
in rectangu
lar and polar form.
Figure Pi 1.19
C H EC K : complex power delivered by source h
Z,
is 10 + 260; VA.
v .Q
CONSERVATION OF POWER
20. This problem should be done without any
phasor voltage or current computations. In the
circuit o f Figure P i 1.20,
= 2300
Figure PI 1.21
C H EC K S: P,
.= 1600 watts, |Ij = 8 .6957 A
at 60
Hz and the following powers (in kVA) are con

sumed by various impedances and resistances:
S , = 20 + y8 , S , = 20 + 7 I 8 , S 3 = 5+ 76 , and
= 3 + j4.
(a)
Find
(rectangular form) and the
complex
power delivered
by
the
POWER FACTOR AND POWER

FACTOR CORRECTION
22 . (a)
load that absorbs 2 kW o f average
source.
(b)
(c)
Determine
in polar form.
Find the complex power delivered to
(b)
(d)
(e)
Find
in rectangular and polar form.
power with pf = 0.90 lagging.

Find the complex power delivered to a
load that absorbs 4 k\V o f average
the group o f impedances Z j, Z-,, and

^4Find V , in rectangular and polar form.
Find the complex power delivered to a
power with pf = 0.90 leading.

23. For the circuit o f Figure PI 1.23, Z j absorbs
1600 watts at pf = 1, while the apparent power

absorbed by Z^ is 1000 VA at pf = 0.8 lagging,
and V;(/)= l20V 2cos(1207cr) V. Find
(a)
the phasor current
(b)
the phasor voltage V|
(c)
the phasor voltage
in
Figure PI 1.20
C H EC K S: (a) Complex power delivered by
source: 48 + y'36 kVA; (b) |lj = 26.087 A
v,(t)
21. In the circuit o f Figure PI 1.21, V^. = 230

V.
at 60 Hz. The complex powers (in VA)
absorbed by the five impedances are S| = 100 +
yioo, S , = 200 +7 IOO, S 3 = 100 + p o , S^ = 400

+J250, and S^ = 800 + J700. Find
C H EC K :
- 1 8 .8 2 3 5 ;
Figure P i 1.23
= 20 - 5; A^ and V , = 75.2941
538
Chapter 1 1 * Sinusoidal State State Power Calculations
24. The circuit shown in Figure P I 1,24 is in

the sinusoidal steady state. Suppose that
absorbs 3000 W of average power at pf =
0.7905 lagging, Ri-^^ = 0.1 2, and
= 120
V rms
(a) Find the average power absorbed in
the transmission line resistance (in
26. The circuit in Figure P I 1.26 operates at CO

= 500 rad/sec and = 100Z0 V ^ . The com
plex power drawn by the load without the
capacitor attached is S^; = 100Z30 = (86,6 +
y50) VA. This constitutes a pf of 0.866 lagging.
(a) Find the values of R and L
(b) Find the value of C in fiF that pro
duces a pf of 0.95 lagging.
W).
(b)
Find
=A cos(1207t^ + 0) V and
the complex power delivered by the
source.
Load
0 -"0 -
!'4<
Transmission
line resistance
'N/S/'------ OR .,
o - -oFigurePll.26
CHECK; 3 pF < C< 5 pF
Figure P I 1.24
CH ECK: (a) 100 watts
25. As shown in Figure P I 1.25b, a capacitor is
put in parallel with a motor using average
power
= 40 kW operating at a power fector of 0.7 laggmg to boost it to a power factor
of 0.9 lading. The voltage across the parallel
motor-capacitor combination is 230Z0 V ^ .
The power relationships are shown in Figure
P I 1,25a. If the frequency of operation is CO =
120tc, compute the proper value of the capaci
tance, C (in mF).
27. The circuit shown in Figure P I 1.27 is operat

ing in the SSS with v^(0 = 120>/2cos(1207cO
V. Device 1 absorbs 360 W with pf =0.9. Device
2 absorbs 1440 W with a pf of 0.866 lading.
Find the value of the capacitor C such that the
magnitude of the source current equals 16 A ^ .
What is the pf of the two-device-plus-capacitor
combination?
Device
1
Device
Figure P I 1.27
CH ECK: 0.108 mF
< C<2 mF
(b)
28. A group of induction motors is drawing 7

kW from a 240 V power line at a power faaor
of 0,65 lagging. Assume CO = 12071 rad/sec.
(a) What is the equivalent capacitance of
a capacitor bank needed to raise the
power factor to 0.85 lagging?
(b) What is the kVA radng of the capacitor
bank of part (a); i.e., what is the reactive
power supplied by the capacitor bank?
Chapter 1 1 Sinusoidal State State Power Calculations
(c)
539
Determine the annual savings from
By what amount in kVAR must
installing the capacitor bank if a
the reactive power be reduced to
demand charge (in addition to the
produce a pf o f 0.94 lagging?
charge for the kilowatt-hours used by
(iii) Compute the needed capacitor
the induction motors) is applied at

S20.00 per kVA per month.
(iv) Compute Z^j()) as the ratio o f
AN SW ERS: (a) 0.1771
current
niF; (b) Minimum
the capacitor phasor voltage to
kVA rating: 3.8457 kVA; (c) S608.14
the capacitor phasor current, at

the indicated frequency.
(v) Compute the proper value o f C in
29. Consider a source that drives an electric
mF.
motor that consumes an average power o f 94

kW (about 125 hp) at a pf o f 0.65 lagging, as
show'n in Figure PI 1.29, where
= 0.07 Cl.
(i)
Compute the new I"^ .
(j)
Compute V 'f":
(k)
Compute the complex power deliv

ered by the source and the new effi
ciency.
oM o to r)
Later
Addition of
Capacitor
for p.f.
Correction
MAXIMUM POWER
TRANSFER
Figure P11.29
The phasor voltage across the motor is êff~
2 3 0 Z 0 V. The sinusoidal frequency is 60 Hz.
(a)
Find the apparent power delivered to

Find the complex power,
(a)
(c)
Find the value o f the load impedance

that will absorb maximum power
at
delivered
(b)
to the motor.
CO =
delivered to the motor.

Compute I
(e)
(f)
Compute V^.
Compute the complex power deliv
(g)
ered by the source.

Determine the efficiency o f the con
= 100
find the average power
absorbed by the load.

u
R.
v .Q
o
figuration, i.e., the ratio o f average

power delivered to the motor to the
100 rad/sec.
Given the conditions o f part (a) and
Determine the reactive power in VAR
(d)
C=
1 mF.
the motor in kVA.

(b)
= 100 Q, R-> = 25
PI 1.30 in which
Figure PI 1.30
C H EC K : 20 + ;1 0 Q
average power delivered by the source

as a percentage.
(h)
Add a capacitor across the motor to
31. In the circuit o f Figure PI 1.31,

v^(r) = 100V 2cos(1000/) V, /?, = 80 Q, R, =
improve the power factor to 0.98 lag
20 Q, Z. = 5 mH. Find the value o f the resist
ging. Then
(i) Compute
ance R^ (in Q) and the capacitance C (in mF)

Compare with
sou
(ii) Recall that the role of the capaci
tor is to reduce the reactive power.
such that maximum average power is absorbed

by the load.
540
Thevenin equivalent, i.e., a volt
Load
age source in series with either a
/Y Y V l
L
series RC or a series RL as appro
R,
priate.
(b)
Compute
the
load
impedance,
Z^(/150), necessary for maximum

power transfer. Show this load as
Figure P i 1.31
either a series RL circuit or a scries RC
A N S W H R : R, = 16 12 and C= 0.2 mF
circuit. Should it be the opposite of
32. The circuit o f Figure P l l .3 2 operates in the

sinusoidal steady state with W = 1000 rad/sec, R
= 1 k n , C = 1 ;<F, I = 0.5 H,
= 3 and
the case in (a)(iii)? Why?

(c)
Compute the average power con

sumed by the load at maximum power
transfer.
2 ^ 0 A ^ (a)
Find the value o f the load imped
ance
for maximum average
power transfer.
(b)
8Q
-L
Find the average power absorbed
3.334 mF - J 1.667 mF
16Q
by the load under the conditions

o f part (a).
Figure Pi 1.33
34. Consider the circuit o f Figure P l l . 3 4 ,

w'hich operates at
CO
=10 rad/sec. Suppose R =
l O a , L = 2 H , and /^(O = I 0 V 2 cos(lOr) A.

(a)
AN SW ER:
Q to deliver maximum average power
= 1 + \.5/ k li, 4000 warts
33. The purpose of this problem is to compute

for maximum powder transfer by following a spe
to the load. What is this maximum

average power?
(b)
(a)
Compute the Thevenin equivalent of

the circuit at terminals A and B:
(i)
Use nodal analysis to compute

Note that there is a floating
If Rj = 30 Q, determine Q for maxi

mum power transfer to the load. What
cific procedure. Consider the circuit o f Figure

PI 1.33 in which /^.(r) = 50V 2 cos(150/) A.
Choose the proper values o f Rj^ and
is this maximum average power?

(c)
If Q = 8 mF, then determine R^ for

maximum power transfer to the load.
What is this maximum average power?
Load
dependent voltage source.

C H EC K : V^^.= 2 8 .4 7 -791.1 V
(ii) Find the Thevenin equivalent
impedance, Z^y^(/150), seen to the
left o f terminals A and B.
C H EC K : Re[Z,//yi50)] =
8.2846Q
(iii) Show' the phasor form o f the
Thevenin
equivalent
circuit.
Then show the circuit form o f the
Figure Pi 1.34
A N SW ERS: (a) 10 12, 5 ml-, 1230 wans; (b) 5
mF, 937.5 wans
35. In rhe circuit o f Figure PI 1.35,
541
= 100
^rms
adjusted to achieve different
goals. Assume R = GO Q., Z q = - ^ 8 0 Q.
(a)
Find the value o f
that maximizes
(C H EC K : 24 Q .) What is the
value o f
(b)
Find the value o f
that maximizes
Figure PI 1.37
V /.,. W hat is the value o f IV,I

L nutx
(b)
Now suppose y ? = 4 Q ,C = Im F , and

/. = 0.1 mH. Choose R^ for maximum
-------------
power transfer and find

38. Consider the circuit o f Figure PI 1.38
where
-------------
= 0.1
at
CO =
10^ rad/sec. Suppose
/?, = 100 a , /?, = 10-1 k n , Zî = -71000.

(a)
Find the impedance
so that P j is
Figure P l l .35
ANSWHR: (a) 48 ih . r .0 3 7 watts
(b)
maximized.
Find the values o f L and C , to achieve
36. The circuit shown in Figure PI 1.36 has
the impedance Zy computed in part
v>v,(/) = 10>/2 cos(60/) V, /?^ = 4
(a).
Find the impedance Z^ such that Pj
(c)
I = 1/120 H, and Q = 1/30 F. What value o f
and IV-,1 are maximized.
R should be chosen so that maximum power is

delivered to the load? Note: It is the source resis
tor here, not the load resistance, that is being
varied. What is this maximum average power
consumed by the load?
Load
/
zrc,
Figure P 11.38
'" " Q
A N SW FRS: (a) Z^ = 100 f ylOOO Q: (b) 99 piand 0.2 niH
39. The series RLC circuit o f Figure PI 1.39 has

Figure Pi 1.36
ANS\\'|-:RS: 0. 25 waits
reached steady state and v^{t) = 110sin(l 207tr)

V.
/
37. The circuit o f Figure PI 1.37 operates in the

sinusoidal steady state with
v^(O = 5 0 V 2 co s(2 0 0 0 r) V.
(a)
6Q
v .(t )Q
-jisn
Choose R and C such that the maximum

average power is absorbed in the load
resistor R^^ = 5
when Z. = 0.1 mH.
What is this maximum average power?
Figure P I 1.39
542
(a)
Find the instancaneous stored energy

and
Lîve
at the moment when the
2(0
terminal voltage o f the source is zero.

(b)
Find the instantaneous energ)^

the moment when
(c)
at
42. (a)
For the circuit o f Figure P H .4 2 a ,

show that the powers absorbed by the
= 0.
impedance Z = R + jX are
Find the instantaneous energ)' \Vf- at
the moment when

= 0.
ANSWHRS: in random order (J), 1.14, 1.65,
and Q =
(b)
= /?|lp
where I is in
For the circuit o f Figure P H .4 2 b ,

show that the powers absorbed by the
4.115, 1.27
admittance Y= G +jB are
= (j|Vp
and Q = 5|Vp where V is in V
THEORETICAL PROBLEMS
L
40.
Let the voltage across a capacitance C b e

v{t) =
sin(cor) V.
(a)
Find p{t), the instantaneous power

delivered to the capacitance and show
that p{t) has a peak value o f
0 .5 c o C (V J2 watts and an average
value o f 0.
(b)
Find
the instantaneous energy
store in C (or, rather, in the electric

field) and show that W^t) has a peak
value o f 0.5C(1/^^)2 joules and an aver
age value o f 0
(c)
. 2
j oules.
Let Q^- be the reactive power absorbed

by C. Show that
WC,ave _
Qc
2(0
41. Let the current flowing through an induc

tance L be i{t) =
sin(co^) A.
(a)
Find p{t), the instantaneous power

delivered to the inductance and show
that
p{t)
has
peak
value
of
0 .5 co(/,)2 watts and an average

value 0,
(b)
Find
the instantaneous energy
store in L (or, rather, in the magnetic
field) and show that W^{t) has a peak
value ot
joules and an aver
age value o f 0.25I(/^,)^ joules.

(c)
Let
be the reactive power absorbed
bv L. Show that
Z = R + jX
o --------(a)
+
V
Y = G + jB
a
(b)
Figure Pi 1.42. Two-terminal elements modeled

via impedance (a) and admittance (b).
Laplace Transform Analysis I:

Basics
HISTORICAL NOTE
The Laplace transform converts a time function into a new function o f a complex variable via an
integration process. The name Laplace transform comes from the name o f a French mathemati
cian, Pierre Simon Laplace (1 7 4 9 -1 8 2 7 ). Pierre Laplace adapted the idea from Joseph Louis
Lagrange ( 1 7 3 6 -1 8 1 3 ), who in turn had borrowed the notion from Leonhard Euler
(1 7 0 7 -1 7 8 3 ). These early mathematicians set the stage for converting complicated diff-erential
equation models o f physical processes into simpler algebraic equations. The Laplace transform
technique allows engineers to analyze circuits and to calculate responses quickly and efficiently.
In turn engineers became better able to design circuits for radio communication and the tele
phone, not to mention other, earlier electronic conveniences. This chapter introduces the notion
o f the Laplace transform, a mathematical tool that is ubiquitous in its application to an army o f
engineering problems.
CHAPTER OBJECTIVES
1.
2.
3.
Explain and illustrate the benefits o f using the Laplace transform tool for solving circuits.
Develop a basic understanding o f the Laplace transform tool and its mathematical prop
erties.
Develop some skill in applying the Laplace transform to differential equations and cir
cuits modeled by differential equations.
SECTION HEADINGS
1.
2
3.
Introduction
Review and Summary o f Deficiencies o f Second-Order Time Domain Methods
Overview o f Laplace Transform Analysis
4.
5.
Basic Signals
The One-Sided Laplace Transform
6.
7.
The Inverse Laplace Transform

More Transform Properties and Examples
Chapter 12 Laplacc Transform Analysis 1; Basics
8.
Solution o f Integrodififerential Equations by the Laplace Transform
9.
10.
Summary
Terms and Concepts
11.
Problems
1. INTRODUCTION
I'his chaprer introduces a powerful mathematical tool for circuit analysis and design named the
Laplace transform. Later, more advanced courses will describe the design aspects. Use o f the
Laplace transform is commonplace in engineering, especially electrical engineering. A student
might ask why such a potent tool is necessary for the analysis o f basic circuits, especially since
many texts use an alternative technique called complex frequency analysis. Complex frequency
analysis does not permit general transient analysis; rather, it restricts source excitations to sinu
soids, exponentials, damped sinusoids, and dc signals. This class of signals is small and does not
begin to encompass the broad range o f excitations necessary for general circuit analysis and the
related area o f signal processing. The Laplace transform framework, on the other hand, permits
both steady-state and transient analysis of circuits in a single setting. Additionally, it affords gen
eral, rigorous definitions of impedance, transfer fimctio}!, and various response classifications perti
nent to more advanced courses on system analysis and signal processing. Introducing the Laplace
transform early allows students an entire semester to practice using the tool and learn about its
many advantages.
Section 2 describes some of the difficulties associated with the methods o f circuit analysis intro
duced in earlier chapters when applied to circuits o f order 3 or higher. Following this, we present
an overview o f Laplace transform analysis in section 3, define important basic signals in section 4,
and introduce the formal definition o f the one-sided Laplace transform in section 5. The inverse
Liplace transform and important properties o f the transform process arc introduced in sections 6
and 7, with numerous illustrative examples. Section 8 applies the technique to circuits modeled
by differential equations. Such models were developed in Chapters 8 and 9.
2. REVIEW AND SUMMARY OF DEFICIENCIES OF "SECONDORDER" TIME DOMAIN METHODS

Recall that the output or response o f a circuit depends on the independent source excitations, on
the initial capacitor voltages, and on the initial inductor currents. Calculation o f the output often
begins with the writing o f an algebraic or a differential equation model o f the circuit for the out
put variable in terms o f the source excitations or inputs and element values. For first- and secondorder circuits with simple .source excitations, such as dc or purely sinusoidal, the solution o f the
differential equation circuit model has a known general form containing arbitrary constants. See,
for example, Tables 9.1 and 9.2. The arbitrary constants depend on the initial conditions and the
magnitude o f the dc excitation or on the magnitude and phase o f the sinusoidal excitation.
Specifically, the steps in finding the response o f a second-order circuit to a constant input are as
follows;
Chapter 12 Laplace Traiisforni Analysis 1: Basics
Step 1. Generate a differential equation model o f the circuit.

Step 2. Compute the characteristic equation o f the circuit/differoitial equation and then compute its
roots {say Aj and X-y) using, for example, the quadratic fonnula in the second-order case.
Step 3. From the location ofthe roots o f the characteristic equation, determine the form o f the solution:
or if A, = X-),
Step 4. Compute the constant D by shorting itiductors, open-circuiting capacitors, and analyzing the
restdting resistive circuit.

Step 5. Compute the constants A and B using the initial conditions on the circuit.
For circuits beyond second order, the approach in the above algorithm tends to break down.
Example 12.1 demonstrates how the approach breaks down with a simple third-order circuit.
As mentioned earlier, the foregoing technique, although quite useful for simple circuits, has seri
ous drawbacks for circuits with more than two capacitors or inductors. This is because higherorder derivatives o f circuit output variables generally have little or no physical meaning. Such
derivatives are complicated linear combinations o f initial capacitor voltages and initial inductor
currents. The following example illuminates the difficulties.
E X A M P L E 12.1. Figure 12.1 shows three
circuits coupled through the use o f dependent volt
age sources. The goal o f this example is to construct a differential equation model, determine the
solution form in terms o f arbitrary constants, and demonstrate the difficulties w'ith the simple
recipe o f the above algorithm by attempting to relate the arbitrary constants to the initial condi
tions.
1Q
1Q
1O
-I-
FIG URE 12.1 A cascade of three RC circuits coupled by means of dependent voltage sources. The
differential equation model of the circuit is third order.
S o l u t io n
Step 1. Construct the differential equation o f the circuit. For this task, first write a dilTerential equa
tion relating
to
Then write one relating
to
and finally, write one relating
Some straightfor^vard algebra leads to the following three differential equations:
to
Chapter 12 Laplacc Transform Analysis 1: Basics
at
+ v c i ( 0 = 0.5iv
( 1 2 .!)
0 . 5 - ^ + it2(/ ) = 0 .5 i'c i
(12.2)
0 . 2 5 ^ . v , ( O = 0 ,5 v
^ , ^ 3 ^
Successively substituring equation 12.1 into equation 12.2 and equation 12,3 into the result pro
duces the input-output differential equation model,
ônt
-7 ^~ôut , 1 A
dt^
dr
, o
(12-4)
d,
Step 2. Compute the characteristic equation and its roots. The characteristic equation for differen
tial equation 12.4 is
^ + 7s^ + \As + ^ = {s - a) {s - b) {s - d) = {)
which has roots a = - \, b = - 2 , and d = - 4 .

Step 3. Determine the form o f the solution. If v-^j^t) =
form
Vou, it) =
+ Be^
then the complete solution has the
+ E = Ae~^ + Be~~' + De~^ +
(12.5)
for r > 0.
Step 4 . Compute A, B, D, and E in equation 12.5. Using the rule o f thumb mentioned earlier, a
simple calculation yields E = 0.125
Calculation o f A, B, and D specifies the solution in equa
tion 12.5. Applying the recipe described earlier, we take derivatives o f equation 12.5. set t = 0, and
relate them to the circuit initial conditions:
ôut (0) = ^
'ôut (0) =
+ dD
Again, one dot over a variable means a first-order time derivative, and two dots denotes a second-order
time derivative. A, B, and D are computed by solving this set o f equations. The difficulty is in specifying
and
First,
is simply the initial capacitor voltage on the third capaci
tor. However, v^^^ît) is proportional to the current through it, which depends on all the initial
capacitor voltages. Further, what is the physical interpretation o f
And how do v^^^{0)
and v^^fiO) relate to the initial capacitor voltages? The relationship is complex and lacks any mean
ingful physical interpretation. Finally, even for this simple example, computation and solution o f
the differential equation 12.4 proves tedious.
Chapter 12 Liplacc Transform Analysis I: Basics
Exercise. For Example 12.1, compute expressions for

and
(0).
A N SW ERS:
= [2iv..(0) - 4
(0) = [\6r^J0) - 1 2 ^ ,.(0 ) + 2/v^(0)]
One o f the advantages o f Laplace transform analysis is that it does not destroy the physical mean
ing o f the circuit variables in the analysis process. Chapter 13 addresses how the Laplace transform
approach explicitly accounts for initial capacitor voltages and initial inductor currents.
3. OVERVIEW OF LAPLACE TRANSFORM ANALYSIS

Laplace transform analysis is a technique that transforms the time domain analysis o f a circuit, sys
tem, or differential equation to the so-called frequency domain. In the frequency domain, solu
tion o f the equations is generally much easier. Hence, obtaining the output responses o f a circuit
to known inputs proceeds more smoothly.
To apply the technique, one takes the Laplace transform o f the time-dependent input signal or
signals to produce new signals dependent only on a new' frequency variable s. In an intuitive sense,
and as precisely derived later, one also takes the Laplace transform o f the circuit. Assuming zero
initial conditions, one multiplies these two transforms together to produce the Laplace transform
o f the output signal. Taking the inverse Laplace transform o f the output signal by means o f known
algebraic and table look-up formulas yields the desired response o f the circuit. The effect o f initial
conditions is easily incorporated.
Figure 12.2 is a pictorial rendition o f the method. As just mentioned, one transforms the input
signal, transforms the circuit to obtain an equivalent circuit in the Liplace transform world, and
computes the Laplace transform o f the output by multiplying the two transforms together.
Inverting this (output) transform with the aid o f a lookup table or MATLAB produces the desired
output signal.
> Output Signal
Input Signal
\
Laplace Transfornn
of Input Signal
'' r
Laplace Transform
of C I R C U I T ^ ,
Laplace Transform
of Output Signal
F I G U R E 12.2 D iagram show ing flow o f Laplace transform circuit analysis.
S4iS
Chapter 12 Liplacc Transform Analysis I; Basics
In a mathematical context, one executes the same type o f procedure on a difFerential equation
model o f a circuit and, indeed, difFerential equations in general. Figure 12.3 illustrates the idea.
Input Signal
DIFFERENTIAL'
EQUATION
> Output Signal
Laplace Transform
of Input Signal
Laplace Transform
of DIFFERENTIAL
EQUATION
Laplace Transform
of Output Signal
FIG URE 12.3 Diagram showing flow of Laplace transform analysis for solution
o f differential equations.
The benefit o f this t)'pe o f analysis lies in its numerous uses. Some o f these uses include steadystate and transient analysis o f circuits driven by complicated as well as the usual basic signals, a
straightforward lookup table approach for computing solutions, and explicit incorporation of
capacitor and inductor initial conditions in the analysis. The forthcoming sections will flesh out
these applications.
4. BASIC SIGNALS
Several basic signals are fundamental to circuit analysis, as well as to future courses in systems
analysis. Perhaps the most common signal is the unit step function.
(r) =
1,
r>0
0,
r<0
( 12.6)
defined in Chapter 8. The bold line in Figure 12.4, resembling a step on a staircase, represents the
graph o f u{i).
u(t)
FKiURE 12.4 Graph of the unit step function. It often represents a constant voltage or current le\'el.
The unit step function has many practical uses, including the mathematical representation o f dc
voltage levels. Any t)'pe o f sustained, constant physical phenomenon, such as constant pressure,
constant heat, or the constant thrust o f a jet engine, has a step-like behavior. In the case o f jet
engine thrust, a pilot sends a command signal through the control panel to the engine requesting
a given amount of thrust. The step function models this command signal.
Chapter 12 * Ijp lacc Transform Analysis 1: Basics
The shifted step, shown in Figure 12.5, models a rime-delayed unit step signal.
u(t-T)
1 --
FIGURE 12.5 Graph of a unit step shifted T units to right.

This function is often used to represent a delayed startup.
Shifted steps, u (t- 7), often represent voltages that turn on after a prescribed time period T. The
flipped step fioiction, u{T - t), o f Figure 12.6 depicts yet another variation on the unit step. Here
the step takes on the value o f unity for time t ^ T. Often it provides an idealized model o f signals
that have excited the circuit for a long time and turn o ff at time T. The key to knowing the val
ues o f these various step functions is to test whether the argument is non-negative or negative.
Whenever the argument is non-negative, the value is 1; when the argument is negative, the value
IS zero.
u(T-t)
FIGURE 12.6 Graph of flipped and shifted unit step. This function is often used to model signals
that have been on for a long time and turn ofTat time /'.
Exercise. Represent each o f the following functions as sums o f step functions;
(0 / ( 0 =
1,
0</<2
0,
otherwise
I, -3</<6
(///) m =
0.
otherwise
1,
/ < -l
0,
-l< r < l
1,
; > 1
ANSVC^R.S: in random t)rder: //(-I - r) +
u {t
- 1), wu! - u{i - 2),
ii{ t
+ 3) - i d t - {>)
The pube function, p-î), o f Figure 12.7 is the product o f a step and a flipped step or, equivalent
ly, the difference o f a step and a shifted step. Specifically, a pulse o f height A and width 7" is
=
Au{i)u{T -
1) =
Au{t) - A u { t - T)
(12.7)
Chapter 12 Laplace Transform Analysis 1: Basics
S50
Pr(t)
>
T
FIGURH 12.7 Pulse of width T and height A.
This function is often used to model signals o f fixed magnitude and short duration.
A signal sharing a close kinship with the unit step is the ramp function r{t) depicted in Figure
12. 8.
r(t)
FIGURE 12.8 Graph of the ramp function, r{t) =

Ramp functions conveniently model signals having a constant rate of increase.
The ramp function is the integral o f the unit step, i.e.,
/(0=
w h e re t
J oo
ii(T)dx =ln{t)
( 12.8)
is s i m p l y a d u m m y v a r i a b l e o f i n t e g r a t i o n .
Exercise. Plot r { - t) and r{t - 2).

ANSW ER: i\-t) is the relleciion of ;(;) about the vertical axis while >\t - 2) is siniplv the shift of
lit) by two units to the rinht.
EXA M PLE 12.2

Express Figures 12.9a and b in terms o f steps and ramps.
f,(t)
y t)
F I G U R E 12,9 T w o signals to be represented by steps and ramps.
5S1
o l u t io n
For the signal/j (/), observe that the signal begins with a ramp with a slope o f 2. Thus we have
f\ U) = 2r(f) + ? . At ; = 7, the signal/j(^) levels off. Since the 2r{t) part o f the signal continues to
increase, the increase must be canceled by another ramp o f slope 2, but shifted to the right by T
units. Thus,/j(r) = 2r{t) - 2r(t - T).
The signal j 2(^) replicates/j(^) up to 3T . After 37', the signal drops to zero. Hence we must sub
tract a shifted step o f height 2 7 from/j(/), Thus fjit) =/j(f) - 2Tu{t - 3 7 ) = 2r{t) - 2r{t - T) -
2T u {t-5 T ).
Exercises. 1. Figure 12.10 depicts a sawtooth waveform denoted hy J{t). Sequences o f sawtooth
waveforms are used as timing signals in televisions and other electronic devices.
f(t)
A
FIGURE 12.10 A sawtooth waveform.

A N SW T .R : / / ) = fit) -;(/- 1) - u{r - 1)
2. For/r) o f Figure 12.10, p lo t/ l - t) and represent the ftmction in terms o f steps and ramps.
AN SW ER: /(I - /) = ;-(l - r) - >i-t) - u{-t)
EXA M PLE 12.3. Express Figure 12.11 in terms o f steps and ramps.
f(t)
> t
FIG URE 12.1 I Triangular waveform to be represented by steps and ramps.
S
o l u t io n
Observe that the signaly(f) begins with a ramp with a slope o f 1 at r = -T . Thus /(f) = r{t 7) +
?. The signal falls o ff with a linearly decreasing ramp for 0 < r < T. If we subtract r{t) the signal
would become flat for r > 0. Thus we subtract 2r{t) to obtain the linear decrease. Hence, y(r) =
Chapter 12 Liplacc Transform Analysis 1: Basics
^52
r{t + 7) - 2>it) + ?. For r > T, this signal, r{t + T) - 2Kr), continues to linearly decrease. Hence
for the signal to be zero for t > T, we cancel this decrease with an additional ramp. T h u s/ r) =
r{t + 7) - lr{t) + r { t - T).
Newtonian physics provides a good motivation for defining the ramp signal. Applying a constant
force to an object causes a constant acceleration having the functional form Ku{t). 'J'he integral o f
acceleration is velocity, which has the form
a ramp function.
A very common and conceptually useful signal is the (Dirac) delta function, or unit impulse
Function, implicitly defined by its relationship to the unit step as
h{cj)dq
(12.9)
The relationship o f equation 12.9 prompts a natural inclination to define

s/ .
, V
I-
u{t)~ ii{t - h)
h
0 {{) = //(/)= hm----------------(h
//->{)
( 12. 10)
Strictly speaking, the derivative o f u{t) does not exist at t = 0, due to the discontinuit)' at that
point. Without delving into the mathematics, one typically interprets equations 12.9 and 12,10
as follows: define a set o f continuous differentiable functions
The derivatives,
as illustrated in Figure 12.12a.
5^(/) = -y//^(/) >of these functions are depicted in Figure 12.12b.

5A(t)
-t
(a)
(b)
F IG U R l'. 12.12 (a) Continuous differentiable approximation to the unit step.

(b) Derivative of
the integral of 6^(/) produces
Clearly, 6^(/-) has a well-def'med area o f 1, has height I/A, and is zero outside the interval 0 s t
s A. In addition, u[t) = lim //^(r), and lim 6^ = b{t) as A - 0.
Hence, although the definition
o f equation 12.10 is not mathematically rigorous, one can interpret the delta function as the limit
o f a set o f well-behaved functions. In fiict, the delta function can be viewed as the limit o f a variet)' o f different sets o f functions. A problem at the end o f the chapter explores this phenomenon.
Despite the preceding mathematics, the delta function is not a function at all, but a distribution,^
and its rigorous definition (in terms o f so-called testing functions) is left to more advanced math
ematics courses. Nevertheless, we shall still refer to it as the delta or impulse function. The stan
dard graphical illustration o f the delta function appears in Figure 12.13, which shows a pulse of
Chapter 12 Laplace Transform Analysis I: Basics
infinite iieight, zero width, and a well-defined area o f unit}', as identified by the 1 next to the
spike. Visualization of the delta function by means o f the spike in the figure will aid our under
standing, explanations, and calculations that follow.
FIGURE 12.13 Standard graphical illustration o f a unit impulse lunction having a wcll-dcfincd area
of 1. The function typically represents an energ)' transfer, large force, or large impact over a very
short time duration, as might occur when a bat hits a baseball.
The unit area property follow's from equation 12.9, i.e.,
bU)(lr=
5(rV/r = / / ( 0 ^ ) - / / ( 0 " ) = l
where 0 is infinitesimally to the left and 0"^ infinitesimally to the right o f zero. I f the area is dif
ferent from unity, a number Kalongside the spike will designate the area; i.e., the spike will be a sig
nal Kb{t).
One motivation for defining the delta function is its abilit)' to ideally represent phenomena in
nature involving relative immediate energ\' transfer (i.e., the elapsed time over which energy trans
fer takes place is very small compared to the macroscopic behavior o f the physical process). An
exploding shell inside a gun chamber causing a bullet to change its given initial velocity from zero
to some nonzero value instantaneously is an example. Another is a barter who hits a pitched ball,
instantaneously transferring the energy o f the s\vung bat to the ball. Also, the delta function pro
vides a mathematical setting for representing the sampling of a continuous signal. Suppose, for
example, that a continuous signal v{t) is to be sampled at discrete time instants t^ r,*
v{t) is to be physically measured at these time instants. The mathematical representation o f this
measuring process is given by the sifting property of the delta function.
v(//)=
(12.11
v(l)5(X
In other words, the value o f the integral is the non-impulsive part o f the (continuous) integrand,
replaced by that value o f r which makes the argument o f the impulse zero, in this case r = tj.
Verify'ing equation 12.11 depends on an application o f the definition given in equation 12.9.
Specifically, if v{t) is continuous at f = t-, then
\'(T )6 (T
1: )dX =
v (T
)5(T 5(1 -
f j )dx
rj)(lx = v(tj)
SS4
by equation 12.9, where
are infinitesimally to the right and left o f f,.
Exercises. 1. Compute the derivatives o f the signals in Figures 12.9a, 12.9b, and 12.11.
A N SW E R S;/,(/) = 2|//(/)-/ / (/ -T ) ] . / .(z) = 2[//(/) - / / ( / - 7 ')| - 276 (/ - 3 7 ').
) = //(/+
T)-2iiii)+
oo
2. Compute the integral

ANSWHR; sin(l.5.T)
//(/-
T)
sin(27T/+ 0.57t)5(/ 0.5)^//.
-oo
5. THE ONE-SIDED LAPLACE TRANSFORM

Intuitively, a transform is like a prism that breaks white light apart into its colored spectral com
ponents. T he one-sided or tmilaternl Laplace transform is an integral mapping, somewhat like a
prism, between time-dependent signals y(r) and functions F{s) that are dependent on a complex
variable s, called complex frequency.
LAPLACE TRANSFORM
Mathematically, the one-sided Laplace transform
J{t) is
( 12. 12)
where s = (5 + ./CO(y = yf-A) is a complex variable ordinarily called a complex frequency im
the signals and systems literature.
As the equation makes plain, the Laplace transform integrates out time to obtain a new func
tion, F{s), displaying the frequenc)' content o f the original time function/r). In the vernacu
lar, F{s) is the frequency domain counterpart of/ r). Analysis using Laplace transforms is often
called frequency domain analysis.
Exercises. 1. Find the Laplace transform o f a scaled Dirac delta function, Kb{t).
ANSWF.R; K
2. Find [sin(2;rr + 0 .5 jr ) 6 ( r - 0.5)].
ANSWHR: sin(l.5.T)<'-'^ ^
Chapter 12 Laplacc Transform Analysis I: Basics
A number o f questions about the Laplace transfbrm promptly arise:
Question 1: Why is it called one-sided or unilateral?

Answer: It is called unilateral because the lower limit o f integration is 0 as opposed to -oo. If the
lower limit of integration were -oo, equation 12.12 would be called the two-sided Laplace trans
form , which is not covered in this text.
Q uestion 2 : Why use 0 instead ofQ * or 0 as the lower lim it o f integration?
Answer: Our future circuit analysis must account for the effect o f instantaneous energy transfer
and, hence, impulses at / = 0. The use o f 0"^ would exclude such direct analysis, since the Laplace
transform of the impulse function would be zero. Using ^ = 0 is simply ambiguous.
Question 3 : What aboutJunctions that are nonzero fo r t < 0.^

Answer: Because the lower limit o f integration in equation 12.12 is 0, the Laplace transform does
not distinguish between functions that are different for f < 0 but equal for ^ > 0 (e.g., (/) and
u{t + 1) would have the same unilateral Laplace transform). However, since ^ = 0 designates the
universal starting time o f a circuit or system, the class o f signals dealt with will usually be zero for
t < 0 and thus will have a unique (one-sided) Laplace transform. Conversely, each Laplace transform F{s) will determine a unique time fiinaion J{t) with the property that f^t) = 0 for ^ s 0.
Because o f this dual uniqueness, the one-sided Laplace transform is said to be bi-unique for signals/^) yfirh J{t) = 0 for ^ < 0.
Question 4 : Does every signalj{t) such th a tfj) = 0 fo r f < 0 have a Laplace transform?
Answer: No. For example, the function f i ) = ^ (/) does not have a Laplace transform because
the integral o f equation 12.12 does not exist for this function. To see why, one must study the
Laplace transform integral closely, i.e..
Observe that e~j^* = cosicot) -jsm{(Ot) is a complex sinusoid. As f approaches infinity, the real and
imaginary parts o f the integrand in equation 12.13 must blow up, due to the
term. Hence,
the area underneath the curve e^~^ grows to infinity, and the integral does not exist for any value
of a.
W h enever//) is piecewise continuous, a sufficient condition for the existence o f the Laplace trans
form is that
I
f.
\f{t)\<kx^^
for some constants
and
(12.14)
This bound restricts the growth of a function; i.e., the fimction can
not rise more rapidly than an exponential. Such a fiinaion is said to be exponentially bounded. The
Chapter 12 * Uplacc Transform Analysis 1: Basics
condition, however, is not necessary for existence. Specifically, the transform exists whenever the
integral exists, even if the function/f) is unbounded. W ithout belaboring the mathematical rigor
underlying the Laplace transform, we will presume throughout the book that the functions we are
dealing with are Laplace transformable.
Question 5: Why does the existence o f the Laplace transform integral depend on the value o f a , men
tioned in the answer to question 4?
Answer: If the condition in equation 12.14 is satisfied, then there is a range o f a s (recall that s =
a + yoo) over which the Laplace transform integral is convergent. This is explained in the follow
ing example.
EXA M PLE 12.4. Find the Laplace transform o f the unit step. By equation 12.12,
{u(i)\ = U(s)= \Z
O'
(12.15)
1
a +
provided that o > 0. Notice that if a > 0, then
+ 7 (0
.V
-* 0 as t
This keeps the area under
neath the curve finite. For a < 0, the Laplace transform integral will not exist for the unit step.
The smallest number Oq such that for all a > Oq the Laplace transform integral exists is called the
abscissa o f (absolute) convergence. In the case o f the unit step, the integral exists for all a > 0;
hence, Oq = 0 is the abscissa o f convergence. The region a > 0 is said to be the region o f conver
gence (RO C) of the Laplace transform o f the unit step. Figure 12.14 illustrates the R O C for the
unit step.
j (o-axis
-f->a-axis
FIGURE 12.14 Region of convergence, a > 0, o f the Laplace transform of the unit step function
(i.e., the Laplace transform integral will exist for all a > 0).
Question 6: Is the unilateral Laplace transform valid only in its region o f convergence?
Answer:
the answer is no. There is a method in the theory o f complex variables called ana
lytic continuation which, although beyond the scope o f this text, permits us to uniquely and anal)aically extend the transform to the entire complex plane.- Analytically means smoothly and also
that the extension is valid ever)^vhere except at the poles (to be discussed later) o f the transform.
Thus, the region o f convergence goes unmentioned in the standard mathematical tables o f one
sided Laplace transform pairs.
Chapter 12 Liplace Transform Analysis I: Basics
EXA M PLE 12 . 5 . Find F{s) iorJ{t) = Ke

So
u{t), where K and <{ are scalars.
lu t io n
Applying equation 12.12 yields
poo /
Ke~^ e~^dt = K
s + ci
(12.16)
The integral exists if Re[j + ^] > 0. If^7 is real, then the R O C is a > -a. As mentioned in the answer
to question 6, by analytic continuation, F{s) = M{s + a) is valid and analytic in the entire com
plex plane, except at the point s = -a . The point s =-a\s a pole o f the rational function M{s + a)
because as s approaches -a , the value o f the function becomes infinitely large.
The preceding discussion and examples set up the mathematical framework o f the Laplace trans
form method. Our eventual focus rests on its application to circuit theor)\ which builds on two
fundamental laws: Kirchhoffs voltage law (KVT) and KirchhofTs current law (KCL). KVL
requires that the voltage drops around any closed loop sum to zero, and KCL requires that the
sum o f all the currents entering a node be zero. For the Laplace transform technique to be useful,
it must distribute over such sums o f voltages and currents. Fortunately, it does.
Linearity property: The Laplace transform operation is linear. Suppose j{t) =
Then
L [ m = L [ a jl{t) +
= ^,[/i(r)] + a.L\f,{t)]
= a^F^{ s ) â, f , St )
(12.17)
This property is easy to verify since integration is linear;

l l/ | (0 +
= l
O'
/]{t)e~^'dt + fl2 [q- fi{t)e~^'dt
This is precisely what equation 12.17 states. Hence, our curiosit)^ satisfied, we may rest peaceful
ly in the knowledge that the Laplace transform technique conforms to the basic laws o f KVL and
KCL.
E X A M PLE 12.6. Find F{s) wheny(r) = K^ti{t) +
So
for real scalars A', ,
and a.
lu t io n
The Laplace transform o f u{t) is \h by equation 12.15 and that of
is ]/{s + a) by equa
tion 12.16. By the linearity property (equation 12.17),

F(.v) = ^
+- ^ ,
s +a
with region o f convergence (a > 0} H {a > - a }, where H denotes intersection.
By analytic con
tinuation, the transform is valid in the entire complex plane except at the poles, s = 0 and s = -a.
(Henceforth we will not mention the RO C in our calculations.)
Exercise. Find the Laplace transforms o f

{i) J{t) =
+ 2 r t ( r ) + 2u(t),
(//)
J{t) = -2u(t) +
(iil)
J{t) = 5u(t) - 4e~û{t) + 2e~"û{t)

?>s - a
AN SW ERS: (/) - .
,v~ - (I~
- 2e~^û{t), and
2
.V/- A
...
.v - (/
2
, + -----,
4
The transform integral o f equation 12.12 has various properties. These properties provide short
cuts in the transform computation o f complicated as well as simple signals. For example, the
Laplace transform o f a right shift o f the s i g n a l a l w a y s has the form e~^^F{s), T > 0. Shifts are
important for two reasons:
1.
2.
Many signals can be expressed as the sum o f simple signals and shifts o f simple signals.
Excitations o f circuits are often delayed from t = 0.
Hence, provisions for shifts must be built into analysis techniques.

Tim e shift property: If [/{i)u{t)] = F{s), then, for T > 0,
[ f { , - T l u i i - T ) ] = r 'rF{s)
(12.18)
Verification o f this property comes from a direct calculation o f the Laplace transform for the shift
ed function, i.e.,
[ f ( t -T )u O - 7)1 = J " / ( r - T )u (t- D e'^ di = J " / (; -T )e~ ^ d t

Let cj = t -T ,W k WJ q = dt. Noting tliat the lower limit o f integration becomes 0 with respect to q,
L\S(t -T )u U - D 1 = r
fUl)e~'ê~^'^dq =
f /(qye-^ 'dq =
O'
F^s)
Observe that if T < 0, the property fails to make sense, since J { t - T )ii{t- 7) would then shift left.
Since the transform ignores information to the left o f 0 one cannot, strictly speaking, recover
J{t) from the resulting transform.
Exercises. I. Find L\J{t - T)] w hen/f) is (i) Ad(t), (ii) Au(r), and (iii) Ae~''u(t).
2. pyU) = /iu(t) - Aii{t - T).
A N SW ERS: In random order:
-----
.\----- ^
s+ a
E X A M PLE 12.7. Using the tim e shift property, find F{s) for the signal
12.15.
sketched in Figure
f(t)
3
2
1
-1
-2
FIGURE 12.15 Signal for Example 12.7.

S o l u t io n
Using step functions and shifted step functions, we obtain

p )
= 5 u {t)-5 u {t-\ )^ 2 u {t-2 )
3 5e~^
2e~
Direct application o f linearity and the time shift propert) yields F {s) = ------------f- -
Exercises. 1. Find the Laplace transform o f the pulse signal o f equation 12.7.
2. Find F{s) when/^) = Aû[t - T^) + A2 ^(t - T-^ + Aû{t - T^).
1
AN SW ERS:
~sl
v7i
.4
P j(s) = A ----- . F ( s ) = ^ -------^ +
---------------------
One more property allows us to revisit the signals discussed in section 3 and take their Laplace
transforms. The new propert)^ is multiplication of//) by t. This always results in a Laplace trans
form that is the negative o f the derivative o f F{s).
M ultiplication-by-f property: Let F{s) =
Then
[tf{t)\ = - F(5)
ds
(12.19)
Verification o f this property follows by a direct application o f the Laplace transform integral to
with the observation that te~^^ = ------ In particular,
ds
S60
-St
oo
n o
j()
dt =
.d s
as
ds J
Table 12.1 lists this transform pair, as well as numerous other such pairs, without mention o f the
underlying region o f convergence. As mentioned earlier, we shall dispense with any mention of
the ROC, assuming that all functions are zero for r < 0.
EXA M PLE 12.8. Find the Laplace transform o f the ramp function, r(r) = tu(t).
S o l u t io n
Using equation 12.19,
M ds \ s ,
R {s)= L\r{t)\ =
E X A M PLE 12.9. Supposey{r) = te
( 12.20)
ds
where a is real. Find F{s).
S o l u t io n
The quickest way to obtain the answer is to apply equation 12.19. Specifically, since
^ r nt , , T
s+a
L
ds s + a
ds
{s + a ) -1
( 12.21)
An alternative, more tedious approach is to use integration by parts as follows:
F {s)= L
where v = t and dti = e
=:
te
oo
fOO
ydii = uv
oo
_
0"
O'
udv
dt. Thus,
oo
le
.V+ a
dt
The RO C is a > -a , in which case the first term on the right-hand side is zero. Thus, in this RO C ,
evaluation of the integral term implies that
F {s)'0
s+a
dt =
(s + aY
Equation 12.21 is a special case ol the more general formula
\t''e-^û(t)] =
nl
(s +a)71+ 1
(1 2 .2 2 )
Chapter 12 Liplace Transform Analysis 1: Basics
561
Exercise. Find the Laplace transform oij{t) = p-e

AN SW ER: F{s) =
(\ + a r
EXA M PLE 12.10. Find F(s) for the signal depicted in Figure 12.16.
f(t)
FIGURE 1 2 . 1 6 A signal to be represented by steps and ramps.

S o l u t io n
From Example 12.2,/ r) = 2r(^) - 2r(t - T) - 2 T u { t - 3 7 ). Hence by linearity, the time shift prop
erty', and equations 12.15 and 12.20,
F {s) = L \ 2rU )- 2r{t - T ) - 2Tu(t - 3 7 )
2 - 2 '^
Exercises. 1. Note that the sawtooth o f Figure 12.17 is/ f) = t[u{t) - u {t- 1)]. Suppose
= z/(r)
- u {t- 1). Compute F (5 ) = - -G ( i')-
d.s
f(t)
2. Use equation 12.22 to compute the Laplace transforms o f/ ;) = tr{t) for the ramp function r{t)
and forjit) = p-r{t).
.s'
AN SW TRS: .
562
EX A M PLE 12 .1 1 . The circuit o f Figure 12.18a has two source excitations, /j(/) and /2W> shown
in Figures 12.18b and c. Compute
V o Jt)
(a)
(c)
FIGURE 12.18 (a) Resistive circuit driven by two current sources.
(b) Triangular signal, /,(r), in A. (c) Pulse signal, ijit), in A.
S o l u t io n
Step 1. Find the form ofV^^^{s). By superposition and Ohms law,

= 10/, U) + 10/2(r)
From the linearity o f the Laplace transform,
= 10/, W + lOAW
Step 2. Compute /,(j) and
Some reflective thought yields /,(?) = 2ti{t) - 2r{t) + 2r{t - 1) A
and ijit) = \.5u{t) - \ 5 u {t- 1) A. From linearity, the time shift propertys and the previously com
puted transforms.
2
h{s) = - ~
S
Step 3. Find V..Js). Since V
l + 2e
s'
and l 2 {s) =
(s) = 10/As) + lOAU), ir follows that
your(s) = - - ^
+ e~^
20
l5^
Chapter 12 * Laplace Transform Analysis I: Basics
S63
Step 4. As an introduction to the next section, by inspection we can compute the time function
o f the output voltage:
= 35(/) - 20r{t) - \5 u (t- 1) + 2 0 r{t- 1)
Exercises. 1. Find the Laplace transform o f (i)^(^) and (iii)^ (r) =
>r e
(ii)
=e
+e
+ te
+ te
2) + (t - 3).
i\NSV('T,RS: in random order: !----- 1-----!-----j------- !------- 1------- ?----- . !------- y 5

.v + c/ .v + /> (.v + f/)(.v + /?)~ (.v + f/) (.v-c/)
- --------h
.v+ ^/
2. Recall that cos(cor) = ------------------ . Show that the Laplace transform ofJ{t) = cos{cot)u{t) is
. r + 0)
3. Recall that sin(CO/) = ------------------ Show that the Laplace transform o f/ r) = sin(ct)f)z^(f) is
(0
F{s) = - ------,
4
4
4. Find the time functions associated with Fi ( j) = , F) (s) = ----------^
^
(s + 2)-
AN SW ERS: /;(/) = Au{t)
= Atr"H{() , /^(/) =
4 . -
= ---------.
-v+ 4
- 4)
We end this section with Table 12.1, which lists a number o f Laplace transform pairs. Some o f
these will be developed later in the chapter and some in the homework exercises. We will refer to
this table in the next section when computing inverse transforms.
.V
KV4
T A B L E 12.1 Laplace Transform Pairs
Item Number
m t)
Ku{t) or K
KIs
m t)
KlP+1
\I {s+ a)
]/{s+ dp-
(OqUP- + OJ^)
cos{ci)Qt)u{t)
10
s!{P- + co^)
(Oo
s2
s in {(O Q t)u {t)
~>
(i + a) + coq
is + a)
11
{s + Cl)~ +
(Oq
2(0o^
12
(.v^ +toi5)
t C O s{0 )Q t)u {t)
13
is +0)o)
sin((W()/ +
14
.vsin((|)) + coo cos(<)))
(j>)tiit)
s~ +(0n
.vcos((t))-a)o sin(({))
0
1
+ (Oq
15
16
te
smiO)Qt)uit)
17
te
cosi(OQt)uit)
s +a
2(0
is + a)~-(OQ
((5 + rt)^ +C0n)^
2(0o
18
[(5+C/)^ +(OoJ
19
20
21
Cl cos(coor) +
2Î a ^ + B ~
Cj ~ C\Ci
sin(o)o/) n it)
(.? + ) +0)^
A + jS
cos (Oq/- tan -1
2Â~ + B- te~^' cos cogr - tan *
C|5' + C 2
' B^
A + jB
.A)j
[s + a + ;cOo)
A j B
A -jB
{s + a - ycoo)"
6. THE INVERSE LAPLACE TRANSFORM

For the Laplace transform tool to effectively anal)'ze circuits, one must be able to uniquely recon
struct time functions /(r) from their frequenc}'" domain partners F{s). Theoretically, this is attained
through the inverse l-aplace transform integral.
INVERSE LAPLACE TRANSFORM

Intuitively, if [/{()] = F{s), then J{t) = X '[F(s)]. Rigorously speaking, the inverse Laplace
transform integral is a complex line integral defined as
/ ( ,) = r V u ) , = ^ J ^ f ( . ) e V ,
over a particular path V in the complex plane. T he path F is typically taken to be the vertical
line Oj + jio where OJ ranges from -oo to +00 and Oj is any real number greater than Oq, the
abscissa o f absolute convergence.
This integral uniquely reconstructs the time structure o f F{s) to obia.\n J{t) in whichy(r) is zero
for ^ < 0. Conceptually, the process resembles the reverse action o f a prism, to produce white
light from its spectral components. An appreciation for the power o f this integral requires a
solid background in complex variables and would not aid our purpose, the analysis o f circuits.
In fact, the evaluation o f the integral is carried out using the famous residue theorem o f com
plex variables. Further discussion is beyond the scope of this text.
Just as the Laplace transform is linear, so, too, is the inverse Laplace transform, as its integral
structure suggests, i.e.,
Also, the unilateral transform pair
[fit), f(j)} is uftique, where by unique we mean the following: let F^{s) = L\f^{t)] and Fjis) =
Z[^(r)] coincide in any small open region o f the complex plane. Then F^{s) = Fjii) over their com
mon regions o f convergence, and/j(f) = f-y{t) for almost all r > 0, Almost all means except for a
small or thin set o f isolated points that are o f no engineering significance. Hence, there is a oneto-one correspondence between time functionsy(f) for whichy(f) = 0 for f < 0 and their one-sided
Laplace transforms. Linearity and this uniqueness make the Liplace transform technique a pro
ductive tool for circuit analysis.
Virtually all the transforms o f interest to us have a rational function structure; i.e., F{s) is the ratio
o f two polynomials. Rational functions may be decomposed into sums o f simple rational func
tions. These simple rational functions are called partial fractions and their sums are known as par
tial fraction expansions. Two o f the more common simple terms in partial fraction expansions
have the form K b and K!{s + a). Such simple rational functions correspond to the transforms o f
steps, exponentials, and the like. Table 12.1 lists these known inverse transforms. With the table,
direct evaluation o f the line integral in equation 12.23 becomes unnecessary. Our goal is to
describe techniques to compute the simple rational functions in a partial fraction o f F{s). Once
these are found, the transform dictionar)- in Table 12.1, in conjunction with some well-known
properties o f the Laplace transform, will allow us quickly to compute the time function y(r).
PARTIAL FRACTION EXPANSIONS: DISTINCT POLES

Our focus will center on proper^ rational functions, i.e.,
+ --- + ^^l-y + ao _
^ (^ )_
+--- + biS + bo
5" +
where m s and p^, ... ,
is - Pi )(.v - 7^2
- Pn )
are the zeros o f the denominator polynomial,
+ ... +
+ l?Q, and are called the finite poles o f F{s). For the most part, rational functions are sufficient for
the study o f basic circuits. There are three cases o f partial fraction expansions to consider:
(1)
the case o f distinct poles, i.e., p - p - for all i
j;
(2)
the case o f repeated poles, i.e., pj = pj for at least one i
(3)
the case o f complex poles. Although case (3) is a subcategory o f case (1) or (2) or both,
j\ and
its attributes warrant special recognition.

If F{s) is a proper rational function with distinct (equivalently, sitnple) poles />j, ... ,
F(s)=K +
(S-Pi)
{S-P2 )
++
(s-p)
then
(12.24)
where
K = lim F(s)
(12.25a)
5-400
The numbers Aj in equation 12.24 are called the residues o f the pole p- and can be computed
according to the formula
A =
[(s - Pi)f^(s)] = [(^ - Pi)Fis)\,^p,
(12.25b)
The rightmost equality o f equation 12.25b is valid only when the numerator factor {s - p ) has
been canceled with the factor [s - p ) in the denominator of F{s)\ othenvise, one will obtain zero
divided by zero which, in general, is undefined. As intimated earlier, this partial fraction expan
sion should enable a straightforward reconstruction o f/ r). Indeed, from Table 12.1, we immedi
ately conclude
f i t )= Kh{t) + Aê^^'uU) +
) + -!- Aeû(t)
(12.26)
EXA M PLE 12.12. Findy(r) when
jr(^+ a )
S o l u t io n
The solution proceeds by executing a partial fraction expansion (equation 12.24) on F(s) to pro
duce the Laplace transform o f two elementar>- signals, a step and an exponential. Specifically,
F{s) = ----------- = - - h
5(5 + a )
s +a
S6 '
where Ah is the Laplace transform o f a weighted step, Au{t), and B!{s + a) is that o f a weighted
exponential,
To find A, multiply both sides by^, cancel common numerator and denom
inator factors, and evaluate the result at j = 0, to produce A = Ma. Similarly, to find B, multiply
both sides by i
cancel common numerator and denominator factors, and evaluate the result
at j = -ay to obtain B = -Ma. Recall that, by iinearit)', X [aF{s)] = aL~^
Hence,
/ (r) = - H(0 - - ( - 'uir) = - (I - e - ' ) (r)
I ^Cl
Exercises. 1. Findy(f) when F{s) = --------- ^

AN SW ER: J{t) = 2u(t) 'j
2. Find/r) when F(.s) =

^ + 3 6 '+ 6
A N SW ER;
{.s + \)(s+ 2){s + 3)
Jit) =
3. Find a partial fraction expansion o f /r^y) =
Â +
s+a
5(i' + a)
AN SW ER: K = c . A = - , B =
(I
4. Find/r) for f{s) from Exercise 3.

AN SW ER: / ( ,) = c 5 ( 0 + ^ K / ) -
E X A M PLE 12.13. Suppose V',(5)= 1 0 ^ +^>y + 2
circuit o f Figure 12.19. Find
and vU) . Assume standard units.

2Q
V ..
'>
80>v^,
FIG URE 12.19 Series resistive circuit.

S O L U T IO N
Step 1. Detemihie
By voltage division,
= 0.8t'y(r), in which case

2 5 + 35 + 2
C'haptcr 12 Liplace Transform Analysis 1: Basics
56H
Step 2. Construct a partial fraction expansion ofV^^^^{s). Since the numerator and denominator are
both o f degree 2,
16.v" + 2 4 i-+ 1 6
.v(.v + 2)
= K + + ------.v + 2
(12.27)
The value o f K in equation 12.27 is determined by the behavior o f F(s) at infinity (equation
12.25a), i.e..
K - lim
^ 1 6 i + 24.v+ 16
i ÔO
i(A- + 2)
To I'lnd A in equation 12.27, we use equation 12.25b:

165 + 24.V + 16
(.v + 2)
Ks + /4 +
,v=()
Bs
.sT 2
.v=0
Similar!),
16.v- + 2 4 a + 16
Ii =
= -1 6
i= -2
Step 3. Find v^^^ît). Using Table 12.1,
v v , ,( o = r '
1
8
161
' 8 '
16
16 + -------------- = r ' [ Li 6 Ji + r
- r
.V .v+ 2.
s
..v + 2.
= 166(r) + 8u{t) - ]6e--'u{t) V
Exercises. 1. Repeat Example 12.13 for \A^^(.y)=

= lOuU) \'
AN SW ER:
*v(.v + )
2. Given the circuit o f Figure 12.20, find a partial fraction expansion o f

/. (.v) = ^
'
+ {a + b + c ).v + b
assuming standard units.
.v(.v+l)
A N SW l'R: /-,(/) =
+ \hn{t) + -\ce~u(t) A
ijt)|
1Q
40
F I G U R E 12.20 Parallel resistive circuit.
E X A M PLE 12.14. Compute the inverse transform oFthe function
F{s) =
So
-e
lu t io n
From Example 12.12,

1
r '
This result and the shift theorem yield
.v(.v+l)
By the linearit}' o f the inverse Laplace transform,
/rt = ( l - e - O u W - ( l - r < ' - ' ) ( ; - ! )
A sketch
appears in Figure 12.21..
FIG URL 12.21 Sketch ofy{/) = [ 1
^](r) - [1 -
!)
PARTIAL FRACTION EXPANSIONS: REPEATED POLES

Proper rational functions with repeated roots have a more intricate partial fraction expansion, and
calculation o f the residues often proves cumbersome. For example, suppose
F{s) =
fijs)
is-a)^d(s)
Chapter 12 Liplace Transform Analysis 1: Basics
S70
- a)^ specifies a repeated root o f order k, d{s) is the remaining fac
where the denominator factor
tor in the denominator o f the rational function F{s), and n{s) is the numerator o f F{s). The struc
ture o f a partial fraction expansion with repeated roots is
Ak
FCv) = - ^ +
where A^, ... ,
(s-af
(12.28)
are unknown constants associated with s - a, ... ,(s - a)^, respectively, and î^(s)
and <^(s) are whatever remains in the partial fraction expansion o f f(s). The formulas for comput
ing the y4y o f equation 12.28 are
A k=(s-arF(s)
n(s)
-*i-a
(12.29a)
f/(5)
n{s)
[d(s)
Â-l = y ( ( . v - / 'F ( ^ ) )
(12.29b)
and, in general,
1
/!
ds '
njs)''
(s-afF(s)
(12.29c)
O f these expressions, only the first looks like the case with distinct roots; the others require deriv
atives o f (.f - a)^ F{s). Computation o f high-order derivatives borders on the tedious and is prone
to error. The above formulas, equation 12.29c in particular, are included for completeness.
Computer implementation circumvents these difficulties. An example that illustrates the use o f
the preceding formulas, as well as a usefiil trick, comes next.
EXA M PLE 12.15. The goal here is to illustrate the computation
F(s) =
when
s +2
.v (.v + l)
.V
.s-
.v + 1
(.v+l)'^
(12.30)
The two easiest constants to find are A2 and Bj, as their calculation requires no differentiation.
From equation 12.29a,
A, =
s-Fis)
-I
5-1-2
-5=0
=2
i=0
and
B 2 = ( s + \r F (s )
j'-i-2
.v=-l
_ s~
.9 = - !
Finding A^ and
is more difficult, since formula 12.29b requires some differentiation.
According to equation 12.29b,
as
571
To implement this formula multiply both sides o f equation 12.30 by

resulting expression with respect to s, and evaluate at i = 0:
d
ds
5+2
_ d
(5+l)-_
y4|5 +
ds
5=0
take the derivative o f the
^ B|iA-y H----------- h
(5+1)
S+\
i= 0
Observe that, on the right-hand side, it is not necessary to differentiate the terms that contain A 2 ,
and ^ 2 since these terms disappear at / = 0, as the formula for
d
ds
'
5+2
(,v+l)2
.v=0
(5+1)"
^+2
'
requires. Consequently,
= -3
(5+l)-\
i=()
Similarly,
Bx=ds
0
( 5 + l ) ^ F ( 5 )
_ d
s= - 1
ds .
+2
5
=
. A--1
^s + 2'
.5 * '
S'
= 3
. ,v=-l
Aj, and B2 were known, a simple trick allows a more direct computation o f
Bji merely evaluate equation 12.29 at j = 1 (in fact, any value o f s, excluding the poles, will do),
Note that since

to obtain
0.75 = - 3 + 2 + 0.25 + 0 .5 5 j
As expected, solving yields Bj = 3. Hence
-3
2
3
= + ^ + ------ +
i (.v + l)5 s + \ (5 + 1 )2
F(s) =
.v+ 2
The above result can also be found with the MATLAB command residue as follows. Let F{s) =
7i{s)ld{s). It follows that n{s) = s + 1 and d{s) =
+ r . In MATLAB,
num = [1 2];
>>den = [ 1 2 1 0 0 ] ;
[r, p, k] = residue(num, den)
The answers from MATLAB are:
r = -3 2 3 1 (the residues associated with the poles)
p = 0 0 -1 -1
and constant
k = 0.
Exercises. \. Find the partial fraction expansion o f F{s) =
2 5 ^ + 2 r S 3 ^ + 35 4-2
.v^(.9+n^
AN SW ERS (residues in random order): 2, 2. 2, - 1 . -1

2. The partial fraction expansion o f a rational function is given by
F{s) = -
3s^ + \0s + 9
+ 45* + 55 + 2
A
5+ 2
+ ------ +
5+1
Compute A, B, and C
ANSWHRS: In random order: 2. 1. 2
3. Use MATLAB to find a partial l^raction expansion o f F{s) =
(Iis)
where n{s) = G{s + 2)^(^ - 2)
and d{s) = s{s + 1)(^ + 4). Hint: Use num = 6*poly([-2 -2 -2 2 2]) and den = poly([0 -1 -1 -4 -4]).
ANSWHR: [r,p,kj = residue(num,den) \'iclds
r = - 4 4 4 8 - 1 6 - 6 12
p = _4 .4 _| _ ] 0
,
n .r
-^8
1 his results m rlie r r h : r(.s) =
-1 6
-6
v+l
(v + l)~
12
-{------------ - h----- H--------------- -----------h 6
_______________________________ '>+ 4
(.v + 4 )
________________________
The derivative formulas o f equations 12.29 are often difficult to evaluate for complicated ration
al functions, such as
s
55^ + 955-^ + 692^^ + 2369.V- + 3715.9 + 2076
F{s) = ---------------------------------------------------- r--------------
(.v+l)(:? + 2)(.s- + 3)(.v+5r

A
s+\
s +2
_ C ^ D\
i + 3
.y + 5
For these functions, it is very efficient to find A, B, C, and
1)2
D3
(.v + 5 ) -
(.^+ 5 )'*'
directly. Then one evaluates F{s) at
two values o f j, e.g., j = 0 and s = 1, to obtain two equations in the unknowns D j and Dj.
Typically, solving the resulting two equations simultaneously is much easier than evaluating Z),
and D j directly by equations 12.29. Alternatively, one can use a software program such as MAT
LAB to compute the answers. In particular, in MATLAB:
n =[5 95 692 2369 3715 2076]
d = [l 21 176 746 1665 1825 750]
[r,p,k]=residue(n,d)
r=
-l.OOOOe+00
-l.OOOOe+00
-l.OOOOe+00
3.0000e+00
2 .0000 e+00
l.OOOOe+00
P=
-5.0000e+00
- 5 .0000 e +00
- 5 .0000 e +00
- 3 .0000 e +00
- 2 . 0000 e +00
-l.OOOOe+00
k=
573
PARTIAL FRACTION EXPANSIONS: DISTINCT COMPLEX POLES

Distinct complex roots present challenges different from those for the repeated root case. Since the
roots are distinct but not real, the methods o f equations 12.25 and 12.29 apply. Unfortunately, the
resulting partial fraction expansion has complex residues, and the resulting inverse transform has
complex exponentials multiplied by complex constants. Such imaginar}^ time functions lack mean
ing in the real world unless their imaginar}- parts cancel to yield real-time functions. When they do,
our goal is to find a direct route for computing the associated real-time signals. To do this, consider
a rational function having a pair o f distinct complex poles as in the following equation:
F(s) =
n{s)
n{s)
|(.v + a ) - + ( 0 -](/(. 5)
(s + a + ju>)(s + a - j w ) d ( s )
(12.31)
Since the poles - a - j c o and - a + jco are distinct, the partial fraction expansion o f equation 12.24
is valid. Since the poles are complex conjugates o f each other, the residues o f each pole are com
plex conjugates. Therefore, it is possible to write the partial fraction expansion of f{s) as
r(s)-
.v + + yto
5 + a -y c o
(12.32)
d{s)
and d{s). As per equation 12.25b, the first residue in equation
for appropriate polynomials

12.32 is
(12.33)
jco
With A and B known, executing a little algebra on equation 12.32 to eliminate complex numbers
results in an expression more amenable to inversion, i.e..
C\S + C->
nîs)
/?|(.v)
F{s) = ------ \2, ~ 2 + ^ 7 T = ^0(') +
(I{s)
d(s)
{s + fl) + to
(12.34)
C, =2/1
(12.35a)
C-, = 2aA + 2 ojB
(12.35b)
where
and
with A and B specified in equation 12.33. W ith Cj and Cj given by equations 12.35, it is straight
forward to show that
F'ois) =
C \S + C
{s + a)~ +oj^
=C
( C2 -C â \
1,
x2
(.9 + a) + 0)
to
to
{s + )"
From Table 12.1, item 19, or a combination o f items 10 and 11,
M O = e'
C| cos(tor) +
(C2-Cici)
sin (to/) Hit)
[ to
)
+03"
(12.36)
574
Exercise. Suppose F { s ) =
. Compute/(r).
a- 2 + 4
AN SW ER: C, cos(2r)//(f) + 0 . 5 ^ sin(2f)(r)
The following example illustrates the algebra for computing C, and Cj without using complex
arithmetic.
EX A M PLE 12.16. Find/^) when
3 .r + 5 + 3
D
A + jB A - j B
D
C^s + C^
F{s) = ------------^------- = -------- + ------ =;- + ------^ = ------ + - ^ ------- (5 + 1 )(5 + 4 )
^+1
s + j2
s-jl
i +1
s +A
, ^
(12.37)
Step 1. Compute the coefficients D, C ,, and C2 in the partial fraction expansion o f equation 12.37.
First we find D by the usual techniques:
3 .r
+ 5
+ 3
.v^ + 4
= 1
s=-\
Given that D = 1, to find C , we evaluate F{s) at j = 0, in which case 0.75 = 1 + O.2 5 C2, or Cj =
- 1 . With D = 1 and C2 = - 1 , we evaluate F{s) at j = 1 to obtain 0.7 = 0.5 + 0.2(C j - 1) or, equiv
alently, Cj = 2. Thus,
V+ 1
s^ + 4
+ 4
Step 2. Compute j{t). Using Table 12.1, items 8 and 9, to compute the inverse transform yields
J{t) = [e~ + 2 cos(2r) - 0.5 sin(2f) ]u{t)

Alternative Step 1. Compute A and B in equation 12.37 by hand or with MATLAB. In MATLAB,
num = [3,1 3];
den = conv([l 1],[1 0 4])
den = [1 1 4 4]
[r, p, k] = residue(num.den)
r=
1.0000 + 0.25001
1.0000 - 0.2500i
1.0000 + O.OOOOi
P=
-0.0000 + 2.0000i
-0.0000 - 2.0000i
1.0000
k =0
This implies that
.y+1
^+
. A -JB
s + j2
s-jl
, l- y '0 .2 5 , l + yO.25
5+ 1
s+ jl
s-jl
(12.39)
Alternative Step 2. One must exercise caution here and note the difference between the MATLAB output and the form o f the partial fraction expansion. From equation 12.39, w = +2, A = 1,
and B = - 0 .2 5 . Again using MATLAB to obtain the form needed in item 20 o f Table 12.1,
K = 2*sqrt(A^2 + B^2)
K = 2.0616
theta = atan2(B,A)* 180/pi
theta = -1 4 .0 3 6 2
Thus
Example 12.16 illustrates not only the computation o f an inverse transform having complex poles,
but also the computation o f Cj and C , without resorting to complex arithmetic, as was needed in
equation 12.32. The trick again was to evaluate F{s) at two distinct ^-values different from the
poles o f F{s). This yields two equations that can be solved for the unknowns Cj and C ,.
5 ^ 8
Exercises. 1. Find Kt) when F{s) = ------- z----------.

s(s-+4)
AN SW ER://) = [1 + 4 cos(2r) - 4 sin(2r)]//(/)
,
5s" - 2 ^ + 5
2. Find/(r) when F ( s ) - ^
.
. v ( r + 2 5 + 5)
A N SW ER://) = u(t) + 4 r [cos(2/l - sin(2/)l/<(/)
7. MORE TRANSFORM PROPERTIES AND EXAMPLES

Another handy propert)' o f the L'lplace transform is the frequenc)' shift property, which permits
one to readily compute the transform o f functions multiplied by an exponential. With knowledge
o f the transforms o f u{t) , sin(o)/), and other functions, computation o f e~û{t) and f"^sin(to/)(^)
becomes quite easy.
Frequency shift property: Let F{s) =
Then
L[ e~^'p)] = F{s * a)
(1 2 .4 0 )
Chapter 12 Laplacc'Iransform Analysis I: Basics
ThisS property can be verified by a direct calculation,

= F (s + c)
Xlc- 7 ( ' ) l =
where we have viewed the sum s + a m tlie integral as a new variable p, which leads to F{p) with
p replaced hy s + a.
EXA M PLE 12.17. Let//) = sin(wr)//(r). D efine^/) = e~" p ) = e~* s\n{iot)u{t). Suppose it is
known that
Compute G{s).
So
lu t io n
By the frequency shift property, G{s) = F{s + a), or

CO
)| = L\e-"\m \ = F (s + ) =
G( j ) = |
CO
5 + (0
Exercise. Lcijit) = cos(cor)u(i) for which r ( s ) = ___ - ___ D efin e^r) = e

Compute G{s).
J{t) = e
cos{LOt)u{t).
+C0
.V+ a
ANSW I-R:
i.s +
a )-
+ (0 "
Another property o f particularly widespread applicability is the time differentiation formula. Its
utility resides not onl)' in obtaining shortcuts to transforms o f signals, but also in the solution o f
differential equations. Differential equations provide a ubiquitous setting for modeling a large
variety o f physical systems mechanical, electrical, chemical, etc. In terms o f signal computation,
recall that the velocity of a particle is the derivative o f its position as a function o f time. The accel
eration is the derivative o f the velocity. After computing the Laplace transform o f the position as
a function o f time, one finds that a differentiation formula allows direct computation o f the trans
forms of the velocity and acceleration. Also, as discussed at the very beginning o f this chapter, cir
cuits have differential equation models. For example, weighted sums o f derivatives o f the response
of- the circuit are equated to weighted sums o f derivatives o f the input signal. Therefore, a differ
entiation formula is an essential ingredient in the analysis o f circuits.
First-order tim e difiFerentiation formula: Let L\j{t)] = F{s). Then
L jfU )
(It
= sF{s)-f{Q-)
(12.41)
The difFerenriarion property is validarcd using integration by parrs as follows:
4/(0V' ' d t = f U ) e
y / (/ )
(It
JO'
The following examples explore some clever uses o f the first-order time difTerentiation formula
E XA M PLE 12.18.^^
Recall that 6f/) = u ( t ) . Using the sifting propert)', a direct calculation yields [(5(/)] = 1. Is
this consistent with the differentiation propert}^? Interpreting the delta function as the derivative
o f the step function and applying the differentiation formula yields
I
d
6 { l ) = - H i t ) = s u{r) - u { 0 ) = s
dt
s)
(12.42)
which demonstrates the expected consistency.
Exercises. I. The Liplace transform o f a signal/r) is F{s) =

f o r m o f e - 2 '4 / W
. What is the Laplace trans
dt'
2s + 4
AN SW I-R: - ^-----
(.Y+ 2 r + 4
2. [sm{wt)6it)] = ?
AN SW ER: u
EXA M PLE 12 .1 9 . Suppose^r) = sin(wr)w(/) and we know (for example, from Table 12.1) that
Compute [cos(wr)//(r)] using the time differentiation formula.
F { s ) = sin((0/);/(/)
.V" + (0
So
lu t io n
Since cos(coO/^0 = ------- sin{a)/)//(/)and sin(wr)(/)]j ^ o = 0 , the differentiation property imme

diately implies that ^ ^^
cos(coO//(0
(0
(0
+C0
.S - + W
Exercises. 1. Express//) = s\n{o)t)u{t) in terms o f the derivative o f ^ r) = cos{(Ot)u{t). Note the

presence o f the delta function.
AN SW ER: - '
(1)
^
di
(ij
= sin (w ;)//(r '
2. Now suppose it is known that cos(co/)/KM
-. Use the result of Exercise 1 and the

+ 0)'
differentiation propert)' to compute the Liplace transform of/r) = sin(oj/)(/) noting that^Ô ) = 0 .
-t
S~
ANSW ER; sin((0/);/(/)
1
(0
(!)
(I)
(0
(0
V +(0'
s' - (O'
Chapter 12 Laplacc Transform AnaK'sis I: Basics
578
EXA M PLE 12.20. Lety(r) and its derivative iiave the shapes shown in Figure 12.22. Th e goal o f
this example is to explore the relationship between the Laplace transforms o f/ r) and f'{t) in light
o f the differentiation property.
FIGURE 12.22 A pulse and its derivative for Example 12.20. Observe how the derivative o f the
pulse leads to a pair o f delta functions.
Using linearity and the shift theorem on j{t) yields
X|/(01 = | ( 0 - - 1)] = [ (f)]- X | h(/ -1)1 = - (I Applying the linearity o f the Laplace transform x.of\t) yields
l / (r)] = i : [ 5 ( 0 - 5 ( f - I ) ] = l - ^ - ^
From the differentiation formula, it must follow that L\f{t)] = sL\J{t)]. Thus,
|/'(/)l = sL U (t)] = j i d - - * ) = I demonstrating consistency.
As might be expected, the formula for the first derivative is a special case o f the more general dif
ferentiation rule:
L
cit
= s"F(s) -
(12.43)
This rule proves useKil in the solution o f general th-order difTerential equations. O f particular
use is the second-order formula:
(1 2 .4 4 )
579
The inverse o f differentiation is int^ration. The following property proves useful for quantities
related by integrals.
Integradon p r o p e r ^ Let F{s) = -[/(/)]. Then for ? > 0,
c Jo-/(9 )d? =
(12.45a)
and
F (s)
(12.45b)
As with many o f the justifications o f the properties, integration by parts plays a key role. By direct
computation (using equation 12.16),
To use integration by parts, let
fiq ) d q and dv = e
u=
Then
-too
rr
JO'
For the appropriate region of convergence, the first term to the right o f the equal sign reduces to
fO'
et
f(q ) d q
f( q ) d q
JO*
Since the second term to the right of the equal sign is F{s)/s, as per equation 12.45a, the proper
ty is verified.
S80
E XA M PLE 12 .2 1 . Find the Laplace transform o f the signal/f) sketched in Figure 12.23a using
the integration property.
f(t)
FIGUIIE 12.23 (a) A triangular signaly(/) for Example 12.21. (b) The derivative
S o l u t io n
Observe that the triangular waveform y{r) o f Figure 12.23a is the integral o f the square wave^^).
Since ^t) is easily represented in terms o f steps and shifted steps as
^t) = u{t) - 2u{t - 7) + u{t - 2 7)

its Laplace transform follows from an application o f linearity and the time shift property:
x u (o i= The integration property implies that
g{q)dq
.S -
EXA M PLE 12.22. This example explores the voltage-current {v-i) relationship o f a capacitor in
the frequency domain by way o f the integration property. Recall the integral form o f the voltagecurrent dynamics o f a capacitor:
1 r'
Taking the Laplace transform o f both sides and applying the integration property produces
[vcit)] =
C->
1 fO-
ic(T)dx = 7T^c(-^) +
Cs
Cs
But this expression depends on the initial condition
), because
S8]
Chapter 12 * Laplacc Transform Analysis 1: Basics
Therefore,
Cs
(12.46)
Equation 12.46 says that the voltage V(^s) is the sum of rvvo terms: a term dependent on the fre
quency domain current I^^s) and a term that looks like a step voltage source and depends on the
constant initial condition V(^0~). The quantity Z^^s) = MCs looks like a generalized resistance
generalized because it depends on the frequency variable s and a resistance because it satisfies
an Ohms law-like relationship, V^^s) = Z^^s) I({s). These analogies prompt a series-circuit inter
pretation o f equation 12.46 as depicted in Figure 12.24. An application o f this equivalent frequenc)' domain circuit to general network analysis appears in the next chapter.
\,{S)
o >+
ic(t)
O >
4-
[ ]
c ----- > VJs)
V,(t)
Cs
O 'f
O------FIG U RE 12.24 Equivalent circuit interpretarion of a capacitor in the frequency domain. This equiv
alent is arrived at by applying the integration propert)' of the Laplacc transform to the capacitor volt
age, seen as the integral of the capacitor current.
A second example interpreting the v-i characteristics o f the capacitor in the frequenc)' domain
ensues from the differentiation rule. Instead o f winding up with a series circuit, one obtains a par
allel circuit. The interpretation is thus said to be dual to the one just described.
E X A M PLE 12.23. This example has two goals: (i) Verify that equation 12.46 is consistent with
the differentiation formula interpretation o f the capacitor; (ii) Build a dual frequenc)' domain
interpretation o f the v-i characteristic o f a capacitor analogous to that o f Example 12.22.
As a first step, recall equation 12.46:
Vc(-
Cs
which, after some algebra, becomes
I(is) = CsVfis) - CvfiO-)
(12.47)
Notice that equation 12.47 is consistent with the application o f the derivative formula to i(^t) =
C[civ(Jdt]. This consistency offers some reassurance in the accuracy o f our development. The
interpretation o f equation 12.47, however, is quite different from that o f equation 12.46. In the
latter equation, the current /^j) equals the sum o f two currents, CsV^s) and -C t/^ 0). This sug
582
gests a nodal interpretation, resulting in an equivalent circuit having two parallel branches. One
branch contains a capacitor with voltage V^s). The other, parallel branch contains a current
source with amperage Cv(\Qr). The current through the capacitive branch is
where
G now acts like a generalized conductance because it multiplies a voltage, similar to Ohms law.
Q is generalized because it depends on s. Figure 12.25 presents the equivalent circuit o f the
capacitor in the frequency domain and is dual to the circuit o f Figure 12.24. Chapter 13 covers
in detail the role o f these equivalent circuits in analysis.
Ic(s)
FIGURE 12.25 Equivalent circuit to a capacitor in the frequency domain

using the differentiation formula.
The last elementar)' property o f the Laplace transform that we consider in this chapter is the timescaling property, also called the frequency-scaling property. Its importance is fundamental to net
work synthesis. Here, numerical problems, such as roundoff, prevent engineers from directly
designing a circuit to meet a given set o f specifications. Instead, the design engineer will normal
ize the specifications through a frequency-scaling technique. Once the normalized circuit is
designed, frequency-scaling techniques arc reapplied in an inverse fashion to obtain a circuit meet
ing the original specifications.
Time-/Frequency-scaling property: Let ^ > 0 and L\J{t)] = F{s). Then
L[J\at)\ = - F
a \ci)
(12.48)
or, equivalently, F{sln) = aL[f{nt)].

Since the proof o f this property is straightforward, it is left as an exercise at the end o f the chap
ter.
E XA M PLE 12.2 4 . Figures 12.26a and b show impulse trains that model sampling in signal-pro
cessing applications. The impulse train o f Figure 12.26b is the time-scaled counterpart to that o f
Figure 12.26a.
^83
f(2t)
f(t)
A
2 --
A
2 --
i<
1
1 --
- -
> t
> t
1
FIG URE 12.26 (a) Unit impulse train, (b) Time-scaled imit impulse train. Unit impulse trains such
as these model sampling in signal-processing applications.
The time-scaled impulse train in Figure 12.26b increases the frequency at which the impulses
occur (twice as often as in the original signal). This is reflected in the Liplace transforms o f the
two signals:
'Z&U-k) = I
k=0
A-=0
\-e'
(12.49)
By the time-scaling property,

[/ (2/ )l = 0.5
l - e -0 .5 5
(12.50)
Notice that what occurs at, say, Sq in equation 12.49 now occurs at 2s^^ in equation 12.50. Hence,
time scaling by numbers greater than 1 concentrates more o f the frequenc)^ contcnt o f the signal
in the higher frequency bands.
Exercise. Verify, by direct calculation, that L\J{2t)\ is given by the right side o f equation 12.50.
Several more properties o f the Laplace transform are germane to our purpose. However, these
properties have a systems flavor and are postponed until Chapter 13. We close this section by pre
senting Table 12.2, which lists the Laplace transform properties and the associated transform pairs.
lA Bl.E 12.2 Liplace Transform Iropertics

Property
Transform Pair
Linearity
L \ j{t-
Time shift
\t/lt)n(t)]=- F'{s)
as
Multiplication by t
n cr r j s )
Multiplication by t
ds"
= Rs + //)
Frequency shift
Tim e differentiation
jfO )
(it
d - f{ t )
Second-order differentiation
= sF(s)-J{0-)
= r F c v ) - 5 / ( ( r ) - / '\ ( D
dr
wth-order differentiation
T> 0
7)1 =
d''fU)
dt'
(i)X
fUl)dq
m dq
Fis)
Time integration
(ii)
Time/Frequency scaling
.-fU l)d q
Fis)
[f{at)\ = MaFista)
S8S
8. SOLUTION OF INTECRO-DIFFERENTIAL EQUATIONS USING

THE LAPLACE TRANSFORM
Differential equarions provide a cross-disciplinary mathematical modeling framework. Although
difTerentia! equation models may represent only the dominant behavioral facets of a circuit or
physical process, their widespread utility and importance to circuits and control systems warrant
special discussion. To begin, recall the time differentiation formulas o f equations 12.41 and 12.43
and the integration formulas o f equations 12.45a and 12.45b. Also, recall that a differential equa
tion relates a sum o f derivatives o f an output signal to a sum o f derivatives o f an input signal. For
example, if the input and output signals are voltages, then the relation
d'\-
+ a..V.
dl
for constants
dt'
+b
and bj might model the behavior o f a linear circuit. We may use the following steps
to solve this differential equation for
using the Laplace transform procedure:
1.
Take the Laplace transform o f both sides o f the equation, using the appropriate deriva
2.
Algebraically solve the resulting expression for
3.
Compute a partial fraction expansion o f the expression for
4.
Inverse-transform the partial fraction expansion to obtain the time function
tive formulas, equations 12.41 and 12.43.
If the equation is an integro-differential equation, i.e., a mixture o f both derivatives and integrals
of the input and output signals, then we simply apply the same algorithm, except we use the inte
gral formula where appropriate. Some examples ser\-e to illuminate the procedure.
EXA M PLE 12.25. Consider the pulse current excitation o f Figure 12.27a) to the RC circuit o f
Figure 12.27b. The goals o f this example are (i) to use and illustrate Laplace transform techniques
to solve a difTerential equation derived from a simple RC circuit and (11) to find the response volt
r > 0, when V(^0~) = 1 V.
age
F IG U R E 12.27 Excitation currcnt (a) fora simple /?Ccircuit (b) for Example 12.25.
S o l u t io n
Step 1. Find L[i{t)]. Since i{t) = 0.5//(r) - 0.5//(^- 1),
X [/ (0 ] = 0 .5
\ -e
Step 2. Find the circuits dijfereutial equation model that links the excitation current i{() to the response
voltage, V(it). Since ijît) = 0.5v(\t) and
= ^.b d v jd t, summing the currents into the top node
of the circuit yields
d\c {t)
dt
After multiplying through by 2, the desired differential equation circuit model is
dv({f) + v c ( f ) = 2 i(t)
dt
Step 3. Take the Laplace transform o f both sides, apply the differentiation rule to the left side, and solve
for V(i{s). Applying the Laplace transform to both sides yields
sVcis) - v^Q-) + V^s) = ll{s)

Solving for V^s) produces
2
v r(0 )
.v+1
i + l
Vcis) = ----- :/(.v) + - ^ ^
\-e ~ ^
= -----------+
^(i-i-1) .v + 1
Some straightforward calculations show that

1
1 _
5(.v+ l)
1
(.v+1)
Thus, with the aid o f the shift propert)' and the transform pairs o f Table 12.1, we obtain
vcU) = r'[V c(s)\ = r '
.v(i + l)
s+\
= / / ( 0 - ( l- c '" ^ ' '^)/K ^-l).

Figure 12.28 presents the graph o f this response. Because o f the initial condition and the magni
tude o f the pulse input, the capacitor voltage is constant for 0 < r < 1 second. At r = 1 second, the
pulse magnitude drops to zero, making the circuit equivalent to a source-free RC circuit in which
the capacitor voltage decays to zero as shown in the figure.
V,(t)
F I G U R E 12.28 T h e response voltage
v^^t)
for Exam ple 12.25.
S8
Chapter 12 Iâplace Transform Analysis 1: Basics
EXA M PLE 12.26. The goal o f this example is to compute the response, denoted here by the
given the scries RLC circuit o f
input current /,(^), to the input voltage cxcitation
Figure 12.29. Suppose the initial conditions are /^(O-) = 1 A and V(^0~) = -2 V.
TYYY
40
1H
FIGURE 12.29 Series RLC circuit for Example 12.26. Here the current //(^) =
So
lu t io n
Step 1. Compute the Laplace transform o f the input. From the tables or by inspection, X[6(r)] = 1.
Step 2. Compute the integro-differential equation o f the circuit o f Figure 12.29. The first task is to
sum the voltages around the loop to obtain
Substituting for each o f the element voltages using the mesh current, ij{t), yields the desired integro-difFerential equation,
diin
(12.51)
Step 3. Take the Laplace transform o f both sides, substitutefor R, L, C,
solve for
Vf^Qr), and
and
W ith the aid o f the differentiation and integration formulas, taking the Laplace
transform o f both sides o f equation 12.51 produces
/?/,(.V) +Uli (,v) - L/i(0-) +^Cs
(s) +
= V;(s)
This has the form

i - + S.s + 7li
L ------ 1----- ^ h U )
= K (S ) + i , / t ( 0 - ) -
V>-(0 )
Plugging in the required quantities and solving for
produces
I
') ~
5 + 4 5 + 4
5 +2
5+ 2
( 5 + 2 )-
Step 4 . F in d iinit). Taking the inverse Laplace transform yields the desired result:
i d t ) = (2 - 2,)e-^-u(t) A
588
A plot o f this response appears in Figure 12.30.
Exercises. 1. An integro-differential equation for an LC circuit is given by
cli,C
dl
C oo
= 0, and V(^0~) = - 1 0 V. Compute
with C = 1 F,
L.(s) r . ( ( ) )
ANS>X^R: .v/,-.(.v) + -^----- + -^-------- = 0
^
.S
= > / ..(/ )= 10sui(/)//{/) A
.V
2. If two signals x{^) and^(^) are related by the equations
dxU)
dt
+ 2v(/) = 45(/) and 2.v(/)-
y{z)dz = 2ii(t)
where x(0 ) = 2 , u{t) is the unit step function, and 8{t) is the Dirac delta function, then findATi).
AN SW ER: X{s) = V
EXA M PLE 12.27. The final example o f this chapter looks at the leaky integrator circuit o f
Figure 12.31, which contains an ideal operational amplifier (op amp).
resistance o f the capacitor. Given C and
represents the leakage
/?, is chosen to achieve an overall gain constant, in
this case, 1. The objective is to compute the response
assuming that t'(j(0~) = 0, and com
pare it with that o f a pure integrator having a gain constant o f 1.
SH)
Chapter 12 Liplacc Transform Analysis 1: Basics
R, = IMegO
V (t) = 5 u (t )
''ou.W
FIGURE 12.31 Leaky integrator op amp circuit.

So
lu t io n
First, note that since the op amp is ideal, -V(^t) =

equation that relates
to z^^-and solve for
The goal, then, is to write a difFerential
using the Laplace transform method.
Step 1. Determine the dijferential equation. Since the op amp is ideal, it follows that ijr= -i^. From
O hm s law, i^ = vJR^^. On the other hand,
/?2
dt
This leads to the difFerential equation model o f the op amp circuit,
dt
where, as indicated before,
I<2C
R^C
= -v^^t). Note that if /?,C= 1 and R2 is infinite, then the cir
cuit works as a simple integrator. The circuit is called a leaky integrator because /?-,C is large but
finite. Since /?, C = 1, one expects the gain constant to be 1 as well.
Step 2. Substitute values, take the Laplace transform o f both sides, a7jd solve for Vg,,f{s). Taking the
Laplace transform o f both sides, one obtains
'w ( 0 " ) +
Since
=- -
= 0, it follows that
-5 0
5( 5 + 0 . 1 )
Step 3. Invert
to obtain
.V
50
.9 + 0 .1
Solving for v^^^^{t) produces
^92
'Problems
5. Find the Laplace transform o f each o f the

following time functions.
f\{t) = Ke-û {t- 7), r > 0
(a)
BASIC SIGNALS, SIGNAL

REPRESENTATION, AND
LAPLACE TRANSFORMS
(b )
(d)
(e)
r > 0
f^{t) = Kte-û {t- 7), T > 0
(c)
r>0
f^{t) =
T>0

following signals assuming T- > 0.
6. Compute the Laplace transform for each of

the following signals.
(a)
(h)
/ 2 W = / / - 7 ) 6 U - 7-,)
(c)
h (D = e-^' cos(0.5rt/ + f )5(2/ -
(d)
/^W = A-,6W + 7 - )
(c)
(f)
(a)
/\{t) = Kt^[uU)-N{t- T)l T>()
(b )
f,U)
sin(2yTr - 2jr)u{t - 1)
(c )
K^6(t-T)
/sW = c o s h (2 / )a (t-2 )^
= s in ( 2 . T / ) ( r ) -
(d )
f^{t) = r [ s i n ( 2 . T / ) w ( r ) s i n ( 2 ; r / - 2jt)u{t - 1)]

f^{t) = 2 s i n ( 4 ; r r ) / / ( r ) ( 2
r)
sinh(/ - 2 r , ) 6 ( f - r , )
(g)
7. Represent each o f the following signals using
fj{t) = sm{2m - Ji)d{2t - 4)
(sums of) steps, ramps, shifts o f basic signals,

etc. Then find the Laplace transform.
following signals. Use Tables 12.1 and 12.2 as
f,{t)
needed.
(a)
f^{t) = lu{t) + u{t - 1) + u{t - 2)
2K -
-Au{t-A)
(b)
f^{t) = 2 r { t ) - l r { t -
l ) - r ( r - 3 )
At - 5)
->
(c)
(d)
t(s)
j\{t) = cosiS)5m)u{t) +
c o s (0 .5 > -r(/ -2 )) a -2 )
(a)
(e)
f,(t)
3. Sketch the indicated waveforms and find the
2K -
Laplace transform. Use Tables 12.1 and 12.2 as

needed.
(a)
(b)
(c)
(d)
(e)
f^{t)
fJ,t)
f^{t)
f^[t)
f^{t)
= ti{t) - l i t - \ ) .
= u{t)-r{t-\)-^r{t-2)
= 2u{t) u{2 - t).
= Ar{t) u{\ - t)
= 2rit)u{\ - t) + u { t - \)r{2 - t)
>
2
(b)
A
K

following time functions.
(a)
f^{t) = Ktu{t-\)
(b)
/,(/) = AT/- ])(/)
(c)
(d)
f^{t) = Ktr\t-\)
f^{t) ^ K{t - \)r{t)
ir
(c)
t(s)
S93
f,(t)
2K
K
> t
>
2T
T
(d)
i k f^ft)
-2
f,(t)
K
> t
> t
-T.
-K
(e)
f(t)
(d)
9. Represent each of the following signals using
(sums oO steps, ramps, shifts o f basic signals,
^-------1-------1------ r
1
>
f,{t)
5
4K -
(f)
8. Represent each of the following signals using

(sums of) steps, ramps, shifts o f basic signals,
2K
-I---------- '----------1--------- 1 t(s)
1 2
3
4
(a)
(a)
(b)
5 94
r>
- t
r^
(0
11. For the circuit o f Figure P I 2 .11, suppose
/?, = 6 0 0 Q, Rj = 1000 Q, and
= 1500 Q.
Use the Laplace transform tables to compute
\ltu{t) + 3e~^û{t) + 18^ ^sin(2jK)W A.

> t
| U t)
i,(t)
(!)
Figure P I2,11
12. For the circuit of Figure P I 2 .12b, R^ = 600

Q, /?2 = 1000 Q, and R^ = 2 4 0 0 Q. Use the
Laplace transform tables to compute Vg^t^s) =
for the input given in Figure P I2 .12a.
10. Represent each of the following signals

using (sums of) steps, ramps, shifts of basic sig
nals, etc. Then find the Laplace transform.
vJt)
f,(t)
2
>
1
(a)
2
> t
o
Figure P I2.12
(a)
13. For the circuit o f Figure P i 2 .13, suppose

/?! = 6 0 0 Q, /?2 = 2 0 0 0 Q, and R^ = 3 0 0 0 Q.
Find the Laplace transform of the voltage
when yjj(f) = 2Ae~^û{i) V and v^2 ^t) =
30^-^Mf) V.
r>
>
t
o
r^
(e)
Using the formulas o f part (c) and the

frequency shift property, compute
(i) L [Ke-^^ cos(wr)]
(ii) L [K e -^^s\n{o)t)]
17. Find a simple expression for each o f the

waveforms shown in Figure P i 2.17 in terms of
sines, cosines, shifted sines, and shifted cosines.
Then find the associated Laplace transforms.
Half-cycle of sin (nt)
LAPLACE TRANSFORMS VIA

TABLE 12.1 AND PROPERTIES
VIA TABLE 12.2
14. Find the Laplace transform o f J{t) =
Ke-"u{t)u{T- t), T> 0, as follows.

(a)
Express Ku{t)u( T - t) zs a difference o f
step functions.
(b)
(c)
Find [Kn{t)u{T- t)].

Apply the frequency shift property on
Quarter cycle of 2cos(0.25m)
your answer to part (b) to compute
W )l
15. Prove the time-/frequenc)'-scaling property
by direct calculation o f the Laplace transform
integral.
16.(a)
Using the famous Euler formula,^'^ =

cos((Ot) + j sin(wf), find an expression
(b)
for cos(a>/) and sin(wr) in terms o f the
Figure P I2.17
complex exponentials
(b)
Determine the Laplace transform o f
(c)
Using the formulas developed in parts
18.(a)
Represent
sin(7cO
0<t< 2
otherwise
(a) and (b), show that
as the difference o f a sine and a shifted
(i) [co s((o r)]=
sine. Find G{s).

(ii) Xfsin(co/)] =
(d)
(b)
>
~>
S +(D
Using the formulas o f part (c) and the

multiplication-by-/ propert)', compute
(i)[/irrcos(w r)]
(ii) L[Kts\n[cot)]
Relate j^r) in Figure P 12.18 to ^ r) o f

part (a). Then find F{s) from G{s).
S)6
Figure P I2.18
19.(a)
22. (a)
Using the formulas cosh(/7r) =

and sinh(^/r) =
(i)
- e~^, find
Idt
using the derivative property.
f(X )dx
(iii) Com putc[//(0] =
multiplication-by-r property, compute

(1 )X [/ T r c o s h (/ 7 r )]
(c)
as a sum o f appropri
(ii) Compute \g {t ) \= /(/)
Using the formulas o f part (a) and the
(2)
Express
ate step functions. Compute F{s)
{\) [cosh(/7r)]
(2) L [sinh(^r)]
(b)
Considery(/) in Figure P i 2.22.
using the integral property.
[Kt sinHat)]
Again using the formulas o f part (a)
(b)
=J{t + 4).
Repeat part (a) for
and the multiplication-by-r property,
f(t)
compute
{\) [K r cosh(/7r)]
(2) [Kr
3 -
s m h ia t)]
2 -
20. Suppose/r) = 0 for r < 0 and F(.v) =
2.V + 4
s +1
> t
Find the Laplace transforms o f the functions
below, identifying each o f the properties used
Figure P I2.22
to compute the answer. Solutions obtained by

finding/f) are not permitted.
(a)
(b)
(c)
(d)
23. The Laplace transform
g^{t) = 5 J { t - T), T> 0

g,{t) = 2r^^t)
gît) = 2e-^ % t - D , T > 0
g^{t) = 5 i f i t - r i , T > 0
F{s) =
2 1. Use Liplace transform properties to find G/is)

4^ + 20
as eiven below when F{s) = --------- :r .
and
(.9+1)-
State each property that you used. Assume that
J{t) = 0 for ^ < 0 . Solutions obtained by finding

J{t) are not permitted.
(a) g^{t) = 0.5^r)
(b)
W
W)
^^W = 2 ^ , ( 2 f - 4 )
(c)
^5W = 2 ( r - 2 ) / ; - 2 )
(f)
\-e
is given as
-(s-a)
s-a
(a)
Find the Laplace transform o f e %t).
(b)
Find the Laplace transform of tj{2t).
AN SW LR: (a)
.s'
24. The Laplace transform
F{s) =
(a)
-V
J{t) is given as
l-e "_ ^
Find the Laplace transform of

withy(0) = 3.
cm
cit
(b)
Now find the Laplace transform o f
(c)
Finally, find the Laplace transform o f
dt
25. Supposey(/) = S(t) d{t T), T > 0.

(a)
Find L\J{2t)\ by direct calculation of
597
transform o f / r ) and then, using the relation

ship, find the Laplace transform o f^ f).
the Laplace transform integral.

(b)
Find L\J{2t)] by computing Hs) and

then using the scaling property.
26. Use only Laplace transform propenies to

answer the following question. Suppose that for
w
Vw/
O
- t
t < 0 , / / ) = eû{-t), and the one-sided Laplace

5 f" 1
transform o f/^ ) is X [ / ( f ) ] = F {s) = ^ .
L et^ r) =J{t)u(t). Find the Laplace transform of

v{t) when
(a)
(b)
(c)
v(t) = 2 g " {t)-g \ t)

=
t/ W = / a )+ J_ ^ q )d q
(d)
- t
It is not necessary to have the answer be a
rational function.
27. (a)
W
O '
O
Find the Laplace transform of the

function f j )
sketched
in
Figure
P12.27a.
(b)
Identify a relationship between

and the function
Figure P i2,28
sketched in
Figure P I2.27b . Use your answer to
29. Develop a relationship betweenj^f) and g(t)
part (a) and the appropriate property
in Figure P I2 .29. Find the Laplace transform
from Table 12.2 to compute the
ofy(/) and then
Laplace transform o f ^ /).
tionship between the two functions. Assume
by making use of the rela
that 0 < A < B < C. Also, determine D and E in

terms o f A, B, and C
O
->
o
Figure P I2.27
28. In Figure P I 2.28, what is the relationship

between j{t) and g^t) ? First find the Laplace
598
33. Find (i) the partial fraction expansion and

(ii) the inverse Laplace transform for each o f
the following functions by hand. Show all
work. (No details, no credit.)
(a) F,Cv) =
s{s
a){sb)
C H E C K : One residue is a.
^
- 7.v^ + 4.V + 2
( b ) F ^ s ) = ---------------------------- ^
Figure PI 2.29
30. Let j{t) and ^t) be as sketched in Figure
P I 2.30. Find G{s) in terms o f F{s).
f(t)
20
10
C H EC K : One residue is - 2 .
. X
s 2 / + 18i-^ + 4 6 5 ^ + 4 4 5 + 12
(c) F2,{s ) = --------------------r---------- z----------(5 + l ) ^ 5 + 2 r
C H E C K : Two residues are at 2 and - 2 .
Remark: Check answers using MATLAB.

Use the help command to make sure you
understand the terms used. For example
for part (b),

>
n = [2 -7 4 2];
d = conv([l -1],[1 -4 4]):
Figure P12.30
INVERSE LAPLACE
TRANSFORMS BY PARTIAL
FRACTION EXPANSION
31. Using partial fraction expansions and your
knowledge o f the Laplace transform o f simple
signals, find j{t) when F{s) equals
(a )
2 5 ^ + 1 3 r + 305 + 32
[r,p,k] =residue(n,d)
34. Find (i) the simplified partial fraction
expansion and (ii) the inverse Laplace trans
form for each o f the following functions by
hand. Show all work.
(a) Fi(.v) =
{a + b ) s + l a b
(5 + fl)(5 + b )
Check: One residue is a.

(b ) F2{ s ) =
( a + b + c ) s ' + ( b e + 2 a b + a c ) s + a b c
s { s + a ) { s + b)
s{s^ + 6^ + 8)
C H E C K : One residue is a.
(b)
- s-6
Cv + 2 ) ( . v - - l )
(c)
cs~ + { a + 2 a c ) s +
2.v'^ + 12.v +22.y + 8
( r +25 + 1)( j + 2)
(d ) F4(s ) =
(d)
il + c ) a ^ - a
(c) F3(.v) =
/ +185^+ 9 8 5 -+ 2085+ 144

(5 + 2 ) (5 + 4 )
/ + 1 2 A - ^ - 2 4 r - 3 2 .y + 16
(e)
C H EC K : Two residues are at 2 and - 2 .

(e ) F ^ ( s ) =
55^ + 1 4 4 5 + 2 0 4
(5+1)
32. Inadvertently left out by the authors.
(5+ 2) + 6 4
599
'w '
Vgf{s) and then find
35. F in d /r) when F{s) equals
(a)
(b)
(c)
(d)
for the
input current
25 + 16
+16
45 + 9
/X
lin ( s ) = 2 0 -------------- 2 2---------
245-72
5^ + 4 5 + 40
Check: One residue is at 20.
25^ + 885
(5 + 4)(5^ +64)
2 5 ^ + 2 5 ^ -2 5 -6
v.(t)
36. Find (i) the partial fraction expansion and
+
20
80
24 0
(ii) the inverse Laplace transform for each of

(a)
the following functions by hand. Show all

work.
, X r ., X
25^ + 125^ + 235 + 17

(5 + l)(5 + 2)(5 + 4 )
C H EC K : Residue at j = - 2 is - 1 .5 .
(b) F2(5) =
25^ + 9 5 ^ + 1 6 5 + 1 1
(5 + i)(5 + 2 y
C H EC K : Two o f the residues are 2 and 1.
5^ + 4 5 ^ - 2 5 ^ - 9 5 - 3
(C) ^3 ( 5 ) = -------------=-------- T-----
( 5 - 1 ) 2 ( ^ + 2)2
CHECK: Two residues are at I and two are at - 1 .
4 5 ^ - 1 2 5 ^ + 325 + 16
(d) /=4(5)=r
(5 + ir+ 1
38. Suppose F {s)
it follows that
(5 + 2)^ + 9
fj) =
cos{(ji)i)u{i) + Ar2^'sin(cflf)(r). Find

tf, A^j, Kj, and CO. Now express J{t) =
cos{(Ot + 9)u{i) by finding K y a, (O and d.
(5 + 2)2 + 16
37. Find the partial fraction expansion and the

inverse Laplace transform for each of the indi
39 .
The
Laplace
cated output voltages or currents. All answers
[A^jf'^'cos(fi)?) +
transform
of f j )
sin(fi)r) + Kê~^û{t) is
must be in terms o f real functions with real

coefficients or symbols. Show a ll work,
(a)
- 5 2 5 + 228
F {s)
For the circuit o f Figure P 12.37a, find

the partial fraction expansion o f
and then find
for die
(5 + 4 ) ( 5 + 0 ^ + 100
Find a, b, K^,
K y and 0).
input voltage
40. Consider the resistive circuit in Figure
P 12.40. Use Table 12.1 and the shift property
5(5 + 4 )
(b)
For the circuit o f Figure P 12.37b, find

the partial fraction expansion
for each Vj^(s) below. Sketch

tôuM) by hand or with the help o f MATLAB.
to find
of
600
(a)
Kv,(.v) =
IOf'"'' +
(b)
\-e'
Vi{s)=\0
- 5e'
..-4 5
-2 0 -
(b)
Use the Laplace transform method to

compute the inductor current,
for t > 0.
(c)
If the input is changed to
1 0 ( r - 7) V, where 7"= 10 msec, find

/^(/), for t > 0. Hint: Your answer
should be a shift o f the answer com
8kQ
puted in part (b).
-n/ V ^
8kQ
(d)
for r > 0.
1 8 kQ / 9kQ,
(e)
for / > 0. Hint: Can you
use superposition?
(0
If ij{0~) = 100 mA and

for t > 0.
tion expansions o f the rational functions listed

below. Then use Table 12.1 and MATLAB to
obtain the associated time function.

i,(t)
(c) F^{s) =
(d) 74(' ) =
(e) F^{s) =
= 10//(r
- 7) V, where 7 = 1 0 msec, find i/it),
41. Use MATLAB to compute the partial frac
(b) F2(s ) =
= 10//(r)
If //(0~) = 100 mA and

V, find
Figure P i2.40
(a) F,(5) =
= 0, find
If /^(0~) = 100 mA and
3.v'^ + 3 0 5 - + 86.9 + 6 4
'">6
.s'* + 8 i - +20.V+ 16
-O
+
v,(t)
- 4 6 .2 5 s - 6 9 2 . 8 125
.s-'* + 14.5.V- + 169.5625s + 510.25
-2.s-^ + 23.V- - 68.V - 3265
Figure P i2.42
.V - S 3 .55 " + 134.V + 797.5

10.5.r'^ + 47.875.s- + 151.875 - 108.5938
+ 6.5
+ 36.5625.v- + 101,5625.v + 2 07.0312
- 1 .5.s-'^ - 25.75.v^ - 127.5.y'* - 2 9 1 .5.y~ - 330^ - 143.75

.s-^ + 10.5.v-'^ + 50.v-^ + 141.V-'' + 250.V- + 262.5.V + 1 2 5
CIRCUIT RESPONSES VIA

LAPLACE TRANSFORM
APPLIED TO DIFFERENTIAL
EQUATIONS
4 3 . For the circuit o f Figure Pi 2.43, suppose R

= 10 2 and C = 0 .0 1 F.
(a)
Show that the differential equation lor
the circuit is
(h'rU)
1
1
^
+ ----- VcU) --------,ll
RC
RC
42. T he input to the circuit o f Figure PI 2.42 is

= 10//(r) V, valid for r > 0, and has an ini
tial inductor current of /^(0 ) = 0. Suppose R =
(b)
30 Q. and L = 0.2 H.
(a)
Show that the differential equation for
for r > 0 when

(c)
R. , .
1
, .
, ({) = - ViU)
L
L
= 10(r) V, valid
for t > 0.
the circuit is
di l it )
(If
Use the Laplace transform method to

compute the capacitor voltage,
Now suppose the initial capacitor volt

age ^'Ô") = 10 V and
= 0.
601
(a)
Find V(^t) , for r > 0.

(d)
Find
the capacitor voltage,
for r > 0, wlien

(b)
= 10/^r) V.
- 1 0 and
Construct a differential equation in

Solve the differential equation by the
Liplace transform method; i.e., show
that V(it) =
R
v jt)
/> 0 and for appropriate constants A'j,
K-,, and to, which are to be found in

terms o f /y,
L and C.
vJt)
sin(ojf) + KjCosUot) for
-O
i,(t)
ijt)
\r
Figure P I 2.43
;r
v^(t)
44. The circuit o f Figure Pi 2.44 has rwo source

excitations, v^^{t) = 10//(f) V and
= lu{t)
A, both applied at / = 0. Suppose /?j = 5 Q, R-,

= 20 Q, and Z = 2 H. The initial condition on
the inductor current is /^(0 ) = I A.
(a)
ANSWl-R; (a)
the circuit, assuming the response is

and
(c)
Find the response due only to
(e)
- \V' = 0 . ( b ) (I) = 7 =
LC ^
J lC
46. The circuit o f Figure P 12.46 is a series (loss

less) LC circuit driven by a voltage source.
Suppo.se
(a)
= 0 and /^(0 ) = 0.
Construct the differential equation o f
assuming /^(0 ) = 0.
the circuit in terms o f the capacitor
Find the response due only to ip{t)
voltage,
assuming /^(0 ) = 0.
(d)
dr
Construct a differential equation for

/^(r). Leave the inputs in terms o f
(b)
Fieurc P I2.45
(b)
Solve the differential equation using
Find the response due only to the ini
the Laplace transform method, and
tial condition ;^(0~) = -1 A, assuming
show that
both inputs are zero.
for t > 0.
10 -
10
c o s(0
.5 jia )
for t
Find the complete response,
l = ih
> 0 by superposition.
^
' ) Q
R.
' . ,(t)
4C=-,F
vJt)
Figure P I2.46
47. A pair o f (coupled) differential equations
Figure P i2.44
that represent a circuit are given as

M t )
45. Consider the Z,C circuit ol^ Figure P i 2.45,

for w'hich /^(O") = 7q and
= V^. Since
and
dt
there is no resistance present in the circuit,

there is no damping; hence, one expects a pure
ly sinusoidal response. Such circuits are called
lossless.
dt
-t-^/|.v(/) = oiyit)
602
with initial conditions ;c(0~) = 1 andyO ) = 2.

Suppose , = 1 ,
= 1. ^3 = 1 . ^3 = 1. and/r)
= 2u{t). Find j/(^) and A-(r).
51. The op amp in the circuit o f Figure Pi 2.51

is assumed to be ideal. /?, = 20
R-, = 40 ki2,
and C = 10 |iF.
(a)
= 2,
48. Reconsider Problem 47 with
Use nodal analysis to construct a first-
= 2,
order differential equation describing
= 4, b^ = 3, and/r) = 2u(t).
the input-output relationship o f the

voltages.
49. The inductor current i{t) in a second-order
If u jt ) = 2uU) V, and j/c(0) = -1 V,
(b)
RLC circuit satisfies the following integro-dif-
and
ferential equation for / > 0.
then
Sketch the response in

MATLAB.
(a)
If v-it) = 2e~^-">tt{t) V and ^CÔ) = 0,
(c)
v'c(() ) + 8 j^ _ //(X )f/r
find
If /(0-) = 8 A and vM~) = - 4 V, find
kis).
(b)
Use your answer in part (a) to find

/,(/).

(a)
Use KVL and KCL to show that the

differential equation relating the input
Figure P I2.51
voltage to the capacitor voltage is
ci\\
d\ 'c
(It-
RC clt
LC
''C
52. Reconsider the RC active circuit shown
LC
Figure
12.1
o f Example
12.1, where we
encountered difficult)' using the single third(b)
Take the Laplace transform o f both
order difTerential approach. Now we will solve
sides o f this equation to show that
the problem with Laplace transforms applied to

three first-order differential
_
Vc(A) =
RC
(c)
( .v + 5 ) v c ( 0 " ) -h v c (0 " )
I
1
s~ + ----- 5 +
1
s + -----
RC
LC
Assuming that vc(0 ) =
) = 0,
equations
LC
(a)
Three node equations in the time

domain have been given in equations
^ = 0.8 Q, Z. = 1 H, C = 0.25 F, and
12.1, 12.2, and 12.3. Take the Laplace
= 5(r), show that
transform o f each o f these three node

vc(/)
equations, accounting for initial con
(b)
L=1H
v(t)
;
R=0.8Q I
= lOu(t) V and the initial
capacitor voltages are ^^(O) = 12 V,

= 6 V, and t/^(0) = 3 V, find
ijt)
ditions.
If
v,(t)
(d)
Now do a partial fraction expansion of
Vgut^^^ and determine

Figure P I 2.50
for /> 0.
Laplace Transform Analysis II:

Circuit Applications
A FLUORESCENT LIGHT
APPLICATION
Fluorescence
is a process for converting one
type o f ligiit into another. In a fluorescent

light, an electric current heats up elcctrodes
at each end o f a tube. T he hot clcctrodes
emit free electrons, which, for a sufficiently
high voltage between the electrodes, initiate
an arc, causing mercury contained in the
tube to vaporize. The energized mercury
vapor emits invisible ultraviolet light that
strikes a phosphorus coating on the inside o f the tube. The phosphorus absorbs this invisible
short-wavelength energy and emits light in the visible spectrum.
A starter circuit must quickly generate a sufficient quantity of free electrons and crcate a suffi
ciently high voltage to initiate the arc that vaporizes the mercury inside the tube. One t)'pe o f
starter circuit contains a special heat-sensitive switch in series with an inductor. We will model this
special switch by an ideal heat-sensitive (bimetal) switch in parallel with a capacitor. The concepts
developed in this chapter will allow us to analyze the operation o f such a starter circuit as set forth
in Example 13.1 1 .
CHAPTER OBJECTIVES
1.
In terms o f the Laplace transform variable
s,
Z (j), and the notion o f admittance, denoted
define the notion o f impedance, denoted
y\s).
Impedances and admittances will sat
isfy a type o f O hm s law. These ideas are generalizations o f the phasor-based notions o f
impedance and admittance introduced in Chapter 10.
2.
Learn the arithmetic o f impedances and admittances in the Laplace transform domain,
which is analogous to the arithmetic o f resistances and conductances in the time domain.
Chapter 13 Laplacc Transform Analysis 11: Circuit Applications
3.
Apply the new concepts o f impedance and admittance to redevelop the notions o f volt
age/current division, source transformations, linearity, and Thevenin and Norton equiv
alent circuits in the /-dom ain.
4.
Define /-domain-equivalent circuits o f initialized capacitors and inductors for the pur
pose o f transient circuit analysis.
5.
Introduce the notion of a transfer function.
6.
Define rvvo special types
7.
Redevelop nodal and loop analyses in terms o f impedances and admittances.
8.
of
responses: the impulse and step responses.
Utilize the l-aplace transform technique, especially the /-domain-equivalent circuits o f

initialized capacitors and inductors, for the solution of switched /?ZC circuits.
9.
Introduce the notion o f a switched capacitor circuit, which has an important place in
real-world filtering applications.
10.
Set forth a technique for designing general summing integrator circuits.
SECTION HEADINGS
1.
Introduction
2.
Notions o f Impedance and Admittance
3.
Manipulation o f Impedance and Admittance
4.
Equivalent Circuits for Initialized Inductors and Capacitors
5.
Notion o f Transfer Function
6.
Impulse and Step Responses
7.
Nodal and Loop Analysis in the j-Dom ain
8.
Switching in RLC Circuits
9.
Switched Capacitor Circuits and Conservation of Charge
10.
The Design of General Summing Integrators
11.
Summary
12.
Terms and Concepts
13.
Problems
1. INTRODUCTION
Chapter 12 cultivated the Laplace transform as a mathematical tool particularly useful for circuits
modeled by differential equations. This chapter adapts the Laplace transform tool to the peculiar
needs and attributes o f circuit analysis. W ith the Laplace transform methods described in this
chapter, the intermediate step o f constructing a circuits differential equation, as was done in
Chapter 12, can be eliminated.
Available for the analysis o f resistive circuits is a wide assortment of techniques: O hm s law, volt
age and cu rren t division, nodal and loop analysis, linearit)', etc. For the sinusoidal steady-state
analysis o f RLC circuits, phasors serve as a natural generalization o f the techniques o f resistive cir
cuit analysis. The Laplace transform tool permits us to extend the sinusoidal steady-state phasor
analysis methods to a much wider setting where transient and steady-state analysis are both pos
sible for a broad range o f input excitations not amenable to phasor analysis. Recall that transient
an;ilysis is not possible with phasors.
Chapter 13 * Laplacc Transform Analysis II: Circuit Applications
T he keys to this generalization are the i-domain notions o f impedance and its inverse, admit
tance. Instead o f defining impedance in terms ofyoj, as in phasor analysis, we will define it in
terms o f the Laplace transform variable
s. This
definition allows the evolution of a frequency- or
j-dependent O hm s law, j-dependent voltage and current division formulas, and ^-dependent
nodal and loop analysis; in short, all o f the basic circuit analysis techniques have analogous
s-
dependent formulations. W hat is most important, however, is that with the ^-dependent formu
lation, it will be possible to define .^-dependent equivalents for circuits containing initialized
capacitors, inductors, and other linear circuit elements. These equivalent circuits make transient
analysis natural in the i-domain.
In the final section o f the chapter, we introduce the notion o f a switched capacitor circuit.
Switched capacitor circuits contain switches and capacitors, and possibly some op amps, but no
resistors or inductors. Present-day integrated circuit technolog)' allows us to build switches,
capacitors, and op amps on chips easily and inexpensively. This has fostered an important trend
in circuit design toward switched capacitor circuits. A thorough investigation o f switched capac
itor circuits is beyond the scope o f this text. Nevertheless, it is important to introduce the basic
ideas and thereby lay the foundation for more advanced courses on the topic.
2. NOTIONS OF IMPEDANCE AND ADMITTANCE

Chapter 10 introduced an intermediate definition o f (phasor) impedance as the ratio o f phasor
voltage to phasor current, and admittance as the ratio o f phasor current to phasor voltage. In the
Laplace transform context, impedances and admittances are j-dependent generalizations o f these
phasor notions. Such generalizations do not exist in the time domain. To crystallize this idea, we
Laplace-transform the standard differential
v-i relationship
o f an inductor,
at
to obtain
Vîs) ^ Lsliis),
assuming /^(0 )
(1 3 .1 )
Ls multiplies an ^-domain current, /^(^), to yield

Vjis), in a manner similar to O hm s law for resistor voltages and currents.
Ls are ohms. The quantit)' Ls depends on the frequency variable s and gen
= 0. Here, the quantit)' Z^(s) =
an j-domain voltage,
T he units o f Z^(j) =
eralizes the concept o f a fixed resistance, and it is universally called an im pedance. This complexfrequency or ^-domain concept has no time-domain counterpart.
Although the inductor served to motivate ^-domain impedance, in general an impedance can be
defined for any two-terminal device whose input-output behavior is linear and whose parameters
do not change with time. A device whose characteristics or parameters do not change with time
is called
time invariant.
606
Chapter 13 Laplacc'Iransforni Analysis II: Circuit Applications
IMPEDANCE
T he
impedance, denoted Z{s), o f a linear time-invariant
rwo-terminal device, as illustrated in
Figure 1 3 .1 , relates the Laplace transform o f the current, /(s), to the Laplace transform o f
the voltage, V^j), assuming that all independent sources inside the device are set to zero and
that there is no internal stored energy at ^ = 0 . Under these conditions.
V{s) = Z{s)I{s)
(13.2a)
and, where defined,
Z(s) =
V(s)
I(s)
(13.2b)
in units o f ohms.
l(s)
Device
vis)
Z(s) orY(s)
FIG U R E 13.1 A two-terminal device having impedance
Exercise.
For an unknown linear circuit,
ANSW'HR;
Vj,j{s) = ------
and
5 + 4
Z{s) or admittance
I- (s)
^(^).
= 2. Com pute
.y" + 4
T he inverse o f resistance is conductance, and the inverse o f impcdance is admittance. For exam
ple, if we divide both sides o f equation 13.1 by
Ls,
we obtain
Ls
This suggests that
\/Ls acts
is defined as follows.
as
a. generalized conductance universally
(1 3 .3 )
called an adm ittance, which
607
Chapter 13 Laplacc Transform Analysis II: Circuit Applications
ADMITTANCE
T he
admittance,
denoted y(^), o f a two-terminal linear time-invariant device, as illustrated
in Figure 1 3 .1 , relates the Laplace transform o f the voltage, V(j), across the device to the
Laplace transform o f the current, /{^), through the device, assuming that all internal inde
pendent sources are set to zero and there is no internal stored energy at f = 0~. Under these
conditions,
i{s) = ns)v{s)
(13.4a)
and, where defined,
Ijs)
(13.4b )
V{s)
in units o f S.
From equations 13.2 and 13.4 , impedance and admittance satisfy the Inverse relationship
Y(s) =
Exercise.
1
Zis)
(1 3 .5 )
16
For an unknown linear circuit,

'
C om pute Ky,//).
(.S-
----
arid
+ 2)
A/l(-^)="
(>v +
2 )(i + 4 )
2
ANSWER:
^+ 4
As a first step in deepening our understanding o f these notions, we compute the impedances and
admittances o f the basic circuit elements shown in Figure 13.2.
i,(t)
ic(t)
O-
O
+
-I-
v,(t)
V ,(t)
v(t)
o >
o-
(a)
FIGURi-l
(b)
1 3 .2
From O hm s law, the resistor o f Figure 13.2a satisfies

sides yields the obvious, K^(j) =
(c)
(a) Resistor, (b) Capacitor, (c) Inductor.
Rljîs).
= Rijît).
Laplace-transforming both
From equations 13.1 and 13.2, the im pedance o f the
resistor is
Zfîs) = R
608
Chapter 13 l-aplacc Transform Analysis II: Circuit Applications
and, from equation 13.5, rhe adm ittan ce o f the resistor is
Here the kinship of impedance/admittance with resistance/conductance is clear.

rh e capacitor o f Figure 13.2b has the usual current-voltage relationship,
(!v(^{ t )
dt
/c ( 0 = C-
Assuming no initial conditions, the Laplace transform relationship is
I^s) = CsV^s)
From equation 13.4 , the ad m ittan ce o f the cap acito r is
Y(is) = a
and from equation 13.5, the im pedance o f the cap acito r is
Z eis) =
1
Cs
Repeating this process for the inductor o f Figure 13.2c ,
)=
the im pedance and
ad m ittan ce o f the in d u cto r are
Z,,U) = U . YLU) = - jLs
Exercises. 1. Given the integral form o f the v-i capacitor relationship, assume no initial stored
energy and take the Laplace transform o f both sides to derive the impedance o f the capacitor. This
provides an alternative, more basic means of deriving the impedance characterization.
2. Given the integral form o f the
v-i
inductor relationship, assume no initial stored energy and
take the Laplace transform o f both sides to derive rhe admittance o f the inductor.
Throughout the rest o f the text, whenever we refer to an impedance the unit o f Ohm is assumed,
and similarly, admittance is assumed to have the unit of siemens (S). The units for KW and
I{s)
are usually not shown, although strictly speaking they are volt-second and ampere-second, respec
tively.
609
Chapter 13 Laplace* Transform Analysis II: Circuit Applications
3. MANIPULATION OF IMPEDANCE AND ADMITTANCE

Recall that the Laplace transform is a linear operation with respect to sums of signals, possibly
multiplied by constants. KVL and KCL are conservation laws stating, respectively, that sums of
voltages around a loop must add to zero and sums o f all currents entering (or leaving) a node must
add to zero. Since the Laplace transform is linear, it distributes over these sums, so the sum of the
Liplace transforms o f the voltages around a loop must be zero and the sum o f the Laplace trans
forms o f all the currents entering a node must be zero. In other words, complex-frequency
domain voltages satisfy' KVL and complex-frequency domain currents satisfy' KCL. Because of
this, and because impedances and admittances generalize the notions o f resistance and conduc
tance, one intuitively expects their manipulation properties to be similar. In fact, this is the case.
M a n ip u la t io n
r u le .
Because impedances map j-domain currents,
I{s),
to ^-domain voltages,
K(j), and because all /-domain currents must satisfy' KCL and all /-dom ain voltages must satisf)
KVL:
1.
Impedances,
Z{s), can
be manipulated just like resistances and, like resistances, have units
o f ohms.
2.
Admittances, K(/), can be manipulated just like conductances and, like conductances,
have units o f S.
This manipulation rule suggests, for example, that admittances in parallel add. T he following
example verifies this property for the case o f two admittances in parallel.
EXA M PLE 13.1. C om pute the equivalent admittance,

general admittances, K,(/),
V^{s), and Y:^{s) in
and impedance,
o f three
parallel, as shown in Figure 1.3.3. Then develop the
current division formula.
Yji-s]
K,(.v)+Ko(.v)+r,(.v)
(1.3.6)
An(-v)
Z Js)
FIGURE 13.3 Three general admittances, Vjis), in parallel,
having an equivalent admittance K- (/) or impcdance
Z- (s).
(>10
Chapter 13 * Liplace Transform Analysis II: Circuit Applications
S o lution
W c seek the relationship
vvhich implicitly defines
o f the admittance o f a two-terminal device, /^(j) =
= AW
for
^3(5) =
From the definition
= 1 ,2 , 3. From KCL,
>3(5))
This relationship implicitly defines the equivalent admittance as
= y i(s)+ y 2 (s)+ y 3 (s)
y i(s)=
affirming that admittances in parallel add. From the inverse relationship
Z; (5 ) = -------------- !-------------K,Cv)+K20v) + y3(.v)

Returning to the relationship
/j^(s) =
we now note that
/;,(,C) = Y / : ( S ) V , ( S ) = y , ( s ) Z , ( s ) / , ; , ( s ) =
Y^{s )+Y 2( s )+Y2{ s )
Equation 13 .6 has the obvious generalization to any number o f parallel elements.
Exercises.
1. Show that for t%vo impedances,
Z^{s)
and
Zjis),
in parallel, Z ,(^) =
2. Show that the equivalent impedance o f two capacitors in parallel is
Z (5 )=
and that the equivalent capacitance is
C]^+C25
(C j+ C 2 )5
2 i ( ^) +
= Cj + C 2 .
3. Derive the following formula for the impedance o f two inductors in parallel:
L^ + Lo
4. A 2 |.iF and a 0 .5 uF capacitor are in parallel. Find the equivalent capacitance.
A N SW ER : 2.5
til
5. A 2 mH inductor is connected in parallel with a 0 .5 mF capacitor. Find the equivalent imped

ance.
ANSWER: 2.^
'
.s- + i ( r "
61 1
Chapter 13 Laplace Transform Analysis II: Circuit Applications
6. In the circuit o f Figure 13.3, suppose Kj(j) =
MR,
K,(j) =
M{L$),
and
Cs,
a resistance,
an inductance, and a capacitance. Derive the relationship
1
fin (s)
L C s ^ + -s +
R
the equivalent admittance,
and find
terms o f I;{s).
in'
Ai\S\V1-:R: y j s ) =
I^{s)
1
^ ^
7. In the circuit o f Figure 13.3, suppose Kj(j) = 0 .5 ,
25+1
in
Find
^
l and /^(.v) =
EXAMPLE 13.2. Compute the input impedance o f the parallel RLC circuit sketched in Figure 13.4.
o
+
VJs)
Z Js) =
Y Js)
FIG U R E 1 3 .4 Parallel
RLC circuit
for Example 13.2.
S o l u t io n
For parallel circuits, it is convenient to work with admittances, since
parallel admittances add.
Thus, for the circuit o f Figure 13.4,

^
r +
RC
.v+'
LC
Since impedance is the inverse o f admittance,

1
C , 2 ^ _ 1L , +
RC
w hich is the equivalent input im pedance o f a parallel
LC
RLC circuit.
(1 3 .7 )
612
Exercises.
1. C om pute the equivalent impedance o f a parallel connection o f three inductors hav
ing values 4 m H , 5 m H , and 2 0 m H .

A N SW ER : 2 x 1()--S2. Com pute the equivalent impedance o f a parallel connection o f six elements: rsvo resistors, o f 6
kQ and 3 kQ; two inductors, o f 3 mH and 6 m H ; and two capacitors, o f 0 .2 |.iF and 0 .0 5 |.iF.
A N SW E R ; 4 X 1()^V(r + 2 x 1
i)-^s + 2 x 1 0 * )
1 he dual o f the parallel circuit o f Figure 13.3 is a series connection o f three impedances as shown
in Figure 13.5. T he following example verifies that impedances in series add, and simultaneous
ly develops a voltage division formula.
E X A M P L E 1 3 .3 . Com pute the equivalent impedance,
general impedances,
Z^(s), Zîs),
and
Z^{s)
and admittance,
o f three
in series, as shown in Figure 13.5. Then develop the
voltage division formula,
Zj{s)
=
Z,(.v) + Z2(.v) + Z3(i-)
+ V,(s) -
V^n(.v)
(1 3 .8 )
+ V^(s)
r i d l J R E 13.5 Series impcdance circuit illustrating voltage division.

S
o l u t io n
O hm s law tell us that

(1 3 .9 )
for / = 1 , 2 , 3,. From KVL
(1 3 .1 0 )
Using equation 1 3 .1 0 and the definition o f input impedance, it follows that
Z, (s) = -^^4^ = Z| (s) + Z , ( i) + Z3(,s)
(1 3 .1 1 )
Chapter 13 * Laplacc Transform Analysis II; Circuit Applications
613
T he voltage division formula o f equation 13.8 follows from a modified form o f equation 1 3 .1 0 ,
and equation 13.9, to yield
Z i (5) + Z2(5) + Z3(5)

T he voltage division formula is easily extended to the case o f w devices in series:
Z (s )
Z ,( 5 ) + Z 2 ( s ) ...+ Z(5)
Exercises.
1. C om pute the equivalent impedance o f two capacitors, C, and C j, in series.

1
AN SW I-R;
C|CS
Cj + C .
2. Show that the equivalent admittance o f rwo capacitors, Cj and C-,, in series is
3. Suppose
Zj(s)
= 10 Q,
Z-,(s)
2s,
and
Z^(s) = 6^ in
Figure 13.5. Find
Y{s) = -------- s.
C\ + Cl
V^-)W, and
.V
A N SW ER S:
Z,,(s)
= 10 + 8^. \ s ( .0 =
4.V + 5
. /-.(/) =
'
Zj(s) = 10 Q, Z-,(s) = 2s,

V^Js).
4. Suppose
terms o f
and
-7
+ 10.v + 2
A N SW ER S: Z,(.v) = ---------------------,
V
Z t^(s )
=-
=
^
in Figure 13.5. Find
Z^-^^(s)
and K^(j) in
-------------- V;(.v)
,v' + 5 .v + !
5. Verify that the equivalent inductance o f two inductors in series is
= -î +
O f course, there are series-parallel connections of circuit elements that combine the concepts illus
trated in Examples 13.1 through 13.3, as set forth next.
EXA M PLE 13.4. Com pute the input impedance Z-J,s) o f a series connection o f t\vo pairs o f par
allel elements, as shown in Figure 13.6, in which
Then compute
in terms o f
If
= 10 Ci, C = 0.1
u{t),
find
Vjit).
= 5 O.,
and Z. = 1 H.
Chapter 13 l^place Transform Analysis II: Circuit Applications
614
FIG U R E 13.6 Series-parallel connection of

S
RC elements
for Example 13.4.
o l u t io n
Conceptually, view the circuit as shown in Figure 13.7.
V,(s)
FIG U R E 13.7 Conceptual series structure of the circuit in Figure 13.6.

Here
Z ,U ) =
10
O.l+O.l.v
.y + 1
and
Z2{s) =
Is
7+2
in which case
10(.y + 2 )+ 2 .s;(.v + 1) _ 2^ + 1Is + 20
(5 + 1 )(5 +
(s + \)(s + 2)
2)
It fo llo w s th at
1// X
(.v + l )(i + 2 )
2.V
.v(.v+l)
2.v + 12.v + 20
-V+ 2
.v + 6 .y + I 0
Chapter 13 * Laplace Transform Analysis 11; Circuit Applications
Finally, if
61 S
= -,
(^+1)
(-^ + 0
^2(^') = -3 ---------------= ---------- ^
.9 + 6 . 9 + 10
( 5 + 3)" +
From Table 12.1, item 19,
Exercise.
Repeat Example 13.4 with the following changes: C = 0.01 F and /?, = 10 Q.
A N SW ER S:
Z Js )
= 10
VS(.v) =
\+10
Another basic and useful circuit analysis technique is the source transform ation property,
exhibited now in terms o f impedances and admittances. The first case we will examine is the voltage-to-current source transformation, illustrated in Figure 13.8,
(b)
(a)
FIG U R E
1 3 .8
Z^[s), as shown in
Z^(s), as shown in part (b).
Illustration of the transformation of a voltage source in series with
part (a), to an equivalent current with a current source in parallel with
Often, voltage-to-current source transformations provide an altered circuit topology that is more
convenient for hand or calculator analysis. Mathematically, the goal is to change the structure o f
a voltage source in series with an impedance to a current source in parallel with an admittance
while keeping both
Vjis)
and / 2 W fixed. To justify this, one starts with Figure 13. 8a, in which
voltage division implies
V,(.v) = ^
^ V^îs)=Z.is)l2is)
Z ,(5 )+ Z 2 (5 )
H en ce,
ZAs) ^ 0 ,
V ;(.v) =
Z2(.V)Z, is )
(V iu (s)\
Z,(.v)+Z2(.v) I 2 ,(5 )J r,(5)+K2(.v) U i(.v )j
(1 3 .1 2 )
Chapter 13 U p lace Transform Analysis 11: Circuit Applications
616
where
Yjis)
[Zj{s)]
^ This equation identifies the parallel structure of Figure 1 3 ./b ; i.e., Figure
13. 8b is a circuit equivalent o f equation 13.8.

Reversing these arguments leads to the current-to-voltage source transformation, illustrated in
Figure 13.9.
(a)
(b)
FIG U R E 13.9 Illustration of (a) current source to (b) equivalent voltage source transformation.
Clearly, the manipulation of impedances and admittances parallels that o f resistances and con
ductances, as suggested earlier. Indeed, for a rigorous statement o f the soiuce transformation tech
nique developed above, refer to the source transformation theorem in Chapter 5 and replace
by
Z{s),
by
and
by
Ijp).
Indeed, all such values in Chapters 5 and 6 have i-domain
counterparts.
This section ends with a demonstration of finding a Thevenin equivalent in the /-dom ain.
E X A M P L E 1 3 .5 . C om pute the Thevenin equivalent circuit o f Figure 13.10.
VJ s )
V Js )
v(s)
(b)
FIG U R E 1 3.10
SOLUTIO N
From the material in Chapter 6, our new concepts o f admittance and impedance, and Figure
13.10b,
(13.13a)
o r equivalently,
o s-B b )
Chapter 13 * L^placcTransform Analysis II: Circuit Applications
61
Now from Figure 13.1 Oa,
(d + 1)/(^(.V)+ GV'j-^C.v)
/(; (.v)+ CV^(5) =
= {ii + i).vc[v;(.9) - \/^(^)] + gv;(.9)

= I {a + 1).vC + C ]
(13.14)
( 5 ) - ( + 1).vCV/^(.v)
Rewriting equation 13.14 in the form o f equation 13.13a, we have
(a + \)sC + G
(13.15)
'
(W +D.9C + G
Com paring equations 13.15 and 13.13a, we identify
(a + 1).vC + G
Exercises. 1.
In Example 13.5, what is the Norton short circuit current,
A N SW ER : /.^.(.v) = U +
2. Find
(fl + 1).?C + G
and
\)s(:Vjj,s)
VgJ<s) for
the circuit in Figure 13.11.

2-2t;\
..
2/ / , (. s)
2 + 2.V +
FIG U RE 13.11
3. For the circuit o f Figure 13.12, use source transformations to find I^p) and Y^jj^s) for the indi
cated terminals.
A N SW ER S: I. is) = 0 . 2 .a ' (>) and
= 0.2.^ + - + 0.4
0.2 F
(V Js )
2.5 Q
1 H
F IG U R E 1 3 .1 2
618
Chapter 13 * Laplace Transform Analysis 11: Circuit Applications
4. EQUIVALENT CIRCUITS FOR INITIALIZED INDUCTORS AND

CAPACITORS
T he notions o f impedance, admittance, and transfer function do not account for the presence of
initial capacitor voltages and initial inductor currents. Hoiv can one incorporate initial conditions
into various analysis schemes? For an answer vve look at the transform o f an initialized capacitor
and inductor and interpret the resulting equation as an equivalent circuit in the complex-frequency domain. For the capacitor and the inductor, rvvo equivalent circuits result for each; a series
circuit containing a relaxed (no initial condition) capacitor/inductor in series with a source, and
a parallel circuit with a relaxed capacitor/inductor in parallel with a source. Example 1 2 .2 3 pre
viewed this notion.
T he capacitor has the standard voltage-current relationship
(it
= /c (0
Taking the Laplace transform and allowing for a nonzero initial condition
Cs
yields
- Cv(J,Q~) = I (is)
(1 3 .1 6 )
T he left side of equation 1 3 .1 6 is the difference o f two currents, one given by the product o f the
capacitor admittance and the capacitor voltage
{CsVf^s))
and the other by Cy^Ô"). Thus the cir
cuit interpretation o f equation 1 3 .1 6 consists o f a relaxed capacitor in parallel with a current

source, as illustrated in Figure 13.13. In the time domain the current source o f Figure 1 3 .1 3 cor
responds to an impulse that would immediately set up the required initial condition.
1^(5)
^
.....................................................
V,(s)
Cs
FIGllRK 13.13 Parallel form of an equivalent circuit for an initialized capacitor. Here, the capacitor
within the dotted box is relaxed while the current source
Cv(4S)~) accounts
for the initial condition.
Rearranging equation 1 3 .1 6 yields
Cs
(1 3 .1 7 )
Example 12.22 previewed this equation by taking the transform o f the integral relationship o f the
capacitor. We observe that the right-hand side o f equation 1 3 .1 7 is the sum o f two voltages, one
o f which is the product o f the capacitor impedance and the capacitor, current and the other
z/^(0)A'. Thus, the interpretation is a series circuit, as sketched in Figure 13.14.
619
V Js)
FIG URE 1 3.14 The series form of an equivalent circuit for an initialized capacitor. Here the capacitor
in the dotted box is relaxed, and the voltage source accounts for the effect of the initial condition.
Initialized inductors have similar j-domain equivalent circuits analogous to those o f the capaci
tor. W ith the voltage and current directions satisfying the passive sign convention, the differen
tial inductor current-voltage relationship is
Transforming both sides yields
( 13 . 18 )
Again, this equation consists o f a sum o f voltages,
Lslîs)
and - Z /^ ( 0 ). Thus equation 1 3.18 can
be interpreted as a series circuit, as depicted in Figure 1 3 .1 5 .
FIG U R E 13.15 Series form of equivalent circuit for an initialized inductor. Here the inductor with
in the dotted box is relaxed; notice the polarity orientation of the voltage source.
To construct a parallel equivalent circuit, divide equation 1 3 .1 8
Ls
by Ls and
rearrange to obtain
(1 3 .1 9 )
T he right side o f Equation 13.19 is a sum o f currents that determines a parallel equivalent circuit,
as sketched in Figure 13 .1 6 .
Chapter 13 * Laplace Transform Analysis II: Circuit Applications
620
F'lGURli 13.16 Parallel form of equivalent circuit for an initialized inductor.

Again, the inductor inside the dotted box is relaxed.
Two examples illustrate the use o f these four equivalent circuits for initialized capacitors and
inductors.
E X A M P L E 1 3 .6 . This example illustrates an ^-domain application o f superposition. In the /?Z,Ccir
= Au{t) V.
cuit o f Figure 13.17, suppose ^(--(0 ) = 1 V, /^(0 ) = 2 A, and
Find Vj{t) for f > 0.
+ vjt) --------- (-
F K iU R E 13.17 Circuit for Example 13.6.

S
o l u t io n
In this example, it is convenient to replace the capacitor by its (series) 5-domain voltage source
equivalent circuit, because the capacitor is in series with the input voltage source. On the other
hand, it is convenient to replace the inductor by its (parallel) /-dom ain current source equivalent
circuit, because the desired output is the inductor voltage. This results in a three-source or multi
input circuit. O nce the equivalent circuits are in place, one can apply superposition to obtain the
answer, although there arc many other ways to solve the problem.
Using the voltage source ynodelfor the capacitor and the current source modelfor the induc
tor, draw the equivalent s-domain circuit. Using the equivalent circuits o f Figures 1 3 .1 4 and 1 3 .1 6 ,
Step 1.
we obtain the circuit of Figure 13.18. Here we note that
V;(.v) = - .
---- = - ,and
-----= - .
621
>
O
+
1 .5 0
V,(s)
0.5s
FIG U R E 1 3 .1 8 j-domain equivalent accounting for initial conditions of the circuit of Figure 13.17.
Step 2 . Fmd the contribution to
from
From voltage division,
S- + 3S + 2
I.5 + - + 0.55
s
from
^ = - . Again, from voltage division,
V l(s) =
-.y
s^ + 3s + 2
X =
.5 + - + 0.55
Step 4 . Find the contribution to Vj{s) from Z,;'^(0 ) = 1. Using O hm s law in the 5-domain,
0.5^ 1.5 + -
s/
V t{s) = -
1.5 + - + 0.55
5
-3 s - 2
^ + 3^ + 2
--------------
X =
Step 5. Su?n the three contributions and take the inverse transform.
V i(s) = v l u ) + v l { s ) + vl(.s) = -
-2
- + 35 + 2
-V+ 2
.V+ 1
in which case
Vj{t) =
Exercise. Find Ij{s) and
2e
^hi{t)
- le
û{t) V
ij{t) for the circuit o f Figure 1 3 .1 7 using the equivalent circuits o f Figures
1 3 .1 4 and 1 3 .1 5 . Hint: Write one loop equation.

ANSW FR:
/ ,(,)=
.//(/) = M - ' i a i ) - 2( " '/M/)

( .V
I )(.V + 2 )
622
E X A M P L E 1 3 .7 . This example illustrates a single-node application o f nodal analysis. In the

circuit o f Figure 13.1 9 , suppose
V(^Qr)
= 1 V, /^(0 ) = 2 A, and
n{t) V.
Find
V(^t)
RLC
t>
for
0.
0.5 H
/Y Y \
-o
1.5 Q
+
v,(t)
v Jt)
1F
MGURH 1 3.19 Circuit for Example 13.7.

S
o l u t io n
In this example, it is convenient to replace the inductor by its (series) coniplex-frequency domain
voltage source equivalent circuit, because the inductor is in series with the input voltage source.
On the other hand, it is convenient to replace the capacitor by its (parallel) complex-frequency
domain current source equivalent circuit, because the desired output is the capacitor voltage. This
results in a three-source, or multi-input, circuit. O nce the equivalent circuits are in place, one can
combine the voltage sources and write a single node equation to find
V(As).
Using the voltage source model for the inductor and the current source ynodelfor the capaci
tor, draw the equivalent complex-frequency domain circuit. Using the voltage source equivalent for
Step 1.
the initialized inductor and the current source equivalent for the capacitor produces the circuit o f
Figure 13.20a. Combining the voltage sources and the series impedance into single terms results
in the circuit shown in Figure 13.20b.
0.5 i^(O-) = 1
/ Y Y V
1.50
_ Q
0.5 s
CvJO-) = l
(a)
FIG U R E 1 3.20 (a) Complex-frequenc) domain equivalent accounting for initial conditions o f the
circuit of Figure 13.19. (b) Circuit equivalent to part (a) with voltage sources combined.
Step 2.
Write a single node equation for V(^s).
623
Summing the currents leaving the top node o f
yields
1
V c ( s ) ---------------- l + 5V c ( 5 ) = 0
1.5 + 0.5.y
Grouping terms produces
V c is )= -:
----- r+ 1
5 (0 .55'+ 1.5 )
+ 5
U . 5 + 0.5^
V(\s)
Solving for
leads to
s~ + 5s + 2
Vcis) =
5
( 5 + 0 ( 5 + 2)
Exemte a partialfraction expansion on V(^s), and take the inverse transform to obtain V(^t).
Step 3 .
Using the result o f step 2,
5^ + 55 + 2
1
2
-2
^(7(5) = -------------------- = I----------- f-
s{s +
1) ( 5 + 2 )
.V +
+ 2
Inverting this transform yields the desired time response,
v^t)
Exercises.
le-'-2e-^~û{i)V
1. In Example 13.7, change the resistance from 1.5 H to 2 .2 5
A N SW ER :
2. Find
= [1 +
v^p)
I^{s)
and
Find
V(\t)
for r > 0.
= [1 0.57l4i>-'^-^'- 0 .5 7 l4 f -^ q (/) V
iît)
for the circuit o f Example 13.7, using the equivalent circuits o f Figures
13.14 and 13.15. Hint: W rite one loop equation.

TV
A N SW ER :
1,
EXA M PLE
1 3
( 5 ) = -------- ----------- .
( 5 + I ) ( 5 + 2)
i,{t)
^
=-
le -ii{t)
Ae--'u{t)
. 8 . T he chapter opened with a discussion o f the operation o f a fluorescent light
with classical starter, com m on in residential usage. For a fluorescent light to begin operating, there
must be a sufficient supply o f free electrons in the tube and a sufficiently high voltage between the
electrodes to allow arcing to occur. During arcing, mercury particles in the tube vaporize and give
off ultraviolet light. The ultraviolet light excites a coating o f phosphorus on the inside o f the tube
that emits light in the visible range.
For a simplified analysis, assume that all resistances are negligible and refer to Figure 1 3 .2 1 . The
source
VjJ^t)
is 120 V, 6 0 Hz, i.e., ordinary house voltage, which is too low to cause arcing inside
the fluorescent tube. Prior to arcing the gas inside the fluorescent tube acts like a very large resist
ance betvN'een the rwo electrodes. W hen the switch is turned on, the starter, a neon bulb with a
bimetallic switch inside, lights up and heats the bimetallic strip. This causes the metal to curl and
624
close the contacc. The bulb then looks like a short circuit, and a large current, limited by the
inductive ballast, flows through the heating electrodes o f the fluorescent tube, making them bet
ter able to emit electrons. During this time the neon bulb is shorted out and the bimetallic strip
cools and opens the circuit after a few seconds. At this point in time, which we will call ^ = 0 , the
inductor has an initial current
Because o f the Z.Ccombination, a very high voltage will then
appear across the electrodes of the lamp, resulting in ignition or arcing. After the lamp ignites, the
voltage between the electrodes becomes small and is insufficient to relight the neon starter lamp.
Hence, the ac current flows between the two electrodes inside the fluorescent tube. The ballast
again serves to limit the current.
Heating
Direction
of curl when
FIGIIRK 13.21 Wiring diagram of simple fluorescent light circuit, including an inductive ballast, a
capacitor, and a starter within which is a neon bulb containing a bimctallic switch.
Suppose
L = 0 .8
H , C = 1 nF, and
= 0.1 A. For r > 0, we find the com ponent o f
due
to the initial inductor current, i.e., the zero-input response. The other com ponent, the zero-state
response, is not as important for ignition purposes. O ur strategy will be to use the ^-domain equiv
alent circuit for
L,
as illustrated in Figure 13 .2 2 .
0 .8 s
Voltage
due to
I,(S)
Li,(0)
= 0.08
intial
inductor
current
'....................................................
u- u
*
High resistance
prior to
arcing
FIGUKI! 13.22 Equivalent complex-frequenc)' domain circuit immediately

prior to arcing and normal lamp operation in fluorescent lighting.
625
Since we are assuming that all resistances are negligible and that the internal resistance (between
electrodes) o f the fluorescent lamp prior to arcing approximates infinity, voltage division in terms
o f impedances yields
+ Ls
Cs
r +
J l .25x10^
+ 1 .2 5 x 1 0
LC
J l . 2 5 x lo '
2 ,8 2 8 ^
o
. r + 1.25 X 10^
Hence, immediately prior to arcing, the capacitor voltage approximates
= - 2 ,8 2 8 sin (3 5 ,3 5 5 /) V
which is sufficiently high to induce arcing and cause the fluorescent lamp to operate.
See the homework exercises for an extension o f this analysis to the case where the ballast model
includes a resistance o f 100
Q.
5. NOTION OF TRANSFER FUNCTION

Besides impedances and admittances, other quantities such as voltage gains and current gains are
critically im portant in amplifiers and other circuits. T he term transfer fu nction is a catchall
phrase for the different ratios that might be o f interest in circuit analysis. Impedances and admit
tances are special cases o f the transfer function concept.
TRANSFER FUNCTION
Suppose a circuit has only one active independent source and only one designated response
signal. Suppose fiirther that there is no internal stored energy at f = 0~. T he transfer func
tion o f such a circuit or system is
H{ s) =Thus if the input
X fdesignated response signal
--------- r------ . V
(13.20)
[designated input signal
and the response is^(f), then y(^) =
which is a handy for
mula for computing responses. Notice that if the input is the delta function, then
and
Y{s)
H{s).
F{s)
= 1
This means that the transfer function is the Laplace transform o f the so-
called im pulse response o f the circuit, i.e., the response due to an impulse applied at the
circuit input source when there are no initial conditions present. The idea is easily extend
ed to multiple inputs and multiple outputs to form a transfer function matrix. This exten
sion, however, is beyond the scope o f this text.
626
Exercise.
A transfer function o f a particular circuit is
H(5 ) =
response. Hint: Review Table 12.1.
. Find the impulse

(5' + ^^)
A N SW ER : ^'-"[cos(Z^/) + S sin(Z^r)]/Hf)
A transfer function, as defined by equation 1 3 .2 0 , has broad applicability to electrical and electro-mechanical systems. For example, the designated output may be a torque while the input
netiuork
driving point impedance, where the
the current source; (ii) driving point
might be voltage. However, in the context o f circuits, a transfer function is often called a
function.
T he literature distinguishes four special cases: (i)
input is a current source and the output is the voltage across
admittance,
where the input is a voltage source and the output is the current leaving the voltage
source; (iii)
transfer impedance,
where the input is a current source and the voltage is across a des
ignated pair o f terminals; and (iv)
transfer admittance,
where the input is a voltage source and the
output is the current through another branch in the circuit. In cases (i) and (iii), the voltage polar
ity must be consistent with the conventional labeling o f sources as set forth in Chapter 2. In gen
eral, however, we will adopt the ordinary language o f transfer function.
EXA M PLE
1 3
. 9 . T he circuit o f Figure 1 3 .2 3 has elements with zero initial conditions at f =
0 Find
ôut
V :Js)
o l u t io n
There are many ways to solve this problem. O ur approach is to execute a source transformation
on the
R-L
impedance in series with the voltage source. After the source transformation, we use
current division to obtain the necessary transfer function.

Step 1.
Execute a source transformation to obtain three parallel branches as per Figure 13.24.
627
FIG U R E 1 3 .2 4 Circuit equivalent to Figure 13.23 after a source transformation.

This circuit has the parallel structure o f Figure 13 .2 5 .
,(S)
V Js ) r A
Y,
sT T
FIG U R E 1 3 .2 5 Parallel admittance form of Figure 13.24.

Step 2 . Use current divisiofi. Since the output current,
^ current through one o f three
parallel branches, the current division formula (equation 13.9) applies, producing
>3(^)
M l
y,(A-)+r2(.s') + >3(-^'V 5 + 1
Hence,
H(s) =
Yîs)
(1 3 .2 1 )
U W + i 2 ( > ^ ) + W / s+ 1
Vinis)
Step 3 . Compute K, (^), Y2 {s), and Y^îs). Because impedances in series add, and admittance is the
inverse o f impedance (equation 13 .7 ), some straightforward algebra yields
K,(.v) = ----- K2 (.v) = ------------. = ----- ^3(5) =

S+\
5+1
5
2.5s
0 .4 5 +
, 2^1
0.4
Step 4 . Substitute into equation 13.21 to obtain H{s):
2.5 s
H(s) =
1
5+1
2.5s
5^ +1
2.55
+ - I X---- \ 5 + 1/
5
5+1
+ 1 + 5 ( 5 + 1) +
s +l
2.5s
2.5s
(5 + 1 ) ( 5 - +
2.5s +
1)
(s +
1)(^ + 0 . 5 ) ( 5 + 2 )
.5 5 ( 5 + 1)
(i2
Exercise.
For
A N S W FR :
H{s)
as computed in Example 13.9, find the so-called impulse response
//{/) =
--C
-0 .5 /
-2 /
h{t)
u(t
E X A M P L E 1 3 .1 0 . C onstruct the transfer function o f the ideal operational amplifier circuit o f

Figure 13 .2 6 , where
Zp)
and
Ip)
denote a feedback impedance and feedback current, respec
tively.
V,(s)
FIG U R E 1 3 .2 6 Simple ideal operational amplifier circuit for Example 13.10.

S
o l u t io n
Since no current enters the inputs o f an ideal op amp, I - p ) = - I p ) . Further, the voltage at the
negative op amp terminal is driven to virtual ground; hence,
V-p)
Z -p )I-p ),
and
Vg^,f{s)
Z p ) I p ) . Combining these relationships with I - p ) = - I p ) yields
Vi{s)
Zj{s)
Yj{s)
(1 3 .2 2 )
Equation 1 3 .2 2 is a verv' handy formula for computing the transfer functions and responses o f
many op amp circuits.
Exercises.
Find R so
1. In the circuit o f Figure 1 3 .2 6 , suppose

that the transfer function is
A N SW E R :
R=
H{s)
-Ms,
Z p)
is the impedance o f a 0.1 mF capacitor.
i.e., an inverting integrator.
1{) kQ
2. In the circuit o f Figure 1 3 .2 6 , now suppose

a 0 .2 m F capacitor, and
the transfer function,
Zis)
consists o f a 10 kQ resistor in parallel with
consists o f a 4 0 k ti resistor in parallel with a 0 .4 m F capacitor. Find
the dc gain, and the gain as x
A N SW E R : -{s + 0.5)/(2y + 0 .1 2S).
-O.S
oo.
629
3. Find the value o f C for which the transfer function o f the op amp circuit in Figure 1 3 .2 7 is
H {s) = ----------- ---------- .

(s + 2)is + 4)
A N SW ER : C = 0 .5 F
0.2 5 Q
FICJURE 1 3 .2 7 Op amp circuit.
E X A M P L E 1 3 .1 1 . T he ideal op amp circuit o f Figure 1 3 .2 8 is called a leaky ifjtegrator. If the

input to the leaky in tegrator circu it is v-^^) = e~û{t), find the values o f /?,,
an output response
Rj,
and Cleading to
= -lte~ u{t), assuming that I'fjCO") = 0.
R,
FIG U R E 1 3 .2 8 Ideal operational amplifier circuit known as the leak)' integrator.

S
o l u t io n
Step 1.
From the given data, compute the actual transferfunction o f the circuit.
By definition o f the
transfer function,
H{s) =
L[resp(mse\
Voui{s)
(s + l)
\input\
Vi(s)
1
.y+1
5+1
( 1 3 .2 3 )
(i3 0
Step 2 . Using Figure 13.28, fin d the transferfiinction o f the circuit in terms o f R^, Rj, and C. Here,
obsenê that Figure 1 3 .2 8 has the same topolog)^ as Figure 13 .2 6 , where
Ky.(,9)= ! =
c j+
Z /(.v )
-^
'
R2
'
i
From equation 13.22 o f Example 13 .1 0 ,
Cs + Step 3 . Match coejfcients in equatiotis 13.23 and 13.24a to obtain the desired values ofR^, Rj, and
C. Equating the coefficients yields

_1_
Cv+
Rl
O ne possible solution is
R^ =
0 ,5 Q, /?, = 1 Q, and C = 1 E If we rewrite equation 1 3 .24a as
H (s)= -
(1 3 .2 4 b )
CRj
other solutions are also possible. For example, for any

=
> 0,
^2new ~
^new
represents a valid (theoretical) solution. In Chapter 14 we encounter a concept called
magnitude scaling.
is called a magnitude scale factor, which leaves this transfer function
unchanged but produces more realistic values for the circuit elements.
Exercises. 1. In equation 13.24b, it is required that C = 10 uE Find appropriate values o f /?j and R-^.
AN SW ER:
R^ =
50
kLl R,
= 100
kLl
2. Given equation 13.24b , compute h{t) = L [//(^)].
A N SW ER :
_JL
t
H,C
tl(l)
631
6. IMPULSE AND STEP RESPONSES

H{s), with ^-domain input
Y{s) = H W F{s). Assuming that
Suppose a circuit or system lias a transfer function representation

denoted by
F{s)
and j-domain output given by y(y) in which case
all initial conditions arc zero, if /(f) =
^{t),
then the resulting^(r) is the system im pulse response.
Some simple calculations verify that the transform o f the impulse response is the transfer function,
1.e.,
X[y(r)] =
H {s)m t)]
His)
Hence, the impulse response o f the circuit/system, denoted /;(/), is the inverse transform o f the
transfer function
Ht) -'[M(s)]
(13.25a)
H{s) = [h{t)]
(1 3 .2 5 b )
and conversely
These equivalences represent another use o f the transfer function concept.
Exercises.
l.T h e transfer function o f a certain linear network is
H{s) = (s + 5)l[{s + 1)(j + 2)].
Find
the impulse response o f the network.

A N SW E R :
[lc^ -
2. If the impulse response o f a circuit is a pulse ^(f) =
u{t) - u {t- T), T > 0,
compute the transfer
function.
A N SW ER : (I
3. Suppose
- e^'^)/s
t/{t) = 2b{r-
ing an impulse response

ANSXXHER: y(r) =
2hU -
1) - 3 6 (/- 3) is the input to a relaxed (zero initial conditions) circuit hav
h{t)
2u(t) - 2u{t5).
Find the output ^(f).
1) - 3A(r - 3)
Why is the impulse response important? hs
we will see, it is because every linear circuit having con
stant parameter values for its elements can be represented in the time domain by its impulse
response. This is shown in Chapter 15, where we define a mathematical operation called
tion
convolu
and show that the convolution o f the input function with the impulse response function
yields the zero-state circuit response. In addition to this significant theoretical result, the impulse
response is im portant for identification o f linear circuits or systems having unknown constant
parameters. Sometimes a transfer function is unavailable or a circuit diagram is lost. In such a
predicament, measuring the impulse response on an oscilloscope as the derivative o f the step
response is quite practical.
What is the step response o f a circuit? T he step response is merely the zero-state response o f the cir
f^t) to the circuit is (/), then F{s) = 1/^ and K(j)
= H[s) (1/^). By the integration propert} o f the Laplace transform, it follows that the step response
cuit to a step function. Observe that if the input
032
is the integral o f the impulse response. Conversely, the derivative o f the step response is the
impulse response. In lab, many scopes can display the derivative of a trace and hence can display
the derivative o f the step response, which is the impulse response. Alternatively, a homework prob
lem will suggest a means o f directly generating an approximate impulse response.
Exercises.
1. If the transfer function o f a circuit is
H{s)
= 1/j, what arc the impulse and step
responses?
2. If the Laplace transform o f the step response o f a circuit is given by
Y{s)
= I/fi'U + I)], what is
the impulse response?

3. If the step response o f a circuit
\sy{t)
= [1 - 0.5^ ^-
cos(2r)]//(r), what is the impulse
response?
u(t),
A N SW ER S: in random order: H/).
cus(2/ + 2 6 .5 7 ')//(r).
E X A M P L E 1 3 .1 2 . Figure 13.29a shows the impulse response o f a hypothetical circuit. If an input
= b{r)
b {r -
1), com pute the response,^(f).
y(t)
A
h(t)
3
A
2
--
1 H-----
(a)
(b)
FIG URE 13.29 (a) Impulse response of hypothetical circuit, (b) Response to 6(r) +
b {t-
1).
S o lu t io n
Since X [6(r)
+ 6(/- 1)]
= I +
e~\ the
response,
is simply the sum
o f /}{[) and h (t- \)u {t-
1).
Doing the addition graphically yields the waveform o f Figure 13.25b.
E X A M P L E 1 3 .1 3 . The response o f a relaxed circuit to a scaled ra m p ,/(/) =

= ( - 6 + 4r + 8 e~' S
Com pute the impulse response,
is given by;/(^)
h{t).
o l u t io n
T he relationship between/( /) and
b{t)
identifies the strateg)' o f the solution. If the step function
is the integral o f the delta function and the ramp the integral o f the step, then the delta function
equals the second derivative o f the ramp. Hence,
the impulse response
h{t)
y\t)
b{t)
= 0 .1 2 5 /" (r ). By the linearity o f the circuit,
= 0.1 2 5 7 "(^ ), and some straightforward calculations produce
= [4 -
+ [ - 6 + 4 / + 8 ^ * -'-
(t)
Chapter 13 Laplace Transform Analysis 11: Circuit Applications
y"{t)
But the right-hand term is zero. (Why?) Hence,
633
= [Sf*
and
hit) = [e-^-
To see the utility o f this approach, try the alternative method o f computing
V(s)/f(s).
F(s), V(s),
and
H(s) =
The algebra is straightforward, but tedious and prone to error.
As a final example, we compute a circuits step response and verify that its derivative is the impulse
response.
E X A M P L E 1 3 .1 4 . C om pute the step response o f the
RLC circuit
o f Figure 13 .3 0 .
/m
R = 40
v
L=1H
C = 0 .2 F
-o
FIG U R E 13.30
S
circuit for Example 13.14.
o l u t io n
From voltage division.
s~ -\ --s+
L
LC
f.v + 2 ) + l
Cs
/^ + L .V +
Cs
(1 3 .2 6 )
From equation 1 3 .2 6 , the Laplace transform o f the step response is
His)
= - +
.y (.y + 2 ) + l l
-.V - 4
(.v + 2)^ + l
Rearranging terms yields
.V
(5 + 2 ) - + 1
{s + 2)-+\
Taking the inverse transform produces the desired step response:

= [1 -
cos(f) -
2e-~ sin(f)];K r)
(1 3 .2 7 )
fi3^
Chapter 13 Laplacc IVanstorm Analysis II: Circuit Applicarions
As a check, obser\'e that the derivative o f equation 1 3 .2 7 is
, , ( / ) = 2 e ""^ [co s(/) + 2 s i n ( 0 k ( 0 -
(It
= 5e
+ 2 c o s(/)] /(/) + (l - 1)(5(/)
sin{t)u{t)
Thus
5
Cv + 2 ) - + 1
in which case
h{{) =
5^" ^sin(r)(f)
as expected.
6. NODAL AND LOOP ANALYSIS IN THE S-DOMAIN

This section develops ^-domain formulations of node and loop analysis. Nodal analysis o f circuits
builds around KCL, whereas mesh/loop analysis utilizes KVL. In Chapter 3 and, indeed, in most
beginning courses on circuits, loop and nodal analysis are taught first in the context o f resistanc
es and conductances and then (in Chapter 10 here) in the phasor context. Recall that KCL
requires that the sum o f the currents leaving any circuit node be zero. Further, KVL requires that
the voltages around any loop of a circuit sum to zero. By linearity, the Laplace transform o f a sum
is the sum o f the individual Laplace transforms. Hence, a KVL equation and a KCL equation have
an j-domain formulation where elements are characterized by impedances and/or admittances.
For loop analysis, one writes a KVL equation for each loop in terms o f the transformed loop cur
rents and element impedances. The set o f all such equations, then, characterizes the circuits loop
currents, which determine the ciu rents through each o f the elements. Knowledge o f the loop cur
rents and the element impedances permits the computation o f any o f the element voltages.
In nodal analysis, one writes a KCL equation at each node in terms o f the Laplace transform o f
the node voltages with respect to a reference, the transform of the independent excitations, and
the element admittances. The set o f all such equations characterizes the node voltages o f the cir
cuit in the ^-domain. Solving the set o f circuit node equations yields the set o f transformed node
voltages. Knowledge o f these permits the computation o f any o f the element voltages. W ith
knowledge of the element admittances, one may com pute all o f the element currents. Since nodal
analysis has a more extensive application than loop analysis, our focus will be on nodal analysis.
E X A M P L E 1 3 .1 5 . Figure 13.31 shows an ideal operational amplifier circuit called the
Key normalized low-pass Butterworth filter.
Sallen and
(See Chapter 19 for a full discussion o f filters.) A nor
malized low-pass filter passes frequencies below 1 rad/sec and attenuates higher frequencies. As we
will see later in the text, the 1-rad/sec frequenc)' cu toff can be changed to any desired value by
frequency-scaling the parameter values o f the circuit. (See Chapter 14 for a discussion o f frequen-
63 S
cy scaling.) The goal here is to utilize the techniques o f nodal analysis to compute the (normal
ized) transfer function o f this circuit.
FIG U R E 13.31 Sallen and Key normalized Butterworth low-pass

filter circuit containing an ideal operational amplifier.
S
o l u t io n
T he solution proceeds in several steps that utilize nodal analysis techniques in conjunction with
the properties o f an ideal op amp. Recall that for an ideal op amp, the voltage across the input ter
minals is zero and the current into any o f the input terminals is also zero. Finally, note that one
does not write a node equation at the output, which appears across a dependent voltage source
whose value depends on other voltages in the circuit.
Step 1. Find Vy. Because the voltage across the input terminals o f an ideal operational amplifier
is zero.
Step 2. 'Write a node equation at the node identified by the node voltage V^. Summing the currents
leaving the node yields
(Va -Vi) + (V a-V t) + V2.v(l/,

Substituting
for
- V,) = 0
and grouping like terms produces
(J2.S + 2 ) V - ( J 2 S + \)V,=V,
(13.28)
Step 3. Write a node equation at the node identified by the node voltage Vy. By inspection, the desired
node equation is
5+1
(1 3 .2 9 )
Step 4 . Write the foregoittg tivo node equations in matrix form. In matrix form, equations 1 3 .2 8 and
13.29 combine to give
- (V 2 . + I)'
-1
-|=5+ 1
IV 2
J
K;
'în
0
(1 3 .3 0 )
6.U>
Step 5. Solve equation 13.30 for
in terms ofV-^^ using Cramer's rule. From Cram ers rule,
del
'{ J lS + l )
V;
-1
(>/ 2 :i- +
-(V
.V + 1 )
del
-1
\42
The resulting transfer function is
K..
[.j2 s + 2 )lj^ s+ \ \ -[y l2 s+ \ )

IV2
s- + yf2s+\
N otice that for small values o f ^ = yto (i.e., low frequencies), the magnitude o f H{s) approximates
1, and for large values o f s = JiO (i.e., high frequencies, where |/b)|
small. Since
1), the magnitude o f H(s) is
= Myo)) ^^^(yco), such a circuit blocks high-frequency input excitations and
passes low-frequency input excitations. As mentioned at the beginning o f the example, the circuit
passes low frequencies and attenuates high frequencies.
T he preceding example used matrix notation, com m on to much o f advanced circuit analysis. In
one sense, matrix notation is a shorthand way of writing n simultaneous equations: the n variables
are written only once. More generally, matrix notation and the associated matrix arithmetic allow
engineers to handle and solve large numbers o f equations in numerically efficient ways. Further,
the theory o f matrices allows one to develop insights into large circuits that would otherwise
remain hidden. Hence, many o f the examples that follow will utilize the elementary properties o f
matrix arithmetic.
T he next example uses nodal analysis to compute the response to an initialized circuit. The exam
ple combines the equivalent circuits for initialized capacitors and inductors with the technique o f
nodal analysis.
E X A M P L E 1 3 .1 6 . In the circuit o f Figure 13.3 2, suppose /y^^(r) = 6(r), /^(O") = 1 A, and

= 1 V. Find the voltages
V(^t)
FIG U R E
and
v^Qi~)
vît).
1 3 . 3 2
Two-node /?Z.Ccircuit for Example 13.16.
Given the indicated current direction of i[{t), what is the implied voltage polarit)' for
v^{t)}
63'
Step 1. Draw the s-domain ecjuivaletit circuit with an eye toward nodal analysis. Inserting the equiv
alent current source models for the initialized capacitor and inductor in Figure 1 3 .3 0 , one obtains
the 5-domain equivalent circuit shown in Figure 13 .3 3 .
V Js)
VJs)
1o
ii(O-)
,Cv,(0 )
O
1o
= 1
1F
1H
FIG U R E 1 3.33 5-Domain equivalent of the circuit of Figure 13.31.

Step 2 . Write two node equations and put in matrix fonn. At the node labeled
V(is)
K C L implies
that
(1 +
s)V(is)
[V(is)
Vj{s)] = 2
Simplifying produces the first node equation:
{s^ 2 )V ^ s)-V i{s) = l

- V(is)] + {\ls)Vj{s) = - (1/i), or, equivalently,
At the node labeled
.9+1
-V c(^ ) +
1
=
T he matrix form o f these uvo node equations is

.v + 2
2 1
-1
_i
.9+1
s
Step 3 . Solve the matiix equation o f step 2 for the desired voltages. Using C ram ers rule, computing
the inverse, or simultaneously solving the equations gives
2{s+\)-\
2
Vcis)-
5+ 2 . 9 + 2
[5 +1
1 1 2
s
1
5+2
(s+Vr + \
(1 3 .3 1 )
_i
C v + D -3
(.V+ 1r +1
Step 4 , Take the inverse Laplace tratisfonn to obtain time domain voltages. Breaking up equation
13.31 into its components yields
Chapter 13 L iplacc Transform Analysis II: Circuit Applications
638
2(^ + 1)________ 1
Vcis) =
(5 + 1)- + !
(5+ 1)^ + 1
in which case
V(^t)
e ^[2 cos(t)
- sin(/)]/^(r)
Also,
(.^ + 1)
(A-+1)^ + 1
(.V+1)^ + 1
V^(s) =
leading to
=e
Figure 13.34 presents plots o f
V(\t)
^[cos(^) - 3sin(f)](r)
and
H G U R E 13.34 Plots of the capacitor and inductor voltages for Example 13.16.
Dual to nodal analysis is loop analysis. In loop analysis, one defines loop currents and writes KVL
equations in terms o f these loop currents. The following example illustrates the method o f loop
analysis for computing the input impedance o f a bridged-T network.
639
E X A M P L E 1 3 .1 7 . Use loop analysis to com pute the input Impedance o f the bridged-T network
illustrated in Figure 13 .3 5 .
/YY\ 1 H
2Q
20
20
0.25 F
Z Js)
FIG U R E 1 3 .3 5 Bridged-T network for Example 13.17.

S
o l u t io n
Define
find
= / , U) in Figure 1 3 .3 4 . Since
I-^j{s)
Step 1.
the goal is to use loop analysis to
= /j(^) in terms o f
Smn the voltages around loop 1.
Dropping the specific j-dependence for convenience, one
obtains
2(/, - / 3 ) + - ( / , - / 2 ) = 2
Step 2.
Sum the voltages around loop 2.
- - / 2 - 2 /3 =
By inspection,
- ( / 2 - / | ) + 2( / , - / 3 ) + 2 / 2 = - - / , + 4
.V
Step 3 .
Sum the voltages around loop 3.
Again by inspection,
2 (73- 1{) + sl^ + 2(73 - I^) = - 2/1 - 11^ +

Step 4 .
/ 2 - 2 /3 = 0
Put the three loop equations in matrix form, and solve.
-K4)73 = 0
In matrix form, the three loop equa
tions are
, 2
.1
/l1
.1
-2
-2
-2
5+ 4
0
0
Chapter 13 * I.aplacc Transform Analysis II: Circuit Applications
(viO
Using Cram ers rule to solve this equation for
in terms o f
= / ] W yields
.v + 4.V + 4
=2 Q
It may be a little surprising that
is independent of
s, despite
the appearance o f an inductance
and a capacitance inside the nerwork. Such a nerwork is called a
constant-resistance network.
problem at the end o f the chapter shows the conditions on the elements o f a bridged-T network
for it to be a constant-resistance nerwork.
Exercise.
The loop equations o f the circuit in Figure 1 3 .3 6 are

T + 25
-2 s
\h]
-Vs:
4
2.V + - + 2
_2
-1
-2 i'
_2
25 + 4
-I
-4
h
V
Find the values o f C and A'in Figure 13.36.

ANSWHRS: c ;= 0.2^ ! A -
M GURE 13.36
7. SWITCHING IN /?iLC CIRCUITS

Switches control lighting systems, furnaces, car ignitions, traffic lights, and numerous other
devices. Switching also takes place inside electronic circuits, as in switched power supplies and
switched capacitor filters. All have a functional element called a
switch
that affects and, indeed,
shapes the behavior o f the circuit. T he switch inside an electronic circuit is a special device that
we will model simply as an ideal on /off switch. This section investigates the behavior of switching
in simple
RLC circuits,
64
as preparation for the understanding o f switching in more elaborate and
sophisticated electronic circuits. O ur immediate task is to apply the Laplace transform method to
com pute the responses of switched
RLC circuits.
The following example motivates a general pro
cedure.
EXA M PLE 1 3 .1 8 . In the circuit o f Figure 1 3 .3 7 , suppose
u{t),
= 0.
= 20 Q ,
R = A Q., C = 0 .2 5
F, and
The switch is initially in position A. T he switch S moves from
position A to position B at / = 1 sec and from position B to position A at ^ = 2 sec, and moves
V(^t)
back to B at / = 4 sec where it remains for all subsequent time. Find
for
t>
0.
v,(t)
v jt)
FIG U R E 1 3 .3 7 Switched /eCcircuit {R^ = 20 Q, R = A Q . , C = 0.25 F) for Example 13.18 in which
V(^Q~) = 0;
the switch S moves to B at ^ = 1 sec, returns to A at ^ = 2 sec, and moves back to B at r =

4 sec, where it remains.
o l u t io n
Because o f the switching at r = 1 sec, the first step is to determine

f < 1, This allows us, in turn, to find
V(-{t) over
the time interval 0 s
which will serve as the initial condition over the time
interval 1 s r < 2. This then produces
(2), the initial condition for the interval
2<. t <A,
etc.
Step 1. Compute the response for 0 ^ t < 1. Over the interval 0 :s r < 1 , the circuit o f Figure 13.37
is the simple
RC circuit
o f Figure 13.38.
20 0
+
V,(s)
0.2 5 F
FIG U R E 13.38 Equivalent circuit of Figure 13.37 over 0 ^
< 1.
For the circuit o f Figure 1 3 .3 7 , voltage division implies
s
20
20
(v + 0.2)(.? + 0 . 1 ) " .v + 0.1 ~ .v + 0 .2
+ -
.V
Hence, for 0 ^ r < 1,
v^t)
20e-^-^^-20e~^-^
20
642
W e note that
Step
(1") =
2 0
^0 . 1 ^ j
7 2 2
Coynpute the response over 1 <. t < 2.
V.
After the switch moves ft-om position A to position B,
the source is decoupled from the right half o f the circuit; the response then depends only on the
initial condition at r = 1 , i.e.,
^ f<
= 1 .7 2 2
, or equivalently, over the inten'al
V. T he goal is to compute
s r' <
where
t' = t - \. The
v^^t)
over the interval
^-domain equivalent cir
cuit that models the behavior o f the time domain circuit o f Figure 1 3 .3 7 over
\ ^ t < 2 has the

Cvq (1) =
form illustrated in Figure 13 .3 9 . We note that the value on the current source is
0 .4 3 0 5 .
FIG U R E
1 3 . 3 9
Equivalent circuit of Figure 13.37 for the time interval 1
^ t <2.
T he equivalent admittance seen by the current source is 0.25^ + 0 .2 5 . Hence by O hm s law,
Vc(.s) =
For 0 s ^' <
s +1
1 72^
X 0 .4 3 0 5 3 = ^
.V +
, taking the inverse transform yields

=
or equivalently, for
^ r<
\.722e-^'ti{t)
V(^t) = \.722e-^^-^\i{t- 1)
W e emphasize that this last equation is valid only for
: r <
. We note that
(2') = 1 .7 2 2 e ^ =
0 .6 3 3 5 .
Step 3. Compute v(it) fo r 2 ^ t <4. For this interv'al, the capacitor is initialized at /' = t - 2 = 0.
Using the parallel current-source equivalent circuit, we note that
(r')
0 .1584, as shown in Figure 13.40. We further note that

iV(r)= vv(r'+2) = ? ; ( ;')=
Thus
Vjfjis)
= 8 .l 8 7 .V - ''' ( r ')
in Figure 1 3 .4 0 is
8 .1 8 7 3
5 + 0.1
(2') =
Chapter 13 Laplacc Transform Analysis II; Circuit Applications
643
FIG U RE 1 3 .4 0 Equivalent j-domain circuit valid for 2 s r < 4.

Using superposition on the circuit o f Figure 13 .4 0 ,
4
0 .1 5 8 4
0 . 2 x 8 .1 8 7 3
0.25.S + 0 .0 5
U + 0 .2 )(5 + 0 .1 )
____s
20
+ -
0 .6 3 3 5
5 + 0 .2
.V
Hence, for 0 s t'< 2,
= l6 .3 7 4 6 ^ -^ -^ ''-1 5 .7 4 ll^ -0 " ^ '

or equivalently for 2 < f < 4,
v^t) = l6.3 7 4 6 < '-0 -i(^ -2 )-1 5 .7 4 llf> -0 -2 (^ -)

W e note that
v^4~) = \63746e-^ ~ -
1 5 . 7 4 1 =
2 .8 5 4 8 V.
Step 4 . Compute the response over 4 ^ t. After the switch moves from position A to position B,
again, the source is again decoupled from the right half o f the circuit. According to Chapter 8, we
can write the answer by inspection: the solution is simply v^^t') = 2.8548^ ^ n{t). Equivalently,
f o r / > 4 , y J r ) = 2 .8 5 4 8 f - ( '- ^ \
Step 5. Combine results into a single expression and plot. I'h e combined expressions for V({i) arc
Os/ <l
1.722e"^ '"W - l )
v c(0 =
2.85486^
A plot o f the response appears in Figure 1 3 .4 1 .
\^t < 2
Chapter 13 * l^placc Transform Analysis II: Circuit Applications
()-i4
FIGURK 13.41 Plot of capacitor voltage for switched circuit of Figure 13.37.
T he extension o f this method to more than three switching times is straightfoi^vard. Although the
preceding example uses an
RLC circuit,
the same strategy or algorithm is applicable to the calcu
lation o f switching transients in any linear dynamic circuit. T he following is a summary o f the
general procedure.
P rocedure for Applying Lapiace Transform M ethod to Switched C ircuits
At / = 0, a switching may or may not occur. For / > 0, denote the switching time instants
sively as
, r-, ,
The goal is to com pute the response
Divide the time axis into intervals (0 ,rj),
successively for
fo r
succes
0 s r < sc.
^ ^ = co), and com pute the responses
each time interval.
P art I. Consider the first time interv'al, (O.fj).
Step I. C onstruct
t = 0~, which are
the ^-domain equivalent circuit, making use o f the initial conditions at

either given or calculated from the past history o f the circuit for ; < 0
(see note at the end o f the procedure). This equivalent circuit is valid for the time inter
val 0 ^ r < / j .
Step 2.
Step 3.
Find the response by the Laplace transform method for the circuit o f step 1.
Evaluate the capacitor voltages and inductor currents at
t = T j.
Part II. For all remaining time intervals:
Step 4. Set the value of subscript / = 1.

Step 5. (1) For t^^ t< t^^^, let t' = t - t- (i.e.,
consider the time interv'al 0 s /' < oo), and
construct the ^-domain equivalent circuit, making use o f the

conditions at /'= 0 (i.e., at r =
t~).
calculated
initial
60
Chapter 13 * Liplace Iransform Analysis II: Circuit Applications
(2) Determine the proper form o f the input excitation(s) (if there are any), in
terms of t .
(3) I'incl the response by the Laplace transform method. Note that the time vari
able is t' . Then obtain the solution in
t by
substituting { t - t'^ for t
(4) If / = w, stop. Otherwise, eN-aluate the capacitor voltages and inductor currents at
(5) Increase the subscript value / by 1 and go to the beginning o f step 5.
Note: In some situations, the first switching occurs at r = 0, but the capacitor voltages and induc
tor currents at r = 0 are not given. Instead, the problem specifies that dc and sinusoidal sources
have excited the circuit for a long time. If the network is passive i.e., if it consists o f inductors,
capacitors, or
lossy elcm ents-then
the circuit will have rcached a steady state at
t = 0 . The
proce
dure then is first to find the steady-state solution and then to evaluate the capacitor voltages and
inductor currents at / = 0 . It is instructive to review the dc and sinusoidal steady-state (phasor)
analysis methods studied in a first course. Recall that under certain stabilit)' conditions (to be stud
ied in Chapter 15) on the network:
1.
For dc steady-state analysis, open-circuit all capacitances and short-circuit all inductors
to find the steady-state voltages and currents.
2.
For sinusoidal steady-state analysis, use the phasor method to find the steady-state
responses.
8. SWITCHED CAPACITOR CIRCUITS AND CONSERVATION OF

CHARGE
In addition to its many uses already described, the Laplace transform method is applicable to a
special class o f circuits called
switched capacitor (abbreviated
SC) networks. These circuits contain
only capacitors, switches, independent voltage sources, and possibly some operational amplifiers.
No resistors or inductors are present. One can dispense with resistors because it is possible to
approximate the effect o f a resistor with rwo switches and a capacitor. Similarly, inductors can be
approximated by circuits containing only switches, capacitors, and operational amplifiers. These
facts, coupled with the easy and relatively inexpensive fabrication ol switches, capacitors, and
operational amplifiers in M OS (metal-oxide semiconductor) technolog}^ have made switched
capacitor filters an attractive alternative to classical filters. Given this scenario, the purpose of this
section is to lay a foundation (i.e., introduce the principles) upon which switched capacitor cir
cuit design builds. More advanced courses delve into the actual analysis and design o f real-world
switched capacitor circuits.
Besides the Laplace transform approach, an alternative method for analyzing switched capacitor
networks builds on the principle o f conserv'ation o f charge.
Principle o f conservation o f charge: T he total charge transferred into a junction (or our o f
a junction) o f a circuit at any time is zero.
646
This principle is a direct consequence o f KirchhofFs current law. For example, in Figure 1 3 .4 2 ,
KCL implies that
(13.32)
/jU) + ijit) + i^{t) + i^{t) = 0
I,
l-IGUlUi 13.42 Node to which KCL applies.

Since charge is the integral o f current over a time interval, integrating both sides o f equation 1 3 .3 2
from -00 to
t yields
f
li,(r) + i 2 ( 0 + h ( 0 + U U m = 0
X
or equivalently.
î(^) + ^2^^) + ^3(^) + ^4(r) = 0

which is just another expression o f the principle o f conservation o f charge. A simple switched
capacitor circuit will now serve as a test bed for comparing the merits o f the foregoing analysis
techniques.
EXA M PLE 1 3 .1 9 . Consider the circuit shown in Figure 13.43a. The switch S is closed at / = 0.
Just before the closing o f S, the initial conditions are known to be
C om pute the voltages
and
= 1 V and
for ^ > 0.
i(t)
^ -----
Vc, -
IF r
IF
-----
+
Vc, ^
C2
IF
IF
---- -----
(a) t < 0
(b) t > 0
= 0.
Chapter 13 Ijp la c e Transform Analysis II: Circuit Applications
64"
l(s)
V,C2
(C )
FIG U R E 13.43 Equivalent circuits for Example 13.19.

S O L U T IO N
M ethod 1. Using the parallel current-source model o f a capacitor, we obtain the ^-domain equiv
alent circuit shown in Figure 1 3.43c. By inspection,
0 .5
s
Therefore,
V(y{t) = 0.5u{t)
V, and
i{t)
s s+ s
= 0.56(^) A.
M ethod 2 . Using the series voltage-source model o f a capacitor, we have the ^-domain equivalent
circuit shown in Figure 1 3.43d . Again, by inspection.
i + i '
S
'
i+ i
This is the same answer as obtained with method 1.

M ethod 3 .
Clearly,
Conservation-of-charge approach.
For r > 0 , the network is shown in Figure 13.43b .
After S is closed, some charge is transferred from C, to C2 . However,
according to the principle o f conservation o f charge, the
total charge
transferred out o f the junc
tion must be zero. Hence,
and
Solving these two equations for the two unknowns,
Vci(() ) - V c 2 ( 0
and
results in
) - ------------7 -------------------C] + C 2
Since there is no external input applied, the voltages remain constant once the equilibrium con
dition has been reached. Therefore,
Clhapccr 13 * I.aplacc Transform Analysis il; Circuit Applications
(vi.S
for r > 0. For the specific capacitance values given in Figure 1 3.43a, we obtain
0 .5 V for r > 0.
Exercise. In the circuit o f Figure 13.43a, let C, = 4 F and C-, = 6 F,
) = 8 V, and v a m
3 V. Find, by at least two methods, the capacitor voltages after the switch is closed. (Rework the
problem if the answers do not agree.)
A N SW ER : S \'
Computationally, the Laplace transform method is more straightforward. On the other hand, the
conservation-of-charge method is more basic and often provides better insight into what happens
to the charges stored in various capacitors. It is particularly useful for the purpose o f checking
answers obtained by other methods: the answers are correct when the conser\'ation-of-charge con
dition is met at every node.
E X A M P L E 1 3 .2 0 . The initial conditions at r = 0 o f an SC network are shown in Figure 13 .4 4 .
Switches Sj and S-, are closed at / = 0, connecting the tw'o dc voltage sources to the nersvork. Find
the node voltages for
t>Q.
lO V
FIG U R E 1 3 .4 4 Switched capacitor nervvork for Example 13.20 at r = 0 .

S
o l u t io n
We first construct the.f-domain equivalent circuit using admittances. The result is shown in Figure 13.45.
fO
15
4s
3s
2s
12
+ A IQ
FIG URE 1 3.45 Equivalent circuit to Figure 13.44, accounting

lor initial condiiions; the admittance ol each capacitor is indicated.
Chapter 13 * l^placc Transform Analysis II: Circuit Applications
6 i9
Applying KCL to node 3 generates the node equation
4s
from which wc derive

verify that
= 9/s and
+ 3.V
10
= 9 V for
+ I s Vt. = 4 + 1 5 + 12
r > 0. Obviously, Kj = 5 V and K2 =10 V. To
= 9 V is indeed the correct solution, we check for conservation o f charge. From
t=
0~ to A= O'*, tiic voltage of the 4 F capacitor jumps from 1 V to 4 V (note that 4 = 9 - 5 ) , indi
cating that 4 X (4 - 1) = 12 coulombs o f charge have been transferred to this capacitor. The volt
age o f the 2 F capacitor jumps from 6 V to 9 V, indicating that 2 x (9 - 6) = 6 coulombs o f charge
have been transferred here. Finally, the voltage o f the 3 F capacitor changes from 5 V to - 1 V (note
that - 1 = 9 - 1 0 ) , indicating that 3 x (-1 5) = - 1 8 coulombs of charge have been transferred to
this capacitor. As a check for conservation o f charge, we have 12 + 6 + ( - 1 8 ) = 0, and the solution
is assured to be correct.
Exercise.
Solve Example 1 3 .2 0 again, with all capacitors initially uncharged.
A N SW ER :
v^(t)
= 5 0 /J V, / > 0
T he preceding examples considered idealized circuits, i.e., no resistances were present. In zny prac
tical circuit, the connecting wires have some resistance. W hat is our interest in the analysis of an
idealized circuit? T he analysis o f an idealized circuit is much more straightforward than that of a
realistic circuit yet provides relatively accurate answers. As a case in point, reconsider Example
1 3
.2
. Suppose we insert a 0.1 Q resistance in series with ever)' capacitor. The resulting transform
analysis would produce a rational function with a cubic denominator polynomial (a third-order
network) whose factorization would require the use of a root-finding program. In sharp contrast,
the idealized circuit o f Example 13.20 was anaK'zed by writing a single first-order node equation,
making a partial fraction expansion unnecessary.
Idealizations o f circuit models sometimes lead to phenomena that defy intuitive explanations. An
interesting case is given by Example 13.20. Before S is closed, the energ)' stored in the electric field
is
0 .5
X 1^ + 0 .5 x 0 = 0.5 joule. After S is closed, the stored energ)' becomes 0 .5 x (0 .5 ) + 0.5 x
(0 .5 ) =
0 .2 5
joule. Apparently, 0 .2 5 joule o f energy has been lost. Since there is no resistance in the
circuit to dissipate the energ)', what accounts for the lost energ}'? Is energ)' not conser\'ed?
An explanation o f this paradox is as follows. Instead o f considering a zero-resistance circuit, place
a resistance R in series with all capacitances. Then analrze the circuit, and let R approach zero. The
result shows that no matter what value R takes on, the total energ)' dissipated in the resistance for
0 < / < ^ exactly equals the difference o f the total stored energies before and after the closing o f
the switch. This accounts for the apparent lost energ)'. In actuality, part o f the 0 .2 5 joule of cnerg) would be lost in the form o f radiated energ)'. However, the principles of field theor)- would be
necessar)' to explain the radiation phenomenon.
Note that for idealized SC circuits, as long as the independent voltage sources arc piecewise con
stant, all capacitor voltages are piecewise constant and all currents in the circuit are impulses. These
650
properties remain valid for more general idealized SC circuits that allow the inclusion o f VCVSs,
C C C Ss, and ideal op amps. T he reason is that the parameters characterizing these components are
instantaneous.
lossy switched capacitor circuit,
dimensionless and hence do not result in a time constant. All voltage changes are
On the other hand, if the circuit contains resistances, we have a
whose voltages arc no longer piecewise constant. The transient analysis o f a lossy SC circuit
requires the usual Laplace transform analysis.
approximate a
approximated by an
One reason for our interest in SC circuits is that a SC combination can be used to
resistor. As a result, any
RC-op
amp circuit used for signal processing can be
S C -o p amp circuit. A study o f the general theor)' o f such S C -o p amp circuits is beyond the level of
this book. We shall merely illustrate the approximation property with a simple integrator circuit.
E X A M P L E 13.21. Consider the RC-op amp integrator circuit shown in Figure 13.46.
R = 5kQ
C,= 1 mF
FIG U R E 13.46 Op amp integrator of Example 13.21.

Using the result o f Example 13.5 and the integration property o f the Laplace transform, it is
straightforward to show that
If the input is a constant voltage

function
vj,t)
v{ =
= 5 V, and if
= 0, then the output waveform is a ramp
= - l ,0 0 0 ///( f ) (as long as the output has not reached the saturation level), as shown
in Figure 13.47.
t (msec)
-
10-
F I G U I I E 1 3 .4 7 R am p o u tp u t o f th e in teg rato r in F igu re 1 3 .4 6 .
6SI
W h at we would like to do is construct an SC approximation to this circuit. T he idea is to replace

the resistor with a switch and a capacitor as in Figure 13.48.
F-IGLIRF, 13.48 Switched capacitor equivalent of Figure 13.46.

The switch S is operated in the following manner:
1.
At r = 0, S is at position a.
2.
At t = T, S is
3.
At r = 27", S is moved to position a.
4.
At / = 37", S is moved to position b, etc.
moved to position b.
T he output waveform may be determined ver)' easily by the principle o f conser\'ation o f charge as
follows: for 0 :s / <
T,
y.
= ,(]= CE, qj-= 0,
and
= 0. At / = f , switch S is moved to position
b. Because the op amp is assumed to be ideal, the voltage across the input terminals is zero, and
so is v^. Thus, C cannot store any charge. T he charge CE previously stored on C must be trans
ferred out o f C. Since the op amp is ideal, the input impedance is infinity and the input current
is zero. Therefore, none o f the charge can flow into the op amp. Instead, the charge must be trans
ferred to the capacitor
This leads to
CE and Vgi'T') = - C /C ^
At r = 27', switch S is moved back to position a, causing C to be charged to E volts again. Since
the charge
trapped on
Cj-, the output voltage
remains unchanged until S is moved to posi
tion b again. At that time, another CE coulombs o f charge are transferred to Cp and
is incre
mented by -C EIC p Subsequent switching is similar.

To make the output waveforms o f the circuits o f Figures 1 3 .4 6 and 13.48
the
average o f the
approximately the same,
charges transferred to (y m u s t be the same in both cases. For Figure 1 3 .4 6 , the
current flowing into Cîs at a constant value o f E/R. Therefore, every I T sec, the charge trans
ferred to Cîs equal to 2TEIR. On the other hand, for Figure 1 3 .4 8 , the charge transferred to Cjevery I T sec is CE. Equating these rwo quantities, we have CE = 2TEIR, or RC = IT . Thus, there
is no unique combination o f C and T that will produce the approximate effect o f a resistance. A
smaller T (i.e., a higher operating frequency o f the switch) in Figure 13.48 produces a staircase
output waveform that closely hugs the ramp output o f Figure 1 3 .4 7 . For the purpose o f com
parison, the output waveform corresponding to 7 = 1 msec and C = 0 .4 pF is shown in Figure
1 3 .4 9 , together with the ramp output from the RC-op amp integrator. It is worthwhile to note
Chapter 13 Laplace Transform Analysis
652
II: Circuit Applications
that each o f the circuits o f Figures 1 3 .4 6 and 1 3.48 drains the same average amount o f charge
from the voltage source and puts the same average amount of charge on the capacitor
difference is that in the former the process is
coutinuous,
Cj; The
only
whereas in the latter the process occurs
in quantized steps.
2 3 4 5 6 7 8 9
19
-I I I I I I I I I It (msec)
Output of
switched capacitor
circuit
FICURU 1 3 .4 9 Responses of circuits in Figures 13.46 and 13.48.
Exercise.
Plot the output waveform of the SC circuit o f Figure 1 3 .4 8 for 0 s
^ 2 msec if 7'= 0.1
msec and C = 0 .0 4 pF. Also, plot the ramp output on the same graph for comparison.
As recently as two decades ago, switched capacitor circuits were considered impractical. No longer!
Advances in semiconductor technology have made the fabrication o f high-speed electronic switch
es and op amps as cheap as resistors. Furthermore, large numbers of switches, capacitors, and op
amps can be fabricated on a single chip. Consequently, switched capacitor circuits hold an impor
tant place in future signal-processing applications. Although we cannot delve into the practical
aspects o f the design of such circuits, we have at least outlined the basic principles needed for their
approximate or exact analysis.
9. THE DESIGN OF GENERAL SUMMING INTEGRATORS

In the design o f active filters and in the design of data acquisition equipment, general summing
integrators play an important role. This section takes up the design o f such op amp circuits from
the transfer function perspective. However, to simplify the presentation, we consider only multi
input transfer functions o f the form
*oiit ( '0 -
(1 3 .3 3 )
where the a- and
are positive gain constants and
653
....
are ^-domain inputs. This
expression represents the ^-domain equivalent o f the following time domain equation assuming
zero initial conditions at ; = 0 on the variables:
'W
( 0
- 1 j j j 'V/1
+ ^
2 /,2(r)ciT
/^ V
W ith a little cleverness, it is possible to design by inspection a general integrating operational

amplifier circuit whose input-output characteristic is precisely equation 1 3 .3 3 . T he four-input op
amp circuit o f Figure 1 3 .5 0 accomplishes this. The dashed lines are present because the admit
tance may or may not be needed. Computation o f the values o f
and
are explained in
design step 2 below.
Design Choices for the General Summing Circuit o f Figure 13.50

The first two design steps constitute a preliminary or protorvpe design, meaning that the two
capacitors are normalized to 1 F. After completing the prototype design, an engineer would scale
the capacitances and resistances to more practical values without changing the gain characteristics.
T he scaling procedure is explained in step 3.
Design Step 1.
Prototype design.
Set C = 1 F and also set
S,
^2
^bl
= ^2 S.
In the derivation described later on, we conclude that a simplified design requires that the total
admittances incident on the inverting terminal equal the total admittance incident on the non
inverting terminal by proper choice o f AG^ or AG^. The proper choices are given in design step
2.
Design Step 2. Prototype design [continued): Computation o f
or AG^ such that the total admit

tance incident at the inverting terminal o f the op amp equals the total admittance incident at the non
inverting terminal. To achieve this equality, recall that in design step I >
- 1 S,
- 2 ^b\
= h, S. G ,, =
S. Define a numerical quantity 6 =
+ a-^) - [b^ + b.^)
6S4
Chapter 13 Laplacc Transform Analysis II; C ircuit Applications
W e distinguish two cases:
Case I: If 6 > 0 , set
= 6 and
Case 2:
= - 6 and AG^ = 0.
If 6 < 0 , set
= 0.
D esign Step 3 . Scaling to achieve practical element values. Multiply all the admittances incident at
the inverting input terminal o f the op amp by a constant
K^.
Similarly, multiply all admittances
incident at the non-inverting terminal o f the op amp by /Q,. It is possible to choose
= /Q,, but
this is not necessary.
E X A M P L E 1 3 .2 2 . Design an op amp circuit having the input-output relationship
(1 3 .3 4 )
S
o l u t io n
Step 1. Prototype design. Using Figure 1 3 .5 0 , choose C = 1 F. W ith all admittances in S, set
Step 2 . Equalization o f total admittances at inverting and non-inverting terminals. Since 6 = (7 + 3)

- (2 + 4) = 4 > 0, set
type design.
= 4 S and AG^ = 0. T he circuit in Figure 13.5 1 a exemplifies the proto
K^ = K y- 10^. This scaling leads

R^2 = 3 3 .3 3 kQ, /?^, = 50 kQ, R^2 =
Step 3 . Scaling. To have practical element values, let us choose

to a design with C = 10 j_iF and resistances
= 1 4 .2 8 kQ,
25 kQ, and A/?^ = 25 kQ.
75
6SS
1 4 .2 8 kQ
lO p F
a2 '
3 3 .3 3 kO
+
5 0 kO
b2
o
2 5 kO
I O mF
2 5 kO-
(b)
FIGURE 13.51 (^) Prototype design of equation 13.34. (b) Final design after scaling with /T = AT^ = 10~5.
Exercises.
1. Show that the normalized realization o f
is given by the circuit of Figure 13.52.
2S
6S
M GURE 13.52
2 . Scale the circu it o f Figure 1 3 .5 1 so that the capacitors b ecom e 1 uF.
656
Derivation of Op Amp Input-Output Characteristic

Referring to Figure 13.50, from the properties of an ideal op amp, V^= V_ and no current enters
the inverting and non-inverting op amp input terminals.
Derivation Step 1. Write a node equation at the non-inverting input terminal of the op amp.
Summing the currents leaving the + node of the op amp yields
Solving for
yields
W b2
(13.35)
Derivation Step 2. Write a node equation at the inverting input terminal of the op amp. Recall that
V_ = Vy Summing the currents leaving the inverting node yields

. AG^V^ = 0
- V^2) Thus,
sC
sC
''al
Gi
Gno +
AG +
(13.36)
sC
n\
Derivation Step 3 . Combine steps 1 and 2 to compute the general input-output relationship.
Substituting equation 13.35 into 13.36 yields
1
/
^ a l T,
G a2 y
I I ^ a\ + ^ a l +
"^ 7 )
If we choose
and AG^ to make AG^ + â\ + âl =
+ ^b\ + ^bV -e- if the total admit
tance incident on the inverting terminal is made equal to the total admittance incident on the
non-inverting terminal, then equation 13.37 simplifies to
I sC
(13.38)
sC
If we let C = 1 F (to be scaled later to a practical value), equation 13.38 further simplifies to
ôut
^b\ +
^bl
(13.39)
Equation 13.39 shows that the circuit of Figure 13.50 is a general summing integrating circuit
whose gains are proportional to the admittances G^- and/or % The sign o f each gain depends on
whether the corresponding input is connected to the inverting or non-inverting terminal o f the
op amp. This completes the derivation of the input-output characteristic of the op amp circuit and
is the basis for the prototype design. The extension to more than four inputs is straightforward.
Chapter 13 * Laplacc Transform Analysis 11: Circuit Applications
6^
It remains to justify the scaling o f step 3 in the design procedure. In step 3 , it is stated that all
admittances incident at the inverting input terminal o f the op amp can be multiplied by a (scal
ing) constant
K^.
Similarly, all admittances incident at the non-inverting terminal may be multi
plied by a (scaling) constant /Q,. It is possible to choose
= /Q,, but, again, this is not necessar}^
The verification that these multiplications will not change the input-output characteristic follows
directly from equation 13 .3 7 . After inserting the scale factors, one immediately sees that they can
cel and have no effect on the overall gain.
11. SUMMARY
This chapter has presented the basic principles and techniques o f circuit analysis in the j-domain.
Impedance, admittance, Thevenin equivalents, superposition, linearity, voltage division, current
division, source transformations, and nodal and loop analysis have ^-domain forms that allow the
analysis o f complex circuits excited by a variety o f waveform types. Indeed, the simple starter cir
cuit o f a fluorescent light points to the usefulness o f Laplace transform methods in the analysis
and design o f simple, everyday electrical conveniences. Various op amp applications were also pre
sented. As subsequent chapters will illustrate, complex circuits and advanced analysis methods
build on these basic principles and techniques.
T he chapter also introduced the notion o f switched capacitor circuits. Integrated circuit technol
og)' has made such circuits easy and inexpensive to produce. Applications include speech process
ing and other t\'pes o f signal processing. Although a full-scale analysis o f such circuits is beyond
the scope o f this text, the basic principles o f their operation arc presented as a foundation for more
advanced analysis tools.
Finally the chapter introduced a general method for the design o f multi-input integrators having
both positive and negative gains. Such circuits can be used for implementing active filters and for
implementing controllers in practical situations.
65<S
Chapter 13 * Lnplacc Transform Analysis II: Circuit Applications

Admittance: the ratio o f the Laplace transform o f the input current to the Laplace transform o f
the input voltage with the two-terminal network initially relaxed.
Admittance o f capacitor: the quantity Cs.

Admittance o f inductor: the quantity MLs.
Admittance o f resistor: the quantity MR.
Current division: a formula for determining how currents distribute through a set o f parallel
admittances.
Impedance: the ratio o f the Laplace transform o f the input voltage to the Laplace transform o f
the input current with the two-terminal network initially relaxed.
Impedance o f capacitor: the quantity MCs.

Impedance o f inductor: the quantit}' Ls.
Impedance o f resistor; the quantity R.
Impulse response: the response o f a circuit having a single input excitation o f a unit impulse;
equal to the inverse Laplace transform o f the transfer function.
Integrator circuit: usually an op amp circuit with transfer function K/s.

Source transformation property: in the ^-domain, voltage sources in series with an impedance
are equivalent to the same impedance in parallel with a current source whose value equals
the transform voltage divided by the series impedance.
Steady-state analysis: analysis o f circuit behavior resulting after excitations have been on for a
long time; often refers to finding the sinusoidal or constant parts o f the response when
the circuit is excited by sinusoids or dc.
Switched capacitor circuit: a circuit containing switches, capacitors, independent voltage

sources, and possibly op amps, but no resistors or inductors.
Time-invariant device: a device whose characteristics do not change with time.

Transfer function; the ratio o f the Laplace transform o f the output quantity to the Laplace trans
form o f the input quantity with the network initially relaxed.
Transient analysis: analysis o f circuit behavior for a period o f time immediately after independ
ent sources have been turned on.
Voltage division: a formula for determining how voltages distribute around a series connection
o f impedances.
6S9
Problems
IMPEDANCE, ADMITTANCE,
VOLTAGE DIVISION,
CURRENT DIVISION, SOURCE
TRANSFORMATIONS,
THEVENIN AND NORTON
EQUIVALENTS
ZJs)
(a)
/Y Y V ^
L
1. Consider the circuits in Figure P I 3 .1 .

(a)
{s) and Y-J^s) as
Find
rational func
,b)
tions for each o f the given circuits.

(b)
Figure P I3.2
Find the ^-domain and time domain

output responses to a step input for
each circuit.
3. In the (relaxed) circuit o f Figure PI 3 .3 , if

= (5(r), then
5 + 40
2.5 0
0.2 F
1H
Y-^^{s) and Z-^^{s).

= 10u{t) V, find
for t>Q.
v jt ) = 20e-^^û{t) V, find i j t ) for
(a)
C om pute
(b)
If
(c)
If
r > 0.
(a)
(d)
0.25 H
J T
If
v jt )
0.
20e~^K(t)
V, find
ijt)
for
ijt)
0.25 F
Z(s)
v.(t) I
20
Figure P I3.3
(b,
Figure P I3.1
C H E C K S : Figure P I 3 . la:
+5
y,(5) = 0 . 4 --------- ^ F i g u r e P I 3 . lb:
( .y + l ) -+ 2 "
4. For each o f the circuits in Figure PI 3 .4 , find

expressions for
(a)
Z Js) and YJs)

and
(as indicated)
J" + 1 6
2. Find the input impedance and admittance in
terms o f
R, L,
and C o f each o f the networks
shown in Figure P I 3 .2 . Then find /-domain

expressions for the output in terms o f the input
for each circuit.
Z(s)
(a)
660
7. Repeat Problem 6 for /? = 6.25
C = 0.01
F, and v-J^t) = 32^^^cos(4r)(/) V.

8. For the circuit of Figure P 13.8, suppose R =
2 il, I = 4 H, and Z{s) = ^ 1 5 . Find Z -(j),
j + 0.5
Voup\
and
when
v-J,t) = 2Qe~^û{t) V, assuming zero
(a)
initial conditions
v-J^t) = 2Ae~^-^û{t) V, assuming zero
(b)
initial conditions
(0
r\
Figure P I3.4
r\
5. In the circuit o f Figure P13.5, suppose
) = 0. Find
and i^^t)
. Z.W.
= 5{t) and (b)
when (a)
50(1 -
n\
mA.
Z.(s)
Y Js)
Figure P I3.8
0.4 F
9. For the circuit of Figure P 13.9,
^m\ ~
^ml ~
= 100 12,
^2
= 1 mF.
Figure P i3.5
(a)
Compute I 2 U) in terms of /,(j) and
6. In the circuit o f Figure PI 3.6a, 7? = 4 0 Q and
(b)
If v-JJ) = e^^û{t) V, compute /j(/).
(c)
If v^{t) =
then compute Z-J^s) and Y-Jj) .

C = 0.01 F. Find Z.W, Y.J,s\ V^s), and
for (a) v-J^t) = (5(r) and (b) v-J^t) as given in
V, compute /j(r).
l,(s)
Figure PI3.6b.
+
JL
l,(s)
R.
V^(t)
(a)
Figure P i3.9
t vJOCV)
20-
C H EC K : Z;(5) = - 1 0 0
5-10
5 + 10
101
10. (a)
(b)
Figure P I3.6
(b)
Find the input impedance of the cir

cuit o f Figure PI 3 .10a.
Consider Figure P I 3 .10b. Use your
answer from part (a) to compute the
responses i-JJ) and
to a step
0(11
input when /? = 3 0 H , Z,|
= Lj=
1 H,
iY Y V ,
and C = 10 mF.
0.1 H
10 pF
I (s)
vjt)
Z,(s)
50 0
K,V,(s)
2,(s)
Y,(s)o-
V,{s)
(b)
Figure P I3.12
13.
Vj(s)
For the bridged-T network o f Figure
P I 3 .1 3 , show that if Z jU )Z ,(j) = i?", then

' V,(s)
^ , w (b)
Z,(s)
Figure P I3.1 0
11. Find theThevenin equivalent impedance o f
z (S)
Z,(s)
each circuit in Figure P 1 3 .1 1 . Flint: Label the

terminals
and an input current to the top
terminal as
determine
or
0.5 H
_ m
v _
Figure P I3 .13
14. Consider the circuit o f Figure P 1 3 .1 4 , in
which
= 10
7^2 = 10 Q , Z = 0.1 H, and C
= 1 mR
0.5 Q
.(i)
(a)
(b)
Find
and
If a voltage source is attached as indi

cated and
(a)
Ifl(S)
(c)
2H
V, find
capacitor are initially relaxed.

(d)
(b)
Find the zero-input response if /^(O) =

0 and
-o
i,(t)
2 + 2.V +
iind
v^{t) =
assuming the inductor and
Figure P I3.11
12. Find
V, find /^(r)
If a voltage source is attached as indi

cated and
0.5 F
o-
ANSWF.R: (h)
are initially relaxed.
o- J \ / v 'n_T Y Y \ .
20
vî)
assuming the inductor and capacitor
for each o f the cir
cuits shown in Figure P i 3 .1 2 .
= 10 V.
R,
/Y Y V
L
'"Q
YJs)
Z,(s)
Figure P i 3 .1 4
Z,(s)
CH ECK:
K, = 1 0
.y -h 200.V + ? ? ? ?
(a)
662
Chapter 13 * Liplace Transform Analysis II: Circuit Applications
15. Consider the circuit o f Figure P i 3 .1 5 , in

which /^, = 16
y? 2 = 4 n , Z, = 1 H, and C =
0 .1 2 5 5 ^ + 4
(c)
Z /.(-v ) =
(d)
yM =-
2 mF.
0 .1 2 5 5 - + 4
Y.î,s) and
(a)
Find
(b)
Find an expression for

of
0 .5 5
0.5 s
in terms
and /.W .
(c)
505
zM
5 -+ 2 5 0
55
R.
(0
R,
ijs) ( t )
V.Js)
(g)
(h)
yM =
5 + 2 5 0
5 + 4
0 .2 5 5 + 4
Z ,v ,(j) = ----------+
0 .5 5
is
5^ + 16
n > ,(^ )=
45
+
'0 .2 5 5 ^ + 16
Figure P I3 . 15
TRANSFER FUNCTIONS , ,
16. Find an
RC circuit
for each o f the follow
ing:
(a)
Z:,As) = 0 .2 5
(c)
Z , ( 5 - ) = 1 0 +
+ 0 .0 2 5
(b) iv,,(/) + 20v,, , ( / ) + 100vv,,(r) = 2 0 0 v v (0 + 10v,(O
Z M
('= )
s+ 2
4s~ +
2 0 . Find the transfer function

36.? + 6 4
RL circuit for each

Z Js) = OAs^ 10
o f the following:
0 .1 5 + 1 .5
2 1 . Find the transfer function,
V
H{s) = - ^
(t) + bî-'y,, (/)

(5 )
-----
differential equations accounting for initial

conditions?
(a)
5+10
5+2
( 0 + 1 Vaut ( 0 + 2 ôut ( 0 = /^2în( 0 + /^|
models o f a physical process. In the case o f (b)
Z , (5 ) = 0 .2 5 + 2 0 +
for the general
and (c), what is the Laplace transform of the

2 0 + 0 .0 4 .V
LC circuit for eac i of- the following:

Zi(s) = 0.05s + ------
^
+ 25
( ;) + 1
5 v / (r)-1 0 jJ.
5+10
5+8
your ( 0
(b)
(c)
20
5
(T)dr =
Vi{T)(h
iaut ( 0
+ 0.1
dinutit)
+ 4 2 .5
r,v,(5) = 0.025 +
+40 J'/^,,,( t Wt = 1 0 v,C/)
18. Find an
(b)
v ^
second-order differential equation
for each o f the following differential equation
5 + 240
V f \
2 2 5 0 V ,. , , ( 0 + l O
impulse response, and step response
70
yM =
v , ( / ) + 1 0 v ,(/ ) + 2 5 .. , (/ ) =
2.V + 8
17. Find an
(a)
( A")
-----,
models o f a physical process.

(a) <o,a(0+ P\Voui(0= KviU)
(5 + 2)(.y + 4 )
(c)
for each o f the following differential equation
0 .2 5 5 + 0 .2
(a)
H{s) =
impulse response, and step response

K J j) = 0 .b + 0 .1
(b)
(a)
19. Find the transfer function,
J
00
= 200V;(/)+20i:';(/)
{T )dT
663
IUU.V
2 6. (a)
.s- + 400
0
(b)
+ 40^^^^
- 0 . 1 / , ( 0 " ) + 4()
(c)
.V
2 2 , An integro-difFerential equation
Given your answers to part (a), com

pute the step responses.
= 1 0 l
.V
Compute the transfer flinctions o f the

op-amp circuit in Figures P i3 .2 6 a and b.
for an
R=
Assuming
1 kH and C = 0 .5 niF,
compute the zero-state response to the
active circuit driven by a current source with C
2u{t) - 4u (t- 2)
input
= 0 .5 F , Z = 1 H, ^ = 2 n i s
(d)
V.
For the circuit o f Figure P I3 .2 6 b ,

assuming
R=
1 kD and C = 0 .5 mF,
compute the output when the input is

-X
as given in Figure P i3 .2 6 c .
(a)
Find the transfer function ^/(.y) =
(b)
Find the impulse response.
(c)
Find the step response.
.
%
He
c
2R
V (t)
2 3. Suppose two signals
y{t)
and
are
related by the equations

(a)
ôut
--- r ; +
dt~
) + 2y(/) = 4 ^
at
He-
and
2Vo,(0 - fy(T)c/T = 2 J \,(T)i/T

()
(a)
+ /
Assuming t/y^(0 ) = 0, find the transfer

function
(b)
v.(t)
O'
He
N(s) =
(b)
v(t)
Find the impulse and step responses.
2 4. A certain circuit has input
co^{t)u{t)
te~ii{t).
C onstruct the
and output
- t
transfer function o f the circuit, assuming that
the circuit had no internal stored energ)- at r =
(0
0~ . Then compute the step response.
Figure P i 3.26
2 5. T he input to a relaxed (no initial condi
e~u{t)
tions) linear active circuit is
V.
2 7 .(a)
Find the transfer function o f the
T he response is measured in volts as
(ideal)
\0e~ s'm{2t)u{t).
P I3 .2 7 a in terms o f 7?,,
(a)
op
amp
circuit
Find the transfer function
it is desired to obtain
with C = 1 IIF, then
Vl{s)
d{s)
tored form.
Figure
and C . If
H{s) =
-2
^
' ^
(b)
Find the response to the input func

tion in Figure P I3 .2 7 b assuming zero
Com pute the response o f the circuit to

the new input y,(f)=
ol
find /?j and /?->.
in which the denominator is in fac

(b)
R-,,
2te~'u{t)
V.
for the initial capacitor voltage.
664
r\
(c)
Repeat part (b) assuming the initial

capacitor voltage is
= -4 V.
-
29. (a)
Find the transfer fimaion of the (ideal)

op amp circuit of Figure PI 3.29 in terms
of /?,, /?2, Cp and Cj. Make the leading
coefficient of s in the denominator 1.
(b)
If it is desired to obtain H(s) = - 4 -------
(c)
with 2 = 100 jaF, find /?j, Rj, and Cj.

j +2
Given H(s) = - 4 -------find the zero-
v,(t) +
5+ 2
s+4
^^
^^
s+4
state response to t',(^)= sin(2r)(f) V.
v,(t)
->t
(b)
Figure P I 3.2 7
28.(a)
(b)
(c)
(d)

(ideal) op amp circuit o f Figure
P I3.28 in terms of /?,, /?,, and C.
s 8
If it is desired to obtain H (j;) = -----with C = 1 nF, find /?, and i?2- ^ ^ ^
Find the response to the input func
tion in Figure P I3.27b assuming zero
for the initial capacitor voltage.
Repeat part (b) assuming the initial
capacitor voltage is
= - 4 V.
30. (a)

(ideal) op amp circuit o f Figure
P I3.30 in terms of R^, Gj, C j, and
C2 .
(b)
If it is desired to obtain
//M = - 4
(c)
(s + 2)(5 + 10)
with 2 = 100 ^F, find 7?j, /?2and C,.
Given the answer to part (b), find the
zero-state response to y,(^)= -2.5e~^^
sin(2/)(/) V.
Figure P I3.30
Figure P I3.28
Gi + G t
s + ^------ ^
CHECK:
H(s) = -------- f
s+
CHECK:
H(s) =
-1
C 2R 1
s+
s+
/ ? ,C j
R 2C 2
665
Chapter 13 Liplacc rransform Analysis II: Circuit Applications
3 1. (a)
C om pute the transfer functions o f

each circuit in Figure P I 3.31 in terms
of
(b)
C,
Suppose
R and K.
C = 1 i.i, K = 3,
3 3 . C onstruct the transfer Kinction o f the d rcuit o f Figure P 1 3 .3 3 , assuming that all op
amps are ideal, as follows:
and
R = R^
= 1 k n . Find the zero-state response to
(a)
Let G| = , C t = , G t, =
(b)
Com pute the ratio
sin(500r)//(r) for each circuit
R.
in Figure P 1 3 .3 1 .
(c)
Repeat part (h) for
ij{t)=
2 sin(l 500/)//(r).
V'i(.v)
(c)
Com pute the ratio
(d)
Com pute
(c)
Com pute
(0
Rack-substitute to eliminate interme
V'2(-v)
(a)
in terms o f
V:^{s)
and
in terms of V^|(i) and
diate variables and find
H{s)
Vi,M)
circuit o f Figure P I 3 .3 2 in terms

o f /?j,
(b)
R-y, Ry R^,
R-y, Ry
H{s) =
(c)
C p and
C-,.
Figure P i 3.33
If Cj = C 2 = 1 t<F, find values for /?j,

and
R^ so
that
ANSW HR:
-5
-2
(.v+lOO)
(.v + 2 0 0 )
3 4
. (a)
Find the impulse and step response.
H(s) =
Ci\
s~ + (j^s +
G -1
For the circuit in Figure P i 3 .3 4 , find

the transfer functions
(b)
H^{s) = - ! - ^ and //^ (.v) =

Vinis)
Vinis)
If
= 10 Q , Yj = 0.1 S, Z , = /, V, =
s, and
= 2 0 S, com pute the impulse
and step responses.
(c)
Given the values o f part (b), compute

the response to the input
-2u{t
2u{t)
- 4) V using linearit)^ and no
further calculations.
666
IJs)
2,(s)
+
Z.(s)
Y,(s)
VJs)
Y,(s)
9 jY
Figure P i3.34
35. (a)
Figure P13.36
For the circuit in Figure P13.35, find

die transfer fiinctions
w
Z,(s) V,(s)
< p
37 Consider the circuit o f Figure P13.37. Find

die transfer fiinctions
(b)
(c)
If Z 2 =
,Z^ = j + 3, Y^=s, Y^=
3. and r ^ = 20 Q, compute the

impulse and step responses.
Given the values of part (b), compute
the response to the input
Kû(t) -Kjuit 7) V using linearity
and no further calculations, assuming
as follows:
(a) Compute the input impedance
_ Vîs)
Zin(s) =
(b)
T>0.
VJ(j)
i(^)
Given the results from part (a), use

voltage division to compute
Vln(s)
Figure P I3.35
36. (a)
(d)
in terms ofZy^(^), J 2W>

Compute ^
(e)
Compute
Find the transfer functions

lM = ^ . 2 W
Hîs) =
(b)
(c)
in terms of
Do not substitute
your answer to part (a) into the
obtained expression.
Compute
v ,w
= ^ .a n d
for the circuit of
^tc.
V ,( 5 )
Figure P I3.36.
Vinis)
Vi(s)
Find the impulse and step responses
and
as
products
of
your
prior
answers.
associated with each transfer
function if
= 8, Kj = 2,
IJs)
s
, _
1
z.(s) [-0
Z,{s)
Z, =
and Z 2 =
j + 5
<$>
Yj(s) V,(s) Y,(s)
K.IJS)
Figure P I3.37
r\
667
CHECK: Z ^ = Z i +
K 1^2
.K
n + ^L
Vout = det(MM)/det(M)
% You have now computed the symbolic
% expression for Vout in terms o f lin so you
% can now identify the transfer function as a
% symbolic ftinction o f the variable s.
% Now lets do some numerical work.
C l= 1; C2 = 2; G l = 1; R2 = 1; G3 = 2; a = 0 .5 ;
vJt)
% Re-enter M above
M = [Cl'*s+Gl 0 1; fill in the rest of your co ^ -
G is mhos; R is ohms
Figure P I3.38
(a)
(b)
Draw the equivalent frequency

domain circuit assuming t'QCO") ^ 0
V and ^^2(0") ^ 0 V.
Define a current Ij^{s) from left to
right through
Then write three
(modified) nodal equations for the cir
cuit o f part (a) similar to the text
example, but accounting for the initial
conditions. Put equations in matrix
form (unknowns on the left side and
knowns, input, and initial conditions
on the right). You should have a 3 by
3 matrix of literals. The unknowns are
two (node) voltages and one current,
cient fnatrix]
% check: the roots o f the determinant should
% be -1 and - 5 .
dt = det(M)
factor(dt)
% Re-enter MM above
% Compute the actual transfer function
Vout = det(MM)/det(M)
% Identify the transfer function
(d)
Find the impulse response h(t) of the

circuit as follows:
% Rewrite the above expression for Vout with
% out lin.
syms s t H h
H = -(1/2V1)/????????
h = ilaplace(H)
Determine the transfer ftinction H{s)

= yôut^s)/IjJs) of the circuit in MATLAB using the following code:
(e)
Now use the MATLAB command

ilaplace to compute the step
response, again with zero initial condi
tions.
syms M MM lin Vout s G l C l C 2 G3 G2 R2
Remark: Use your newfound MATLAB knowl
a t,H
edge o f symbolic computation to carry out the

following tasks. Hand computation is not
acceptable.
(f) Find the response due only to the ini
tial condition on Cj assuming " a (0 )
= 16 V.
(g) Find the response due only to the ini
tial condition on C2 assuming
(c)
M = [C l*s+G l 0 1; fill in the rest of your co ^ -
cient matrb^
'w ^
% Transfer ftinction computation assumes all

% ICs are zero
% Using Cramer s rule, define MM by
% replacing the middle column o f M
% by the right-hand side o f your matrix nodal
% equation with all IC s set to zero.
M M = [C l*s+G l lin 1; fill in the rest of your
matrii^
dt = det(M)
dt = coUea(dt)
= -8 V .
(h)
Write down the complete response

assuming the input is a step.
66S
3 9 . In the so-called feedback configuration o f

Figure P I 3 .3 9 ,
E{s)
is the Laplace transform o f
the error between the reference signal
e{t) =
the response^(^), i.e.,
}i{s)ld{s)
and
-y{t).
G, W =
is called the plant o f tiie system and rep
resents a physical process or special circuit.

(a)
(b)
e{t)
x^^,j{t) = Kfû{t)'
conditions does e{t)
00
Under what
0 as / -
(d)
XrefU)
CH ECK :
Under what conditions does

0 as f -
(c)
H{s) =
F ig u re P I 3 . 4 0
E{s)
00
where F (s)G i(.s ) = -------
djis)cl^(s)
from
E{s)
H{s)X^^jsy.
d{s) + n{s)
= --------
d(s)
RESPONSE CALCULATION
WITH INITIAL CONDITIONS
4 1 . For the circuit o f Figure P i 3 .4 1 ,
C = 0.1 F,
E(s)
1 + F(.v)Gi(.9)
when
Under what conditions and how could
_ d{s)Xr(.f{s)
Xrefis)
when
you determ ine
X..,(s)
E{s) =
-> Y(s)
G,(s)
V(is)
and
initialized
L = 0 .5
H,
i^{0~) = 2 A, and vîO~) = 2 V. Find

vît) using the et]uivalent models for
L and C in the j-domain.
ijt)
Figure P I3.3 9
ch ec k
+
L
l-t-G](5)
d{s) + n{s)
4 0. In the so-called feedback configuration of

Figure P i 3 .4 0 ,
E{s)
nj,s)ldjj)
x^^t)
- y{t).
i.e.,
e{t) =
and
=
is the plant o f the system (described
in the previous problem) and
F{s)
)tj(s)lci^s)
is
a feedback controller to be designed.

(a)
00
when
that accounts for the initial condition, write a

single node equation in
0
R1
x^^J(t) =
e{t) -*
0.4 H
A'y(^)?
0-2 v , Y
2Ci
Under what conditions and how could

you determ ine
e(0'^)
from
E{s) =
F igu re P i 3 .4 2
A N S \ V F :R :
(e)
and then find
50
e{t) -*
x^^Jit) =
^ 0 as r - X when
(d)
A. Draw an equivalent circuit for the inductor
H{s) = E{s)IX^^,J(s).

as r
(c)
4 2 . For the circuit o f Figure P 1 3 .4 2 , /^(0) = 2
Find the transfer

function
(b)
Figure P I3.41
is the Laplace transform of
the error between the reference signal

the response
v,(t)
Suppose
d^s)
had a pair o f poles on
the imaginary axis. How could you

design an
F{s)
to cancel these poles?
669
43. For the circuit o f Figure P I 3.43, assume

that i'q ( 0 ) = -4 0 0 mV and V(^{0~) = 0 V.
Find
t> 0.
2mF
CHECK: /q = 2.5
46. In the circuit o f Figure P I 3.46, 7? = 20 Q,
C = 0.2 F, and the capacitor is initially charged
at V(^Qr) = 10 V.
(a) If v.{t) = 20[(f) - u{t - 20)] V, find
I({s), V^s)y if^t) , and vîj) for r > 0.
Plot V(jJ) and v-J^t) on the same graph
for 0 < f < 40 sec using MATLAB or
its equivalent.
(b) If
V, find i({t)
and V(4J^ for ^> 0. Plot v^{t) and v-JJ)
on the same graph for 0 < ? < 20 sec
using MATLAB or its equivalent.
44. Consider the circuit o f Figure P I3.44, in

which C = 20 mF and R = 100 O.
(a) I f vÔ-) = 10 V and
=
200e~^'^û{t) mA, find Vf^s) and V(^t)
for f > 0.
= 200
(b) I f v^0~) = 10 V and
cos(2f)(/) mA, find V(^s) and V(4J)
for ^ > 0.
ic(t)
+
Vc(t)
Figure P I3.46
47. Consider the circuit of Figure P I3.47, in
which R^=l k2, R^ = S kH, and C = 50 ^F.
(a) Find the transfer fiinction H{s).
(b) Compute the step and impulse
responses.
Compute the response to v-^(t) =
(c)
Figure P I3.44
(d)
45. Consider the circuit of Figure P13.45, in
which R = 200 2.
(a) Compute the transfer function in
terms o f L.
(b) If the response to the input i-JJ) =
Intuit) A is ijij) = {2.5t - 0.025 +
Vw>
(c)
O.O25e"^0(^) A, assuming /^(O") =

0, find the values of L and /q.
Assuming ij{0~) = 1 A and i-JJ) =
2.5 sin(100f)W A, find
(e)
(0
10r'2-5^cos(25^)W V.
Find the response to V(iO~) = 20 V.
Using the principle of superposition,
find the response to the input of pan (c)
with the initial condition o f part (d).
From the principle o f linearity, what
would the response be to the input
= 20-^^-^^cos(12.5/)(/) V and
y j^ 0-) = 1 0 V ?
+
v Jt)
lV (t)
Figure P I3.47
u (t)
C H EC K :
Figure P I3.45
H{s) =
20
2 5 + 25
670
48. Consider the circuit o f Figure P I3.48, in

which C = 2.5 mF,
= 200 Q,
= 50 2, /?
= 5 n , Z = 0 .2 H , and^ = 0 .5 S .
(a) Find the transfer functions
(c)
Find the response i({t) when v-JJ) =

0, V({Q) = 10 V, and /^(O) = 0.
//2 (5 ) = - ^ ^ , a n d
VcisV
Figure P I3.49
W
Given N(s),
(b)
find the zero-state

response; i.e., assume /^(0) = 0 and
= 0 when
= 25(1 V.
Find the zero-input response
if
/^(0) = 6 A a n d t;c(0 ) = -1 0 V .
Find the complete response (combine
parts (b) and (c)) and identify the
transient and steady-state parts o f it.
(Note that the transient pan is the
part that is not constant or not period
ic. Usually the transient part converges
to zero.)
(c)
(d)
'
g.Vc(s) 1
C = -
9
R^ = 0.5 2,
= 5 2, C2 = 0.05 F, and a = 4 .
Compute Vciis),
assuming diat
v jt ) = 36u{t) V. /^(O-) = 0. v c ^ m = - 1 8 V,
and vufiO~) = 0 V. (Hint; You must construct
the equivalent circuit in the j-domain, account
ing for initial conditions. Consider a source
transformation on v-J^t), and then draw the
equivalent circuit in the j-domain so that you
can combine sources in the front half o f the cir
cuit.)
1 L
'1
50. In the circuit o f Figure P I 3.50, suppose L

= 0.04 H.
1
1
Figure PI3.50
Figure P I3.48
49. Consider the circuit of Figure P 13.49. Note
that the computation o f transfer functions pre
sumes no initial internal stored energy.
(a) Use a source transformation and the
current divider formula to show that
the transfer function between V-JJ)
(the input) and I(\s) (the output) is
1
s +
RC
(b)
1
LC
If Z = 0.2 H, C = 25 mF, R = 4/3 Q,

and v-^{t) = \Qe~^^û{t) V, find /^j)
and i({t).

Suppose /?j = 2 2, /?2 = 6 2, i?3 = 3 2,
C = 0.125 F I = 1.6 H.
Suppose v-JJ) = yj(/) + 1/2W = -1 0 (-f) +
10 u (t) V (plot this input function so that you
know what it looks like).
(a) Compute V(iO~) and /^(O").
(b) Compute the zero-input response
for ^> 0; i.e., assume you have the ini
tial conditions computed in step (a)
and that the voltage v-^(t) = 0 for r > 0.
First draw the equivalent j-domain cir
cuit accounting for the initial condi
tions.
(c)
(d)
(e)
(0
Compute the transfer function H{s) of

the circuit valid for ^ > 0.
Compute the zero-state response,
i.e., the response to the circuit o f the
input v-JJ) = V2 {t) = 10(f) assuming
all initial conditions are zero.
Compute the complete response.
Identify the transient and steady-state
parts of the complete response.
671
Plot the resulting time function in

MATLAB for 0 < r < 1 sec. The partial
fraction expansion is most easily com
puted using MATLABs residue
command. {Note: Practicing hand cal
culation is important for the exams.)
Now forget about the partial fraction
expansion and instead use MATLAB s
command ilaplace to compute the
time function
You should define
H, Vc, s, t, lin, and vc as symbols
using syms. The program should be
something like the following:
R = ? ; L = 1 ; C = ?;

Suppose Z = 1 H,
= 15
C = 0.01 F, and
^2 = 5 Q. Suppose
is the desired output,
(a) Find the input admittance
(b)
in terms of literds and then substitute

numbers. Combine terms to form a
rational function.
Find the input impedance
syms Yin Zin H Vc s t Is vc

lin = ?
% H will be the transfer function defined in
terms o f s, a symbol, and R, L, C, and K.
% MATLAB will fill in the numbers.
H = ????
Vc = H*Iin
vc = ilaplace(Vc)
.( .) 0
99
L
's ^
(c)
(d)
V^
(e)
V^
5+
1
LC
in which case V^(j) = ????(write down

the formula).
Use voltage division to express K^j)in
terms o f V^{s) and then in terms o f
Z-JJ) and IfJ). Hence determine the
transfer function H^{s) = V(^s)IV^{s).
The leading coefficient o f the denom
inator polynomial is to be 1.
Compute the impulse and step
responses.
Suppose the input is
= 2e~^û{i) A.
Find the partial fi^ction expansion of
V^s) and then compute V({t) assum
ing zero initial conditions on L and C.
Figure P I3.52
53. Repeat Problem 52 using ij{i) as the
desired output.
54. Consider the circuit in Figure P13.54, in
which /?j = /?2 =
Suppose V(\t) is the desired
output.
(a) Find the input impedance in the form
Z /rt(j)
in terms
poles />j
negative
ances in
S-p\
S -P 2
of R, L, and C. Identify the

and p 2 - They should have
values. Recall that imped
series add and two imped-
672
r>
(b)
ances in parallel satisfy the product

over sum rule.
Compute the transfer function
+
V^(t)
99
V,(s)
(c)
id)
(e)
i,(t)
RC
LC
in terms of R, L, and C
The roots of the characteristic equa
tion (denominator o f H{s)) are to be at
- 2 y4. If Z = 0.2 H, determine R and
C. Then specify the transfer function
with the proper numerical coeffi
cients.
Given the values o f part (c),
(i) Compute the impulse response of
the circuit.
(ii) Compute the step response of the
circuit.
Suppose the input is vfi) = 10aT*%(/)
A- Find the partial fraction expansion
o f V({s) using MATLABs residue
command. Now forget about the par
tial fraction expansion and instead use
MATLABs command ilaplace to
compute the time function vît).
Note you should define H, Vc, s, t,
Vin, and vc as symbols using syms.
For example,
R = ?; L = 0.2; C = ?;
K = 1/(R"C)
syms H Vc s t Vin vc
Vin = 10/(s+ 10)^2
% H will be the transfer function defined in
terms o f s, a symbol, and R, L, C, and K.
% MATLAB will fill in the numbers.
H = ????
Vc = H^Vin
vc = ilaplace(Vc)
Figure P 13.54
r\
55. This problem is to be solved using nodal

analysis. Consider the circuit o f Figure P I 3.55,
in which Gj = (^2 = 1 S, G3 = 4 S, Z = 1 H, C
= 2 F, and g^ = 2 S .
(a) Draw the equivalent frequency
domain circuit assuming /^(0) = 2 A
and
= 2 V.Since you
are using nodal analysis, what type of
equivalent circuit might be best to use
to account for initial conditions?
(b) Write two nodal equations for the cir
cuit o f part (a) in terms o f the node
voltages V.J,s) and
and, of
course, the input I-JJ) and initial con
ditions. Simplify. Put equations in
matrix form.
Determine the transfer function of the
(c)
circuit. Use Cramers rule.
(d) Find the step response of the circuit.
(e)
tial condition on the inductor.
(0 Find the response due only to the ini
tial condition on the capacitor.
(g) Find the complete response if the
input is a step.
r\
n
o
r\
r>
r>
r\
r\
Figure P I3.55
56.(a)
(b)
For the circuit o f Figure P I3.56, cal

culate the transfer function.
Suppose i-Jit) = <5(/). Find and plot
r-s
n\
673
(c)
Now suppose
i-p) = 16[u{t) - u {t- T)]
mA, where 7'=
1 0
If v jt) = l[u{t)
(d)
msec. Find and plot
0.3 H
- u { t- 0.5)]
V, y^(0") =
\/ and /yr(0~) = 0, find
0.1 H
ijt) 0 1 5 0
90 n
lO Q
Figure P I3 .56
Figure P I3.58
C H E C K : (a)
{s +
100)(5 + 3 0 0 )/[(j + 200)(^ +
4 0 0 )]
59. in the circuit o f Figure P 1 3 .5 9 , /?j = 2 Q,

C, =1 F, /?, = 1.75
57. Consider the bridged-'F network in Figure
(a)
Vj{t)
= 10(1 -
e--îi{t)
V({s)
and C , = 2 /7 F.
Show that the transfer function o f the

_ ^C2
circuit in literal form is H (.v) =
P I 3 .5 7 . Let C = 0 .2 5 F. Assuming no initial

conditions, find /j(s ), ^ (s ), and
Q,
when
V. You might use
R\C,
C ram ers rule and the symbolic toolbox in

M A FLA B to solve the problem. Now find
.v +
U 2C2 ^
"JO (b)
If
|C|/?2C ,
= 15(/) V and the capacitors
are initially relaxed, find
V(^[t)
for
t>
0. Plot the input and the response on

the same graph using M ATLAB or the
equivalent.
(c)
Z Js)
(d)
Figure P I 3
= 0, ^^^(0 ) = 0 , and i^Q(0~)
(e)
which /?^ = 2 k n , C = 2 0
80 n , and
}d,
If
= 20
Q,
If
for
t>
0.
= 0 , t^<^'i(0 ) = 0, and /^ ^ (0 )
= 15 V, find
.5 7
V(^{t)
for r > 0.
= 15//(/) V, i^Q(0 ) = 15 V, and
^(^(0 ) = 15 V, find
58. Consider the circuit o f Figure P I 3 .5 8 , in
(a)
if
= 15 V, find
Vf^{t)
for r > 0.
=5 H.
Com pute the transfer function

/V,(.v) =
(b)
v;,,(.v)
(c)
Vf(.v)
0
Figure P i3.59
A N SW ER S: (b)
y.
Vi Js)
- ~e^ûu) V; (c,[^-0.25;^
V; (e) sum ol parts (b). (c). and (d)
674
60. In the circuit o f Figure PI 3 .6 0 , /?, = 0.5
Rj
=1 H,
(a)
= 1.75 a
(c)
and Z , = 7 /8 H.

circuit in literal form is
Now suppose
A,
n {s) =
) = 3 V,
?>u{t) V,
R ecom pute
the
v,(t)
L jG2 ^2
+
>v+
v
-> 0 .1 H
0.5 n
= \5u{t)
If
V and the inductors

for
the same graph using M ATLAB or the

equivalent.
If
6 2 . Reconsider the circuit o f Figure Pi 3 .6 1 .

(a)
If
(e)
If
= 0 , /^ ,( 0 -) = 0 , and
1 5 A, find
=
an equivalent circuit accounting for

the initial conditions
mesh equations w'ith
V, /^ ,( 0 -) = 15 A,
and /^-,(0 ) = 15A, find
iu^t)
for
t>0.
/Y Y \
0,
/^(0 ) = 3 A. Com pute the associated
lor / > 0.
\5u{t)
Apply a source transformation to the

independent current source and draw
= 0, /^.(O -) = 0, and /> ,(0 " ) =
15 A, find /-,(/) for r > 0.

(d)
( t ) ' . ,(t)
1Q
Figure P I3.61
t^
0. Plot the input and the response on
(c)
Vjft).
^2^1/
are initially relaxed, find
equations
V .(t)
0.8 F
(b)
nodal
i^{Q) ) = 0
= 5u{t) A.
accounting for the initial capacitor
voltage. Then find
- +
^C|L|
and
/^(s) as one o f the
mesh currents.
(b)
If rîO-) =
3iiU) V, and
0,
i^(O-)
= 3 A,
/;.(/) = 3 (/) A, find /^(r).
6 3 . In the circuit shown in Figure P I 3 .6 3 ,
i\U)
= 12 V for r > 0 , i/^ ,(0 ) = 6 V, and z^^,(0) = 2 V.
' 6
(a)
C onstruct the equivalent circuit in the

Laplace transform domain, account
ing for an initializ-ed capacitor.
Figure P i3 .60
(b)
Write a nodal equation for the circuit

constructed in part (a).
A N SW FR S: (b) (30 (c)
A:
(c)
Find l/^ W .
(d)
Find
A; (d)
7e^\u{t) .A;
6 1 .(a)
for
t> 0 .
(e) sum o f parts(b). (c). and (d).
In the circuit o f Figure P 1 3 .6 1 ,

and
V(^t)
represent node voltages with
respect to the bottom/reference node.

Show that with zero initial conditions
the nodal equations are
O . B r + 2.9 + 10
-1 0
(b)
- 1 0 V'c'
2\ /,-
-h i
.v + 1 0
Find the response
with /.ero ini
tial conditions and with % ](/) = 3//(r)

V and
i^2 ^t) = 'iu{t)
A.
Figure P i 3.63
CH ECK:
form
should contain terms o f the

and
K~)U{t).
6 4 . Consider the third-order RC circuit o f

Figure P I 3 .6 4 . Suppose that the initial capaci-
tance voltages are t/c,(0) = 0, t/Q(0) = 6 V,
675
equations for the circuit of part (a)

only in terms of the variables V({s)y
*;C3(0) = 2 V .
(a)
Construct the Laplace transform

domain equivalent circuit.
(b) Find Vf^{s) in terms o f V^J^s) and the
initial conditions.
(c) For the given initial conditions, if
v-^{t) = \2u{t) V, find
for r > 0.
Hint: After obtaining Vq W, use the residue
conmiand in MATLAB to obtain the partial
flection expansion. You might also investigate
the use o f the command ilaplace.
(c)
(d)
(e)
^2^
Initisl
conditions. Simplijy each equation.
Put equations in matrix form.
Assuming /jr(0) = 0,
(0~) = 5 V,
W = 5<5(/) A, and
(^) = \06{t) V,
use Cramers rule to find the current
(f) and then i^ {t).
Now suppose that ii(0~) = 0,
= OV.
= OA. and () = 10(() V.
Find 1^2 W-
1 0 I.
Figure P I3.64
Figure P I3.66
65. In the circuit of Figure P I3.65,

= 0.25
S, C2 = 0.25 F, and /? = 1 Q. Write a set o f mod
ified loop equations and solve for the three loop
currents and V, assuming V(^0~) = 4 V and
v jt ) = \2u{t) V.
67. Consider the circuit of Figure PI 3.67. This

problem is to be solved using (modified) nodal
analysis.
(a) Draw the equivalent frequency
domain circuit assuming /^(0) = 0
and 'ci(0) = 2 V.
(b) Write three nodal equations for the
circuit o f part (a) only in terms o f the
voltages
Ifis), and the
initial conditions. Simplify your equa
(c)
(d)
tions.
Put equations in atrix form.
Using Cramers rule, find the transfer
fimction
_
H (s)^
(e)
'O
66. Consider the circuit of Figure PI 3.66. This

problem is to be solved using (modified) nodal
analysis.
(a) Draw the best equivalent frequency
domain circuit for nodal analysis
accounting for the as-yet-unknown
initial conditions ifiO') and i/q(0 ).
(b) Following the procedure explained in
the text, write three (modified) nodal
(0
(g)
I;in(s)
o f the circuit.
Find the impulse response h{t) of the
circuit.
Find the response of the circuit to
i.J^t) = -8u(t) A, assuming the initial
conditions are zero.
tial condition on the capacitor. A sim
ple observation leads to the answer
direcdy.
66
= 190 a, L =20 H,
{()
and
100 V
for all
time.
L + V .(t) -
Figure 1M3.67
v jt)l
A N SW ERS:
and
/-(/)
0 .2 5
H {s)=
4,s(.v+ 1)
Figure P i3.69
1 )//(/)dcrivcd from
( 2s +\ )
7 0 . The switch in the circuit o f Figure P I 3 .7 0

-4
2.V
is in position A for a very long time. It moves
t=
to position B at time
I,
1 sec, back to A at r =
2 sec, and then back to B at r = 4 sec, where it

remains forever. Suppose /?, = lOOQ,
/? 2
Q.,C^=2 mF, and C2 = 2 niF. Com pute

when V- (t) = 20z/(^) V. Plot for 0 < r < 5
A
{t) = 2{t - \ + e
CvÔ-)hit) = (e- - \)u{t).
Finally
= 500
(/)
sec.
and
V \ /V
+
R,
SW ITCH IN G PROBLEM S
6 8 . The switch in the circuit o f Figure P i 3 .6 8
c,
'1
----- 0
Figure P i3.7 0
is in position A for a very long time and then

moves to position B at time
t=I
sec, back to A
at f = 2 sec, and then back to B at / = 3 sec,
= 5 0 0 Q.,
where it remains forever. Suppose
= 1 k n . and C = 2 mF. Com pute

when
= 10//(r) +
5u{t -
1) + 5 ( / - - 2) -
7 1 . Repeat Problem 7 0 for
20u{t)
2 0 ( f - 1) V.
7 2 . Repeat Problem 7 0 for
R^ =
500
Q. and
y-(r) = 20//(r) + 2 0 (r- l)V .
2 0//(r - 3)V. Plot for 0 < r < 5 sec.

R
+
V (t)
Figure P I3.68
7 3 . In the circuit o f Figure P i 3 .7 3 ,
Rs = Ri,=
lo g ,( 2 )
Cj = 1F, and
Cj = 1 F.
1 .4 4 2 7
iX
Suppose switches Sj and
S2 have been closed for a long time. At r = 0, S2

6 9 . The switch in the circuit o f Figure P i 3 .6 9
is opened. At / = 1 sec, $ 2 is closed and Sj is
has been in position A for a very long time. At
opened. Will any capacitance voltages change
/ = 0, the switch moves to position B. Compute
abruptly in this circuit?
/?, over [0, oo). Assume
energy dissipated in
(a)
Find i/,(0 ") and
R^ = 2 Q, R^ = 8 Q., R^
(b)
Find
and
for 0 < r < 1 sec.
(c)
(d)
(e)
677
Find v^{\~) and ^2( 1~)

Find
and ^2( 1^).
Find v-^{t) and ^2(^) for 1 ^ ^sec.
i'
+
v,(t)
Figure P I3.75
32 V
c,
Figure P I3.73
74. In die circuit of Figure P13.74, Vq = 10 V,
= 4 Q,
= 3i2, /?2 = 24Q, Zj = 3 H, and
^2 = 6 H. Suppose switch S j has been closed
for a long time while S2 has been open. Ax. t =
0 S2 is closed, after which Sj is opened. At ^ =
1 sec, S j is closed and S2 is then opened.
(a) Find /j(0") and
(b) Find z2(/) and
for 0 < f < 1 sec.
Find Zj(l) and
(c)
Find
and /2(1'^).
(d)
and
for 1 < /sec.
(e) Find
76. In the circuit in Figure P I3.76, the switch

has been in position A for a long time. The
switch moves to position B at r = 0, back to A
at ? = 2 sec, and finally back to B at r = 5 sec,
where it remains. Suppose
= 2.5
=
1.5625 H, R2 = 2.5 Q, ^2 = 1 H, C = 0.04 F,
and
= 10 V for all t.
A
vjt)l
(a)
(c)
(d)
Figure P I3.74
75. Consider the switched RLC network of

Figure P13.75, in which
= 2.5
Z, = 1 H,
and C = 0.04 E
(a) Suppose the switch S has been closed
for a long time and opens at r = 0 . If
v.^{t) = -40(-/) V, find V(^t).
(b) Suppose the switch S has been closed
for a long time, and it opens at r = 0
and recloses at r = 1 sec. If
=
-40u{t) + 40(/) V, find
:\{t)
Figure P I3.76
(b)
CHECK: (b) /jW has a pole at ^ = - 1 , and I2

is) has a pole at ^ = - 4 ; (e) !^{s) and /2W both
have poles at j = - 2 and s = -5-
Find Vf^s) and

for 0 < / < 2 sec.
Draw the equivalent frequency
domain circuit that is valid for / > 2
sec.
Find an expression for
with the
switch in position B and then deter
mine
for 2 < f < 5 sec.
Compute V(jJ) for / > 5 sec. Hint:
This can be done without any further
computation.
77. Consider the circuit o f Figure P I 3.77.

Suppose R^ = 2Q, /?2 = 4 Q,
= 4 Q, Cj = 0.3
F, C2 = 0.25 F, and v-JJ) = -30(-/) + 30(/) +
30u (t- 10) V. The switch has been in position
A for a long time and moves to position B at ^
=
0.
(a)
(b)
(c)
(d)
Find
and y ^ (0 ).
Draw the ^-domain equivalent circuit
valid for t >0 .
Write a set o f nodal equations in terms
of Vî W and
Solve the set of nodal equations con
structed in part (c) and determine
Chapter 13 * Uiplace Transform Analysis II: Circuit Applications
67H
for 0 < r < 10 sec.

(c)
(e)
solving the equations o f part (e) for
Com pute the initial conditions at / =

10
sec,
i.e.,
find
^ ^ (1 0 ")
Set up the Cram ers rule equations for
and
y ^ ( 1 0 ). Also assume that at r = 10

sec, the switch moves back to position
domain valid for r > 10 or r' > 0.
SW ITCH ED CAPACITOR
NETW ORKS
Com pute
7 9 . Consider the circuit o f Figure PI 3 .7 9 .
A. Draw the equivalent circuit in the sfor r > 1 0 sec.
(a)
If
v^{t)
= 25 V and the 100 m F capac
itor is uncharged at r = 0 , find

2 sec <
(b)
V(^t) for
t.
vp) = 25 V and
vît) for 2 sec < t.
Now suppose
10 V. Find
t=2 sec
Figure P I3.7 7
C H E C K : (d)
v^Jt) = [ 1,846^(-2-5')
' " '6

7 8 . Consider the circuit in Figure P I 3 .7 8 .
R
150 mF
100 mF
v,(t)
Figure P i 3.79
21,(s)
8 0 . In the switched capacitor circuits o f

Figure P i 3 .8 0 , the switch closes at r = 0.
l.w
2V Js)l
:v,u)
ijs)
2Q
Suppose
+
V fs)
(a)
= 5 7 2 mV.
For the circuit o f Figure P I3 .8 0 a ,

compute
for ^ > 0.
(b) The circuit o f Figure P I3 .8 0 b differs

from Figure P I3 .8 0 a in that the
Figure P i 3.78
v-^j^t)
-source is replaced by a 10 //F capaci
Part 1.
The switch is in position A at r = 0 and
tor with an initial voltage i^^/0) = 5 7 2
t'Ô ") = 2 V, /^(0) = 0 and for 0 < r < 10,
mV while all other capacitor voltages
= 2 e ^ u{t) A .
are zero at r = 0~. C om pute
(a)
Draw the frequenc)' domain equiva
r> 0.
lent circuit o f the capacitor and com

pute
V(4,s).
(b)
C om pute
(c)
C om pute
Part 2. \i t =
Approximate t^^^lO).
V^Js)
(d)
lu { t -
v^Jt).
10 sec, the switch moves to posi
new
input is given by
three
(modified) nodal
tion B. Assume that the

/.,(r) =
and
10) A.
W rite a set o f
equations in the ^-domain in terms o f

file equa
tions in matrix form. Your answer
should contain
parameter.
R as
an undetermined
(b)
F igu re P I 3 .8 0
for
G79
8 1. For the circuit in Figure P I 3.81 ^^(0') = 0.

At time
t=0'
both switches arc flipped to posi
itor voltages are zero at f = 0
. Com pute
and //2 (^) at / = O'", r = 2~ , and
v\(t)
t = 2* .
tions A. At r = 1 sec, both switches move back

to their original positions, and then back to
positions A at r = 2 sec.
(a)
Determine the vakie o f yj(O^).
(b)
Determine the value of
lO F
85. In the switched capacitor circuit o f Figure
20 5(t)
5F
CH ECK :
5F
5F
20u(t)V
P I 3 .8 5 , all capacitor voltages are zero at r = 0~.

Switches S arc moved to positions A at r = 0 s
Figure P i 3.81
and to positions B at /= I sec, and then back to
=3 V
A at / = 2 sec. C om pute and plot
v^{t) and V2 {t)
for 0 > /> 3 sec.

82. Repeat Problem 81 for the circuit o f Figure
V,
3 mF
P 1 3 .8 2 .
2m F
20 V
40 8(t)
(!)
2F
6 mF
Figure P I3.85
4F
2F
1 mF
> r i4 F
4F
2m F
10u(t)V
86. In Figure P I 3 .8 6 ,
v^{t) = - 2
V,
= 0,
Figure P I 3.82
and the switch S is in position A. At r = 0 , S is
83. For the circuit shown in Figure P I 3 .8 3 ,
back to A at / = 1 sec and then back to B at r =
moved to position B, after which it is moved

= 0. The switch moves from position A
2 sec. Com pute and plot
for 0 > r > 3 sec.
to position B at r =1 sec, back to position A at
t =2
I mF
4e-
sec, and finally back to position B at ^ = 3
sec, where it remains forever.

(a)
Com pute
(b)
Com pute t^^;(3.5).
(c)
Plot
+
v.Jt)
for 0 > r > 4 sec.

2 mF
rlO F
lO F
lO F
f +
Figure P I3.8 6
20u(t)V
87. In Figure PI 3 .8 7 , suppose

Figure P 13.83
A N SW E R : rJ3.5^) = 6.25
and
v^{t) = - 2
V, / = 1,
the capacitors are initially uncharged.
Suppose switch S is moved to position A at r = 0,

2, 4 , . . . (even integer values) msec and to position
84. In the switched capacitor network o f Figure

P I3.84, the switches S are moved to positions
A at r = 0 and to positions B at r = 2. Ail capac-
B at r = 1, 3, 5 ... (odd integer values) msec.

(a)
Find
0 < t < 20
and plot the waveform for

msec.
Chapter 13 Laplace'Iransform Analysis II: Circuit Applications
680
(b)
Repear part (a) for ^ = 0 .5 .
(c)
Repeat parr (a) for
increasing bur less rhan 8 0 V, and is

moved to position B when
k = 2.
v^^ reaches
80 V. (2) S remains at position B when
kC
is decreasing bur is greater rhan 5 V,

and is moved to position A when
reaches 5 V. Find
and make a
rough sketch of the waveform for one

cycle o f operation.
C H E C K : charging time s 4.?? msec;
discharging time s 3? jisec
M ISCELLAN EO U S
8 8. Use the material on op amp integrator
design to achieve the following input-output
characteristics. In your final design, capacitors
should be 100 nF.
0 .5
(a) V,u,b) = - \<l\

S
(b)
(C ) V ,
0
+ - V ;b\
(0 = -
Figure P i3 .8 9 Switching circuit lor gen

erating a sawtooth waveform.
I',h\
9 0 . (Fluorescent light) Reconsider the
0 .7 5
+
K .-
s
b\
fluorescent
light
starter
circuit
hi
described in the text (Example 13.8).
In this problem, suppose the ballast is

realistically modeled as an ideal
inductor in series with a 100 Q resistor,
as shown in Figure P i 3 .9 0 . Using SPIC E or

89. (Sawtooth waveform generation) This prob
some other circuit simulation program, com
lem uses some simple switching techniques in an
pute the starting voltage,
RC circuit
tial condition on the inductor, as depicted in
to generate an approximate sawtooth
due to the ini
waveform. Sawroorhs are common to a number
Figure P i 3 .9 0 . Estimate the starting voltage
o f devices, such as televisions and test equipment.
from a plot o f the response over one half-peri
Consider the circuit of Figure P i 3.8 9 .
od. Com pute the difference between the peak
(a)
Assume the circuit is initially at rest.
present. Is the lossless circuit o f Example 13.8 a
alternately closed to the left position. A,
good approximation o f the starter response?
for 1 msec and then to the right posi

tion, B, for 50 |isec. Find z-'Q(r) for 0 s r
^ 1.05 msec (one q^cle of operation),
and sketch the waveform.
CH ECK :
(b)
voltages with and without the 100 Q resistor
Beginning at time / = 0, the switch, S, is
19 V
T he circuit in the shaded box is a

crude model o f a neon lamp (costing
less than a dollar) and operates as fol
lows: (1) S is at position A when
is
681
1000
0.001 pF
Li,(0)5(t) =0.088(t)
Time domain representation of
voltage due to initial inductor
current when heat sensitive
switch opens
Very high
resistance prior
to arcing
Figure P I3.90 Model of fluorescent light starter circuit that includes a ballast resistance of 100 Q.
Note that in the time domain, the effect of the initial inductor current appears as an impulse in this
model.
Laplace Transform Analysis IIITransfer Function Applications

APPLICATION TO ELECTRIC M OTO R ANALYSIS
Electric motors turn fans, run air conditioners, pull trains, rotate antennas, and help us in a wide
variety o f ways by efficiently converting electrical energy into mechanical energy. These electro
mechanical devices are o f two general types: ac and dc. D c motors are ordinarily used in electricpowered transit cars, often with rwo motors per axle to propel the car. A typical rating o f such a
m otor is 140 hp, 3 1 0 V, 2 ,5 0 0 rpm. Another type o f dc m otor is a high-performance dc servo
motor, found in computer disk drives and microprocessor-controlled machinery. These motors are
very useful in applications where starts and stops must be made quickly and accurately. As an
application, we will represent a dc m otor by an equivalent circuit-like model and analyze its oper
ation using the transfer function method.
O f the several kinds o f dc motors, the t)'pe most pertinent to the analysis techniques o f this chap
ter is a permanent magnet type typically found in low-horsepower applications. They are reliable
and efficient. Further, for a permanent magnet dc motor, a plot o f the torque produced on the
rotating shaft o f the motor versus the input current to the motor is almost a linear curve. Hence,
the m otor has a linear circuit-like model that most nearly describes its performance over a large
range o f operating conditions. Since the output o f the m otor is a mechanical quantity, e.g., angu
lar velocit)' o f the m otor shaft:, the transfer function is the only viable modeling tool available to
us at this stage o f our development. Section 9 presents the circuit model o f the m otor and devel
ops the transfer function analysis o f its operation. Several problems at the end o f the chapter
extend the analysis. In more advanced courses, time domain analysis is developed and used.
CHAPTER O BJECTIVES
1.
2.
Characterize the transfer function of a circuit in terms of its poles, zeros, and gain constant.
Use knowledge o f the pole locations o f a transfer ftjnction to categorize generic kinds of
responses (steps, ramps, sinusoids, exponentials, etc.) that are due to different kinds o f
terms in the partial fraction expansion o f the transfer function.
3.
Identify, categorize, define, and illustrate various classes o f circuit responses, including
fi.S-4
Chapter 14 Laplacc Transform Analysis III: Transfer Function Applications
the zero-srate (zero initial conditions) response, the zero-input response, the step
response, the impulse response, the transient and steady-state responses, and the natural
and forced responses.
4.
Define the notion o f the frequency response o f a circuit, explore its meaning in terms of
the transfer function, and introduce the concept o f a Bode plot, which is an asymptotic
graph o f a circuits frequency response.
5.
Introduce the notions o f frequency and magnitude scaling of circuits.
6.
Illustrate the applicabilit)- o f the transfer function concept to a circuit model o f a dc

motor.
SECTIO N HEADIN GS
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Introduction
Poles, Zeros, and the 5-Plane
Classification o f Responses
Computation o f the Sinusoidal Steady-State Response for Stable Networks and
Systems
Frequency Response
Frequency Scaling and Magnitude Scaling
Initial- and Final-Value Theorems
Bode Plots
Transfer Function Analysis o f a DC Motor
Summary
Terms and Concepts
Problems
O ur experience o f using Laplace transforms to calculate responses makes clear that the pole-zero
structure o f the transfer function sets up generic kinds o f circuit behaviors: constants, ramps,
exponentials, sinusoidals, exponentially modulated sinusoids, etc. Such knowledge leads to a qual
itative understanding o f the circuits response. For example, if a pole is in the right half o f the com
plex plane, then we know that the response will grow with increasing
t.
Identifying this kind of
behavior allows us to define the notion o f stabilit)' o f a circuit or system. Generally, the pole-zero
locations allow us to categorize and com pute various special t)'pes o f responses, including tran
sient, steady-state, natural, forced, step, and impulse responses. Coupling the transfer function
with the presence o f initial conditions in the circuit permits us to define two further types of
responses fundamental to both this text and advanced courses in circuits, systems, and control: the
zero-input response (due only to the initial conditions o f the circuit or system) and the zero-state
response (due only to the input excitation, assuming that all initial conditions are zero).
These time domain notions are balanced by the concept o f the frequenc)' response of the circuit
or system. Briefly, the frequenc)' response is the evaluation o f the transfer function,
Jii).
Since
H{Jw)
H{s),
for
s=
has a magnitude and phase, the frequenc)' response breaks down into a magni-
Chapter 14 * Liplacc Transform Analysis 111: Transfer Function Applications
6S"S
rude response and a phase response. A technique for obtaining asymptotic (straight-line) approx
imations (called
Bode plots)
is also outlined in this chapter.
As a final introductor)^ remark, unless stated otherwise, all circuits in this chapter are linear and
have constant parameter values. Such circuits are said to be
linear
and
time invariant.
Also, for
convenience in this chapter, the symbol Z. will be used to denote either o f two things: (i) the angle
o f a complex number,
=
L{a + jb)
= arg(/z + jb) =
z^n~^{bla), with
due regard to quadrant, or (ii) Z.({)
The context will determine the actual usage o f the symbol.
2. POLES, ZEROS, AND TH E S-PLANE

In all our circuits, impedances
Z{s),
}\s),
admittances
and transfer functions
H{s)
are rational
functions of^, i.e., they are ratios o f a numerator polynomial ;/(j), divided by a denominator poly
nomial,
r/{s).
Mathematically,
(i{s)
where
s = pj
is a finite pole o f
H{s), s = zj
{ s - p^ ) (s - P2 )...{S - P)
transfer function. We assume that common factors

satisfies M /,) =
H{s), and A" is the gain constant of the

o f n{s) and d{s) have been canceled. A finite pole
is a finite zero o f
which is shorthand for //(s) - x as j />y, and a finite zero satisfies
the transfer function takes on the value zero at each
z-.
If pj = pj,
j,
= 0. i-e.,
the pole is a repeated pole. A
pole repeated twice is second order; one repeated three times is third order, etc. The terminology is the
same for zeros. Also, transfer functions sometimes have infinite poles or infinite zeros. If m
-*
-
00
, then
m at
H{s)
H{s) has a zero

rn - n at infinit)'.
0, suggesting the term zero at infinit}'. In such cases,
infinit)'. If, on the other hand,
n < m, H{s)
has a pole of order
< n and s
n
of order
O ut o f all this terminolog)' comes one striking fact: transfer functions, impedances, and admit
tances are characterized by their finite poles, their finite zeros, and their gain constant.
E X A M P L E 1 4 .1 . A transfer function
H{i)
has poles at j = 0, - 2 , and - 4 , with zeros at j = 1 and
- 3 . A t x = l , / / ( j ) = 4 /3 . Fin d M ^ ).
S
o l u t io n
From equation 14.1 and the given locations o f its poles and zeros, the transfer function must have
the form
(.S -- 1 K5 -C
^ (i-H l)(5 + 3)
H(s) = K ------------------------ ;--------- T

{s - /;, )(5 - P2 )(s - />3 )
Since
H{\) = 413,
.v(.v + 2 )(.v -I- 4 )
evaluating equation 14.2 at ^ = 1 yields
l(l + 2 ) ( l - h 4 )
15
This implies that K = 2 .5 . Equation 14.2 specifies the transfer function with K = 2 .5 .
686
Chapter 14 Laplace Transform Analysis 111; Transfer Function Applications
Exercises.
0 0 , H{s) -*
1. Suppose a transfer function has a zero at j = 1 and a pole at x = - 1 and that as ^
H{s).
His) = -Ms - 1 )!{s +
- 3 . Find
A N SW E R :
2. A transfer function
known that
H{0)
H{s)
1)
has poles at j = - 1 and - 2 and zeros at j = - 3 and - 5 . It is further
= 15. Find //(oo).
A N SW E R : 2
Because the essential information about transfer functions resides with the poles and zeros, a plot
of these locations in the 5-plane, called a pole-zero plot, proves informative.
E X A M P L E 1 4 .2 . Draw the pole-zero plot o f the transfer function o f Example 14.1.

S
o l u t io n
T he transfer function given in equation 14.2 has the pole-zero plot shown in Figure 14.1,
JW
>k
j
-o--------- 'G
-4
-3
-2
-1
...J
FIG URE 14.1 Pole-zero plot of //(j) given by equation 14.2, where the poles arc flagged by
and
the zeros by o. Since j = a + ya), the real axis is labeled o and the imaginar)' axisyw.
As a second illustration, consider the transfer function

(5 + 1)(.V4-3)
(1 4 .3 )
l(5 + l) -H l](5 + 2 )
which has the pole-zero plot shown in Figure 14.2.
jco
A
-j
-o---- ----- o-4
-3
-2
X
FIG U R E 14.2 Pole-zero plot of
H{s) given
> a
-1
...J
by equation 14.3. Again, the a-axis represents the real
part of the pole or zero and the /o -axis represents j times the imaginary part of the pole or zero.
Chapter 14 Laplace Transform Analysis III: Transfer Function Applications
68
Plots such as those in Figures 14.1 and 14.2 com m unicate much about the nature o f the imped
ance, admittance, or transfer function o f a circuit. For example, an
RC input
impedance,
satisfies the following properties: (i) all o f its poles and zeros are on the non-positive C7-axis o f the
complex plane; (ii) all o f its poles are simple (o f multiplicit)' 1) with real positive residues, i.e., the
coefficients in a partial fraction expansion are real and positive; (iii)
does not have a pole at
j = CO; and (iv) poles and zeros alternate along the a-axis. Proofs o f these assertions can be found
in texts on network synthesis.
Exercise.
C om pute the input impedance o f the circuit in Figure 14.3, and show the pole-zero plot
if / ? ,= / ? - ,= 1
Q,
Cj = 0 .2 5 F, and C2 = 0 .5 F. Are the poles on the non-positive a-axis? Do the
poles and zeros alternate? Is there a pole at
s=
oc?
J_
c.
A N SW E R :
A.C ,
More commonly, pole-zero locations provide important qualitative information about the response of
the circuit. Pole locations determine the inherent, natural behavior of the circuit, and the poles are
commonly called
tmttiralfrequencies.
Howe\'er, the complete set o f natural frequencies o f the circuit
may be larger than the set o f poles o f the transfer function. 'Fhis is because there might have been a
pole-zero cancellation in constructing the transfer function. The canceled pole would amount to a nat
ural frequency o f the circuit that is not present in the poles o f the transfer ftmction.
T he terms in a partial fraction expansion o f the response establish the types o f behavior present in
Kh,
pj real in each case. Figure 14.4 sketches each o f the associat
ed responses. In Figure l4 .4 a , the term Kh leads to a dc response and KlJ to a polynomial response
proportional to
. In Figure l4 .4 b , the term K l{s- p-) leads to an exponential response that is
increasing if pj > 0 and decreasing if pj < 0. Finally, in Figure l4 .4 c , if pj < 0, the response curve
the response. Each term has only one o f several possible forms. Four ver)' com m on terms are
KjJ,
K/(^ - />y), and
K/{s -
with
has a hump.
These qualitative behaviors suggest that one important application o f the transfer function is
determining the stability o f the response; i.e., under what conditions will the circuit response
remain finite for all time?
688
T ransform
(a)
T im e R esp onse
(b)
V-
Pi
(c)
(S -
Pi)
FIGURE 14.4 Response t)'pcs common to partial fraction expansion terms, (a) The term
response and
KJ^ to a polynomial
response proportional to
(b) The term
KJ{s
Kh leads to a dc
leads to an expo
nential response diat is increasing i f > 0 and decreiising if pj < 0. (c) If pj < 0, the curve has a hump.
Chapter 14 Laplacc'Iransform Analysis III: Transfer Function Applications
68^)
In addition to the preceding response t)'pes, there is the sinusoidal response associated with terms
o f the form
/4.V+ B
-----------^------- T
f.v + a r +
Here, i f - o < 0, the response is an exponentially decaying sinusoidal, as sketched in Figure l4 .5 a ,

and if a = 0, the purely sinusoidal response o f Figure 1 4 .5b results. If - a > 0, the exponentially
increasing sinusoidal response o f Figure l4 .5 c occurs.
FIG URE 14.5 Various sinusoidal responses, (a) Exponentially decaying, - a

(b) Pure sinusoidal, 0 = 0. (c) Exponentially increasing, - a
> 0.
< 0.
Chapter 14 l^placeTransform Analysis III; Transfer Function Applications
690
Referring again to Figure 14.5 , the real part o f a pole, i.e., - a , specifies the
response. Often, the word
damping is
decay rate
o f the
used. I f -O < 0, then the response is damped and the oscil
lations die out. T he farther -O is to the left o f the imaginary axis, the greater the damping. If a =
sustained oscillation. I f - a > 0,

unstable And increases without bound.
0, there is no damping and the response is a

tively damped, i.e., the response is
the response is nega
O ne concludes that pole locations specify the type o f time domain behavior o f a circuit or system.
A very important type o f circuit behavior characterized by the pole locations is
stability.
STABILITY AND BO U N D ED N ESS

A circuit represented by a transfer function
H{s)
is called
stable if ever)' bounded input signal

say,J{t), is bounded if |/(^)| < K <
yields a bounded response signal (B IB O stability). A signal,

00
for all t and some constant
K. In other words, a signal
is bounded if its magnitude has a max
imum finite height. Interpreting this definition in terms o f the poles o f the transfer function,
one discovers that a circuit or system is stable if and only if all the poles o f the transfer func
tion lie in the open left half o f the complex plane. This makes sense, because if any poles were
in the right half o f the plane, the response would contain an exponentially increasing term; if
any were on the imaginary axis with multiplicity 2 or higher, then the response would contain
an unbounded term proportional to
fory > 2; and, finally, if there were an imaginary
axis pole with multiplicity 1, excitation o f the pole by an input o f the same frequency would
yield a pole o f multiplicity 2. The corresponding response term would be proportional to
fcos(o)r + 0 ), which grows with time an unstable behavior. W hat this means is that, for exam
ple, a unit step current source in parallel with a 1 F capacitor would produce a voltage pro
portional to
tu{t). This
voltage grows without bound and would destroy the capacitor and pos
sibly the surrounding circuitry if left unchecked. Such phenomena are considered unstable.
Despite the need for stability, some circuits utilize an unstable-like response for a finite duration.
Circuits that exhibit both stable and unstable-like responses are studied in electronics courses.
A transfer function with first-order poles on the imaginary axis is sometimes called
metastable.
Such a classification has no practical or physically meaningful significance, since the ubiquitous
presence o f noise would excite the mode and cause instabilit}' o f the circuit. Moreover, in power
systems engineering, i.e., the study o f the generation and delivery o f electricit)' to homes and
industry, transfer function poles that are in the left half-plane, but close to the imaginary axis, are
highly undesirable. Such poles cause wide fluctuations in power levels. The situation is analogous
to the way a car without shock absorbers would bounce. Much work has been done on how to
move the poles that are close to the imaginary axis farther to the left. Moving these poles to the
left increases the damping in the system and maintains more stable power levels. Summarizing,
the requirement that the transfer function have no poles on the imaginar) axis is both theoreti
cally and physically meaningful.
Exercises.
1. If
H{s)
691
find a bounded input that will make the response
unbounded.
2. If H{s) =
= l/(^^ + 1), find a bounded input that will make the response unbound
ed. Use M ATLAB or some other program to plot the response for 0 ^ r s 10 s.
A N SW E R S: fa) I/- (.v) =
io r k ^
l ;( b ) V;(a) =
lor
> 1
(.V- + 1)
E X A M P L E 1 4 .3 . During a laborator}' experiment, a student tried to build an inverting amplifi
er, as shown in Figure l4 .6 a . T he student accidentally reversed the connection o f the two input
terminals and obtained the circuit o f Figure l4 .6 b . The student was gready surprised that the cir
cuit did not behave as expected. Explain this phenomenon in terms o f the stabilit)'^ theory just
developed.
R^ = 4 k n
= 4kO
(a)
(b)
FIGURE 14.6 (a) Correct wiring of op amp circuit, (b) Accidental, improper wiring of op amp circuit.
S
o l u t io n
Assume that the op amp is modeled as a voltage-controlled voltage source with a finite gain o f 10"^
and that there is a very small stray capacitance o f 1 pF across the input terminals. Figure 14.7 illus
trates the equivalent circuit model for each o f the circuits in Figure 14.6.
= 4kO
R^ = 4kQ
FIG U R E 14,7 (a) Model of correctly wired op amp circuit,

(b) Model of incorrectly wired op amp circuit.
6 )2
P art 1:
Analysis o f the correctly wired op amp.
W riting a node equation at
to compute the trans
fer function o f Figure l4 .7 a yields
-av
- v ;)+ ^ (- v , - v) =0
A|
A2
After some algebra,
Y
Substituting
v.,- v,- v = o
.V + - - - - - - - +
CRi
CR| C , j
''
CR,
x 10^ for Ky produces the stable transfer function
\
His) =
10 13
Vi(s)
R^C
1
l+ I O '
.y + ------ +
R^C
with approximate dc gain

P art 2 :
^ + 2 .501 X 10
(1 4 .4 )
12
RjC I
as expected.
Analysis o f the incorrectly wired op amp.

Vj.
To com pute the transfer function o f the circuit o f
Figure l4 .7 b , we write a node equation at
C i V , , + ^ ( V ^ - V '. ) + -^ (V r f -V ' ) = 0
Kj
/<2
which produces the transfer function
H{s) =
1013
10"
Viis)
R^C
1
1 -u r
5 + -------- h
. y - 2 .4 9 9 x 1 0
12
(1 4 .5 )
T he transfer function o f equation 14.5 has a right half-plane pole, in contrast to that o f equation
14.4. This implies that the incorrectly wired circuit is unstable, which explains the students con
cern over the surprising performance o f the op amp.
A brief interpretation o f the zeros o f a transfer function ends this section. This is best done in
terms o f a simple example. Suppose
His) =
Let
v- it)
e sin(/)/^(f)
( .v + l )- + l
V;.,
(5 + l)U + 2)(^ + 3)
V, so that
Assuming that the system is initially relaxed, i.e., all initial conditions are zero, we obtain
V^,As) = H is )V :Js ) =
1
(.9 + l)(5 + 2 )C y + 3)
Chapter 14 Laplace Transform Analysis 111: Transfer Function Applications
693
The time response is
for appropriate constants
A^-.
Observe that the response dies out very quickly and does not have
any term similar to the input signal. This follows because tiie input signal has transform poles,
= -1
J,
that coincide with the zeros o f the transfer function. One can think o f the pole locations
in the transform o f the input signal as identifying frequencies that are present in the input. Hence,
the effect o f these input signal frequencies (poles) is canceled out by the transfer function zeros,
eliminating them from the circuit response.
3. CLASSIFICATIO N OF RESPONSES
In addition to the various response behaviors discussed in section 2, there are other general
response classifications. Three fundamentally important general response classifications germane
to all o f circuit and system theory are the zero-input response, the zero-state response, and the
complete response.
Zero-input response: T he response o f a circuit/system to a set o f initial conditions with the

input set to zero.
Zero-state response;The response o f a circuit to a specified input signal, given that the ini
tial conditions are all set to zero. Figure 14.8 illustrates this idea.
Output
Input F(s)
Relaxed Circuit
H(s)
>
V(s) = H(s)F(s)
Zero-State Response
FIGURE 14.8 Relaxed circuit having transfer function H(s) and zero-state response.
Complete response: The response o f a circuit/system to both a given set o f initial condi
tions and a given input signal. For linear circuits, the complete response equals the sum o f
the zero-input and zero-state responses.
Recall that a circuit is linear if, for any tv.'o inputs,yj(f)
and/jU),
whose zero-state responses are
and_y2 W> respectively, the response to the new input [AT|/j(r) +
K^ 2 ^t)]
is
where /T, and A'-, are arbitrary scalars. The circuits studied in this book are linear unless
otherwise stated.
T he decomposition o f the complete response into the sum o f the zero-input and zero-state
responses is important for three reasons:
1.
It is defined for arbitrary input signals.
6 9
~^[H{s)F{s)],
lôr tiie arbitrary i-domain input
F{s).
2.
T he zero-state response is given by
3.
It illustrates a proper application o f the principle o f superposition for linear dynamic net
works having initial conditions.
T he following example illustrates point 3.
E X A M P L E 1 4 .4 . Consider two linear networks: (i) a linear resistive network, as sketched in

Figure 14.9, and (ii) a linear dynamic network, as sketched in Figure 14.10.
= 6u(t)
FIG U R E 14 .9 Linear resistive network.
20
Va(t)
+
Vo(t);
= 3u(t)
2 0
v jt)
1F
= 6u(t)
FIG U RE 1 4 .1 0 Linear dynamic network.
Response o f linear resistive network o f Figure 14.9. For the resistive network o f figure 14.9,
vj^t) due to vj^t) with
= 0 is v^J^t) = l//(f), and the contribution due to
VjJ^t) with vj,t) = 0 is
= lu{t). By superposition.
P art 1:
the contribution to
For this type o f circuit there is no initial condition and the complete response consists o f only the
zero-state response, which decomposes into the superposition o f each source acting alone.
P art 2 :
Response o f linear RC network o f Figure 14.10.
Now consider the dynamic network o f
Figure 14 .1 0 . Suppose the capacitor has an initial voltage o f 2 V at f = 0, i.e., ^^(0) = 2 V.

Step 1. W ith
applied, y^(0) = 2 V, and
Vf^set
to zero, the response is
v j t ) = {0.5e- ^ \.5)ii{t)W
Step 2 . W ith the input
Vy{t) applied,
y^(0) = 2 V, and
= {-e-^
set to zero, the resulting response is
5)u{t)
Step 3 . An
incorrect application
695
o f superposition implies that
T he last answer is wrong because the response due to the initial condition has been added in twice.
Step 4 . A correct application o f superposition would entail:
(i)
computation o f the zero-state response due to
(ii)
computation o f the zero-state response due to
(iii)
vj^t),
Vy{t), and
computation o f the zero-input response due to y^(0).
By superposition, the complete response is the sum o f all three. In particular, the zero-state
response due to
is
v J t ) = \.%\-e-û{t)V
and the zero-state response due to
is
Vobit)
= 3(1 -
e-^n{t)
Hence, the complete zero-state response, by superposition, is the sum o f

the zero-input response is
le~Ui{i).
and
Further,
Hence, the complete response is

= (4-5 - 2.5rO /K r) V
It is important to note that the transfer function is defined only for circuits whose input-output
linear input-output behav

H{s), then H{s)[K^Fj^{s) + K2 Fi{s)] = K^H{s)F^{s) +
Yf,s) = H{s)Ff^s) is the zero-state response o f the network
behavior is linear. In terms o f the zero-state response, if a circuit has a

ior characterized by a transfer function
K2 H{s)p2 {s) = ATj KjU) + KjYîs), where

to F^^s). This says that the zero-state response
to [A"j/j(r) +
is
l^^^^ce,
the transfer function model reflects the underlying linearit)^ o f the circuit.
The complete response has a second structural decomposition in terms o f the transient and steadystate responses. T he notion o f a periodic signal is intrinsic to these classifications. A s ig n a l/r) is
periodic if there exists a positive constant
T such
thaty(r)
=j{t + 7)
for all / > 0. (The restriction
to r > 0 exists because our Laplace transform analysis implicitly constrains our function class to
those that are zero for t < 0.) If a signal is periodic, there are many positive constants for which
j{t) =J{t + 7) for all t> 0. For example, i f / / ) =J{t + T) for some T and for all t> 0, then it is true
for IT, 5T, etc. We define the fundam ental period, often simply called the period, o f / f ) to be
the smallest positive constant T for w h ic h /f ) = J{t + T) for all t > 0. Sinusoids are periodic sig
nals: sin(2Jtr) = sin(2Jt/^ + 2 k ) with fundamental period T= \. The square wave o f Figure 14.11 is
periodic with fundamental period 7"= 2.
Chapter 14 Liplace I'ransform Analysis ill: Transfer Funciion Applications
f(t)
>k.
1
-1
FIGUllH 14.11 A periodic square wave with fiindamental period 7'= 2.
This notion of periodicity and, by default, non-periodicity allows us to define the transient and
steady-state responses o f a circuit.
Steady-state response: Those terms o f the complete response that satisfy the definition o f
periodicity for r > 0. This includes a constant response.
Transient response: Those terms o f the complete response that are not periodic for / > 0 , i.e.,
that do not satisfy' the definition o f a periodic function for / > 0. Note that a constant response
satisfies the definition o f a periodic function.
A circuit response may have no transient part, as illustrated by the sustained sinusoidal oscillato
ry response o f the circuit given in Figure 1 4 .1 2a. Further, the steady-state part o f the response may
be zero, as in the circuit o f Figure 1 4 .12b, where
is a damped sinusoid. If the a circuit is
unstable, the transient response may blow up, overwhelming the constant or periodic part o f the
complete response, as in the case o f
{e^ cos(10/)
+ 1 5)u{t) V, where the steady state is 15(/)
V. Note that transient here does not mean something that diminishes in importance with time.
Most circuits have both a transient and a steady-state response. When the input is constant or peri
odic, the circuit response approaches the steady-state response asymptotically for large
t, i.e.,
as the
transient dies out, only if the circuit is stable. For such circuits, the steady state is crucial. Further,
when the input is sinusoidal, the steady-state response is easily computed via the transfer function,
//(s ), or by the phasor method. Details o f the calculation are presented in section 4.
1F
(a)
1 H
1F
1O
1 H
(b)
FlG U lll{ 14.12 (a) Unstable circuit illustrating the possibility of no transient response,
(b) Stable circuit having a zero steady-state response.
Chapter 14 Laplace Transfomi Analysis III: Transfer Function Applications
6^)'
E X A M P L E 1 4 .5 . Com puting the response o f the circuit o f Figure 1 4 .1 3 provides a simple illus
tration o f the decomposition of the complete response into tiie sum of the zero-input and zerostate responses. Also, some rearrangement o f the terms identifies the transient and steady-state
responses.
-O
R .
>
P
>
ZJs)
1
(a)
FIG U R E 14.13 /?Ccircuit for Example 14.5. (a) Time domain circuit.
(b) Frequency domain equivalent, accounting for initial condition.
S
o l u t io n
Step 1.
Computation o f the zero-state response.
The input impedance can be viewed as a special type
o f transfer function. For Figure 1 4 .13a the output
V(^s)
is the voltage appearing across the input
current source. Hence,
__
R
Letting
i-^^{t) =
/q/K^) and
= 0, then zero-state response is
Rl,,
Rio
.v-i-
Step 2 .
RC
= RI,
Ki t )
RC
Computation o f the zero-input and complete response.
Now, supposing that ^(^0)
0 , the
zero-input response is the inverse transform o f //(i)[C y^ '(0)] as per Figure 1 4 .13b. Hence, by
superposition, the complete response is
RC
u{!)
zero-state response
Step 3 .
}e
zero-input response
Decomposition into transient and steady-state responses. As
a final point, since the input is
dc, a step function, the complete response decomposes into its transient and steady-state parts as
''c ( 0 = ( v c ( 0 " ) - / ? /o ) ^
transient response
Rl(Mn
steady-state response
698
Chapter 14 Laplacc Transform Analysis 111: Transfer Function Applications
Observe that from the above example the transient and zero-input responses are ordinarily different.
E X A M P L E 1 4 .6 . It is sometimes mistakenly said that the zero-input response contains only those
frequencies represented by poles o f the transfer function. To see the fallac)' o f this statement con
sider the y?C bridge circuit o f Figure 14 .1 4 . Com pute the zero-state and zero-input responses.
o l u t io n
The transfer fiinction o f this circuit is
H{s)
zero-state response is always zero. On the other hand, if

'o u t ( 0 = V c ( 0
where
= 1 Q. Thus,
= 0> which has no poles. Hence, the

0, then
)e x p
the zero-input response is a decaying exponential whenever
v^^Qr) ^
0 and
C > 0. Notice that the transfer function has no poles. As a side remark, in this case, the zero-input
response is also the transient response, with the steady-state response being zero.
T he phenomenon illustrated by Example 14 .6 occurs because the symmetry o f the resistor values
precludes excitation by the current source. Moving the current source to a different position, say,
in parallel with one o f the resistors, or changing the value o f one o f the resistors to 0 .5 Q will result
in a nonzero transfer function.
E X A M P L E 1 4 .7 . This example looks at a simple initialized series
RL
circuit driven by a cosine
wave, as show^n in Figure 1 4 .1 5 . Let /^(^) denote the circuit response. Suppose
= Acos{t)u{t)
V and /y(0~) = 1 A. The objective is to isolate the transient and steady-state responses from the
zero-input and zero-state responses.
6 9 )
USO-) =
> m v
1 H
1Q
v.(t)
(a)
FIG U R E 1 4 .1 5
RL circuit
. (b).
for Example 14.7. (a) Time domain series
RL circuit,
(b) Frequency domain equivalent, accounting for initial condition.

S
o l u t io n
Using Figure 1 4 .15b and the principle o f superposition leads to the response
4s
T he term
the term
response
+ l ) ( r + 1)] is the Laplace transform o f the zero-state response, and
Y-^^{s)Li^{Qr) = \!{s + 1) is the transform o f the zero-input response. Thus,

is e~û{t). A partial fraction expansion o f the zero-state response yields
4.V
-2
(.s + l)(5 - + l)
5+1
the zero-input
2.V +2
r
+ 1
It follows that the zero-state response is -2e~^ti{t) + 2[cos(r) + sin(^)](/). Notice that both the zeroinput and the zero-state response contain a transient part, the part proportional to
e~^. A little
rear
ranging shows that the complete response is
i^{t)
-e~u{t)
+ 2[cos(r) + sin(/)]{r) A
which implies that the transient response o f the circuit is
e~û{t)
and that the steady-state
response o f the circuit is 2[cos(/) + sin(f)]/<(r) A.
Exercises.
1. An
RLC network
has transfer function
cos(?)(t) A is applied, then for very large
t,
H{s)
= l/(^ + 1). If an input
approaches a cosine wave o f what form?
A N S W E R : 0 .7 0 7 cos(/ - 45'")
RLC network has transfer function

e~^^û{t) A is applied, then for very large t,
2. An
A N SW E R : A constant
//(s ) =
\l{s +
approaches what?
1). If an input
= (1
'()()
natural response and

e^, where k is possibly
, in which case /. = j.
Many books on elementary circuits contain two other notions o f response: tlie
forced response.
To explain this we use the term exponent to mean
in
complex. For example, if sin(r) is part o f a response, then it comes from
Natural response: Fhe portion o f the complete response that has the same exponents as the
zero-input response.
Forced response: 'I'he portion o f the complete response that has the same exponents as the
input excitation, provided the input excitation has exponents different from those o f the
It would seem natural to try to decompose the complete response into the sum of the natural and
forced responses. Unfortunately, such a decomposition applies only when the input excitation is
(i) dc, (ii) real exponential, (iii) sinusoidal, or (iv) exponentially modulated or damped sinusoidal.
Further, the exponent o f the input excitation, e.g.,
a m f{t)
e'^ti{t),
must be different from the
exponents appearing in the zero-inpur response. T he natural and forced responses are properly
defined only under these conditions.
T he decomposition o f a complete response into a natural response and a forced response is impor
tant for two reasons. First, it agrees with the classical method o f solving ordinary differential equa
tions having constant coefficients, where the natural response corresponds to the complementary
function and the forced response corresponds to the particular integral. Students fresh from a
course in differential equations feel quite at home with these concepts. The second reason is that
the forced response is easily calculated for any o f the special inputs dc, real exponential, sinu
soidal, or damped sinusoidal. For example, if the transfer function is
H{s)
and the input is
then the forced response is simply
H{a)V^
(1 4 .6 )
To justify equation 14.6, note that the Laplace transform o f the input is
V!{s a).
Since the com
plete response is the sum of the zero-input and zero-state responses, we have
Complete response = [zero-inpnt response]
+iT '
The zero-input response terms all have exponents different from

second term,
\H{s) Vl{s- //)],
Hi s ) -----s- a
a,
the exponent o f the input. The
has only one term with exponent equal to
a.
fraction expansion o f this term yields
K
H{ s) ------- = -------- + [terms corresponding to poles of H{s)]
s - (I s - a
Using the residue formula to calculate A'leads
H{s)
s-a
s- a
to K = H{a)V.
Thus,
+ [terms corresponding to poles o f H{s)]
Executing a partial
and
L'^[H{s)VI{s - a)]
has a term,
which we identify as the forced response.
By using exacdy the same arguments, it is possible to show that if the input is a complex
exponential function
Ve, where
both Kand
are complex numbers, then the forced response is
simply
H{s^) \W
A complex exponential such as
is a mathematical entity. It cannot be generated in the labora
tory. However, the real part, Re[V^T (or the associated imaginary part), is simply an exponential
ly modulated sinusoidal signal, as shown in Figure 1 4 .5 , and is readily generated in a laboratory.
A derivation similar to the preceding leads to the conclusion that, if the input is Re[V^^, then the
forced response is
R e[//W K ^ 1
This relationship o f the input to the forced response prompts some textbooks to define the trans
fer function
H{s)
as the ratio o f the forced response to the input, under the condition that the
input is a complex exponential
Ve^. This,
however, is not natural and makes one wonder at the
applicability o f such a definition to the broad class o f inputs for which the transfer function is
most naturally defined, as covered in Chapter 13 o f the text.
4. CO M PUTATION OF TH E SIN U SO ID AL STEADY-STATE

RESPONSE FOR STABLE NETW ORKS AND SYSTEM S
Suppose that a transfer function
H{s)
models a
stable linear
circuit containing a total o f
n capaci
tors and inductors. In addition, suppose there are no comm on factors in the numerator and
denominator o f H{s) and that the
degree o f the denortiinator ofH{s) is ri. (This means that the effect
H{s).) T he goal o f this section is to develop the fol
o f each capacitor and inductor is included in
lowing formula: if H{s) satisfies the aforementioned assumptions, and the input to the circuit has
the form
A cos(to/ +
0 ), then the steady-state circu it o u tp u t response has the form
B cos(o)r
+ (}))
(14.7a)
where the
magnitude o f the
response is
B = A\H{jco)\
(14.7b )
and the phase
shift is
(j) = 0 + L H { f o )
Here we assume that to is some fixed, but arbitrarA^ value.
(l4 .7 c )
'
Chapter 14 Laplacc Transform Analysis III: 'Iransfcr Function Applications
From an input-output viewpoint, these formulas imply that the frequency response is the steadystate response o f a circuit to sinusoids o f varying frequencies. To construct this formula, suppose
H{s).
again that a linear circuit has a stable transfer function model

input is a s i n u s o i d w h o s e Liplace transform is
F{s)
Suppose also that the circuit
with zero-state response
Y\s), as
illustrated
in Figure 14.16.
F(s) ---------->1^
H(s)
-------- > Y(s)
Transfer Function
FIGURU 1 4 .1 6 Frequency domain representation of hypothetical circuit.
Since
H{s)
H{s) has
H{s) will
poles, p - <
is stable, all poles lie in the open left half o f the complex plane. Assume that
real, distinct poles labeled />j, ...
and complex poles labeled
-(tj
Consequently,
have a partial fraction expansion containing only two types o f terms; those having real
0, and those having complex poles with negative real parts, i.e.,
A\
A-)
H{s) = -----
^ + = -+
s -p \
S -P 2
...+
/4,,,
a- < 0.
Ci.v+D|
+ --------- 7 - 7
s - p , (i- + a , r + ( P
real poles with

negative real parts
Specifically,
T+ -"
i)"
(1 4 .8 )
complex poles with

negative real parts
It is easy to account for higher-order poles.

Suppose now that the circuit is excited by a sinusoidal input o f the form
/(/) =
Acoai OJ/ +
0) =
A-------------- ---------------
having Laplace transform
0.5 Aej^ 0 .5
^(v) = --------: + ---------:
.V - yco
s + jo i
Then a partial fraction expansion o f the Laplace transform o f the zero-state response, F(j) =
H{s)F{s),
has the form
A\
At
Y(s) = ----- ^
+
s - p\ s - P2
C|.v + D\
R\
R')
+ . . . + --------- ^
^----- T + . . . + ----- ^ +
(5 + a 1 j" + ( Pi j"
- yw
A--I- yco
real poles with

negative real parts
complex poles ivith

negative real parts
steady-state
contribution
=
In the steady state, i.e., for large
t,
the only residues o f interest are
and /?2 , because the part o f
the time response due to the other terms decays to zero with increasing
o f complex variables, we obtain
t.
By the usual methods
Chapter 14 Laplace Transform Analysis III: Transfer Funaion Applications
H(s)(s - j(o
J0.5Aej^
703
0.5Ae~j^\
jw
s + jm
S = J(0
= 0.5AH( j(n)e^ = 0.5A\H(
and
/?2 =
................... ( 0.5AeJ^
0.5Ae-j^\
H (s)(s-j(j})\ ------- ; + -------- ^ I

\ s - yo)
s + j(a }
S= J(0
= 0,5 AH (-j( 0 )e~-^ = 0.5A\H(-j(o)\e^^^~j'Û~j^
0.5A\H(-j(o)\e^^^^~j^^^^
But \Hi-p)\ = |//(;a))| and LH{-j(a) = -Z.H(j(a); hence,

/?2 = 0.5A\H(J(a)
^)
Consequendy, die Laplace transform o f the steady-state response when all initial conditions of the
circuit are zero is
s-
yo)
S+ jO)
In fact, this is the Laplace transform o f the actual steady-state response, provided that the zeroinput (nonzero initial conditions) response makes no additional contribution. The zero-input
response makes no contribution to the steady-state response when one or more o f the following
reasonable conditions on the circuit are met:
1 . The network has only practical passive elements, meaning that there are always stray
resistances present.
2 . The circuit may have active elements in addition to passive elements, but remains stable
in the sense that every capacitor voltage and every induaor current remains bounded for
any bounded circuit excitation.
3. The circuit contains a total o f n capacitors and induaors, and the stable transfer func
tion, H{s), has n poles.
Under conditions 1 through 3,
yj,t) = A\H{jwi)\ cos(o)f + (Z.//(/o)) + 0)) = 5 cos(cor + <j>)
704
Chapter 14 Laplace Transform Analysis 111: rransfcr Function Applications
Hence, if the input to the circuit lias the form

response has the form
0 +
B cos(ojr
A cos(o)r
+ 9 ), then the steady-state circuit output
+ (})), where the magnitude
The next question concerns the numerical calculation o f

uate
B = A\H{Jlo)\ and
the phase shift ({) =
//( .
H{s)
at
s = JiO.
B and
One method is simply to eval
W ith a calculator that easily accommodates complex numbers, this is quite
straightforward. An alternative method is to use the graphical technique o f the next section.
At this point, it is instructive to illustrate equation 14 .7 and at the same time compare it with the
phasor method studied in a first course on circuit theory.
E X A M P L E 1 4 .8 . In the circuit o f Figure 14.17, ji = 0.5 and
= cos(2r) V. Find
v^{t) for large t.
o l u t io n
P art 1.
Phasor method.
From the principles o f phasor analysis detailed in Chapter 10, the phasor
domain circuit o f Figure 1 4 .1 7 at (O = 2 rad/sec is given by the circuit of Figure 14.1 8 .
The single node equation lor
is
1_
=0
./4 - y
T he phasor solution to equation 14.9 is

V ,=
Therefore, for large
1 + yO.75
= 0 .8 Z - 3 6 .9 '
t
v^{t) = 0 .8
c o s ( 2 r - 3 6 . 9 ) V
(14.9)
P art 2 .
Laplace namform method.
OS
Tlie first step here is to construct the j-domain equivalent cir
cuit, which is given in Figure 14.1 9 .
The single node equation for
in the ^-domain is
V i-V ;-
V i-0 .5 V |
+
=0
2.V + 2
2
Solving for the transfer function yields
v,(s)
(1 4 .1 0 )
(.v + ir
From equation 1 4 .1 0 , at j = yto = j l ,
Hijco) = H{jl) =j4l{J2
+ 1)2 = 0 .8 ^ - 3 6 . 9
According to equation 14.7, it follows that
v^(t) = 0 .8
c o s ( 2 /- 3 6 .9 ) V
In this example, the two methods give the same answers, as expected. Since complex numbers are
easier to manipulate than rational functions, what is the motivation for such an analysis using the
transfer function
H{s)^ W hy
not stay with the phasor method? T he next example answers these
questions.
E X A M P L E 1 4 .9 . In Example 14.8 with the circuit o f Figure 1 4 .1 7 , let the value o f p be increased
to 1.5. Find y,(r) for large
S
t.
o l u t io n
P art 1.
Phasor method.
Since only the response for large
t is desired,
the problem appears to be one
involving sinusoidal steady-state analysis. The phasor domain circuit o f Figure 1 4 .1 8 yields the
single node equation
V, - 1
V, - 1.5V,
= 0
Solving again for V . yields

V, = ------ !------ = 0 .8 Z 3 6 .9
'
l - /).7 5
Therefore, for large
t,
vAt)
= 0.8cos(2^ + 3 6 .9 ) V
A beginner who has just learned sinusoidal steady-state analysis by the phasor method might
accept this answer. Unfortunately, the answer is not the voltage
v^{t) for
large r! T he reason is clear
from the Laplace transform analysis, which follows.

Part 2:
Laplace transforyn method.
From the ^-domain equivalent circuit o f Figure 14.1 9,

v j - y /- ^
25+2.
.9
Solving for the transfer function yields
His) =
%{s)
Viis)
-2s
(s-\ r
o f H{s) in the right half-plane, the circuit is unstable. As f becomes very large,
v^{t) approaches infinit)', instead o f 0 .8 cos(2r + 3 6 .9 ") V, as calculated by the
Since there are poles

the magnitude o f
phasor method.
This analysis demonstrates that the unstable behavior o f a circuit cannot be determined by the
phasor method. It is desirable to know when to use a particular method in order to avoid unnecessar)' complicated calculations. The following guidelines help:
1.
W hen the stability o f the circuit has been assured by some means, and co has a specific
numerical value, the phasor method is the better method to use for computing the
response for large
t,
which is also the steady-state response in this case. Circuits whose
stability is guaranteed include those with only passive elements, such as resistors, capaci
tors, and inductors; and amplifier circuits o f well-established configurations.
2.
The circuit is known to be stable, but CD is variable. In this case, the

rior to the phasor method. To say the least, we need only write
sL
H{s)
method is supe
instead ofy'coZ.. There
are other advantages to be gained from knowing the pole-zero plot o f
H{s)
that are not
possible with the phasor method. The examples o f frequency response calculations given
in the next section clearly demonstrate this point.
3.
If the stability o f the circuit is not yet determined, then
H{s) should
be calculated and its pole
locations checked for stability. Then step 1 or step 2 should be referred to, as appropriate.
Exercises.
1. Suppose a second-order linear circuit having the transfer function

/Y(^) =
V :Js)
S--0.5S + 5
5 - +0.5.V + 5. 7321
Chapter 14 Laplace Transform Analysis III; Transfer Function Applications
is driven by a sinusoidal input
j2cos{2t + 45)t^{t).
707
Show that the steady-state response
is given by
2. Consider the
L C circuit
o f Figure 1 4 .2 0 , in which
= 0,
= 0, and
= 100(r).
Show that the largest voltage to appear across the capacitor for r > 0 is 2 0 0 V. Hint: Show that
V(\t)
= 100(f) - 100cos(f)(f).
L= 1 H
v,(t)
C=1F
(a)
FIG U R E 1 4 .2 0 (a) Time domain
LC circuit.
(b) Frequency domain equivalent, accounting for initial conditions.
5. FREQ U EN CY RESPONSE
The frequency response o f a linear stable circuit having constant parameters characterizes the
input-output behavior o f the circuit to unit magnitude sinusoids, cos(oj/), as ca varies from 0 to
00
. This extremely important concept plays a major role in the analysis and design o f circuits and
systems. In terms o f the transfer function, the frequency response o f a stable circuit is the evalua
tion o f
H{$) zi s =jw.
In terms o f phasor analysis, studied in an introductory course, the frequen
cy response o f a circuit corresponds to the ratio o f the output phasor to the input phasor.
From the steady-state analysis perspective o f the previous section, if an input has the form
cos(ojQr), then the steady-state response (i.e., the response for large
out) has the form
cos(ojQr + (j)), where
t, after
all transients have died
= |//(/(Dq)| and the phase shift ({) = Z.//(/Wq). Here
|//(/Wq)| is the magnitude o f the complex number ^(/W q), and Z.A/(/0)q) is the angle o f the com
plex number //(/cOq). Thus, //(/(o ), for 0 s to < x , defines how a linear circuit adjusts the magni
tude and phase o f an input sinusoid to produce a steady-state output sinusoid o f the same fre
quency, but possibly with a different magnitude and phase.
An example o f practical importance is the specification o f a stereo amplifier. Here one specifies the
gain, gain-magnitude
\H{jLo)\,
to be more or less constant from 20 Hz to 2 0 kHz. W hy? Because
musical signals are composed o f sinusoids o f different frequencies. Accurate amplification o f the
music requires that all component sinusoids be amplified with equal gain.
OS
Frequency response: The frequency response of a stable circuit or system represented by a transfer
function
H{s)
is the complex-valued funaion //(/w ) for 0 s OJ < oo. The magnitude (frequency)
response is |//(^)| for 0
to < co, and the phase (frequency) response is
A complex-valued function
H{jw)
a-^ +
number. A complex number,
is a function such that for each value of to,

has a polar form,
(j), the phase angle o f the number. Moreover, if

( j + y /?i)( 2
+ y' ^ 2
for 0 ^ co < co.
Hijoi)
is a complex
in which p , is the magnitude and

another complex number, then
+Jb2^ =
and
a2 + p2
P2
In polar form, the frequency response as a function o f (O is

H(Ju)) = p(to)e-^^^'^^
where p(to) = |//(yto)| denotes the m agnitude response and

({>((0 ) =
is the
angle or phase o f
/LH( jio) -
tan
-I
the frequency response. As in other books,
magnitude response means
the
magnitude o f the frequency response. Typically, frequency response computation requires a calcu
lator or computer. We now illustrate the idea o f frequency response with two so-called band-pass
transfer functions in which a band o f frequencies is passed with relatively little attenuation while
frequencies outside the band are significantly attenuated.
E X A M P L E 1 4 .1 0 . Consider the two transfer functions
Q.25s
and
.v^-h 0 .2 5 5 + 1
^ _________________ 0 -0 6 2 5 . v^_________________
^
+ 0.35355.s-^ + 2.0625.V - + 0.35355.V + I
Using the M ATLAB code

nl = 0.2 5 *[1 0];
dl = [1 0 .2 5 1];
>>n2 = 0 .0 6 2 5 * [1 0 0];
>>d2 = [1 3 .5 3 5 5 e -0 1 2 .0 6 2 5 3 .5 3 5 5 e -0 1 1];
>)w =0.2:0.005:2;
hl = freqs(nl,dl,w );
h2 = freqs(n2,d2,w);
plot(w,abs(hl),w,abs(h2))
grid
xlabel(Frequenc}' r/s)
Chapter 14 Liplace Transform Analysis 111: Transfer Function Applications
"O ')
yIabcI(Magnitude response)
gtext(2nd O rder BP)
gtexr(4th O rder BP)
we obtain the magnitude response plot given in Figure 14.21,
O)
1/1
c
CL
t/i
O
)
u.
a;
-o
13
c
cn
(TJ
Frequency r/s
FIG U R E 14.21 Magnitude responses of a second-order and a fourth-order band-pass type transfer
function. Tlie fourth-order response has steeper sides (sharper cutoff) and a flatter top.
From the transfer function and by interpolation on the plots, one observes that at co = 0 the mag
nitude is zero and at co = x the magnitude is also zero; at (O = 1 the magnitude peaks, and this frequenc)' is called the center frequency. This is characteristic o f a band-pass type o f response. T he
fact that the response peaks at OJ = 1 rad/sec means that the transfer functions are normalized.
Transfer functions o f practical band-pass circuits have much higher center frequencies. Such fre
quencies can be obtained by the technique o f frequenq' scaling, taken up in the next section.
A very important system theoretic relationship is that o f the poles and zeros o f the transfer func
tion to the magnitude and phase responses. In iJie above example for
H^{s),
the pole-zero plot is
given by Figure 1 4 .2 2 . One immediately notices that the poles are very close to the point yl on
the imaginary axis with the magnitude response peaked. Further, the zeros at w = 0 and
where the magnitude response is zero.
co = x
are
710
Pole-zero plot
- I --------------------- !--------------------- 1--------------------- :
1^
0.8
:
0.6
0 .4
S. 0.2
:
0
-
0.2
-0 .4
-
0.6
:
0.8
-1
-1
-0 .5
0 .5
1.5
I--------------------- --------------------- --------------------- :
-1 .5
Real part
FIG U RE 1 4 .2 2 . Pole-zero plot of H^{s) in Example 14.10.
To emphasize this qualitative discussion we consider the following example.
E X A M P L E 1 4 .1 1 . Iwo circuits have transfer functions
H^{s)
and / / 2 W, with the pole-zero plots
shown in Figures 14 .2 3 a and b, respectively, and gain constants o f 1. Qualitatively speaking (with
out doing any com putations), what can we deduce about the magnitude response o f each circuit?
To verify our qualitative deductions we will use M ATLAB to construct the exact magnitude
response.
JCO
A
.. 12j
jO)
A.
X
- 12j
8j
8j
4j
4j
---------- >
II
-4j
-4j
-8j
- -8j
- .- 1 2 j
T
- .-1 2 j
>r
(a)
(b)
FIG U R E 1 4 .2 3 Pole-zero plots for Example 14.11. (a) H^{s) (b) Hjis).
11
o l u t io n
P art 1.
H^{s)
Qualitative analysis o f Figure 14.23a. This figure tells us that there are r\vo finite poles o f
\2j, but in the left half o f the complex plane, and that there is a finite zero at the ori
near
gin. So there is also a zero at y = x . Thus
\H{jto)\ - 0 as ( 0 - 0
\H{j(o)\ -* 0 as (O - oo
(iii) \H{jco)\ = maximum value
(i)
(ii)
as co - 12
In the case o f point (iii), we can say in general that in some neighborhood o f to near the pole, the
transfer function peaks in magnitude.
Qualitative atialysis o f Figure 14.23b. This figure tells us that there are again rvvo finite poles
Hjis) near 12y, but in the left half o f the complex plane, and that there arc two finite zeros at
P art 2 .
of
the origin. So there is no zero at ; = oo. Hence
\H{jco)\ - 0 as oj - 0
\H{jo))\ constant as to -
(iii) \H{jo})\ ~ maximum value
(i)
(ii)
oo
as co
12
Again we cannot make a stronger general statement in point (iii) above.

P art 3 .
Magnitude plots via MATLAB.
Suppose the poles are at -0 .1 y l2 and we desire the actu
al magnitude response plot for A/, (yco), 0 < co < 3 0 rad/sec. To construct the plot shown in Figure
1 4 .2 4 , we use the following M ATLAB code:
w = 0:. 1:2 0 ;
n = [1 0];
d = poly([-0.1 + 12*j -0 .1 -1 2 * )]);
h = freqs(n,d,w);
plot(w, abs(h))
grid
xlabel(Frequency in rad/sec)
ylabel(Magnitude H 1 (jw))
FIG URE 14.24 Magnitude frequenc)' plot, |//,(yo))| vs. to.
12
Exercises.
H^{s)
1. Suppose the poles o f
are at -0 .1 j\ l and that we desire the actual magnitude
response plot for 0 < to < 3 0 rad/sec. Use iMATLAB to construct this plot. Verif)' the accuracy of
the qualitative predictions.
2. Show that the transfer function
>
.V" +
H{s) =
o f an unknown circuit with
R=
1
s~ +
LC
.y+
RC
(1 4 .1 1 )
LC
Z. = 0.1 H , and C = 1 mF has the band-reject magnitude
response depicted in Figure 14 .2 5 . Use M ATLAB and the freqs command.
Frequency in r/s
FIGURH 1 4 .2 5 Plot of the magnitude response of the transfer function of equation 14.11.
The example below further illustrates this relationship with a pedagogicaily useful graphical tech
nique. Mastering this technique helps concretize the meaning o f magnitude and phase and rein
forces the qualitative discussion above on using pole and zero locations to compute the magnitude
and phase.
E X A M P L E 1 4 .1 2 . 'lb better grasp the ideas o f the magnitude, |/y(/'to)|, and the phase, Z.//(ja>),
o f a frequency response, suppose a transfer function has the form
H{s) =
where Zj =
2j and
= -2j.
(s-Zi)(s-Z2)
is+]){s + \ + j ) { s + \ - j )
Figure 14 .2 6 a shows the pole-zero plot o f
(1 4 .1 2 )
H{s).
Chapter 14 Laplace Transform Analysis ill; Transfer Function Applications
~ 13
JO )
JO )
()
2j
-1
X
-J
-2j
(a)
FIG URE 14.26 (a) Pole-zero plot
of H{s) as given
by equation 14.12. (b) Measuring distances
graphically from zeros and poles to the point j\.

S
o l u t io n
The plan o f this example is to compute the magnitude o f

pute
\H{J\)\. This
H{jio) graphically
for oj =
, i.e., to com
computation entails the following steps.
Step 1. Observe that j\
plex plane. Similarly,yi -
defines a complex number. Think o f J\
Zj and J] - p-, where /> ] = !
, as a vector in the com
and p 2 ,^ = - 1 7 . define vectors. Each vec
tor has a length that can be determined either graphically, by physically measuring the distance
with a ruler, or by the Pythagorean theorem. Figure 14 .2 6 b illustrates the idea.
Step 2 . Following from step 1, the magnitude
\j\ -y'2| x
N y i)| =
o f //(/I )
has the form
1/1 +y2|
1x3
I./1 + l|x 1/T -h 1 + ./|x 1/T + 1 - ./ I
= 0 .9 5
V I X V5 X 1
H{j\) graphically.
- z - ory ( 0 - p- can be
Step 3. Suppose we wish to compute the phase or angle o f
In Figure l4 .2 6 b ,
observe that each complex number viewed as a vector jio
represented in the
form
x^\
where ip is the angle the vector makes with a horizontal line passing through its base.
For example,
{j\ - J2)
= -j\
From basic complex number theory, the angle o f the prod
uct o f two complex numbers is the sum o f the angles, and the angle o f the ratio of two complex
numbers is the angle of the numerator minus the angle o f the denominator. Fience, from the
angles shown in Figure 14 .2 6 b ,
LH ijl) = L { J \ - jl)
L{j\
+jl)
- a j\
+ 1) -
= _ 9 0 " + 9 0 - 4 5 - 0 - 6 3 .4 = - 1 0 8 .4
Uj\
+ 1
Uj\
+ 1 -j)
To extend the ideas o f Example 1 4 .1 2 , let
(1 4 .1 3 )
{s - P\){S - P2 )---{s- P)
Because the magnitude o f a product o f complex numbers is the product o f the magnitudes o f the
numbers, and because the magnitude o f the ratio o f two complex numbers is the ratio o f the mag
nitudes o f the numbers, the general form o f the magnitude response o f equation 14.13 is
Zx
|yto-C2
- 1>\ j M - f h
(1 4 .1 4 a )
,/ w - /^
Similarly, since the angle o f the product o f complex numbers is the sum o f the angles o f the num
bers, and since the angle o f the ratio o f two complex numbers is the difference in the angles o f the
numbers, the general form o f the phase response o f equation 1 4 .1 3 is
= [Z-(/to -
Thus, qualitatively speaking,
H{jn))
Zj) + + /-(/CO M jco-pj) + ' + /- ijco-p^)]
+ /-(yw -
[Lijco-p^)
LK\
(1 4 .1 4b)
tends to have a large magnitude for jcos near poles and a small
magnitude for jcos near zeros. As mentioned earlier, this can be used to advantage in estimating
the magnitude response and phase response o f a transfer function.
Exercise.
Draw an estimate o f the general shape o f the magnitude and phase response o f the
Butterworth normalized low-pass transfer Kinction
M i> =
5 +
Jls
-I- 1
Com pute the exact magnitude and phase at lo = 1. W hat happens to the magnitude and frequen
cy response if H{s) is changed to
H^{s)
H{sl\Q)
and
H^{s)
= / / ( / / 1 00)?
A N SW ER S: in random order: - 4 5 . O.TO"^, the general shape is the same wich
= /7 i(/]()0 )
= 0 .7 0 7 .
6. FREQ U EN CY SCALIN G AND M A G N ITU D E SCALIN G

Design of filters and amplifiers often begins with a design template in which almost all parameter
values are normalized. In particular, a source or load resistance is often set to 1 t l . Also, a critical
or im portant frequenq^ is set to 1 rad/sec. Such circuits are called
normalized.
W ith the comple
tion o f a normalized design, engineers can frequency-scale to obtain realistic frequency responses
and magnitude-scale to obtain reasonable impedance levels as necessary to meet power and energ)' restrictions.
FR EQ U EN C Y SCA LIN G
E X A M P L E 1 4 .1 3 . The circuit o f Figure 1 4 .2 7 realizes the transfer function o f equation 14.11
from the previous section for /? = 1 Q , Z, = 0.1 H , and C = 1 mF. We pose the following question.
Suppose
->
----- , C
Xf
and
R remains
the same in the circuit; what happens to
H{s)
and the frequency response plot?
FIG U R E 1 4 .2 7 Band-reject
type circuit
that realizes the transfer function of equation 14.11.
S O L U T IO N
Step 1.
Calculate the circuit transferfunction.
H{s) =
R +Cs
Step 2 .
By voltage division,
LC
RC
------
LC
Ls
Incorporate the effect o f /y.R
R
R+
R+
s +
s\
Kf
s V
LC
We conclude that when
^ andA T
C
K
with
R unchanged,
the new and old transfer functions are related in a ver>' simple way:
\
.y
A'
I )
Step 3. Plot the magjiitude response. Rather than go directly to M ATLAB, consider that
to
Kf I
ôld
Table 14.1 depicts this relationship for specific values o f id.

TABLE 14.1
Transfer Function Relationship
to (rad/s)
0
I
1
~ ôld J
\ ^ Jf /
^ newU^f
Xf
2f
)~
^ôUl ( / " ^ l = H .,u(j)

\f]
\
' 2K j
ôld
= w (y 2 )
/
In general, the new transfer function is related to the original
s by s!K^. T he
H^[J,s)
through the replacement of
implication o f this relationship is that whatever happened at ^ =
must now happen at
s=
where
the term
before scaling
frequency scalhig.
In terms
o f the plot o f our specific transfer function, then, we obtain from Figure 14.25 the new plot o f
Figure 14. 28 by inspection.
F I G U R t 1 4 .2 8 T h e new m ag n itu d e respon se after freq u en cy sca lin g by
Kj-
Chapter 14 Laplacc Iransform Analysis 111: Transfer Function Applications
For general linear circuits, vvc will only
state the
procedures for scaling and their effects, leaving a
rigorous mathematical justification to a more advanced text on circuit theory.

Procedure for frequency scaling: To frequency-scale a linear network by a positive scale factor
1.
Divide all inductances and capacitances by / y
2.
Leave all resistance values and controlled source parameters unchanged.
Kj-
Effect o f frequency scaling:

1-
which means that

3.
has zeros and poles at
KjrZ- and
The magnitude and phase response cur\'es o f
for / = 1, . . . , w and
are those o f
k=
with the fre
quency' scale multiplied by / y Conversely, the magnitude and phase o f //(/to ) are the
same as
Exercises.
1. Fill in the details o f the derivation o f

/
2. Given
point 3 above.
in step 2.
K f Z \ ) " { s K fZf/i)
--------- ---------------------------U- K
- K^p)
\ //-m { s
= KlK r)
3. Given the circuit transfer function
H(s) = K
, verify the converse statement in
(5 - /? l)( i -/> 2 )
find the new pole and zero locations if the circuit is frequency-scaled by
1000.
AN SW I-RS: l0{)0/>,. !()()()/> and HHHb,
E X A M P L E 1 4 .1 4 . T he circuit o f Figure 1 4 .29a has transfer function
OAs
i* + 0.1
Figure 14 .2 9 b shows the pole-zero plot and Figure 14 .2 9 c shows the magnitude response with
peak value at
= 1 rad/sec.
(a) Frequency-scale the circuit by the factor /y = 10^ and compute the new transfer function.
(b) C om pute the new pole-zero plot and the new magnitude response curve.
/YY\
-I-
V J
>
0.1 Q >
(a)
718
JCO
FIG U R E 14.29 (a) Series resonant circuit, (b) Pole-zero plot, (c) Magnitude response.
S
o l u t io n
(a) Dividing the capacitance and inductance by
Kj-=
10*^ yields the circuit o f Figure 14. 30a.
JO )
(-0.05+ j0.999)10^
X
for
Hnew(s)
o
{-0 .0 5 -j0 .9 9 9 )1 0 "
(b)
FIG U R E 1 4 .3 0 Demonstration of the effects of frequency scaling, (a) Frequency-scaled circuit, (b)
New pole-zero plot, (c) New magnitude response curve.
Using effects o f frequenc)' scaling property 2 , the new transfer function is
10^/
10^ /
/ 5 ^
( S
+ 0.1
+1
lio^/
UOî
(b) The ideas stated in part (a) are borne out by comparing the pole-zero plot and the magnitude
response curve for
shown in Figures 14. 30b and c, with their counterparts for
Figure 14 .2 9 . Notice that the pole locations have been scaled by
Kj-^s
H{s)
in
per property 2. A direct
computation in M ATLAB verifies this:

pnew = r o o t s ([l/le l2 0 .1 /l e 6 1])
pnew =
-5 .0 0 0 0 e + 0 4 + 9 .9 8 7 5 e + 0 5 i
-5 .0 0 0 0 e + 0 4 - 9 .9 8 7 5 e + 0 5 i
pold = roots([l 0.1 1])
pold =
-5 .0 0 0 0 e -0 2 + 9 .9 8 7 5 e -0 1 i
-5 .0 0 0 0 e - 0 2 - 9 .9 8 7 5 e -0 1 i
% Multiply pold by Kf and check with pnew
le6*pold
ans =
-5 .0 0 0 0 e + 0 4 + 9 .9 8 7 5 e + 0 5 i
-5 .0 0 0 0 e + 0 4 - 9 .9 8 7 5 e + 0 5 i
Also, in Figure 1 4 .2 9 c for
3 0 c for
H{s),
the peak response occurs at co^/^= 1 rad/sec, whereas in Figure 14.
the peak response occurs at
= 10^ rad/sec.
At this point we round out our discussion o f frequency scaling by relating it to the time domain
via the time/frequency scaling property o f the Laplace transform:
or, equivalently.
K fl
=H
( s \
\ j /
where H{s) is the circuit transfer function and h{t) is the circuit im pulse response.
'2 0
E X A M P L E 1 4 .1 5 . Consider the transfer function
H(s) =
30
10
10
(5 + 2 ) ( 5 + 5 )
{s + 2)
(5 + 5)
whose impulse response is
h{t)
\Oe-">)u{t)
Find the impulse response when the transfer function is frequency-scaled by

S
Kj:
o l u t io n
T he scaled transfer function is

10^ ,
f-lneAs) =
(5 + 2 /r y )(5 + 5/C|)
ioa :,
(^ + 2A:/-)
(5
/:^ )
Taking the inverse Laplace transform yields

/W (0 =
- \QKfe~^Âu{t)=Kf h[Kfi)
Suppose for the sake o f argument that / y = 10^. Let us plot

Figures 1 4 .3 la and b.
t In mllli-sec
(a)
h{t)
and
as has been done in
^21
t in milli-sec
(b)
FIG U R E 14.31 Illustration of the effect o f frequency scaling on impulse response.
(a) Original h[t). (b) h^J,t).
Observe that the plots are structurally the same. However, the time scale in Figure 14.30b is
in milliseconds and the magnitude has been scaled by
Exercises.
Kj~=
10-^.
1. Use the Laplace transform time scale propert)' to show that in general
2. In Example 14 .1 5 , suppose the poles are changed from - 2 and - 5 ro - 3 and - 6 , respectively.
Suppose further that the transfer function is scaled by
10. Find the impulse response and the
new impulse response after scaling.

A N SW ER : /;(/) -
- 10^--^0//(/) and
h,
(/) = (lOOt"-^"' - I
M A G N ITU D E SCA LIN G

Frequency scaling has allowed us to obtain realistic frequency responses from normalized circuits.
Another technique in achieving a realistic design is magnitude or impedance scaling. A simple
example illustrates the idea.
E X A M P L E 1 4 .1 6 . Figure 14 .3 2 a shows a normalized voltage divider circuit having voltage ratio
ôuJîn = ^-5
= 2Q.
and consisting o f two 1 Q resistances. T he input impedance is
Suppose
both resistances are made 1 000 times larger, as shown in Figure 14 .3 2 b . W hat happens to the volt
age ratio and input impedance?
lOOOQ
1O
------0
+
10000
1O
-O
(a)
(b)
F K iU R E 1 4 .3 2 (a) A normali7.ed voltage divider circuit, (b) The voltage divider of part
(a) magnitude-scaled by 1000.
S olution
By direct calculation, the voltage ratio remains unchanged, i.e.,
1000
out
\ V-in
I new
10004-1000
=0.5 =
out
\V
''in
old
T he new input impedance is

z/;f"'(.y) = 1000 X 1 -h 1000 X 1 =
which is 1000 times larger. T he elements in Figure 14 .3 2 b are said to be
magniti(de-scaled (by
1000) from those o f Figure 14 .3 2 a .
The above example motivates a more general discussion. Suppose each impedance in Figure
14 .3 3 a is scaled (multiplied) by
to yield the circuit o f Figure 14 .3 3 b . As in Example 14.16,
the voltage ratios remain the same for both circuits:

1
()
Further, the input impedances o f Figures 14 .3 3 a and b are related as
4 , (,) = K,Z^ W
K,Z2U) = K,Z (,v)
Similar to the case in Example 1 4 .1 6 , the input impedance is increased by the scale factor
The
network o f Figure 14 .3 3 b is said to be obtained from that o f Figure 14 .3 3 a by magnitude scaling

with scale factor
tance is
=
If Zj is an inductance, then
mtiltipliedhy K^.
= sl{CIK^, i.e.,
Z^{s)
Ls and
On the other hand, if Z , is a capacitance, then

the capacitance is
dividedhy
i.e., the induc
Z^{s) =
1 /G and
723
Z,(s)
K .Z ,(s)
-o
+
V r A
J
---w
Z
Z,(s)
-o
(b)
(a)
FIG U R E 1 4 .3 3 . (a) A general voltage divider, (b) Magnitude scaling of the circuit in part (a) by
T he above discussion suggests that magnitude scaling can be achieved by multiplying resistances
and inductances by
and dividing conductances and capacitances by
This and several other
properties are stated next.

P rocedure for m agnitude scaling: To magnitude-scale a linear network by a scale factor
K^.
1.
Multiply all resistances and inductances by
2.
Divide all capacitances and conductances by
3.
For current-controlled voltage sources (C C V S ), i.e., the
4.
For voltage-controlled current sources (V C C S), i.e., the^^^ typ^> divide the parameter
5.
by
Parameters for voltage-controlled voltage sources and current-controlled current sources
type, multiply the parameter
remain unchanged.
6.
Ideal operational amplifiers remain unchanged.
The effect o f magnitude scaling on a transfer function is set forth below, with its verification left
for more advanced courses on circuit theory.
Effect o f m agnitude scaling on transfer functions: If
magnitude scaling has no effect on
H{s).
H{s)
is a voltage ratio or current ratio,
If H{s) has units o f ohms, then the magnitude-scaled net
If H{s) has units o f siemens or mhos, then the magnitude-scaled net
work has
work has
E X A M P L E 1 4 .1 7 . The series circuit o f Figure 14.34a has input impedance

,
Z{s)
I
s 2 + 0 .I ,v + l
= i-5-k 0.1 =
and transfer function
0.l5
v;-,,
(a) Magnitude-scale the network by

(b) Calculate
and
Z(5)
.r-hO.I.y+1
= 1000.
to verify the effects stated above.
/24
Chapter 14 Liplace Transform Analysis III: Transfer Function Applications
_JYY\
Im F
V
1000 H
V.
Znew(s)
(b)
FIG U R E 1 4 .3 4 Magnitude scaling, (a) Original network, (b) Network magnitude scaled by 1000.
So
lu t io n
(a) Following the procedure for scaling
Rs, L's,
and C s, we obtain the scaled network shown in
Figure 14 .3 4 b .
(b) For the simple circuit o f Figure 14.34b , the new input impedance is
Ze,,.(s) =
+ 1 0 0 0 s + 1 0 0 = 1 0 0 0 ^ + 0 - 1 ^ + ^ ^ l O O O Z (s ) =
K,Z{s)
By the voltage divider formula, the new transfer function is
100______________ l l i
1000
.y + 0 .h + l
.V-+0.1.V + 1
= H(s)
Because the transfer function is a voltage ratio, there is no change. These results clearly illustrate
the stated effects o f magnitude scaling.
Exercises.
1. Magnitude-scale the circuit o f Figure 1 4 .34a by
= 5.
A N SW ER : O.S 12. S H, 0 .2 12. Two parallel resistors
and
/? 2
are magnitude-scaled by
Hint. Use the formula for parallel resistance to compute

that to
Verify that
old
Rf,q =^mêq
and then relate
COMBINED MAGNITUDE AND FREQUENCY SCALING

Moving from normalized circuit design to realistic circuit design ordinarily entails both magni
tude and frequency scaling. This subsection provides several illustrative examples.
EXAM PLE 14.18. The amplifier circuit shown in Figure 14.35a consists of r\vo stages. The first
stage is a Sallen and Key low-pass active filter. The transfer function of the Sallen and Key circuit is
T he gain o f the second stage is
H2(s ) =
V
Ôllt____-'>
V,
.v +
Therefore the transfer function o f the amplifier is
His) =
V
Ôlll _ V
* out. V
^(1^ =
...........................
y.
y y
-1
- " ^ 1
------------------------------------------------------------ {s + 1) (.v + ^ + 11 5 ' + 2 . r + 25 + 1
FIGURE 1 4 .3 5 A third-order Butterworth low-pass filter, (a) The original network, (b) The scaled
network with
K^= 2000ti
and
= 10,000;
= 3 1.4 uS.
The overall transfer function o f the two-stage amplifier is a third-order, maximally flat
(Butterworth) low-pass filter, which will be studied in Chapter 19. The magnitude response o f the
transfer function has a 3 dB down frequency at OJ = 1 rad/sec, or l/(2ji) Hz = 0 .1 5 9 Hz an
extremely non-useful audio frequency. In fact, when the 3 dB down frequenc)' is at (o = 1 rad/sec,
the magnitude response is said to be
normalized.
N ot only is 1 rad/sec not useful, but the element
values are unsuitable for practical applications. However, using both magnitude and frequency
scaling, this reference or normalized amplifier circuit can be made into a very practical filter.
Suppose we wish to have the 3 dB down frequency a t^ ^ ^ = 5 0 0 Hz and the largest capacitor at
10 nF. Such a filter could be used to direct the low-frequency content o f a music signal to a woofer.
O ur goal requires that we frequency-scale the circuit by
Kj-= IOOOtt and
the 2 F capacitor (the largest) becomes 10 nF; i.e.,
must satisfy
10 X K )-'' =
Solving for
K jK ,
lOOOrcK,,,
magnitude-scale such that
yields
K^,
m = 6 .3 6 6 X lO'^
T he scaled circuit meeting the requirements is shown in Figure 14 .3 5 b , where
Exercise. A circuit has transfer function H{s) =
(5")
cs
= -^ ------------- ,with
Avj
s + fl.v + h
After both magnitude and frequency scale changes o f

and
= 4 and
= 3 1 .4 |iS.
a = Q.\, b = Aand c = ^ 0 .
Kjr= 2,
compute the new
a, b,
c.
A N SW E R : in random order: - 3 2 0 . 0 .2 , and 16
Examples 1 4 .1 5 and 1 4 .1 6 illustrate one o f several reasons for scaling a linear nersvork. Chapter
19, on elementary filter design, will derail further applications.
7. INITIAL- AND FINAL-VALUE THEOREMS

In system theory, and especially in control theory, engineers want the output signal o f a circuit or
system to track a given reference signal. T he idea behind the term
nack
is that for large t, the ref
erence signal and the circuit output are more or less indistinguishable. To accomplish this,
design/control engineers generate an error signal,
put and
e{i) = y(i) ~ yref^)^ where
is the circuit out
is a given reference signal. Since much o f the analysis is done in the frequency
domain, one often knows
E{s)
Y{s) -
without knowing the related time functions.
Ordinarily, the design engineer needs to know the initial error, if(0), and the final error, f(co).
Available to engineers are the initial-value theorem and the final-value theorem, which permit the
computation o f these quantities in the frequency domain.
Initial-value theorem: Let
= F{s)
be a strictly proper rational function o f j, i.e., the numer
ator and denominator o f F a r e both polynomials in
s,
with the degree o f the numerator less than
that o f the denominator. Then

lim 5fCO = / ( 0 ^ )
Proof: The.
quantity x/'(y) suggests a derivative operation
sF {s)-f(Q -)= L
on j{t).
y/(0
dt
(1 4 .1 5 )
Specifically,
/ 27
Applying the definition o f the Laplace transform to the right-hand side and taking limits as
approaches infinity implies that

l i m [ j f ( .v ) - / ( ( ) " ) ] =
lim
s-*
,V - X
(1 4 .1 6 )
lim C h O e -^ 'd t
v-oc J O
-t- lim
S-X
C h t)e-^ 'd t
J{)
where the dot over the function/(r) indicates the derivative o f the fimction. Observe that
fiOe'^'dt
lim
S -*
and that, because
/(/)
Wm (e~")dt = 0
30
is continuous at r = 0,
lim
.V-*30
lim
j! r
s-*x
e -.90 f j . /(r)(/f = / ( 0 " ) - / ( 0 ).
(1 4 .1 7 )
Hence, the left-hand side o f equation 1 4 .1 6 equals the right-hand side o f equation 14.17.
Equation 14.15 follows from equating these two sides and canceling th e /(0 ) terms in both.
E X A M P L E 1 4 .1 9 . T he Laplace transform o f a capacitor voltage is given by

2
.V
5i + 2
Find the initial capacitor voltage

S o l u t io n
By direct application o f the initial-value theorem,

Vf-fO
EXAMPLE 14.20.
)= lim
i->oo
.vV(^('.yJ=
lim
2-
= 2 - - = 1.8 V
5
Let the Laplace transform o f the velocity of a certain projectile be given by

,
500^ + 20
V{ s) = ----------------------------i-(5 5 -h 2 0 )(1 0 .v + l)
Find the initial velocity, y(0'^), and the initial acceleration, //(O"^).
S o l u t io n
l b find the initial velocity, we directly apply the initial-value theorem:

v(O^) =
lim sV(s) = lim

S-^ o o
s * x
5 0 0 ^ -f 2 0
(55 + 20)(10. v+ 1)
=0
Since acceleration is the derivative o f velocity, and the initial velocity is zero, from the time dif
ferentiation propert}'^ o f Table 12.2, assuming the velocity is continuous at f = 0,
2S
/\(.v) =
sV{s) -
500.V + 2 0
v(o^
(5.s + 2 0 )(1 0 .v + l)
From the inirial-value rheorem in standard units,
20
f/(0 ) =
Urn
.V-
5 () 0 i- +
20 s
(5.V + 20X10.S + 1)
Urn
i-ôo
5.V +
20 \
s!
.y /
Exercise.
Suppose
F{s)
= (8j +
2)1{Isr
10
IO.V+-
+ 85 + 3). Fin d y(0^).
.\N SW ER : 4
T he initial-value theorem and the examples that follow it illustrate the computation o f initial val
ues. To compute final values vve use the next theorem.
Final-Value T h eorem : Suppose
F{s) has
poles only in the open left half o f the complex plane, with
the possible exception o f a single-order pole at ^ = 0. Then
lim sF{s) = Urn / ( / ) .

(1 4 .1 8 )
Proof:T\\Q condition
o f the theorem, i.e.. the condition that
F{s)
has poles only in the left half of
the complex plane, with the possible exception o f a first-order pole at the origin, guarantees that
the limit on the right side o f equation 1 4.18 exists. This is because a partial fraction expansion of
F{s)
leads to a time fu n c tio n //) that is a sum o f exponentially decaying signals and at most one
constant signal. Since the right-hand limit is well defined,

/l//l . v F ( .v ) - / ( 0
s-^0
) = /mi r " / ( f k - ^ ' c / r

s-*Q JO
= { lim
\/-X
/ ( / ) ''- / ( ( ) ) .
/
This implies equation 14.18.
E X A M P L E 1 4 .2 1 .
As in Example 1 4 .2 0 , suppose the velocity,
v{t),
Laplace transform
V ( .) =
500^ + 20
.v(55 + 20)(10.v-h l)
with the Laplace transform o f the acceleration, a(r), given by

500.V + 20
( 5 i + 2 0 ) ( 1 0 .v + l )
Find the final values o f iÛ) and a(r) if possible.
o f a certain projectile has
29
S o l u t io n
K(y) and
A{s)
have poles that meet the conditions o f the final-value theorem. Hence,
lUn \'(t)= Urn ^ y ( 5 ) = lim

I-*X
.v-^0
,v-*-0
5005+20
= 1
(55 + 2 0 X 1 0 5 + 1 )
and
lini a(t)= lim

i-* x >
5y\(.0 =
.v-()
lim
.v-ô
5 0 0 5 + 20 5
= 0
;5 5 + 2 0 )( 1 0 5 + 1)
Observe that a constant final velocity implies a zero acceleration as these expressions indicate.
E X A M P L E 1 4 .2 2 . W hat if the conditions o f the fmal-value theorem are not met? W hat would
go wrong? A simple example illustrates the problem. Let
Fis) = -
5 + 1
which corresponds
toJ{t)
= sin(r)//(r). Then
lim
sF{s) =
lim ^ 5 ----- = 0 ,
1
,v-*0 5 " +
but
lim
/{[) =
lim
t-*x
sm(t)ii{t)
is undefined, i.e., it does not exist. The theorem, however, presupposes that both limits exist.
Again, the condition of poles in the leh half complex o f the plane with at most one pole at the
origin is necessary and sufficient for both limits to exist.
Exercises.
1. If
F{s)
= (6y + 10)/(2^^ + 4^), then, by the final value theorem,
value for large r?

A N S W E R : 2.5
2. The Laplace transform o f a signal,
5y{t- 2)u{t-
2), is
1 5 t ^ - - - '( 5 - + 5 - 2 )
.v(5" + 5 5 + 6 )
Find the value of^(f) for very large

A N SW E R : - 5
t.
f{t)
approaches what
"30
8. BODE PLOTS
Section 5 described the use o f the poles and zeros o f N(s) to compute the frequency response o f a
circuit. In this regard, Hendrik Bode developed a technique for computing approximate or asymp
totic frequency response curves. These so-called Bode plots can be quickly drawn by hand. A
description o f the technique requires the introduction o f some terms widely used in the engineer
ing literature.
Let //(s) be a transfer function that is a dimensionless voltage ratio or a current ratio. As explained
in section 4 , for sinusoidal steady-state analysis, one replaces s by yoj to study the circuits magni
tude response, |//(/w)|, and phase response, Z .//(/co). For convenience, let |//(/to)| be a voltage
gain, |K,/Kj|. T he ga/;/
in dB
(decibels), denoted by
is defined by the equation
= 2 0 log,|//(/o)|
(14.19)
For convenience, whenever we write log(.v), we will mean log,Q(>;). Solving for |//(/w)| in equation
1 4 .1 9 yields the inverse relationship
(1 4 .2 0 )
Table 14.1 presents some pairs o f |/y| and
we may say that
is 2 0 dB
above
Thus, instead o f saying that |K,| is 10 times |Kj,
K,, or that Kj is 2 0 dB
below 2 - Similarly,
to say that
V-,
is 3
dB above K, means that |K,| is 1.414 times |Kj|.

TABLE 14.1 Transler Function Gain in Magnitude and in Decibels
^dB
Exercise.
V2
10
100
1000
s3
^6
20
40
60
A certain amplifier has a dc gain o f 80 dB. W h at is the actual voltage gain?
A N S W E R : H)-'
O ne of the reasons for using the dB terminology is that it simplifies the analysis and design of
multistage amplifiers. Suppose an amplifier has three stages with voltage gains equal to 2 0 , 100,
and 10, respectively. The overall voltage gain is the
product o f
the gains of each individual stage,
which is 2 0 X 100 x 10 = 2 0 ,0 0 0 . Using the dB specification, the overall gain in dB is the
sum of
the dB gains o f the individual stages. This is (2 6 + 4 0 + 2 0 ) = 8 6 dB. It is easy to justify this claim.
First,
|H| = |//,| X |//,| X |//,
731
Taking the logarithm o f both sides and multiplying by 2 0 yields

20 log|/^ = 20 log|//J + 20 logl/Zjl + 20 log|//3|
or
^dB
^2,dB + ^5,dB
This summation has pronounced advantages for repetitive calculations at many frequency points,
as when plotting a magnitude response such as equation I4 .l4 a . We could, of course, convert this
equation to an equation having all terms in dB. However, with an eye toward a further simplifi
cation, it is desirable to first rewrite H{s) in a slighdy different, but equally general, form, namely
+ 1
H(s) = Ks' j
'
(14.21)
-Z 2
^ - + l'
V -P i
where \p^, p 2 >} are those poles o f H{s) that are not at the origin and {zp 2^, ...} are those zeros
of H{s) that are not at the origin; if a is positive (negative), then H{s) has a zeros (poles) at the
origin. For example, a transfer function
(^ + 5 0 ) (5 + 2 0 0 )
(5 + 8)( j + 6 0 0 )
has the equivalent form
O '
H {s) = 2.5
^50
+ 1
Uoo
(f-l ^,600
Observing that
+ 1
(14.22)
+ 1
= 10 log^^|//(70))p, setting s = /a) in equation 14.21, and noting that the mag
nitude squared o f a complex number is the imaginary part squared plus the real part squared yields
0.5
HdBi(o) = \K dB
0)
(0
dB +
dB
2
,0.5
0.5
v2
\Z2/
+ 1
dB
(14.23)
-,0.5
( co\
dB
Equation 14.23 suggests that we may compute the dB vs. (O curve for each term on the right-hand
side and graphically sum the curves to obtain the desired H jg vs. (O curve. However, each indi
vidual curve is reasonably sketched by using log(o)) instead o f 0) as the independent variable. This
amounts to using semi-log paper to plot the dB vs. (O curves. Such a plot, with a linear scale for
the dB values on the vertical axis and a logarithmic scale for O) on the horizontal axis, is called a
Bode magnitude plot, in honor o f its inventor. Similarly, a plot of LHijiiS) vs. O), with a linear
scale for Z.//(/co) and a log scale for co, is called a Bode phase plot. Note that, because of the log
arithmic scale for the (o-axis, the actual distance on the paper between (O = 1 and O) = 10 is the
same as that between o) = 0.1 and O) = 1. (See Figure 14.36a.) Note also that the (O = 0 point will
not appear on the graph, because log(co) approaches -oo as O) approaches 0.
W
"3 2
Chapter 14 LiplaceTransform Analysis III: IVansfer Function Applications
With log(oj) chosen as the independent variable, the plot of each term in equation 14.23 either is exact
ly a straight line or is a curve having two straight line asymptotes. I'his is illustrated in Figure 14.36.
FlG lIR Ii 1 4.36 Bode magnitude plots for three basic transfer functions,
(a)
H{s) = K> 0.
(b)
H{s)
sin, a > 0.
(c)
H{s) - s!a +\, a > Q.
733
In Figures 15.36b and 15.36c, the rising asymptote has a slope of 20 dB/decade, which means that
along this line, an increase in frequency by a factor of 10 causes an increase in gain of 20 dB. Here,
the word decade (abbreviated dec) simply means 10 times. Another way to express the same
slope is to indicate the increase of gain in dB when the frequency is doubled, or increased by an
octave, in music terminology. It is easy to see that 2 0 dB/dec is equivalent to 6 dB/octave. In gen
eral, if a frequency, (Oj, is d decades above another frequency, coj, then, by definition, (cOj/cOj) =
lO'^. Conversely, if (0)2/t0 j) = r, then we say that (Oj is log(r) decades above (Oj.
In Figure 14.36c, the left asymptote is a horizontal line and hence has a slope of 0 dB/dec. The
point where the two asymptotes intersect is called the breakpoint, and the corresponding fre
quency is called the l>reakfrequency or comer frequency.
The derivations o f the true curves and asymptotes in Figures 14.36a and 14.36b are very simple
and are left as exercises. For figure 14.36c,
in which case
+1
+ 1
and
| H ( | ^ =
For 0)
201o g
( l) '
+1
a.
|H(j<o)|jB = 201o g [-y r] = 0
indicating that \H{j(a)\^g approaches the left asymptote in the figure. For co
|Oo))|dB = 201og ,1 - )
V\a )
(I)'
+1 as 20 log -
a,
=201og(co)-201og(a)
indicating that |^(/(o)|^^ approaches the right asymptote in the figure. The two asymptotes inter
sect at the point ( cd = a,
= 0). At this corner frequency, the largest error, 3 dB, occurs
between the true value of |//(/co)|^^ and the value read from the asymptotic curve. The error at
twice or half the corner frequency is about 1 dB.
The following variations of the three basic Bode magnitude plots o f Figure 14.36 are easily
derived:
1.
If H{s) = {sld)^y the Bode magnitude is similar to that shown in Figure 14.36b, except
that the slope is now 2 0 dB/dec. The curve still passes through the point (co = a,
2.
If H{s) = {sla + 1 ) , the Bode magnitude is similar to that shown in figure 14.36c, except
that the right asymptote has a slope o f 2 0 dB/dec. The breakpoint is still at (o) = a,
|//(/co)|^^ = 0), and the error at the comer frequency is 3 dB. If n is negative, the right
asymptote points downward.
734
Let us now consider the Bode plot for a general transfer function
H{s).
After expressing
H{s)
in
the form of equation 1 4 .2 1 , we can draw the asymptotes for each term in equation 14.23 with the
aid o f Figure 1 4 .2 6 . The asymptotes for the
log(o)) curve can then be constructed very eas
ily by graphically summing the individual asymptotes. Since the asymptote for each term in equa
tion 14.2 3 is a piecewise linear curvey the graphical sum o f all the asymptotes is also a piecewise lin
ear curve. Accordingly, it is not necessary to calculate the sum o f dB values at a large number o f
frequencies. If there are
n break frequencies,
then the summation need only be carried out at these
frequencies and for the slopes o f the leftmost and the rightmost segments o f the piecewise linear
asymptote. T he following example illustrates this procedure.
E X A M P L E 1 4 .2 3 . Obtain the asymptotes for the Bode plot o f
H{s)
o f equation 1 4 .2 2 rewritten
below:
I s
+1
H (s)= 2.5
V50
-M
/ V200
/
+ 1
V8
Solution
There are four break frequencies. Rewriting

cies appearing in
H{s)
/ V6 0 0
as a product o f five factors with break frequen
ascending order yields

-1
H{s) =
2.5 - + 1
.8
+ 1
.5 0
.2 0 0
+1
.V
.6 0 0
+1
i}4.24)
= / / , X / / . X //3 X //4 X
Figure 14.3 7 a shows the asymptotes for the five individual terms in equation 1 4 .2 2 , and Figure
14 .3 7 b shows the asymptotes for
^^5
CO
(rad/sec)
(b)
FIGURE 14.37 Asymptote and Bode magnitude for H{s) of equation 14.22 or, equivalendy, equadon
14.24. (a) Bode plots for / / , through
Hy
(b) Bode plot for
equal to sum of those for
through
T he calculation o f the asymptote in Figure 14 .3 7 b proceeds as follows:

1.
T he slope o f the leftmost segment, i.e., the segment to the left o f co = 8, is obviously zero,
2.
There is a breakpoint Pj at co = 8. T he only factor contributing to a dB value at this fre
3.
There is a breakpoint ? 2 at OJ = 50. Since
from Figure 14.3 7 a .

quency is //j(s ) = 2 .5 for which
= 2 0 log(2.5) se 8.
H^{s)
contributes - 2 0 dB/dec to the slope for
to > 8, and since 50 rad/sec is 0 .7 9 6 decade above 8 rad/sec (from lo g (50/8) = 0 .7 9 6 ),

the dB value corresponding to
Pj
is
8 - 2 0 X 0 .7 9 6 = - 7 .9 2
4.
There is a breakpoint
5.
There is a breakpoint
the slope for co > 50,

slope for
0) > 6 0 0 ,
additional 2 0 dB/dec to
resulting in a slope o f zero, we have - 7 . 9 dB for P^
at co = 6 0 0 . Since H^{s) contributes an additional 2 0 dB/dec to the
at co = 2 0 0 . Since
contributes an
resulting in a slope o f 20 dB/dec, and since 6 0 0 rad/sec is 0 .4 7 7 decade
above 2 0 0 rad/sec (because lo g (600/200) = 0 .4 7 7 ), the dB value corresponding to
P^ is
- 7 .9 2 + 2 0 X 0 .4 7 7 = 1.63
6.
Finally, consider the slope o f the rightmost segment. Since
tional-20
contributes an
addi
dB/dec to the slope for co > 6 0 0 , the resulting slope o f the rightmost segment
is zero.
T h e com p lete specification o f the piecewise linear asym ptote is shown in Figure 14 .3 7 b .
'3 6
Exercises.
Chapter 14 Laplacc Transform Analysis 111; Transfer Function Applications
1. Given a firsi-order low-pass characteristic o f tiie form

_
(') =
show that (i) the dc gain is
K and
(ii) for
out _
în
s = ja,
1 4 -
i.e., at co =
a,
the gain is 3 dB down from the dc
H{s).
function H{s)
value. I his to is called the 3 dB frequency or 3 dB bandwidth of

2. Construct the piecewise linear asymptote for the transfer
= 40j-/[(i- +
2){s +
2 0 )].
O nce the piecewise linear asymptote for the Bode plot has been constructed, the true curve can
be sketched approximately by noticing that the error at each corner frequency is about 3w dB (pro
vided that the two neighboring corner frequencies are more than five times larger or smaller).
Figure 14 .3 7 b shows such a rough sketch. In the pre-com puter era, the ability to draw such a curve
by hand even a crude approximation was considered valuable. Nowadays, with the wide avail
ability o f personal computers and C A D sofuvare, one might just as well get the exact plot with
out bothering to look at the straight-line asymptotes. From the perspective o f circuit analysis, the
value of the ability to construct the asymptotes for a Bode plot is greatly diminished, but the tech
nique is still important for its application in the design o f feedback control systems. Such an appli
cation utilizes both the Bode magnitude plot and the Bode pha.se plot. Some background in con
trol systems is required for one to understand the use o f the Bode technique in this kind o f appli
cation. We will relegate the discussion of the topic to a more advanced course in feedback control.
O ur objective in mentioning it in this text is twofold: (1) to introduce the definitions o f some
com m oni) used terms, such as
decibels, decade,
and
octave,
and (2) to demonstrate a highly sys
tematic procedure for adding up several piecewise linear curves to obtain a desired curve, as
described in Figure 14.37.
9. TRANSFER FUNCTION ANALYSIS OF A DC MOTOR

A permanent magnet dc motor, an electromechanical device, converts direct current or voltage
into mechanical eêrg^^ The shaft of the motor rotates freely on bearings. Mounted on the shaft
within the housing is a wire coil called the
armature winding o f the
motor. Surrounding the coil
are permanent magnets that interact with a magnetic field produced when a current flows through
the armature winding. This interaction o f magnetic fields forces the shaft to rotate in a process o f
energ)' conversion. Here current flowing through the armature coils rotating through the magnetic
field produced by the permanent magnets encasing the coils produces a torque on the motor shaft,
which drives a load. Power is delivered from the source to the load.
The electromechanical characteristics o f the permanent magnet dc m otor have a simple circuit
like model amenable to Laplace transform analysis. The model is given in Figure 14.38 and con
sists o f an adjustable dc voltage source in series with a resistor
labeled m otor. Here,
an inductance
and a device
represents the resistance present in the armature winding, and
resents the equivalent inductance o f the wire coil. The device labeled m otor has the current
rep
i J^t)
as an input and the angular velocity co(r) o f the rotating shaft as an output. T he interaction
between the electrical part o f the model and the mechanical part o f the model occurs at the loca
tion o f this symbol. T he voltage
is an induced voltage proportional to the angular velocity
oj(r). Because co(/) is not a circuit variable, classical notions o f impedance, admittance, voltage
gain, etc., do not fit the problem, whereas the more general notion o f transfer function does, forc
ing us to move slightly beyond the confines o f circuit theory proper to analyze the system.
The torque
J\t)
produced on the rotating shaft by the current flowing through the armature coils
is proportional to the armature current /^(/), i.e..
Tit)
(1 4 .2 5 )
T he mechanical rotation o f the m otor affects the electrical portion of the system by inducing a
voltage
This voltage is proportional to the m otors rotational speed, or angular velocity, co(r),
i.e.,
(1 4 .2 6 )
Since the m otor converts electrical energy to mechanical energ^^ conservation of energy dictates
that the constant o f proportionalit)' be equal to the same constant that relates torque and current
for a lossless motor. Specifically, electrical power-in must equal mechanical power-out or, in equa
tion form.
= m coit)
which forces
k.
O ur first goal is to find the transfer function o f the motor,
H{s)
[o){t)]/V^{s).
For convenience,
let Q(^) denote X [o)(r)]. As a first step, sum the voltages around the loop of elements in the cir
cuit model of Figure 14.3 8 . This results in the differential equation
V (/)
at
(1 4 .2 7 )
A ssum ing zero initial con ditions, the Laplace transform of equation 1 4 .2 7 is
V^{s) = {R^ + Lj)I^{s)
kQis)
(1 4 .2 8 )
738
From basic mechanics, the differential equation governing the mechanical portion o f the system
r(o = y ^ + B < o (o
at
where
is the moment o f inertia of the combined armature, rotating shaft, and load; B is the
coefficient of friction; and 7\t) is the torque produced by the motor to turn the load. Recalling
equation 14.25, T\t) =
the Laplace transform o f equation 14.29, assuming zero initial con
ditions, is
(14.30)
Solving equation 14.30 for IJis), substituting the result into equation 14.28, and then solving for
Q(j) leads to the expression
k
2 V,{s)
Q(s) =
(14.31)
Equation 14.31 characterizes the pertinent dynamics of the permanent magnet dc motor and
allows one to find the angular velocity o f the motor shaft as a function of time for given inputs.
To see how the motor responds to step inputs, suppose vjit) = Ku{t). As objectives, let us find (1)
the steady-state value, or final angular velocity, o f the shaft and (2) the steady-state value o f the
armature current. The final speed o f the shaft is important because, for example, one needs a fixed
speed for the rotation o f a compact disk or a fen. The final or steady-state current is important for
determining the power needed from the source.
If Vj(t) = Ku{t), then V(^s) = KIs. It follows from equation 14.31 that
S +
S+
RgS + k
2\
(14.32)
J
Applying the final value theorem to equation 14.32 implies that
(Ovv =
lini
co(r) = ------- y K
To isolate the armature current IJJ), again apply the final-value theorem to determine the steadystate value of i {t). Combine equations 14.30 and 14.31 to obtain
1
5+
4 (^ ) =
2 's
LaJm
( 14.33 )
Equation 1 4 .3 3 allows us to find the armature current as a function o f time for a given input volt
age. As above, if the input is a step, i.e., if
v^{t) = Ku{r),
B
s + --------^
------------ t t K.
S+
\
it follows that
R,,B + k~\
(1 4 .3 4 )
O nce again, application o f the final-value theorem to this expression leads to the value o f the
steady-state armature current:
Q
hi . ss ~
^Cl
B+ k
T he preceding analysis suggests the utility o f the Laplace transform as a tool for analyzing the
dynamic behavior o f electromechanical systems. In fact, system transfer functions o f the form of
equations 14.31 and 14.33 are often starting points for further analysis. Extensions of the above
analysis can be found in the homework exercises.
10. SUMMARY
This chapter has expanded the notion o f transfer function from its definition in Cliapter 13 into
a tool for modeling not only circuits, but other practical systems, such as a dc motor. The trans
fer function characterizes circuit or system behavior by the location o f its poles and zeros. For
example, if a transfer function has a pole on the imaginary axis or in the right half-plane, the asso
ciated circuit or system is said to be
unstable,
because there is an input or, possibly, simply an ini
tial condition (as in the case o f a second-order pole on the y'oj-axis) that will excite the pole and
cause the response to grow without bound. Further, the ubiquitous presence o f noise will always
excite poles on the imaginary axis and cause the response to be unstable.
This chapter categorized various types o f responses: zero-state and zero-input responses, natural
and forced responses, transient and steady-state responses, etc. Recall that the zero-state response
is the response to an input assuming zero initial conditions, which is the inverse Laplace transform
o f the product o f the transfer function and the Laplace transform o f the input excitation. Recall
that the zero-input response is due only to initial conditions on the capacitors and/or inductors o f
the circuit. The complete response for linear circuits having constant parameter values is simply
the sum o f the zero-input and zero-state responses. This decomposition generalizes to the broad
class o f linear systems studied in advanced courses. Under reasonable conditions, other decom po
sitions are possible, such as a decomposition into the natural and forced responses or transient and
steady-state responses. Other important responses are the impulse and step responses.
For stable circuits, the frequency response provides important information about the circuit.
Recall that frequency response is a plot o f the magnitude and phase o f //(/w ) as to varies from 0
to cc. T he Bode plot is a plot o f gain in dB vs. frequency represented using a log scale. In this con
text it is relatively straightforward to construct an asymptotic approximation using straight line
segments, from which the actual plot is easily sketched by hand. T he information in such a plot
tells us how a circuit behaves when excited by sinusoids o f different frequencies.
(0
Chapter 14 Ijp la c c Transform Analysis III: Transfer Function Applications
Lastly, this chapter has introduced the initial- and final-value theorems, which provide a quick
means of computing the initial and final values o f a time function from knowledge o f its Laplace
transform. Such theorems have wide application in control system analysis, as evidenced in our
analysis o f the dc servo motor.
TERMS AND CONCEPTS

3 d B frequency for low-pass characteristics: the frequency at which the gain is 3 dB down from
the dc gain.
Asym ptote: a limiting straight-line approximation to a curve.
B ode m agnitude plot: a plot o f gain in dB vs. frequency represented on a log scale.
B ode phase plot: a plot o f phase in degrees vs. frequency represented on a log scale.
B ounded: the condition where a signal
satisfies [/(r)|
< K <<x> for
all
f,
i.e., it has a maximum,
finite height.
B reak point: the point at which two asymptotes o f a Bode plot intersect.
C om p lete response: the total response o f a circuit to a given set o f initial conditions and a given
input signal.
C o rn er (break) frequency: frequency at which two asymptotes o f a Bode plot intersect.
D ecade: a frequency band whose endpoint is a factor o f 10 larger than its beginning point.
D ecibel (d B ): a log-based measure o f gain equal to 2 0 logjQ|//(yw)|.
Final-value theorem : a theorem stating the following: suppose
F{s)
has poles only in the open left
half o f the complex plane, with the possible exception o f a single-order pole at ^ = 0. Then
the limit
oij{t)
as /
oo equals the limit o f sF{s) as i 0.
Forced response: the portion o f the complete response that has the same exponent as the input
excitation, provided the input excitation has exponents different from those o f the zeroinput response.
Frequency response: measure o f circuit behavior to unit magnitude sinusoids, cos(w t), as O) varies
from 0 to
00
. Equal to the evaluation o f the transfer function
H{s)
at ^
for all co.
Frequency scaling: lor a linear passive network, dividing all inductances and capacitances by a
factor, Ayr while keeping all resistance values unchanged.
Fundam ental period o f p e rio d icy (r): the smallest positive number
T such
that J{t)
=j{t + T).
Im pulse response: assuming all initial conditions are zero, if the circuit or system input is 6(r),
then the resulting^(r) is called the
impulse response.
I he inverse transform o f the transfer
function equals the impulse response.

Initial-value theorem : a theorem stating the following: let
L\J{t)] = F{s) be a strictly proper

F{s) are both polynomials
rational function o f s\ i.e., the numerator and denominator o f

in
s,
with the degree o f the numerator less than that o f the denominator . Then /(O*") is
the limiting value o f sF{s) as ^
oo.
L in ear circuit: circuit such that lor any rwo in p u ts/j(r) an d/^ (f), whose zero-state responses are
^ j(/) andjsCO. respectively, the response to the new input
A'2 >'2 (^)]. vvhere ATj and
Kj
[Kjf^{t)
+ A'2 /S(r)] is [A',^,(/) +
are arbitrar)' scalars.
M agnitude response: the magnitude o f the frequenc)' response as a function o f O).

M agnitude scaling: for a linear network, multiplying all resistances and inductances by a factor,
and dividing all conductances and capacitances by
for dependent sources, this
means all parameters having ohms as units would be multiplied by

siemens as units would be divided by
and those having
dimensionless parameters are left alone.
N atural response: the portion o f the complete response that has the same exponents as the zeroinput response.
Octave: a frequenq^ band whose endpoint is twice as large as its beginning point.
Op amp open-loop gain: the gain of the op amp when no feedback paths to the input terminals
are present.
Periodicy(/); function satisfying the condition that there exists a positive constant, T, such that
/ f ) =fijt+ 7) for all f > 0.*
Phase response: the angle o f the frequency response as a ftinction o f (O.
Piecewise linear curve: an unbroken curve composed of straight-line segments.
Pole (simple) o f rational function: zero of order 1 in denominator polynomial.
Poles (finite) o f rational {unction: zeros o f denominator polynomial.
Ramp function, K^)j integral o f unit step function having the form Ktu{i) for some constant K.
Rational fim cdon: ratio o f two polynomials; a polynomial is a rational function.
Stable transfer function: a transfer function for which every bounded input signal yields a
bounded response signal; i.e., all poles are in open left half o f the complex plane.
Steady-state response: that part of the complete response which either is constant or satisfies the
definition of periodicity for t> 0 .
Step response: the response of the circuit to a step function, assuming all initial conditions are
zero.
Transient response: those terms o f the complete response that are neither constant nor periodic
for ^ > 0; i.e., the transient response does not satisfy the definition of a periodic function
for f > 0.
Zero-input response: the response of a circuit to a set of initial conditions with the input set to
zero.
Zero-state response: the response of a circuit to a specified input signal, given that the initial con
ditions are all set to zero.
Zeros o f rational function: values that make the numerator polynomial zero.
O '
O '
* This (nonstandard) definition has been adapted for one-sided Laplace transform problems.
O '
742
Chapter 14 * Liplacc IVansform Analysis llh lr a n s fe r Function Applications
PROBLEMS
3. Consider the pole-zero plot o f the transfer

function
POLES AND ZEROS

1. For the pole-zero diagram shown in Figure
P 1 4 .1 ,
C om pute the transfer function
(b)
C om p ute
the
impulse
H{s).
and
step
(b)
C om p ute
response.
(c)
If the input is
itive number
find the pos
such that the response
impulse
response.
parts o f the response.

(c)
Com pute the step response. Identify

the steady-state and transient parts o f
the response.
CH ECK .- Your answer to (b) should be the

derivative o f your answer to (c), since the delta
function is the derivative o f the step function.
(d)
does not have a term o f tlie form
Ke~u{t).
the
H{s).
Identify the steady-state and transient
responses. Check that the impulse

response is the derivative of the step
given in Figure P I 4 .3 .
If the dc gain is 2 , find
HiO) = 8.
(a)
H{s)
(a)
itive number
Find the response under this
\Qe~^ti{t)y find the pos

a such that the response
If the input is
does not have a term o f the form
condition.
Ke~'û{t).
jto
Find the zero-state response
Identify the
transient and steady-state responses.
(double pole)
jO)
-2
-1
)C4j
Figure P I4.1
-q>a
-2
2. Consider the pole-zero plot of a transfer

function
H{s)
given in Figure P I 4 .2 .
(a)
If the dc gain i s - 1 0 , find/yW -
(b)
C om pute the impulse response.
(c)
C om pute the step response.
-A]
Figure P I4.3
C H E C K ; Your answer to (b) should be the
4. Consider the pole-zero plot o f a transfer
function
function is the derivative o f the step km ction.
(a)
(d)
If the input is
itive number
find the pos

such that the response
H{s)
given in Figure P i 4 .4 .
If the gain
H{s)
is 4 .8 at ^ = 1, find
His).
(b)
C om p ute
the
impulse
response.
does not have a term of the form
Identify^ the steady-state and transient
Ke~û{t).
Find the zero-state response
(c)
C om pute the step response. Identify'

the steady-state and transient parts of
jO)
the response.
C H ECK .- Your answer to (b) should be the
-e o
-2

function is the derivative o f the step function.
(d)
:-4j
Figu re P I 4 .2
If the input is
10.9
find w, such that the response does

not have a purely sinusoidal term.
743
Find the zero-state response under this
(c)
If the input to the circuit is
condition. Identify the transient and
2tu{t), use MATLAB to compute the
steady-state responses.
partial fraction expansion and the

resulting time response. Sketch the
jw
approximate response for large t.
ik
oj
(d)
------- 1> o
-2
Repeat part (c) for v-J^t) = lp-u{t).
7. The input impedance o f each network

shown in Figure P I4 .7 is 2 (5) =
s+a
O-j
with a > Q. Which network is stable and why?

Figure P14.4
Compute the step response of the stable net

work. Identify the transient and steady-state
5. The pole-zero plot of the transfer function
His) is given in Figure P 14.5.

(a)
If the dc gain is - 1 , find H{s).
(b)
Compute the impulse response.
(c)
Compute the step response. Identify the
Z (s)
(a)
Compute the response to the input

sin(4/)(r).
(e)
Z (s)
transient and steady-state responses.

(d)
(b)
Figure P I4.7
If the input is e~^û{t), find the posi

tive number a such that the response
8 . Reconsider the linear active networks of
does not have a term of the form
Figure P I 4.7, which have input impedance
Ke~^û{t). Find the zero-state response
function Z{s) with poles at j = - 1, - 3 y 4 and
zeros at 5 = 2, p .. It is known that Z(0) = 8 .
jco
>
(a)
Write Z{s) as the ratio of two polyno

mials in s.
(b)
If the network of Figure P I4.7b has

zero initial conditions and
e------ ------>o
-2
2
lu{t) A, write down the general form

of
writing
down the forms, simply leave the con
C-4j
stants as literals without calculating

Figure P14.5
6 . The transfer function H{s) of a particular
numbers. Does the output remain

finite as
co>
(c)
If the network of Figure PI 4.7a has zero
active circuit has poles at - 1 2j and - 2 . At j
initial conditions with input
= - 1, the transfer function gain is found to be
20u(t) V, find the general form of
8 , and //(oo) is known to be finite. When this

= sin(2r)tt(/) V and v-JJ) = u{t) V, it is found
and then
Again leave the con
stants in these forms as literals. Does the
output remain finite as f
00?
that the output is zero after a long time.

(a)
Find the zeros of the transfer function.
9. Repeat Problem 8 for when the zeros of Z{s)
Construct H{s) and then compute the
are j = -2, j2 . Also, answer the following ques
step response.
tions:
circuit is excited by the input waveforms
(b)
^44
Chapter 14 Laplacc'I'ransform Analysis 111: Transfer Function Applications
(a)
W hich configuration shows a stable

circuit?
(b)
If one o f the configurations is an

unstable circuit, what input will cause
the zero-state response to grow with
out bound as /
x?
10. Reconsider Figure P l4 .7 a with impedance
Z{s) having poles at ^ = 2, 6 y8 and zeros at s

= - 1 , - 2 , 4. It is known that Z (0 ) = 3 2 Q.
(a)
Write Z{s) as the ratio o f two polyno

mials o f s.
(b)
If the network has zero initial condi

tions, and
û(t)
V is applied to
the two terminals, find the current

/
it)
flowing into the network for
t > 0.
11. Show the pole-zero plot o f each o f the fol

lowing
transfer
functions,
and
determine
whether the system (circuit) is stable. Hint: Use
Figure P I4.1 2
the M A rij\ B command roots to obtain the
7X . T O S
.ind poles of k I, transfer ainction, even

though there are mediods that can determine sta-
,,
Consider the circuit o f Figure P I 4 .I 3 .

(a)
Show that the transfer function is
bilit)^ without calculating the exact pole locations.

(a) /Y,(.v) =
(b)
Is -
I^
R^C2
H(s) =
s -+ ^s +\2
^ ( /?,Ci
1.5.V- -Q .5 .V -H .2 5
H2{ s ) =
(b)
r V 2 . v + 7.5.vH-6.5
R2C2 ^ R\C2 j
If Cj =
S+
= 1 mF, find
K\C\R2C2
and
Rj
so
that the poles o f the transfer function

(c)
H-^{s) =
are at 5 = - 0 .2 1 9 2 2 and - 2 .2 8 0 8 . The
1.5.V- -H2.25.S + 4. 4062
problem o f finding element values to
5-V4.3I25.V + 5.3125
2s^ + 2.5s-
- 17.V + 220.5
/ - h 7 r V 5 5 . v - + 5 6 9 5 + 520
realize a given transfer function
synthesis problem.
R.) = (4 k^, 0 .5 kI2)
A N SW ER ;
2
kLl)
12. For each o f the op amp circuits o f Figure

P I4 .1 2 , find the transfer functions and the poles
is
called the
He
and zeros in terms o f the indicated literals.
+
v.(t)
Figu re P I 4 .1 3
or (1 k il,
Chapter 14 iîplacc Transform Analysis 111: Transfer Function Applications
STABILITY PROBLEMS
14. For each of the circuits below, compute the
indicated transfer function and determine the
range
or
for which the circuit as m od
eled by the transfer function is stable in the

BIBO sense.
(a)
For the circuit of Figure P I 4 .1 4a the

transfer function is
R =4a,
(b)
His)
Vi s )
16. (a)
The pole-zero plot o f a transfer func

tion
c = 0 . 2 5 F.
is given
in
Figure
P i 4 . 16a.
Construct an input that will cause the
For the circuit o f Figure P I 4 .1 4b the

V
(.v)
transfer function is His) =
.
response to be unbounded (unstable),

(b)
Repeat part (a) for the pole-zero plot

o f Figure P i 4 .1 6b.
(c)
For the circuit o f Figure P i 4 .1 4 c the

H{s) =
Imag Axis
Ojw,
finis)
Real Axis
------o - >
-2
-jco,
(a)
Imag Axis
(a)
Figure P I4 .1 6
17. The circuit in Figure P i 4 .1 7 is to be stabi
lized in the BIBO sense.
(a)
If y? = 2
range o f
Figure P i4.14
(b)
\/
15. Given that
H(s) = - ^
V
*-111
complete
cuit is stable.
range of
in the circuit in
R for
= 0 .5
S,
which the cir
required for stability.
= 0 .5 , determine the complete
range o f
Figure P 1 4 .1 5, where C = 2 F and

find the
If
determine the complete
R required
for stabilit}'.
746
r-s.
(1) Find the zero-state response, the
in tiie circuit o f
Figure P i 4 . 18; then find the
range
plete response.
for whicii the circuit is sta
Now suppose tiiat
il.
= -2
complete range R for
which the circuit
ble, assuming
(b)
zero-input response, and the com
complete
R=
(2) Find the steady-state response and

the transient response.
S. Find the
is stable.
(3) Find the forced response and the

natural response.
(c)
Repeat parr (b) for

V and
v-^^{t) = 20e~^^û{t)
= 0.
Is the forced
response well defined?
H in t:
M A T L A B s
[r,p,k]
com m and
Use
=
residue(n,d) to com pute the partial

fraction expansions.
R.
19. For the circuit o f Figiu-e P i 4 .1 9 find the
-C
R, <
transfer function
v;-(.v)
and the range o f a for which the transfer func
tion is stable.
Figure 1P14.20
2 1 . Find the complete response for r > 0 for the
circuit o f Figure PI 4 .2 0 for each o f the follow
ing circuit conditions. (If you properly utilize
the results o f Problem 20 and the principle o f
linearity, the answers to this problem can be
written down directly.)
(a)
= 6 V,
(b)
Figure P I4 .1 9
response
has been found to be

zero-input response =
(a)
Determine the transfer function.

If v-^^{t) = 10 co s(10f)(/) and I'Ô) is
changed to 8 V, com pute the complete
response
i + io
= 50
/?, = 2 0 0 Q ,
Suppose
v^Q-)
Vj^^{t)
= 8 V.
for / > 0. Identify the
transient and steady-state parts o f the
R^^^= R^
C = 2 . 5 niF.
(b)
e~^û{t) V
4e~~û{t) V
(b)
response.
when
V, and ^f;(0")
zero-state response = 10(1 -
2 0 . Consider the circuit in Figure P I4 .2 0 .
RC
v-^j{t) = 10(^)
= 4 V; a decomposition o f the complete
RESPONSES AND
CLASSIFICATIONS
H{s) =
5 sin(5r)w(r) V
2 2 . Consider the network o f Figure P I 4 .2 2 .

Suppose C = 0.1 F,
(a)
v-^j{t) =
I 6 V > . , W = 8 f - ' M r t V
= 10 sin(5/)//{/) V, and
74
Chapter 14 Laplacc Transform Analysis III: Transfer F u n aio n Applications
Pure
resistive
network
vJt)
sC + +
Figure P I4 .22
2 3 . In the circuit shown in Figure P I 4 .2 3 ,
= 20 a
v jt ) = 25u{t)
V(^0~) = 10 V.
C = 0 .0 5 F,
v^2 i^) = 20e~û{t)
V, and
.v-f-
/?2
5-t- 2
CR.(1/
Draw its pole-zero plot.

Express the
(b )
/? 2
.V+ 1.6
C/?,
H (i) =
/?1
5 a
.v +
AC H------
v,(t)
given
in Figure
P24.b as a sum o f possibly shifted sim
V,
ple functions such as steps, ramps, etc.

Then find the zero-state response.
(c)
Draw the equivalent frequency domain

circuit, and find the zero-input response
when
'' 6
V(^0~) = -1
V. Hint: The response
RC circuit
R^^C.
of an undriven first-order
where x =
(d)
(e)
is
Com pute the complete response.

If
were changed to the input shown
in Figure P l4 .2 4 c , what would be the
Figure P I4.2 3
response of the circuit? (Hint: W hat is

(a)
(b)
(c)
Find
for r > 0 by the Laplace
relationship
between
P I4.24b and P I4.24c? Note that differ
Write
as the simi o f two com
entiation and integration are linear oper
ponents: the zero-state response and
ations when the initial conditions are
the zero-input response.
zero and all circuit parameters are con
Separate the zero-state response into
stant. For linear constant parameter cir
one term due to
cuits, then, a linear operation on the
and another term
input induces the same linear operadon
v^.
Separate the complete response into
on the response, provided that the initial
the steady-state response and the tran
conditions are zero and the circuit
sient response.
parameter values are constant. Combine
Utilize the answers o f previous parts to
this concept with the structure o f the
write down the complete response for
decomposition to obtain the answer.)
(e)
r > 0 for the following two cases (all

values in volts):
Case 1
Case 2
V(;iO )
v jt)
Vs2it)
20
\Ou{t)
\5e^û{t)
lOuit)
dOe^ûit)
R.
R, <
(a)
2 4 . Consider the circuit o f Figure P I4 ,2 4 a .

Suppose C = 25 mF, /?, = 25 n ,
and
100
n,
A v(t) (V)
20
Figu re P I 4 .2 4
R^^ = R^ 11 R j.
(a)
Figures
transform method.
due to
(d)
the
-t
(c)
7AH
Chapter 14 Lapbcc IVansform Analysis 111: Transfer Function Applications
25. For the circuit o f Figure P 1 4 .2 5 , Cj = 0 .2

mF and
C-,
= 0 .5 mF. The initial conditions arc
;-'q (0 ) = 8 0 mV and
v jt ) = OAu{t)
v^{0~)
i,(t)
= 0. The input is
V.
V ,.(t)
(a)
(b)
Find the zero-state response.
K.U'I
(c)
Find the zero-input response.
(d)
Find the complete response.
(e)
Specify the general form of the natural
2 7 . Consider the circuit o f Figure P i 4 .2 7 .

(a)
response.
(0
Specify the transient and
H{s) =
is)
KM
steady-state responses.
LC
.9 -I-
- v +
R2 C}
.v +
LC
R j)
16
r +
4.V + 20
when /?, = 1 Q ,
F, and
(b)
=4
LX C =
0 .1 2 5
Find the zero-state response to the
input
(c)
/? 2
= 0 .5 H. ^
IO(r) V.
Draw an equivalent j-plane circuit

that accounts for the initial condi
tions. There are four possibilities, but
Figure 1M4.25
one is superior.
(d)
2 6 . Consider the circuit in Figure P i 4 .2 6 .

Suppose /e, = 10
C= 1 /40 F,
+ i>2 (() =
Q,
= 3 0 LX
= 15
il,
(e)
function so that you know its shape).

C om pute
Com pute
and /^(0).
(0
response
vc{0~)
for
= 4 V.
Find the complete response for
5u{t),
= 2 V, and /^(O") = - 2 A.
Hint: Use linearit}'.
and /^(f) for r > 0; this

------- _ _ o
is the complete response.

(c)
zero-input
Find the zero-input response for /^(0~)

= 0 and
V (plot this input
(a)
the
/Y-(0) = 0 and /^(0) = 1 A.
and Z. = 8 H. Suppose
(b)
Find
Identify' the part o f the response due
i,(t)
only to the initial conditions at r = 0

that result from
think
of
this
one could
as
the
zero-input
Figure P I4 .2 7
response.
(d)
Identify' the part o f the response due

only to the part o f the input for r > 0,
i.e., due to 25//(^) assuming zero initial
conditions; one could think o f this as
the zero-state response.
(e)
Identif) the transient and steady-state

parts o f the complete response.
28. The purpose o f this problem is to illustrate the

computation o f the zero-input response using the
transfer Rmction concept. This requires proper
application of the equivalent initial condition cir
cuits and the use o f a transfer fiinction for each
initialized dynamic element. Consider the circuit
of Figure P I4.28a.
Chapter 14 Liiplacc Iransform Analysis III: Transfer Function Applications
(a)
Find che transfer function
H\ (j)
jco
if the
input is a properly directed current

source connected between nodes
(double pole)
and 2. T he value o f the current source,

accounts
for
the
initial
-> o
-2
capacitor voltage y ^ 0 ).
(b)
<l>-2j
Figure P 14.29
Find the transfer function / / 2 W if the

input is a properly directed current
source connected between nodes
and 3 . T he value o f this current
3 0 . For the pole-zero diagram o f
source,
H(s) =
s
accounts for the initial inductor current.
(c)
For the circuit o f Figure P I4 .2 8 b , uti

lize the transfer functions obtained in
shown in Figure P I 4 .3 0
H{2)
cos{4t)u{t)
=2
V.
jO)
parts (a) and (b) to find
>k.
0 4j
the zero-input response for Vq^^{)
(i)
= 10. Com pute
the sinusoidal steady-state response to
for / > 0, assuming initial condi

tions
= 4 V and /^(0) = 5 A;
>a
(ii) the zero-state response for

for ^ > 0; and
0 -4 j
(iii) the complete response for

for r > 0.
Figure P I4.3 0
ANS\\T-:R: 0
/
2H
3 1
4S <
,2 S
4v.
2F
. Find the phase and magnitude in sinusoidal
steady state for each o f the given transfer func

tions when excited by the indicated sinusoidal
mput.
(a)
(a)
H{s) =
2s+ 6
Vi{s)
.V--1-4.9-h i6
= 2 cos{2t
is excited by the input

+ 4 5 ) V.
4S <
(b)
Suppose
5
=3
output
(b)
( 6 + 4 )
has
cos(3f + 4 5 ) V.
Find the sinusoidal input
that
gives rise to this output.
Figure P i4.2 8
(c)
Let Z = 0 .2 H, C = 0.2 5 F, and /?j =
STEADY-.STATE CALCULATION
= 7? = 1 n
2 9. For the pole-zero diagram o f
P i 4 .3
in the circuit of Figure
. Compute the transfer function
H{s) =
shown in Figure
ViU)
P i 4 .2 9 H{0) = 4.
Com pute
the sinusoidal steady-state response to

sin(2f)(f) V.
.V -I-------5 -i-
RC
LC
750
Find the phase and magnitude o f the

output Vfî) in sinusoidal steady state
to
(d)
= 20 cos(4?) V.
Repeat pan (c) for the output ijjJ).

+
Figure P I4.32
V ,( t )
33. Consider the circuit o f Figure P I4.33.

Given that C2 = 1 F, find Cp R^ and /?2 so
that the transfer function is H (s) = - 5
(a)
Figure P I4.31
Find the steady-state response
Kcos{2t+ 0) when
2 sin (2t)u{t).
32. Find the phase and magnitude in sinusoidal

steady state for each of the circuit transfer func
(b)
s+4
=
Now find the complete response to
tions (or circuits) below when excited by the
the input
indicated sinusoidal input.
identify the transient and steady-state

parts, assuming no initial conditions
16^ + 96
(a)
Vi(s)
are present.
s^ + Ss + 64
is excited by the input
= 2 sin(2^)(^) and
= 4 cos(4?
(c)
Verify that your answer to part (a) and

your steady-state answer to part (b)
(b)
coincide.
Suppose H (s) = 4>/5

(5 + 2 r + 16
has output
= 25 cos(3^ + 100)
V. Find the sinusoidal input
=K
cos(fiW + 0) that gives rise to this out
put.
(c)
G>nsider the circuit o f Figure P I 4 .32.

Suppose Z = 1 H ,
F, and R^ =
= 15 n , C = 0.01
O.. Suppose
voltage + to -
the
across the current
source, is the desired output. The
Figure P I4.33
transfer function
_
'in
hnis)
?
1
+ 5 + -----
work transfer function H(s) =
is shown
in Figure P 14.34. If the input is a sinusoidal
L
LC
Find the phase and magnitude o f the
voltage o f 1 V amplitude, then, in the steady
= Kcos(cot + 6),
state, the output voltage has the greatest ampli
output cosine,
in sinusoidal steady state to

(d)
34. The pole-zero plot o f a certain RLC net
99
= Z/n(^) =
tude at approximately what CO?
At what
200 cos(10/) mA.
approximate value o f (0 does the amplitude of
Repeat part (c) for the output
the output voltage dip to its lowest value? Give

qualitative, not quantitative, justification for
your answers.
Chapter 14 biplacc Transform Analysis III: Transfer Function Applications
C om pute the phase and the magnitude o f the
jco
steady-state response for each o f the following
jlO
inputs:
= 20
(a)
j3
---- -5
cos(4r + 4 5 ) V
(b)
y (/) = 4 co s(4 0 ^ ) V
(c)
v,{t) = 4cos{2t)V
Suggestion: One can use the com m and freqs(
-2
n,d,w ) for a rational function.

3 9 . Following are the transfer functions o f
X -
some linear networks that contain controlled

sourccs. The networks are initially at rest, i.e.,
Figure P i 4.3 4
with zero initial conditions. An input
cos{2r)u{t)
35. A linear circuit witii a transfer function
2v+ 4
H{s) = ^ =
v;
residue comm and in M ATLAB to generate the

partial
fraction
expansion,
=4
determ ine
the
approximate form o f the output,
+ 5^ + 6
very large values o f

has an input
=8
V is applied at f = 0. Using the
cos(2f + 4 5 ). Com pute
t.
for
Hint: For stable circuits,
what part o f the response is meaningful, and for
the magnitude and phase (in degrees) o f the
unstable circuits, what part o f the response is
output o f the circuit in the steady state.
dominant?
3 6 . A stable active circuit has the transfer func
(a)
- 3 .7 5 j + 5
//,(^) = ^ ---.9 "-1 -4 5 -I-3
tion
H(s) =
V.
V:.
5 + 4
(b )
- ^ 5 :
5- + 2.V + :)
2.5s- 3
H2(s) = -^
s" -t- 2 i 4- 5
Com pute the phase and the magnitude o f the

steady-state response for each o f the following
. .
(c)
Ht,( s )= -^
s
7 r + .v -h 4
-l- 5 ~ + 9 i + 9
inputs:
(a)
= 4 cos(2^ ) V
(b)
U ^ ) = 4 c o s (4 0 V
(d )
H^(s) = -^
75^ + 5 + 7.75
(e)
5 ' V 3 5 + 1 1 . 2 5 5 + 18.5
^ --0 -5 5 + 5
5 -h0 .5 5 + 5 .7 3 2 1
=2^2
The input to the circuit is
cos(2r +
3 0 ) V. Knowing that the response has the

form
A cos{2t
+ (j)), find
A and
(j).
(f)
5^ - 5 . 5 5 - +
A N SW E R : (b) I'h c result follows from the fol

lowing X'lATLAB code:
3 8. A stable active circuit has the transfer func
nin = 8*[1 0 ];
tion
n = [2.5 -3]:
V:Js)
165^ + 4 4 5 + 1 2 8
85- + 365 +
80
145-12
^6(*v) =
5- + 5 .5 5 + 1 4 5 + 12
din = [1 0 4];
K,ur(^)
=5
. s ' + r + 7 .2 5 5 4 -1 8 .5
3 7 . A stable circuit has the transfer function
v^îs)
14,s-2-23.V + 20
d = [l 2 51;
nn = conv(n,nin);
dd = conv(d,din);
'5 2
[r,p,k] = residue(nn,dd)
(a)
Find the transfer function H{s) =
(b)
If the circuit is initially at rest, find an
r=
V^{s)IV^{sY
-4.0000e+ 00 - 7.0000e+00i
-4.0000e+ 00 + 7.0000e+00i
expression for Vj (/) that is valid for all
4.0000e+ 00 + 4.0000e+00i
t (small or large). Use the residue com
4.0000e+ 00 - 4.0000e+00i
mand in MATLAB to do the needed

partial fraction expansion.
P=
-l.OOOOe+00 + 2.0000e+00i
-1 .0 0 0 0 e + 0 0 -2 .0 0 0 0 e + 0 0 i
-7 .7 7 l6 e -l6 + 2.0000e+00i
(a)
- 7 .7 7 l6 e -l6 - 2.0000e+00i
Compute the transfer function. (It

should be first order.)
k=D
(b)
Find an expression for the natural
(c)
Suppose that for t = 0 , each capacitor
>>K = 2*abs(r(4))
response o f the circuit.
K = 1.13l4e+01
Phi = atan2(imag(r(4)),real(r(4)))* 180/pi
voltage is 1 V and the inductor current
Phi = -4.5000e+01
is zero.
Compute the zero-input
response. Observe that this response

(Apply item 19 o f Table 12.1.)
has a steady-state part. This phenome

non illustrates why equations 14.7
12140. Following are the transfer functions of
require more than just a stable transfer
some linear networks that contain controlled
fijnction as an underlying assumption.
sources. The networks are initially at rest, i.e.,
It is necessary that the transfer fiinc-
with zero initial conditions. An input
=8
tion include all the dynamics of the
sin(40f)(r) V is applied at f = 0. Using the
circuit, which is not the case in the cir
residue command in MATLAB to generate the

partial fraction expansion, determine the
cuit of Figure P I4.42.
approximate form o f the output,
for
very large values of t. Hint: For stable circuits,

what part o f the response is meaningful, and for
unstable circuits, what part o f the response is
dominant?
(a )
(b )
H^(s ) = - y
H2(s) = ^
80
+4^ + 3
80
43. The circuit shown in Figure P I4.43 has zero
initial conditions, yÔ") = 0, and //^(0) = 0).
80
(a)
(c )
s + 2 j + 160l5
(b)
(d)
If
= K(j(r), find V({s) and
5 ^ + 8 5 + 1 616
and determine the largest voltage that
5^ + 155-^ + 8 3 5 ^ + 1 9 9 5 + 170
can appear across the capacitor. At

what times does this occur? This phe
H{s) = - , ------- g------- 2-----------------
y 4 l . In Figure 14.17 o f Example 14.8, let ji=
nomenon
1.4 and recall that v-{t) = cos(2r) V.
design o f insulators that support trans
has application
in the
mission lines. For example, a 100 kV
Chapter 14 I^placc rransform Analysis III: Transfer Function Applications
voltage transmission line would need

insulators that can handle at least
1) V, find
for r > 0 by the
Laplace transform method.
2 0 0
kV.
(c)
If the impulses are applied to the cir

cuit ever}^ second, i.e., if
= [6(^)
+ 6 (r - 1) + 6 (r - 2) + - ] V, find the
output
for w < t < ( + 1) when
the integer ;/ becomes very large (i.e.,

after the circuit reaches a steady state).
Sketch the waveform.
Figure P I4.4 3
4 4 . Suppose the voltage source in Problem 4 3 is
changed to
For
t >0,
ti{t)
10 sin
V.
determine a tight upper bound on the
largest voltage that am appeiir across the capacitor.

ANSWHR: 111.2 V
4 5
. In the circuit o f Figure P i 4 .4 5 , all stored
energ)^ is zero at r = 0. The current source is

=
A. Determine
Vgn,{t). W hat
Figure P I4.4 7
is the largest
voltage that can appear across the capacitor?
FREQUENCY RESPONSE
4 8 . Draw to scale the pole-zero plot for
H{s)
(i + l)/[i(j^ + ^ + 10], and graphically compute

the magnitude and phase o f //(/co) for to = 0 .2 ,
0
Figure P i4.45
46. At r = 0, the energy stored in the
LC elements
in Figure P I4 .4 6 is nonzero. The current source

is
cos{lt)ti{t)
A. Is the circuit stable?
Determine the behavior o f
. 5 , 1, and 10. W hat are the limiting values of
the magnitude and phase, i.e., for co = 0 0 ?
4 9
. Draw to scale (on graph paper) the pole-
zero plot o f the transfer function
for large t.
((.v+ l)2 + 1 6 ) ( . r - l )
His) =
( ( a--I-1) + 4)((.v + l) -1-36)

and determine graphically the magnitude and
phase o f
H(joj)
for co = 0, 2, 4 , 6, and oc. Do
your answers make sense physically? Explain.
Figure P i4 .4 6
. (This problem focuses on a qualitative under
4 7 . This problem illustrates the condition o f
5 0
steady state, but not sinusoidal steady state. For
standing o f poles and zeros o f a frequenc)-
the circuit shown in Figure P 1 4 .4 7 , the op amp
response.) You have determined the pole-zero plot
is assumed to be ideal.
of a band-reject filter as shown in Figure P I4 .5 0
(a)
(b)

If
H{s).
= 0 and y- (f) = 6(/) + 6(r -
(or so you think, according to die qualitative

boardwalk suggestions o f the professor).
'^ 4
jco
jw
J
X
<>8j
'l6j
-> a
---- 1-------0.2
*-8 j
-2
-1
8j
II1fV ^W
-Bj
-16j
Figure P I4.51
Figure P I4.5 0
To make the gain o f the transfer func
5 2. Figure P 1 4 .5 2 shows the magnitude fre
tion 1 for large frequency you must
quency response o f a transfer function
choose what value o f A" in
whose numerator and denominator are both
( .9 -
fourth order. Qualitatively determine the struc

ture o f
( . V - /?,)
(b)
1I
X
(a)
1
I
H{s)
H{s)
approximately.
Using M ATLAB, compute the

frequency response (magnitude
and phase) over the range 0 < (o
< 2 0 rad/sec. W h at range o f fre
quencies are rejected, approxi
mately?
51. (This problem focuses on a qualitative

understanding o f the poles and zeros o f a
frequency response.) You have come up
with the pole-zero plot o f Figure P 14.51
for a possible band-pass filter with a sec
ond-order zero at the origin.
(a)
Figure P i4.5 2
1
->
To make the gain o f the transfer func

tion 1 for
CO
= 6 you must choose what
value o f A'in
ANSWI-H
(.v~ + l)~
(.v + O . l r - f l
(5 - z, )
5 3. Figure P i 4 .5 3 shows the magnitude fre

(b)
Using M ATLAB, compute the fre

quency
response
(m agnitude
quency response o f a transfer function
and
phase) over the range 0 < co < 2 0

rad/sec.
(c)
Is the filter o f the band-pass type?

W hat range o f frequencies are passed,
approximately? W h at ranges o f fre
quencies are rejected, approximately?
Frequency-rad/s
Figure P i4.5 3
H{s).
Chapter 14 Laplacc Transform Analysis III; Transfer Function Applications
7SS
W hich o f the following is the best candidate for
equivalent, the magnitude and phase
H{s)>
responses 0 < to < IO.OOOti rad/sec for
( 1) K
(2) K
(3) K
(4) K
(5) K
(6 )
the second-order low-pass Butterworth
(^ + 0 .5 + ;5 ) ( 5 + 0 . 5 - y 5 )
filter
l(s) = --------------- , --------^------------------------ .

(5 + 0 .5 + y5)(i- + 0 . 5 - y 5 )
+ 1.414
{s + j l ) { s - j 2 )
j(5 + 0 .5 + y5)(.? + 0 . 5 -
which has a 3 dB down
j5)
____
the
(5 + 0 .3 + y4)(.v + 0 . 3 - y 4 ) ( i + 0 .8 + y6)(.v + 0 .8 - y6)
{ s + j 2 ) { s - j 2 ) ___________________
.3 + y 4 ) ( 5 + 0 . 3 - 7 4 )(5 + ().8 + y6)(.v + 0 . 8 -
(s+j2)(s-j2)
(7) K
5
( 5 + 0 .8 +
7 6
X5 +
7 6
steady-state
Butterw orth
phase
and
filter when
the
input is (i) cos(5007rr), (ii)
(5 + 0 .3 + y 4 )(5 + 0 . 3 - y 4 ) ( 5 + 0 .8 + y 6 ) ( 5 + 0 . 8 - y 6 )
5 (5
at 1000 Hz.
magnitude o f the output o f the
s { s + j 2 ) ( s - j 2 ) __________________
___________________
[X )in t
(c) Referring to part (b), find
(s+j2){s-j2)
__________________
+ 1
2000:1/
\ 2 0 0 0 Ji/
cos(20007rf),
and
(iii)
cos(40007tr).
)
Do your answers to part (c)
make sense? Why?
.8 -
7 6
Remark: T he
of
A N SW E R : (5)
gain,
or the magnitude
is t}^pically given in decibels
and computed as gain (dB) = 20
54. T he circuit in Figure P I 4 .5 4 is to be fre

quency and magnitude scaled so that the value
o f H{s)
ats= j5
s = yiOOO.
becomes the value o f
56. (a)
at
Plot, using M ATLAB or the equiva

lent,
If the final value o f the capacitor is to
the
magnitude
and
phase
responses for the second-order nor
be 1 mF, then what is the new final value o f the
malized high-pass Butterworth filter
inductor,
transfer function
(in H)?
H {S ) =
1Q
- + 1 .4 1 4 5 + 1
4H
over the range 0 < (O < 5 rad/sec.
1F
(b)
Repeat part (a) for the denormalized

high-pass filter
Figure P I4.54
55.(a)

lent,
the
magnitude
and
2000;t
H(s) =
phase
+ 1.414
responses for 0 < to < 5 rad/sec for the

2
second-order
normalized
low-pass
(c)
(b)
" + 1 .4 1 4 5 + 1
+ 1
\ 2 0 0 0 jt/
over the range 0 < (jl) < 10,000jr rad/sec.
Butterworth filter transfer function

H(.5) =
OOOJ1 /
Referring to part (b), find the steadystate phase and magnitude o f the out
put o f the Butterworth filter when the
Denormalization to achieve a different 3
input
dB down point is done by frequency
cos(20007tf), and (iii) cos(4000Tr/).
scaling. Plot, using MATLAB or the
is
(i)
cos(5007ir),
(ii)
~S6
57. Frequenc)' responses are typically plotred as
malized high-pass Chebyshev filter
\H{j(S))\ vs.
transfer function
logjQicol or as gain = 2 0 log,()l//(yo))l
clB vs. logjQCO. These plots are com m only called

(a)
~T
0 .7 0 7 9 .S -
4 - 0 .6 4 4 9 5 +
Repeat parts (a) and (b) o f Problem

55, except plot only the gain in dB vs.
logjolcol.
(b )
0 .5 0 1 2 .r
H{s) =
Bode plots.
over the range 0
< f<
1 Hz.
Repeat part (a) for the denormalized
(b)
Repeat parts (a) and (b) o f Problem
high-pass Chebyshev filter
5 6 , except plot only the gain in dB vs.

0 .5 0 1 2
logjolcol.
H{s) =
5 8 .(a)

lent,
the
m agnitude
responses for 0 <

second-order
CO
and
0 .7 0 7 9
\200 0 7 1 /
phase
low-pass
(c)
Chebyshev filter transfer function
(b)
put o f the Chebyshev filter when the
0 .5 0 1 2
Referring to part (b), find the steadystate phase and magnitude o f the out
---------------
H{s) =
200071/
over the range 0 < / < 1 0 0 0 Hz.
< 5 rad/sec for the
normalized
\ 2 0 0 0 Jt/
input
+ 0.6449.s + 0 .7 0 7 9
is
(i)
cos(5007rr),
(ii)
cos(20007t^), and (iii) cos(40007i^).
Denormalization to achieve a different

3 dB down point is done by frequency
6 0 . The circuit in Figure P I 4 .6 0 , with ^ = 1 Q,
scaling. Plot, using M ATLAB or the
C = I F, and Z, = 2 H, has a so-called third-order
equivalent, the magnitude and phase
Butterworth low-pass characteristic. T he trans
responses 0 <
fer function o f the filter is
CO
< 1G.OOOtt rad/sec for
the second-order low-pass Chebyshev

filter
s +2i +25' + l
0 .5 0 1 2
H{s) =
(a)
-h 0 .6 4 4 9
+ 0 .7 0 7 9
H{s)
. Verify
that their magnitude is 1, meaning
2000 Jt
200071/
Calculate the poles o f
that they lie on a circle o f radius 1 in

which has a 3 dB down point at 1000 Hz.
(c)
Referring to part (b), find the steady-
Using MA'l'LAB or the equivalent,
state phase and magnitude o f the out
plot the frequency response for 0
put o f the Chebyshev filter when the
1 Hz.
input
(d)
the complex plane.

(b)
is
(i)
cos(5007tf),
(ii)
< f<
\NSWH
cos(20007tr), and (iii) cos(40007rr).
- [0:0, 01:
Do your answers to part (c) make
= 0 .5 ;
sense? Why?
Remark; The
gain,
or the magnitude
o f //(/C O ), is t)'pically given in decibels
= freqsi n,d,\v);
p lot(w /(2 pi). abs(h))
and computed as gain (dB) = 2 0

(c)
If
is chosen so that
Kjr= (??);rso
S ]59.(a) Plot, using M ATLAB or the equiva
lent,
the
magnitude
and
phase
responses for the second-order nor
R=
10
find
th a tC = 1 5 .9 1 5 /iE W h at
is the new value o f Z? The filter trans

fer function becomes
U.5
H (s ) =
(a)
\3
+2
+2
(ss)
^>/
Assuming an ideal op amp, show that

the transfer function is
+1
s-
Plot the new frequency response
s+
(magnitude and phase) for 0 s / s

1 0 0 0 // .
(d)
(b)
From your plots in part (c), determine
1
RC
1
RC
Assuming that R = 20 kQ and C = 1

fiF, use MATLAB or the equivalent to
approximately the magnitude and phase
plot the magnitude and phase (in
of the steady-state output voltage when
degrees) responses for 0.1 < (o < 1000
the input is (i) 10cos(20;cf),
rad/sec.
(ii) 10cos(200;rz), (iii) 10cos(1000;ir),
response, it is apparent why this cir
(iv) 10 cos(2000;iy), (v) 10cos(5000;cf).
cuit is called an all pass network.
From
the
magnitude
'.w Q
Figure P I4.60
61. The circuit of Figure P I 4.61 has transfer
W
fiinction
H{s) =
R\Ci
1
^/?lCi
1
R i^ i
s+
^ iQ /
0 6 3 . Figure P I4 .6 3 is a circuit that realizes a
^ iQ ^ 2 ^ 2 band-pass characteristic. Use SPICE to deter
With Cl = C2 = 1 mF, R^=A kH, and /?2 = 0.5

kQ, plot the magnitude o f the frequency
response; i.e., plot |//(/to)| as a function of 0)
over the range 0 < O) < 5 rad/sec and 0 < O) <
50 rad/sec
mine the frequency response o f the circuit and

verify that it is a band-pass characteristic. Note
that the computation o f the transfer function
for this circuit is quite involved.
+
V .Jt)
'" 6
Figure P I4.61
62. The circuit of Figure P 14.62 is called an all
pass network This means that the circuit does

not alter the magnitude of a sinusoidal wave
form, but it does introduce a new phase shift
depending on the frequency of the sinusoid.
Figure P i4.63
Si64. A certain band-pass circuit has transfer
flmction
758
+ 1 3 9 .9 0 f - ^ )
VlOO^
^VlOO/
H{s) =
5
5^ + 2 .7 9 9 6 f )
UOO/
+ 0 .9 9 8 9 f ;^ ) ^
VlOO/^
4
+ 1 3 .0 5 l f )
3
+ 1 7 .2 7 lf )
\m J
VlOO/
+ 1 3 .0 5 1
+ 2.7991 (
) + 1
UOO/
100^
Plot the magnitude frequency response in dB
tie attenuation. This can easily be seen, since
on a semilog scale to verify a band-pass charac
the transfer function has the form

o
teristic. Find an appropriate range of (0.
65. The circuit o f Figure P I 4.65 is the Sallen
H{s) =
Vo(s)
+
and Key low-pass filter, which can be used to
RC
eliminate unwanted high-frequency noise.

(a)
5+
In l
R^C
o
V.
v:-in
___ K
Observe that if s = ycOg = //RC, then
RxR'i C\C2
\-K\
1
//(/COq) = 0, i.e., Vo(/(Oq) = 0 for
+
/?2 Q
input sinusoids o f the form
s+
y4sin(cOo^ + Cp).
^2^2/
where K = 1+/?^^^. Hint: You will

(a)
Use nodal analysis to derive the trans
the indicated nodes.
fer function.
Now let Cj = C2 = 1 F, /?j =

=
1 Q, and R^ = 0.8 Q. These values
(b)
Show that the poles (zeros o f the
represent a normalized design. Use
(c)
need to write three node equations at

(b)
denominator) are ( - 2 S y R C .
L e t / ? = 5 / J t 1 .5 9 1 5 l a n d C = 0 .0 1
MATLAB or the equivalent to plot the
F. Plot the magnitude and phase
magnitude and phase responses for 0 <
responses over the range 0 to 2 0 0 Hz.
CO< 5 rad/sec.
---------- r ^ ( -
"
---- 1--1
R
r\
r> 0.5R
v^(
1 2C <
-----L---------- n
r\
Figure P I4.66 Twin-T RC circuit.
67. Suppose the impulse response of the net

work in Figure P I4 .6 7 is
Figure P i4.65 Sallen and Key low-pass filter.
= [2e^^+ 3cos(2/) - 1.5sin(2/)](r) V
66, Figure P I4 .6 6 depicts the so-called twin-T
(a)
Compute the transfer function, H{s).
RC circuit. This network is often used as a
(b)
Find all poles and zeros, and draw the
band-reject filter, i.e., a circuit that stops or

rejects signals at certain frequencies and allows
signals at all other frequencies to pass with lit-
pole-zero plot.
(c)
If there is no initial stored energy in

the network, i.e., no initial conditions.
759
Chapter 14 Laplace Transrorm Analysis 111: Transfer Function Applications
com pute the network response to
(b)
Find the response o f this same relaxed
(d)
= b(t)
circuit to the input
= sin(r)//(/) V. Does your answer

(c)
make sense? W hy or why not?
V.
= 5b(t-
cos{lt)n{t) V.
For large t, show that Vg,{t) ~ Kt
cos(2t + (j)) V for appropriate K and (J).
70. For this problem you are to use the proper
Does the response remain finite as r
ties o f Laplace transforms as listed in Chapter
Now suppose
00
? Is the network stable or unstable?
2) V.
12. Suppose the Laplace transform o f the

impulse response,
h{t),
W(.v) =
vJt)
o f a circuit is given by
s+2
(,s + l)(.v + 3)
H^{s) and Hf{s)

h^{t)=h{At)
(b)
hît) =
Find
when
(a)
8)
Figure P i4 .6 7
property that its zero-state output is always
INITIAL AND FINAL VALUE

PROBLEMS
equal to the derivative o f its input shifted in
7 1 . (a) The Laplace transform o f a signal y(r) is
68. A time-shift differentiator circuit has the
T > 0. Hence, its impulse

response is h{t) = h'{t - 7 ), a shifted version o f
the derivative o f the delta function. Suppose T
time by some
( 5 .v - l ) ( 4 .v - 5 ) ( 6 .v - 2 )
5 (2 5 -h l)(3 .s -F 2 )(5 i + 4 )
= 1 and the input to the circuit is given byy(f),

as sketched in Figure P I 4 .6 8 . Com pute the
Find the values o f /(O^) and /(co).

(b)
zero-state response^Cr) o f the circuit at ^ = 2 .5

and f = 6 seconds.
Repeat part (a) when
the Laplace
transform o f a s i g n a l / r -
2)u{t-
2) is
^_ 2 , (l(Xs - l ) ( 8 . v - 5 ) ( 1 2 . v - 2 )
^
.v(5.v+l)(3.s'-h2)(i- + 4 )
7 2 . The output,
o f a particular circuit is
engineered to track a reference signal,

After the circuit overheated, it was found that
1
4s- - 4s s-
+ 2i' + 2
Figure P i4.6 8
A N SW ER ; - 2 , 1
Find the error, error(r) =

large
6 9. The Laplace transform o f the response o f a

relaxed circuit to a ramp
v^{t)
t.
for
W hat was the initial error, error(0")?
C H E C K : for large
t,
error is 3
= r ( f - 1) V is
73. Given the following fimctions
F-{s),
find
and/y(3o) for those cases where the initial(a)

v-Û) = u{t)
V.
value and final-value theorems are applicable (a

root-finding program is useful in this endeavor).
Chapter 14 U place Transform Analysis 111: Transfer Function Applications
APPLICATIONS AND BODE

TECHNIQUES
2 l.v + 5
O.I.v*- +4.V + 3
7 6 . Obtain the asymptotes for the Bode plot o f

2.v^ + 7.v + .y + 4
F,{s) =
(b)
s{s^
+ .V " + 7 . V
H(s)=
(c)
F,is) =
.y(.v +
+ 1
+ 6 )
s~ +4.9 + 3
5s^ + 5s^ + 4.V +
iO-
4)
100
.V
+
V16
\ 400
1200
+ 1
Then verify your asymptotic plots by using

7 4 . The capacitor voltages,
o f various
M ATLAB to generate the true Bode plot,
networks have the Laplace transforms given

below. Use only the initial- and fmal-value theorems to determine
and
7 7 . Obtain the asymptotes for the Bode plot of
for each
o f the transforms. If either theorem is not
H(s)=
applicable, explain why.
10
+ 1
V16
1200
+ 1
(a)
(i 4 .v - i r
100
s~(ls+2)
+ 1
+ 1
V400

M AI LAB to generate the true Bode plot.
(b )
(c )
(d )
7 8 . Obtain the asymptotes for the Bode plot o f

, ^ ^ a ) . l . v - l X 0 . 2 i + lJ
zUU---------- ^----------------.y(.y + 144j
(1 6 5 -1 )
(2 5 + 1 ) -
H(s)=
25-
25+1
V25 + 1
10-
+ 1
+ 1
\ 1200
+ I
4 5 -1 \ -'
16
100
+ 1
V 400
V 25+ I/
I'hen verify your asymptotic plots by using
7 5 . The capacitor voltages,
o f various
MA'I'LAB to generate the true Bode plot. Note
nerworks have the Laplace transforms given
that the shape o f your asymptotic Bode plot
below. Use only the initial-and fmal-value theorems to find /f;;(0') and
i(^^)
for each o f the
resembles that o f Problem 77except that the

slopes are steeper,
transforms given that C = 0 .2 F. If either theo

rem is not applicable, explain why.
7 9 . Obtain the asymptotes for the Bode plot of
10
(a)
5(5
+ 1
+ 2)
H(s}=
(b )
5 (25 + 2 5 )
(c )
20
(^ -2 )
5(25 + 25)
10
\ 20
200
2400
+ 1
+1
+1
800

MA^LAB to generate the true Bode plot.
> 80. Suppose a dc motor is modeled as per
(c)
= 25 Q . / = 0.005 kg-m^, k = 0.0 2 N -
m/A, B = e~^ N-m-s, and
= 100 mH.
W hat effect does R^ have on the

steady-state speed for a step input of a
Figure 14.38 o f the text. The motor parameters

are
"^61
fixed amplitude?
(d)
W hat effect does R^ have on the rate at
Calculate the steady-state angular velocity of
which the armature approaches a
the motor and the steady-state armature cur
steady state for a step input o f fixed
rent for the input 10(r) V.
amplitude?
> 81. Consider the dc motor modeled in
> 8 2 .(a) For the dc motor of Figure 14.38 of
Figure 14.38 o f the text.

(a)
Using the same parameters as in
the text, suppose k = 0.0 5 N-m/A and
Problem 80 with
R^ = 50 Q. Plot the steady-state cur
= 5u{t) V, com
rent as a function of 5 as 5 ranges

from 0 to c.
pute the armature current and the

angular velocity as a function of time.
(b)
Change
(b)
Using /?^ = 25 Q and B = e~^ N-m-s,
to 50 Q, and recompute
plot 0)^ as ^ ranges from 0 to oo. Recall
the armature current as a function of
that increasing k increases the torque
time.
per ampere o f armature current.

Explain why increasing k reduces co^^.
O '
C H A P T E R
Time Domain Circuit Responsi
Computations: The Convolution Method
AVERAGING BY A FINITE TIME INTEGRATOR CIRCUIT
Sometimes one must compute the average o f some quantity, sucii as the average vakie o f the light
intensity on a solar cell over the last fifteen minutes, or the average value o f the temperature in a
room over the last hour, or the average value o f a voltage over the last 50 milliseconds. W hen such
an average is updated continuously, it is termed a running average. The idea is that the readout o f
the device that averages these quantities always produces an updated value valid over a specified
prior time interval. If a voltage represents the value o f the quantity to be averaged, then one can
build a circuit whose output voltage is the required average. This is done by observing that an aver
age value o f a continuous time variable is simply the integral o f the variable over the proper time
interval, divided by the length o f the time interval. A device that integrates a variable over the last,
say,
T seconds
is called a finite time integrator. As an application o f the ideas o f this chapter, we will
look at a finite time integrator circuit and how it can be used to compute the average value o f a
quantit)^ The convolution concept directly leads to the required transfer function o f such a circuit.
CHAPTER OBJECTIVES
1.
Introduce the notion o f the convolution o f two signals.
2.
Using the notion of convolution, develop a technique o f time domain circuit response
computation that is the counterpart o f the transfer function approach in the frequency
domain, presented in Chapter 14. In particular, we seek to show that the convolution o f
an input excitation with the impulse response o f a circuit or system produces the zerostate circuit or system response.
3.
Develop objective 2 from r\vo angles: first, from a strict time domain viewpoint, and sec
ond, as a formal theorem relating convolution to the transfer function approach.
4.
Develop graphical and analytical methods in particular, an algebra for evaluating the
convolution o f two signals.
5.
Illustrate various properties o f convolution that are pertinent to block diagrams o f series,
parallel, and cascade interconnections o f circuits or systems modeled by transfer func
tions.
^6'i
Chapter 15 T im e Domain Circuit Response Computations: T h e Convokition Method
SECTION HEADINGS
1.
Introduction
2.
3.
4.
5.
6.
7.
8.
Definition, Basic Properties, and Simple Examples

Convolution and Laplace Transforms
Graphical Convolution and Circuit Response Computation
Convolution Algebra
Circuit Response Computations Using Convolution
Convolution Properties Revisited
Time Domain Derivation o f the Convolution Integral for Linear Time-Invariant
Circuits
9. Summary
IL
Problems
1. INTRODUCTION
At the beginning o f Chapter 12, we claimed that circuit response computations could take place
in either the time domain or the i-dom ain. Yet, except for the solution o f some very elementary
differential equations, circuit response computations have relied almost exclusively on the Laplace
transform technique. This chapter develops and explores the time domain counterpart o f the
Laplace transform method by introducing the notion o f the convolution o f two signals to produce
a third signal. We then show that the time domain convolution o f an input excitation with the
circuits impulse response yields the zero-state circuit response. Pictorially, the idea is expressed in
Figure 15.1, which is an update ot Figure 12.3.
f(t)
Circuit Impulse
>
Response h(t)
Input Signal
y(t) = h^f
Output Signal
Transform of
Circuit Transfer
Transformed-
Function H(s)
Input Signal F(s)
Output Signal
Y(s) = H(s)F(s)
riCnJRF, 15.1 Diagram showing the symmetric relationship of time and hequenc)'^ domain analyses.
The upper part of the diagram asserts that the convolution, denoted
with the circuit impulse response
h{t)
h * f, of the
input signal/(f)
produces a third signal, ^(r). This third signal is the zero-state
response, which equals
Chapter 15 Time Domain Circuit Response Computations: The Convolution Method
Justification of the diagram o f Figure 15.1 begins with the formal introduaion o f the notion of
convolution. Mathematically, the convolution o f two fiinctions h{t) andy(f), denoted hy h * f or
h{t) * j{t)i results in a third function through the integration process,

>-(0 =
j^ ^ h {t-x )f{T )d T
(1 5 .1 )
The second step in verifying the diagram is proving that the convolution of the input signal with
the impulse response produces the zero-state circuit response. There are two approaches. One is
to work stricdy in the time domain and construct the actual zero-state response from the impulse
response and an arbitrary input excitation. The second approach is to prove that
is
the convolution of the signal y(/) with the impulse response h{t). Because o f the Laplace transform
development o f the last three chapters, this direction seems the most painless and will be taken up
V^
in section 3, after an introduction to the basic ideas o f convolution in section 2. Section 4 will
look at the graphical method of convolution, which helps in visually grasping the definition.
Section 5 will describe a convolution algebra, which yields a harvest o f shortcuts for evaluating the
convolution o f certain types o f signals. Section 6 looks at the computation o f circuit responses
using the convolution method, and here the averaging circuit is developed. Section 7 overviews
various properties o f convolution. The respective properties lend themselves to different structures
(e.g., parallel or cascade) for designing interconnected circuits. Section 8 describes the rudiments
o f the construaion o f the zero-state response o f a circuit working stricdy in the time domain
toward the development of a convolution integral.
Before closing this introduction, we should consider the question, why is convolution important^
One reason is that it allows engineers to directly model the input-output behavior of circuits and
general systems in the time domain, just as transfer functions model circuit behavior in the sdomain.
A second reason is that circuit diagrams are sometimes unavailable or even get lost. How would
one generate a circuit model for analysis? One way is to display the impulse response on a CRT
and approximate
O '
by some interpolation function such as the staircase approximation illus
trated in Figure 15.2. The process o f constructing an approximate impulse response o f an

unknown circuit or system is called system identification and is a vibrant area o f research. By storing the measured impulse response data as a table in a computer, one can numerically compute
the zero-state response of the circuit or system to an arbitrary input signal. This numerical simulation process lets an engineer investigate a circuits behavior off-line. For example, simulating a
circuit destined for use in a hazardous environment provides a cost-effective means of evaluating
its performance in a simulated environment. Such performance evaluations often identify needed
design improvements prior to the constructing and testing of prototypes.
766
FIG UIIE
Chapter 15 * T im e Domain Circuit Response Computations: T h e Convolution Method
1 5 .2
A rectangular approximation to a hypothetical impulse response obtained from meas

ured data on an unknown circuit or system.
A third important reason for studying convolution in the context o f circuits relates to a deficiency
o f the one-sided Laplace transform: function segments that are nonzero for f < 0 are ignored by the
one-sided Laplace transform technique. Hence, time domain convolution offers a direct means of
computing the circuit response when signals are nonzero over the entire time interval, -<
< t< cc.
2. DEFINITION, BASIC PROPERTIES, AND SIMPLE EXAMPLES

As mentioned, convolution is an integral operation between two functions to produce a third; i.e.,
the convolution o f two functions
h(t) and J{t)
produces a third function j = h* f. One might expect
that the convolution h * /e q u a ls the co n v o lu tio n /* h, i.e., the operation o f convolution is com
mutative. In fact, this is the case. To emphasize this property, we restate equation 15.1 in its more
general form:
CONVOLUTION
T he convolution o f two sig n a ls/r) and
h{t)
produces a third signal, ^(/^), defined according to
the formula
v (/ ) = J
J A t - r)f{T)dT =
T)dx
(1 5 .2 )
which is well defined, provided that the integral exists. This formula emphasizes the proper
ty that convolution is a commutative operation, i.e.,
h * f = f * h.
T he equivalence expressed in equation 15.2 comes about in a straightfonvard manner by the

change o f variable t ' = r - t . In addition to being commutative, convolution is associative,
h * { r g ) = { h- f i * g
(1 5 .3 )
and distributive,
(1 5 .4 )
Chapter 15 T im e Domain Circuit Response Com putations: T h e Convolution Method
Exercises.
1. Verify equation 15.2 using tiie change o f variable T
'=t
767
- t .
2. Illustrate the associative property o f equation 15.3 with a specific example.

3. Verify the distributive property o f equation 15.4. Hint: Integration is distributive.
4. Verify that
a{h *f) = {af) *h =f*{ah).
Hint: Use equation 15.2.
Another very useful property o f convolution is the so-called tim e shift property: if
h{t) * j[t)
then
h{t-T^)^j{t-T^)=g^t-T,-T,)
(1 5 .5 )
Some simple examples serve to demonstrate the actual calculation process. These examples will
utilize the sifting p rop erty o f the delta function: if h{t) is continuous at / =
//(7') = J*^_ h(r)d{T - r ) d T =
h{T
T,
then
-T)d{T)dT
E X A M P L E 1 5 .1 . C om pute the convolution o f an arbitrary continuous function y{r) with

So
6{t) .
lu t io n
By the definition o f equation 15.2 and the sifting property o f the delta function,
y{t) = f * 5 =
-T)d(r)dT
= fit
= f it )
(1 5 .6 )
Example 15.1 makes the point that 6 {t) acts like an identity' for convolution, i.e., it always returns
the function with which it is convolved. T he next example indicates that the convolution o f an
arbitrar)' continuous function y{/) with
ad(t - T)
produces a scaled and shifted version
as
given below in equation 15.7.
E X A M P L E 1 5 .2 . Compute the convolution o f an arbitrar)' continuous fTinctiony(/) with

So
a d {t-
7).
lu t io n
By the definition o f equation 15.2 and the sifting property o f the delta function.
y(t) = f it ) * [adit - T)] = C lJ i t - T)diT - T)cIt = a f i t -
t)^ ^ =
Exercise. Check the associative property o f convolution by

ii)fxit) = d i t - 1) ,^ ( r ) = u{t) - ui t-2 ) , a n d ^ (f) = dit -
co m p u tin g /j
(/;)/; (r) =
+ T^)
d it-
7 , ) , / , ( / ) = u i t ) - u { t - 2 ) , a n d / 3(r) = d i t
3)
*fi*
afit - T)
(1 5 .7 )
:"6 8
Chapter 15 'Fimc Domain Circuit Response Com putations; T h e Convohition Method
Ler us a|3ply the results o f the above two examples to a simple staircase function.
EXA M PLE
1 5
. 3 . Com pute the convolution o f the functiony(/) shown in Figure 15.3a with h { t )
^ ld {t) -2 d {t -
1).
f(t)
iL
2 --
1-
H -------
(a)
FIG URE 15.3 (a) Functiony(f) for Example 15.3. (h ) 7 = / h.
So
lu t io n
By the definition of equation 15.1 and the sifting propert)' of the delta function,
v(/) = / * / / = 2 j
=
- T) d( T - l)r/r
- t ) 6 ( t ) cI t - 2 j
( 15. 8)
/ < r - T - 2 / ( / - T)]^,| = 2 / ( / ) + ( - 2 / ( / - 1))
Graphically combining the result given in equation 15.8 yields the relation in Figure 15.3b.
Exercise. C om pute
25{t-2).
the convolution o f the functiony(r) shown in Figure 15.3a with
h{t)
26{t) -
AN SW 'FR: l f { t ) - l / { r - 2) =
EXA M PLE
1 5
. 4 . Find the convolution o f u { t ) with itself, i.e., n { t ) * u { ( ) .
S o lu tio n
By the definition o f equation 15.2,
ii(i)*u {t)= C
J -o c
u(t-T )u(T)dT =
Jo
n (t-T )d r
(1 5 .9 )
Chapter 15 T im e Domain Circuit Response Com putations: T h e Convolution Methotl
"6 9
since
TÔ
|0
T< 0
We now break equation 1 5.9 into two regions: (i) t > r, in which case
tion 15.9 zero; and (ii) 0 s t s /, in which case
u{t- t)
= 1. W hen 0
u{t- t) = 0 , making equa

t, equation 15.9 reduces
to
u ( / ) * i ( ( f ) = f'i/(r - t ) c/ t = r' (It = tu{t) = r{t)

%/()
In other words,
u{t)
u{t) = r(t),
(1 5 .1 0 )
/()
the ramp function.
W e now apply the time shift propert}' o f convolution to the result o f Example 15.4.
E X A M P L E 1 5 .5 . Find
y = h /w h e n h{t) = ii{t)
a n d /r) =
ii{t + 1)
u{t-
1), as shown in Figures
15.4a and b.
u(t+ D - u(t-l)
>k
>^ u(t)
--- 1
-1
1
1
1
2
> t
^ 1
1
(b)
(a)
y(t)
FIG U R E 15.4 (a)
h{t)
= h(/), the step function. ( b ) / r ) =
u{t +
(c) Resulting y(r) for the convolution /;*/.'
1) - (r - 1).
Chapter 15 * T im e Dom ain Circuit Response Com putations: T h e Convolution M ethod
So
lu t io n
Using the distributive law o f convolution,
y{t) = u{tY[u{t^
1) -
ii{t-
1)] =
u{tYu{t^
1) -
u{tYu{t-
1)
(1 5 .1 1 )
Now using the time shift property o f convolution and equation 1 5 .1 0 , we conclude that
y{t) = u{t) * (r
+ 1) -
1) = K / + 1) -
r(t-
1)
which is plotted in Figure 15.4c. We note that this result was achieved without any direct inte
gration.
E X A M P L E 1 5 .6 . Com pute the convolution^ =
h */ôf the
two waveforms
/j(t) and/(t)
in Figure
15.5.
h(t)
(a)
(b)
y(t)
> X
FIG U R E 15.5 (a), (b) Waveforms for Example 15.6.
{c)y = h * f
S o Lu n o N
We first note t h a t / r ) =
tive law o f convolution.
ii{t)
+ (r - 1 ) -
lu{t -
2) and
h{t)
u{t) - ii{t -
1). Using the distribu
y{t)
[u{t)
= [(^) + ( / - 1) -
u{t-
1) -
l i i i t - 2)Yti{t)
2u{t-2)V[u{t) - u{t- [//(r) +
ti{t-
1) -
1)]
2)]*//(/- I)
lu {t-
Now using the time shift property o f convolution and equation 1 5 .1 0 , with a further application
o f the distributive law, wc conclude that
y{t) = [r{t) + r { t - \ ) - 2r{t- 2)] - [r(t- \) + r{t- 2) -
2 r ( r - 3)] =
r{t) - M t - 2 )
2 r it-
3)
which is plotted in Figure 15.5c. We again note that this result was achieved without any direct
integration.
E X A M P L E 1 5 .7 . C om pute the convolution
u{t-
h * /o f
the signals
h{t)
a n d /(r ) =
e ^[u{t)
7 )], as given in Figure 15.6.
h(t) >k
0
(a)
FIG U R E 15 .6 (a), (b) Fu n ction s/r) and
h{t)
for the convolution of Example 15.7.
(c) The resulting convolution.

So
lu t io n
Step 1.
Apply definition and adjust limits o f integration.
Applying the definition o f equation 15.2
yields
\{t) = h * f =
-T)f{T)dT =
i/(t
-T)]dT
Chapter 15 T im e Domain Circuit Response Com putations; T h e Convolution Method
Since ;/(x) - z/(x - 7 ) is nonzero only for 0 ^ x :s T, the lower and upper limits o f integration
T,
bccome 0 and
respectively:
y{t) = h * f = ^ ^
Step 2.
e~û{T-t)clT
(1 5 .1 2 )
Detennirie the regions o f t over luhich the integral is to be evaluated. From

t), there are three regions o f interest: (i) r < 0, (ii) 0 s ^ s
gration and the w(x -
the limits o f intex <
T, and
(iii)
^ t.
Step 3.
Evaluate the convolution integral, equation 15.12, over the given regions.
Case I: t < 0.
Z and
Here, r < 0 implies that x - f > 0 over 0 s x s 7". Hence, u(t - /) = 1 over 0 s x ^
j.
y( 0 =
Case 2: 0 ^ t < T.
For this c;ise, u { t
- t)
e~'" c/t = 1- e~^
in equation 1 5 .1 2 is nonzero only when
i >t.
Hence, in
the region 0 ^ f < 7 , it must also be true that 0 < r s x < 7 for the integral o f equation 1 5 .1 2 to
be nonzero. Thus, the lower limit o f integration with respect to the variable x becomes
v (/) = J ^
Case 3: t>T. A simple

Step 4 .
Plot y{t).
e 'û{T-t)dT =
calculation shows that
_^ n 7 -
=e
- e
f.
= 0 in this region.
Combining the results o f step 3 with 7 = 1 implies that y(r) has the graph
sketched in Figure 15.6c.
Exercise.
Find the convolution, say, y(/), o f the signal J{t) =
A N SW E R :
Ka
'//(- /) f
e ^hi{t)
with
h{t)
Ku{-t).
Ka
3. CONVOLUTION AND LAPLACE TRANSFORMS
As claimed in the introduction, circuit analysis in the time domain by convolution and circuit
analysis in the frequenc)' domain by Laplace transformation are equivalent in terms o f zero-state
response calculations. T he purpose o f this section is to rigorize the equivalence between the time
and frequenc)' domain analysis methods by formally show'ing that
presents a time domain rendition.
L[h ^j] = H{s)F{s).
Section 8
CONVOLUTION THEOREM
h{t)
Supposey(/) = 0 and
= 0 for / < 0. Then

(1 5 .1 3a )
i.e., convolurion in the rime domain is equivalent to multiplication o f transforms in the fre
quency domain; or equivalently,
hit) ^j{t)
r^[H{s)F{s)]
(1 5 .1 3 b )
T he justification o f this theorem proceeds as follows:

Step 1. Given equation 15.13a, the first step is to write down the transform o f
h[t) * J{t).
Specifically,
f /;(/)*/(Ol =
-T)//(/ -T)/(T)^/rje " J/
(1 5 .1 4 )
A couple of points are in order: (i) the inner integral, surrounded by parentheses, represents the
convolution
h{t) * J{t);
emphasize the fact that
and (ii) the presence o f the step function
h{t) = 0
u{t - r)
is added as an aid to
for r < 0.
Step 2 . The goal at this point is to manipulate the integral o f equation 1 5.14 into a form that can
be identified as the product ol the Laplace transforms o f two functions, i.e., as
H{s)F{s). The
only
possible strateg)' is to interchange the order o f integration and group appropriate terms. Note that
the Re[^] must be chosen sufficiently large to ensure the existence o f the Laplace transforms of
both
h{r)
Under certain conditions that are t\'pically met, it is possible to interchange the
order o f integration within a com m on domain of convergence of
H{s)
and
F{s).
Interchanging the
order and regrouping the f-dependent terms inside a single parenthetical expression produces
( / ? { / ) * / ( / ) ] = J ^ _ / ( t ) ^J^_/;(/ - T) / / ( /
(15. 15)
Step 3 . Observe that the interior integral, surrounded by parentheses, in equation 15.15 is sim
ply the Laplace transform o f a time-shifted
e~^H{s)
i.e.,
[M{t -
r)] =
Substituting
for the interior integral leads to the desired equivalence:

|/ i ) / ( O I = j ; f ( T ) H U ) e - 'dr =
= H ( . v ) f ( ,v )
This theorem asserts the equivalence o f convolution o f one-sided signals with multiplication of
their transforms in the /-dom ain. For our purposes,
h{t)
assumes the role o f the impulse response
o f our circuit andy(r) the role of the input excitation. Accordingly, the convolution o f the impulse
response o f a circuit or system with an input signal, a time domain com putation, equals the
inverse transform o f the product o f the respective Laplace transforms. In other words, the diagram
o f Figure 15.1 is correct, as claimed under the conditions o f the theorem.
Chapter 15 * T im e Domain Circuit Response Com putations: T h e Convolution Method
Exercises.
1. An unknown relaxed linear system has impulse response
the response to the input signaly(f) = l u { t
- 1)
b{t)
n{t) - u{t -
1). Find
using convolution and check your answer using

A N SW E R :
h it -
1) - 2 r ( / - 2)
2. An unknown relaxed linear system has impulse response

sig n a l/r) =
e~û{t).
Find the response ^(f) =
h{t) = (a - b)e~û{t),aj? > 0
and input
h{t) * fit).
A N SW ER :
T he conditions o f the theorem are somewhat restrictive in terms o f computing circuit responses
strictly in the time domain. Specifically, it is the one-sided Laplace transform that does not rec
ognize function segments over the negative real axis hence the condition on the input excitation
t h a t / r ) = 0 for /< 0. This restriction does not lend itself to the computation o f initial conditions
and circuit responses due to input signals extending back in time to / =
In general, the con
volution o f an input excitation with a circuits impulse response presupposes no such restriction.
However, justification o f the computation o f zero-state responses due to input excitations extend
ing back in time to / = -oo cannot be based on the convolution theorem o f the one-sided Laplace
transform. A justification o f the time domain convolution approach to computing circuit response
is reserved for the last section o f this chapter, due to its complexit)^ Nevertheless we will use the
result as necessary, such as in the next exercise.
Exercise, T he
transfer function o f a particular system is
H{s)^
.9+ 0.2
C om pute the convolution o f the impulse response with the input

A N SW ER : lOlI -
+ 1) - 1()[1 -
')],,(/
v{t),
shown in Figure 15.7.
v(t)
J <
,1
-1
FIG U RE 15.7
4. GRAPHICAL CONVOLUTION AND CIRCUIT RESPONSE

COMPUTATION
T he convolution integration formula, although explicit, has many layers of interpretation.
G raphical convolution is a pen-and-pencil technique for determining the convolution integral o f
simple, squarish waveforms. T he technique often leads to a more penetrating insight into the con
volution integral.
There are four key ideas in the graphical procedure: flip, shift, multiply, area. The following is a
detailed description o f the procedure.
To compute
t = T,
given by the convolution integral o f equation 15.1 or 15.2, for a specific value o f
perform the following steps:
1.
Plot
h{x)
vs. T andy(T) vs. T curves.
2.
Flip t h e / r ) curve about the vertical axis (t = 0) to obtain t h e / - r ) vs. T curve.
3.
T to obtain t h e y ( r Multiply. Plot the product h{z)j{T - r) vs. x curve.

Area. Calculate the area beneath the hir)j{T - r) vs. x curve for x < t
Shift th e y (- t) curve to the right by the am ount
4.
5.
t)
t curve.
v s.
< oo. The result
is7(T ).
We make the following remarks with regard to the above steps:
(1)
The roles o f /;( t ) andy(T) may be interchanged because o f the commutative propert)' of
convolution, equation 15.2. In other words, we may flip and shift /?(t) instead oF /( t).
Usually we flip and shift the simpler waveform.
(2)
In step 3 , if T is negative, the shift is actually to the left.
(3)
As r i s varied from -oo to +oo, a complete plot of^(^) is obtained. In shifting and find
ing the area, we often have to divide T in to separate inter\'als, becausc each interval may
require a different formula to com pute the area beneath the
t, h{t-T)
vs. x curve.
t moves
from
to cc,
x) moves along the x-axis from x = -oo to x = oo; a simple illustration is h(t - x) =
u{t -
x), w'hich
With regard to step 2, for each
h{t -
h{T)j{T- r)
is a shifted horizontal flip o f/;(x): as
is sketched in Figure 15.8. Part (a) o f the figure shows (x), part (b) depicts
u{t-T),
which slides to the right along the x-axis as
ftuKtion
h{t-x)
u{t-
t increases.
u{-x),
and part (c) plots
For comparison, we can consider the
1 - t), whose three forms are given in Figures I5.8d through f
u(-l) ><
>k u(x)
u(t-x)
>k
increasing t
----w -
---------------- ^ ^
(b)
(a)
(C)
increasing t
u(-T-l) A
A u (x -l)
t- 1
-1
(d)
(e)
h{t-x)
h{t-x) = u{t-x - 1),
(f)
FIG URE 15.8 (a)-(c) For the function
= ( r - x ) , plots of (a) (x), (b) ( - t ) , and (c) ( r - x ) .
(d)-(O For the function
plots of (d)
tt{x -
1), (e) //(-x - 1), and ( 0 (^-x - 1).
Chapter 15 T im e Domain Circuit Response Com putations: 'I'he Convolution Method
E X A M P L E 1 5 .8 . Graphically compute the convolution y = / / / o f the two waveforms
t)
and JU) =
2u{-t)
h{t) = u{-
shown in Figure I 5.9.
f(t)
"
k.
f(T)
I
h(t-T) = u (t-1 )
h(t)
increasing t
y
(a)
(b )
FIGURE 15.9 Waveforms for Example 15.8.

So
lu t io n
T he graphical solution to this convolution depends on a partitioning o f the time line into special
segments over which the graphical convolution is easily done. There are two regions to consider:
(1)
-0 0
< / :s 0 and (2)
t > 0.
Step 1. Consider the region r > 0. Figure 15.9b shows that h{t - t ) J { t ) = 0 for all r > 0 and all T .
Thus the convolution integral is zero and
= 0, r > 0.
Step 2. Consider the region -oo < ^ ^ 0. From Figure 15.9b , the product h{t
=1 x 2 = 2 for
r ^ T s 0 and zero elsewhere. Consequently, the area under the nonzero portion o f the product
functions
h{t -
is
y (0 =
hit - T )f{ T ) d r =
dr = -It
Step 3. Combine the foregoing calculations into a plotted waveform. Figure 15.10 shows the func
tion j/(r) resulting from the convolution.
y(t)
F IG U R E 1 5 .1 0 P lo t o f th e resu ltin g fu n c tio n , ^(/).
Chapter 15 T im e Domain Circuit Response Computations: The Convolution Method
Exercise.
In Example 15.8, supposey(/) is changed
ioj{t)
2u{t) - 2ii{t 1).
M ndj'(r) at r = 0 .5 ,
1.5, and 2 .5 sec by the graphical convolution method.

A N SW ER : 1 , 1 , 0
Another, more complex example will end our illustration o f the graphical convolution technique.
E X A M P L E 15.9. Com pute the convolution
j{t)
o f the triangular pulse
h{t)
with the square pulse
as sketched in Figure 15 .1 1 .
f(t)
T
(b)
/ \
/ \
\
/I
\
/
/
/
-1
Time (seconds)
(c)
FIG U R E 15.11 Convolution of triangular signal (a) with square pulse

(b) to produce the signal (c) for T'^ 2 for Example 15.9.
T he goal, o f course, is to graphically evaluate the convolution integral o f equation 15.2 using the
following steps:
Chapter 15 Tim e Domain Circuit Response Com putations: T h e Convolution Method
h{t-T)
1.
Since x is the variable o f inregrarion, draw
2.
Evaluate the product
3.
Com pute the area under the product curve for each region determined in step 2.
Step 1.
Drauf h{t -
h{t -
a n d / x ) on the x-axis.
x)J{t ) for various regions o f
t.
x) audJ{T) on the T-axis for f < 0 and compute the area o f their product. Figure
15.12 shows h{t - x) andy(x) on the X-axis. From the figure, it is clear that h{t 0; hence,
x )fj)
= 0 for ? <
= 0 in the first region.
FIG U R E 1 5 .1 2 Graph of h{t-x) and/(x) on x-axis for r < 0.

Step 2.
Consider the region 0
^ r <
as illustrated in Figure 15.13.

h(t-x)f(T)
A
-- 2T
Area
2T-t
t-2T
FIG U R E 1 5.13 Graph of h{t-x)fix) on the x-axis for 0 s / <
T.
T he shaded area o f the figure is the difference between the area o f the large triangle, defined as
Area A = 0 .5 (2 7 )^
and the area o f the smaller triangle to the left o f the vertical axis, defined as
Area B = 0 . 5 ( 2 7 -
Hence,
j/(r) = Area = Area A - Area B = 2 Tr - 0 . 5 r
for 0 :s r <
T.
Alternatively, one may use the formula for the area o f a trapezoid, i.e., the average
height times the base, in which case one immediately obtains 0 . 5 ( 2 7 +
I T - t)t
= Area =
ITt-
0.5/^.
Step 3 .
Now consider the region
t <2T, as depicted in Figure 15.14.

h(t-T)f(T)
A
-- 2T
Area
t-2T
> I
FIG U R E 1 5.14 Graph of h{t-x)j{T) on the x-axis for
T^ t <1T.
In this figure, the shaded area, which determines j'(f), is again the difference o f two triangular
areas, specifically.
y{t) = 0 . 5 [ r - { t - I T ) ] for
- 0 . 5 [ 2 r - r]2 = 2 .5 7 ^ -
Tt
T^t<2T.
Step 4 . Figure 1 5 .1 5 shows the next region,
2T ^ t < 3T.
Another straightforward calculation
yields
y{t) = 0 .5 ( 3 r - /)2 = 4.5T^ - 5 T t ^ 0.5t^
for 2T^ t < 3T.

h(t-T)f(x)
A
- 2T
Area
> T
t-2T
F IG U R E 1 5 . 1 5 Graph o f
h{t-T)j{T)
on the x-axis
ioxlT^ t < 5 T
780
Step 5.
Consider the region 5T^ t.
Here, the product
h{t-T )/(t)
= 0 , in which
caseyit)
= 0 for r
>37:
In sum,
0,
v(0 =
/< 0
2Tt-0.5r,
0^t<T
2.5T^-Tt,
T^[<2T.
4.5T- -3T f + 0.5r, 2 T ^ t < 3 T

0.
A plot of^(r) appears in Figure 15.1 Ic For
T=
2.
Exercises. 1. Repeat the calculations o f the preceding example, except flip and shift J{t) instead of
h{t). Here, one looks zx. j{t - r) sliding through h{x). The calculations should be easier and the
result the same.
2, Find the output o f a linear and relaxed circuit with in p u t //) and impulse response
h{t)
(shown
in Figures 15 .1 6 a and b) at time r = 2 .5 seconds. Hint: Use graphical convolution.

A N SW E R : !
MGURF. 15.16
3. T he impulse response o f a particular circuit is approximately measured on a scope as illustrat
ed in Figure 1 5 .1 7 . If the input to the circuit \s J{t) = u{t - 1 ) -u{t
response, y{t), at / = 3 sec and r = 3 sec.
/VNSW ER S: 2, 1
2), find the value o f the
81
Chapter 15 * Tim e Domain Circuit Response Com putations: T h e Convolution Method
4. For
h{t)
andy{r) as sketched in Figure 1 5 .1 8 , fin d ^ (l) for the convolution ^(r) =
h{t) j{t).
A N SW ER : 4
h(t)
-1
FIG U RE 1 5 .1 8
Sometimes the foregoing graphical techniques prove difficult to execute. Nevertheless, an under
standing o f them offers fundamental insight into the meaning o f the convolution integral. A use
ful set o f techniques for quickly evaluating convolution integrals arises from the properties o f a
convolution algebra, discussed in the next section.
5. CONVOLUTION ALGEBRA
A set o f functions, together with operations called addition and multiplication, is called an
bra,
alge
assuming certain conditions are satisfied. The set o f all functions that can be convolved with
each other also constitutes an algebra with respect to the operations o f addition (+), and convolu
tion (*). This set, together with the two operations, is called a convolution algebra. In this con
text, operations such as differentiation and integration are inverses o f each other. For example,
integrating a function and then differentiating the result returns the original function. W ithin the
convolution algebra, the co n v o lu tio n /*
the derivative o f
g is equivalent
to the convolution o f the integral o f/w ith
The advantage here is that, by successive integration and differentiation, it is
often possible to reduce an apparently difficult convolution to a simpler one.

For a set o f functions to be an algebra with respect to + and *, several arithmetic operations must
hold. In particular, + and must be both commutative and associative. The commutativit}^ and
associativity o f + is clear:/ + g = ^ + / a n d / +
ciativity o f * is equally clear: f *
g =g * / a n d
functions that are mutually convolvable must

/*
h.
(g + h) = ( f ^ g) + h. T he commutativity and asso

{g* h) = {/* g) * h. To be an algebra, the set o f all
also satisfy the distributive la w ,/* {g+ h) = f * g +
In addition to obeying these laws, algebras o f functions must contain identity elements. For
+, the zero function ser\'es as the identity. For convolution, the delta function plays this role. The
delta function is an identity element because o f the sifting property as set forth in Example 15.2
and equation 15.7.
For our purposes, the interesting aspects o f a convolution algebra o f functions rests with the inter
relationship o f convolution, differentiation, and integration. To map out this kinship, we use the
following notation
/
/(t)A
(1 5 .1 6 )
and
/,< !)(,) = ^
(1 5 .1 7 )
dt
It can be shown that
/ * / 7 = / ( ) * // - I )
if /(-o o ) = 0 and
t=
(1 5 .1 8 )
exists. T he co n stra in t/(-o o ) = 0 implies that the derivative o f / / ) is zero at
-oo and the constraint that
exists means that the integral o f
semi-infinite interval (-oo, /] for all finite
t.
h{t)
has finite area over the
Similarly,
/* /,= /(- )
(1 5 .1 9 )
if /;(-oo) = 0 a n d /^ '^ exists.
u{t)
u{t
T'he goal is to use equation 1 5 .1 8 to evaluate the convolution, i .e .,/ *
=/^'^
E X A M P L E 1 5 .1 0 . Find the convolution
y =
f o r //) =
- 1 ) and
h{t) = uU) -
//(/ - 1 ) , as sketched in Figures 15.5a and b.

So
lu t io n
= 5(r) + 6 ( / - I)
-2 6 { t -
2) and
r{t) -
, where/^*^(/)
K ^-1), as presented in Figure 15.19.
Chapter 15 * T u n c Domain Circuit Response Com putations: T h e Convolution Method
/H3
(b)
FIG U R E 1 5 .1 9 (a) The derivative ofy(r) =
(b) The integral of h{t) =
Since f^\t) =
+ 6{t - \) -26
u{t) - u {t -
n{t)
+ (r-l) -
2u{t -
2),
1), as given in Example 15.10.
(/ - 2 ), a sum o f impulse functions, the sifting propert)- o f the
impulse function implies
1) -
yU) =f^
W ith the picture o f g i v e n
1)
(1 5 .2 0 )
in Figure 15.19b , the right-hand side o f equation 1 5 .2 0 can be
interpreted as a graphical sum o f (shifted) versions o f //~ '^ (/), as illustrated in Figure 15.20.
FIG URE 1 5 .2 0 y(r) equal to superposition of shifted replicas of
Equations 1 5 .1 8 and 1 5 .1 9 , as illustrated in the preceding example, are special cases o f more gen
eral formulas. Specifically, let
y =f * h- Then
yij-^ k) -J-iJ) - fj{k)
(15.21)
Chapter 15 Tim e Domain Circuit Response Com putations: riie Convolution Method
where j and
k are
integers and the n o t a t i o n m e a n s theyth integral o f/o v e r [ - 0 0 ,
the yth derivative if j

where j =
-k
EXA M PLE
> 0.
t] if j < 0,
and
O f c o u r s e , = /. An application o f this formula to the special case
with y = 2 is given in the following example.
1 5
. 1 1. Find the convolution
^t)
= :r
cos{m)u{t)
withy(f) =
r{t) - r{t-
2 ), as sketched
in Figure 15 .2 1 .
S o lution
Some preliminary thought suggests that evaluation o f the convolution integral might proceed
more smoothly via equation 1 5 .2 1 ; i.e.,
where
=
TC~
J0
cos(jTq)dq = JTsmiJTt)u(t)
and
f _n\n(jtq)dq = {\ - coîJtt)]u{[)
J0
Differentiatingy(r) twice leads to
Hence,
f*g =f^-'> *
But 1 1 - cos(M ^~ 2))]u (r -
2)
= [1 - cos(jrr)]K(r) - [1 - co s(7 r(r-
2))\ii{t -
2)
is just a right shift by one period o f [1- cos(;rr)](f). Therefore,

= [1 - c o s (;rr)] (/)//(2 -
T h is function is plotted in Figure 1 5 . 22 .
t)
Chapter 15 T im e Domain Circuic Response Com putations: T h e Convolution Method
Tim e (seconds)
FIG URE 15.22 Graph of the convolution
for Example 15.11.
Convolution algebra and graphical convolution lend themselves to a second application o f the
convolution technique: the computation o f circuit responses from a staircase approximation to a
circuit impulse response. If a circuit schematic is lost, such an approximation could result from a
C R T readout o f the circuit impulse response measured in a laboratory. The following example
illustrates this application.
EXA M PLE 15.12. Suppose the schematic diagram o f a very old linear circuit is lost. However,
the circuit impulse response is measured in the laboratory and approximated by the staircase wave
form o f Figure 15.23. If the input to the circuit is
at r = 0 and at r = 0.5 sec.
= 100(^), compute the output voltage,
h t)
1.0
0 .8
--
0 .6
0.4 -
0.2
0.2
0.4
0.6
fI h H I h
0.8
1.0
1.2
H I I I I h -l1.4
1.6
1.8
2.0
FIGURE 15.2.3 Staircase approximation to circuit impulse response.
'
S olution
The objective is to convolve the input with the impulse response. The technique o f convolution
algebra where one differentiates the impulse response to obtain a sum o f shifted impulse functions
and integrates the input to obtain a ramp function seems to be the most straightforward approach
for this calculation.
First, observe that
h{t) = 0.1
- 0.1) +
- 0.2) + 0.1u{t - 0.3) + 0.2(r - 0.4)
(15.22)
+ 0 .2 U - 0.5) + 0 .1 (r -0 .6 ) - 0 .1 ( r - 1) - Q>.2u{t- 1.3)

- 0.2//(r - 1.5) - 0.2(/ - 1.7) - 0.2(^ - 2) - 0.1 u{t - 2.2)
Hence,
h'{t) = 0 . l 6 ( r - 0.1) + 0 .2 6 ( r - 0.2) + 0 .2 6 ( r - 0.3) + 0 .2 6 ( f - 0.4)

+ 0 .2 6 U - 0 .5 ) + 0 . l 8 ( r - 0 . 6 ) - 0 . l 6 ( r - l ) - 0 . 2 6 ( r - 1.3)
- 0 .2 6 ( t - 1 .5 ) - 0 .2 5 ( f - 1 .7 ) - 0 .2 6 ( r - 2 ) - 0 .l 6 ( f - 2 .2 )
Now, since the integral o f the input is 100r(^), we can compute the output voltage as
= 1Or(? - 0.1) + 20r(r - 0.2) + 20r(r - 0.3) + 20r(r - 0.4)
+ 20K/^- 0.5) + 10r(^ - 0.6) - 1 0 r ( r - 1) - 2 0 ;-(r- 1.3)
- 20Kr - 1.5) - 20K r - 1.7) - 20r(^ - 2) - 1OKr - 2.2)
At f = 0,
= 0, and at / = 0.5, ôiJO.5) = 4 + 6 + 4 + 2 = 16.
O f course, it is possible to obtain the solution just as easily in this case using the graphical method.
Simply flip the
curve and slide it through the /?(f) curve. The area under the product curve
is simply the sum o f the rectangular areas, which are easy to compute.
Exercises. 1. Compute
for Example 15.12 at
1 and
1.5 sec.
A N SW ERS: 65, 106

2. As an alternative to the convolution algebra, find the convolution
j{t) = u{t) with h{t) given
in equation 15.22. Hint: Use the result o f Example 15.4 or equation 15.10, and the time shift the
orem o f convolution, equation 15.5.
ANSWF.R: Replace the step functions in equation 15.22 with ramp functions.
6. CIRCUIT RESPONSE COMPUTATIONS USING

CONVOLUTION
This section contains a series o f examples that illustrate the convolution approach to computing
zero-state circuit responses.
Chapter 15 T im e Dom ain Circuit Response Com putations: T h e Convolution Method
/87
EXA M PLE 15.13. Consider the RC circuit o f Figure 15.24, whose impulse response is h{t)
e~û{t). If the input is
find
when 7" = 0 and T > 0 for <7 = 0 and <7 > 0.
10
IF
FIG URE 15.24 /?C circuit for Example 15.13.

So
lu t io n
Case 1: a = 0. Here,
= u{-t) V. Since the capacitor voltage,
is the convolution o f the
input with the impulse response,
^'oia (7) = r
J 00
= f
J-00
h{T -
t)v ,
{T)dT = f
J oc
/Kt)v,(T - T)dr
e~îf(T)i/(T - T )dr = f e~û{T-T)dT
Jo
Consider the case o f 7'= 0. Then

^ W (0 )=
e~'î((T)dT = 1
This makes sense because with a = 0,

= u{-t) V is a 1 V constant, and the capacitor looks
like an open circuit at r = 0. On the other hand if r > 0,
- T)dx = e~^
^ W (^ ) = Jq
This also makes sense physically because at 7 = 0, the initial capacitor voltage is 1 V and the capac
itor discharges with a time constant o f 1 sec since there is no further nonzero input.
Case 2: a > ^. Similarly to case 1,

'W ( ^ ) =
- r)dT =
- T)dx
Consider the case o f T = 0. Then

.-0^^)rdT =
1+ a
This initial voltage depends on the history o f the excitation and can be computed only by convo
lution because the circuit is not in steady state at T = 0.
For 7 > 0, then, since the input is zero, we know that
e-'^
Vout^T) = Vc(0)e-û{T) = -------u{T)

\+ a
Exercises. 1. For case 1 in Example 15.13, find
for 7 > 0.
AN SW l-R: I
2. For case 2 in Example 15.13, find
for T > 0.
AN SWF. R: ^
1+i/
3. In Figure 15.24, suppose the resistor has value /? > 0 and the capacitor C > 0. Show that the
impulse response o f the circuit is
h{t) = ^ e
aC
4. For the general impulse response computed in Exercise 3, suppose
= e^'u{-t) V. Compute
1C
RC
ANSWF.RS; 1,
1
----- + a
RC
E X A M PLE 15 .1 4 . The goal o f this example is to design a circuit that computes the running aver
age o f a voltage
over the interval [/^- T, t] given a specification o f the necessary impulse
response, h{t).
S o l u t io n
A circuit that computes a running average must have the input-output relation
From our development o f convolution, such a circuit must have an impulse response h{t) satisfy
ing the relationship
Vnu,iO =
Now, h{t - t) must be a window function that captures the portion o f
Ts
^ r. Figure 15.25 depicts the proper forms o f
t)
and /;(?).
over the interval t -
Chapter 15 * T im e Domain Circuit Response Compurations: T h e Convolution Method
789
>^h(t)
h(t-i)
T
1
T
-
--t-T
FIG U RE 15.25 (a) The window function h{t - r). (b) The impulse response h{t).
The circuit design problem reduces to developing a circuit that integrates the function segment
V- {t) over t - T <, T
To achieve this integration, note that

v, { T ) d r =
V,-,, { T ) d T +
V,. { T ) d T
This relationship leads to

(15.23)
For the second integral on the right-hand side o f equation 15.23, let /I = r + 7 . Then
T
=dr,
= A + r , and
(15.24)
Since both X and x are dummy variables o f integration, we may replace A by ra n d rewrite equa
tion 15.23 as
(15.25)
where Vj^^{T - 7) is simply a delayed replica o f t^/(x) and where, for practical reasons, we can
replace the lower limit of-oo by /q, the time the actual circuit turns on.
As a convenience, we will define a device called an ideal delay o f T seconds, whose input is yy,/^)
and whose output is
T). Figure 15.26 shows the ideal delay as a device having infinite input
impedance combined with a dependent voltage source whose output is a delayed version o f the
input. Such a device can be achieved by storing the values of
in a digital computer or, for
small T, by the use o f an analog delay line.
F lG U llE 15.26 Depiction of a device called an ideal delay of

T seconds whose output is a delayed replica of the input.
790
Chapcer 15 T im e Domain Circuit Response Computations: T h e Convolution Method
All the pieces are now available; it is merely a matter o f putting them together. Integration can
occur using an inverting ideal op amp circuit having a capacitive feedback and resistive input. The
input to this ideal op amp integrator can then scale and sum the voltages
produce the desired running average,
and
T) to
by setting RC = T. This will guarantee the correct
scaling to achieve the desired average, since the transfer function o f the integrator will be M{RCs).
Figure 15.27 shows a circuit that will realize the desired running average.
FIG U RE 15.27 Op amp circuit that produces a running average of the input waveform,
provided that RC = T.
For the circuit o f Figure 15.27, the input drives the first inverting op amp circuit and also feeds
the ideal delay. T he second op amp circuit is an ideal inverting integrator whose inputs are
and
T). With RC = T the output is the required running average.
7. CONVOLUTION PROPERTIES REVISITED

From the perspective o f the impulse response theorem, the convolution properties o f commutativir)^ associativity, and distributivity have important implications in terms o f circuit and system
configurations. For example, if /?, {t) and hjit) are the impulse responses o f two systems, then com
mutativity says that
means that the order o f a cascade o f circuits or systems is
mathematically irrelevant, provided that there is no loading between the circuits. The idea is illus
trated theoretically in Figure 15.28.
Chapter 15 * T im c Domain Circuit Response Com putations; T h e Convolution Method
h,(t)
h,(t)
'9 1
(a)
h,(t)
h,(t)
(b)
FIGURE 15.28 Interchanging the order of impulse responses, in which the equivalence of parts
(a) and (b) follows from the commutativity of convolution.
Although mathematically parts (a) and (b) o f Figure 15.28 produce the same result, practical con
siderations often dictate a more careful realization process. For example, one circuit may have a
high input impedance w'hile the other has both a low input impedance and a low output imped
ance. In this case we would put the first high-input-impcdance circuit at the front end and the
other circuit at the output end.
A cascade op amp realization o f a transfer function,
H {s) =
100
(15.26)
(.y-Kl)(.9-H20)
illustrates commutativity nicely. A designer may use either Figure 15.28a or Figure 15.28b to real
ize H{s). Magnitude scaling, say by
= 10^, will yield more realistic resistor and capacitor val
ues. Recall that in magnitude scaling,
and
20
0.05 0
(a)
92
Chapter 15 I1 m c Domain Circuit Response Com putations: T h e Convolution Method
20
0.05 Q
(b)
FIG U RE 15.29 Realization of interchanging the order oFcascaded circuits.
Exercise. Verify that each o f the circuits o f Figure 15.29 realizes the transfer function o f equation
15.26. Magnitude-scale by
= lO' to obtain more realistic parameter values. If only 1 [.iF capac
itors were available, you would need two scale factors
for the first op amp stage and
for the second op amp stage. What are the rwo scale factors?
AN SW ERS: Multiply each resistor by I O' and di\'ide each capacitor by 10'' to obtain the new val
ues. II only 1 uF capacitors are available, then for Figure 15.29b,
For Figure 15.29a, the two scale factors are interchanged.
Theoretically speaking, the associative propert)', /;, * (A, *
= 10*^ and
= 0.5 x 10^.
= {h^ * h-y) * hy means that multi
ple cascades o f circuits or systems can be combined or realized in whatever order the designer
chooses. This assumes that there is no loading between the circuits or systems. Op amp circuits
called buffers or voltage followers having gains o f 1 can be used to isolate stages. On the other
hand, practical constraints may impose a condition on the realization o f a circuit for which the
mathematics o f the associative property does not account.
Finally, we consider the distributive property o f convolution: /;, * {hj +
= {h^
(/?j * h^. One interpretation o f this property is that the superposition o f the input signals
h-^) +
and
is valid. Flowever, Figure 15.30 presents rwo block diagrams with different interpretations.
Here one sees rwo possible topologies for realizing a system.
~93
y(t)
FIGURE 15.30 Two possible block diagram interpretations of the distributive law For convolution.
Exercise. In terms o f reliability, i.e., possibly continued partial operation in the face o f a circuit
failure, suggest reasons why one realization in Figure 15.30 would be superior to the other. In
terms o f minimum number o f components, suggest reasons why one realization would be better
than the other.
8. TIM E DOMAIN DERIVATION OF THE CONVOLUTION

INTEGRAL FOR LINEAR TIME-INVARIANT CIRCUITS
As mentioned, a deficiency in the one-sided Laplace transform technique is its inability to deal
with signals whose time dependence may extend back to -<. This section develops the zero-state
system response as the convolution o f a not necessarily one-sided input excitation with the
impulse response o f the circuit or system. Throughout the development, we will assume that the
circuit or system under consideration is linear, i.e., composed of linear circuit elements. This
implies that the zero-state response o f the circuit satisfies the conditions o f linearity; i.e., if the
zero-state response to the cxcitation fj{t) \s yj{t) for / = 1 ,2 , then the zero-state response to the
input excitation
addition, we assume that the circuit or sys
tem is time invariant, i.e., each circuit element is characterized by constant parameter values.
Mathematically, this means that
j{t) is the input to a circuit element and y(/) the zero-state

1 0 j{t - T) for all T and all possible
response o f the circuit element, then y{t - 7) is the response
794
input signals y(^). intuitively speaking, time invariance means that if we shift the input in time,
then the associated zero-state response is shifted in time by a like amount. These properties under
lie the development that follows.
Rectangular Approximations to Signals
Let us define a pulse o f width A and height 1/A as 6^(r). In particular.
4A
0 < /< A
(15.27)
othenvise
0
Mgure 15.31 sketches
for several A values.
u(t) - u(t-A) ,
, .
6A(t) = ---- 7---- for several As
1
A.
> t
1
FIG URE 15..31 6^(/) for several As.

Shifting the pulse o f equation 15.27 yields
as represented in Figure 15.32.
5^(t-kA)
kA
FIGURE 15.32
- M ) = b(^{t -
(k-Hl)A
is a shifted version of the pulse 6a(^)> where
= M.
For convenience, let

= M so that 6^ (r - M ) = 6/^(/ - tj^. Figure 15.33 shows a rectangular
approximation, v{t), to a continuous waveform v{t).
79 S
^ v (t)
FIGURE 15.33 Rectangular approximation, v{t), to a continuous signal v(t).

Expressing the rectangular approximation indicated in the figure analytically, using shifted ver
sions o f the pulse functions defined in equation 15.27, leads to the infinite summation
HD= 2
v(A-A)d^(r-A'A)A= ^
v{ti. )d^(t - tf^)A
k=-oa
k=-x
(15.28)
Hence, for sufficiently small A, it follows from equation 15.28 that

X
v{t)
V(/) =
v{tf, )6^ (/
tf, )A
(15.29)
k=-oo
One concludes that if v{t) is continuous, then
X
y
A-*{) A'=-00
^
- ti.)A= C
v'(/)= lim
J-oo
\iT)d{l - t ) ( I t
(15.30)
where we have interpreted the delta function as a limit o f short-duration pulses whose height is
inverse to the width so that the area is constant:
d(t)=
lim d ^ ( { )
AÔ
Observe that the right-hand integral o f equation 15.30 is precisely the convolution v{t) * S(r)
C om putation o f Response for Linear Tim e-Invariant Systems
Suppose
is the zero-state response o f a linear time-invariant circuit to the pulse
If the
circuits impulse response satisfies smoothness conditions, i.e., if it has sufilcient continuous deriv
atives, then the circuits impulse response is the limit o f the /ja U) values as A goes to zero. In par
ticular,
//(/)= lim//^(/)= lim f
A - * ()
/?(f -
T )d^ (T )c/r
= f
A - * ()JO
where the fact that

h i^ {t)= ^ ^ _ h {t -T )d f ^ {T )d T
h{t -T)d{T)dT,
(15.31)
follows from the convolution theorem. Equation 15.31 restates the law that the zero-state response
o f an input to a linear time-invariant circuit is the convolution o f the input with the impulse
response.
Now, by the assumption o f time invariance,
linear time-invariant circuit to
kA) is the zero-state response o f a well-behaved
- M ). Suppose further thatj'(f) is the zero-state response of
the same circuit to v{t). (See Figure 15.34.)
LinearTlme
8,(t)
Invariant Circuit
LinearTime
8.(t-kA)
!< = -CC
Invariant Circuit
(v(gA)8,(t-tJ
LinearTime
k = -cc
Invariant Circuit
LinearTime
v(t)
y(t)- / v(t) h(t-x) di
Invariant Circuit
/-oc
h
FlCirilF. 15.34 Zero-state responses to a particular linear time-invariant s)^stem, showing the framework
of the derivation. Note that the bottom condition follows because as A
0, A
di,
x, and H
j.
It is now possible to use the approximation for v{t) given in equation 15.31 to generate an approx
imation to^(^) in terms o f a summation o f terms o f the form v{kA)hi^{t - kA)A. Taking the limit
as A
0 will yield
as the convolution o f v{t) and h{t).
To derive this, note tliat for each k\ v{kA) = v{ti^ is a scalar. Hence, the zero-state response to
tfÂ is
tf^)A. By the linearity assumption, which implies superposition, the

zero-state circuit response to v{t), equation 15.29, is
(15.32)
k--x
In the limit, as A approaches zero,
approaches a continuous variable t and A - ch. Hence, if
the impulse response is sufficiently smooth i.e., if it has sufficient continuous derivatives then
v ( / ) = lim
v(A-A)/2 a ( / - A'a )A = r
v{T)h{t-T)ch
Thus, we conclude that for a linear time-invariant circuit, the zero-state response
(1 5.33 )
to an input
excitation v{t) is the convolution o f the input v{t) with the impulse response h{t). We will refer to
equation 15.33 as the impulse response theorem,* which says that the zero-state response o f a lin
ear time-'mvariant circuit or system to a (possibly tivo-sided) input signal is the convolution o f the input
with the impube response o f the circuit.
9 . SUMMARY
This chapter has introduced the concept o f the convolution o f two signals. The convolution can
be evaluated analytically (by direct computation o f the convolution integral) or graphically. Often,
by applying the properties o f convolution algebra, it is possible to implement shortcuts for calcu
lating the convolution o f two signals, resulting in the simplification o f the analytical calculation
or o f the graphical calculation.
Using the notion o f convolution, the chapter developed a technique o f computing circuit respons
es in the time domain. This technique is the direct counterpart o f computing the transfer func
tion in the frequency domain, the approach presented in Chapter 14. Using the convolution
approach, one can compute the zero-state response o f a circuit excited by signals that extend back
in time to -oo, something not directly possible with the one-sided Laplace transform. However,
for one-sided signals which constitute the great majority o f signals that are relevant to circuit
analysis the convolution and Laplace transform approaches are completely equivalent, as
demonstrated by the convolution theorem. The chapter gave an example o f designing a circuit to
compute a running average. In addition, it presented an application of the convolution technique
to the computation o f circuit responses for a circuit whose impulse response is approximated on
a CRT. Future courses will expand the seeds planted in this chapter. For example, convolution is
pertinent to an understanding o f radar techniques, commonly used to identify speeding motorists.

Algebra: a set o f functions with respect to two operations, + and *, satisfying the commutative,
associative, and distributive laws. In addition, there must be an identit)' with respect to
each operation. For addition, the zero function serves as the identity. For convolution,
the delta function plays this role. The delta function is an identit)' element because o f its
sifting property.
Associativity: for convolution, the property h * (/* ^ = {h * f } * g.
Com m utativity: for convolution, the property h ' f = f * h.
Convolution: an integration process between two functions to produce a third, new function in
accordance with equation 15.1 or 15.2.
Convolution algebra: the algebra o f functions with respect to the operations o f addition (+) and
convolution (*).
798
Convoladon theorem: for.one-sided waveforms
ancIy(^), L{h{j^ * J{t) ] = H{s)F{s).
Distributivity: for convolution, the property h * {/ ^ = h* f + h * g.

Graphical convolution (fiip-and-shift method): a pen-and-pencil technique for determining the
convolution integral o f simple, squarish waveforms.
Ideal delay o f T seconds: waveform with inputy(/) and o u t p u t 7), a delayed replica ofy(f).
Impulse response theorem: theorem stating that the zero-state response o f a linear time-invariant circuit or system to a (possibly two-sided) input signal is the convolution of the input
with the impulse response o f the circuit.
Linearity: property whereby, if the zero-state response to the excitation f^t) is
then the zero-state response to the input excitation
for / = 1 ,2 ,
Running average: average that is updated continuously.

Sifidng property: the property o f a delta function for simplifying an integral, i.e.,
4-00
f{T )=
- T )d t , provided t h a t / 7) is continuous at T.
*
Time invariance: property such that, i f / / ) is the input to a circuit element and^(^) is the zero-
state response to the circuit element, th e n ^ (/- 7) is the response
^__
7) for all T and
all possible input signals//).

Zero-state response: response o f a circuit or system to an input when all initial stored energy is
zero, i.e., all initial conditions are zero.
* T h e derivation o f this result is, o f course, not rigorous. A rigorous justification is given as theorem 4 o f Sandbergs
Linear Maps and Impulse Responses,
pp. 2 0 1 -2 0 6 .
IEEE Transactions, on Circuits and Systems, vol. 3 5 , no. 2, February 1988,
Problems
799
CONVOLUTION BY INTEGRAL
1. Leijit) = K^5{t - r , ) ,
plot in terms o f A', and
> 0. Compute and
- t
the results o f the fol

Figure P i5.5
lowing convolutions:
(a)
(b)
P )'u {t^ 2 T ^ )
J{t)'r{t^ T ^ )
6. L et/ f) = Ke~u{t)y ^ > 0. Use the convolu
(c)
r ,)- (r -4 r i)]
tion integral to compute the convolution in
(d)
/ f ) M K r + 2 r , ) - r ( ^ - 2 7 ', ) ]
part (a). Then use the results for (a) along with
(e)
7 ' , ) - ; < r - 2 7 ,)]
the various convolution properties, especially

the time shift property, to compute the remain
2. Let^(f) = 25(/ + 2) - 2 5 { f - 2). Compute and

plot the following convolutions:
J{t)'u {t)
(b) j{t)-r {t)
(c) J { t ) - [ u { t ) - u { t - A ) ]
id)
J{t)-[r{t)-r{t-2)]
(e) J { t ) - [ r { t ) - r { t - 2 ) - u { t - 4 ) ]
(a)
ing convolutions.
(a )
t({r) * f i t )
(b)
u { t ^ T) * J{ t)
[u{t^ T ) - u { t - T ) ] ' P )
[uir^T^)-u{r-T,)]*J{t-T^),T^
> 0, Tj > 0,
>0
uir^ l)*[J{t)-e-^'% -T)]
(c)
(d)
(e)
plot the following convolutions:
=Kê-îi{t), a > Q, and f^{t)

-K-,e~Ûiii), b > 0. Use the Laplace transform
(a)
J{t) - cos{m)u{t)
method to compute the following convolutions:
(b)
J{t)
3. Let//) = 25{t + 2) - 2 6 {t- 2). Compute and
- sin(;rr)(r)
7. Let f i t )
(a )
fir)* fir)
4. Compute and plot the results o f each o f the
(b) T iW / z W
Now use the results for parts (a) and (b) and the
following convolutions:
various convolution properties to compute the
(d)
u{t) * u(t - 2)
n { t - 2)
u(^)*[u(t)-;^(t-4)]
u(t^2)*lu(r)-u(r-2)]
(e)
[ ( f + 2 ) - ;v U ) ] * [ / / W - ( r - 2 ) ]
(a)
(b)
(c)
following convolutions:
(c) f i r ^ T ) * f i t - T ) , T > 0
(d)
(e)
fit^ T )*f^ [t-T )

fit+T^)-f^{r-T,),T^,T^>Q
5. Use the convolution integral to compute the
8. Compute and draw the convolution y{t) =

f t ) * v{t) o f Figure Pi 5.8 with v{t) = 5{t + 4) -
convolutions in parts (a) and (b). Then use the
5 (f+ 2 ).
results for (a) and (b) to compute the remain
f(t)
ing convolutions. (Parts (a) and (b) can be used

to solve many o f the subsequent problems also.)
(a)
(b)
(c)
(d)
(e)
(/) * Hr)
Hr) r(r)
u(r)*[r(r)-r(t-4)]
r (r )* [ r ( r ) - K r - 4 ) ]
M r + 2 )-/ / W ] * l u ( r ) - u ( r - 2 ) ]
(0
[K ^ + 2 )-K r)]*[K r)-K r-2 )]
(g)
/^(r) * fjir) for the functions in Figure

P15.5
>k
3 2
Figu re P i 5 .8
Chapter 1 5 * Time Domain Circuit Response Computations: The Convolution Method
800
9. (a)
Compute the convolution y{t) = f j ) *
(b)
Repeat the calculation in part (a)
v{t) for the functions in Figure PI 5.9

using the fact that r{i) = (/) * u{t) and
using graphical convolution for K = 2 .
the properties o f convolution.

(b)

using graphical convolution for 7 = 1
and/r= 1.
. v(t)
2K
K
(a)
H------h
1
3
(b)
2K
Figure P15.11
2T
12. (a) Compute the convolution j>(/) = j{t) *
>t
-T
v{t) o f the functions in Figure PI 5.12.
--K
r\
Hint: Use the results o f Problem 5 for
(b)
u{t) * r{t).
Figure P I5.9
(b)

using graphical convolution for
lO.(a)
Compute the convolution oiy{t) = fj)
v{t) in Figure PI 5 . 1 0 using the fact

,f(t)
that lit) = (/) * u{t) and the proper

1*1
1
/*
usmg graphical convolution ror 7 = 1.
i. v(t)
ties o f convolution.
(b )
=2
and 7 = 1.
2A
.f(t)
/N
2T
(a)
4T
(b)
2+
Figure P i 5.12
13.
> t
> t
-T
2T
2T
PI 5 .1 3 . Hint: Use the results o f
Figure P I 5.10
Problem 5.
(b)
1 l.(a)
Compute the convolution o^y{t) = fjj

* v{t) in Figure P I 5 . 11 using the fact
that r(/) = u{t) * (/) and the properties
o f convolution.
(a) Compute the convolution
J{t) * v{t) o f the functions in Figure

using graphical convolution for
A = 2 and T= I.
17. I f / r ) = Kû{-t), and h{t) =

0, a > 0, compute j'(^) = h{t) * j{t).
f (t )
a >
t 18. Compute the impulse response,
>t
-T
801
2T
h{t),
of the
circuit of Figure PI 5.18. Express h{t) in terms of
b = l/(/?Q . Supposeyj(/) = K^^û{t)y a > 0,
(a)
andy^(^) = K 2 e~^*u{t), a > 0 and a
Figure P I5.13
b. Compute
each o f the indicated convolutions using equa

14. Find the convolution of j{t) in Figure
P I 5.14 with
v{t) = d{t) + 5(f + 2) (Hint: Use graph
(a)
ical convolution.)
h{t) =
(b)
tion 15.2.
(a)
^j(r) = h(t) * u{-t), and sketch for

< r < 00.
(b)
yjit) = h{t) *
^<
+1) + 5(? - 1) (Hint: Use

(c)
graphical convolution.)
and sketch for - o o <
00.
y^{t) =
h{t)
*^ (/ ).
and sketch for - o o
<
r < 00.
/(^) = u{t + 3) - (? + 1)
(c)
-o o
(d)
(e)
y^{t) = h{t) *[/i(r) ^

and sketch
for -CO < t < 0 0 .
yÛ) = h{t) *[ u{-t) +^(^)], and sketch
for -oo < / < 00.
'" 6
Figure P I5.18
19. A particular active circuit has the transfer
function
Let h{t) = 4(/). Find^(/) = h{t) * fît)
15.(a)
H (s) =
s+a
w ith ^ (/) as shown in Figure PI 5 .15a.
(b)
Sketch y{t). Verify your answer using
where a> 0. Suppose the input to the circuit is

Repeat part (a) for /j ( /) in Figure
v{t), shown in Figure P I 5.1 9 with 7 > 0. Using
PI 5 .15b.
A
convolution techniques, compute the response
yit)-
f ,( t )
V (t)
>
>t
- t
-T
Figure PI 5.19
Figure P I 5.15
16. I f / / ) = Kêûi-t), ^ > 0, and h{i) = K2 u(t),
compute ^(/) = h(t)
20. Consider the RLC circuit in Figure PI 5.20.

(a) Compute the transfer fiinaion //(j),
and express H {s) =
s+ a
s+ b
802
the circuit is
Then compute the impulse response
response ^(r) using MATLAB as follows:

tstep = 1;
hit).
(b)
Compute y{t) = h{t) *
where
= v{t) is given in Figure P I 5.19.

(c)
If the input voltage is
V,
com pute y{t) = h{t) *
(d)
= ti{t) -u {t - 1) V, plot the
using the indicated element values.
vin = [1];
h = [0. 2, 3, 1, 1];
y = tstep*conv(vin, h)
y = [0 y 0];
Repeat (c) if the input voltage is
t = 0:tstep:tstep (length(vin)+length(h));
= 5e^ûH) V.
plot(t,y)
grid
/
o
+
2H
5Q
0.5F
h(t)
Vou,(t)
3 --
2-
Figure P I5.20
21. The transfer function o f a particular time-
> t
invariant linear circuit is
H{s) =
(a)
(b)
+
.? + 1
i- + 2
Figure P I5.22
A+ 4
Find the impulse response o f the cir
23. Repeat Problem 22, for the input
cuit.
u{t) + u { t - 2) - l u { t - A ) V. Hint: You need to
Find the transfer function o f the cir
change only the specification o f vin in the
cuit as a ratio o f polynomials by using
MATLAB code o f Problem 22.
the residue command o f MATLAB as

follows:
24. This problem formalizes the use o f the
(i)
MATLAB code above to compute the convolu
define the array p = [-1 , - 2 ,- 4 ] ;
(ii) define the array r = [2 ,-2 ,4 ];
tion o f two piecewise constant time functions.
(iii) define k = 0;
The steps below develop a formula for the con

=
volution o f two staircase time functions v{t)
residue(r,p,k) to obtain the coef
and h{t) (shown in Figures P I5 .2 4 a and b)
(iv) use
(c)
the
command
[n,d]
ficients o f the numerator and
using polynomial multiplication. In general, we
denominator polynomials.
assume h{t) and v{t) are zero for negative t and
Find the step response o f the circuit by

the convolution method.
(d)
Compute the zero-state response o f
have a finite nonzero duration.

(a)
Let y{t) = h[t) * v{t) and let the time

step T = \. Using Example 15.4 and
the circuit to the inputy(r) = 8 (-f) +
the convolution properties, argue that
?,u{t-T)y T > 0.
y{t) has a piecewise linear structure as

shown in Figure P i5 .2 4 c. Find the
breakpoint coordinates y i, y2 y3 and
GRAPHICAL CONVOLUTION
22. The impulse response o f a particular circuit
y4, first in terms o f the literal levels vO,

v l, ho, hi, and h2, and then in terms
is measured approximately on an oscilloscope,
o f their actual numerical values.
as illustrated in Figure P I 5.22. If the input to
(b) Define three polynomials in
a-
803
using the coefficients from Figures
h(t)
*<
P I5.24a, b, and c as follows:
5 -hO
p(x) = vO X + vl
q(x) = ho x + h 1 X + h2
h2
r(x) = yl x^ + y2 x^ + y3 x +y4
Show that rix) = p{x)(]ix) with yl through y4
assigned the values computed in part (a).
- 2 -1
-
Remark: The result o f step (b) indicates that the
(b)
coefficients [yl, y2, y3, y4] in K-v) or the break
y(t)
point values in y{t) o f Figure P I5 .2 4 c can be

obtained from the coefficients [vO vl] and [hO
hi
h2] by polynomial multiplication. The
M ATLAB command conv performs the

desired polynomial multiplication. In particu
lar, the code for obtaining the plot o f Figure
> t
P I5 .2 4 c is
V = [ 2 4];
h= [ 3 - 2 1];
T = 1;
tstep = T;
Figure P I5.24
y = [0 conv(v,h)*tstep 0];
% The additional beginning and
ending zeros are added to indicate
% that the initial and final values o f
the convolution are zero, due to the
% finite duration assumption,
t = 0: tstep : tstep* (length(v) +
length(h));
A N SW ERS: (a) yl = vO x hO = 6, y2 = vO x h 1
+ vl X ho = 8 , v3 = vO X h2 + vl x h 1 = - 6 . and
v4 = vl x h2 = 4
25. Repeat Problem 24 for the waveforms in
Figure PI 5.25. When using the MATLAB
code, it is necessary to account for/j(/) being
plot(t, y)
nonzero for negative t. This is easily accom
grid
plished simply by shifting/J(r) to the right by

one unit and then doing the indicated convo
lution; the proper result is obtained by a left
v(t)
shift o f one unit.
5 -vl
442
f,(t)
vO
(a)
-4..
(b)
(a)
Figure P i 5 .2 5
.so^
26. The schematic diagram o f a very old linear

circuit is lost. The impulse response, h{t), is
measured in the laboratory and approximated
by the staircase waveform o f Figure P I 5.26.
Based on the available information, solve cach
o f the following problems.
(a)
If the input is
output
= 100(r), find the

(b)
at r = 0, 0.5, 1, and 1.5
sec by the convolution method.

(b)
Figure P I5.27
If the input is 100f(/), find the out

put at r = 1 by the method o f your
(c)
choice.
CONVOLUTION ALGEBRA
Use the MATLAB code o f Problem 24
28. (a) Using the techniques o f convolution
to solve part (a). Hint: Although the
algebra, find j^(r) = /j(r) *
step function is not o f finite duration,
plot^ (r) vs. t, where/j(r) and^(^) are
after 2.2 seconds the output,^(r) = h{t)
as shown in Figures P i5 .2 8 a and b.
* v-^j{t)y does not change; also, the time

step 7'= 0.1 sec.
and
Hint: Graphically integratey^(r).

(b)
Repeat part (a) with/j(/) changed to

the waveform o f Figure Pi 5.28c for T
= 2 and /I = 2.
A
Figure P i5.26
27. A crude approximation to the impulse
response o f an active RLC circuit is given by
h{t)y as sketched in Figure Pi 5.27a. II the input
signal is v{t), as given in Figure P i5 .2 7 b , find
the response y{t) using MATLAB and the
> t
method developed in Problem 24. Plot the

response for 0 < ? < 6 sec. Verify using the tech
-4 . .
nique o f Example 15.4.
(b)
-T
2T
(c)
Figu re P i 5 .2 8
Chapter 15 * Time Domain Circuit Response Computations: The Convolution Method
29. Repeat Problem 28(a) w ith^(t) as given in
O '
Figure P I 5.29. Hint: Graphically int^rate^(/).
32. Consider Figure P I 5.32.

Find the transfer function H{s) and
(a)
the impulse response h(t) of the circuit
fj(t)
in Figure PI 5.32a.
(b)
Find the step response using convolu

tion algebra methods.
(c)
Find
due to the rectangular
pulse input in Figure P I5.32b . Hint:
> t
Again, use convolution algebra meth

ods, and sketch the output waveform
(or T = 2n^l(LC) .
-4 --
iYYV
Figure P i 5.29
30. Suppose the impulse response of a circuit is

given by
pulse
ôuM^~
= (P'e~^û{i) and the input is a
= u{t+ T) - u {t- T), T > 0 V. Find
(a)
using convolution algebra
(b)
Figure P I5.32
techniques.
31. Use convolution algebra techniques to
determine y / ) = h{t) * j{t) for each of the func
tio n s //) given in Figure P I 5.31 where h{i) =
7^ cos{m)u{t). Plot the resulting j^(/) using

MATLAB or its equivalent.
33. For the circuit o f Figure PI 5.32a, find the

convolution^/) = h{t) * J{t) for eachJ{t) given
in Figures 15.31b and c.
34. Consider the circuit presented in Figure
PI 5.34a.
At)
>t
(a)
(a)
Find the transfer function H{s).
(b)
Compute the impulse response h{t).
(c)
Use only convolution algebra tech
(d)
for the input

in Figure P I5.34b.
Plot your answer.
niques to compute the response
w
w
(a)
A(.)A
w
Vw^
Vw>
'O
> t
2
1
> t
- tse c
2n
4n
6n
(b)
Figure P I 5 .3 4
806
35. Compute the convolution of^(/) = h{t) *
37. This problem repeats the trick o f Problem
for each j { t ) below, given the h { t) o f Figure
36 of using the Laplace transform method to
j{t)
P 15.35.
(a) j{t) = o r cos{ojt)u{t)
(b)
(c)
compute responses when the input is nonzero
for - T < t < 0 when T > 0. Suppose a circuit
J{t) = (ur s\n{(Ot)u{t)

J{t) = aê-^'u{i)
has the transfer function
H(s)=
Vin
A
s +a
Suppose the input to the circuit is
shown
in Figure P I 5.37.
(a)
Define
(/) = Vj(t - T )and compute
(b)
Compute
due to the input
(c)
To compute
V/(.y) and then
CONVOLUTION BY INVERSE
LAPLACE TRANSFORM
).
due to
use
time invariance, i.e.,
36. This problem illustrates a trick for using the

Laplace transform
responses when the input is nonzero for - T

/ < 0 when
method for computing

q
<
> 0. Suppose a circuit has the
transfer function
>t
-T
V
Figure P I5.37
s+2
Vin
Suppose the input to the circuit is

in Figure P I 5.36.
(a)
Define
shown
38. The ideas presented in Problems 36 and 37

can be generalized as developed in this prob
(t) =
V/( t
- T )
and compute
lem. Starting with the definition of convolu

tion integral, equation 15.2, prove the time
(b)
Compute
) due to the input
(a) f i t ) *
Vjnis) and then

(c)
To compute
shift properties for convolution as follows:
due to
use
(b)
time invariance, i.e.,
Vout(^)=ôut(l + T')-
- T^) * }>(t)
/)^,'(0 = [ /
f=t+T,
- r , ) * g(r - r ,
39. Two active circuits with impulse responses
h^{t) = 2e~û{t) and h-,{t) =
v.(t)
are cas
caded as shown in Figure P I 5.39. No loading

occurs between stages. Compute the zero-state
response for
>t
-T
here?
Figure P I5.36
ANSWHK: (c)
= ti{t + 2) V. Would it be
advantageous to use the convolution method
= (I-
+T ) \
v.Jt)
SO
h,(t)
h,(t)
' 6
Figure P I 5.39
40. This problem shows the advantage o f the
Figure P I5.41
Laplace transform method over the convolu

tion method in the context o f a very common
42. Consider the circuit o f Figure P I 5.42,
situation. After solving this problem, try evalu
which has the transfer function
ating the convolution integrals, but dont spend
s
R,C
much time on this. Consider Figure P i 5.40,

which shows the cascade o f three circuits
with indicated impulse responses h^{t) =
2u{t), , h2 {t) = 2Qe~'^û{t), and

(a)
= 2tu{t).
S~ +
R\C\R2C2
Find the impulse response o f the

cascade, i.e., h{t) = h^{t) * h-y{t) *
(b)
H(s) =
Let C, = 0.5 mF, Cj = 1 niF, /?, = 2
(a)
Find the step response o f the cascade.
kQ, and /?2 = 1 kH. Find the poles and
Hint: You might want to use the residue com
zeros of H(s), a partial fraction expan
mand in MATLAB to compute the partial frac
sion of H{s), and the resulting impulse
tion expansions.
response, h(t).
(b)
v.(t)
h,(t)
h,(t)
h3(t)
Find
= 54trû{-t)
when
V. Then find the (zero-input) response
due to VoutÔ~).
(c)
Find the zero-state response when

= 54te~^'u{t) V. Would the time
Figure P I5.40
domain convolution method be desir

able for this calculation?
41. Consider the circuit o f Figure P 15.41.

(a)
Determine the value o f
from
(d)
Find the complete response, given

vour answers to parts (b) and (c).
= \Qe^û{-t) V. Can
the input
were doubled and
this be done using Liplace transform
72te~û{t) V, find
techniques? It so, how?

(b)
the
complete
response without doing any further
Find the transfer fijnction
calculations.
C,
H(s) =
ê
and the impulse response h{t) in terms
v(t)
o f R and C.
(c)
Suppose R = 10 n and C = 10 mF.
Given
(0") computed in part (a),

find the response due to the input
Fig u re P i 5 . 4 2
te~u{t) using the Laplace

transform
method.
Evaluate
the
advantages o f this method over the
MISCELLANEOUS
time domain convolution method.

43. Figure PI 5.43 shows a configuration for an
interconnection o f active circuits. The function
inside each box is the impulse response o f the

sub-circuit or subsystem. Suppose A,(r) = Au{t),
h-î) = 4 ( f - 2 ), and
(a)
(b)
47. Repeat Problem 4 5 for h^(t) = 4(/), ^2(^) =

56(^), hît) = 5 6 { t - 2 ) , and h^{t) =
= k cos{,m)u{t).
Using any of the convolution tech
4 8 . Figure P I 5.48 shows a configuration for an
niques you have learned, compute and
interconnection o f active circuits. The function
sketch the overall impulse response.
inside each box is the impulse response of the
circuit. Suppose h^(t) = 46{t), hjit) = 4 8 {t - 1),
niques you have learned, find y{t) for
hît) = 4d{t - 2 ). h^(t) = 45{t - 4), and h^{t) =
j{t) = 6 (r).
2 cos(;rf)tt(r).
h,(t)
(a)

niques you have learned, compute and
sketch the overall impulse response of
each configuration.
f(t)-
(b)

niques you have learned, find y{t) for
h,(t)
M = 6u{t).
Figure P I5.43
44. Repeat Problem 43 for ^2(^) = 4 6 {t-2 ) and

with the other impulse responses remaining the
same.
45. Figure P I 5.45 shows a configuration for an
interconnection of active circuits. The function
inside each box is the impulse response of the
sub-circuit or subsystem. Suppose h^{t) = 4(/),
hjit) =
h^{t) =
and //4(f) =
6 (r). Using any o f the convolution techniques

you have learned, compute and plot (using
MATLAB) the composite impulse response of
the configuration.
4 9 . Lety(f) be an arbitrary signal having a welldefined Laplace transform. Let

00
K t)= y
> y(t)
f(t)
6{t-kT)
itti
be a so-called impulse train. Find L
Note that if
z, a complex variable, then
your answer should be of the form

Figure P I5.45
46. Repeat Problem 45 with h^{t) changed to
h^{t) = 4[u{t) - u (t- 4)]. Having done Problem
4 5 , can you do this problem without any fur
ther calculations?
00
f(k T ) z -'
it=0
which is the so called Z-transform of the

sequence \J{kT) |^ = 0 ,1 ,2 ,...}.
809
C!liapccr 15 * Tim e Domain Circuit Response Com putations: T h e Convolution Method
50. The circuit o f Figure Pi 5.50 is initially at
51. Consider the circuit o f Figure P I 5.50
rest. The input
again. Let the input
v , (0 = ^ 6 { t + kT)
A=()
is a periodic impulse train with T = 1 sec.
(a)
be the impulse train
shown in Figure PI 5.51, i.e., analytically
cc
' ,)= ' ^ d U - k T )
k=0
Show that the impulse response is h{t)
with r = 1. Find
for r > 0 and plot the
resulting waveform.
(b)
Find the exact solution o f
for 0
<t< \.
(c)
Find the exact solution o f

< /< 2
for 1
(d)
Repeat part (c) for the interval 4 < r <

5. You should wind up with the
expression
VoJt) = 0.5^-0-5('-'^)(1 + ^-0-5 +

( ,.- 0 .5 ) 2 ^ (^ -0 .5 )3 ^ (^ -0 .5 )4 )
(e)
Simplif}' the expression computed in

part (d) by making use o f the sum for
mula for a geometric series:
//-I
=
1 -A
A -O
\.
provided
(f)
1 -A
Find the solution o f
for < / <
(;; + 1). K4ake use o f the techniques set

forth in parts (d) and (e) to simplify
your expression.
(g) Sketch the waveform for large >i.
This is the so-called steady-state solu
tion.
'"6
20
IF
Figure P i 5 .5 0
(1)
-4
(1)
-3
(1)
-2
(1)
-1
(1)
-> t
Figure P I5.51
AN SW ER: i>Jr) = 1.27t'*'-^^ for r > 0
Band-Pass Circuits and

Resonance
HOW A TOUCH-TONE PHONE SIGNALS THE NUMBERS
DIALED
Calling friends and others by phone occurs daily. When we dial a number, the information is sent
to the central office by one o f two methods: fast tone dialing or the much slower pulse dialing. For
example, electronic processing o f the pulse-dialed long-distance number 555-555-5555 requires
about 11 seconds, while electronic processing o f the tone-dialed takes only about 1 second.
The keypad o f an ordinary touch-tone phone has 12 buttons arranged in four rows and three
columns, as shown in the following diagram:
row 1
697 Hz
row 2
770 Hz
row 3
852 Hz
row 4 941 Hz
colum n 1
colum n 2
colum n 3
1209 Hz
1336 Hz
1477 Hz
keypad
Chapter 16 Band-Pass Circuits and Resonance
Pressing any one button generates two tones, with the frequencies selected by an electronic circuit
inside the telephone set. For example, pressing the number 5 generates tones at 770 and 1336 Hz.
The row and column arrangements and the dual-tone method permit the representation o f 10
digits (0 ,... , 9) and two symbols (*, #) using only seven tones. These seven tones are divided into
two groups: the low-frequency group, from 6 9 7 to 941 Hz, and the high-frequenc)' group, from
1209 to 1477 Hz.
"
Such tones are easily produced by an LC resonant circuit. The four tones in the low-frequency
group are produced by connecting a capacitor to four different taps o f a single coil (inductor). A
similar connection generates the three tones in the high-frequenq' group. When a button is
pressed to the halRvay point, a dc current from the central office is sent through the coil in the
tank circuit. When the button is fully pressed, the dc current is interrupted. This action initiates
sinusoidal oscillations in the LC resonant tank circuit at a frequency inversely proportional to
V L C . The presence o f small resistances causes the oscillations o f the tank circuit to die out.
However, pressing the button fully also connects the tank circuit to a transistor circuit that replenishes the lost energy and sustains the oscillations.
At the central office, the equipment used to detect the presence ol the tones and to identify their
frequencies is much more sophisticated. Two filters are required, one for each ol the frequency
groups. Each filter must pass the frequencies within 2% o f their nominal values (697 to 941 Hz
for one filter and 1209 to 1477 Hz for the other) and reject the signal if the frequencies are out-
side o f 3% limits. The output tone from each filter is then proce.ssed digitally to determine its
frequency.
The concepts and methods developed in this chapter will allow us to understand the properties o f
resonant circuits and the design o f various basic t}^pes o f band-pass circuits, as used in the touchtone telephone system.
CHAPTER OBJECTIVES
1.
2.
Describe and characterize the ideal band-pass filter.

Understand band-pass circuits from the viewpoint o f transfer functions and pole-zero
3.
plots.
Investigate the basic band-pass transfer function and its realization as a parallel or series
RLC circuit or op amp circuit.

4.
Investigate band-pass circuits having practical capacitors and inductors in contrast to
5.
ideal capacitors and inductors

Describe the phenomenon o f resonance and investigate the properties and applications
6.
o f resonant circuits.
Investigate general second-order transfer functions having a band-pass type o f frequency
response.
SECTION HEADINGS
1.
Introduction
2.
3.
The Ideal Band-Pass Filter

The Basic Band-Pass Transfer Function and Its Circuit Realizations
4.
5.
6.
Band-Pass Circuits with Practical Components

The Resonance Phenomenon and Resonant Circuits
General Structure of the Band-Pass Transfer Function with One Pair o f Complex
7.
8.
Poles
Summary
Terms and Concepts
9.
Problems
1. INTRODUCTION
How is it possible to listen to a favorite radio station by merely pushing a button or two or sim
ply turning a dial? Why do some very expensive receivers have very clear reception, while with
some cheaper models other stations chatter in the background? What circuitry inside the radio
makes this difference? The ability to clearly select a particular broadcast station depends on the
design o f an internal band-pass circuit. Such a circuit will pass signals within a narrow band o f fre
quencies while rejecting or significantly attenuating signals outside o f that band. To understand
why this is important, note that audio signals have significant frequency components up to about
3 kHz for voice and up to about 15 kHz for high-fidelity music. These frequencies are far too low
for wireless transmission. In (wireless) AM radio transmission, the audio signal modulates the
amplitude o f a carrier signal that is suitable for wireless transmission. The carrier signal is a highfrequency sinusoidal waveform bet\veen 500 kHz and 1650 kHz. The modulated waveform con
tains many frequency components centered about the carrier signal frequency, but extending over
a range o f frequencies equal to twice the highest audio frequency. For example, the radio station
WBAA, at Purdue University, has a carrier frequency o f 920 kHz and occupies a band from
approximately 915 to 925 kHz. To select WBAA from all the carrier signals received by a radio
requires a good band-pass filter to pass the frequency band o f 915 to 925 kHz while rejecting sig
nals outside this band. This chapter introduces the idea and properties o f a band-pass filter.
In its simplest form, a band-pass circuit consists o f only one capacitor, one inductor, and one resis
tor, connected either in series or in parallel. In the first half o f the text, we analyze simple RLC cir
cuits where we emphasize (1) transient behavior under dc excitation and (2) sinusoidal steady-state
behavior at a single frequency. This chapter investigates the behavior o f circuits over bands o f sinu
soidal frequencies. iMany useful results may be obtained with the phasor and impedance concepts
studied earlier. However, rapid advances in technology have made it possible to have a band-pass
circuit without any o f the usual RLC circuit components. Therefore, a study o f the band-pass
propert)' o f a transfer function H{s) dominates the material o f this chapter. The resulting analysis
is readily applicable to general linear systems, whether they be electrical, mechanical, or otherwise.
Although transfer functions underlie our approach, circuit realizations with ideal and practical
components illustrate all the basic concepts and properties.
S14
2. THE IDEAL BAND-PASS FILTER

This section sets forth the ideal band-pass characteristic and shows how a basic second-order trans
fer function can be used to approximate the ideal. Specifically, let H{s) be a voltage gain or some
other type o f network function. As we know, the curves for \H{j(a)\ vs. (O and /.//(/) vs. OJ are
called the magnitude (frequency) response and phase response, respectively. Figure 16.1a shows an
ideal band-pass (rectangular) magnitude response curve. Here, ideal means that all frequency
components o f the input signal within the range 0)j < to < (o^ are amplified with equal gain (in
magnitude), and all frequency components outside o f the range are totally eliminated from the
output. Actually, for a band-pass circuit to pass a signal with frequency components in the range
0)j < CO < 0J 2 without distortion, there is also a requirement on the phase response that is ordi
narily studied in a course on signal analysis.
(b )
S IS
(C)
FIG URE 16.1 Definitions o f peak frequency, co^, and bandwidth,
(a) Ideal band-pass character
istic. (b) Approximate band-pass characteristic of simple RLC circuit, (c) Band-pass characteristic of a
more complex circuit.
Unfortunately, an ideal rectangular band-pass characteristic is not realizable by a rational transfer
function or any circuit. One can, however, approximate the ideal characteristic with a simple
tuned circuit whose transfer function produces a bell-shaped magnitude response curve as illus
trated in Figure 1 6 .lb. With more complex circuits and more complex transfer functions, one can
improve the approximation as shown in Figure I6 .1 c. (How to do so is a topic to be studied in
advanced courses.) The bell-shaped curve o f Figure 16.1b has several important features. The fre
quency at which |//(/w)| reaches its maximum value,
denoted by
is called the peak frequency and is
The two side frequencies at which |//(/w)| is 1/V 2 o f its maximum value are
called the half-power frequencies, denoted tO| and 0^2 - The term half-power comes from the
fact that if the output is a voltage across a fixed resistance, then a drop in voltage by the factor
1/-v/2 means a drop in power equal to ( l/ yfl] = 0 .5 .

Half-power frequencies are also called 3 dB (down) frequencies. Recall from Chapter 14 that dB
gain is defined as
dB {gain) = 201ogjQ |Myw)|
If a gain o f is reduced by the factor \/j2 , then the dB reduction in gain is given by
- 3 d B = 2 0 log 10
H en ce the terminolog}^ 3 dB d o w n .
71
Chapter 16 Rand-Pass Circuits and Resonance
For obvious reasons, 0)| is called the lower half-power frequency and
frequency, and their difference
the upper half-power
= CO2 - co, is the half-power bandwidth (or simply bandwidth)
o f the band-pass circuit/transfer function. Band-pass circuits are designed so that all frequencies
o f interest fall within the pass band [(Op co^].
One way o f categorizing and comparing different band-pass circuits/transfer functions is by their
selectivity, i.e., a circuits relative capability to discriminate between frequencies inside the pass
band and signals outside the pass band. The selectivit)' is measured by the quality factor, Q, o f a
band-pass circuit/transfer function. The quality factor, Q, is the ratio o f the (geometric) center fre
quency {yj(JJiC0 2 ) to the bandwidth,
. For second-order circuits and transfer functions with a
bell-shaped magnitude response (Figure 16.1 b), the center frequency and peak frequency
cide, as we will show. In this case Q =
coin
where Q denotes the pole Q, to be defined in equation 16.1. For the circuit realization o f the
transfer function, Q is sometimes denoted by
or
narrow band o f frequencies relative to yJcOiOh or
A high-Q circuit passes only a very

whereas a low-Q circuit has a broad (pass)
band and a less selective characteristic.

It is important to note that the concepts o f
B^^, and Q o f a circuit are all based on the magnitude
function |//(/o)| and, therefore, on how the transfer function H{s) is defined. Even for the same circuit,
these values are different when the output is associated with different branches or when the input is
changed from a voltage source to a currcnt source. Further, for the investigation o f the frequency-selec
tive characteristic of the circuit, the foregoing definition o f Q is most appropriate because it directly
assesses the sharpness o f the magnitude response cur\'e. As such the definition allows Q to be experi
mentally determined. In certain other applications, where only one fixed frequency is o f interest, there
is another definition o f a circuits Q based on an energ)' relationship that is inadequate for general band
pass circuit design.
3 . THE BASIC BAND-PASS TRANSFER FUNCTION AND ITS

CIRCUIT REALIZATIONS
The most basic second-order band-pass transfer function having the bell-shaped curve o f Figure
16. lb has a pair o f complex poles and a single zero at the origin, i.e.,
(.V - p\){s- P2)
(s + G p - j w j ){s + Op + jco^j)
(16.1)
+ 2apS + a l + m ; ,
r + 2ct,,.v + (o^
Qp'
"
Figure 16.2 illustrates the pole-zero plot o f the transfer function o f equation 16.1 as well as the
relationships among the various parameters. Note that there is a finite zero at the origin. Further,
because the numerator has degree strictly less than the denominator, there is also a zero at ^ = co.
In both cases the magnitude o f the transfer fijnction is zero.
SI
Chapter 16 Band-Pass Circuits and Rcsonancc
JO)
FIGIJRK 16.2 Pole-zero plot of a band-pass H{s) with a single zero at origin as per equation 16.1.
Our next goal is to derive the peak frequency
the half-power frequencies tOj and t02,
the maximum gain
=|
the circuit Q, and the frequency at which the angle of
/y(/to) is zero. The first step in this derivation is to write down
K
(16.2a)
2 o + 7 ----------- (JO
whose peak magnitude response is
m ax
(I)
1^1
H(Jco)
I CO., - 0 )
l\~\
I
4a+
mm
0)
(0
)
9 \(OZ, - U>~ \
\K
2a,
(16.2b)
CO
Note that since the numerator is a constant, the maximum occurs when the denominator is a min
imum, i.e., when co = (O^. Therefore we conclude that
(16.3)
COI,
Noting that Qp ~
, we further conclude
H(jco)
(16.4)
To find the half-power frequencies lOj and co-,, the maximum gain must be reduced by the factor
l/V 2 . Considering equation l6 .2 b , this occurs at those cos for which
818
(col-co-)
CO
(16.5)
= 4a
This fourth-degree equation in to reduces to two quadratic equations defined by

cOp - c o = 2opO) => co^ + 2o^,(o -co^, = 0
(16.6)
We denote the two positive solutions by cOj and a>2. with CO2 > o jj. Solving equation 16.6 using
the quadratic formula, we have
yJcTp+OJ-p = CO.
(
(16.7)
To compute the bandwidth, we have

CO,
B,^=co2-C0^=20p=-^.
Exercise. Show that for high-Q^^ transfer functions/circuits (say
(16.8)
> 8),
(16.9)
We can deduce one more important property from the above equations. When equation 16.5
holds, the real and imaginary parts o f the denominator in equation 16.2a are equal. Hence the
denominator has an angle o f 45^^. Thus the angles o f
H(jcOi) = ----2CT. + 7
(cor-col\
CO:
(/ = 1, 2) are 45^" for K> 0, and 4 5 - 180 (or K < 0.
Exercises. 1. Given the transfer function H{s) =
--------- , find
5 ^ + 2 ^ -h 2 5 6
A N SW I'RS:
2. Suppose that
= 16 r.id/sec, //,,, = 0.5,
, (0,^ ,
, and a>i 2 .
= 2, o), = 15.031. and (o^ = 17.031
= 8 and 10^ = 1000 rad/sec. Find the exact to^ ^ from equation 16.7, the
from equation 16.8, and the approximate Wj 2 from equation 16.9. Now compute the magnitude
o f the percent errors between the approximate and exact values o f tOj 2A N SW ERS: in random order; 125. 939.^^. 1064.5, 1062.5. 937.5. 6 .1833, 0 .2077
3. Given equation 16.9, show that for a high-(^ transfer function/circuit (say
to^ s 0.5(cOi + W2)
> 8),
819
i.e., for the high-Q case the geometric and arithmetic means of 0)j and 0J 2 are approximately equal.
The next two questions are: (i) what is the geometric center frequency o f the magnitude response
and (ii) is it equal to to ? Here, using the square o f the geometric center frequency,
CO1 CO2 = -O p + J a j - c o l
Op +
p+<4
= - Gp + Op + ( O p = ( O p
we conclude that for the above transfer function and associated circuit realizations, the geometric
center frequency, the pole frequency, and the peak frequency coincide, i.e..
Îoj\co2 = COp = CO.

As mentioned earlier, the circuit/transfer function Q is the ratio o f the center frequency to the
bandwidth. Given equations 16.8 and 16.10, we have
Q = Q ..- Q ,, = ^
(16.11)
Finally, the nonzero frequency for zero phase shift occurs when the transfer hmction is purely real,
i.e, when the imaginary part is zero. This occurs at CO = CO^; hence
(16.12)
zero phase-shift frequency = co
E XA M PLE 16.1. Compute
= |Myto^^^)|, Q,
O),, (1)2 (exact and approximate
values) for the simple parallel RLC circuit shown in Figure 16.3a where the current source is the
input and the voltage across the input nodes is the output. Then compute and label the magni
tude response and verify that it has the bell shape o f Figure 1 6 .lb.
S20
.(b )
FICiURE 16.3 (a) A parallel /?/,C circuit, (b) Magnitude response curve.
So
lu t io n
The key to the solution o f the example is the computation o f the transfer function as follows:
1
Y(s)
^
s
C
C5 + +
Li
.V + - .y-l-
RC
LC
Then, according to equation 16.1,
H{s) = K
=K
we conclude that Op = ^
, cOp = j- , and
2 R C " " ' - 1 W
iOn
RC
Q = Q, = ^
= 4^ = R
c
L
(16.13a)
Thus from equations 16.3 and 16.4,

(16.13b)
and
2a^
( 1 6 .1 3 c )
Chapter 16 Bund-Pass Circuirs and Resonance
i '- 1
= R.
Note that this impHes that

To compute the bandwidth
equation 16.8 implies that
li, = 2 0 p = - ^
(16.14)
Similarly from equations 16.7 and 16.9, the exact half-power frequencies are
tO\ H---------- 1- 4
IR C V \1RC}
LC
(16.15)
while the approximate half-power frequencies for the high-(^ case are
f/Ji s p = ----------,
VZC
IRC
W o s j^=H --------- .
ylLC
(16.16)
IR C
Finally, to obtain the frequency response we note the values o f the above computations, and the
fact that at
CO
= 0 and
co
= oo the magnitude o f the frequency response is zero, and the bell-shaped
curve o f Figure 16.3b results.
Exercises. 1. Compute the exact values o f
tOp O),,
and Q for the circuit o f Example 16.1
when R - 2.5
Z, = 0.1 H, and C = 0.1 uF.
AN SW ERS: (0,,, = 10"^ rad/sec, o , = 12.198 x lO 'l to. = 8.198 x 10\
= 4 x \o\ and Q = 2.S
The preceding example and exercise demonstrate that for high-Q circuits (Q > 8), there is really
no need to use the exact equation 16.15 to compute cOj and co-,, as the much simpler estimates
given by equation 16.16 are sufficiently close to the true answers. Indeed, sometimes it is con
venient to use the approximate formulas when Q > 6.
In many practical circuits, the independent source could be a voltage source in series with a resis
tor. Before applying any o f the foregoing formulas, it is necessary to transform the circuit into the
form o f Figure l6.3a by the use o f the Norton equivalent circuit studied in a first course on cir
cuits. The resistance R in Figure l6.3a then is not a physical resistor, but rather, the equivalent
resistance of several resistances in parallel. The following example illustrates this reformulation.
EXA M PLE 16.2. In the circuit o f Figure 16.4a, an independent voltage source
in series with
an internal resistance R^ = 40 kQ models a real-world sinusoidal excitation. Suppose L = 20 mH,

C = 0.05 l i F, and Rj = \0 kQ.
(a)
Find the exact values o f co^^^, B^^, and Q.
(b)
Estimate the values o f OOj and co-,.
(c)
Find I
S22
+
V
(a)
FIGURE 16.4 (a) Tuned circuit driven by a practical voltage source.

(b) Equivalent circuit showing Norton equivalent for the source.
So
lu t io n
The solution proceeds by using the formulas developed in Example 16.1 after replacing the prac
source by its Norton equivalent and identifying R in Example 16.1 as the parallel com
tical
bination o f
and /?^. By the usual formula for two parallel resistors,
R,R,
4 0 .0 0 0 x 1 0 ,0 0 0
_
R=
=
:------ = 8 ,0 0 0 Q
Rs + R l
4 0 ,0 0 0 + 10,000
This results in the transfer function
s
C
'm
(a)
R,C
\
RC
LC
s " + -------s ' - + -------------------------------------- S +
RC
Notice the new value: K =
LC
R,C
From equations 16.13b, 16.14, and 1 6 .13c,
W,=(Op =
^
= 3 1 ,6 2 2 .7 7 rad/s
V 0 .0 2 x 0 .0 5 x 1 0 " ^
1
Bco =
RC
8,000x0.05x10
-6
= 2,500 rad/s
823
and
6
. ^
3 1 .6 2 2 .7 7 ^ , , ^
Z50 0
(b)
Using the approximate equations for high-Q circuits, we obtain the half-power frequen
cies as
^
= 31,622.77 - 1,250 = 30,372.77 rad/s
+^
= 31,622.77 + 1,250 = 32,872.77 rad/s
and
"2
(c)
Now we use equation 16.4 to obtain

a
.
2cr^
R,CB^
,
/?,
Notice that we did not use equation 1 6 .13c because the input in Example 16.1 is a current source
and not the voltage source o f Figure l6.4a. Hence the maximum value is not equal to R.
This example demonstrates that putting an external resistance in parallel with the LC tank circuit
reduces the value o f R, which in turn causes a larger bandwidth and thus a lower circuit Q while
keeping the peak frequency
unaffected.
Exercise. Repeat Example 16.2 with the element values changed to R^ = 36 kQ, L = 40 mH, C =
0.25 pF, and R^ = 4 kH.
A N SW ERS: 10,000 rad/sec, 1111.11 rad/sec, 9, 9 4 44.44 rad/sec, 10,555.55 rad/sec.
The examples so far have illustrated only the analysis o f parallel IU.C circuits. In the design of a par
allel-tuned circuit, we must also pay attention to other factors, such as available component sizes,
desired voltage gain, and cost. In practice, design specifications ordinarily impose a small number of
constraints relative to the number o f circuit parameters to be determined. Consequendy, realistic
design problems usually do not have a unique answer, as illustrated in the next example.
E X A M PLE 16.3.
Design a parallel RLC circuit, as shown in Figure l6 .4 a , to have a magnitude response with^^^ =
200 kHz and a bandwidth o f 20 kHz. Only inductors in the range 1 to 5 mH are available. The
source has a resistance R^ = 50 kH.
So
lu t io n
For the circuit o f Figure l6 .4 a there is a restriction on the available inductors. Hence, we keep L
as a variable, subject to the condition that 0.001 < L < 0.005 H. Using the specified peak fre-
.S2-
quency and rhc fact that
LC
= 6 .3 3 x 1 0 "'^ L
C = - r r = - -r
From the specified bandwidth and equation 16.14,
= L^2li^ liC ^ =l.2 5 7
= ^
x l0 U
2 ;r x 2 x lO '^
As explained in Example 16.2, R is the parallel combination ol
or
1 -_L
_L
R ~ Ri_
R, '
R[
R,
and R^, i.e.,
Once a specific value o f L is chosen, we can calculate successively the values of C, R, and Rj. Since
R, which is the parallel combination ot R^ and R^, must be no greater than
= 5 x lO'^ f2, the
upper limit for L is
5 0 .0 0 0
R.
^ma\ ~
1 .2 5 7 x 1 0 '
= 0 .0 0 3 9 8 H .
1.257 x 1 0 '
Numerical values corresponding to the extreme values ot L are given in Table 16.1.
TABLE 16.1
L (mH)
C (p F )
R(VQ)
Rl i ^ )
633
12.57
16.77
3.98
159
50
infinite
Table 16.1 clearly shows that there is no unique answer to the design problem. 'Fhe freedom in
choosing a value for L in the range 1 to 3.98 mH can be utilized to accommodate another design
specification, such as a value for
Exercises. 1. In Example 16.3, find the maximum and minimum values o f |//(/to^^^)|.
AN SW ER: 0.251, 1.0
2. In Example 16.3, if the bandwidth requirement is changed to 10 kHz, determine the minimum
and maximum possible values o f L.
AN SW ER: 1.0 mH and 1.99 mH
H2S
Dual to the parallel /?/.Cof Figure 16.3a is the series y?Z.Cof Figure 16.5, which has a voltage input
and a current output. Although we can use clualit)', to infer frequency response behavior, we pre
fer the direct transfer function approach.
E XA M PLE 16.4. For the series RLC circuit o f Figure 16.5, let 1^ be the desired output. Find
(a)
The transfer function H (>y) =
(b)
The peak frequency
(c)
The bandwidth,
(d)
The circuit Q
(e)
The half-power frequencies, to, and (o-, (exact and approximate values)
(0
(g)
W, = |M;co)|,,= |//(/0|
The frequency at which the angle o f H{jio) is zero
R
V
FIGURK 16.5 Series RLC circuit having a band-pass transfer function.

SO L U T IO N
(a) Find the t r a n s f e r By inspection,
z , ( .)
(16.17)
-)
.y + .v-l- -
C ,.'
= K S~ + l O n S + Clj:
=K
5 + ^S + W~
Q
Thus, A-= ML,. 2a^ = //.,, co^ =
and
'
= /,
(b )-(f) Inspecting the transfer function o f equation 16.17, we immediately have

Peak frequency: (o, = Wp =
C ircu it Q:
=Qp =
lc
(16.18)
^p^S
( 1 6 .1 9 )
Cs
826
Half-power bandwidth:
(16.20)
Half-power frequencies (see equation 16.7):
(0 ,2 = 0 3 ^ ,
iV
4Q -
2Qpj
2L,
where the approximation for the high-Q^ case is
-B co
C0\ ') =' (0, H-------- ,
i -------
^ l^s
2L,
/, /'
0 6 -2 2 )
From equation 16.4,

_ |/f| _ \K\Q,, _
2<J,,
(o,
1
(16.23)
R,
Finally, from equation 16.12,

zero phase-shift frequency = iOp
Exercise. For Example 16.4, suppose

nitude. Find o)^^,
AN SW ER:
= 5 O.,
= 1 mH,
= 0.1 |.iF, and
has a fixed mag
and the approximate half-power frequencies.
= 10 rad/scc.
= SOOO rad/sec, Q.-^ = 20. 97,500 rad/sec, and 102,500 rad/sec.
The above derivations and calculations yield formulas and numbers. To add some meaning to the
concept o f circuit Q and its relationship to bandwidth, we provide a plot o f the normalized |//(yco)|
vs. (0 curve for different values o f
in Figure 16.6. The ordinate is the ratio
|//(yco)|/|//(yo))|^^^^^, while the abscissa shows the ratio to / to T h e se curves are called universal res
onance curves because they are applicable to parallel RLC circuits, to series RLC circuits, or to any
system having a transfer function o f the form o f equation 16.1. Observe that as Q increases, the
bandwidth decreases, indicating a better selectivit)^
82'
(O/Wm
FIGURE 16.6 Normalized magnitude response for equation 16.1.
To conclude this section, we present an example o f an active band-pass circuit that avoids the use
o f inductors. The band-pass circuit illustrated here is only one o f more than a dozen configura
tions in use. This example illustrates the possibility o f eliminating inductances while producing
the same kind o f frequency response as the parallel or series RLC circuit. You can learn a lot more
about these active filters in a more advanced course.
E X A M P L E 16.5. The operational amplifier in Figure 16.7 is assumed to be ideal.

1.
2.

A band-pass circuit is to have peak frequency
= 1000 rad/sec and a bandwidth
100 rad/sec. Find Q for the desired transfer function. Then realize (find values for /?p Rj,
Cp and C-y) for the normalized transfer function
H.
Ks
,{s) =
S
5 +
,
1
3.
under the condition that Cj = C2.

Magnitude- and frequency-scale the circuit to achieve the correct center frequency with
4.
q = C2 = 1 1-iF
Verify the frequency response with a SPIC E simulation o f the real circuit.
82.S
MGURE 16.
S o lu t io n 1.
An active band-pass circuit without inductance.
Find the circuit transfer function
From the assumption that the operational amplifier is ideal, we have K_ = 0 (virtual ground) and
/_ = 0 (infinite input impedance). Applying KCL to the inverting input node V_, we obtain
'C , K , + - ^ = o ,
A-7
which yields
v;. = -
Next, we apply KCL to node
to obtain
V
^'ci - V
* -in
/e,
+ iC | V + iC 2 (K -V ,)= ()
Substituting the previous expression for
ôut
ou t
into this equation, regrouping terms, and solving for
in the desired transfer function:
-s
~RxC^
+
(16.24)
1
,/?2C,
R iC iV
I^\Ri C\C2
This is precisely the form o f equation 16.1.
l.Computation o f normalized transfer fimction to be realized. The desired transfer function is

Ks
H{s) =
V'/(^V)
+ Ba) S + 1 0 ;
Ks
2
i-" +
Ks
S + iO
.v^ + 100 .v + (1000)
We now frequency-scale s -*
= 1000 rad/sec. Thus uhe normalized transfer function
with
is
_
=
X ------- r r ^ --------= TTrrr x

lei
?>.Computation o f element valuesfor normalized transferfiinction. Equating the circuit transfer func
tion with the normalized transfer function, we have
--------------- = 1 rad /sec
(16.25)
R\
and
H---------- --- 0.1 rad /s e c .
(16.26)
Under the condition that

C, = C, = 1 F
we have, from equation 16.26,
From equation 16.25 we obtain
Ri R2C\C2
R[20
Note also that
= I H U co, ) I =
= ^
=^
= 200
RjC^ ^ RoSi
4. Frequency and magnitude scaling. As per the problem statement, we desire
Hence we frequency-scale with Kj- = 1000. It is further required that
= 1000 rad/sec.
= 1 liF. Hence,
we solve the following for K^^-.

C. = f f -
KfjjKj
K, =
10
X 10
= 1000
It follows that
c , = 1
50n,
iOOOR2M- 2 0 n
The ratio o f /?, to R^ is very large and unrealistic. It turns out that this circuit is best suited to lowQ transfer functions.
5. Verif' frequency response. To verify the frequency response for a realistic implementation o f the
above circuit, we consider a SPIC E simulation using the standard 740 operational amplifier as
shown in Figure 16.8. Observe that the maximum value o f the magnitude response is 200, as
expected, and that the bandwidth is about 16 Hz, which translates to about 100 rad/sec.
830
Exl 6.5-Small Signal AC-3

+50.000
+100.000
+150.000
+200.000
Frequency (Hz)
+250.000
+300.000
FIG URE 16.8 Magnitude response plot resulting from a

SPICE simulation o f the 740 operational amplifier.
Exercises. 1. In Example 16.5, find the circuit Q and the approximate half-power frequencies
CO
1,2AN SW ER. 10,950 rad/sec and 1050 rad/sec

2. In Example 16.5, if Cj = 0.5 uF and C j = 1 f.iF, find the values o f /?, and Rj required to meet
the specifications on
AN SW ER: A, = 66.67
and
R,= 30 kU
Throughout this section we have assumed the use o f ideal inductors and capacitors. Practical
inductors have complex circuit models to account for real-world behavior. The next section takes
up an approximate analysis o f circuits containing simple models o f practical inductors and capac
itors.
4. BAND-PASS CIRCUITS WITH PRACTICAL COMPONENTS

Quality Factor o fL and C Components
How can we analyze band-pass circuits in the presence o f practical (non-ideal) inductors and
capacitors? Practical inductors and capacitors have models consisting o f their ideal cousins and
other ideal (parasitic) elements to account tor losses and coupling effects. Figure 16. 9a illustrates
8 31
a simple model o f a practical inductor for low to medium frequencies while Figure l6 .9 b is a rea
sonable model for high frequencies.
FIGURE 16.9 Two models of an inductor, (a) For low to medium frequencies, (b) For high frequencies.
The primary parameter is, o f course, the inductance L. The remaining elements, R^,
and C^,
account for undesirable yet unavoidable practical effects and are called parasitic. Since an inductor
usually consists o f a coil o f wire, R^ represents the wires resistance. Also, a capacitance is present
between adjacent turns o f wire. Hence
models this parasitic capacitance. The resistance R
accounts for the energy loss in the magnetic core material (if present) inside the coil. Complex mod
els such as Figure 16.9b, although important, if used for every inductor would unduly complicate
the analysis o f a band-pass circuit. Fortunately, for low to medium frequencies (up to a few mega
hertz), the simpler model o f Figure l6.9a suffices and hence underlies the material that follows.
Figure 16.10 shows two models o f a practical capacitor. Again, the primary parameter here is the capac
itance C; R , Rp and
are parasitics. The leakage resistance, R^, accounts for the energy loss in the
dielectric; tne inductance

At frequencies above \j
and resistance R^ are due mainly to the connecting wires o f the capacitor.
, the capacitor actually behaves as an inductor! For frequencies of up to
a few megahertz, the simpler model o f Figure 16.10a suffices, and it is used for the analyses o f this text.
(a)
FIGURE 16.10 Two models of a practical capacitor, (a) For low to medium frequencies.
(b) For high frequencies.
832
How close is a pracrical inductor (Figure l6.9a) or a practical capacitor (Figure 1 6 .10a) to the
ideal? The so-called element quality factor provides a quantitative measure. To develop the qualit)' factor, consider that each practical inductor or capacitor has an impedance
Z(;(o) = RejZ(>))} + j Im {Z(>))} = /^(oj) + >V{(o)
For both the practical inductor and capacitor, the frequenc)-dependent reactance X is the primary
parameter of concern, whereas R represents the unavoidable parasitic effect. In the ideal case, R
is not present. Hence, the ratio
provides a measure of how close the network model is to an ideal inductor or capacitor. The larg
er the ratio, the better the element behaves as an ideal inductor or capacitor, (ideally, R = 0 and
the ratio is infinite.) This suggests defining the quality factor associated with a practical inductor
or capacitor having impedance Z(yco) as
,
UCw)!
The inclusion o f co in equation 16.27 is to emphasize the fact that Qy depends on the frequency'
o f operation. The subscript Z indicates a generic impedance and may be replaced by more specif
ic descriptors such as coil or capacitor. Unlike the Q o f a circuit, which depends on the values
of the elements of the circuit and on the circuits configuration, the quality factor Qj/oi a practi
cal inductor or capacitor varies with the operating frequenc)' OJ and remains unchanged irrespec
tive o f its connection in the circuit. Any element with a finite ratio given in equation 16.27 is
termed a lossy component, as are all real-world components.
For the practical inductor model o f Figure l6.9a, equation 16.27 reduces to
R^
^
where we may synonymously denote
as
(16.28)
As mentioned, higher
implies a better-qual-
it)' coil in the sense that the energ)' loss in the component is smaller. Infinite
inductance L, which is lossless. In the audio frequency^ range,
represents a pure
may vary from 5 to 20, where
as in the radio frequency range, it may exceed 100 in practical applications. Although R^ here
varies with frequency (due to the skin effect), at the level o f this text, we treat R^ as a constant inde
pendent o f (0 .
Similarly, the capacitor model o f Figure 1 6 .10a has impedance
I
Z ( = - j -----------=
Hence, the factor Q^ or
"
' , (l - JioK c ) m,o)+JX(cu)
calculated in accordance with equation 16.27 is
=S =
(6.29)
S33
A higher
implies a better-qualiry capacitor, in the sense that the energ)' loss in the device is
smaller and closer to ideal. Infinite Q^-represents an idealized, lossless capacitor modeled by a pure
capacitance. In practice, Q^;;is usually much greater than Q^, meaning that
is more critical for
circuit performance, i.e., Q(;;is often assumed to be infinite. The reciprocal o f Q^^ is called the dis
sipation factor o f the capacitor and is denoted by cIq A lower dissipation factor means a betterquality capacitor.
The determination o f
and
requires specification o f the operating frequency co. If the value
o f 0) is unspecified, the analysis o f a band-pass (or tuned) circuit proceeds under the assumption
that
= Q^(cOq) and Q ^= QôOq), where C0q = M \[l C . We will discuss the meaning o f oJq in
section 5, on resonance.
Reductioit o f Band-Pass Circuits to Approximate Series or Parallel RLC Circuits

When one uses the practical inductor and capacitor models o f Figures 16.9a and 16.10a, the series
RLC and parallel /?ZC band-pass circuits with practical sources have the more complex configura
tions o f Figures 16.1 la and b, respectively. Since they are no longer series or parallel RLC, their
transfer function is not the ideal band-pass type of equation 16.1. Hence the associated formulas
for peak frequency, bandwidth, etc. are not directly applicable.
FIG URE 16.11 (a) Model of a parallel tuned circuit using practical inductor and capacitor models,
(b) Model of a series tuned circuit using practical inductor and capacitor models.
834
Topologically speaking, the circuits o f Figure 16.11 are series-parallel because the input imped
ance seen by the source consists o f a sequence o f series connections and parallel connections o f
simple networks. Exact analysis o f such series-parallel band-pass circuits to obtain
and
is
cumbersome, especially in light o f a simpler, more efficient method widely used by engineers to
compute approximate solutions. The approximate analysis relies on the conversion between a
series circuit and an equivalent parallel circuit at a particular frequency. This conversion process
depends on the component quality factors developed in the previous subsection. See Problem 85
for the development of this equivalence. The next example illustrates the conversion process for
an inductor.
EXA M PLE 16.6. When the ideal components o f a parallel RLC circuit are modeled with a prac
tical inductor and a practical capacitor, the circuit is no longer parallel. However, by converting
the series inductor model to an equivalent parallel model, we can proceed w'ith our standard
analysis. To illustrate this conversion, consider the practical inductor model o f Figure 16.12a and
the equivalent parallel configuration in Figure 1 6 .12b. The goal o f this example is to find
L in terms o f
and
and
at a particular O).
(a)
(b)
FIG URE 16.12 Conversion o f an inductor model from (a) a series connection to
(b) a parallel connection at a fixed frequency.
So
lu t io n
For Figure 16.12a,
Y<ko) = _____!____
For Figure 1 6 .12b,
yrU (o) = - j
Rp
Equating the real and imaginary parts o f the above t\vo admittances, we obtain
Rp =
and
ly
= Rs[' +Q h<o))
(16.30a)
83^
co"-L,
q I uo)]
where
(oL
At any particular frequency (O, if Qy (to) is sufficiently large (say Qj (o)) > 8), then equations 16.30a
and b suggest that
R., a
{(o)Rs = Q i( co ) x( oL
(16.31a)
^p-^s
(16.31b)
and
Conclusion: In a "paraller RLC with a practical inductor, we can replace the practical inductor by its
parallel counterpart valid in a neighborhood o f a single frequency and analyze the circuit to obtain
approximate values o f peak frequency, bandiuidth, etc.
Exercise. A 2 mH coil purchased from an electronic parts store has a
50 at 100 kHz. Find
the element values in the series representation and the parallel representation at 100 kHz as shown
in Figure 16,35.
A N SW ERS: l.^ = 2 mH, R^ = 25.13 LX
s 2 mH, R^ = 62.83 \dl
A similar derivation can be done for the practical capacitor model o f Figure 16.13a. The details of
the derivation are left as a homework problem. The exact conversion equations for a specific to are
(,
and
\
(16.32b)
As before, for high
(say >8), these equations reduce to the simpler forms o f

(16.33a)
and
1
Q c (c o )
(1 6 .3 3 b )
(a)
(b)
F-KiriU-. 16.13 Conversion of a capacitor model from (a) a parallel connection to
(b) a series connection exact at a fixed fi-equency.
Table 16.2 summarizes the various conversion formulas for both parallel to series and series to par
allel inductor and capacitor models.
I'AMI.F 16.2 Conversion of Models for Practical Inductors and Capacitors
Approximate
Original Circuit
Exact Equivalent
Circuit at (d,,
Equivalent Circuit, for High
{Q, > 8
> 8)
and
and u) within (1 0 .0 5 ) oJq
Q^R.
=Q,.o)oL
R (i+ Q l)
Hf-
Q,((i)^ = o)R,C
Q;
CL
1 + 1/Q'
1
Qc(o) = (oR,C
Q,
S '-
The next example illustrates the application o f the conversion formulas to a practical circuit to
obtain approximate band-pass characteristics.
EXA M PLE 16.7. The circuit o f Figure 16.14a contains practical components that make the cir
cuit not amenable to the formulas for ideal series and parallel RLCs developed earlier. For exam
ple, the sinusoidal source is represented by an independent voltage source
internal resistance
tion factor o f
= 40
in series with an
The practical capacitor has a value o f C = 0.05 |.iF and a dissipa
0.01 at cOq. Here cOq represents the undamped (no resistance) natural frequen
cy o f the circuit, i.e.,
^
' = 7 Z c -
The practical coil has an inductance o f 20 mH and Q/(cOq) = 40. The external load resistance is
/?^ = 10 k n . Our goal is to find approximate values o f
cOj, CO2, and
(a)
leakage
signal source
(b)
FIG URE 16.14 Approximate analysis of a high-Q circuit, (a) Original circuit,
(b) Approximate parallel /?Z,C circuit.
Solution
Step 1. Compute cOq.
= 31,623 rad /s
COq =
V 0 .0 2 x 0 .0 5 x 10"^
83R
Step 2. Find the parallel equivalent circuit values for the practical inductor. Since Q^(cOq) = 40, from
equations 16.31 and column 3 o f Table 16.1,
Rp = Q/(Wo)WoZ. = 40 X 3 1 ,623 x 0.02 = 25,298 Q.

and
L r-L
Step 3. Represent the capacitor model by a parallel RC. First, the dissipation fiictor tells us that
Qc ^^q) - ~T~ - 100

dc
From row 3 and column 4 of Table 16.1, we note that the parallel resistance associated with the
capacitor is the so-called leakage resistance
^ ^
^^
clcCo^C
^ ------------ ^ 0 ---------- ^ ^
o jj:
3 1 6 2 3 x 0 .0 5 x 1 0 " ^
Step 4. Replace the practical source with its Norton eqidvalent and compute the eqtdvalent parallel
resistance, denoted as R. Replacing the practical source with its Norton equivalent and incorporat
ing the results o f steps 2 and 3 produces the network o f Figure 1 6 .14b. The parallel combination
o f R^, R^, ^leakage* and R^ is (using the notation // to indicate the parallel combination)
R = 40,000//25,298//63,247//10,000 = 5545 n
Step 5. Approximate analysis o f the circuit o f Figure 16.14a using Figure 16.14b. Our formulas from
the ideal parallel RLC case now apply, but the results are, of course, approximate for the circuit o f
Figure 16.14a:
0)^^^ =
1
RC
co^ = 3 1 ,6 2 3 rad/sec
=
------- = 3 ,6 0 7 rad /s ,
5 , 5 4 5 x 0 . 0 5 x 1 0 ^
3,607
and since Q^.-^ is sufficiently large.
- - ^ = 3 1 ,6 2 3 - 1 ,8 0 3 .5 = 2 9 ,8 1 9 rad/s,
0 )2
+ - ^ = 3 1 ,6 2 3 + 1,803.5 = 3 3 ,4 2 6 rad / s
Finally,
^max =
---- = ------------ 5------------------- Z----------- = 0 .1 3 8 6 2

4 0 X 10^ X 0.05 x 10"^ x 3607
R,CB,,
839
5. THE RESONANCE PHENOMENON AND RESONANT

CIRCUITS
The term resonance has different meanings in different disciplines. From Webster's Collegiate
Dictionary, in the field o f engineering, resonance refers to the phenomenon o f a vibration o f large
amplitude in a mechanical or electrical system caused by a relatively small periodic stimulus o f the
same or nearly the same period as the natural vibration period o f the system. In this section, we
shall investigate this notion o f resonance and its manifestation in RLC circuits, in both cases uti
lizing the theories studied in previous chapters. The main applications o f resonant circuits are for
the filtering and tuning purposes. Additionally, resonant circuits can be used to transform a resist
ance from one value to another value (at a single frequency) to achieve maximum power transfer.
This matching application is discussed in the last subsection.
The Resonance Phenomenon

A child on a swing knows by instinct when to flex his or her knees in synchrony with the swings
pendulum motion to make the swing go higher and higher. This activity illustrates the phenom
enon o f resonance, in which small and quick leg movements at Just the right moment produce the
large pendulum-like motion o f the swing. If the child stops his or her leg movements, the swing
ing motion gradually dies down.
Let us approximate the pendulum motion o f the swing using a linear system with transfer function
+
(16.34a)
h it)= K e~ '''' s\nico,it)ii{i)
(16.34b)
and having the associated impulse response
Mathematically, the periodic leg movements are modeled by the periodic impulse function
x(r) = /l [ 6 ( t )
+d{t-T)-^6{t-n)^
...]
(1 6 .35 )
where T = 2 7 t/ o )T h is choice coincides with the period o f the natural response o f the system
transfer function. The build-up in magnitude can be seen quickly through superposition. The
contribution to the output, sayj/(r), due to the first Impulse at / = 0 in equation 16.35 Is
\\){t)=ÂKe
sln(f/j^//)//(/)
(16 .3 6)
Figure 1 6 .15a shows the form o f this equation. Note that successive positive peaks {Kq, aV^, a^V^,
a^Vf^, ...} decrease geometrically by a = e~^P^. In fact, the waveform o f each period replicates the
waveform o f the prior period scaled by the factor a.
From time invariance, the contribution to the output ^(/) due to the second impulse at r = 7'in
equation 16.35 is simply J'q(/ - 7), as shown in Figure 16.1 5b. Similarly, for the third input
impulse zi t =1T, the response 1s7q(^- 2T), as shown In Figure 1 6 .15c.
8-10
From superposition, the response over [0, 3T) \s simply the sum
i.e., for 0 < r < 37'. This sum is illustrated in Figure I6.15cl.
a^)
F IG U R E 1 6 .1 5
Chapter 16 Rand-Pass Circuits and Resonance
In Figure l6.15cl \vc observe thar successive positive peaks arc given by
(1 +
(1 + a +
In general one can show that the th positive peak is given by
= 1
rt-1
11 + ^/+ t/ H------ ~
(16.37)
l- "
k=0
For our purposes n =
< 1 since
> 0. Hence as n gets large,
\-a
1-
This means that the waveform reaches a steady-state periodic response in which the positive peaks
have value
v
\-a
Exercises. 1. Consider the slightly damped second-order system with transfer function
H(s) = ------ ----(s + 0.1)" + 4

having a periodic input
x{t) = 2[6(f) + 6 ( r - . t ) + 6 { r - 2;r) + ]

(a) Find the first five positive peaks.
(b) Find the positive peak in steady state.
ANSWHRS; (a) 3.698, 7.116, 10.28, 13.2, 15.9: (b) 48 .9 6
2. Suppose in the previous exercise
is changed from 0.1 to zero, i.e.,
5 +
(a) Find the first five peaks for the input o f the previous exercise.
(b) Is the system stable? Why or why not?
ANSWHRS; (a) 4, 8, 12, 16, 20. (b) The system is unstable bccausc there is a pole on the imagi
nary axis of the complex plane.
Another interesting application o f resonance is in product security in stores. The securit)' tag is an
RLC circuit with very small R. The circuit is excited by a radio wave at its resonant frequenc}' by
the security panels in front o f the exit doors. If the circuit has not been destroyed at the checkout
counter, it begins to resonate as one approaches the exit and transmits a signal back to a detecting
device at an amplitude much higher than the original transmitted signal. Fhis sets ofi an alarm.
The easiest way to understand this phenomenon is by way o f frequency response. Let us consider
the second-order transfer function
H(s) = --------- ---------(5 -h 0 .1 )-+ 4
84 2
The magnitude frequency response is given in Figure 16.16. Notice the sharp peak at about 2
rad/sec, which is 10 times the dc gain o f about 1. Thus sinusoidal inputs at frequencies close to 2
rad/sec produce a steady-state response with magnitude almost 10 times larger. This is precisely
the type o f resonance phenomenon that occurs with the security tags at stores.
Frequency in rad/s
F IG U R E 16.16
Series and Parallel Resojiant Circuits

The simplest stable linear circuits capable o f producing the resonance phenomenon are the series
and parallel resonant circuits shown in Figures 16.17a and b, respectively.
/Y Y ^
L
R
V
(a)
:D
L
>R
^
(b)
F I G U R E 1 6 .1 7 Series (a) and parallel (b) reson an t circu its.
843
Using the frequency response approach o f Figure 16.16, we see that small inputs in these circuits can
produce large outputs when the conditions are right: (i) a high-Q circuit and (ii) input signal fre
quency and the peak frequency coincide. To develop a circuit-theoretic perspective on resonance we
take a more basic approach using the sinusoidal steady-state analysis ideas o f Chapter 10.
For any circuit containing one inductance L and one capacitance C, denote by (Oq the frequency
at which the two reactances
= -1/oJqC and
= (OqZ have equal magnitudes, i.e.,
(16.38)
in which case
0)^ =
(16.39)
lc
In Chapter 9, on second-order RLC circuits, cOq = I I JT C is called the undamped natural fre
quency. The name stems from the fact that if all resistive elements are absent (i.e., the circuit is
undamped), then a parallel or series connection o f L and C produces a natural response o f the
form K cos((Oq^ + 0). In the jargon, the parallel LC circuit o f Figure 1 6 .17b is called a tank cir
cuit, and 0)q is called the tank frequency.

Since
and jXj^ =
have opposite signs and equal magnitudes, if the elements L
and C are connected in series, then at (Oq the resulting impedance,
0 and hence is equivalent to a short circuit. Similarly, if the elements L and C are connected in
parallel, then at (Oq the resulting admittance is zero and hence equivalent to an open circuit. These
properties are illustrated in Figure 16.18.
OZ(jO)o)
Z(jCOo)
O(a)
o-
O-
Y(jcOo)
Y(jcOo)
o-
O(b)
FIG URE 16.18 (a) Series- and (b) parallel-connected ZCelements with coq =
>/Zc
With the short circuit and open circuit equivalents sho\vn in Figure 16.18, we can easily deduce die fol
lowing properties for the series and parallel resonant circuits of Figure 16.17 operating a t : coq =
8 ''il
Chapter 16 * Band-Pass Circuits and Resonance*
(1)
The magnitude o f the impedance seen by the voltage source in the series RLC o f Figure
16.17a is minimum, and the impedance is a pure resistance equal to R.
(2)
The magnitude o f the impedance seen by the current source in the parallel RLC o f Figure
(3)
The magnitude o f the voltage across C o r I in the series /?ZCof Figure 16.17a is Q^./;.(Wq)
1 6 .17b is maximum, and the impedance is a pure resistance equal to R.

= oJqZ.//? times the magnitude o f the source voltage.
(4)
The magnitude o f the current through Z, or C in the parallel RLC o f Figure 1 6 .17b is
O^.y^(ojQ) = cOq RC times the magnitude o f the source current.
To establish property' (3), we use voltage division and propert) (1):
> ()C
M )L
COqL
^ rO o )
A similar derivation yields property' (4).
Exercise. Derive property (4) using current division and property (2).
Properties (1) and (2) lead to a general definition o f resonant frequency, denoted to^.. Specifically
is the frequenq' at which the source sees an impedance or an admittance that is purely real, i.e., pure
ly resistive, despite the presence o f capacitors and inductors. For the series and parallel RLC circuits
For more general circuits,
oJ q.
Circuits operating at ojâre said to be at resonance.
Exercise. A sinusoidal voltage source at 1 MHz is applied to a series RLC circuit. If Z. = 300 uH,
/? = 5 n , and C is adjustable, what value o f C produces resonance?
A N SW ER: 84.4
p 1-
Using property (2) above, the calculation o f
in many o f the earlier examples can be done with
very little effort. For example, in Example 16.2, at resonance.
RL
10'
R, + Ri^
4 x 1 0 -^ + 10-^
Exercise. Use propert}' (2) to calculate

A N S W l - K : 1//^
0.2
in Example 16.4 for Figure 16.5.
The resonance condition co^ = (O^ = 1/V LC can be achieved by vary'ing one o f the three parameters
oJq, C, or L. When C or Z, or both arc adjusted to achieve resonance, the circuit is often called a tuned
circuit. The next example illustrates how this tuning can be used in a practical application.
84^
EXA M PLE 16.8. Figure 16.19 displays an amplifier model containing a V CC S with g^ = 2 mS
(milli-siemens) and
= 20 k fi. The applied sinusoidal voltage,
V at 10 MHz. The load is modeled by the parallel combination o f
has a magnitude o f 0.1

and the 40 pF capacitor;
the capacitance accounts for such real-world phenomena as wiring capacitance, the device input
capacitance, and other embedded capacitances. This capacitance cannot be removed from the cir
cuit and often has deleterious effects on the amplifier performance.
FIGURE 16.19 Amplifier circuit illustrating the application of the tuned circuit conccpt to elimi
nate undesirable capacitive effccts.
The example objectives are
(a)
With the load connected directly as shown (without I ) , find the magnitude o f the out
put voltage.
(b)
If an inductance L is connected across the load to tune out the effect o f the capacitance,
find the value o f L and the resulting
that will show that the amplifier gain at 10
M Hz is greatly increased.
So
lu t io n
(a)
At 10 MHz, the load impedance is
3 p r = 3 9 7 . 8 Z - 8 8 .9 Q .
0 .00005 + y 2 jrx 10 X 4 0 x 1 0
Therefore, since jV'il =
|= 0.1 V, the magnitude o f the output voltage is

I
^ 0-002 X 3 9 7 .8 = 0.0 7 9 6
Here the voltage gain is 0.0796/0.1 = 0.796 due to the low impedance o f C at the high operating
frequency.
(b)
By tuning out the effect o f the capacitance, this poor gain response can be eliminated.
The inductance needed to tune out the capacitance is calculatcd from equation 16.39:
L = A - = = 7------- n-^------------ T T T = 6.33 X 10"^ H

ojqC
4 ji
10
X 4 0 x 10
W ith a 6.33 uH inductor connected across the load, the parallel Z.C behaves like an open circuit
at 10 MHz and the load looks like a pure resistance o f 20 kQ to the amplifier. The new output
voltage magnitude is
846
Iôut
1^ 0-002 X 2 0 ,0 0 0 = 4 V
wich a resultant voltage gain o f 4/0.1 = 40.
Exercises. I. Suppose the frequency o f the input is changed from 10 M Hz to 5 M Hz in Example

16.8. Find (a) the gain without the inductor connected, (b) the value o f the inductor to tune out
the capacitance, and (c) the resulting gain with the capacitance tuned out.
ANSWHRS: (a) 3.18, (h) 2S.33 ul 1. (c) -40
2. In Example 16.8, if the embedded capacitance is 63.3 pF instead o f 40 pF, what inductance is
needed to tune out the capacitive effects?
AN SW ER: 4 IIH
Series-ParalelResonantCircuits
For parallel and series RLC circuits, co^ = cOq. For series-parallel circuits containing only one L and
one C when it exists to^ ^ (jd^j in general, but it may not exist at all. The next example illustrates
the point.
E X A M P L E 16.9. Find the resonant frequency,
and the input impedance at co^for the circuit
shown in Figure 16.20.
/YYY
L
Z(jo))
FIG URE 16.20 A variation of a parallel resonant circuit.
SO L U T IO N
Step 1. Calculate the admittance "looking into the input node pair o f the circuit o f Figure 16.20. By
the usual techniques.
(16.40)
K(/oj^) = ----- !----- = M C + ------- !-------=

Z O ,)
R + ji^ ,L
y ? ^ + (a ),L r
Step 2. Set the imaginary part to zero. Resonance occurs when Kis real, i.e., when
\m{Y}= o ) ,C ----- ^ = 0 .
R~ -t- (oifL)
Solving for
and then expressing it as a function o f Wq = \JyjLC yields
0).
R-
1-
CR-
(16.41)
84
The rightmost term shows how the resonant frequenc)' is scaled away from the parallel or series
ideal cases where to^ = cOq.
Step 3. To obtain the values o f the admittance and impedance at resonance, substitute this value
o f (Of into equation 16.40 to obtain
Yij0^r) =
RC
RC
(16.42)
Three conclusions can be drawn from equations 16.40 through 16.42:

1.
If {CR^)IL > 1, then there is no real solution for (0^ this means that the source voltage
and source current cannot be in phase at any frequency.
2.
If {CR~)IL < 1, then there is a unique nonzero resonant frequency O)^ that is strictly small
er than Wq.
3.
When O)^ exists, i.e., there is a real solution to equation 16.41, then at (O = CO^the source
sees a pure resistance, the value o f which equals {LIRQ and is greater than R.
Exercises. 1. Fill in the details o f step 3 above.

2. In Figure 16.20, let Z. = 1 H and C = 1 F. Find the resonant frequency co,. for (a) R = 0.8 Q,
and (b) 7? = 2
Compute the resistance seen by the source in each case.
AN SW ERS: (a) 0.6 rad/scc, 1.23 H; (b) (O^. docs not exist.
Results similar to equations 16.40 through 16.42 can be derived for the circuit o f Figure 16.21.
/YY\
Z(jco)
L
R
FIGURE 16.21 A variation o f a series resonant circuit.
Exercises. 1. For the circuit o f Figure 16.21, let G = MR. Show that the input impedance is
./
G " + ( ojC )
coC
coZ-- ^------7
G~ + (coC)
2. Again for Figure 16.21, show that the resonant frequency is

0),. = to,
1 - ^
(16.43)
CR'
3. Finally, show that the input impedance at the resonant frequency is
Z(jo),) =
L
RC
( 1 6 .4 4 )
848
Chapter 16 Band-Iass Circuits and Rcsonancc
In the above example and exercise, vve concluded that

When co-exists, i.e., there is a real solution to equation 16.41, then at to =co^
the source sees a pure resistance, the value o f which equals {URQ.
This property finds application in maximum power transfer from source to load in fixed-frequency situations, as illustrated by the next example.
E XA M PLE 16.10. The output stage o f a certain radio transmitter is represented by a 1 MHz sinu
soidal voltage source having a fixed magnitude o f 50
and an internal resistance o f 100 Q as
shown in Figure 16.22a. A load resistance Rj models an antenna connected to the transmitter also
shown in the figure. The purpose of this example is to shov/ how a m atching network based on
the principle o f resonance can be designed to maximize the power delivered to the antenna.
(a)
If Rj is adjustable, find the value o f
(b)
by the load. What is the value o f (/*jr)niax^

If Rj = 2Q Q. in Figure 16.22a, find the value o f P^.
(c)
Suppose that Rj^ is fixed at 20 O., but a coupling net%vork consisting o f LC elements is
yielding the maximum average power
absorbed
inserted between the source and load to increase the power P^ as shown in Figure 16.22b.
Choose values for L and C in the circuit so that {Pi),j,^j^ o f part (a) is again obtained.
1000
lOOO
/YYV
L
50 V
20 0
1MHz
IM Hz
(b)
(a)
FIG URE 16.22 Matching load to source using a resonant circuit, (a) Load connected direcdy to
source, (b) A coupling network designed to maximize the load power.
So
lu t io n
(a)
From the maximum power transfer theorem,
Ri = R^ = 100 n
and
2,500
AR,
(b)
400
= 6.25 W.
With R^ = 20 Q., by voltage division
20
in which case
100 + 20
50 = 8.33 V,
3 .4 7 2 w .
20
(c)
S h9
If we can make the impedance at the input terminals o f the LC coupling network equal
to (100 + yO) LI, then maximum power will be drawn from the source. Since Z.Celements
consume zero average power, the same maximum power will be delivered to the load
resistance. The resonant circuit shown in Figure 16.22b provides a possible design. The
LCR" load circuit is the one analyzed in Example 16.9. Hence, to calculate the element
values, we use equations 16.41 and 16.42 as follows:
7
6\
1
20
(oT = 2 ji X 10 j -----------(from equation 16.41, squared)
LC
and
Z{jiD^)= 100 =
(from equation 16.42)

20 C
Solving these equations simultaneously results in Z = 6.37 uH and C = 3.18 nF.

Some remarks about this design are in order:
1. fg= \I{2ii J T c ) = 1.12 MHz, and fj. = 1 M Hz i f j . The 20 D. resistance is transformed
into a 100 Q resistance at/ = fj. (not at/ = J^).
2.
Should the source resistance be smaller than the fixed load resistance, C is moved to be
in parallel with
In that case, equations 16.43 and 16.44 are used in place o f equations
16.41 and 16.42.
Exercise. Redesign the coupling network in Example 16.10 if the resistors are
= 300 Q and Rj
= 50 n .
AN SW ERS: 17.79 ul 1 .nui 1.18 nl-
A variation on the computation o f the resonant frequency is the design o f a circuit to achieve a
desired resonant frequency using a variable capacitor; this is the design that underlies the tuning
o f many AM radios.
E X A M PLE 16.11. Consider the series RLC o f Figure 16.23. Here, the voltage source has a fixed
magnitude |V^| and a fixedfrequency O). W ith R and L fixed, we seek the value o f the variable capac
itance C that maximizes the magnitude o f the voltage across the capacitor.
F I G U R E 1 6 .2 3 A d ju stin g
C fo r
m a x im u m o u tp u t voltage.
850
The first step is to compute the magnitude o f the voltage across the capacitor using vohage divi
sion:
I, , I
Vc
o
R -+ U)L-
1
coC/
\/
Maximizing |K j is equivalent to maximizing \V(^ or minimizing------- . To obtain this last expres
sion, we square the above expression for \V^ and rewrite it as
/ (C ) = ------ ^ = (coCRr + (CO^LC - 1)l^ c l"

To m in im izey(0, s e t / '( Q to zero, i.e.,
/ '( O
+ 2 ( lo ^ L C - \ W L
= 0
Solving for C produces
C=
R~ +(coL)
Zxo
+1
Although this value o f C produces a maximum capacitor voltage, the circuit is not hi resonance., as
the value o f 1/VLC is not equal to the signal source frequency o). However, given this equation
for C, for a high-Q circuit (ojZ//? > 8), the condition is practically the same as l/VTC = to.
E X A M PLE 16.12, In Figure 16.23, let

Find I
= 1 V, (O = 10^ rad/sec, ^ = 5 H. Let C b e variable.
and the corresponding value o f C for each o f the following cases: (a) Z, = 1 mH; (b)
L = 100 uH.
SO L U T IO N
(a)
For L = 1 mH,
= toZ,//? = 10^ x 10"^/5 = 20. This is a high-Q circuit. Therefore, C is
given approximately by
C = - \ - = ^ ------ r = ( ) . l x l O ^ F ,
to -L
and I Vy
lO '^ x lO " '
is given approxmiately by
l^ c U = Q Jn i = 2ov
Exact solutions follow from equations 16.48 and 16.47:
0 .00 !
an d
x-6
.
^ = 0 .0 9 9 7 5 x 1 0 " F ,
5^ +(10^ X 0.001 r
n
J5 +
7 T
1 0 ^ x 0 .0 0 1 -
8S
10-' X 0 .0 9 9 7 5 X 10
10^ X 0 .0 9 9 7 5 X 10"^ J
Plainly, the approximate solutions are very close to the exact solutions.
(b)
For = 100 uH,
= coZ//? = 10^ x 10^/5 = 2. This is a low-Q circuit, requiring the
use o f equations 16.48 and 16.47. Here,

0-0001
C=
5
5 2 + ( 1 0 ^ x 0 .0 0 0 ir
,^ _ 6 ^
0 .8 x 1 0
F,
and
= 2 .2 3 6 V .
2
5^+
lO-'^ X 0 . 8 x 1 0 ^
iO-'^x 0 .0 0 0 1 1 0 ^ x 0 .8 x 1 0 " ^ /
Exercise. For part (b) o f Example 16.12, compute cOq = 1/ V lC . Is this value equal to the signal
frequency? Why or why not?
AN SW I:R. {.)(, = 1 .1 1 8 x 1 0 ^ rad/scc
6. GENERAL STRUCTURE OF THE BAND-PASS TRANSFER

FUNCTION WITH ONE PAIR OF COMPLEX POLES
W ith the experience gained from the analysis o f the first five sections, we now present a general
transfer function approach to the analysis o f band-pass type circuits. The circuits considered in
this section will contain only one inductance and one capacitance. Hence any associated network
function will have at most a second-degree polynomial in s as the denominator and numerator.
Thus, the general transfer fiinction H{s) (which includes the impedance function Z{s) and the
admittance fiinction F(j) as special cases) has the biquadratic form
u( A -
d(s)
s~ + 2 o pS + io~p
(16.45)
A reasonably sharp band-pass characteristic requires that H{s) have complex poles, i.e.,
<
O)^.
The finite zeros o f H{s), which are roots o f n{s) = 0, may be real or complex. The case o f complex
zeros, corresponding to more advanced filter characteristics such as the inverse Chebyshev or ellip
tic types, is beyond the scope o f this text. For practical reasons we focus on the case where H{s)
has one real zero or no zero.
When H{s) has one real zero and complex poles, then equation 16.45 reduces to
+
_
s- + 2 o ^ s + (o l
a\S + ao
+
"
Chapter 16 Band-Pass Circuics and Resonance
SS2
with the pole-zero plot shown in Figure 16.24.

jco
A
p, = -^ p + i d
\p
s-plane
(0p2 = ap2 + C02
-a
(0
-e -
-a
Q =
P2 = -^ p -j d
FIG UIIE 16.24 Pole-zero plot of the transfer function H{s) of equation 16.46.
Equation 16.46 contains the usual qualit)' factor, Q , mathematically called the pole
defined as
CO,
(16.47a)
= arcsin
(16.47b)
\^P/
where the angle 0 is as shown in Figure 16.24. Relative to the pole-zero plot,
measures how
close the pole is to the/o-axis: a higher Q means a smaller 0, implying a pole closer to they'coaxis. As we already know,
is related to the circuit Q, i.e.,
the sharpness o f the response curve. For some special cases,
In Figure 16.24, another new quantity,
and serves as a quick estimate o f
appears. To ascertain the meaning o f to^y, recall from
Chapter 15 that the impulse response o f a system characterized by equation 16.46 has the form
hit) =
r^
{/ / (.V )} =
Ke'""'" c o s ( w , y
d)
The waveform h{t) is a damped sinusoid, and the quantit)^

(not co^ specifies the frequency o f
oscillation. For this reason, co^ is referred to as the damped oscillation frequency.
For a transfer function o f the form o f equation 16.46, our goal is to determine several key quan
tities that are indicative o f the circuits behavior:
frequencies.
the bandwidth
and the half-power
Case 1. No finite zeros. When the transfer function o f equation 16.46 has no finite zero, then it
reduces to
H{s) =
K
(16.48)
.V" +
The pole-zero plot o f this H(s) differs from that o f Figure 16.24 only in that the single zero is now
absent. The series RLC circuit o f Figure 16.25a with the capacitor voltage as the output and the par
allel RLC o f Figure 16.25b with the inductor current as the output both fall into this category.
fYYVV.
(b)
(a)
FIGURE 16.25 (a) Scries RLC with capacitor voltage as output.

(b) Parallel RLC with inductor currcnt as output.
The transfer function o f equation 16.48 can represent either a low-pass or a band-pass character
istic depending on the value o f
and the value o f H{0). In fact, the transfer function can display
both characteristics as well as intermediate behaviors. For example, if

o ,,,
_________OOOO)-
V +
1000
VI
then the maximally flat low-pass characteristic o f Figure 16.26a results.

On the other hand, if
(1000)^
H{5) =
.v +
1.31
5 + (1 0 0 0 )-
then something between a low-pass and a band-pass characteristic results, as shown in Figure
16.26b. Here low frequencies are still passed, yet the characteristic has a selectivit)' propert}^ result
ing from the pole
o f 1.31. The ratio o f the maximum gain to dc gain is yfl . This means that
the peak is 3 dB above the dc gain. Now, with increasing pole Q the selectivity goes up, as does
the maximum gain, and the characteristic looks more and more like a pure band-pass. Finally, if
H{s) =
(1000)*-
1000
.v + -------- s
10
(1000)*
then the approximate band-pass characteristic o f Figure 16.26c results. Here the pole Q is 10 and
the ratio o f the maximum gain to dc gain is also 10. Although low frequencies are not attenuat-
854
ed, the characteristic is highly selective and frequencies near co^ are highly amplified, so that for
all practical purposes the characteristic is identified as band-pass.
Frequency in rad/sec
(a)
(b)
Frequency in rad/sec
(c)
FIGURK 16.26 (a) Low-pass characteristic, (b) Moderate Q^, resulting in characteristic exhibiting
both low-pass and band-pass behavior, (c) High-Q^ case, showing a marked band-pass characteristic.
8^5
The important aspects of the curves in Figure 16.26 are co^,

and then
and
We first compute O)^
At the peak value, the derivative o f |//(/o))| is zero. Setting the derivative of |//(/to)|
to zero and solving for O) produces the exact formula ^
< B = 0)p
W
l - ^
^ P ^ co s(2 0 )
To compute
consider
'm
^(Op - ct)^ co s(2 0 )j + (U p 4sin ^ (0)co s(20)

(o = o )p y Jco s(2 d )
ct)p4sin (0 )c o s (0 )
(Op sin (26)
Hence
K ____
^/=
2 .
in(2d)
H (0)
sin(20
0 6 .5 0 )
Similar derivations with more complex algebra yield the half-power frequencies,
<^1,2 = (Opyl cos(2d) T sin(20)
(16.51)
^(o = p ^ 2 (c o s ( 2 0 ) - 7 c o s ( 4 0 ) )
(16.52)
and the bandwidth,
provided there exist real solutions.

's-/
Referring back to equation 16.49, if Q < l/V 2 = 0 .7 0 7 or, equivalently, if 0 > 4 5 , then there is
no real solution for (o^. In this case, almough the poles are complex, the magnitude response dis
plays no peak at any firequency. Rather,
= |//(/))| and |//(/o))| decreases monotonically to zero
as (0 increases, as demonstrated in Figure 16.26a.

When
W
is greater than l/V 2 or, equivalendy, if 0 < 4 5 , the magnitude response of equation
16.48 starts fiom a nonzero value at co = 0, rises to the peak value at co = O)^, and finally decreas
es to zero as CO - 00. This behavior is illustrated in Figures 16.26b and c. If
is only slighdy
greater than 0.707, then the magnitude response is essentially that of the low-pass type, with a
small hump in the pass band.
For the high-(^ case ( ( ^ > 8), the magnitude response near co^ approximates that of a pure band
pass circuit. The preceding exact expression for (o^ reduces to O)^ a o)^ for the high-(^ case.
856
Similarly, for high Q^,
t^\Qp
s 2a^,
. and ojj 2 -
I'"*
error o f these
approximations is less than 0 .5 % .
Exercise. For
= 8, (o^ = 1000 rad/sec, and A^= in equation 16.48, computc the exact and approx
imate values o f
and
Then compute the percentage error in the approximation:
\exact - approximate
X 100
\exacf\
AN SW liRS: Exact values arc
Approximate values are
= 99 6 .0 9 rad/sec, B^^^ - 125.5 rad/sec, and
1000 rad/sec, B^^^ = 125 rad/sec. and
- 8.015^.
= 8. Percentage errors are,
respectively, 0 .3 9 ^(), O-.V)'*/!), and 0.196% .
EXA M PLE 16.13. Consider the Sallen and K eyzcw e network o f Figure 16.27, which can be used
to realize the transfer function o f equation 16.48. As per a homework exercise in Chapter 14, the
1
H(,v) =
Ri R2C[C2
' nut
U ,c ,
R2 C J
R\R2<îC2
Suppose Cj = 2 uF, C , = 5 nF, /?j = /?2 = 10 kQ.

and K o f equation 16.48.
(a)
Compute co^,
(b)
Compute exact values o f
(c)
Compute approximate values o f
(d)
(e)
Plot the magnitude response curve o f the transfer function.

Simulate the circuit using SPICE.
and
0)j, and co,.
FIGURE 16.27 Sallen and Key active net\vork for realizing the transfer
function o f equation 16.48.
So
8^'
lu t io n
For the given parameter values,

_ V, _
(1 0 0 0 )2
'm
(a)
By inspection (o^ = 1000 rad/sec,
(b).
From equation 16.48,
I --------V =
(JO,,, = o)p
11
Numerically evaluating N(s) at s =
5 + (1 0 0 0 )
10
= 10, and K =
= 10^.
9 9 7 .5 rad/.sec.
~^P
we obtain
= 10.013 via the following MATLAB code:
n = wp^2;
d = [1 w p / Q p w p ^ 2 ];
Hm = abs(pol\âl(n,j*wm))/abs(pol)âl(d,j'wm))
Hm = 1.0013e+01
(c)
The approximate values are given by

s
2 a ^
1 0 0 , 0 ),
co ^ -
= 9 0 0 , CO
2^
1 1 0 0 ,
all in rad/sec.
(d)
'lb obtain the magnitude response for the given transfer function, as shown in Figure
16.28, we use the following MATLAB code:
n = wp^2;
d = [1 w p / Q p w p '^ 2];
f = logspace(l ,3,600);
w = 2*pi*f;
h = freqs(n,d,w);
semilogx(f,abs(h))
grid
ylabel(Magnitude H(jw))
858
Frequency in Hz
FIG URE 16.28 Magnitude response o f high-Q^ active circuit of Figure 16.27.
(e) A SPIC E simulation yields the corresponding plot in Figure 16.29.
Sallen and Key-Small Signal AC-2

(V)
+10.000 -H 5.849
Frequency(Hz)
+25A 19+39.811 + 6 3 .0 9 6 + 10 0.00 0+ 1 58 .48 9+ 2 51 .1 89 + 3 98 .107+ 630.957
F I G U R E 1 6 .2 9 SPIC E sim u la tio n o f activ e circu it o f F igu re 1 6 .2 7 .
8S9
Exercise. Suppose Cj = 0.2 j.iF, Cj = 0.5 nF, and /?j = ^2 ~

tion 16.48, along with exact values o f
A N SW ERS:
A"of equa
Find approximate values o f B^, Wp and Wj.
and
= 10, K = I0 \
= 10^ rad/sec,
= 9975 rad/sec. /-/,,, = 10.013, /?, =
10-^
rad/sec, 10, 2 9000 rad/scc, to, = 1 1,000 rad/scc
Case 2. A single zero o jf the origin. In equation 16.46 with a^^
0,
i=- 0, and
and <7, o f the
same sign, a zero is present in the left half-plane. A second form o f the band-pass transfer func
tion o f equation 16.16 is
s +a
H{s) = K
=K
CO,
s +
Qn
s+ a
s~ + 2 0 p.v + oj
(16.53)
Figure 16.32 sketches the pole-zero plot for this transfer function. Intuitively speaking, the closer
the zero is to the origin, the more the magnitude response resembles the response o f the case with
a single zero at the origin. For the transfer function o f equation 16.53, derivations o f
of
values
and the zero-phase-shift frequency are possible. (See the homework problems.) The results
are
+ J(o )l + a -y -(2 o^ ay ,
to.., =
CO (for zero phase shift) =
(16.54)
- 2o^,a
(16.55)
No exact expressions are available for the half-power frequencies and the bandwidth
case o f high
and a
For the
approximate answers are

1^1 ,a n d f i ^ ^ 2 a ^ ,
2o,
for
>8
(16.56)
To see why, observe that

jc o +
A.
9
(
+j2oxj^^
jo3 + a
K
IS.
-)
(yco) + j 2 o i( j j , + co
co
(16.57a)
>
1+
/ c o - - c o '\
(16.57b)
2 o + ./
03
The second factor is approximately 1 for frequencies near co^ by virtue o f our assumption a OJ^.
Hence the properties o f the magnitude response reduce approximately to those o f the first factor
in equation 16.57b. These properties are those o f a single zero at the origin. Hence for high-Q^^
circuits, the relations o f equation 16.56 approximate the single zero at the origin case.
(S60
1^1
Exercise. Use equation 16.57b to show that when
> 8 and a to , H
?Q
When exact values are desired, one must resort to SPIC E or MATLAB to obtain the frequency
response from which values can be graphically determined. SPIC E has the advantage o f not hav
ing to compute the transfer function o f the circuit; MATLAB requires this computation.
7. SUMMARY
This chapter began with a study o f simple series and parallel
band-pass circuits. Because o f
its generality, we set forth a transfer function approach to the analysis and design o f these band
pass circuits. Specifically, we first investigated a second-order transfer function with a single zero
at the origin for which
His) = K ---------------------= K - ------------------- ^

,v- + 2 o .v + co^,
Qp
'
For this transfer function we derived various formulas for determining band-pass parameters, such
as the peak frequency,
the peak value,
IG p
and the bandwidth,
to,
a~p
These formulas have a special form for the parallel RLO.
to,,, = 0)^ =
4 lc
which is also the resonant frequency of the associated impedance; the bandwidth is found to be
^ (0 = t0 2 -0 J, = - ^
while the circuit Q is
0 =
= u . R C = ^
The resonance phenomenon of a second-order RLC circuit was then set forth from a frequency
domain perspective. One application was the design o f a matching network that produces maxi
mum power transfer, at a single frequency, from a source with fixed internal resistance to a fixed
resistance load.
iSOl
After this we cook up the consideration o f more general second-order band-pass transfer func
tions for example, those with a single zero o ff the origin,
H(s) = K ------ ------------ = K

^
, 2
5 + 2 a.9 + Q)
.y + - s + oi~
P
P
a
or with no finite zero:
Qp
"
Other cases are left to more advanced courses on filter design.

Active bajid-pass circuit: a circuit containing operational amplifiers and no inductors that
achieves a band-pass characteristic.
Band-pass circuit: a circuit that passes signals within a band o f frequencies while rejecting other
frequency components outside o f the band.
Bandwidth (3 dB bandwidth),
CO2 - cOp the difference between the two half-power fre
quencies.
Dam ped oscillation frequency, (0^ frequency given by the condition that if the transfer func
tion o f a second-order linear circuit has complex poles at j = -a ^ jio^, then the impulse
response has the form K cos(co^f + 0). The constant (jo^ is called the dayyiped namralfre
quency.
Half-power frequencies; see lower and upper half-power frequencies.
L C resonance frequency: frequenc)' at which the reactances o f L and Chave the same magnitude;
equals M-JLC rad/sec. (Sec also tank frequency.)
Lower half-power frequency, COj: the radian frequency below the center frequency at which the
magnitude response is 0.707 times the maximum value.
M atching network: an LC net\vork that transforms a resistance
into a resistance o f a differ
ent, specified value at one frequenc)- or a band o f frequencies.

Peak frequency (center frequency),
the radian frequency at which the magnitude response
curve reaches its peak.
Qp (pole Q): for a pair o f complex poles j =

Qp = 1/(2 sin 0).
Quality factor Q (Qr/>) o f a band-pass circuit: the ratio o f the center frequency to the band
width, i.e., Q =
Quality factor Q {Qcoit) o f a coil: for a coil modeled by an inductance L in series with a resist
ance Rs y Qi =
and is frequenc}^ dependent.
Q uality factor Q c iQcap)
capacitor: for a capacitor modeled by a capacitance C in parallel
with a resistance R^^, Q c~
frequency dependent.
Q uality factor
o f a reactive com ponent: for an impedance expressed zs Z = R + jX, Q^ =
Reactance: in sinusoidal steady-state analysis, the imaginary part o f an impcdance. For Z,, the reac
tance is
for C the reactance is -l/(coC).
862
Chapter 16
Band-Pass Circuits and Resonance
-----------------------------------------------------------------------------------------------------------------------
Resonance frequency, 00^ the unique radian frequenq^ at which the input impedance of a twoterminal linear circuit becomes purely resistive.
Selectivity of a band-pass circuit: The circuit Q, defined as the ratio of the center frequency to
the bandwidth. A higher Q corresponds to better selectivity.
Susceptance: in sinusoidal steady-state analysis, the imaginary part of an admittance. For C, the
susceptance is coC For Z, the susceptance is -l/(coZ).
Tank circuit: the parallel connection of an inductor and a capacitor. In the idealized case (no
resistance), the total enei^ stored in a tank circuit remains constant, although there is a
continuous interchange of the energy stored in the various components.
Tank frequency, (Dq: defined as H j W in this text, regardless of the connection of the single L
and single C with other components in the circuit.
Tuned circuit: a second-order circuit containing one inductance and one capacitance, at least one
of which is adjustable to reach a condition of near resonance.
Undamped natural frequency: the natural fi-equency of a circuit consisting of lossless inductors
and capacitors. For the case of one inductor and one capacitor connected together, this
frequency is the same as the LC resonance frequency or the tank firequency and is equal
to m J l c .
Universal resonance curve: a normalized magnitude response curve of a band-pass transfer func
tion having one pair of complex poles and a single zero at the origin. The magnitude is
normalized with respect to the maximum gain, and the frequency is normalized with
respect to the center frequency.
Upper half-power frequency, (Oj: the radian frequency above the center frequency at which the
magnitude response is 0.707 times the maximum value.
r> .
n
o
* W e would like to thank Les Axelrod o f the Illinois Institute o f Technology for providing derivations that led to
this formula and those o f the bandwidth and half-power frequencies.
863
C H EC K ; poles are p, 2 =
'Problems
5 . Suppose a basic band-pass transfer function
BASIC BAND-PASS TRANSFER

FUNCTION PROBLEMS
has
= V s 200 rad/sec,
(a)
(b)
= ^ p
i+ -
4G,-;
= 10, and
8 rad/sec.
1. Fill in the details o f the derivation of
0) 1.2 = a^; +
Find H{s), its poles, and (X.

Find approximate values For cOj and
(O2. Using MATLAB, plot the magni
tude response and verify your calcula
22,
tions.
from
, find a parallel
N(Jco) = K ^
T
i j o i ) + j l w o p+CO p
K
=
2 o -h y
/ 2 _ 2\ circuit realization o f H{s).

I UJ
uJ p \ _____ _
C H EC K : Q^= 11.32
transfer function o f equation 16.1 has poles at
6 . Suppose a series RLC circuit has

= 80
rad/sec,
= 8 , C = 1/80 F, and output l^{s)
-1 j5 and the gain at
with voltage source input
2. Suppose the basic second-order band-pass

is 2 . Find K, H{s),
^1
^2quency response using MATLAB and verify
(a)
Write down the transfer fijnction for the
your answers.
(b)
Find the values o f R and L.
C H EC K : cOj = 2.831 rad/sec
(c)
Find
3. Suppose the basic second-order band-pass
(d)
series RLC circuit in terms o f R, L, and C.
and approximate values for
OJj and 0)^.
answers to part (c).
- 8 0 y l5 9 9 and the gain at co^ is 6.25. Find
K, H{s)y
Q,
exact and approximate
Use SPIC E to obtain the magnitude

(frequency) response and verify your
transfer function o f equation 16.1 has poles at
C H EC K : /?= 0.125 n
values for tUj and Wj, and the relative percent

error in the approximate computation o f CO2.
7. Suppose a series RLC circuit has
= 100
Using MATLAB, plot the magnitude response
rad/sec,
and verify your calculations.
with voltage source input

(a)
Write down the transfer function for
C H EC K : t02 = 1681 rad/sec while the approx
= 4, C = 0.01 F, and output
the series RLC circuit in terms o f R, L,
imate value o f CO2 is 1679 rad/sec
and C
Find the values o f R and L.
4. Suppose a basic band-pass transfer function has
(b)
= 10, and B = 2 rad/sec.
(c)
Find
(d)
cOj and CO2.

to
= V S 2 rad/sec,
(a)
(b)
Find H{s) and its poles.

Find tOj and CO2. Using MATLAB,
plot the magnitude response and veri
fy' your calculations.
Vguri-^) , find a parallel
(d)
m(^)
circuit realization o f M{s).
frequency response and compare it
with your answer in part (b).
and approximate values for
(frequenqO response and verif}' your

answers to part (c).
C H EC K : I = 0.01 H and
=4 S
8 . Consider the circuit o f Figure P I 6.8 for

which R^ = 40 k n , R, = 10
and C = 1 liF.
/ = 10 mH,
S64
(a)
Find the exact value o f the maximum
P I 6.10, w h e r e = 2 mS,
voltage gain and the corresponding
= 10 \dl.
frequenc)' (in Hz).

(b)
Find the exact 3 dB bandwidth
(in
100 Q, and
(a)
Find H{s) in literal form.
(b)
If A = 100 u H ,^ = 1040 kHz, and Bjr

= 10 kHz, find C, R, and
Hz).
(c)
(c)
Find the circuit Q.
(d)
Find approximate values o f the upper
If^^^ = 920 kHz, Bj= 10 kHz, and C

= 250 pF, find A, /c, and
(cOo) and lower ((o,) half-power fre

quencies (in Hz).
(e)
Use iVIATLAB or its equivalent to v

plot
(0
|//(_/0))|
vs.
CO.
What is the new bandwidth if the

input is changed from
to an inde
pendent current source

(g)
If the circuit is frequency scaled by

= 10, find the new values of
to-,,
and (Oj.
AN SW ERS; ui) 234.2 pi-, 6~.96 k ii; (b) 119 7

pH, 6 3 .66 k n
11. For this problem consider the circuit o f
Figure P 1 6 .1 1, in which A = 300 uH. The fol

lowing additional ideal components are avail
able;
Variable capacitors (in pF); 2 0 -2 0 0 , 3 0 -2 0 0 ,
Figure P I6.8
3 0 -3 0 0 , 4 0 -4 0 0
ANSWFR^.: (a) 0.20 at 1591 Hz; (b) 19.) Hz;
(c) 1601 M Hz. I 5 8 1.6 11/.; (d) 80; (0 1S.) 1 Hz
Resistors; all values

(a)
Using these components, design a par

allel resonant circuit such that;
(i)
9. I'he parallel /ÂC circuit o f Figure 1M6.9 has

to^^^ = 1 Mrad/sec,
The circuit can be tuned from

550 to 1650 kHz (standard AivI
= 20 kH, and
broadcast band).
|Z(0.9(o^^^)| = 10 k n . Find the transfer func

tion, M{s) = Z{s), R, A, C, B^^, Q, and approxi
(ii) When the circuit is tuned to 920

kHz (WBAA, at Purdue), the
mate values for to, and (1)2.
bandwidth is 20 kHz.
(b)
With the components as selected in

part (a), find the bandwidth when the
ZIjco)
> R
circuit is tuned to the low end and

then the high end o f the AM broad
cast band.
Figure P I6.9
AN SW ERS; 20 kLl. 2.438 mH, ^10.2 pF.
121.9 krad/sec. 8.2
10. 'Fhe equivalent circuit o f a radio frequenq-
Figure P 16.11
AN SW ERS; 3 0 -3 0 0 pF. 9 .'8 kLl, 7.1 6 kHz.
amplifier in an AM receiver is shown in Figure
64.2 kHz
865
12. For this problem again consider the circuit

o f Figure P 16.11. The following ideal compo
Z(s)
nents are available:
Variable capacitor (in pF): 3 0 -3 0 0

Figure P i6.13
Resistors: all values

(a)
Using these components, find the
A N SW ER: (a) 2.S mH .uui 530 12
range o f allowable inductance so that

the circuit can be tuned from 550 to
14.
1650 kHz (standard AM broadcast
P 16.14, H{s) = Z{s), the peak f r e q u e n c y = 10
band) as follows:
kFIz, and the bandwidth B^ = 3 kHz. If C = 0.1
(i)
For the two-terminal circuit in Figure
Find the largest value o f L
l-iF, find the corresponding values o f L, R, and
allowable. Hint: For the smallest
the circuit Q.
value of C, the inductor must be
chosen so that the circuit can be
TYYY
R
tuned to 1650 kHz.

Z(s)
(ii) Find the smallest value o f Z.

allowable. Hint: For the largest
value o f C, the inductor must be
Figure P i6.14
chosen so that the circuit can be
C H EC K : Q = 3.333
tuned to 550 kHz.

(b)
(c)
For
find the range o f capaci
tance utilized in the tuning.
15. Consider the circuit o f Figure P i 6.15, for
For
which R^ = 40 kQ, Rj = \0 kH, /. = 10 mH,
find the range o f capaci
and C = 1 uF.
tance utilized in the tuning.
(a)
kHz
(b)
Find (o,,
(WBAA, Purdue), the bandwidth is
(c)
C H E C K : 30 pF s C s 270 pF
(d)
For L = 295 pH, find R so that when

the circuit is tuned to 920
With the components as selected as in
(d)
(the corresponding
rad/sec and Hz.
the circuit is tuned to the low end and

(e)
Use MATLAB or the equivalent to

plot |//(yto)| vs. (I).
cast band?
What is the new bandwidth if the
(f)
input is changed from

to an inde
pendent current source / ?
13. For the rwo-terminal parallel /?Z,Ccircuit in

Figure P i 6.13, H{s) = Z{s), the peak frequency
10 kHz, and the bandwidth Z^y= 3 kHz. If
C = 0.1 uF, find the corresponding values o f L,
R, and the circuit Q.
and

and lower half-power frequencies in
part (d), what is the bandwidth when

then to the high end o f the AM broad
and
quantities, in Hz, to
20 kHz.
(e)
Find
Q, and
6
Figure P i 6 .1 5
866
peak frequency (in Hz).
AN SW ERS: O),,, = 10^' and Q = 80

(c)
16. Again consider the circuit o f Figure P 16.15.
Now suppose
= 100
Q = 10, and^^ = 10
kHz.
Find the exact 3 dB bandwidth (in

Hz).
(d)
Find Q o f the circuit.
(e)

and lower half-power frequencies.
(a)
If /?^ = 00, find C and L.
(b)
Now suppose Rj^ = 1000 Q. Again
(f)
Plot the magnitude response using

find C and L.
C H EC K : (a) C = 1.59 uF
C H EC K : (e) 9.9 and
1 0 .1
krad/sec.
17. For the series resonant circuit shown in
20.
For the circuit in Figure P I 6 . 2 0
Figure P16.17, /?, = 40 ^ Z- = 0.8 H, C= 1.25
Z. =
10
uF, and R j= 160 Q.
op amp is ideal.
(a)
niH, C =
0 .1
mF,
100
kQ,
and the
Construct the transfer function H{s) =
(a)
(b)
Find the exact value o f the maxirnum
= -Zj(s)IZ,,{s) in terms
o f the circuit elements R-^^, Rp L, and
C, and put it in the general form
peak frequency (in Hz).;

(c)
Find the exact 3 dB bandwidth (in
s ~ \-la p S + {H~p
Hz).
(b)
Find the values o f K, 0)_, the circuit
(c)
Q, and
= |M;co,)|.
Compute the value o f the half-power
(d)
Find the exact values o f the upper and
(e)
Find Q o f the circuit.
(0
Plot the magnitude response using
bandwidth
quencies
lower half-power frequencies.
m
R.
CO,
and the half-powcr fre

and oyj. (Approximate
values are acceptable.)
(d)
Sketch the pole-zero diagram that rep

resents the circuit, and note the exact
locations o f all the poles and zeros.
(e)
the magnitude o f Vg^^f{t) in steady
Figure P i6.17
state.
A N SW ERS: (b) 0.8 at 159.15 Hz: (c) 39.79

Hz; (d) 180.29, 140.51 Hz; (c) 4
18. Repeat Problem 17 for /?, = 80
Z = 0.25
= 3200 rad/sec.
19. Reconsider the series circuit shown in

Figure P16.17. Let /^, ^ A iX L = 0.1 H, C =
0.1 liF, and /?2 = 16 Q.
(a)
(b)
(0
Use SPIC E or the equivalent to gener

ate the magnitude response plot for 0
< / < 500 Hz.
H , C = 0 . 5 u F , and/^2 = -'^20
C H EC K :
If v-J^t) = lOOsin(lO^f) mV, determine
Find the exact value o f the maximum

R,
867
C H EC K : Q = 100
circuit should be 0.1 uF. Hint: Follow the pro

cedure described in Example 16.5. Use SPIC E
21. For the circuit shown in Figure P i 6.21:

(a)
or its equivalent to generate the magnitude
I (^) response plot for 900 < / < 1200 Hz.
Find the transfer flinCTion H {s) = --------
(b)
(c)
Now supposey^^ = 200 Hz and Br= 20

Hz are desired. One has available a 1
H inductor, a 10 jiF capacitor, and
arbitrar) resistors.
(i)
Determine the necessary value o f (3.
(ii) Determine the value o f R .
ANSWER: (b) w, -
f r
V LC '
Figure P I6.23
24. For the circuit transfer function o f Problem
23,
''out
+
Figure P I6.21
s +^2^2/
22. For the circuit shown in Figure P I6.22:

(a)
(b)
Find
(c)
Now supposey^^^ = 200 Hz and Bj-=20
and Q.
^ 1^ 2Q ^ 2
show' that
(i) H. =
C,
Co
-hi
and (ii)
Q- ..c,
R'
^2
Hi
C,
Hz are desired. One has available a 1
Hence, for high-Q transfer functions, this cir
H inductor, a 10 jxF capacitor, and
cuit is undesirable.
arbitrary resistors.
(i)
Determine the necessary value o f j.i.
(ii) Determine the value o f R.
25. Reconsider the active band-pass circuit o f

Figure P16.23. The filter is to pick out the
midrange o f a typical audio speaker, in which
case/; = 500 Hz ,/2 = 3200 Hz, a n d ^ = 1265
Hz. The gain
is to be 10. Find Bj'd.nd the
circuit parameters. Hint: Follow the procedure

described in Example 16.5. Use SPIC E or its
equivalent to generate the magnitude response
plot for 1 < / < 5000 Hz.
Figure P I6.22
C H E C K : R = 795.77
C H EC K : Q < 1 and R^ > 2, slightly

26. Consider the band-pass circuit o f Figure
P I 6.26, which contains two op amps rather
23. Design the active band-pass circuit in

Figure P I 6.23 to have^^^^ = 1000 Hz and Bj- =
12.5 Hz. The final value o f Cj for the actual
than the one in Problem 25(a)

circuit is
86S
28. Consider the circuit in Figure P i 6.28.
2G:
H{s) =
Model of Coil
Hints; (i) Use the properties o f ideal op amps,

'
(ii) Use voltage division across Gy (iii) Write
two nodal equations and solve with unknowns

K_...and
K,.
out
(b)
(c)
Figure P i6.28
Design an active band-pass circuit to

have^^^ = 1000 Hz and B^= 12.5 Hz.
(a)
Find Hîs) = 7^/1/and H^{s) =
The final value o f C should be 0.1 jdF.
(b)
For //j W, find the exact values of
(c)
For
Compare the resulting resistor values

with those computed in Problem 25.
^(0'
Q-consider the case where the
ac voltage source has fixed
and
CO,
but the capacitance C is adjustable.

Find the exact value o f C (in terms o f
R, L, and co) such that I i s
maxi
mized. Show that if the coil has a high

Q, then
This result provides a practical way o f measuring
It also provides a practical means for
generating a very high short-duration voltage

from a relatively small voltage; this has applica
tion in ignition circuits.
27. Consider again the band-pass circuit of

Figure P I6.26. In Problem 16.26 it was shown
that
2^3
H{s) =
C
J9 i\
\c)
(a)
Design an active band-pass circuit to

have peak frequency
and bandwidth
= 1000 rad/sec
= 100 rad/sec. The
final value of C should be 1 Ii/".

(b)
Compare the resulting resistor values

with those computed in Example
16.5.
BAND-PASS CIRCUITS WITH

PRACTICAL COMPONENTS
29. Consider the /?/.Ccircuit in Figure P16.29,
in which 7?^ = 50
Z, = 1/9 H,
= 0.08 Q,
F
and C = 1/9 F.
1
(a)
Find the coil Q at Wg =
(b)
Find an approximate parallel represen

tation o f the coil near (Oq .
(c)
Find the circuit Q.
(d)
If the circuit is high Q, compute approx
(e)
Suppose
= 2cos(0),^/) A. Find,
approximately,
(in volts) in steady
imate values for
state.
y fiC '
CO,, and CO2.
869
A N SW ER S: (a) 12.S; (c) 10; (cl)
20
(a)
(b)
Find the capacitor Q at Wq = j-----.

Find an approximate series
representation o f the capacitor near ( jO q .
(c)
(d)
Find the approximate value o f Q.

If the circuit is high Q, compute
approximate values for
0)j, and CO2.
CYY\
L
Coil
Figure P I6.29
= 3750 Q., L =
=2

0.1 mH, 7?^ = 4 a
Capacitor
C = 10 nF, and
cos(oj^^^r) mA.
Figure P I 6.33
C H EC K ; Q = 15 and
= 3 cos(co^^^^) V
31. Consider the RLC circuit in Figure P 16.31,

in which
L = 0.5 mH,
= 20
Q., and C = 0.5 nF.
= 20
^
C H EC K ; CO, = 9975 rad/sec
34 . Consider the circuit o f Figure P I 6.34.

Suppose R^ = 0.6 O., L = \ mH, R^- = 2.5 kT2,
and C = 0.1
------V LC
(a)
Find the coil Q at wq =
(b)
Find an approximate parallel represen

tation o f the coil near cOq.
Find the approximate circuit Q.
(d)
(b)
Find an approximate series
-------
representation o f the capacitor near

(d)
o jq

(Oj, and to^.
o jj, and t02AN SW ERS; (a) 12.5; (c) 10; (d)
Find the capacitor Q at (Oq -
(c)
(c)
l i F.
(a)
= 20
c o s ((t) / ) \ '
R.
' 6
Figure P I6.34
C H EC K ;
100
Figure P I6.31
3 5 . Repeat Problem 34 for R^ = 0.32 D., L = \6

32. Repeat Problem 31 for R^ = 40 k ii, L = 4
mH,
= 100 a , and C = 0.25 nE
C H EC K ; 1 >
l i H , /?c= 800 a , and C = 0.25 uE

C H EC K ;
= 20
> 0.5
36. Consider the circuit of Figure P i 6.36.
33. Consider the circuit o f Figure P i 6.33.
Suppose L = 0.5 mH, C = 1.25 uF> Q_i^ = 20,
Suppose R^=2 k n , L = 1 mH, and C = 10 tiF.
and
= 80.
870
Chapter 16 <>Band-Pass Circuits and Resonance
r>
(a)
Find
(b)
Convert the capacitor to a series
and
model at
model at (Oq. Denote the resulting
oO
q.
Denote the resulting
(c)
series resistance by
1 series
(d)

/\
(e)
Lossy Inductor

co^^,
cOpandcDj.
n ,0icap
-r I
n
^cotl^c
Verify that Qdr
Q coil
(c)
(d)
series resistance by RpamFind the approximate circuit Q.
Qcap
0)p and (02-
(e) Verify that

C H EC K :
= 20
, O)^,
q
+ Qcap
40. (Design) Consider once more the circuit of
Lossy Capacitor
Figure P I 6.38. Recall that for high-Qcompo-
n
n
^coil^cap
~n
-I- n
WjcoiY ^cap
;c
Suppose you have been asked by your supervi

sor to use this circuit in a band-pass design with
f = ^ M Hz and a bandwidth o f 20 kHz and a
Figure P I6.36
lossy capacitor with C = 2 0 0 pF and having
C H EC K : Q ,= 16
37. Again consider the circuit o f Figure PI 6 .36.
(a)
Determine the circuit Q, i.e.,
Verify that for high
(b)
Find the necessary
and high
at cOq,
to achieve the
desired
Q coilQ cap
(c) F
P16.38.
38. Now consider the circuit o f Figure P I 6.38.

Verify that for high
and high
ind the inductance L o f the lossy coil

and then find R^ as shown in Figure
Qcoil + Qcap
at (Oq,
*(d)
Suppose we now desire to double the

bandwidth by adding a resistor, R^g^^
QcoilQcap
a
Q coil
Qcap
in parallel with the lossy coil (i.e., in

parallel with the current source).
W hat is the proper value of R^source'
^ ^
Lossy Inductor
C H EC K : (c) R^<SQ.
41. Repeat Problem40 under the conditions'^
= 1.6 MHz, a bandwidth o f 3 0 kHz, a lossy
capacitor with C = 100 pF, and
= 200.
42. (Design) Consider the circuit o f Figure

Figure P I6.38
P I 6.42. Recall that for high-Q components,

^
3 9 . Consider again the circuit o f Figure

P I 6.38. Suppose L = 0.5 m H , C = 1.25 jiF,
= 30, and
= 60 at cDq.
(a) Find R^ and R^.
(b)
Conven the inductor to a parallel
QcoilQcap
Qdr =
.f /
^coil **^cap
Suppose you have been asked by your supervi
sor to use this circuit in a band-pass design with
= 1 MHz, a bandwidth o f 2 0 kHz, and a
8 71
lossy inductor I = 100 uH having Q^gj(= 100.

Determine the circuit Q, i.e., Q^j^.
(a)
Find the necessary
to achieve the
(b)
and the corresponding frequency (in Hz),
desired
(c)
Find the capacitance C o f the lossy

capacitor and then find
(d)
Check your results by doing a SPICE (or

equivalent) simulation of your circuit.
as given
in Figure P I 6.42.
(d)
Suppose we now desire to double the

bandwidth by adding a resistor,
^source
series with the lossy coil (i.e.,

in series with the voltage source).
What is the proper value of

Lossy Inductor
amplifier
Lossy Capacitor
coil
capacitor
Figure P I6.44
A N SW ERS: (a) 50; (b) 1 mH, 50 kQ; (c)
6 3 6 .6 Hz. 10, 15.91 kHz
:C
45. Repeat Problem 44 for the circuit shown in

Figure P I 6.45, where /?, = 500 Q.,
= 2 kQ,
R^ = 5 k n , R^ = 2Q., L = \ mH, C = 1 uH, and
a = 25.
Figure P I6.42
C H E C K : (c) 130 k n > /?^ > 115 kQ
43. (a)
Repeat Problem 42 (except for part

(d)) under the conditions
= 1.6
MHz, a bandwidth o f 25 kHz, and

a lossy inductor Z, = 100 j.iH hav
(b)
ing ^ oii= ^00.

Now suppose you were only able
to purchase a capacitor having
Figure P I6.45
= 200; you must achieve the

required bandwidth by adding a resis
C H E C K : 650 Hz > Bj- > 630 Hz
tor in parallel with the lossy capacitor.

What is the value o f this new parallel
resistance?
C H EC K :
46. Consider the circuit in Figure P i 6.46.

(a)
Find the input admittance o f the cir

cuit, ^y(^), and compute its poles and
= 64
zeros.
44 Consider the amplifier circuit shown in
Figure P I 6.44.
Find Q o f the coil at to = HP rad/sec.
(a)
Represent the coil by a parallel RL cir
(b)
cuit that is valid for frequencies near
10^ rad/sec using column 3 o f Table
(c)
(b)
Determine
and the approximate 3
dB bandwidth and half-power fre

(c)
quencies o f the circuit.

Determine
and
at cOq. Then
use column 3 o f Table 16.2 to find the
approximate bandwidth and half-
16.2.
power frequencies. Compare the result
Find approximate values of the 3 dB
with that obtained in part (b).
bandwidth.
8 " 2
(cl)
Cheek your results by doing a SPICE (or

I. r~
______
10.000 rad/scc.
= 10^ rad/sec. (O, a
H).
)S0.0()0
rad/sec. (u-, s 1,050,000 rad/sec; (c)
coii
0.01 H
A N SW ERS: (b)
(i^ s
equiv;ilent) simulation of your circuit.
0.1 sin(oj,-r) V
750
10F
3000
RESONANT CIRCUITS WITH

APPLICATIONS
SOOkO
48.
capacitor
For each two-terminal circuit in Figure
P I6.48, the resonant frequency/^ = 10 kHz.

If C = 0.1 iiF and R = 1 kH, find the corre
Figure P I6.46
sponding values o f L. For each circuit, deter
47. Consider the circuit in Figure P i6.47, which
mine the input impedance at resonance.
contains a non-ideal capacitor, a non-ideal induc

tor, and a meter to measure the current response,
Z(j(o)
iofU)- The 1 Q resistor representing the meter is

a precision resistor. The voltage across the resistor,
equals the current through the resistor.
(a)
'Fhus, a practical way o f measuring current is by
measuring the voltage across a small resistance in
TY"YV
R
the circuit. If the resistance o f the meter is suffi
Z(jto)
ciently small, it should have little effect on the

behavior o f the circuit. Nevertheless, in analyzing
the circuit, account must be taken o f the resist
(b)
ance o f the meter.

(a)
Use the approximation techniques o f

column 3 ofTiible 16.2 to develop an
Figure P i6.48
AN SW ERS: (a) 2.S niH; (b) 2.5 mi l
approximating series RLC circuit for

the given circuit.
49. For each circuit in Figure P i 6.49,
(b)
Compute approximate values for
(c)
Q,
to,, and co,.
At resonance, fmd the approximate
(a)
L, and C. Verif}' that

10^ = iOq. (This always follows when L
and C are either in parallel or in
steady-state current response,

when
the
input
voltage
is
10
sin(to^r)(r).
(d)
series.)
(b)
Check your results by doing a SPICE (or
that the input impedance is independ

ent of L and C at co^.. (Again, this
inductor
^
Now find the input impedance at to^

in terms o f /?,, R-,, L, and C. Verif)'
equivalent) simulation of your circuit.

m eter
Find the resonant frequency', co^, in

terms o f /?j,
capacitor
always follow's when L and C are either

in parallel or in series.)
(c)
In the case o f Figure P I6.49a, find
Vo,{^) in steady state in terms o f /^,,

R^, and
if v-^^{t) =
sin(w/). In
the case o f Figure Pi 6.49b, find /,^,(/)
in steady state if
Figure P i 6 .4 7
sin(co^).
shown, find the value o f L to maxi

mize the voltage gain
/
Vy^^(/to)| at/= 10 MHz. Verify that the
+
Z(jw)
amplifier gain at 10 M Hz is greatly

increased.
(c)
(a)
fyy\L
(!)
Plot the frequency responses o f the cir

cuit over 1 M H z < / < 10 M Hz with L
and without L.
i (t)
Z(j(o)
(b)
Figure P I6.49
amplifier
50. Figure P I 6.5 0 displays an amplifier

model containingo a V CC S with ofn
?, = 4 mS
(milli-siemens) and R/ = 40 kf2. Suppose the
applied sinusoidal voltage,
has a mag
nitude o f 0.2 V at 20 MHz.

(a)
W ith the load connected directly as
load
Figure P i6.51
52. Find the resonant frequency (O^., in rad/sec,
and Z{jio) o f the circuit in Figure P i6.52 for R
= 2.8
C= 0.2 mF, and L = 20 mH. Verify that
shown (without L), find the magni
lie '
tude o f the output voltage.

(b)
If an inductance L is connected across
Now compute
the load to tune out the effect o f the
and phase responses from 0.75co^ to
capacitance, find the value o f L and
Verify that at co^, the phase angle o f
the resulting
zero.
that will show that
the amplifier gain at 10 MHz is great
and plot the magnitude

is
R
(t)
ly Increased.
1 . 2 5 c d ^.
Figure P I6.52
AN SW ERS: 480 rad/sec, 35.71 LI
Figure P I6.50
53. Consider the circuit o f Figure PI 6.53. The
51. Figure P I 6 .5 1 displays an amplifier model

containing a V C C S with
= 4 mS (milli-
matching network when the values are proper
CL part o f the circuit can be thought o f as a
siemens) and Rj^ = 20 k^2. Observe that the
ly chosen. In case 1 we will see that maximum

power transfer is achieved, wherea.s in case 2 it
amplifier has an input capacitance of 0.2 nF.
is not.
The inductor L is used to tune out this effect.

(a)
(b)
(a)
Find the resonant frequenc\\
Without L find the magnitude o f the
for the cases
voltage gain at 10 MHz.
(i)
If an inductance L is connected as
(ii) /^^ = 6 0 2
Rj =
80
LI
and
874
(b)
Now suppose
= 250
with co =
0)^.. Then compute the average power
/Y Y V
delivered to Ri for each o f the cases in

Z(j(o)
part (a).
0.1 H
125Q
/Y Y V
Figure P i6.55
V cos(o)t)
)
Z(jw)
IO mF
56. Find the resonant frequenc)'
in rad/sec,
and Z(Jo}j-) o f the circuit in Figure Pi 6.56 for R =

12.5 Q, C= 0.2 mF, and L = 20 mH. Verify that
Figure P I6.53
54. For the circuit o f Figure P I 6.54, R = 800
Q, L = 0.2 H, and C = 0.25 liF. For the two

cases (i) Rj = 1 k fl and (ii)
= 800
find co^
and at resonance, find
the voltage across
y jlc '
Now compute Z-^^{s) and plot the magnitude
and phase responses from 0.5w^ to 2to^. Verify
that at (0^, the phase angle o f
the capacitor in the steady state, due to the

input
/Y Y V
= 10 cos(o);-f) niA. Finally, find the
average power delivered to R^ in each case.
is zero.
VJj(o)
Z(j(o)
Figure P i6.56
57. Consider the circuit o f Figure Pi 6.57 for R^
= 80 a , R^ = 125 n , C = 2 uF, and Z. = 20 mH.
The LC part o f the circuit can be thought o f as
a matching network when the values are prop
erly chosen.
(a)
Figure P I6.54
ANSWHR: Case I: 201)0 rad/sec, ^ cos(2000/)
V, 12.5 m\V
Find
the
resonant
frequency
CD^,
Z,(;cop, and Zijia).

(b)
Now suppose
= 250 V^^ with ti) =
Compute the average power deliv

55. This problem develops formulas similar to
ered to
equations 16.41 and 16.42 for the
more power to
circuit
by varying the val
ues o f L and C?
o f Figure P i 6.55.
(a)
Is it possible to deliver
Prove that, for the circuit shown in

Figure P I 6.55,
:..................... -
^
- J
LC
I '------L
Z(jco)
c
<
CR-
and
Z,(jw)
Z{j(0,.) =
Figure P i6.57
CR
58. A two-terminal network has input imped
ance
61. This problem uses an LC coupling network
s~ + 4 s+ S
= 1, find co^ and
to maximize power to a load. The LC coupling

network always has a series connection for L\
the parallel capacitor is always closest to the
(a)
If
(b)
Find the range o f /7 > 0 for which the
larger resistance in this case the load resist
impedance has a real resonant fre-
ance. Problem 62 will consider the general case.
quenc)'.
Suppose the voltage soQrce in Figure P I6.61

has value
= 1 0 0 V 2 cos(27i x lO^r) V.
ANSWHRS: (a) 2 rad/scc and 0.25 i h (b) 0 <
Compute the values o f L and C such that the
a <1
average power delivered to the load resistance
RtL is maximized. What is Pttuix

59. For the circuit shown in Figure P I 6.59,
where
.
205
y i s ) = --------- ,
5 - + 100
find the value o f C that makes the circuit reso
nant at
CO
/Y Y \
son
v jt )
300 0
= 30 rad/sec and then find the value
ofr.,(/-30).
Figure P i6.61
OY(s)
AN SW FRS: P .9 uH. 1,186 pF 50 W

Y(s)
62. Equations 16.41 through 16.44, or those

derived in Problem 55, can be combined to
produce a set o f design formulas for a lossless
Figure P i6.59
network that matches two unequal resistances

at a single frequenc)^ The matching network
60. This problem uses an LC coupling network

to maximize power to a load. The LC coupling
consists o f only one capacitance C and one

inductance L, as shown in Figure P I6.62. At a
network always has a series connection for L\
specified frequency OJ, it is desired to have
the parallel capacitor is always closest to the
matching at both ends, i.e.,
larger resistance in this case the source resist

ance. Problem 62 will consider the general case.
Z , (yo)) = /?, + /O and Zjijiyit) = R2 + P
Suppose the voltage source in Figure P I 6.60

has value
= 100V 2cos(27t x 10^/) V.
Compute the values o f L and C such that the

average power delivered to the load resistance
Let R^jîi denote the smaller o f (/?p /?-,) and

the larger. Prove that
(a)
is maximized. What is
Cshould be connected in parallel with
^large'
^ should be connected
between the rsvo top terminals and
(b)
thus in series with R^j^nThe element values are given by
large
Rsmall
and
Figure P I6.60
A N SW l-RS: 17.79 uH, 1.186 pF, 8.3.^33 W
C=
(oRla/ge
Rsmall
8:^6
AN SW ERS: 12.5 V I /S = 21.65 rad/s cfc 22.2 Q.

5\/21 = 22.91 rad/s. 4.44 Q.
BAND-PASS TRANSFER
FUNCTIONS WITH NO ZEROS
OR A SINGLE ZERO OFF THE
ORIGIN
Figure P i6.62
63. The purpose o f this problem is to show that
the resonant frequenc)' (0^ depends on the

(a)
choice o f the input terminals and that (0^^ (Oq

= 0.25
His).
(b)
I f / ? = 5 a , Z = 0.1 H ,an d C = 10|iF,
(c)
Compute
rad/sec. Find the resonant frequency (0^ if the

input is connected across
(a)
the
output, compute the transfer function
in general. Consider the circuit o f Figure

P 1 6 .6 3 , which has Wq = l/ V LC
With V^{s) as the input and
verify that the circuit is high Q^.

exactly and verify that
A and B
0)^^ = CO^. Then compute the maxi
(b)
B and C
mum gain.
(c)
A and C
(d)
40 n
With the values given in part (b),

compute approximate values for
10Q
CO,, and CO2.

'0 .8 H
2m F
Figure P I6.63
ANSWURS: 2S, I2.5\/3 = 2 1 . 2 8 . 8 6 . all in
rad/sfc
64.
The
circuit
= l/ 'Tl C
of
Figure
Pi 6.64
has
(a)
depends on the choice ot the input terminals and

03^ in general. Find the resonant frc-
quenc)' 03^ and the equi\'alcnt impedance seen at

(a)
E and F
D and E
as the input and I^{s) the
His).
(b)
I f / ? = 4 k a Z . = 0.1 H ,an d C = 10).lF,

verify that the circuit is high
(c)
Compute (0^^^ exactly and verify that

= CO^. Then compute the maxi
the terminals if the source is connected across

(b)
With
output, compute the transfer function
= 25 rad/sec. The purpose of this
problem is to show that the resonant frequency (0,.

that (0^
Figure P I6.65
mum gain, H^^j.

(d)
With the values given in part (b),

compute approximate values for
100
CO,,
and 0) 2-
. 0.8 H
2m F
50 0 /
Figure PI 6.64
F igu re P I 6 .6 6
67. Consider the Sallen and Key active network

of Figure P I6.67, which realizes the transfer
function
H(s)=
f
UiC
____1
H{s) =
LC
]
+
/?2Q/
s+
=
Recall that cOq = lly J h C . By the use of
(b)
equation 16.49, show that
= 50
kfl.
1 --
Compute (O^ and H^.
(c)
Find approximate values of
1-
0)j,
4lc\
and (Oj.
R^C^+s
\
2LC
c2 +
4 .
Plot the magnitude response curve of

the transfer fimction.
(e)
2LC
Compute O) , Qp, and K of equation

16.48.
(b)
(d)
LC
/?l/?2C|C2
Suppose Cj = 0 .2 |jiF, C2 = 0.5 nP,

(a)
S +
/YY\
Simulate the circuit using SPICE.
o
+
Figure P I6.69
70. Consider a transfer function with a single
zero off the origin, i.e.,
Figure P I6.67
s+2OpS+(0p
H{s) = K - 2
------j -
68. Consider the Sallen and Key circuit of

Problem 67. Suppose Cj = 50 nF, Cj = 3.125
nF,
= y?2 = 8 1 ^ .
(a)
Compute O) , Qp, and K o f equation
(a)
Derive equation 16.54, i.e.,
o ) =
16.48. Verify that the circuit is not

high Q.
(b)
Compute the exact values o f (O^ and
Hint: Instead o f maximizing |//(/(o)|, try to

maximize l//(/to)p, and consider co^ as the
(c)
m
Find the exact values o f
0)o.
(d)
Plot the magnitude response curve of
o)j, and
the transfer fimction.

69. For the circuit shown in Figure P I 6.69:
(a) Show that the transfer function is
independent variable.
(b)
Derive equation 16.55, i.e.,
0) (for zero phase shift) = ^(Op - 2opO
Hint: Write \/H(s) as [{l/K +

Im{/^(/0))} = 0, and solve for co.
set
878
73. The y?/.C circuit o f Figure PI 6.71 has trans-
P 16.71.
fer function
(a)
\/
H {s) = Zi{s) = ------- ^

'
s+
H {s) = Z;As) =
L}
[X pC
+ ''S s + 11+
L
\
LC
D \ 1
Rf,C
(b)
R.
s+
L)
Let /?. = 100 k^2, L = 0.225 H, R, = 1000 Q,

s + l+^.s_
RPI LC and = 0.5 [.iF.
(a)
By the use o f equation 16.54 or part
Compute to and verify that the cir

cuit is low Q^.
(a) o f Problem 65, show that
(b)
Compute
is quite different from

1
0).,, =
(c)
1+-
Plot the frequency response for 0 < to

< 40 0 0 rad/sec. Compute the maxi
R r
mum gain and verify the result from

the plot. Also compute the zero o f the
transfer function as -Zy
(d)
Define the coefficients o f the numera

tor and denominator o f your transfer
function in MATLAB as n = [1 zl]/C
and d = [1 ? ?]. Now use the com
mands below to compute the impulse
Coil
and step responses o f your transfer

Figure P I6.71
function.
72. The RLC circuit o f Figure P 16.71 has trans
sys=TF(n,d)
fer function
impulsc(sys)
pause
.
s '-
(a)
step(sys)
L}
H {s) = Z U s) =
(
-I-
1_
AN 'SW FRS: wp = 2 .9 9 6 3 e+ 0 3 ,
s+
KjLC
If /?^ = 50 a , Z- = 119 H, R^ = 0.08 2,

and C = 1/9 F, verify that the circuit is
high
(b)
and that a
Compute
to^.
to^^ a (i)^. Then, compute the maxi

mum gam.
(c)
Compute approximate values for
(d)
tOj, and C02Suppose diat at to^, |/,(/toJ | = 1 A.

Determine, approximately, the value
ANSW'HRS: (b) to,,^ = 9.0071 rad / sec; (d) 10

\'
6.7115C -01, 7.1 = 4 .4 4 4 4 c+03, wm =

1.0l28e+ 03
74. Again consider the circuit o f Figure P 16.71,

in which Z, = 1 mH. This inductor has
40 at 100 kHz. The magnitude response is to

have a peak at
s 100 kHz.
(a)
Specify the value o f the capacitor C.
(b)
Specify the value o f Rp so that the

bandwidth is approximately 10 kHz.
(c)
What is
A N SW ERS: 2533 pF, 83'8 Q, 10
8~9
element values and repeat parts (c)
75. Consider the circuit in Figure P i 6.75. Let

= 0.08 Q, R c = 0.02 a
Z. = 1 H, C = 1 F,
and (d) for the scaled circuit.
and Rs = 40 ^2.
(a)
R, = 0.05 n
o u r ''" ' nt'
(b)
(c)
Find
and approximate answers for
O btain
/Y Y V
" P 1 2magnitude frequency
,0
response plot to graphically verify
R, = 50 O
L=1 H
C =1 F
R, = 0.05 n
your answers.
Figure P I6.76
R
MISCELLANEOUS
77. Consider the idealized (tank) circuit o f
' 6
Figure P I 6,77. The moment the inductor cur

rent passes through zero with positive slope is
taken as the reference point, r = 0. At this time
instant the capacitance voltage is E volts.
Find V(\t) and i^{t) for / > 0 by the
Figure P i6.75
(a)
76. The analysis o f the non-series-paraliel cir
(b)
Find the energy stored in C as a func

tion o f t.
(c )
Find the energy stored in L as a func

tion o f t.
(d)
Show that the total energy stored in
Laplace transform method.

cuit shown in Figure P i 6.76 requires writing
node or loop equations. Because there are no
series-parallel connections, one cannot apply
the equivalents o f Table 16.2.
(a)
Let r , = 1//?,, ^2 = i//?2,
= \/L^s,
the LC tank is constant and is equal to

0.5C 2.
= i//?3,
= C 5J. Using nodal
analysis, show that the transfer func

tion is
His) =
hn
>2 (>i + y4){y3+>5 ) + y?>Y4{y\+>5 ) + y ^ y s + 1^4)
Figure P I6.77
.y-hO.05
0 .1 2 0 3 8 5 + 1.009
(b)
Obtain the pole-zero plot o f H{s).
(c)
Find
(d)
bigb Q^Find approximate values o f (O^,
(e)
Check your results by doing a SPIC E
and verify that the circuit is
78. Consider the circuit in Figure P I6.78,

which contains a non-ideal inductor and a vari
able capacitor. Suppose Z = 100 iiH and 7?^ =
5.4 Q.
o jj, and oj-,.

(or equivalent) simulation o f your cir
cuit.
(0
Frequency-scale the circuit by ^ =

1000 and
= 100. Show the new
Figu re P I 6 .7 8
880
(a)
Determine the range of the capaci
80. Consider the circuit o f Figure P I6 .80, in
tance C, i.e., Cq < C < C j, such that
which R^= \ 0 n , L= 1 H, /e/= 0.8 n , and C
the circuit can be tuned to resonance
= 1 F.
over the AM radio band (from 550
(a)
Show that the transfer function H{s) is
kHz to 1650 kHz).

(b)
that given in Problem 71.
When the circuit is tuned to 550 kHz
Find the exact values o f to^, coy,,
(b)
and 1650 kHz, determine the circuit
CO,, and co^^^. Since this is a low-Q cir
Q's, the two bandwidths, and the
cuit, do not use the high-Q approxi
lower and upper halF-power frequen
mations.
cies for each bandwidth.

7 9 . 1'his problem illustrates the conceptual dif
ferences
among
the
various
frequencies
encountered in this chapter. For practical high

er circuits, the numerical values of these fre
quencies are all very close. To see the differ
ences, we choose a low-Q circuit. Consider the
Figure P i6.80
circuit shown in Figure P I 6.79.

(a)
Show that
R ,C [
R\
AN SW FKS; (b)
= 1, co^ = 1.0392.
=
0 .9367, (-),. = 0.6, and
= 0 .9602, all in
L)
rad/sec
.V H------
.V-l- 1+
(b)
R .) LC
81. Suppose the current source in the circuit o f

Problem 80 is /^^(/) = 2 cos{t) A and that C is
I f = 5/3 2 ,1 = 1 H, R = 0.8 Q., and
variable. Find the value o f C such that |
C = 1 F, show that:
maximum. Verif}' that Cj3q is not equal to the
The LC tank frequency
= 1 rad/sec.
is
source frequency in this case.
The resonant frequency co^ = 0.6

rad/sec.
82.
The peak frequency
Consider the circuit in Figure P I 6.82, in which
= yjl9 / 25 = 1.077 rad/sec.
(Experimental
C = 0.1 uF, L = 1 niH, and R^ =
of
Q)
Q.
The pole frequency
(a)
^p ~
= 1.2165 rad/sec.
The natural damped frequency co^y =
(b)
Compute the poles and zeros o f H{s).

From these, approximately compute
0.995 rad/sec.
, and
co^^^,
Hints:
1. To find
0
measurement
use equation 16.41
To find OJ , use equation 16.54.
(c)
Is this a high-Q cir
cuit?
rhe impulse response for the circuit is
o f the form
h{t) = Ae~" cos(cor) +
Show that \A\
|i9| for the high-Q case and,
therefore, h{t) s
(d)
Figu re P I 6 .7 9
Plot
sin(co/).
cos((i)r).
h{t)
Ae~^cos{iOt)
using
MATLAB or its equivalent. Show that
ss;
Chapter 16 Band'Pass Circuits and Resonance
the peak will decrease to 1/^ o f the first

peak in approximately QIti cycles of
oscillations. (This is how one experi
mentally determines Q.)
(e)
Find approximate values of O)^,

and Q by changing the series L-R
Figure P I6.83
to an approximate parallel L-R con

nection. Do the results agree with
those obtained in part (a) by the trans
fer function approach?
84. The opening seaion of this chapter dis

cussed the generation of dial tones by resonant
circuits. The circuit for generating the three
O '
tones in the high-frequency group is shown in

Figure P 16.84. Using the results of Problem 71,
explain how the tone generation circuit works.
Your explanation will include an oscillation of
an undriven LC circuit that is coupled to the
transistor circuit inside the box by transformers
w
Coil
studied in the next chapter.
Figure P i 6.82
frequency determining
resonant circuit
83. The switch S in Figure P I6.83 has been

closed for a long time and is opened at ^ = 0.
(a)
Use the Laplace transform method of

Chapter 14 to show that if Q = 0 ) ^ C
> 0.5, then for / > 0, the capacitance
voltage is
sin (o )/ + 0)
where
a =
and
to.
2(2
1-
1
4(2^
(b)
Show that the peak amplitude o f the

damped sinusoidal waveform of part
all switches open when no button

is pressed
pressing a button in column k
closes the switch SCK
when any button is pressed halfway,
S is closed to contact H; when the
button Is fully pressed, S moves away
from H and makes contact with F
(a) decreases to Me = 0.368 of the
Figure P i6.84
highest peak approximately after Q/ji

cycles if Q is large (and, hence,
'w '
co^).
(O^ es
85. For the circuit o f Figure P I6.85a, we have

the impedance
and for the circuit of Figure 16.85b, we have

the admittance
W
882
r\
YijiO) =
Find expressions for the conductance
the susceptance
in terms o f
and
and
r\
so that
the two circuits have the same impedance at a
r\
single frequency O). Then find expressions for

and
in terms of G^ and
under the same
condition.
>G.
iBp
JX.
r>
-------- J
o
(b)
(a)
Figure P i6.85 Conversion of (a) a series RfX^

combination to (b) a parallel G^-B^ combina
tion. No specific functional form is imposed
upon X^ or B^ here.
86. Reconsider Problem 85 for the case where
jCsO)
andy5^ = jC^fa. Find formulas for
and R^ in
terms of C and 7?^ and thus specialize the for

mulas o f Problem 85 to the case of a trans
forming a parallel RC to a series RC at a single
frequency (o. Now do the converse.
C H A P T E R
Magnetically Coupled Circuits
and Transformers
WHAT IS INSIDE THE AC ADAPTOR?
Most electronic equipment operates with dc power sources. For portable equipment, such as a
cordless phone and a cordless electric drill, batteries supply the dc power. Using non-rechargeable
batteries becomes expensive. Furthermore, replacing batteries in special equipment is a task not
easily handled by ordinary consumers. These two factors have prompted manufacturers to install
rechargeable lithium-ion batteries in portable equipment. By connecting several batteries in series,
the available dc voltage may range from 1.5 to 12 V. Whenever the battery runs low, it must be
recharged.
Recharging a battery requires a low dc voltage source (1.512V). An adaptor houses a device called
a transformer that changes the 110 V ac voltage at the household outlet to a much lower ac volt
age. The lower ac voltage is then rectified to become a dc voltage that charges the battery. Some
adaptors contain the transformer only, while others may also contain the rectifier circuit.
Typical specifications appearing on the casing o f an adaptor may be as follows:
model: AC9131
model: KX-AIO
input:
ac 1 2 0 V ,6 0 H z,6 W
input:
output:
output:
ac 3.3 V. 500 mA
ac 1 2 0 V ,6 0 H z,5 W
dc 12V, 100m A
The concepts and methods developed in this chapter will allow us to understand how a trans
former works to change the ac voltage level and also to perform some other important functions
in electronic equipment.
8iS4
Chapter 17 Magnetically Coupled Circuits and Transformers
CHAPTER OBJECTIVES
1.
Understand how the mutual Inductance M, between two inductances
and Z ,,
accounts for an induced voltage in each inductor due to the change o f current in the
other inductor.
2.
Develop a systematic method for writing time domain and frequency domain equations
for circuits containing mutual inductances.
3.
Understand why the mutual inductance is less than or equal to the geometric mean of
the individual self-inductances using an energy perspective.
4.
Expand the repertoire o f basic circuit elements to include ideal transformers, and learn
how to analyze circuits containing ideal transformers.
5.
Learn how to model a pair o f coupled inductors by an ideal transformer and at most two
self-inductances.
6.
Learn how a practical transformer can be modeled by an ideal transformer and some
additional RL elements.
7.
Investigate some important applications o f transformers and coupled inductors in power

engineering and communication engineering.
SECTION HEADINGS
1.
Introduction
2.
3.
4.
5.
Mutual Inductance and the D ot Convention

Differential Equation, Laplace Transform, and Phasor Models o f Coupled Inductors
Analysis o f Coupled Circuits with Open-Circuited Secondary
Analysis o f Coupled Circuits with Terminated Secondary
6.
Coefficient o f Coupling and Energy Calculations
7.
8.
9.
10.
Ideal Transformers
Models for Practical Transformers
Coupled Inductors Modeled with an Ideal Transformer
Summary
11.
12.
Terms and Concepts

Problems
1. INTRODUCTION
You may recall from your high school or grade school science class that if iron filings are sprinkled
on a piece o|- paper and a magnet is moved around beneath the paper, the iron filings move in con
cert with the magnet because the magnetic field induces a force on the iron filings. Similar to the
magnet and the iron filings, a changing current in one coil that is ver)' close to another coil induces
a voltage across the terminals o f the other coil.
Figure 17.1a shows rwo unconnected coils o f wire in close proximity. Figure 17.1b shows two
unconnected wire coils wound around a single ferromagnetic core, hi both cases, a voltage source
excites coil 1 while coil 2 is left open-circuited. Experimental evidence shows that a change in the
current z, generates a voltage Vj, called the induced voltage, across the open circuit; the induced
voltage is proportional to the rate o f change of/ j. Each pair o f coils in Figures 17.1a and 17.1b is
said to be magnetically coupled.
Coll 1
+
V ,( t )
Coil 2
(a)
(b )
FIGURE 17.1 Induced voltage in coupled coils, (a) Two coils in close proximity, (b) Two coils
wound on the same ferromagnetic core.
How does one quantitatively account for magnetic coupling? The strategv' is to introduce a new
circuit quantity called mutual inductance for coupled coils; specifically, similar to the v-i rela
tionship o f a single coil, the induced voltage satisfies the equation
V-7 = M 21
where
cli^
dt
(17.1)
> 0 is the proportionality constant callcd the mutual inductance from coil 1 to coil
2, and the sign, here , depends on the relative winding directions o f the coils. Dot markings indi
cate the relative winding directions. With reference to Figure 17.1, a dot is placed on coil 1 for
reference; if the dot on coil 2 is in position A, the sign on equation 17.1 is +, and if the dot is in
position B, the sign is - . A description of the general dot convention is presented in the next sec
tion.
The situation illustrated in Figure 17.1b is motivated by an extremely important magnetically
coupled device called a transform er, which is used to transform voltages and currents from one
level to another. In electric power systems, transformers are used to step up ac voltages from 10
kV at a generating station to over 240 kV for the purpose o f transmitting electric power efficiently
over long distances. At a customers site, such as a home, transformers step these high voltage lev
els down to 220 V or 110 V for safe, everyday uses. In addition, transformers have numerous uses
in electronic systems, including (1) stepping ac voltages up or down, (2) isolating parts o f a cir
cuit from dc voltages, and (3) providing impedance level changes to achieve maximum power
transfer berween devices, and tuning circuits to achieve a resonant behavior at a particular frequenc)'. After the basic analysis methods are set forth, some examples will illustrate these uses.
886
2. MUTUAL INDUCTANCE AND THE DOT CONVENTION

Experimental evidence demonstrates that if the two coils in Figure 17.1 are stationary, the induced
voltage,
is proportional to the rate o f change o f i . e . , the induced voltage
\>2{t) = M2\
du{t)
dt
as set forth in equation 17.1. Note, however, that coil 1 with inductance Z,, continues to act as an
inductor for which
v,(/) = Li
di\{t)
dt
There are two effects present in the circuits o f Figure 17.1: an induced effect and the usual v-i
relationship o f an inductor. Similarly, Figure 17.2 shows the reverse coupling to that o f Figure
17.1. For the circuit o f Figure 17.2, with the reference dot placed at the top o f coil 2,
v, ( 0
= M ,2
d ilit)
(17.2)
dt
where + would be used if the dot on coil 1 were in position A and - if in position B. As in figure
17.1, the dots indicate the relative directions o f the windings o f the two coils. Also as before, the
second coil continues to act as an inductor, for which
viit) Lo
d ilit)
dt
Secondary
Coil 2
v,(t)
(l )
FIGURE 17.2. Coupling from coil 2 to coil 1 (winding directions not explicidy shown).
As will be verified in a later section,
^ \ 2
>
Hence we designate the positive constant

(17.3)
as the mutual inductance (in henries) o f the coupled inductors.

Figure 17.3 shows a composite o f Figures 17.1a and 17.2 where currents are present in both coil
1 and coil 2. Four effects are now present in the circuit: two induced effects and the tw'O usual self
inductance effects. Linearly superimposing the effects (superposition), we obtain the equations o f
the mutually coupled inductors:
(It
dt
887
(17.4a)
dt
(17.4b)
~ dt
Two questions remain: (i) When is the sign positive and when is the sign negative? (ii) How is the
value o f M determined (experimentally)?
Primary
+
i,(t) 0
Secondary
A
v,(t)
v,(t)
B
FIGURE 17.3. Coupled coils with current excitations present on primar}' (coil 1) and secondar}' (coil 2).
The following rule, identified with equation numbers, governs the choice o f sign lor the induced
voltage.
RULE FOR THE INDUCED VOLTAGE DROP DUE TO

MUTUAL INDUCTANCE
The voltage drop across one coil, from the dotted terminal to the undotted
terminal, equals M times the derivative o f the current through
(17.5a)
the other coil, from the dotted terminal to the undotted terminal.
Or, equivalently,
The voltage drop across one coil, from the undotted terminal to
the dotted terminal, equals M times the derivative o f the current
(17.5b)
through the other coil, from the undotted terminal to the dotted terminal.
With reference to Figure 17.3, if the dot is in position A, all signs are positive, whereas if the dot
is in position B, the sign on M is negative. This rule gives the voltage drop due to the mutual
inductance. To obtain the total voltage drop across an inductor that is coupled to another, one
must add in the voltage drop induced by the self-inductance o f the individual coil, which depends
on whether the labeling is consistent with the passive sign convention.
8 S8
EXA M PLE 17.1. For the configurations o f Figure 17.4, determine the pair of equations that spec
ify the relationship betAveen the voltages and currents.
+
1
+
FIGURE 17.4. Two scenarios for setting up coil equations.

S o l u t io n
First we consider Figure 17.4a. The voltage z^,(r) and the current /, (r) as well as the voltage V2 U)
and the current ijit) satisfy the passive convention. For each coil acting alone,
cji^
(It
for k = 1 ,2 . However, the voltage induced in coil 1 by the current in coil 2 is negative relative to
the indicated polarity on v^{t) as per rule 17.5b, i.e., /-,(r) enters the dotted terminal so that-/'2W
can be viewed as entering the undotted terminal. Fience
Vt(/)= L |- + M ------ ^ = L i
(If
(it
' (li
(It
Using rule 17.5a, the same arguments apply to coil 2, in which case
U )= M -------+ L-) = -M + L-y ^

(It
~ (It
dt
~ (It
Now we consider Figure 17.4b. Here observe that neither pair (i^j, /,) nor {vj, h) satisfy the pas
sive sign convention; hence
Vk(t) = -L^
(iik
dt
for ^ = 1 ,2 . However, the voltage induced in coil 1 by the current /-,(/) satisfies rule 17.5a. Hence
dt
dt
On the other hand, the voltage induced in coil 2 by iAt) satisfies rule 17.5b. Hence
8S9
Exercise. For the configurations o f Figure 17.5, determine the pair o f equations that specify the
relationship between the voltages and currents.
M
+
V,
(a)
F IG U R E 17.5. Two more scenarios for setting up coil equations.
A N SW E R S: (a)
di]
vAr) = - L - +
di
and
di
(//
di-,
di
= -/V/ +
(If
^ ( b )
dt
,
, , di]
, di->
and v^(t) = - M - + 1.-,^^
"
dt
~ dt
EX A M P L E 17.2. This example presents the procedures for marking the dots on an unmarked
pair o f coupled inductors and for determining the value o f M. Consider the configuration o f
Figure 17.6, in which a current source is exciting terminal A ot coil 1 o f the coupled inductors
with unknown M.
DVM
Also marked
com, neg, low
increasing current
F IG U R E 1 7 .6 . D iag ram fo r d e te rm in in g d o t p lacem en t and m u tu al in d u cta n ce.
890
Part I: Dot Placement

Step 1. Place a dot on the terminal at which /j (r) enters.
Observe that we have inserted a dot at terminal A.
Step 2. Apply an increasing current /j(r), i.e., a current for which
> q fo r all t.
clt
For example, one could set /,(/) = 10/(/) mA for 10 seconds. We know that /,(/) induces a volt
age at the terminals o f coil 2 according to
(It
where M > 0. If we put the leads o f a voltmeter across the terminals C -D as suggested in Figure
17.6, the reading will either be Vjit) > 0 or v^{t) < 0. Suppose the reading is V2 {t) < 0. If we reverse
the leads o f the voltmeter by putting them across the terminals D -C , the reading will have the
opposite sign, i.e., v-,{t) > 0.
Step 3. Reconnect the voltmeter leads until V2 {t) > 0, i.e., the reading is positive. Place a dot on the
terminal o f coil 2 for which the voltmeter lead is marked + {or "plus" or high or pos'), i.e., at the
terminal o f higher potential.
For the situation described above, the dot would be placed at terminal D. However, the dots on
coils 1 and 2 could be simultaneously moved to the opposite terminals o f each coil without chang
ing the relative information they convey. (Problem 1 confirms this statement.)
Part 2: Determining M
Again, \'2(0
v-){t) = M
implies that
(If
where M > 0. Taking absolute values o f both sides o f this equation
M=
^2(0
clidt)
dt
Exercise. In figure 17.6, /j(/) = 2tu{t) A and a voltage

the placement o f the dots and the value o f M.
AN SW ERS: Dots are at terminals A and
= \Ou{t) mV is measured. Determine

j
or at B and D: M = ------ = 0.005 H.
Compared with the ramp o f Example 17.2, a more practical input signal and measurement
scheme uses a triangular waveform, as displayed in Figure 17.7.
891
Chapter 17 * Magnetically Coupled Circuits and Transformers
E X A M PLE 17.3. Figure 17.7 shows a circuit and two waveforms, /,(r) and V2 {t), as might be dis
played on an oscilloscope. Determine the placement o f a dot either at position C or at position
D, and the value o f the mutual inductance.
i,(A)
FIG URE 17.7 Circuit and waveforms for Example 17.3.

So
lu t io n
For the time interval 0 < r < 0.5 msec, /,(/) is a ramp function and Vjit) is constant. The infor
mation is similar to that given in Example 17.2 and the solution method is the same. First we place
the dot at terminal A as the current enters A. The current /, is increasing, and ^2 is positive. We
must now determine if the dot goes at terminal C or D. Since the current i^{t) enters the dotted
terminal and is increasing over 0 < r < 0.5 msec, its derivative is positive over 0 < r < 0.5 msec.
Also, the voltage iQ^t) is positive for 0 < r < 0.5 msec with the indicated polarities. Hence accord
ing to rule 17.5a, the dot goes at terminal C and
di^
dt
To determine M, consider that the measured values during 0 < r < 0.5 msec give Vj =
and
di^ldt= 1/0.0005 = 2000 A/sec. Thus, M = 2/2000 = 0.001 H.
EXA M PLE 17.4. In the circuit o f Figure 17.7, suppose the dot positions are at A and C. If /jW
A, find Vjit).
= 2(1
So
lu t io n
In this case, = 200t^"^"//(/;. From Example 17.3, A/= 0.001 H. Hence
dt
v9(/) = M = 0 .2 e ^'//(/) V
dt
Exercise, In the circuit o f Figure 17.7 with the dots in positions A and C, if /j(t) = 0.01
sin(1000?)(^) A, find v~,{i) tor r > 0.
ANSW ER:
= 0.0 Tcosi 1000.^)/,u) A
The preceding treatment o f the mutual inductance has not referred to the physical construction
o f the coils, although we have described procedures for measuring M when the coupled inductors
are assumed to be enclosed in a sealed box. However, for designing a pair ot coupled inductors, or
for a better understanding o f mutual inductance, one must relate the coil construction to the val
ues o f Z.J, Ljy and M. A rigorous study o f this problem requires a background in field theory and
magnetic circuits, which are covered in advanced texts or physics courses. Nevertheless, we set
forth here a few basic properties with reference to Figure 17.8.
M
Coil 1
Coil 2
(b)
FIGURE 17.8. (a) Coupled coils in close proximity, (b) Coils couplcd through magnetic core.
1.
In Figures 17.8a and b, the number o f turns for each o f coils 1 and 2, respectively, is TVj
and Nj. Then the self- and mutual inductances have approximately the ratio
L, :L 2 :M = n } \NI\N^N2
2.
If two coils/inductors are placed in a nonmagnetic medium (e.g., air), bringing the
inductors closer together increases the value o f M.
3.
If one inductor o f a pair is rotated, then a larger value o f M results when the axes o f the
inductors are parallel to each other. The smallest value o f M occurs when the axes are per
pendicular to each other.
4.
Changing the core on which the two inductors are wound from a nonmagnetic material
(e.g., air, plastic) to a ferromagnetic material may increase the values o f Z,j, L-,, and A/by
a factor o f several thousand.
As a final note, the development above presupposes linearity. If the rvvo inductors are placed in a
nonmagnetic medium, this holds true. If the inductors are coupled through a ferromagnetic medi
um (e.g., an iron core), then the linear relationships o f equation 17.4 hold only if both currents
are sufficiently small that the magnetic medium avoids saturation, a phenomenon discussed in
other courses or more advanced texts. Our investigations consider only the linear case.
3. DIFFERENTIAL EQUATION, LAPLACE TRANSFORM, AND

PHASOR MODELS OF COUPLED INDUCTORS
Figure 17.9 shows a pair o f coupled inductors. As developed in section 2, since each coil contin
ues to act as an inductor, the terminal voltage depends on the derivative o f the current through
the coil plus an additional induced voltage due to the changing current in the other coil. This led
to the set o f differential equations 17.4, repeated below as 17.6, where the plus sign is used when
the secondary dot is in position A and the minus sign is used when the secondary dot is in posi
tion B.
dt
dt
dh
dh
V2(t) = M ^ + L 2 ^
dt
dt
(17.6a)
(17.6b)
FIGURE 17.9 A pair of couplcd inductors.

Assuming that there is no internal stored energ}^ in the coupled inductors at time r = 0, then the
Laplace transform o f equations 17.6 yields
Vîs) =
Msl. is)
Vîs) = Msl^ is) + L^sUs)
(17.7a)
(17.7b)
Equations 17.7 represent the x-domain model o f the coupled inductors. Further, if one is con
cerned w'ith the sinusoidal steady state, then replacing s by jui yields the following equations for
analysis using the phasor method:
894
(17.8a)
V j = yojz.ji, yojyv/i2
(17.8b)
V 2 = y to M j +
For equations 17.7 and 17.8, the plus sign is used when the secondar)' dot is in position A and
the negative sign otherwise. These three sets o f equations constitute the core o f the examples and
analyses presented in the remainder o f the chapter. Throughout, the j-domain method is the pre
ferred method, but each pair o f equations has its specific uses.
EXA M PLE 17 . 5 . A pair o f coupled inductors are connected in two different ways, as shown in
Figure 17.10. Find the input impedances,
inductances,
and -^,2(^)> and the corresponding equivalent
and L^^2 '
(b)
FIGURK 17.10 Impcdanccs and equivalent inductances of two series-connected inductors.
(a)
+ Z2 + 2 ^ - (b)
+ >^2 - 2 ^ -
S o lution
For Figure 1 7 .10a, label the inductor voltages

o f each inductor. Observe that
j and
with positive reference on the left side
enters the dotted terminals o f both coils. Hence
= I-jis). Direcdy applying equations 17.7, we obtain
în (-^) =
(-^) +
^L2 (^) =
(^ 1 +
= {L ^ + L 2 + 2 M )sIi {s)
It follows that
im plying that
Ms) Ii,j (^) + {Ms +
) I i n (^)
= I^{s)
êq\ -
89S
+ ^2
= I^{s)
Similarly, for Figure 17.10b, with the same voltage definitions, we observe that again
= Ijis). Also note that /2W enters the undotted terminal. Applying equations 17.7, we obtain
y-m is) = ^ L \(-^) + ^ L 2 (-^ ) = (^1^- - M s ) I i { s ) + ( - M s + U s ) l i { s )
= (L , + L o -2M ).9/ ,.(5)
It follows that
Z,2(^>=
ImU)
= [L\ + h - 2 ) s =
implying that
Exercise. Suppose the circuits o f Figures 17.10a and b are connected in parallel. If Z,, = 40 mH,
Z,2 = 60 mH, and A/ = 25 mH, find the equivalent inductance o f the parallel connection.
AN SW ER: . r . 5 ml I
The preceding example implies that

=
0 7 -9 )
This relationship suggests another way o f determining M and dot markings from measurements.
If an instrument for measuring self-inductance is available, we can use the instrument to measure
Z.^^j and L^^2} from equation 17.9, the difference is AM.
4. ANALYSIS OF COUPLED CIRCUITS WITH OPEN-CIRCUITED

SECONDARY
In the next example we compute the voltage across the secondary o f a coupled inductive circuit
using a high-resistance voltmeter.
EXA M PLE 17.6. In the circuit o f Figure 17.11, assume that the meter resistance
is ver)' large
and looks like an open circuit to the secondary. The switch S is closed at r = 0. Find ij{t) and Vy{t).
8%
FIGURU 1 7 .11 A coLiplcd induccor circuit that might be lotind in a lab to determine dot position.
So
lu t io n
Since the secondary looks like an open circuit, /-> = 0 and K, = Msly Hence if we find /j we can
easily find K,.
Applying KYL ro the primary loop, using equation 17.7a and the fact that /, = 0 produces
sms+R)
.V
Therefore
R
/l(0 =
\ -e
Since
V/^ = yW.s7, = Ms
MZiq
_ iW/To/Li
s { L^s +R) ~ m s + R ) ~
It follows that
H(/)
h
Note that u-y(t) > 0 by virtue o f the way the meter leads were connected to the secondary. Hence
we conclude that the dot is indeed on the upper terminal of the secondary.
89'
Chapter 17 Magnetically Coupled Circuits and IVansformers
Exeixise. Suppose in the circuit o f Example 17.6

is 1 msec, and the voltage
at
= 10 V, /^= 2 Q, the time constant o f the circuit
= 1 msec is 3679 V. Find L^, M , and the ratio M : Lîo see the
factor by which the small, safe voltage
is amplified when the switch is closed. The point here is that
when switching with inductors occurs, a small voltage ma)' produce a dangerously iiigh voltage.
AN SW ERS: /., = 2 mi l. M = 2 H, and M : A, = 1000
An interesting application to older car ignition systems also uses an open-circuited secondary to
produce a very high voltage from a small one to fire a spark plug.
EXA M PLE 17.7. Figure 17.12a shows an automobile ignition system found on older cars while
figure 17.12b shows a simplified equivalent circuit model. Todays ignition systems use electronic
switching. Specifically, the block with the condensor (capacitor) and ignition point is replaced by
something referred to as an ignition module. The module contains a power transistor circuit to
perform the switching action electronically, with the trigger timing typically actuated by a sensor
that measures the position of the cam shaft cither optically or magnetically.
CONDENSOR
SPARK
PLUG
fL
IGNITION f
SWITCH
AL
i
IGNITION
POINT
+
---- n
IGNIT10N(f
COIL
o)
12V
DISTRIBUTOR
n---H I
t J
BATTERY
(a)
point
I mF
primary
to spark plugs
through a
distributor
secondary
(b)
F I G U R E 1 7 .1 2 (a) A n a u to m o b ile ig n itio n system , (b) S im p lified eq u ivalen t c ircu it m od el.
898
In Figure 17.12a, the ignition coil is a pair o f inductors wound on the same iron core, which cre
ates a strong coupling between the coils. The prim ary coil is connected to the battery, while the
secondary coil is connected to the spark plugs or load. The primary has a few hundred turns o f
heavy wire, the secondary about 20,000 turns o f ver)' fine wire. When the ignition point (or con
tact) opens by cam action, a voltage exceeding 20 ,0 0 0 V is induced across the secondary, causing
the spark plug to fire. The generation o f a high voltage to cause the spark plug to fire is accom
plished by a basic RLC circuit containing a switch that represents the point o f the ignition system.
Since the secondary is open-circuited, it has no effect on the solution for the primary current. Let
us do the analysis using the equivalent circuit model o f Figure 17.12b. Suppose that the switch
has been closed for a long time. Accordingly, at r = 0~, we have /j = ElR= 12 A. Using the model
for an initialized inductor given in Figure 14.18 results in the ^-domain equivalent circuit o f fig
ure 17.13.
FIGURE 17.13 ^-Domain circuit for Example 17.7.

The primary current is simply the net driving voltage divided by the total impedance in the series
circuit o f Figure 17.13, i.e..
------
t- L i
R+
sC
+ sL
(17.10)
.V + .9
R
Li
+L ,C
Suppose now that E = 12 V, /? = 2 Q, Z.j = 2 mH, M = 0.5 H,
L~,
= 130 H, and C = 5 lîF.
Substituting the given component values into equation 17.10 yields

6 (5 + 1 0 0 0 )
.v + 10()0i + 10^
(.V + 5 0 0 ) + 0 . 0 0 5 X 9 , 9 8 7 . 5
(.v + 5 0 0 )- + ( 9 ,9 8 7 .5 ^
(17.11)
Taking the inverse Laplace transform o f I^{s) yields

/j(r) = 6^-500^ c o s (9 9 8 7 .5 f ) + 0.03^ -5 M n (9 9 9 8 7 .5 r )
s
cos(lO.OOOr)
(17.12a)
(1 7 .1 2 b )
899
Having obtained
we calculate V2 {t) from the basic relationship o f equation 17.6, using the
plus sign for the dot positions:

V2(/) = M
di^
(It
= 0.5 X 6
cos(10,000/) - 1
si n(10, 000/)
s - 3 0 ,0 0 0 e " ^ ^ ' sin(l 0,000/ ) V, for t > 0.
From this expression, the voltage
reaches a magnitude o f 3 0 ,0 0 0 V in about 157 psec (one-
fourth o f a cycle o f the oscillations). This voltage is high enough to cause the spark plug to fire.
After the spark plug fires, the secondary is no longer an open circuit; the above methodology ceas
es to hold during the firing o f the spark plug.
Exercise. Compute the energy stored in
just prior to the switch across the capacitor opening.
AN SW ER. 0.036 1
In going from equation 17.12a to 17.12b, we neglected the second term in 17.12a, retaining only
the first. W ith the practical component values used in ignition circuits, this approximation is usu
ally valid. In terms o f equation 17.11, the approximation is as follows:
R \
S H------- I
H----------------
2 lJ
2L ,
2 R
S + ----- 5
+ ---------
L]
5 +
E
R
L,C
(17.13)
1 R
1
s~ + s + -----L,
L]C
Applying the inverse Laplace transform to equation 17.13 yields
/ j(/ ) ai
e
R
cos[o)jt)u (t)
where
and
^d =
\
z .,c
U tJ
I
^/z;c
For the first few cycles o f oscillations, the value o f

is nearly zero and the value o f the expo
nential factor is nearly 1. Therefore, /j(/) may be further simplified to
')()()
/|(/) s COS
r \
for small t > 0
Finally, we compute V2 {t) from equation 17.6:

-1
. i
\'o{t) = M = M
^sm
dt
R Ju c
Hit)
for small r > 0. Thus,

V-7(/)
-
E
1
a A/----- f =
= QE
Z.|
(17.14)
approximates the maximum value o f t'2 >where
Q=
is the quality factor o f the series RL\ C circuit. (See Chapter 16.) Later we will show that M!L\ is
approximately equal to the ratio o f the number o f turns o f the secondary to that o f the primary.
Equation 17.14 is a simple formula for estimating the maximum voltage that occurs at the sec
ondary. It shows that although the battery voltage is only 12 V, what appears at the secondary for
a brief moment is very much higher due to the switching action. The voltage is stepped up due to
two factors: (1) the Q o f the series RL^C circuit and (2) the turns ratio o f the ignition coil. From
equation 17.14, a smaller R produces a higher voltage across the secondar)\ But a small R causes
a larger current to flow in the primary circuit and therefore shortens the life o f the breaker point.
In practice, when the engine is running, a resistance wire or an actual resistor is placed in the pri
mary circuit to limit the amount o f current flow through the breaker point. The capacitor (con
denser) serves a similar purpose that o f protecting the breaker point by suppressing the arc that
results when the point opens.
Exercise. The ignition circuit o f Figure 17.12b has = 12 V, Z,j = 1 mH, and C = 0.01 [J.F. If the
total resistance in the primary circuit is /? = 8 Q and M/Zj = 50, estimate the maximum voltage
appearing at the open-circuited secondary when the s\vitch opens.
AN SW ER; Alniui l .s r k\'
')()!
Chapter 17 Magnetically Coupled Circuits and Iransform ers
5. ANALYSIS OF COUPLED CIRCUITS WITH TERMINATED

SECONDARY
To begin this section, we would like to develop some general formulas that make the analysis of
doubly terminated coupled inductors straightforward. By doubly terminated wc mean the con
figuration o f Figure 17.14 where there is a source and source impedance connected to a load
impcdance through a pair o f coupled inductors.
EX A M PLE 17.8.
Consider the circuit o f Figure 17.14. Show that
(i) Z i {s)= L ^ s(ii)
M~s~
L'fS + Z-)
is) = Z, (.V) +
Z i
(s) = Z, (.9) + L^s Z ,M
(iii) G ,2 = ^ =
L-)S + Zt
M~ s + L.\Z~)
VI
'
V\
(iv) the voltage gain Gj,| = =
Z;
(v) the overall voltage gain Gy = =
Z.v
ZoM
An + 2 ]
(^1^2 - M -).v + L,Z2
FICU RE 17.14. Coupled inductor circuit.

So
lu t io n
'1 V /
Vi(^)
Part 1: Find Z,(5) = -------
by two loop equations in the y-domain, converting to matrix form,
and solving.
Step 1. For the primary loop,
K, =
+ Msl^
Fo r the secondary loop
0 = ^2 + Z ,/ , = Msl^ + L^sl, + Z. A =MsI^ + m j + Z p /2
902
In matrix notation,
L ji
Ms-
/r
My
L^s + Z 2
V,0
Step 2. Solve for /j and /j. Using the well-known formula for the inverse o f a 2 * 2 matrix, we
have
/r
L,.v
/2
A/s
-1 p
Mv
+ Z-)
L|.v(L2.v + Z 2 ) - M~s~
L-) s + Z-)
Ms
rv,
-M s
L^s
Hence
L 2 S + Z2
^1 ------------------------------- T T M
LiS{L2S + Z 2 )-M \ s-
which implies that

V'l
W^.v-
L 2 S + Z2
(17.15a)
(a) = Z, (.v) + Z,(.v) = Z ,( 5 ) + L,5 -
(17.15b)
/|
Part 2: Obviously, then,
Also, we note that
-M
L\(L2S
(17.16)
Z 2 ) - M~s
Part 3: Find C ^.2 = Using expression 17.16 and the fact that Vi = - ^2-^2
Vi
V2 _ -Z y l.
^v2 -
Ki
V,
conclude that
Z-yM
Z^M
m L2S + Z 2 ) - M - s
( l , L 2 - M ).v + L,Z2
(17.17)
V'l
Part 4: Find Gj.i = .B y voltage division

Z;.,
Z,
z,- + z ,
Part 5: Find
(17.18)
- Gv^Gv^7.
G y = G ^ lG y 2 =
ZoM
(1 7 .1 9 )
903
Exercise. In Example 17.8 find the currcnt gain ratio,
-M v
Note that the answer can be obtained directly be writing a loop equation for the secondar\'.
The next example applies the foregoing development to a specific case and at the same time
expands the developed formulas to new cases.
EXA M PLE 17.9

Consider the circuit o f Figure 17.15, in which \,(/) = 60\/2 cos(10/) V , /?y = 10 Q
, = 5 Q,
C = 0.01 F, L, = 2 H, I , = 0.5 H, and M = 0.5 H. Find (i) Z.(/10), (ii) V,(/IO), (iii) V.CylO),
(iv) Ij(/10), (v) l 2(/10), and (vi) the average power Pi{j\0) delivered to the C-R^ load.
M
V,
R,
-----------------------------+ 1
------------- ---------- +
FIGURE 17.15 Coupled inductor circuit.

So
lu t io n
First we note that co = 10 rad/sec. AJso, using the effective value for the phasor voltage o f the source
yields W = 60 Z. 0 " V. Thus all calculated voltage and current phasors are effective values.
Step 1. F
i n
Observe that
Z2(yiO ) =
CylO +
10
}_
R,
O.ly + 0.1
7 = 5 - 5 ./
1+ 7
From equation 17.15 and the fact that M = 0.5 H, we have
Z,(ylO) = L|ylO- - ^
. . = Uj\{) + 2 0 M - = 5 + y2()
0.5 X ylO + 5 - ^ 7
904
Chapter 17 Magnetically Coupled Circuits and'Fratisformcrs
Step 2 . /vW Vj(yiO ). Using voltage division,
= i ,S l o T 4 .
= 5 4 ^ jl2 ^ 2 1 .8 7 Z 1 2 .5 .r
Step 3. /vW V 2(/'10). From equation 17.17,

V 2 = G .oV , =
ZyM
V, =
- M )./10+L|Z2
0 .5 ( 5 - 5 ./ )
- ( 5 4 + j\ 2)= 1 8 - 6 7 = 1 8 .9 7 / 1 -1 8 .4 4 ''
7.5y + 10 - IO7
Step 4. /7W lT(ylO ). From equation 17.16,
u =
-M
-0.5
V, = ------- ^----- (54 + y l2 ) = - 2 . 4 - 1 . 2 7 = 2 .6 8 Z .- 1 5 3 .4 "
10
S ;
(L ,L 2 - M - ) 7 lO + L ,Z 2
1 0 - 92.57
Step 5. F/W Ij(yiO ). From Ohms law,
I. =
^1
-"54+12./
Z,(ylO )
5 + y20
= 1 . 2 - 2.4./ = 2.683Z. - 63.44'^
Step 6 . Fitid the average power delivered to the load. The complex power delivered to the load is
S(;10) = V.OlO) X [-I^l/lO)]* = (18 - 6;) X (2.4 - 1.2;) = 36 - ; 3 6
'Faking the real part yields an average power o f //(ylO) = 36 watts.
Sometimes the application o f the formulas developed in the previous rsvo examples, although
straightforward, is not the simplest route to the answer. The following example is a case in point.
EXA M PLE 17.10. Find the steady-state components o f z/j(r) and V2 {t) at the frequency 1 rad/sec
for the circuit o f Figure 17.16, in which ^^^^(t) = cos{t)u{t) V. Note that because a resistance is pres
ent, the circuit responses will contain both a transient and a steady-state component.
1Q
M = 2H
------ O
+
2F
V,
IH
5H -
-o
F I G U R E 1 7 .1 6 C irc u it fo r loo p analysis o f co u p led in d u ctor.
90S
Chapter 17 * Magnetically ('ouplcd Circuits aiulTransform ers
So
lu t io n
Since only the sceady-stare responses are required, we use rlie pliasor method and write two loop
equations given that co =1 rad/sec and
= 1 Z. 0^^ V.
The two loop equations are
and
0 =
12 + ./5I 2 + y2I,
= y2I, + y4.5l2
Putting this in matrix form, we have

T
|+y
jl
j l - *ll
y-4.5
I2
II
74.5
- 0 .5 + y'4.5
-j-
-P -
;4 .5 -
- 0 .5 + y4.5
.-j-.
It follows that
v,=v,,-i, =i-
J4.5
0. 5
- 0 .5 + ./4.5
0.5 - ./4.5
= 0 .1 1 0 4 3 Z .8 3 .6 6 "
-./9
_9
- 0 .5 + J4.5
1 - ./9
-j2
v, = - - L i , = _ _ x
jl
j2 - 0 .5 + 74.5
= 0.22086Z . - 96.34"'
Observe that
V , = -2 V ,
in which case
= 0 .2 2 0 8 6 c o s (r - 9 6 .3 4 ) V and
= 0 .1 1 0 4 3 cos(; + 8 3 .6 6 ) V
Exercise. In Example 17.7, write the simultaneous equations in matrix form and then solve by
Cramers rule. Which method is easier?
Example 17.10 is a drastic case contrived to bring up an underlying propert)': the dot markings
for coupled inductors (/.,, A,, M) determine the sign in the equation relating v-^ to eli^ldt and
the equation relating
to di-)ldt. No intrinsic relations ber^veen the polarities o f
and v-, are
conveyed by the dot convention. For most practical circuits, hoAvever, it is true that the voltage
drops o f coupled inductors from the dotted terminal to the undotted terminal are in phase or
nearly in phase.
906
As a practical example ro end this section, we analyze the circuitry typical o f the front end o f an
AM radio receiver.
EXA M PLE 17 .1 1 . Figure 17.17a shows circuitry typical o f the front end o f an AM radio. Figure
17.17b shows a simplified RLCM model o f this circuitr}'. Here,
and the resistance
= 300 ^2
together represent the antenna. Typical parameter values might be /?2 = Rj = 14,7 k fl, Z,j = 50
nH, L-, = 2450 nH, M = 350 nH, and C = 104.5 pF. Find the transfer function
H {s) =
rhen using the methods developed in Chapter 16, compute
|, Q,
coj, (JL)-,, and the maximum value and the bandwidth o f the magnitude response (i.e., the curve o f
vs. to).
antenna
RF
Amplifier
M
coupling
coils
tuning
1capacitor
SR
X
in
' 6
(b)
(a)
R = 300 Q
' 6
R = 14.7 kO
2
= 98nH
= 2612 pF
(c)
FIG U RE 17.17 Circuits for Example 17.11. (a) The original circuit, (b) A simplified circuit model
in which L^L-y = M~. (c) A design without couplcd inductors.
So
lu t io n
The first step in the solution is to observe that L^L-, = Afi. Then from equation 17.19, the trans
fer function is
Z-yM
V ,{s)
where Z , = /?, and
Z , + Z ,
(z,,L 2 - M ) j + L , Z 2
^m + ^l
Z^y = R-y I !C = ----- . Also, from equation 17.15,

s+
R-yC
^1
(1 7 .2 0 )
90
2 ,2
V
L2 S + Z 2
Loi + Z j
Substituting the expression for Z-^^{s) in equation 17.21 into equation 17.20, then dividing
through by Z j and simplifying, we obtain
CR^Lo
H (s) =
_y2 ^ / _ _ _ _ h ^ _ L _ \ ^ ^ _ _ L
\CRiL2
R2 C )
L2 C
4 .5568x10^ 9
5^ + 1 .3 0 2 0 X 10^^ + f 6 .2497 x 1 0 ^ ( 1 7 . 2 2 )
'
Ks
S
+ B^^.S + fO,
Ks
^2 ^
^^ ^ 2
Q
By inspection, then,
co,
= 6.2497 X 10"^ rad/s,
TU
= 13.02 x 10^ rad/sec,
=
P
and
= 3.5 , 0
= 4 8 ,'
XI.
w
(0, 2 =co, + 0.55(^ = 6 1 .8 5 x 1 0 ^ , 6 3 .1 5 x 1 0 ^ rad/s.
Instead o f the coupled inductors o f Figures 17.17a and b, the same
and
can be obtained
with the parallel resonant circuit o f Figure 17.17c. Following the design method described in
Example 16.3, we find the required element values to be Z, = 98 nH and C = 2612 pF. However,
the maximum voltage gain would have a much lower value, l^(;w,)| = 14,700/(14,700 + 300) =
0.98, compared with 3.5 for the coupled circuit o f Figure 17.17a.
The higher voltage gain achieved in Figure 17.17b can be explained by the concept o f nuiximum
power transfer. A routine analysis would reveal that at to = to^^, the input impedance seen by the
source is a pure resistance equal to the source resistance R^, i.e..
Z ,v ,0 , ) =
Hence the sources impedance o f /?j sees a load
= 300 Q
equal to itself. Thus, maximum power has been
extracted from the source. Since the inductors and the capacitor together form a lossless coupling
net\vork (see the homework problems), the same maximum power is transferred to the load resis
tor Rj. In Figure 17.17c, the reflected load impedance at/o^^^ is not 300
and hence maximum
power is not transferred.

Figure 17.17b contains two inductors and one capacitor. Accordingly, one would normally con
sider it a third-order circuit. Yet the denominator o f equation 17.22 is only a second-degree poly
nomial in s. This simplification is due to the condition M~ =
The significance o f this con
dition will be discussed in the next section.

The general analysis given above is quite useful for complex circuits. However, for many applica
tions the rwo coils have a common terminal, as shown in Figure 17.18a. Figures 17.18b and c show
908
rhe T- and Tt-equivalent circuits char often allow the application o f series-parallel techniques to
simplify the analysis. With the common terminal the coupled inductors have only three accessible
terminals instead o f four. Such an arrangement is called a three-terminal device. The reader might
observe that if A/> Z, or M > L~,y one o f the three inductances in the T- or Tt-equivalent circuit may
have a negative value. This negative inductance appears in a mathematical model and is not the
inductance o f a physical component. Also, the equivalent circuits shown in Figures 17.18b and c
are for the specific dot locations indicated in Figure 1 7 .18a. A change o f one dot location in Figure
17.18a will result in a change in the sign in front of M in Figures 17.18b and c.
2
L ,- M
L ,- M
L^-M
A = L,L^-M2
(3)
(b)
(c)
FIGURE 17.18 (a) Coupled coils with a common terminal, (b) T-cquivalent circuit.
(c) 7t-equivalent circuit.
Verification o f the parameter values in the equivalent circuits in terms o l l j , Z.,, and M is left to the
homework problems. We now illustrate their use with a simple equivalent inductance example.
EXA M PLE 17.12. In the circuit o f Figure 17.19, all initial conditions are zero. If i-^(^t) = u { t - 1)
A, find the response,
0.5H
2H
10
IH
2.5H
3H
FIG URE 17.19 RLM circuit to be simplified using the T-cquivalent.

SO L U T IO N
The key is to find the input impedance seen by the current source. Apply the T-equivalent circuit
to the coupled inductors o f Figure 17.18b. Using series-parallel techniques, the equivalent induc
tance in parallel with the 1 H resistance is
25+ 1
T h e re fo re
v,ît) =
~^h({t -\)
V.
= 2 H. T'hus
2.V -Hi
5 + 0.5
909
Chapter 17 * Magnetically Coupled Circuits and 'Iransformers
Exercises. 1. Repeat Example 17.12 with the dot on the 3 H inductor moved to the bottom.
AN SW ER: Same as in example.
2. Find the input impedance
for the circuit o f Figure 17.20. In the circuit, k =
is called the coupling coefficient, to be studied in the next section.

ANSW ER:
v ^ l^ 2
2.V+ 0.5
3H
3H
0.50
o oFIGURE 17.20
6. COEFFICIENT OF COUPLING AND ENERGY CALCULATIONS

The first part o f this section justifies our assumption that
One can justify that
A/j2 = A/t] = M by the principles o f magnetic circuits, but this is beyond the scope o f this text.
Our justification stems from the physical property that a pair o f stationary coupled coils cannot
generate average power. We will also show that the mutual inductance M has upper bound
> i.e., the mutual inductance can never exceed the geometric mean o f the self-inductances.
Justification o f Myj = ^ 2 1 ~ ^
This property is a consequence o f the principles o f electromagnetic field theory, which are beyond
the scope o f this text. To make our approach accessible to the beginning student, we build our jus
tification on the passivit)' principle for inductors.
THE PASSIVITY PRINCIPLE FOR INDUCTORS

A pair o f stationary coupled inductors is a passive system; i.e., they cannot generate energy
and, hence, cannot deliver average power to any external network.
Our technique is a so-called proof by contradiction; i.e., to show that

= ^ i\ =
s equiva
lent to showing that if A/, 2
then the passivit}' principle is violated. As a first step, suppose
that yV/p M-yy Then, instead o f having equations 17.6, the differential equations for the cou
pled inductors with the standard dot locations must take the form
10
, du
d ii
(17.23a)
~ dt
(17.23b)
~ dt
Let us apply /j = sin(/) and /, = cos(r) to the inductors. From equations 17.23, the terminal volt
ages are
dt
" dt
Vj = Li s in (/ )+ M p c o s ( / ) = Li c o s ( / ) - M p s in (0
and
Vo = yUî s i n ( 0 + c o s ( 0
~ dt
~ dt
= M ^ i c o s i t ) - Losin(t)
The total instantaneous power delivered to the coupled inductors is the sum o f the powers deliv
ered to the inputs, i.e.,/>(r) = v^{t)i^{t) + V2 {t)i2 it)- Therefore
p{t) = yj/j + Vjio =

To calculate
cos(f)sin(t) -
sin^(r) + M 21 cos^(r) - L-, sin(r)cos(t)
(17.24)
the average power delivered to the coupled inductors, we use the identities
sin(r) cos(f) = 0.5 sin(2f), sin^(r) = 0.5[1 - cos(2r)], and cos^(r) = 0.5[1 + cos(2/)]. It follows imme
diately that the first and the last terms in equation 17.24 make no contribution to
the terms involving
whereas
ô
=
(17.25)
This result shows very clearly that if M j 2 > ^21

excitations /, = sin(f) and ij = cos(r) will
lead to a negative
violating the passivity principle. Similarly, if M p <
then the new
(transposed) excitations Zj = cos(f) and
12
= sin(r) will again lead to a negative
violating the
passivity principle. Therefore, we conclude that A/p = J^ 2 r
W ith
= A/-7] = M, the average power, P^^^, is always zero for arbitrary sinusoidal excitations.
CALCULATION OF STORED ENERGY

Having proved that A/j2 =
= M, we shall now show that there is a limit to the value o f M
that is attainable once I j and L 2 are specified. Again we use the passivity principle and the fact
that stored energy is the integral o f the instantaneous power.
11
Consider the coupled inductors shown in Figure 17.9. The voltage-current relationships at the terminals are given by
^
v.(r) = L , 5 M &
(17.26a)
V2(/) = M ^ + t 2 ^
(17.26b)
and
^
with the upper sign (+) for the dot in position A and the lower sign ( -) for the dot in position B.
O '
Let us assume that the inductor currents are initially zero (at r = 0). In this state, there is no ener
gy stored in the system.
o
Recall that the energy stored by any device over a time period [0, 7] is the integral o f the instan
taneous power over the interval. For a pair o f coupled inductors with no initial stored energy, we
have
dt
o
= i L f (r )+ i z | (r ) j [ M (/,f + ,-,5 ),
o
^
4 L ,f ( r ) 4 i ,/ | ( r ) M j ;'3 '^ ( ,V ,)
2
2
J/,(0)/2(0)
= U ,i f ( T ) + ^ L ^ il ( T ) M i t { T M T )
O '
(17.27)
where ^(/|/2) is the total derivative o f the product /j /2 and is equal to /jc/f^ +
For the specific scenario above, we apply driving sources to the inductors to bring the currents up
VO
to /|(7) = /j and i2{1) = /j . Then, the energy delivered to the inductors during the time interval
(0, 7) is, by equation 17.27,
W (J) = ^ W \
^ hh
(17.28a)
A couple of things about this result are worth noting:

1.
The final integral in equation 17.27 depends only on the final values, which in the case
of equation 17.28a are 7j and /j . The exact waveforms o f /j(f) and t2 it) during
< t< T
are immaterial.
2.
The energy W{ T) delivered by the sources during 0 < f < T is not lost, but merely stored
in the system.
To grasp property 2, we may adjust the sources so that the currents are brought back from 7j and
/ j at / = r to zero at some t = T > T. Then the energy delivered by the sources to the inductors
during T < t < T may be calculated in a similar manner, to obtain
'
912
Chapter 17 * Magnetically Coupled Circuits and Iransformers
W iT') - \\\T) = - - L d }
7
(17.28b)
with the upper (-) sign for the dot in position A and the lower (+) sign For the dot in position B
in Figure 17.9. Equation 17.28b is precisely the negative ol equation 17.28a. Thus, all o f the ener
gy delivered by the sources during 0 < t < 7 has been returned to the sources during T ^ t ^ T\
For this reason, the energy given by equation 17.27 is called the stored energy. Another way of
recovering the stored energ)' is described in Problem 48. The physics o f the situation shows that
the energ)- is stored in the magnetic field produced by the currents in the inductors.
Upper Bound for M and the Coefficient o f Coupling
The energ)' W[T) must be nonnegative for arbitrar) values o f /, and /-,. Otherwise, the inductors
will ht g e n e r a t i n g during the time interval 0 < t< T , which would violate the passivity prin
ciple. To ensure a nonnegative W{T) for all /] and I 2 , the values o f Zp L j, and M must satisfy the
inequalit)'
(17.29a)
or
M ^ 7 ^ ,L 2
(17.29b)
To show this, we rewrite equation 17.27 in the form

2
W(T) = - l
where x = ^\l^2
U2J
[h )
L|.v
i
1 2M.X + L ^
(17.30)
^ current ratio. Equation 17.30 shows that Wi'f) is negative whenever^(a;) is
negative. Now,y(,v) is a second-degree polynomial in a: with a positive coefficient for the
term.
Consequendy, the curv'e ofy(.v) vs. a: will be a parabola opening upward. From analytic geometr)',
depending on the sign o f the discriminant D = {M~ - L^L-,), the curve may or may not intersect
they{.v) = 0 axis, as illustrated in Figure 17.21. From the figure, it is obvious that if D > 0, there
will be some current ratio that yields a negative^(a:) and hence a negative W{T), again violating
the passivity principle. Therefore, D = {M~ - L^L^) < 0, which yields equation 17.29.
f(x)
13
The degree co which M approaches its upper bound yJL\l^ is expressed by a positive number
called the coefficient o f coupling, defined as
k ^ - j= .
(17.31)
yjL\L2
From equations 17.30 and 17.31,

(17.32)
When /^ = 0, Af is also zero, and the inductors are uncoupled. When ^ = 1, the inductors have
uuity coupling, an idealized situation impossible to realize in practice.
E XA M PLE 17.13. For both situations o f Figure 17.9, suppose that

H, /,(7) = /, = 2 A. and
= 20 H, A/ =
7) = I^= 4 A. Find:
1.
The coupling coefficient k.
2.
The stored energy- at r = T.
S o lu tio n
1.
= 5 H,
For both dot positions, k
= - 7- -------- =
0.8
V 5x20
2.
For the dot in position A in Figure 17.9,
Stored energy =
= 0.5 X 5 X 2^ + 0.5 X 20 X 4^ + 8 X 2 X 4 = 234 joules
For the dot in position B,
Stored energ)' = 0.5/.j/| +
- A/Zj/j
= 0.5 X 5 X 2^ + 0.5 X 20 X 4^ - 8 X 2 X 4 = 106 joules
Exercise. Repeat Example 17.13 with Z,j = 4 H,

AN SW ERS: in random order: 112, 0.~5. 16
= 16 H, A/ = 6 Fi, 7j = 4 A, and 7, = - 2 A.
914
EX A M PLE 17.1 4 . In the circuit o f Figure 17.22, /j(7) = /j= 6 A. Find the minimum value o f
the stored energy and the corresponding value o f /2(7) =
A.
M = 3H
9H
4H
FIGURE 17.22 Coupled inductors for calculating the stored energy in Example 17.14.
S o lu tio n
From equation 17.27,

W = 0 .5 L ,/ f+ 0 .5 L 2 / | -M / ,/ 2
= 0.5 X 4 X 36 + 0.5 X 9 X /2^ - 3 X 6 X /2
= 4 .5 / 2 - 18/2 + 7 2 = /(/2 )
Following the standard method in calculus for finding the maximum and minimum, we set
dW Idlj to zero and solve for Ij-
d l2
clh
This yields 12 = 2 A, and the corresponding minimum stored energy
= 18 - 36 + 72 = 54 J.
Exercise. Repeat Example 17.14 for /j = 12 A.

AN SW FRS: 4 A. 216 1
7. IDEAL TRANSFORMERS
Developing the Equations for an Ideal Transformer
Two coupled dijfereutial equations containing three parameters Z.j, Z<2, and M characterize the cou
pled coils of Figure 17.9. By imposing two idealized conditions on these parameters, the pair o f
differential equations can be approximated by a pair o f algebraic equations.
Idealization 1: The coupled inductors have unity coupling; i.e., M~ =
and the cou
pling coefficient ^ = 1.
Effect o f Idealization 1: With ^ = 1, the coupled coils have the voltage transform ation
property
chapter 17 * Magnetically Coupled Circuits and Transfonners
915
where a\ s 2i constant and both v-^ and V2 are the voltage dropsfrom dotted {position A) to
undotted terminals o f the coils, as per figure 17.9.

To derive the condition of equation 17.33, note that the constraint
= ZjZ2 implies that Lj/M
= M /Zj. Let the ratio

M
Then Zj = oM and M = aLj. Substitute these relationships into equations 17.7. This leads to the
following sequence of equalities:
vi(t) = Li +M ^ = aM- + aL2^ = a M- + L2- =av2(t)

dt
dt
dt
dt
dt
d t}
Although the idealization o f a unity coupling coefficient is not achievable in practice, coupling
coefficients near unity are achievable by winding the turns o f two inductors very closely together
so that n^ l y all the flux that Unks one coil also links the other coil. The constant a = L^/M =
M/L^ = V V Z ^ in equation 17.33 reduces to the ratio of the numbers of turns, denoted by N-^
and N 2 , of the two coils, i.e., a = //j/A^2* Thus a is called the turns ratio. When ^ = 1, it is possi
ble to show that
W ith unity coupled coils, equation 17.33 indicates that the voltages v^{t) and V2 {t), both from dot
ted {position A o f Figure 17.9) to undotted terminals, always have the same polarity. W ith coupling
less than unity, it is possible for v-^ and V2 to have opposite polarities at some time instants, as
shown in Example 17-10.
Idealization 2: In addition to unity coupling, the coupled coils have infinite mutual and
self-inductances.
Effect o f Idealizations 1 and 2: W ith k= \ and idealization 2, the pair o f coils has the cur
rent transformation property
i2 {t) = -ai^{t)
where both / j and
(17.34)
are the currents entering the dotted terminals of the coils, as per Figure
17.9, dot in position A.

To derive equation 17.34, we again make use of the unity coupling condition L^/M = M/L 2 = a^
and rewrite equations 17.7 as
d(Ly.
dt\M ^
-------- ^/l + /9
(17.35a)
916
Reasonably assuming that the voltage vît) is bounded, i.e., there is some finite constant
that |i'|(f)|s
for all t, then
lim
M-*oo
such
\M
A similar relationship holds for

(17.35b)
V2
where if |y2(/)|^
for all t.
lim
L^-*o
In other words, as Zj <, Z2
<, and Af - <, equations 17.35a and 17.35b each lead to
^ (a ii + / 2 ) - 0
(17.36a)
ai^{t) + i2 {t) = C
(17.36b)
whose solution is
for some constant C. Now assume that the coils are unenergized prior to the application o f exci
tations; i.e., at some time f = Zq in the past, /j(Zq) =
which must be true for any real cir
cuit. It follows that the constant C in equation 17.36b is zero, and consequently /2(^) = -^^|(^)We remark that this derivation is valid for non-dc voltages and currents. The negative sign in equa
tion 17.33 implies that at any time, if a current enters one coil at the dotted terminal, then the
current in the other coil must leave the dotted terminal.
The condition o f infinite inductances (Z,p Lj, M) is another idealization that is not realizable in
practice. However, this condition can be approximatedhy using a magnetic material with very high
permeability as the common core for the two coils.
r\
917
Chapter 17 * Magnetically Coupled Circuits anti Transformers
THE IDEAL TRANSFORMER

Two coupled coils satisfying the relationships
V2 (0
N2
/tCO
Ni
are said to be an ideal transformer, as shown in Figure 17.23a.
FIG U RE 17.23 Symbol and defining equations for an ideal transformer, (a) Both currents enter
the dotted terminals, (b)
enters the dotted terminal and
leaves the dotted terminal.
In Figure 17.23 two vertical bars serve as a reminder o f the presence o f a ferromagnetic core in the
physical device. The word ideal may or may not appear in the schematic diagram. Again, the
mathematical model o f an ideal transformer depends only on the turns ratio a : 1 and the relative
dot positions. To avoid the negative sign in the current relationship, an alternative labeling o f volt
ages and currents as shown in Figure 17.23b may be used. The subscript p stands for the primary
coil, which is connected to a power source, and s for the secondary coil, which is connected to a
load. Note that
is entering at the dotted terminal and
is leaving the dotted terminal. The nota
tion o f Figure 17.23b is more commonly used in the study o f electric power flow.
One important simplification resulting from the idealizations is that an ideal transformer is char
acterized by two algebraic cc[\.\7iX.\ons in terms o f its terminal voltages and currents through a sin
gle parameter a, the turns ratio. This is to be contrasted with a pair o f coupled coils, characterized
by two differential equations containing three parameters, Z,p L-,, and M.
018
Chapter 17 Magnetically Coupled Circuits and Iransformers
EXA M PLE 17 .1 5 . For the circuit o f Figure 17.24, suppose /?^ = 10 0.,a = 0.1,/ ?^ = 1000 Q, and
v^{t) = 20 cos(300;rr) V. Find /?,,
/j(r),
and /-,(/).
i,(t)
ijlt)
a: 1
FIG URE 17.24, Doubly terminated circuit for Example 17.15.

SO LU TIO N
From the definition o f resistance,
V ,(0
a V 2 (/ )
= =
'l(')
2 V2 ( 0
2n
n n .
in n n
,n
= a ----- = a Ri = 0.01 x 1000 = 10 Q
-'2 (')
Using voltage division,
V](/) = ------ -
/?_, + /?,
= X 20cos(300> t/) = 1 0 co s(3 0 0 ;r/) V

2
It follows that
vAr)
10cos(300;rr)
/,(/) = -^ ----------------------- = cos(300;r/) A
10
Using the ideal transformer equations,
viU) =
= 100co s(3 0 0 ;r/) V and /,(/) =
Exercises. 1. If in Example 17.15
= -0 .1 cos(300;rr) A
= 0, find v^{t), /j(f), V2 {t), , and
AN SW ERS. 10 i l , 20 cos(300.t/) V, 2 cos(300.Tr) A, 200 ôs{Mnnt) V, - 0 .2 cos(300.t/) A

2. If in Example \7.\5 R^= 10 ^2 and
= 400 0
find /?,,
/j(r), V2 {t), and ijit).
A N SW ERS. 40 2. 16 cos(300.t/) V, O.-i cos(300.Tr) A. U)0 a)s(3()0;r/) V,-0.04cos(300.t/) A
3. In Example 17.15, find the average power absorbed by R^.
ANSW ER. 5 watts
919
The next example generalizes Example 17.15.

E X A M P L E 17.1 6 . For the circuit o f Figure 17.25, find input impedance Z-JS) and then the
transfer functions
V,(s)
V,(s)
ViCs)
Vj(.s)
a:1
<
+
Z,(s)
:
ideal
FIG URE 17.25 Doubly terminated circuit for Example 17.16.

SO L U T IO N
Using the definition o f an impedance and the properties o f the ideal transformer.
^ ^ ^ V^(s)
aVjis)
2
.
Z:,,{s) = =
= a ---= a ZAs)
/|Cv) J i i s )
-hAs)
^
Therefore, by voltage division,
c rz ^ is)
Vi(s)
Z,(>v) + Zi{s)
z,{s) + a -Z ^s )
Further, by the properties o f the ideal transformer,
V|(5)
in which case
Vsis)
Z ,(s) + crZ L (s)
Finally,
, , , ( , ) . Z2<) =
= ----------=1^^--------Z i,(.?)V j(s)
Zi_U)
Z,(.s) + a-Z^(s)
920
Exercises. 1. Repeat the derivation
Z-^p) in Example 17.16 For the case where the dot on the
secondary is at the bottom.

ANSW ER: Z J s ) = a"Z,[s)
and Z^{s) = R + L^. Find the frequency oJq at
2. In Example 17.16, suppose Z/ (.v) =
which the primary is resonant, i.e., the source sees a pure resistance.
ANSW ER.
a
J lc
3. In Example 17.16, suppose a = 2, Zîs)
Zp) = 10 12, and

^
Find/jW and/j(r).
=
^
ANSW ER: '

s + 0.4
The calculation o f
in Example 17.16 is so important (especially in maximum power trans
fer calculations) that we call it the im pedance transform ation property o f an ideal transformer.
IMPEDANCE TRANSFORMATION PROPERTY

If the secondary o f an ideal transformer is terminated in a load impedance Zj{s), then the
impedance looking into the primary is
Z Js ) = a^~Z^{s)
(17.37)
where a is the turns ratio taken in the direction from source to load. (See Figure 17.26; the
dots are not marked on the figure because their positions are immaterial for this application.)
Chapter 17 Magnetically Coupled Circuits aiul Transformers
921
10.V
Exercise. In Figure 17.26, suppose n = 0.1 and
Find
ANSW ER.
.s- + 10
E X A M PLE 17.17. For the circuit o f Figure 17.25, find the instantaneous power delivered by the
ideal transformer and the energ)' stored by the ideal transformer.
SO L U T IO N
With conversion o f the voltages and currents to the time domain, the power delivered to ideal
transformer is by definition
pit) = V j/|
i't
+ vo/^ = (c /i'-)) =- +
~ \
= 0
a}
Hence, considered as a single unit, an ideal transformer neither generates nor consumes instanta
neous power; whatever instantaneous power is received at one side must transfer to the other side.
Furthermore, since p{t) is identically zero, its integral with respect to t is also zero. Thus, an ideal
transformer cannot store any energy.
In summary, the instantaneous power dehvered to one side o f an ideal transformer is transferred
to whatever loads the other side, and as a result no energy is stored in the ideal transformer.
Exercises. 1. In Figure 17.26 suppose the primar)' o f the transformer has <7 = 0.1 and is connect
ed to a source having
= 10 Q. Find Zj{s) for maximum power transfer to the load.
ANSWHR. ZÛ) = 1 kLl
2. In Figure 17.26 suppose the primary o f the transformer again has a = 0.1 and is connected to
a source having Z^^(/100) = 10 + J5 0 Cl. Find ZîJlOO) for maximum power transfer to the load.
AN SW ER. Z^(/100) = 1000 -y '5000 LI
Limitations o f the Idealization

Although mathematically direct, the derivation o f equation 17.34 does not provide insight into
the limits o f practical transformer design and operation. Let us reconsider the second idealization
using the 5-domain coupled transformer equations,
Vîs) = L^slîs) + Mslîs)
922
Divide rhe first equation by Ms and the second equation by L 2 S. To observe the frequenc)'-dependent behavior, set s =j(x) to obtain
Mjw
L 2 J (0
= ^ I \ i j ( o ) + I- , ( j co ) = a I i { j c o ) + 12 (j co )
= - ^ /1 (yw ) + / 2 (7 w ) = ^//1 (yw ) + / 2 ( )
Lj
Practical transformers have rsvo customary properties: (i) the frequency content o f the voltages and
currents are band limited, meaning that the transformer is guaranteed to operate only over a
restricted frequency range, 0 <
^ O) :S
ages are specified, i.e., |Vj(/to)| :s
that docs not include dc; and (ii) maximum volt
and |K,(/a))| rs
Under these practical conditions we
see that it is necessar)' to make M and L-> (and since a =

frequency range
^ co :s
I M,
also) sufficiendy large over the
so that
Viij(o)
Mco
0 and
V2 O )
Li(o
sO
Thus we will achieve the current transformation property /|{/(jd) =
over the frequency
range o f interest provided the voltage magnitudes are less than

Som e Practical Applications o f Ideal Transformers
We now indicate some practical applications o f ideal transformer properties.
E X A M PLE 17 .1 8 . In Figure 17.27 an ideal transformer steps down the voltage o f a 2400
source to pow'er 10 incandescent lamps in parallel, each drawing 0.5 A.
(a)
Find the voltage (magnitude) across each lamp, and
(b)
find the current (magnitude) delivered by the source.
transformer
source
2400 V
rms
load
10
1760
turns
88
turns
10 incandescent lamps
represented by resistors
FIGURE 17.27 A transformer is used to step down a source voltage to meet the lamp specs.
S olution
Since only magnitudes are involved in this problem, dot positions on the transformer and refer
ence directions for voltages and currents arc immaterial. The turns ratio is a =1760/88, Since 2400
=
the voltage across each lamp is
923
88
^load
1,760
2 .4 0 0 = 120 V,
Since 10 lamps use 0.5 A each, the total current used is 5 A. Using the current transformation
property o f the ideal transformer,
/.V
1,760
X 50 = 0.25 A
Exercise. In the circuit o f Figure 17.26, represent each lamp by a 200 H resistance. Suppose the
source is limited to delivering a maximum o f 3 kVA. How many lamps in parallel can be con
nected to the secondary o f the transformer?
ANSWER: 41
E X A M PLE 17.19. Figure 17.28 shows a simplified model o f an audio amplifier containing an
ideal transformer. The input voltage is 1
at 2 kHz. T he load is a loudspeaker, represented by
z 4 Q resistance.
(a)
Find the average power delivered to the 4
load if it is connected directly to the ampli
fier (i.e., with the transformer removed and the resistor connected across A-B).
(b)
With the transformer and load connected as per Figure 17.28, with turns ratio a= 5, find
the average power delivered to the load.
(c)
If the turns ratio a is adjustable, what value allows maximum power transfer to the load?
W hat is the value o f the maximum power?
a:1
speaker
FIGURE 17.28 An ideal transformer used for maximum power transfer.
So
lu t io n
(a) No transformer and resistor connected across A-B. First observe that the magnitude o f the current
through the 4 Q resistor is
200
4Q
9 0 0 -^ 4
= 0 .2 2 1 2 A
T h e re fo re , the average pow er absorbed is
/^4fi = (0.2212) x4 = 0.1958W
) 2 i
(b) Transformer reconnected, a - b . K\ the primary o f the transformer,

Z^. = //2 x 4 = 100
From the voltage divider formula, the voltage across the primary is
100
pri
(200|V ,|)=20 V
900+100
T he voltage across the secondary is

= 351=4v
5
Therefore,
4
= 4 W
(c) For the maximum power transfer, the turns ratio a should match the secondary impedance to
that o f the primary; i.e., the 4
resistance reflected back to the primary should be 900
to match
the internal resistance o f the amplifier. Hence

900 =
W ith
= 15, the reflected impedance is 900 Q. and

100
15
^ = 15
|= 100 V, meaning that
f A ^
= 6.667 V. Thus
--------= 1 1 .1 1 W.
Exercise. Repeat all parts o f Example 17.19 for a loudspeaker with a resistance o f 16 ohms.
ANSW ERS: i).-(,28 W.
W, 11.11 W
8. MODELS FOR PRACTICAL TRANSFORMERS

The ideal dc voltage source was one o f the first basic circuit elements studied in this text. Ideal
means that it maintains a constant voltage for arbitrary loads connected across its terminals. Such
a voltage source does not exist in the real world. However, because an ideal dc source in series with
a resistance will approximately represent practical sources such as a battery, its use is very impor
tant. Similarly, although an ideal transformer does not exist in the real world, it can approximate
ly represent a practical transformer with the addition o f some other ideal circuit elements.
92 S
Because o f the ferromagnetic core used in its construction, a practical transformer is inherently a
nonlinear device. Nevertheless, a first level o f approximation represents a practical transformer, by
the circuit o f either Figure 17.29b or Figure 17.29c. Figure 17.29a is ideal, whereas Figure 17.29b
contains extra inductances to account for leakage fluxes and other magnetic phenomena. These
inductances give rise to a finite usable bandwidth for a practical transformer. Figure 17.29c shows
additional resistances to account for internal power losses. Thus the ideal transform in conjunc
tion with other circuit elements can be used to approximate a real transformer. Specifically, trans
formers do not work for dc or for very high-frequency signals. The circuit model o f Figure 17.29b
makes this behavior clear: the inductance
and
causes a short circuit at dc, and the inductances
produce open circuits at high frequencies. A second level o f approximation would begin to
account for nonlinearities, a topic left for more advanced courses.
N ,:N ,
+
(a)
-I-
N.
V.
N,
N.
ideal
N ,:N ,
L /Y Y \
-h
(b)
V,
ideal
L.
+
V,
Ideal
F IG IJR F 17.29 Linear transformer models, (a) Ideal case, (b) Extra inductances to account for leak
age fluxes and other magnetic phenomena, (c) Extra resistances to account for internal power losses.
In Figures 17.29b and c, the parameters
and
are called primary and secondary winding leak
age inductances, respectively. They are due to that part o f the magnetic flux that links w'ith one
winding but not the other.
is called the magnetizing inductance and is due to that part o f the
magnetic flux that links both windings. Practically speaking,
and
926
The model o f Figure 17.29c incorporates the resistances
and
which are the resistances o f the
primary and secondary windings, respectively. R^^ accounts for the power loss in the iron core due
L^, and
to hysteresis and eddy currents. These resistances as well as the inductances
can be
calculated from knowledge o f the physical layout and materials properties used in the construc
tion o f the transformer. This task usually requires a large set o f design formulas and (empirical)
charts. After a transformer is constructed, these parameters can be experimentally determined.
Books on ac machinery set forth such experimental techniques.
From a circuits perspective, once the model parameters in Figure 17.30 are known, the steadystate analysis o f a circuit containing practical transformers requires no more than the phasor
method or the Laplace transform method for its analysis. l\vo examples will now be given.
Example 17.20 shows the effect these parameters have on frequency response.
E XA M PLE 17.2 0 . Example 17.19 considered an audio amplifier with an ideal coupling trans
former. Figure 17.30 shows the same amplifier with a practical transformer in place o f the ideal
one. O ur objective is to illustrate the effect a practical transformer has on the power delivered to
the 4 Q. speaker as a function o f frequency. We assume our turns ratio ^ = 15, which produced
maximum power transfer in Example 17.19. We use the model o f Figure 17.29c with
(equivalent winding resistance), ^ ^ = 8 mH (effective leakage inductance),
and
= 40 Q
1 M ii (core loss),
= 1 H (magnetizing inductance). For simplicit)^ we set Z,^= /?^= 0 in the model o f Figure
17.29c.
Lp= 8mH
1 5 :1
40
FIGURE 17.30 Audio amplifier with practical transformer.

S o l u t io n
Our analysis begins with the writing o f two loop equations in terms o f /,(;) and l 2 {s):
-20()K = [ , + /? . + L,,^ +
" M
N ote that since R^^^= 1 M^^ and
"
] /, - - V V L / ,
R , ^ L , s }
R, + L,s -
= 1 H,
j
1 -I-
R,
92-7
for s =yto and w in the audio frequency range 0 to 2 ;r x 20 krad/sec. Hence
- 2 0 W , .( R , .
For the second loop, again neglecting
we have
In matrix form
- 2 0 0 V/,,-
(/?! + Rp + {Lp +
LjS
-L,s
/l1
{R^q + L,s)
From Cramers rule

- 2 0 0 L, 5
h =
LpL,S^ +[{R\+ Rp)L, + Req{i-p +
Vin
+ (Ri + f<p)Req
Our purpose is to show how average power to the load varies as a function o f frequency in a prac
tical transformer. To this end, for v,(/) = yj2cos(cot) V, the power delivered to the load is the
power delivered to the primary o f the ideal transformer, which is
/> ,> ) = ,,|/2(/ )P
Recall that the effective value o f the input voltage is 1 V. To obtain the desired plot o f P iif) (dis
played in Figure 17.31), consider the following MATLAB code.
R1 = 900; Rp = 40; Lm = 1; Lp = 0.016;
Req = 900;
num = [-200 0];
den = [Lp*Lm ((Rl+Rp)*Lm +Req*(Lp+Lm )) (Rl+Rp)*R eq];
w = linspace(0,2*pi*le4,500);
I2 = freqs(num,den,w);
I2mag = abs(I2);
PL = Req*I2mag.^2;
plot(w/(2*pi),PL)
grid
ylabel(Average Power to Load)
928
Chapter 17 Magnetically Coupled Circuits and Iransformcrs
TD
TO
O
4-'
$
0
Q.
q;
cn
(Z
I
01
>
<
Frequency in Hz
(a)
"O
fD
OJ
5
Cl.
0)
cn
2
oj
<
Frequency in Hz
(b)
FIG URE 17.31 Plot of the power delivered to the speaker load as a function of frequency in Hz. (a)
Overall response, showing sharp decrease with increasing frequency, (b) Low-frequency response,
showing significant reduction in power transfer.
The plot shows that the power is down from the peak by about 0.5 W at 4 5 0 0 Hz, and, at 10
kHz, the power delivered to the load is down about 2.5 W. Also note that a practical transformer
does not operate at dc and at very low frequencies, as demonstrated in Figure 17.31b.
In the next example we consider voltage drops in commercial power lines using practical trans
formers.
)2 9
EX A M PLE 17.21. Fhe pniccical transformer in Figure 17.32a is designed for operation at 60 Hz
and 1100/220 V, i.e., ideally a 5:1 step-down transformer. Using the model o f Figure 17.29c, we
represent the circuit in Figure 17.32b with the following parameters:
winding resistance),
age reactance),
= 0.002 1^2 (secondary winding resistance),
= 0 .050 H (the primar)^
= coZ,^ = 0.4 Q (primary leak
= (oZ,^ = 0 .016 ^2 (secondary leakage reactance), and
= 250 Q (mag
netization reactance). R^^îs very large and its effect is neglected. If the load draws 100 A at a power
factor o f 0.6 lagging, i.e., I 2 = 100 Z. -53.13^^ A, and the load voltage is V , = 220 Z. 0^ V, what
is the magnitude o f the needed source voltage to achieve the desired load voltage-current values in
the presence o f non-idealities o f the transformer?
ac
High Voltage
Source
Low
Voltage
Load
(a)
Load
5:1
ac
High Voltage
Source
ideal
(b)
FIGURE 17.32 Analysis of a circuit containing a practical transformer, (a) A practical transformer
for stepping down ac voltage, (b) A circuit model utilizing an ideal transformer.
So
lu t io n
Using phasors for all voltages and currents, we have

Load voltage = V 2 = 220 Z.0 V
Load current = I 9 = 100 Z .-53.13^ = 60 - ySO A
(from cos(53.13) = 0 .6 , the given power factor).
Voltage drop across the impedance
R^ + jX^ = 0 .0 0 2
+ y'0.016 Q
= (60 - y 8 0 )(0 .0 0 2 +y0 .0 1 6 ) = 1.4 + >0.8 = 1 .6 1 2 Z 2 9 .7 V
930
Voltage across the secondary o f the ideal transformer

= (1.4 + J0 .8 ) + 220 = 221.4 + yO.8 = 221.4^/0.207 V
c^1
J lOO
..
.
Since !- = --------= 5 (eiven turns ratio), we have
yV2
220
Voltage across the primary o f the ideal transformer
= 5 X 2 2 1 .4 ^ 0 .2 0 7 = 1 1 0 7 ^ 0 .2 0 7 V
Current through the primary o f the ideal transformer
= 0 .2 x ( 6 0 - ; 8 0 ) = 1 2 - y l 6 A
Current through the magnetizing inductance:
1 107Z 0
^------- = 4 .4 2 8 Z. - 8 9.8 = 0 .015 - y 4.428 A

J2 5 0
Current through the primary winding:

= (12 - J 1 6 ) + (0.015 - J 4 .4 2 8 ) = 2 3 .7 ii- 5 9 .5 4 0 A
Voltage drop across the impedance
+ jX^ = 0.05 + yO.4 Q.
= 2 3 .7 ^ -5 9 .5 4 x (0.05 + ;0 .4 ) = 9 .5 5 ^ 2 3 .3 4 V
Therefore, the source voltage is
9.55Z -23.34 + 1 107Z .0.207 = 1 1 15 .8 z l0 .4 V
Thus, the magnitude o f the source voltage is 1115.8 V, which is 15.8 V higher (to overcome the
non-ideal effects) than what would be needed for an ideal transformer.
9. COUPLED INDUCTORS MODELED WITH AN IDEAL

TRANSFORMER
When a circuit contains coupled inductors, loop analysis is a natural way to analyze the circuit.
One then solves the resulting simultaneous equations by any o f the techniques studied earlier.
Although very general and systematic, such methods have extensive mathematical operations that
obscure the essential physical properties o f the circuit. In this section, we shall present some mod
els for a pair o f coupled inductors that utilize an ideal transformer. Since the three basic prop
erties (voltage, current, and impedance transformations) o f an ideal transformer are relatively easy
to comprehend, using such models in place o f a pair o f coupled inductors helps us to more easily
understand the circuit behavior, without complicated mathematics.
931
As a first case, consider a pair o f inductors with unity coupling (i.e., k = 1, or
as
shown in Figure 17.33a. Figures 17.33b and 17.33c show two equivalent circuits, each consisting
o f one inductor and one ideal transformer whose turns ratio depends on the values o f Z,j and Z,,.
k=1
1:
o+
+
V,
ideal
(b)
FIGURE 17.33 Ideal transformer models for unity coupled inductors.

(a) A pair of unity coupled inductors, (b) A model consisting of one inductor and one ideal trans
former. (c) An alternative model.
T he proof o f the equivalence is easy. The v-i relationships for the circuit o f Figure 17.33a are given
by differential equations 17.6. We need only show that the circuits o f Figures 17.33b and 17.33c
have the same v-i relationships. Consider Figure 17.33b first. In this case, A/= JL]L-> . Using the
current transformation property o f the ideal transformer, we obtain
d
Zo .
Vj = L i\ + , h L\------ JL\L-t = L i -------- h M
'
' dt
cit
^
dt
di
dt
V
Next, using the voltage transformation propert)',
^+U = M + U
dt
dt
dt
These two equations are exacdy the same as equations 17.6. Hence, the circuits o f Figure 17.33a
and Figure 17.33b are equivalent. A similar derivation proves the equivalence o f the circuits o f
Figures 17.33a and 17.33c.
Exercises. 1. Verify the equivalence o f the circuits o f Figures 17.33a and 17.33c.
2. In Figures 17.34a and b, let Z, = 2.4 H, Z-2 = 6 mH, L j = 3 mH, and
= 0.2 H. Find
for each circuit using the equivalences in Figure 17.33.

ANS WK KS . O.Ss, 0 . 1 6s.
k=1
k=1
z in
^ --
>
L
(a)
(a)
FIGURE 17.34
E X A M PLE 17.2 2 . The circuit o f Figure 17.35a has a unity coupling coefficient. Find the band
width, the center frequency, and the maximum voltage gain.
300 Q
50 nH
M = 350nH
2450 nH
104.5 pF
(a)
soon
1 :7
14.7 kn
300 0
5 1 2 0 .5 pF
+
V
5 0 nH
V.
300 0
(0
FIGURE 17.35 Analysis of a unit)' coupled circuit, (a) A circuit containing unit)' coupled inductors,
(b) An equivalent circuit utilizing ideal transformers, (c) Equivalent circuit after reflecting load
impedances to primary.
So
lu t io n
The coefFicient o f coupling is = M l -y/^1^2 = 3501 yjSO x 2 ,4 5 0 = 1. Replacing the coupled coils
with the equivalent circuit o f Figure 17.33b yields the circuit o f Figure 17.35b. From the imped
ance transformation property o f an ideal transformer, looking into the primary we see an imped
ance 1/49 times the load impedance. Therefore, looking into the primary, we see a resistance o f
14,700/49 = 300 Q. in parallel with a capacitance o f 49 x 104.5 pF = 5120.5 pF.
Figure 17.35c captures the new equivalent circuit. Th e band-pass characteristics o f H {s) =
Vl(s)
Vsis)
follow the analysis done in Example 16.2. Consistent with the notation o f Example 16.2, we have
/?= 300//300 = 150 0 , 1 = 50 nH, C = 5,120.5 pF
CO,=Wp =
*
= 62.5 X 10^ rad/sec
V 5 0 x 10"^ X 5 ,1 2 0 .5 X 1 0 "*
Q _ ___ _ ___________
RC
150 X 5 ,1 2 0 .5 x l 0 '
\H(/
I
'io)\
'huav =
ID = 0 .5
From the voltage transformation propert)' o f an ideal transformer,
only quantities affected for the transfer function
are the maximum values:
is simply 7 times Vy So the
)34
The above example illustrates how a simple substitution o f an equivalent model for coupled
inductors can reduce a circuit to a simpler form for which the analysis is straightforward.
Exercise. Solve Example 17.22 again using the equivalent circuit o f Figure 17.33c.
ANSW'HRS: I he same as given in Example 17.22, of course.
As mentioned earlier, a unity coupling coefficient is an ideal condition impossible to achieve in

practice. It is desirable, therefore, to modify the models o f Figure 17.33 to account for a coupling
coefficient k < \. The resulting equivalent circuits are shown in Figure 17.36.
M :L
M :L
< o
+
k'L.
ideal
ideal
(b)
(c)
o
+ i.
O
+
L.:M
/YY\
< o
+
V,
k'L
ideal
ideal
(d)
(e)
FIG U RF 17.36 Four different models for coupled coils with /'< 1, cach consisting of one ideal
transformer and rwo self-inductances, (a) Coupled inductor with k < 1. (b) One model using an
ideal transformer and two inductances, (c), (d), (e) Three alternative models.
Exercise. In Figures 17.37a and b, Zj = 2.4 H,

Find Z -is) for the circuits.
A N ^ W 'E R S . 1.6>-. O .S j
= 6 mH, L^ = 5 mFI,
= 1.6 H, and k =
j = .
93
Chapter 17 iMagnctically Coupled Circuits and Transformers
(a)
(b)
FIGURE 17.37
The equivalent circuits of Figure 17.36 are derived as follows. Since
it is possible to subtract
a small inductance from Z,j such that the remaining inductance, Z,'|, satisfies the condition L
= M".
In other words, the new inductance, L' ^ together with L-, and M, forms a unity coupling system. Since
Z,'j = MÎL) = k~L^, the induaance to be subtracted is equal to Zj - Z .'j = (1 -
This inductance
must be added back in series with Z,', to obtain a model for the original coupled inductors. The mod
els o f Figures 17.36b and c result. Repe-ating the process on
yields the models o f Figures 17.36d and
e. A total o f four equivalent circuits is possible. Each consists o f two uncoupled inductors and one ideal
transformer. Clearly, when k= 1, the models in Figure 17.36 reduce to those in Figure 17.33.
As an application o f the models o f Figure 17.36, let us reconsider Example 17.10.
and Vjit) at the frequency 1 rad/sec
E XA M PLE 17.23. Find the steady-state components o f
= cos{t)u{t) V. Note that because a resistance is pres
for the circuit o f Figure 17.38, in which
ent, the circuit responses will contain both a transient and a steady-state component.
1Q
----------O ----------+
M = 2H
(a)
i
2 :5
+
V,
2F
(b)
F I G U R E 1 7 .3 8 Stead y -state analysis o f co u p led in d u cto rs u sing th e ideal tra n sfo rm er m od el.
S olu tion
,.2
For the circuit o f Figure 17.38a, k" = ------- = 0 .8 . Substituting the model o f Figure 17.36b
L| Li
for the coupled inductors in Figure 17.38a, we obtain the equivalent circuit o f Figure 17.38b.
From the impedance transformation property o f the ideal transformer,
It follows that
z . , , a ) = 0 .8 y , i z , ,, ,( ;) = 5
; j ^
= - 0 .0 8 8 9 ,Q
Using the voltage divider propert)',

V,, = ---------yOX)889-----^
I + 0.2./ - yO.0889
^ 0 .0 8 8 3 4 5 Z - 96.34'^ V
Further,
^
0.2y 7 0 .0 8 8 9
1 + 0.2./ - 7X).0889
^ ^ ^ I0 4 2 Z 8 3 .6 6 ^ V
Using the ideal transformer properc)',

V , = 2.5V^, = 0 .2 2 0 8 6 ^ - 96.34^ V
I'his result agrees with the solution given in Example 17.10.
SUMMARY
This chapter has examined the phenomenon o f induced voltage in one inductor caused by a
change o f current in another inductor. A new parameter called the tnutttal inductance (A/) between
the coils was introduced. M was defined as a constant and is present in the coupled differential
and 5-domain equation models o f the coupled inductors. As illustrated in the chapter, M can be
determined experimentally. This treatment has avoided digressing into the study o f magnetic cir
cuits, which in fact underlies a rigorous development o f the concepts in this chapter. From the cir
cuit analysis perspective, this mathematical treatment is adequate. However, for a deeper under
standing o f the physical phenomena, one must study the principles o f magnetic circuits.
O f foremost importance in analyzing a circuit containing mutual inductance is the formulation o f
correct time domain or frequency domain equations in particular, the correct signs for the
mutual terms. For this reason, considerable time was spent on the dot convention. Once the equa
tions are correctly written, we may use any o f the techniques (loop or node analysis) studied ear
lier to anal)'ze the circuit.
A proof o f M p =
= M was given that made use o f the passivit}' principle and some trigono
metric identities. With the establishment o f this equalit)', computing the energ)' stored in the cou
pled inductors follows from simple integration. The expression for the stored energ)' and the passiyitv principle then led to an upper bound for the value ot the mutual inductance, namely M <
yj^ ^2 1 he coefficient of coupling was then defined as k = M
l .
An ideal transformer was defined as a device satisfying both the voltage transformation and cur
rent transformation properties. For practical transformers, these two properties hold only approx
imately for example, over specified voltage and current ranges as well as over specified frequen
cy ranges.
Transformers have broad applicability, h'or example, in power engineering, transformers are used
to step up or step down the voltages. In communication engineering, transformers are used to
change a load impedance for the purpose o f maximum power transfer.
Although ideal transformers can only be approximated in the real world, they nevertheless remain
an important basic circuit element because a practical transformer or a pair o f coupled inductors
can be modeled by an ideal transformer and some additional R, L, and/or C elements. The use o f
such models simplifies many analysis problems and gives physical insight into the operation o f a
circuit.
TERMS AND CONCEPTS

CoefFicient o f coupling: usually denoted by /r, equal to M /yj ^1^2
Coupled inductors (coils): rwo inductors having a mutual inductance M ^ 0.
Current transformation property: for unity coupled inductors having infinite inductances, the
ratio |/|(/) : /2(r)l is a constant equal to the turns ratio
D ot convention: a commonly used marking scheme for determining the polarit)' o f induced volt
ages.
Energy stored in a pair o f coupled inductors: 0.5/.]/| +
A//,/-, joules, with the +
sign for the case where both currents enter (or leave) dotted terminals and the sign for
the case where one current enters a dotted terminal and the other leaves the dotted ter
minal.
Ideal transformer: rwo network branches satisfying both the voltage transformation and current
transformation properties exactly.
Impedance transformation property: when the secondary o f an ideal transformer is terminated
in an impedance Z(;), the input impedance across the primary is equal to {N^lN-,)^Z{s)
= a^Z{s).
Models for a pair of coupled inductors: a pair o f coupled inductors can be represented by one
ideal transformer and one inductance for the case ^ = 1. For coupling coefficients k less
than 1, the representation requires one ideal transformer and two self-inductances.
Mutual inductance: a real number, usually denoted by M, that determines the induced voltage
in one coil due to the change o f current in another coil.
r>
Chapter 17 Magnetically Coupled Circiiits and Transformers
938
71-equivalent for coupled inductors: if two inductors have one terminal in common, then the
three-terminal coupled inductors are equivalent to three uncoupled induaors (one of
which may have a negative inductance) connected in the 7t-form.
Primary: The winding (coil) in a transformer that is connected to a power source.
Secondary: The winding (coil) in a transformer that is connected to a load.
T-equivalent for coupled inductors: if two inductors have one terminal in common, then the
three-terminal coupled inductors are equivalent to three uncoupled inductors (one of
which may have a negative inductance) connected in the T-form.
Transformer: a practical device that satisfies approximately the voltage transformation and cur
rent transformation properties.
Unity coupling: coefficient of coupling k= I.
Voltage transformation property: for unity coupled inductors, the ratio |i/j(/) : V2 {t)\ is a con
stant equal to the turns ratio NÎNj =
r>
n
o
r>
939
3. For each circuit shown in Figure P i 7.3:
Problems
(a)
Express vj^t) and
in terms o f /^(r)
and i^{t) for the dot in position C.
DOT PLACEMENT, M, AND

BASIC EQUATIONS
(b)
Obtain the ^-domain equations that

contain the initial currents /'^(0)and
1. For each o f the circuits in Figure P i 7.1,
/'^(0) by taking the Laplace transform
determine the proper signs on the equations for

coupled inductors:
o f the equations o f part (a).

(c)
Repeat parts (a) and (b) with the dot

moved to position D.
(It
V9 (r) = M
M = 2H
20
(It
L9 ^
dt
~ dt
4H
3H
Observe that some o f the labelings do not con

form to the passive sign convention.
M
Figure P I7.3
M
4. For the circuit shown in Figure
P17.4;
(a) Express vj^t) and v^{t) in terms o f
/^(^) and /^(r) for the dot in position
C.
(b) Obtain the ^-domain equations
containing the initial /^(0~) and
by taking the Laplace trans
form o f the equations o f part (a).

(c) Repeat parts (a) and (b) with the
dot moved to position D.
Figure P i7.1
2. For the circuit o f Figure P I7.2, A/ = 2 H;
compute v^{t) when
= I t cos(20/)//(f) mA
for the dot in position A and then in position B.

M
Figure P i7.4
5. In the circuit o f Figure PI 7.5, if/j(^) = Atu{t)
A, a voltage
= 40//(r) mV is observ'ed.
Determine the placement o f the dots and the
value o f M. Repeat if
observed.
Figure P I 7.2
= -80u{t) mV is
940
(b)
Compute V2 {t) for /* > 0.

M
i,(t)
>
Figure P i7.5
6. Consider the laboratory setup o f Figure
PI 7.6a.
(a)
(b)
Figure P i7.7
ANSW1-:RS; (a) 4 H: (b)
The two waveforms, /j(r) and v-^{t), in

Figures P I7.6b and c are shown as dis
8. Write three mesh equations in the time
played on an oscilloscope attached to
domain for the circuit shown in Figure P i 7.8.
the circuit o f Figure PI 7.6a. Determine
Be particularly careful about the signs o f
the placement o f the dots and the value
induced voltages due to the mutual inductance.
o f the mutual inductance.
Apply the rule given by equation 17.6 if you
Now suppose the signal source pro

duces /jCr) = 2(1 -
have any doubt about the signs.
A. Find
D
(a)
i,(A)
C H E C K : Coefficients for derivative terms in
1>
t (msec)
1\ / ^ 2
Kc
loop a, ( I , , - Zp -
loop b,
(-Z p - Z j, - A/), loop c,
-1
(b)
v,(V)
t (msec)
----
-2
(c)
COMPUTING EQUIVALENT
INDUCTANCES, Z,.^(s), AND
RESPONSES FOR SIMPLE
CIRCUITS
Figure P I7.6
9. (a)
7. Consider the circuit shown in Figure P i 7.7.
to find the impulse and step respons
I'h e 2 H inductor is short-circuited. The cur
es.
rents in the coupled inductors for t> 0 are /,(r)

= 2e~ A and i^{t) = e~^ A.
(a)
Find the mutual inductance M (in

henries).
Find the equivalent impedance o f the

circuit o f Figure PI 7.9a. Use the result
(b)

P I 7.9b .
M l
M = 2H
T Y V -^ Y Y V SW
3H ^
v.(t)
M = 2.5H
30 n
2H
r m
IF
''.w
Figure P I7.11
(a)
AN SW ERS: (a)
.(t)
/Y Y V TYYV
3H ^
(D
= 1 H, (c) i^ J i) = U-'uU)
A, (d) 1 rad/sec, (e) B
= 2 rad/sec, Q = 0.5
5H
M = 2.5 H
30 0

2H
/Y Y V
P17.12.
(a)
Determine the input impedance and

input admittance.
(b)
Figure P I7.9
10. (a)
C H E C K : admittance poles at - 0 .5 and - 3

(b)
Determine the transfer function
Find rhe equivalent impedance o f the
H{s) =
circuit o f Figure PI 7.10 when

(i)
the dot is in position A
(ii) the dot is in position B

(b)
(c)
= 30(1 -
If
V, com
Use the result o f part (a) to find the
pute
using the ilaplace com
transfer function with the dot in posi
mand in MATLAB.
tion A.
(c)
Vin(s)
0.5H
Use the result o f part (b) to find the

impulse and step responses with the
dot in position A.
(d)
If v^t) = 2 cos(2f) V, find the steady-
v ,Jt )
(n
0.1 H
state value o f
/Y Y \
2H
Figure P I7.12
'y
.w o
IY Y \
13. Figure P I 7.13 shows the two ways to con

2H
Figure P i7.10
nect a pair o f coupled inductors in parallel. For

each case, find
0---
11. For the circuit shown in Figure P 17.11,

find (a) the equivalent inductance, Z,^^, o f the
coupled coils; (b) the transfer Rmction o f the
circuit; (c) the step response; (d) the resonant
frequenq' o f the circuit; and (e) the bandwidth
and Q o f the circuit.
(a)
(b)
Figure P I 7.13
942
A N SW ERS; (a)
(b)
A/-)/(A, + / . . + 2/\/)
- 2A/);
dotted terminals connected together. If it is

known that the primary has the larger induc
tance, find Z,j, Ljy and M.
14. Find the input impedance o f each circuit in
A N SW ERS: 2.6 H. 0.5 H. 0.3 H
Figure P I 7.14. Hint: Make use o f each

17. Find
found in Problem 13.
for each circuit o f Figure P i 7.17.
Does the answer depend on the positions o f

M
the dots?
Z(s)
(b)
(a)
Figure P I7.14
AN SW ERS: (a) Z,(.v) =
RL.s
-cir
(a)
(b) Same as (a) with cilHcrcnt

15. For each o f the circuits in Figure P 17.15, R
= 1 n , I j = 2 H, I , = 10 H, M = 2 H. For fig
ure P I 7 . 15a, C = 0.8 mF; for figure P i 7 . 15b,
C = 2 F. Find the values o f O) at which reso
(b)
nance occurs. For each case find
Figure P I7.17
i f
ANSW1{RS: (a) L\ -
ZJs)
Z Js )
M-
independent oi dor positions:

L . + 7-3
independent ol dot positions
(a)
Figure P I7.15
18. Find the input admittance o f each o f the
AN SW ERS: (a) 25 rad/sec, 0 12; (b) I rad/sec,
Z j j i u ) ^ 0.5 Q
coupled inductor circuits in Figure P I7.18.

M = 4H
16. A handy dandy henry counter is used to

measure the inductance o f a pair o f coupled
OZ,(s)
inductors in various configurations. After three

experiments the results are as follows: equiva
lent inductances are (i) 3.7 H after series-aiding
connection; (ii) 2.5 H after series-opposing
connection (dotted to dotted connection); and
(iii) 121/250 H after parallel connection with
6H
8H
(a)
943
A N SW ER:
C.v-(. v -!-4) + .V+2
C = 1 / 6 F ; Z;(.v) =
Cv(.v + 2)
M = 2H
OZ,(s)
22. For the circuit o f Figure PI 7.8, find

3H
4H
Zin(s) =
(b)
when
Figure P I7.18
= /?^ =
= 1 n , Z,j = Z,-, = 2 H, and
A/ = 1 H. Hint: After writing loop equations

with numbers, consider using the symbolic
AN SW ERS: (;,) , (b)

4.S
i.v
toolbox in MATLAB (or the equivalent) to
19. Find the input impedance
o f the cir
evaluate determinants in a Cramers rule solu-
cuit shown in Figure P I 7.19. Does the answer
non.
depend on the positions o f the dots?
AN SW ER: Z,(.v) = - ^
9.V- +6s+\
3.S- + 10.V + 3
GENERAL ANALYSIS OF
CIRCUITS WITH COUPLED
INDUCTORS
23. For the circuit shown in Figure P i 7.23,
= 1
AN SW ER;
(a)
M-
.v
V.
If a dot is placed at position A, com
pute the zero-state response.
Li + Ljj
(b)
independent o f the dot positions
If a dot is placed at position B, com

M= 1
20. Find the zero-state response

input
to the
V for the circuit o f Figure
P17.20.
2H
2H
24. The switch S in the circuit o f Figure PI 7.24

is closed at f = 0. If
= 10 V, find
Figure P i 7.20
21. For the circuit shown in Figure P I 7.21,
find
(j) and the value o f C (in F) that caus
es resonance to occur at
oj
= 2 rad/sec.
v.(t)
Figure P i7.24
25. If a 900 Q resistor is connected across the
F igu re P i 7 .2 1
secondary o f the circuit o f Figure P I 7.24 and
28. For the circuit shown in Figure P i 7 .28,
= 10 V, find the voltage gain
compute the zero-state response for
V2(.v)
G ,{s) =
and then compute Vjit), t > 0. Use the formu
C H EC K :
(a) and (b).
mesh equations and solve by
Cramers rule.
(d)
9000.V
C^.=
= 30e-^',Ki) mA
/^^(/) = 60(1 - e~^û{t) niA. Hint;
Apply linearit)' to the answers to parts
las developed in Example 17.8 or, alternatively,

write two
= 30u{t) mA
(a)
(b)
(c)
Suppose Vf^Q~) = 30 V and all other

initial conditions are zero. Find the
( a + 500)(5 + 9000)

PI 7.26.
(a)
Suppose the circuit is relaxed at r = 0~.

If
= \2n{t) V, find the time con
stant o f the circuit and
(b)
Suppose
(c)
Suppose
0.625 H
-L
for / > 0.
= - \lu{-t) V; find v^{t)
<^10
0.1 F
for r > 0.
= - 6 u{-t) + Guit) V;
find V2 U) for /^> 0. Hint: Use lineari

ty.
Figure P I7.28
AN SW ER rd): v^ Jt) = 1.6<---'//(r) - GAe-^'u{t)
\
200 0
M = 0.5 H
o
+
vJt)
0.2 H'
v,(t)
29 . In the circuit shown in Figure P I7.29, all

initial conditions are zero at r = 0~.
(a)
5H
o
If
= 2u{t - 1) A, compute the
response,
Hint: Use the result
o f Problem 19.
Figure P i7.26
(b)
Find /,(r).
(c)
Find /,(r)
i j t ) = 100(r) niA.
(a)
If a dot is placed at position A, com

(b)
If a dot is placed at position B, com

IH

M = 0 .i H
o+
v Jt)
ANSWl-R: (b) /,(/) = 2(1 -
0.1 H
B
50 mF
Figure P i7.27
A N S W E R (b ):
')Ma-1 ) A
\0.2 H ''
= 0 .1
30 , Consider the circuit shown in Figure

P I 7 .30 , having zero initial stored energ)'. Let
/?! = 9 a
sin ( l()/)//(/) \
(a)
(b)
Find
F in d Z J ^ ) = /? + Z .,W .
(c)
Find
945
(d)
Y = ..........
H = det(Y)/det(A)
F in d G 2 = ^
simple(H)
(e)
Find Gy =
(f)
Find the response,
to
input
v j t ) = \6 .8 le^^û{t) V.
(b)
(^ + 1X^ + 3)
Find the impulse response, ^(f), again

using the symbolic toolbox in MAT
LAB:
R.
+
V,(t)
'.w Q
-3 s
C H EC K : H (s) =
syms t h
200
h = ilaplace(H)
0.2 H
(c)
ZJs)
Using
the
residue
command
or
ilaplace, show that the zero-state

response, V2 (f), is
Figure P17.30
31. Suppose that
= 3yf20cos{t)u(t) A
V2 (t) = 18A12e~^û(t) - 10.06e~û(t) +

9 co s(r-1 5 3 .4 ) W
and the initial stored energy is zero in the cir

cuit shown in Figure P17.31.
(a)
(d)
Compute the transfer function,
Identify the steady-state and transient

responses.
using the symbolic toolbox in MAT-
1.5 0
LAB as follows:
(i)
1H
Write down the mutual induc

tance equations in s, a set o f two
equations in
Kj,
Kj, / p
l2-
Figure P I7.31
Rewrite each equation in the

formO = ? X Kj + ? X /j + ? X K2 +
32. Consider the circuit o f Figure P I 7 .32.
^ ^2(ii) Use KCL and Ohms law to write
output volt^ e is zero. The input voltage is
the relationship in the form ? x

= ? x Vi + F x / j .
(iii) Similarly construct the terminal
Before the closing of the switch at r = 0, the

v ^ (0 = 1 2 V 2 V .
(a) Write two loop equations and find the
matrix Z(s) such that
constraint equation of the form 0

= ? X V^2 + ? x / 2.
Z(s)
(iv) Write the preceding four equa
h (s y
? ? h is y
h (s )
? ?
h is )
Vinisy
0
tions in matrix form, Ax= b.

(v) Use MATLAB as follows:
(b)
Using Cramers rule find / 2 (s). Then

determine K2W.
syms s lin A b Y H
A = Coefficient Matrix in part 4
% Y is the coefficient matrix A with the column
corresponding to V2 replaced by the % vector b
with /y^ (s) set to 1.
(c)
Find V2 (f).
C H EC K : V2 (f) =
946
+
V,(t)
33. Consider the RLC circuit shown in Figure
3 5 . Consider the circuit o f Figure P 17.35, in
P 17.33.
(a)
Find
which Z, = 0.9 H , Z2 = 0 .4 H , Af = 0.2 H , and
(b)
andZ.W .
Find the poles and zeros of Z-JJ). Find

the impulse response.
(c)
T?2 = 80 Q. Suppose
V.
(a)
Compute the currents in the two
(b)
Write two differential equations for
Find the zero-input response if
inductors at ^ = 0 .
= 10 V with all other initial conditions

zero.
(d)
the coupled circuits in terms o f /j(r)
Find the zero-state response if

cos(100/)(/) A.
(e)
= 120(-r) - I 2 0 u{t)
and
(c)
Take the Laplace transform of the
Find the complete response of the cir
equations computed
cuit.
accounting for the initial condition.

(d)
Find V2(t) for t > 0 .
(e)
Now suppose
V, Find
j
/2 5 0
in part (b),
= \2u{-t) for / > 0.
-----
L (t)(
S
0 .2 m F
0.2 H '
L
v jt )i
Figure P I7.33
34. Consider the circuit o f Figure P I 7 .3 4 , in
which
= 0.9 H , 2 = 0-6 H,
0.2 H , and
/?2 = 100 Q. Let
= 120 V. The switch S
has been closed for a long time (i.e, the circuit
has reached equilibrium) and is then opened at
f= 0.
(a)
After the switch is opened, write two

differential equations for the coupled
circuits in terms of /j(t) and ijit).
(c)
Take the Laplace transform o f the

equations
(d)
36. The two circuits shown in Figure P I7 .3 6

have the same transfer function
Compute the currents in the two

inductors at ^ = 0 .
(b)
Figure P17.35
computed
in
part
(b)
accounting for the initial condition.

Find V2 {t) for f > 0.
Find the values o f the three uncoupled induc

tances assuming A / = 2 H. Hint: Use the Tequivalent circuit of Figure 17.18 with the
indicated labeling in Figure P I7.36a; node 3 is
the common terminal of the coupled inductors.
94'
Figure P I7.38
.ANSWI-R: k = 0.2S
30
fY Y Y jy Y V
39. Consider the circuit o f Figure PI 7.39. Find
the value o f Csuch that the voltage gain is zero

at
20
CO
= 3 3 3.33 rad/sec. Use the 7i-equivalent
circuit o f Figure 17.18c.

no coupling betw een inductances
(b)
Hh
k=0.5
Figure P i7.36
20 0
AN SW ER; L^.= 2 H. A., = 3 H, and
= 1H
COUPLING COEFFICIENT
PROBLEMS AND ENERGY
CALCULATIONS
40 0
0.27 H
0 .1 2 H .
200 0
Figure P I7.39
A N SW l'R: 3 3 .33 ul-
37. Consider the circuit in Figure P i 7.37. Find
40.
the value o f Csuch that the voltage gain is zero
find the coupling coefficient k such that at co =
at
CO
= 333.33 rad/sec. Hint: Use theT-equiva-
lent circuit.
For the circuit shown in Figure P I 7.40,
10"^ rad/sec, the voltage gain
is zero.
Use the Ti-equivalent circuit o f Figure 17.18c.

0.5 pF
k=0.5
20 0
40 0
0.27 H
0.12
0.6 mF
T l
Figure P I7.37
AN SW ER: 0.1 ml-
Figure P I7.40
AN SW ER: k = 0.946

determine the coupling coefficient k such that
41. Consider the circuit o f Figure P 17.41.
at
CO
=10"^ rad/sec, the voltage gain
(a)
Find the mutual inductance M so that

the coupling coefficient k = \.
(b)
Obtain the transfer function using the
is
zero. Hint; Use theT-equivalent circuit for cou

pled coils and recall that a series LC behaves as
formulas o f Example 17.8. Make use
a short circuit at w = \NLC.
o f your answer to part (a).
9i.S
(c)
Q, and H
Obtain values for (u ,
40 n >V
(s)
Figure P i7.42
43. Let /(r) = 2 cos(10r + 3 5 ) A in the circuit
Figure P i7.41
in Figure P i 7.43. Find the maximum instanta

0) .
ANSW l-K: (c)

- 2.5 rad/scc,
(I) = 10 rad/sec./-/= 17.H8
= -^ neous stored energy.
42. In die circuit shown in Figure P 17.42, the

coupling coefficient k has been adjusted for the
circuit to be resonant at OJ = 10 rad/sec. /?^ = 0.1
Q., L^=2 H, and L-, = \ H while C = 20 niF.

(a)
Find k.
(b)
Construct the transfer function

H{s) =
Figure P i7.43
K ..,(5 )
44. In the circuit o f Figure P 17.44, Z.j = 0.8 H,

and then determine its poles and
(c)
zeros. Hint: Use the T-equivalent for
Z.2 = 0.45 H ,M = 0.175 H ,/?= 12 ^ Ify/r) =

30cos( 1Or) V, find the maximum instantaneous
coupled coils to simplify calculations.
steady-state power delivered to R when
For a sense o f the circuit behavior,
(a)
A dot is in position A.
obtain
(b)
A dot is in position B.
magnitude
frequency
response (using MATLAB) o f the cir
fv c t
cuit for 1 rad/sec < (U < 20 rad/sec.

Observe the band-pass characteristic
behavior.
(d)
To obtain approximate values o f

2 ^^^ctor H{s) =
Q>
where
H^{s)=
*
Figure P i7.44
A+ 0.05
Note that |//, (/10)| 2

2 ^ order
A N SW l'R: (a) 48 w atts, (b) 27 w atts
45 . Consider the coupled inductors shown in

1 so that the
characteristic
o f //2W
Figure P I7.45.
(a)
If /, = 2 A and
approximately characterizes the band

pass behavior.
= - 3 A, find the
stored energy.
(b)
If /j = 2 A, find the value o f I 2 that

will minimize the stored
A N SW ERS: (a) ^ 2 . , (d)

Q a 200.
= 100
= (^-05 rad/scc,
(c)
energ)' W. What is the value o f

Plot W as a function o f ^ for - 3 < ^
< 3 A using MATLAB or the equiva
lent.
(d)
Determine the coefficient o f coupling k.
currents are /j(0) = 7| and /-,(0) = Ij- Also sup

pose /j(7) = 0 =
Between 0 and 7', /,(a)
and i-y(t) have arbitrary waveforms. Show that

the energy delivered by the inductors to the
current sources during the interval 0 < t < T
4H
equals ( 0 .5 7 ,7j2
9H
0 .5 ^ 2 ^ +
dots are now moved to A-C, show that the final

result is (0.57.j7,2 + Q.5L,l2^ Figure P i7.45
48. In contrast to Problem 47, where the cou
pled inductors delivered power to the sources,
ANSWHRS; (a) 6 6 .S j . (h) 2/3 A and 6 J, (d)

U.S
we can connect two resistors 7?, and Rj to the
46.
inductors and let the current decrease exponen

tially to zero. In particular, for the circuit o f
For the coupled inductors o f Figure

= B
Figure P I7.48, it is possible to show that the
cos(o)/ + q)). What is the period T in terms o f
energ)' delivered by the inductors to the resis
P I7.46, /,(r) = A cos(o)/ +6) and

CO?
Regardless o f whether the dots are at A-B or
tors during 0 < ^ < x
the coupled inductors is zero; i.e., show that
equals (0.57,,7,^ +
A/7,7-,) depending on the position
A-C, sliovv that the average power delivered to
o f the dots. Demonstrate the validity o f this

assertion for the specific parameter values and
-Lii\ ( 0 +
initial currents 7.j = 10 H, 7-, = 2 H, A/ = 3 H,

/?, = /e, = 1 Q., /,(0) = 1 A,uid /2(0) = - 3 A,
with dots in position A-B, as follows:
(a)
Calculate the stored energ)^ at r = 0.
(b)
Calculate /, (r) and
(c)
Evaluate the integral
for / > 0.
Hints:
1.
+V'2/2)^//
Write down the coupled inductor

equations in the time domain with
and comparc with the result o f part (a).
for dots in B and C.

2.
Recall that
1 J
= f (vi/, + v^h)dt\
CHECK.- (a) 5 J, (b) /,(r) = e Ui{t) A and i^{t)

= -?>e-u{t) A
substitute and simplify.
JL
,{t)
L,
L,
R, >
v,(t)
L,
v,(,) 0 1 , ( 0
Figure P I7.48
Figure P I7.46
49. Mr. A claimed he had construaed a pair o f
4 7 . Reconsider the circuit o f Figure P i 7 .4 6
A7= 9 H. Rebut his claim by showing a specific set
with dots at A -B. At r = 0, suppose the initial
o f (7,, I-,) for which the stored energ\^ is negative.
coupled inductors with Z, = 10 H, L, = 8 H, and
950
ANALYSIS OF CIRCUITS
CONTAINING IDEAL
TRANSFORMERS
vJt)
50. In the circuit o f Figure P i 7.50,
R^=\50 Q, /?, = 600 D .,R^=\2 kQ,

and v,(/) = 5 yj2 c 0 s{(0 t) V,where O)
= 200071 rad/sec.
(a)
mum power transfer to Rj^.

(b)
Figure P I7.52
Find the turns ratio, Wj : ft-, for maxi

Given your answer to part (a), find
ANSW l-R; (a) Z , = 10 k il, Z, = 500 k il: (h)
vît) and v^ Jt).
0 .625. 0.1, 0.1, 0.0 0 6 2 5 ; (c) 5/1, -50/1, (d)
(c)
Given your answer to part (a), find
0.5
and /2(r).
Find the power delivered to the load.
53. Repeat Problem 52 for ?n = 20 and n = 5.
(d)
CH EC K :
0.2
= 300 kH, R =
10 k n , R^ = 100 Q, ti = 0.2, and m = 0.1.

C H E C k 'z , = 80 n ,
= -0 .0 1 3 3 3 3
55 In the audio amplifier circuit shown in

Figure P i 7.55, both transformers are ideal,
R^ = 50 Q, and v^.(0 = O.1V 2 cos(w/) V.
The 4 Cl resistor represents a loudspeaker load.
Figure P I7.50
(a)
Find the turns ratios a and b such that

the average power delivered to the 4 Q
51. Repeat Problem 50 for R^ = 400 Q, R^ =

1200 D.,R^ = 5 iX and
(/) = 10>/2 cos(cot )
V, where oj = 20007T rad/sec.
loudspeaker is as large as possible.

What is the maximum power?
(b)
The turns ratios are those determined

in part (a) for maximum power trans
52. In the circuit o f Figure P I 7.52, Rj = 100
fer. Now suppose that a loudspeaker
n , R = \ 0 kH; R^ = 300 kI2, and w = n = 10.

Compute Z j and Z ,.
(a)
with 16
(b)
Compute the voltage gains
resistance is used in place o f
the 4 Q speaker. What is the power

consumed by the 16
speaker?
Vjlvxy v-ijv2 , and

(c)
Com pute
the current gains
(d)
Compute the power gain.
_ V3/3
po w e r
Vi/
l'//i
phono
input
cartridge
transformer
Amplifier
output
loud
transformer
speaker
Figure P I7.55
A N S W E R : (a) 2 0 , 5, 100 W ; (b) 6 4 W
O '
95 1
o
56. The circuit shown in Figure P I 7.5 6 crude
such, it can be represented approximately by an
ly models an audio amplifier circuit. Set
ideal transformer in conjunaion with some
20
inductances and resistances. This problem illus
trates the use of such a model for analysis pur

poses.
K = .
Sw/
Each 8 Cl resistor models a tweeter, and each 16

n resistor models a woofer. Suppose the left
A certain calculator that normally uses four 1.5
and right speakers each consume 80 watts of
V batteries comes with an adapter whose
average power. Determine:
approximate
(a)
w
P I7.57a, where
average power delivered to the speak
and R = 200 Q.
ers is as large as possible.

(b)
(c)
(d)
(e)
(f)
The
voltage
across
and
(a)
is shown
= 0.9 H,
in
Figure
= 9.6 H, = 20,
If the calculator is not connected to
current
the adapter output, but the adapter is
through (rms values) each woofer and
plugged into a household ac outiet of
tweeter.
110 V ^ , what average power is con
i and then (rms values).

The power delivered by the dependent
sumed by the adapter?

(b)
If the output o f the adapter is acciden
source.
tally short-circuited, what are the
The power delivered by the independ
magnitudes (rms) o f the ac current
ent source.
drawn from the wall outlet and
The power gain o f the circuit, i.e., the

ratio of the power delivered to the
O '
model
The turns ratios a and b such that the
through the short circuit?

(c)
If a typical load, represented by a 50 Q
speaker load to the power delivered by
resistor, is conneaed to the output of
the independent source.
the adapter, what are the approximate

Right
Speaker
magnitudes (rms values) of

the voltage across and the
current through the load?
(d)
Source
If the
adapter
is
'O
treated as a pair of coupled
O '
inductors as shown in
Figure P I 7 .57b, deter
L2> and M.
mine the parameters Zj,

Figure P I7.56
O
C H EC K : i = 4 . . ^ ^ = 8 V ^ ,
Pl.m = 27.778
57. Today almost every electronic
O
gadget comes with its own ac

adapter. An ac adapters main func
tion is to step down the household
ac voltage o f 110
to, say, 3
rms or 6 V__
rmsbefore conversion to
a low dc voltage by the use of a rectifier. An
adapter is basically a (practical) transformer. As
O
O
AC adapter as
a pair of coupled
inductors
(b)
AC Adapter as a
practical transformer
(a)
Figure P i7.57
952
AN SW ER: (a) 0.154 warts; (b) 0.279 A, 5.59

A, 15.6 watts; (c) 4.98 V, 0.0 9 9 6 A; (d) 10.5 H,
(c)
24 mi l, 0.48 H
=10/^^) V.
Find the steady-state response to
= 20 cos(20r) V using the phasor
method.
(d)
P I 7.58 with dots in positions A-B.

( b)

Compute the step response.
(c)
If the capacitor is initialized at
(d)
Compute the zero-state response to
(a)
Repeat steps (a) through (c) with the

dots in positions A-C.
lO m F
1 :10
lo n
= 16 V, find the zero-input response.

0.1 F
ZA^(t) = 20 cos(2t) V. Identify the tran

sient and steady-state responses.
Figure P i7.60
Repeat steps (a)-(d) with the dots in
(e)
position A-C.
61. Figure PI 7.61 contains a linear circuit with

a pair o f coupled inductors. Suppose k = 1
(unity coupling) and
i l = ^ = = O.I
M
ideal
L2
The circuit is in the sinusoidal steady state and
Figure P I7.58
has known phasor values V| = 1 V and I| = j

A.
(a)
former) and 0) = 10 rad/sec, find the
P 17.59.
(a)
If A'f is infinite (i.e., an ideal trans
magnitude o f the current ratio
(b)
Compute the step response.
Compare your answer with the value
(c)
Compute the zero-state response to
= 0.1.
= 26 cos(2t) V. Identify the tran
ANSW ER: 0.1
sient and steady-state responses.

(b)
1:10
10 rad/sec. Hint: Use the phasor equa
20
v(s)
Repeat part (a) for M = 10 H and to =

tions in section 7 to calculate the cur
1F
rent ratio I^/Ip

AN SW ER: 0.09
0.01 F
Figure P I7.59
(c)
Suppose the transformer dots are in positions
A-B.
(a)
H {s) =
(b)
C o m p u te
the
V,U)
zero-state
response,
Repeat part (a) for A/ = 10 H and (O =

1207T rad/sec.
A N SW liR: 0.0 9 9 7 3 5 . praciicall\' the same as

for ^/ = 0 . 1
953
Consider the circuit shown in Figure
64.
Rest of Circuit
PI 7 . 6 4 .
(a)
Find the output impedance,

as
a function o f the turns ratio a and the
resistance R.
(b)
Find the Thevenin equivalent circuit

in terms o f a, R, and
(c)
Figure P I7.61
and ^ = 10.
Compute the transfer function H{s) =
Compute the transfer function. Note
(d)
locations of the transfer function.

From the pole locations determine the
the
impulse
and
step
responses.
(c)
the transient and steady-state responses.

C H EC K :
W ( 0 = ( - 3 . 2 2 9 8 f ^ ^^^^' + 5 . 1 1 8 1 f - - - " " ' -
and Q^. Is
Find
Given the parameter values of part (c)

with
= 3^/io sin(VIO 0 V, find
the zero-state response,
Identify
cally damped, overdamped, or under

Com pute
for v-Jt) = 15u(t).
toolbox o f MATLAB.
(e)
t)'pe (impulse/step) o f response (criti

damped).
Compute
Hint: Use ilaplace in the symbolic
there are no finite zeros. Find the pole
(b)
and a = 2. Suppose the
with Z = 0.1 H and C = 0.1 mF.
62. In the circuit o f Figure Pi 7.62, /?. = 0.2 Q

(a)
Let /? = 13
output is terminated in a series LC
high? If
highQ^,, find approximate values for

(0^^,
and the half-power frequen
1. 8 8 8 3 c o s ( 1 0 V i 0 r ) -h 9 . 0 9 4 8 sin(10>/i0r)]//(O
cies OJ, 2- Recall equations 16.47,
Q =
Plot the magnitude response o f the
transfer
function
to
verif}'
your
answers.

find
circuit.
63. Repeat Problem 62 for R^= 2 Cl. This will
be a low-Q^ case. Recall equation 16.47:
e
lo ,
'
2sin(^^)
A 6a = arcsm
and
to PJ
and for exact answers to part (c) recall equa

tions 16.49 through 16.52.
and the Thevenin equivalent
9S4
Figure P I7.67
Figure 1 17.65
AN SW ER: Vv(.)
(H =
^th (
it
68. hi the circuit o f Figure P I7.68,

and
144
/?^ =
^(/)= 10>/2cos{l0-^/) V.
(a)
73.5
Suppose
= 25 H,
and
= 1 (equivalent to a
straight-through connection with dc
^ "I-------
isolation).
Find
66. In the circuit shown in Figure P I7.66,
^2, find
ijjit), and /^(r).
(b)
maximum
^ 2
power transfer to
= 5 cos(2/) V. Find
256
For this value o f
and |V^ |.
Suppose
= 1 (equivalent to a
straight-through connection with dc
isolation).
Find
for maximum
power transfer to /?y. For this value o f

(c)
Suppose
and
^^re both adjustable.
Find a set o f values o f
Pi
and a-j so that
is greater than the values com
puted in parts (a) and (b). Find this

(d)
ideal
transformer
value o f PL.ave'
What is the maximum value o f P Lave
if//, and a-^ are both adjustable?
Figure P I7.66
67. In the circuit o f Figure P 17.67, suppose
= 10 n ,
(a)
(b)
= 40 U, and R^ = 2 0..
Figure P I7.68
ANSWL'.R: (a) 0.1479 watt, (b) 0.64 watt, (c)
- 2,
for
= 0.1463. 0.87671 watt, (d) 1 watt
If the load is identified as the termi

nals A-B, find a for maximum power
69. Consider the network configuration in
transfer.
Figure
PI 7 .6 9 .
The
source
voltage
v^(t) = y / l C 0 S ((0 / ) V, where to = 10^ rad/sec.
If the load is identified as terminals CD (i.e.,

is the load), find a for max
imum power transfer.
(a)
Design a lossless network N such that

maximum average power is transferred
to the 10
load at to = 10^ rad/sec for
Chapter 17 Magnetically G>upled Circuits and Transformers
955
O '
each o f the cases cited below.
Case 1 : N consists of one inductor L

and one ideal transformer. You will
need to determine the configuration
of the inductor and ideal transformer
as well as the values o f L and the turns
ratio a.
Figure P I7.70
Case 2 : N consists of a pair of coupled

inductors, in which case you must
determine the values o f
L^, and M.
Case 3: N is simply an LC network.

(b)
For each case compute the transfer
71. The front end o f a radio receiver uses a

band-pass circuit as shown in Figure P I 7 .7 la
consisting of a capacitor and coupled coils
whose mutual inductance is determined by the
fimction
tap position. For analysis purposes, this circuit

is modeled using the equivalent circuit of
Figure P I 7 .7 lb. Since the circuit is connected
Vsis)
Plot the magnitude response for 10^ s

(0 10^ rad/sec using a semilog scale.
to an RF amplifier, we will assume that the

amplifier input impedance is infinite to avoid
Which circuit has the better perform
loading. Our goal is to obtain the frequency
ance over the range o f frequencies and
response o f this circuit and verify that it is a
why?
band-pass circuit. Assume C = 100 pF and
500 a
1 kO
^
O '
''.W
Ak. Ak.
(p
Lossless
N
10 nF
(a)
Compute the coupling coefficient k.
(b)
Find the transfer fimction H{s) =

Suggestion: Use one o f
100
the models of Figure 17.33 for the

coupled coils.
(c)
Figure P I7.69
70. Consider the resonant circuit shown in
Figure P I7 .7 0 containing a pair of coupled
inductors. Suppose
O
= 20
Zj = 1.5 mH, M
= 3 m H, Z.2 = 6 m H , and R i = 320 Q.

Find the coupling coefficient k.
(a)
(b)
Model the coupled inductors using

one o f the equivalent circuits in Figure
17.33. Find the impedance yj(f) and
then Kj(/o)) as a function of C

(c)
O
o
o
< J
Determine the value o f C that leads to

a resonant frequency of O)^ = 10
O
(d)
krad/sec.
Find the voltage gain at resonance;
i.e., find
J| ,
and Q of
the circuit.
O '
W
Compute (O^,
Remark: We could have designed a parallel res

onant circuit to achieve the frequency-selective
property of the circuit in Figure P I 7 .7 la.
However, because of the mutual inductance, we
can achieve a much larger voltage gain than
that achievable by a simple parallel RLC. This
becomes clear when one uses a model for the
coupled inductors o f Figure P I 7 .7 lb; this
model contains an ideal (step-up) transformer
having a turns ratio of 1:10.
(C H EC K : 600 nH) in the model con

structed in part (b) can be neglected.
Find
approximate
|//(/(0^^^)|,
values
of
and Q for the voltage
gain H{s) =
Plot the magni
tude frequency response for 2 M Hz <

/ < 16 MHz.
Hint: Reflect the voltage source and 1.2
M= IOmH
resistance to the right side o f the ideal trans

former in the equivalent circuit model o f part
(b). That is to say, replace the circuit to the left
o f the ideal transformer by its Thevenin equiv
alent.
(b)
(d)
Figure P I7.71
Including the effects o f
and the
6 00 nH shunt inductance, use a cir

A N SW ERS: (a) 1: (c)
X I t P nitl/scc,
cuit simulation program to find more
= lo " rad/scc. li.. =
= SO, ; i i k1
accurate values for
= 10
and Q o f part (c) and compare.
APPLICATIONS
72. Recently, coupled coils have been manufac
tured on printed circuit boards by printing two
spiral coils on opposite sides o f a board. Such
coils are called coreless PCB transformers and
can be used to effect electrical isolation as well as
energ}' and signal transfer. This problem illus
trates the use o f the models developed in this
chapter for an approximate analysis o f the fre
quency response o f coreless PCB transformers.
AN SW ERS; k - 0.5. (b) l,slmni= 600 n il.
Figure P i 7.72 shows an equivalent circuit o f a
6 .0 8 6 Mrad/:,CC. 7^,., = 3.333 .Vlrad/j,ec. Q =
PCB transformer for which Z., = /-2

= Rj = 1.2
which rep
18.26.
l.sfrics= 4^0 nH, turns ratio a = 2, (c)
= 9.13
M = 300 nH, and
resent the winding resistances. Suppose tiie sec
73. Figure P I 7.73 shows two series RLC cir
ond coil is terminated in a capacitance o f C =
cuits physically isolated but magnetically cou
600 pF in parallel with a resistance of /?y =
pled. The input is a sinusoidal voltage sotirce

cos(cor). We will investigate some
2000 a .
steady-state voltage gains. This investigation

(a)
Find the coupling coefficient k.
(b)
Construct a model for the coupled
the coupled inductors represent a magnetic
inductors using Figure 17.35 of the
coupling between an implant and an external
text.
For a first analysis, assume that the
circuit.
(c)
effect o f Rj^ and the shunt inductance
has apphcation to biomedical implants, where
V;
(c) Suppose L = 1 mH, R = 5 O., C =

0 . l/.< F, 0) = 10 0 krad/sec, and
= 0 .1
V. Find the value o f the mutual induc

tance M needed to maximize
Figure P I7.73
If the coupling coefficient ^ = 0 (or Af = 0),
then the steady-state output magnitude is
= 0. Further, as shown in Chapter 16, [V qI =
Ql
V^. As k increases from zero to its maximum
value o f 1, |Vq| decreases monotonically,

whereas lV^,J increases to a maximum value
and then decreases. Exact formulas for several
critical values will now be derived.
(a)
|VJ.
What is the maximum value o f |Vout\'

(c)
Q = 20. M
= 0 . 05
mi l .
r nttijnux = 1
74 . Example 17.7 provided an approximate

analysis o f a car ignition system. This problem
asks for a SPIC E simulation o f the example.
Reconsider the circuit o f Figure 17.12b with
out the switch. Show that the initial current
through the primary o f the coupled inductors is
Show that if the rsvo series RLC cir

cuits in the primary and secondary are
identical, i.e., Z,j = Z2 =
C,
2nd
ANSWTRS:
Cj = C2 =
/^l(0~) = 12 A. Since the secondary is open-circuited, there is no initial current there.

Generate a SPIC E simulation for 0 < f < 1
msec. Verify that |^'2(^)|,^v - 36 x 10^ V.
7 5 . In the car ignition circuit o f Figure 17.12b,

suppose the resistance and capacitance values
are changed to 2 Q and 0.25 uF. Find approxi
then
mately the maximum voltage appearing cross
kQ^
'out
1+ (A Q )and
76. Again consider the car ignition circuit o f
*c\
i+{kQf
(b)
the spark plug.

A N S W E R ; 3 6 . 0 0 0 \'
Show that if k is the only adjustable

parameter and Q > 1, then
max
'out
when
k=~,
and at this degree o f coupling
Figure 1 7 .12b. Suppose the resistance is 2

= 1 i-tF, and
= 10 mH; find the approximate
value o f the mutual inductance M to achieve

= 38
10^ V. W hat is the approxi
mate minimum corresponding value o f L j, the

inductance o f the secondary o f the coil? Verify
your circuit operation using SPICE.
C H A P T E R
Two-Ports
THE AMPLIFIER: A PRACTICAL TWO-PORT

An actress speaks into a microphone. Speakers instantly replicate her voice, which resounds
throughout the auditorium. W hat happens between the speaking and the hearing? A microphone
produces a voltage signal that changes in proportion to the tenor and loudness o f the voice o f the
actress. Amplifiers magnify this changing voltage signal perhaps a hundred or a thousand times to
drive speakers whose cones reverberate in proportion to the changing voltage signal. The cones
then cause the air to vibrate intensely, also in proportion to the tenor and loudness o f the actresss
voice. Her words become heard by thousands because o f the amplifier.
Amplifier circuits are found in instrumentation and a huge number o f appliances. In radios, radio
frequency amplifiers first magnify signals from an antenna. Special circuitry then extracts the
audio portion from these antenna signals. Other circuitry amplifies the audio to drive speakers.
Video signals from a video cassette recorder are amplified by special circuits for connection to a
monitor. Amplifier circuitry enhances signals coming from sensors in various manufacturing
processes. Repeater circuits, among other things, amplify phone signals whose magnitudes have
attenuated during microwave transmission. There are a large number o f other applications o f
amplifiers.
From the preceding discussion, one can surmise that an amplifier circuit has an input signal and
an output signal. This configuration is represented by a device called a two-port that has an input
port for the input signal and an output port for the output signal. The following figure represents
the idea (one o f the homework problems asks for an analysis o f this amplifier):
960
Chapter 18 Two-Ports
TRANSISTOR AMPLIFIER CIRCUIT

FOR MICROPHONE
Input
Signal
Output
Signal
Often the circuit berween the ports is highly complex. This chapter looks at shorthand methods
for analyzing the input-output properties o f two-ports without having to deal directly with a pos
sibly highly complex circuit internal to the two-port. The chapter will provide a variet)- o f meth
ods for analyzing amplifier circuits.
CHAPTER OBJECTIVES
1.
Provide a general setting for one-port analysis.
2.
Extend one-port analysis to the analysis o f two-ports.
3.
Describe the input-output properties ot two-ports in terms of four sets of characteristic

parameters: impedance, or z-parameters; admittance, or ^'-parameters; hybrid, or hparameters; and transmission, or r-parameters.
4.
Develop specific formulas for the input impedance, output impedance, and voltage gain
o f two-ports driven by a practical source voltage and terminated by a load impedance.
5.
Introduce and interpret the notion of reciprocity in terms o f the various two-port parameters.
SECTION HEADINGS
1.
Introduction
2.
O ne-Port Networks
3.
4.
5.
Tw o-Port Admittance Parameters

Parameter Analysis o f Terminated Two-Ports
Two-Port Impedancc Parameters
6.
Impedance and Gain Calculations o f Terminated Two-Ports Modeled by zParameters
%1
Chapter 18 'Iwo-Ports
7.
Hybrid Parameters
8.
Transmission Parameters
9.
General Relations am ong Two-Port Parameters
10.
Reciprocity
11.
Parallel, Series, and Cascaded Connections o f Two-Ports
12.
Summary
13.
Terms and Concepts
14.
Problems
1. INTRODUCTION
Figure 18.1 shows a general one-port, whose two terminals satisfy the propert)' that for any volt
age Kj across them, the current entering one terminal, say, /,, equals the current leaving the sec
ond terminal. A resistor is a one-port: the currcnt entering one terminal equals the current leav
ing the other terminal. A capacitor and an inductor are also one-ports. K general one-port contains
any number o f interconnected resistors, capacitors, inductors, and other devices. In a one-port,
only the relationship between the port voltage and current is o f interest. For example, the port
Cdv^Jdt
voltage and current in a resistor, capacitor, and inductor satisfy the relationships
- Iq and Ldi^ldt=
respectively. Practically speaking, one-ports are macroscopic device models
emphasizing input-output properties rather than detailed internal models. Thevenin and Norton
equivalent circuits determine one-port models when only a pair o f terminals o f a network is o f
interest.
O+
V.
One
Port
O+
V,
Two
Port
-O
+
V,
-o
(a)
(b)
FIGURE 18.1 (a) General one-port, (b) General two-port.

A two-port is a linear network having two pairs o f terminals, as illustrated in Figure 18.1b. Each
terminal pair behaves as a port; i.e., the current entering one terminal o f a port equals the current
leaving the second terminal o f the same port for all voltages across the port. Coupling networks
such as transformers have two pairs o f terminals, each o f which behaves as a one-port. Hence,
transformers are two-port devices. In modeling a rsvo-port, one must define a relationship among
four variables. Different groupings o f current and voltage variables lead to different t}^pes o f char
acterizing parameters. For example, adm ittance parameters (termed ^-parameters) relate the two
(input) voltages,
and V-,, to the (output) port currents, /j and 7^. Impedance parameters
(termed 2-parameters) relate the two-port (input) currents to the two-port (output) voltages.
Other t}pes o f parameters investigated in this chapter are hybrid or A-parametcrs, and trans
mission or /-parameters.
962
Chapter 18 Two-Pons
Source One-Port
FIG U RE 18.2 A general rwo-port connccced to a source one-port (represented by a Thevenin equiv
alent) and terminated in a load one-port (represented by a Thevenin equivalent impedance).
In practice, two-ports often represent coupling devices in which a source delivers energy to a load
through the two-port network as suggested in Figure 18.2. For example, stereo amplifiers take a
small low-powcr audio signal and increase its power so that it will drive a speaker system.
Determining and knowing ratios such as the voltage gain, current gain, and power gain o f a twoport is ver\' important in dealing with a source that delivers power through a t\vo-port to a load.
This chapter develops various formulas for computing these gains for each type o f two-port
parameter.
2. ONE-PORT NETWORKS
Basic Im pedance Calculations
As mentioned in the introduction, the current entering one terminal o f a one-port, illustrated in
Figure 18.1, must equal the current leaving the second terminal for any voltage
across the ter
minals. We begin study o f one-ports by exploring two impedance calculations o f a transistor ampli
fier circuit that are pertinent to ba.sic electronic analysis. The impedance or admittance seen at a port
is fundamental to the behavior o f a network to which a one-port or two-port is connected.
EXA M PLE 18.1. The circuit o f Figure 18.3 is pertinent to a simplified model o f a common-collector stage o f a transistor amplifier circuit. Specifically, the one-port o f Figure 18.3 models the input
impedance. In a common-collector amplifier stage, the impedance Z j is very large and is often neg
lected, i.e., we assume |Z,| = =c over the useful bandwidth o f the circuit. The goal o f this example is
to compute Z
and interpret Zy^^ in terms o f a transistor current gain parameter denoted by (3.
FIG U llE 18.3 One-port represennng the input characteristic of a

common-collector transistor amplifier circuit.
963
So
lu t io n
We attach a hypothetical vohage source, say,
across the port terminals, top to bottom being
plus to minus, to induce a hypothetical current ly Neglecting Z ,, from KCL,
Since the input impedance is the ratio o f the input voltage to the input current,
Zi=-^=(P+l)Zi,
Thus, for a large P, say, 150, the input impedance can reach very high levels for reasonably sized
impedances Z^. When amplifying voltage signals, we desire to have the ratio o f the internal source
impedance to the amplifier input impedance be small. Conversely, the amplifier input impedancc
should be large relative to the internal source impedance.
Exercises. 1. In the circuit o f Figure 18.3, suppose P = 100, Z j = oo, and the load resistance Z^ =
16 Q. If a sinusoidal voltage
the average powers
and
= 15^ 2 sin(5000r) V is connected to the input terminals, find

and the power gain.
AN SW ER: 139 mW. 14.06 W, 101

2. Consider the circuit o f Figure 18.4. T he one-port shown here models the output impedance of
a common-collector stage o f a transistor amplifier circuit. Show that
7
'o i l ! _
Io,
(P + l ) '
FIGURE 18.4 One-port representing the output characteristic o f a common-collector transistor

amplifier circuit.
3. In the circuit o f Figure 18.4, suppose P = 99 and Z j = 1 kH. If a current source
is connected to the one-port terminals, find
ANSW ER: 2 V. 2 mA
and |/j|.
= 200 niA
A second example that is common to basic electronic analysis depicts a circuit used in the analy
sis o f field-cfFect transistors (FETs).
E X A M P L E 18.2. Figure 18.5a represents a simplified model o f the input impedance o f a fieldeffect transistor circuit. As in Example 18.1, we neglect the large impedance in parallel with^^^K,.
Compute the input impedance
in terms o f Z j, Z 2, Z 3, and
FIC'ilRE 18.5. Simplified input impedance model of a field-effect transistor (FET) circuit.
S o lu tio n
As illustrated in Figure 18.5, assume a hypothetical voltage source
(plus to minus is right to
left) has been attached across the port terminals. At the bottom node, /, + g,V-[ + ^ = 0. Since
K, = Z,/, and /2 =
Now, since Z, =
+ K,)/Z2, it follows that
VAvi
Z ; = Z , + (1 + g Z | ) Z2 = Z j + (1 +
Exercise. In the circuit o f Figure 18.5, suppose
Z j) z ,.
= 0 .002 S, Z-,= 500 Q, and Z j is the imped
ance o f a 0.1 l-iF capacitor. Find the frequency (in Hz) at which the magnitude ofZy^^ is 707 Q..
ANSW i-R; 1(I03.J 11/
Thevenin and N orton Equivalent Circuits

Recall from Chapters 6 and 14 that the Thevenin equivalent o f a network, as seen from a pair o f
terminals (i.e., from a one-port), is a voltage source in series with the Thevenin equivalent imped
ance Z,!,is simply the equivalent impedance o f the one-port when all internal independent
sources are set equal to zero. The value o f the voltage source,
equals that voltage appearing at
the open-circuited port terminals. A source transformation on the Thevenin equivalent produces
the N orton equivalent, which is simply a current source in parallel with the Thevenin impedance.
T he value o f the current source is 7^^ = VgJZ^f^. This is the current that would pass through a short
circuit on the port terminals.
96 S
EXAMPLE 18.3. Consider the circuit ofFfigiire 18.6. Find the Tlievenin and Norton equivalent circuits.
I.
I.
1 :a
----->--------+
9
>
+ C
A +
FIGURK 18.6 Transformer circuit for Example 18.3.

S o l u t io n
The voltage across A-B satisfies
Kw = - ^ I a + V 2 = ^ I a * V i
Cs
Cs
To determine the Thevenin equivalent, we express Kj in terms o f

the current relationship /j = -nl^,
and /^. From Ohms law and
Therefore,
yAB =
+ R cr
Cs
^)
Thus Z,|^{s) = + R cr, V^J,s) = aRIi^j, and /^(jr) = V ( ------------------- , as set frorthL m cFigures
18.7a and b.
Cs
Cs
1 + Ra^
aRCs
RCa^s + 1
(a)
Cs
+ Ra2
(b)
FIGURE 18.7 (a) Thevenin and (b) Norton equivalent circuits of Figure 18.6.
Exercise. Suppose R = 100

and zeros o fZ ,,,
= 5, C = 10 mF, and
AN SW hR: I 'j. = lO/.v; pole at x = 0 ,ukI zero ai
= U.04.
= Au{t) mA. Find
and the poles
966
A technique known as matrix partitioning often simplifies the calculation o f Thevenin and
Norton equivalent circuits. Example 18.4 illustrates the technique.
EXA M PLE 18.4. Find the Thevenin equivalent circuit for the network o f Figure 18.8.
So
lu t io n
Step 1. Construct the loop equations. By inspection, the loop equations for the circuit satisfy
v,
+ 18
2.V
-8
-10
-8
18
-4
-10
-4
16
/3 .
13 6
s
0
The goal is to find
(18.1)
in terms o f /p the matrix equation 18.1 is partitioned accordingly. To avoid
dealing with numbers, let us rewrite this partitioned matrix equation as
r\v,i
\V,2
vvv>
where
= 136/j,
(18.2)
= 0, and the ^Vjjs represent their analogous (partitioned) counterparts in
equation 18.1. In general, Vj is the net voltage in loop 2 and
is the net voltage in loop 3.
Step 2. Solve the partitioned matrix equations for K, in terms o f /j arid the vector [ K,,
The
matrix equation 18.2 may be rewritten as a set o f two equations, namely

V|
and
+'^12
h
h
(1 8 .3 )
967
V2 I
= W21/1 + VV22
(18.4)
^3
Since
is invertible (for all passive networks), solving equation 18.4 for the vector [/j
yields
V2 I
(18.5)
Substituting equation 18.5 into equation 18.3 produces
[V2
U/,, - W^2W22^V2^ ] /i + U'12^^22'
(18.6)
^3
MATLAB or its equivalent, or possibly a symbolic manipulation software package, allows the
matrices in this equation to be conveniently computed. Comparing equation 18.6 with the struc
ture o f a Thevenin equivalent circuit,
allows us to conclude that the Thevenin impedance is

iV,, -VV,2VV27VV2|
(18.7)
and the open-circuit voltage is

V,
Step 3. Compute Z I and
Z , ( .) =
(18.8)
From equation 18.7,
Is
-1 -8
-1 0
18
-4
-4
18
18
-4
-4
16
-1
- 8
-1 0
.
_ .V+ 0 .0 9 5
= 5 .2 6 5 --------------
and from equation 18.8,
- 10 ]
-1 1361
s
0
Figure 18.9 shows the resulting Thevenin equivalent circuit.
84
^
Q
968
5.3
s + 0.095
+ \- M
FIGURE 18.9Tlicvcnin equivalent circuit of Figure 18.8.
Exercise. Suppose the capacitor in Example 18.4 is changed to a 2 H resistor. Compute the new
Z,,[s).
AN SW l-ll: Z^,,{s) = 7.2647 i l via the MA'I'LAB code
AVI 1 = 20; W 12 = [-8 -10]:
\V21 = [-8: -101: \V22 = [18 -4:-4 16]:
vZth = \V11 - \Vl2'inv(\X/22)*\X'21
3. TWO-PORT ADMITTANCE PARAMETERS

B asic D efinitions a n d Exam ples
As mentioned earlier, rwo-port representations allow us to deal only with the terminal voltages and
currents o f the two-port, depicted in Figure 18.1b and repeated in Figure 18.10.
1.
O+
V.
o
Two
Port
-O
+
V,
-o
FIG URE 18.10 Standard two-port configuration having four external variables: /j, I^, Kj, and V-,.
969
ADMITTANCE PARAMETERS
'rhroughout this and later discussions, assume that the two-port ot Figure 18.10 has no inter
nal independent sources and that all dynamic elements are initially relaxed, i.e., have no ini
tial conditions. Under these assumptions, the admittance parameters or ^-parameters o f a
two-port are expressions for the terminal currents, /j and I 2 , in terms o f the port voltages,
and ^2
/r
-Vli
V|2' -\/,-
^2
.^21
>22
(18.9a)
^2
or, in scalar notation,

^1 Al 1^1 "*.'12^2
(I8 .9 h )
I 2 = V21V 1 -I- >22^2

Using either o f these sets o f equations, one can define each admittance parameter,
lows:
-II
V,
as fol
( 18. 10)
vs=o
Vl=0
V2 I =
V, =0
Since each admittance is defined with regard to a shorted terminal voltage
= 0, the^^y are
often called short-circuit adm ittance parameters. I'he unit o f an admittance parameter is S.
Some examples will illustrate convenient methods for computing the ^-parameters.
EXA M PLE 18.5. Compute the short-circuit admittance parameters o f the circuit in Figure 18.'
FIGURE 18.11 Circuit for Example 18.5.

S o l u t io n
The overall strateg}' is to write equations for /, and /, in terms o f

Accordingly, by Inspection,
and
/, = K, I/, + ^3 (Kj - V,) = (K, + Y,) K, - K, 1/

and
using nodal analysis.
970
* y 2 V ._ * K, ( K , - K,) = ( g ^ - r ,) . (K j * K,)
In matrix form, these equations are
h
K, + K3
I2
A ' -^3
Yl
V/,-
A -Vii
V21
.^2.
V,2
V/,-
V22
vs.
in which case
y\ \ - ^1 + ^3>^12 -
^V>'21 -
~ ^3 '*^^>'22 " ^2
^3'
Exercise. Suppose the controlled current source in Figure 18.11 is reconnected across
with the
arrow pointing to the left. Find the new ^-parameters,

A N S W E R : n , = K, +
= g ,~
and y ,, = )\ +
In the next example, we combine the method o f matrix partitioning with the use o f nodal equa
tions to compute the y-parameters.
EXA M PLE 18.6. Compute the ^-parameters for the circuit o f Figure 18.12.
^TT
FIGURE 18.12 Three-node circuit for Example 18.6.

S o l u t io n
Computation o f the j'-parameters will again proceed by the method o f nodal analysis. Consider,
for example, node 1, in which the current /, must be determined, /j has the form
where each variable is understood to be a function o f s. The coefficient K j, is simply the sum o f
the admittances incident at node 1. The coefficient
admittances between nodes 1 and 2, and
nodes 1 and 3. Similarly, I 2 = V-y]
is simply the negative o f the sum o f the
^ is the negative o f the sum o f the admittances between

+ Y-^^Vy where Yjj is the sum o f the admittances
incident at node 2, etc. Hence, in matrix form, the node equations o f the circuit o f Figure 18.12
are
97
.v+ 3
w
=
-s
: -2
- .V
5 + 21 -1
r\y,i
iM
11 : M/pi
VV2V V w ji
V2
This nodal equation matrix is symmetric, because the /^LC network contains no dependent
sources. Using the method o f matrix partitioning introduced by equation 18.6 yields
/r
/2.
-\/,-
.^ 2 .
-li
-
5 + 3
- 5
i - + 2.2
-(5 + 0 .4 )
-(5 + 0 .4 )
5 + 1.8
'
1 T
+ 2 5
Exercise. The circuit o f Figure 18.12 is modified by adding a 2
[2
1]]
V,-
resistor in parallel with the
capacitor. Find the new ^/-parameters.

r
.ANSWFR:
+ 2 .7
- ( 5
- ( .s + 0 .9 )
+ 0 .9 )!
.s + 2.3
A last example couples a transformer with a resistive n-net\vork.

EX A M P L E 18.7. Compute the ^-parameters o f the circuit in Figure 18.13.
FIGURE 18.13 A resistive :r-network coupled to an ideal

transformer circuit for Example 18.7.
S o l u t io n
Find the y-parameters by using nodal analysis in conjunction with the ideal transformer equations.
First, at port 1,
/, = K, + (V , - V,) = 2Vi -
= 2V , - - ^ 2
(1 8 .1 2 )
Now considering that al 2 = -/, a node equation at the primary o f the transformer yields
(V i-V i)
h = h = -a
a
= - - V ,+ 4 - V - ,.
a~
(18.13)
where the last equality uses the relationship aV^ = V2. Putting equations 18.12 and 18.13 in matrix
form yields the ^'-parameter relationship
r\/,i
a
Exercise. In the circuit o f Figure 18.13, the top resistor is changed from 1 Q. to 0.25
turns ratio a =2 , find the new ^/-parameters.
A N SW I- R ;
-2
1.25
With the
.S.
Tw o-D ependent Source E quivalen t Circuit

The key to engineering analysis rests with the interpretation o f appropriate mathematical equa
tions. The key to two-port analysis is the interpretation o f the two-port equations. Take, for exam
ple, the first admittance equation,
1 =.^11^1 +>'12^2
One circuit-theoretic interpretation of this equation has the port current /j equal to the port volt
age Kj times an a d m i t t a n c e ] in parallel with a voltage-controlled current source^, 2 ^^2- ^
ilar interpretation o f the equation
A =^21
yields an admittance branch
+J'22 ^2
in parallel with a voltage-controlled current source^ 9, K ,.
These interpretations lead to the tw o-dependent source equivalent circuit o f a rwo-port repre
sented by the short-circuit admittance parameters. (See Figure 18.14b.) This equivalent circuit
aids the computation o f input and output impedances and voltage gain formulas. Note: in this
chapter and later chapters the standard resistance symbol often is used to designate a general
impedance or admittance.
-o
O+
V.
o-
+
V,
y y.2
L
V2,
V22 J
(a)
FIGURE 18.14 (a) Short-circuit admittance parameters, (b) Their two-dependent source equivalent
circuit. The dotted line at the bottom of (b) indicates that the two halves are not necessarily connected.
4. Y-PARAMETER ANALYSIS OF TERMINATED TWO-PORTS

This section takes up the task o f analyzing term inated tw o-ports modeled by ^-parameters. A
rvvo-port is terminated when it has source/load admittances. Any circuit or system in which a
source provides an excitation signal to an interconnection network that modifies the signal and
drives a load impedance can be represented by a terminated rwo-port. Such a scenario is common
to numerous real-world systems. For example, the utility industry delivers power to a home from
a generating facility through a transmission network, and a telephone system delivers voice infor
mation from the phone through a transmission network to a receiver.
Input and Output Admittance Calculations

The input and output admittances o f a terminated two-port are important for determining power
transfer and for various gain computations. In what follows, we will show two different methods
for computing the input admittance o f the terminated two-port illustrated in Figure 18.15.
Computation o f the output admittance is left as an exercise.
O
+
-o
+
V,
Vn y ,2
L V21 V22 J
FIG U RF 18.15 Two-port terminated by a load admittance

The first method for computing
is a matrix method. Recall equation 18.9a,

/l'
Vll
>'12
\V
.'21
V22
V2
Using the terminal conditions imposed by the load
in Figure 1 8 .1 5 , we obtain
97-1
Incorporating this terminal condition into the two-port ^-parameter equation yields
>11
>12 '
V/,-
/l'
>11
Vl2
>21
>22
.^^2.
>21
y ii +
Using Cramers rule to solve for
det
V. =
Because
^2
in terms o f /j results in
/l
0
>12
>22 + y'L
>12
det >11
>21 >22 +
is the ratio o f /, to
rv,
(>22 + ^l ) ^1
>l 1(>22 +
" >12>21
the input adm ittance o f the two-port o f Figure 18.15 is

Vl2>'21
(18.14)
>22 +
E X A M P L E 18.8. Derive the input admittance o f equation 18.14 using the two-dependent
source equivalent o f Figure 18.16. Here we avoid the solution o f simultaneous equations while
increasing insight into the operation o f the two-port.
I.
FIG URll 18.16 Input admittance calculation using twodependent source equivalent circuit.
S o l u t io n
W ith regard to the right side o f Figure 18.16, the current ^^21
iTiust equal the current through
the parallel admittances j /22

->'21^. = - ( 7 2 2 - ^ / .) ^2
Hence,
v-, = -
.V2 I
(18.15)
Equation 18.15 says that the voltage V2 equals the current, -^21
Now consider the left side o f Figure 18.16. Here
>l I -
impedance,
>'12>'2I
>22 +
Again, y,>, =
leads to the same formula as equation 18.14.
Vi
/
1
+ K^)
9 5
Exercise. Let
denote a source admittance. Show that the output admittance o f Figure 18.17 is
J 12J 2 I
(18.16)
V-
-o
-O+
Vu Vl2
V.
V,
V2, V22
-o
2
FIGURK 18.17 Input-terminated two-port for output admittance calculation.
G ain C alculations
Our objective now is to derive a formula for the voltage gain of a doubly terminated r\vo-port, as
illustrated in Figure 18.18. Again, t h e s y m b o l s denote general admittances rather than the tra
ditional conductances.
yi2^2
V.
V,
y
'22
V2/,
FIG URE 18.18 Doubly terminated rwo-port driven by voltage source,
symbols denote general
admittances.
The specific goal is to derive the voltage gain formula
Ys
->21
(18.17)
The overall gain calculation breaks down into two cascaded gain calculations as follows:
^
V,
V/.
V/,
Gi/ = = X
v;
V,
V/,
(18.18)
Computation o f the gain, ( 7^^ = l l ^ follows direcdy from voltage division:

(1 8 .1 9 a )
Chapter 18 Two-Forts
VS
To compute the gain G ^.2 = from directly from equation 18.15,
V /.=------>22 + Yl
which implies
^
V21
^ v2= = M
>22 +
(I8 .1 9 b )
Equation 18.17 follovv.s by substituting equations 18.19a and 18.19b into equation 18.18.
Exercises. 1. In the circuit o f Figure 18 . 18, 7 ,, = 5 ,^ p = -0 .2 , y 21= 50, j/2-) = 1 (all *n mS),
1 k n and /?^ = 2 kQ. Find Z- and G y = VJV^.
A N SW ER:
= 8 5 .' U. 6\ -
2. Compute the current gain,
2.(>32
o f the circuit o f Figure 18.18.
ANSW^HR;
5. TWO-PORT IMPEDANCE PARAMETERS

Definition and Examples
The 2-parameters, or impedance parameters, relate ^-domain currents to ^-domain voltages, as one
would expect. The z-parameters are the inverse o f the ^'-parameters in most cases.
IMPEDANCE PARAMETERS
For the nuo-port o f Figure 18.10, the z-parameters zij relate the port currents to the port voltages
according to the mati ix equation
Vj-
^11
-12'
V2
.^21
^22_
(18.20)
under the assumption o f zero initial conditions and no internal independent sources. Therefore,
from equation 18.20, each individual z-parameter is defined according to the formulas
h h=0
-1 2
/,=0
(18.21)
Z2\ =
^22
h=0
/,=0
Since each zij is defined with one o f the ports open-circuited, i.e., /j = 0 or /-, = 0, the z ip are
called open-circiiit im pedance param eters. Their unit is the ohm.
EXA M PLE 18.9

Compute the 2-paranieters for the circuit o f Figure 18.19.
FIGURE 18.19 A simple 7'-circuit for computation o f z-parameters in Example 18.9.

S o l u t io n
Rather than apply the z-parameter definitions o f equation 18.21, we will use mesh analysis to
obtain equation 18.20 directly.
Step 1. A loop equation at port 1 yields
.y
.V
s/
(18.22)
Step 2. Similarly, a loop equation at port 2 produces

V9 =
+ (/] + h ) = /i + f 3 + ] h
s /
(18.23)
Step 3 . By inspection of the right-hand sides o f equations 18.22 and 18.23,

10
Z=
^11
- 12'
^21
~22
s
]0
.y
ay + 10
In the next example, we utilize the technique o f matrix partitioning to compute the z-parameters
o f a n:-network.
E X A M PLE 18.10. Compute the z-parameters o f the Jt-network o f Figure 18.20.
F I G U R E 18.20. Ji-network for Exam ple 18.10.
9~S
S o l u t io n
It is straightforward to write the following three loop equations:

l/ ,= Z ,/ ,- Z ,/ 3
= Z 2/2 + Z 2/3
and
0 = - Z / i + Z 2/2 + (Z, + Z 2 + Z^)I^
Putting these equations in matrix form and partitioning the matrix appropriately yields
^2
0
z,
0 1
-z ,
Z, j
Z2
^2
Z|
Z2 j Z| + Z2 + Z3
Hence, using the matrix partitioning formula derived in Example 18.4, we obtain
^11
^12 '
^22
1
0
z,1
0-
^ 2.
Zi + Z 2 + Z 3
- Z iZ i'
Z] + Z 2 + Z 3
^2
^2
-Z ,Z 2
^2
Z | (Z 2 + Z 3 )
^ 1^2
Z|Z2
Z 2 (Z ]+ Z 3 )
z, + z , + z .
Exercises. 1. In Example 18.10, suppose Z j = Z-, = 1 H and Z 3 is a 1 H inductor. Find the zparameters.
AN SW ER:
-11
" i:
-21
-22
.V+ 1
.v+2
1
.v + 1
2. Now suppose that Z , and Z 2 are changed to 1 F capacitors and Z 3 is a 1 H inductor. Find the
new 2-parameters.
A.N'SWER:
-II
- 12'
"21
-22
.v(.v-+2]
.V -
+1
1
I
.V -
+ 1
979
Relationship to j-Paxam eters

Since the z-parameters relate port currents to port voltages and the j/-parameters relate port volt
ages to port currents, one might expect that the ^-parameter matrix and the ^-parameter matrix
are inverses o f each other. Specifically, if
V,V2_
Z21
^12'
^22
then perhaps it follows that
h'
-11
^12'
>21
Z22
^2
This relationship is valid provided Zj

^11
^21
and
-1 Vi'
^12'
'y\\
312 -Vi-
.>21
.''22 .^2.
^ ^ and^j
-1
^22_
V|1
>12'
>21
>22
^11
^12'
>11
>12'
>21
-22.
>21
>22.
Exercise. A certain two-port has z-parameters ^11

^21
which case
(18.24a)
-1
(18.24b)
212'
5 + 1
^22
.9-^2
1
5+ 1
Find the ^-parameters. Can you construct a three-element passive circuit that has these ^/-parameters?
ANSWER: ;
H inductor.
>11
>12 _ 1 '.v + l
>'21
>22
.V -1
-1
.v + l
: sec 1-igure 18.20 with Z, = Z t = 1 2 and Z^ changcd to a 1
Despite this inverse relationship, some circuits have z-parameters but not ^-parameters, and vice
versa, as illustrated by the following example.
EXAM PLE 18.11. Compute the z-parameters o f the circuit of Figure 18.21. Do the/-parameters
exist?
O+
-O
V.
V,
o-
-o
FIG U R E 18.21 Resistive two-port having z-parameters but not ^parameters.
980
S o l u t io n
By inspection, uhe 2-parameters o f the circuit o f Figure 18.21 are

V,-
R-
vs.
A
^2
The 2-parameter matrix, [z-^, is singular, since d e t[2 j =
I2
= 0. Because the [z-^ matrix does
not have an inverse, the circuit has no y-parameters. One can check j/, j directly to verify this claim.
Consider Figure 18.22. Because Vj = 0, there is also a short circuit across Kj, making the ratio
vi 1 =
Vi
^-,=0
undefined.
O-
V,
v ,= o
OFIGURE 18.22. Equivalent circuit for c o m p u t i n g i n which port 2 is shorted, so that
= 0.
The Two-Dependent Source Equivalent Circuit

As with thej)/-parameters, the z-parameters have a two-dependent source equivalent circuit inter
pretation, illustrated in Figure 18.23b. Consider first the equation
Here,
equals the sum o f two voltages: z^ j/j plus the voltage due to a current-controlled volt
age source given by z^-Îj- This is precisely the left-hand portion o f Figure 18.23b. A similar inter
pretation follows for Vj = -221A
O
+
V,
^2, ^22
(a)
Mr-
-0
^12
-
yielding the right-hand side o f Figure 18.23b.
V,2
.
-0
o
^22
V,
+
V,
"211
(b)
F IG U R E 1 8 .2 3 (a) z-paramcters for modeled nervvork. (b) Tw o-dependent source equivalent circuit.
98
This equivalent circuit proves useful for computing voltage gains and input and output imped
ances o f terminated two-ports. It should also be noted that there are other, equivalent circuits that
interpret the 2-parameters. A similar remark can be made for j/-parameters.
6. IMPEDANCE AND GAIN CALCULATIONS OF TERMINATED

TWO-PORTS MODELED BY Z-PARAMETERS
Input and O utput Impedance Com putations
Earlier we derived the formula for the input admittance o f a terminated two-port in terms o f the
V-parameters, leaving the output admittance calculation as an exercise. Here, using ^-parameters,
we derive the output impedance of the terminated two-port o f Figure 18.24 and leave the input
impedance calculation as an exercise. Specifically, our first step is to derive the output impedance
formula
ôiit ~ ^22
-12-21
! 1 + ^v
(18.25)
as a function o f the net^vork z-parameters and the source impedance.
Beginning on the right-hand side o f Figure 18.24,
Vj = Z22^2
In calculating
source,
^21^1
(18.26)
which is the Thevenin impedance seen by the load, the independent voltage
is set to zero. Hence,

-(Z^ + Z j,) /j = 2,2 h
Solving for /, and substituting into equation 18.26 yields the output impedance formula o f equa
tion 18.25.
Exercises. 1. Repeat the preceding derivation using a matrix method and Cramers rule.
2. Derive the following formula for the input impedance o f a terminated rwo-port:
^12^21
A n = ^ \ ---------(18.27)
<22 +
Gain Calculations
The next phase o f our rwo-port analysis is to repeat the ^'-parameter derivation o f the voltage gain
o f the rwo-port in the context o f 2-parameters. Specifically, our aim is to compute
The ratio
^2
^2
C 8-28)
follows from voltage division:

V^l _
Gvi = ^ =
To compute the gain Q ^ = Yl.

V,
Z/,
(18.29)
first apply voltage division to obtain
-----^21^1
+ ^22
^2
From the definition o f input impedance, /j =
Hence,
+ <-22 în
Substituting equations 18.29 and 18.30 into equation 18.28 yields the voltage gain:
-21
Gy = Gy2Gy\ =
\Z22 + ^ L / \ în +
(18.31)
/
An application o f this formula and its derivation to a cascaded network of two-ports (two transis
tor amplifier circuits) appears in the next example.
E X A M PLE 18.12. Consider the nerwork o f Figure 18.25, which represents a two-stage (transis
tor) amplifier configuration. Each stage utilizes the same transistor in a difiêrent circuit configu
ration. The first stage is an amplification stage that will amplify a small source voltage to a much
larger one. The second stage is an impedance-matching stage used to match the load to the out
put impedance o f the amplifier circuitry to achieve maximum power transfer, at least approxi
mately. The 2-parameters, in ohms, for each stage in Figure 18.25 are given by
983
350
2.667-
-1 0 ^
6667
21 =
; Z2 =
1.0262 X 10^
1.0258
K/
6 ,7 9 1 '
(18.32)
6 ,7 9 4
(a)
Compute the input impedances,
and
(b)
Compute the voltage gain,
(c)
Check the matching o f the load and output impedance o f the amplifier circuit.
160
FIG URE 1 8.25 Two-stage amplifier network.

S o l u t io n
Step 1. Determine
^ straightforward application o f equation 18.27 using the z-parameters
o f stage 2 produces the impedance
An! = ^11----- = 1.0262 X 10^ -
6 7 9 1 x 1 .0 2 5 8 x10^
+ Z,
Step 2. Compute the voltage gain,

ticular,
K..
>^2
~ ^0 1 1 / ^ 2
Z,
z,|
Z,,_ + Z22 Z/,.2
(18.33)
6794+16
18.30 applies here. In par

I 6 ( i .0 2 5 8 x I0*>)
(16 + 6 ,7 9 4 )3 ,2 6 2
(18.34)
The gain here is small, but remember that this stages real purpose is impedance matching, not
amplification. By proper choice o f the z-parameters, the output impedance will approximately
match that o f the load. This allows us to dispense with an expensive and bulky impedance-match
ing transformer.
Step 3. Compute
fo r stage 1. Here, observe that
(i-quation 18.33) in parallel with the 2
kD resistor between the two stages acts as a load to stage 1. Z-^^j

the load to stage 1, denoted Zj^ = 1239.8 H.
parallel with 2 kQ becomes
The input impedance seen at the front end o f stage 1 follows from equation 18.27:
4 , =
- ^ 1 2 ^ 1 - = 350 +
Z22+^
li
^
6 . 6 6 7 + 1 ,2 3 9 .8
= 6 8 7.3 n
(1 8 .3 5 )
Chapter 18 Two-Iorts
Step 4. Compute the voltage gain, G,,,

applying equation 18.29 yields
^vl - T 7 -
V ,/VsJ o r stage 1. Using the result o f equation 18.35 and
j" ^21
^
\ Z^. + Zj,j) \^/J + ^22/
-10^
^(
1,239.8
[ 7 5 + 687 .3 ] 1 1 ,2 3 9 .8 + 6.6 6 7 ;
(18.36)
= -2 0 5 .7 .
Here the large gain of stage 1 leads to significant amplification o f the input signal. For example,
a 40 mV sine wave would be amplified to a little over 8 V, which can drive a small speaker.
Step 5 . Compute the overall voltage gain, Gy=
The desired gain is simply the product o f
equations 18.34 and 18.36, i.e., G^/=

x
= 2 0 5.7 x 0.7388 = - 1 5 2 , which remains fair
ly large. Indeed, a 40 mV sine wave would be amplified to approximately 6 V. Other amplifica
tion stages could be added to further increase the overall gain.
Step 6. Verify that the load matches the amplifier circuitjy to a reasonable degree. In this task, one
first computes the output impedance,
nation of
o f stage 1 using equation 18.25. The parallel combi
with 2 k n , denoted Z p = 1.732 kH, becomes the source impedance to stage 2. It
is then easy to compute the output impedance o f stage 2, again using equation 18.25. The answer
is
= 17 Q. The details are left as an exercise.
Exercises. 1. For the circuit o f Figure 18.25, verify that the output impedance equals 17 Q.
2. In Figure 18.25, call the 2 kQ resistor R-,. Find a new value of Rj so that
exactly equals 16 Q.
ANSW ER: 1.783 k il
3 . Suppose the 75 O. source resistance in Example 18.12 is changed to 300
which would represent a
twin line connection bet\veen an ideal voltage source and the first amplifier stage. Redo the example.
ANSW ER: Numerical values are obtained using the h)llowing M AFI.AB code;
1^ = 300;
z z ll = 1.0262e6: zzl2 = 6791;
7.7.21 = 1.0258e6; /.z22 = 6794;
x.ll = 350; vM = 2 .6 6 "; z21 = -le 6 ;
z22 = 6667; R2 = 2e3; ZL = 16;
Zin2 = zzl 1 - z/.I 2'/.z21/(/z22 + ZI.)
(;2 = (zz21/Zin2)-ZL/(Zl.+zz22)
ZLl = R 2 Zin2/(R2+Zin2)
Zin = z ll - zl2'z21/(z22 + Z L l )
G1 = (z21/(Rs+Zin))-ZLl/(ZI.Uz22)
Gv = G 1 *G2
Zoutl = z22-zl2*z21/ (7ll+ R s)
Zs2 = R 2*Zoutl/(R 2+Zouil)
Zout = zz22-zz 12' zzl I /(zz 1 1+Zs2)
985
7. HYBRID PARAMETERS
Basic D efinitions and Equivalences
As we have seen, some circuits have ^-parameters but not z-parameters, and vice versa. A circuit
element that has neither is the ideal transformer.
EXA M PLE 18.13. This example shows that the ideal transformer o f Figure 18.26 has neither znor/-parameters. From the definition o f an ideal transformer (Chapter 17), V2 = a V and /j =
-al-). Clearly, K, and
cannot be expressed as functions o f /j and I 2 , nor can /, and
be
expressed as functions o f K, and Vj- Hence, an ideal transformer has neither j/- nor z-parameters.
1 :a
o+
V,
FIGURE 18.26 Ideal transformer, having neither j'- nor z-parameters.
Two-port circuits having neither z- nor/- parameters require an alternative modeling technique.
The hybrid parameters offer one o f several alternatives.
HYBRID PARAMETERS
Hybrid parameters, h-jy are a cross between y- and 2-parameters: a voltage
and a current
are outputs, with /, and Vj as inputs. Specifically, if the rwo-porr o f Figure 18.10 contains no
internal independent sources and has no initial stored energy, then the hybrid parameters are
defined by the matrix equation
Vi-
/'II
im -
/'21
hi
V2
(18.37)
This mixture o f variables actually arises as a simplification o f a model o f a common emitter

configuration o f a bipolar transistor. From the context o f equation 18.37, />], has units o f
ohms,
and ^21 are dimension less, and h i has units o f S.
As with both/- and 2-parameters, we interpret equation 18.37 as a two-dependent source equiv
alent circuit, as illustrated in Figure 18.27.
)86
Chapter 18 'Two-Ports
'2
+
V.
-o
+
h,2~
_ h21 h22 _
V.
o-
-o
(a)
FIGURE 18.27 (a) Hybrid parameters, (b) Two-dependent source equivalent circuit.
Exercise. Justify the two-dependent source equivalent circuit interpretation o f Figure 18.27; i.e.,
apply KVL and KCL to the circuit o f Figure 18.27b to derive equation 18.37.
The definition o f each /^-parameter follows directly from either the preceding equivalent circuit or
from equation 18.37. For example,
(18.38a)
V'-,=0
Because Aj| is the ratio o f an input voltage to an input current, it is an input impedance. Since
^2 = 0, /?], is termed the short-circuit input impedance. Notice, however, that A, j is simply
related to both the
and the z-parameters as follows:
I
/ , = ^
'I I
-12^21
(18.38b)
The second //-parameter, />2 p is called the short-circuit forward current gain, since it is the ratio
o f I j to /] under the condition K, = 0, i.e.,
(18.39a)
V^=0
From the 2-parameter equation Vj = ^21A
pretation.
^'>2-^2
^2
*2 , has a simple z-parameter inter
(18.39b)
^-,=0
^22
The third //-parameter is
/,=0
(18.40a)
Since it is the ratio o f Kj to Vj under the condition that port 1 is open-circuited, i.e., /j = 0, it is called
the reverse open-circuit voltage gain. Interpreting //p in terms o f the y-parameters, we obtain
987
>'12
I,=0
(18.40b)
'"'I
Finally, we note that the open-circuit output adm ittance is
h
l' n = -
(18.41a)
/,=0
and has units o f S. The word open-circuit suggests a z-parameter interpretation. Considering
that V-y = 221-^1
h ~
1 _
V12J 21
- >'22-----------:22
311
/,=0
(18.41b)
This relationship is similar (notice the subscripts) to equation 18.38b, which determines the shortcircuit input impedance.
We will return to these equivalences later, after we gain some computational experience.
Com putation o f /^-Parameters

The first example o f this subsection demonstrates a circuit that has neither z- nor ^/-parameters.
E XA M PLE 18.14. Consider the two-port o f Figure 18.28, whose front end is a short circuit and
whose secondary is an open circuit. Thus it has neither z- nor ^-parameters. Our objective is to
compute the /^-parameters.
I
L
-O
+
O+
V,
-o
FIG URE 18.28 Simple two-port with /^-parameters but neither z- nor ^-parameters.
SO L U T IO N .
By inspection, the A-parameters are
V,-
O' /r
/2.
V2
T he second example illustrates the computation o f /^-parameters for an ideal transformer circuit.
Chapter 18 Two-Porcs
EXA M PLE 18.15. Find the /^-parameters o f the c%vo-port in Figure 18.29.
1 :a
O--------+
------o
*+
A
V.
V,
V,
FIGURE 18.29 ideal transformer circuit for Example 18.15.

S o l u t io n
Step 1. Construct an equation fo r K,. From the primar}' and secondary voltage relationship o f an
ideal transformer,
V, = - v - ,
By KVL at port 2, =
RI->, in which case

n
V, + - / o
a ~
Step 2. Construct an equation fo r
a ~
(18.42)
in terms o f the other variables. From the primary and second
ary current relationship o f an ideal transformer,
/ 2 = -/ ,
Cl
(18.43)
Step 3. Write equations 18.42 an d 18.43 in matrix form, and solve fo r Kj and I-, in terms ofl^ and
Vj. In matrix form, equations 18.42 and 18.43 are
Solving for the vector [ Kj
rv/,1
/o
- a
/ll
(18.44)
produces the /^-parameter equation

-1
1
-V,-
h
0
r 0n
- -
-M
h'
VS
\R
r
a~
\'
Cl
-1
^2
Exercise. In the circuit o f Figure 18.29, suppose the resistor R is connected in

parallel (instead o f in series) with the secondar)' winding o f the ideal trans
former. Find the new />parameters.
ANSWMR: /,, = 0. /,, = Mu. A,, = -\hi,
= MR
EX A M PLE 18.16. Find the /;-paramcters o f the circuit o f Figure 18.30.

L
R.
O----+
V,
'2
o-
V,
-o
FIGURE 18.30 Simple circuit for illustrating A-parameter definitions.

S o l u t io n
Step 1. Find A, j using equation 18.38a. With V2 = 0, i.e., port 2 shorted,
= R^
v-,=o
Step 2. Find
using equation 18.39a. Again with
= 0, i.e., port 2 shorted, the current I 2 =
-/ j since all current flows through the short circuit. Hence
V-,=0
Step 3. Find A,2 using equation 18.40a. W ith /j = 0, i.e., port 1 open-circuited, K, = V2 since
there is no current through Ry Hence,
/,=0
Step 4. Find A22 using equation 18.4 la. Again with /j = 0 , /2 = (C^ +
= Cs + Gt
/,=0
In summar\-,
\Rx
-1
Cs + G^
in which case
990
Exercises. 1. If all parameters in Example 18.16 are 1 with proper units, find the /;-parametcrs.
ANSW ER:
2. If the capacitor in Example 18.16 becomes an Z. H inductor, find the new A-parameters.
ANSW ER:
-1
+ G-.
Lv
Impedance and Gain C alculations o f Terminated Two-Ports

This subsection analyzes the impedance and gain properties o f a terminated two-port character
ized by /^-parameters as in Figure 18.31, similar to the analysis done with both the z- and )'-parameters.
FIG URE 18.31 Doubly terminated hybrid equivalent circuit of a two-port.

Vi
Recall that Zj{s) = . From the right half o f Figure 18.31,
h
It follows immediately that
/hi
From the left-hand side o f Figure 18.31,
(18.45)
= A|j/| +
Substituting for Vj using equation
18.45 implies that the input impedance is
1,22+ Yl
/,
(18.46)
Exercise. Show that the output admittance is given by

I2
K)ut ~
~ ^*22
//|o/h 1
h\ 1 +
(18.47)
991
Knowledge o f the input and output impedances/admittances permits us to derive various gain for
mulas in terms o f the //-parameters. For example, consider again the left half o f Figure 18.31. Since
the input impedance is known from equation 18.46,
h=^V \
(18.48a)
Thus, from equations 18.45 and 18.48a,
^2 = ------ ^ / | = ------ V,
1122+Yl
în
(18.48b)
implying the voltage gain formula,

V2
^v2 = 7 7 =
+ W
^1
(18.49)
Voltage division at the front end o f Figure 18.31 yields the other voltage gain formula,
G, _
Zi + Z^
(18.50)
The overall voltage gain is the product o f equations 18.49 and 18.50, i.e.,
^ ^
7 7 = C j.|Gi,2 = -
Exercise. Compute the current gain

A N SW ER:
/?21
^2 ^ ^lM\
/1 h->'^ + )' /
8. TRANSMISSION PARAMETERS
Transmission or f-parameters were first used by power system engineers for transmission line
analysis and are still so used today. They are sometimes called ABCD parameters.
V92
Chapter 18 Two-Iorts
TRANSMISSION PARAMETERS
T he /-parameter representation has the matrix relationship
with the matrix, T =
Vj-
^11
hi
V2
/ l.
fl\
t il
-I2
( 18.52)
called the /-parameter matrix. As with th e ^ , z-, and /^-parameters,
the entries t-j are defined as follows:
^ " ^ 2 /, = 0
^-,=0
(18.53)
v^=o
/,=0
The matrix equation 18.52 leads directly to the relationships o f equations 18.53 by setting
the appropriate quantity, I 2 or Vj, to zero.
In computing a single t-j vs'ith equations 18.53, some care must be exercised in exciting the circuit.
By definition,
h= ()
The ordinary interpretation o f this equation is: apply an input V-, and find an output
the condition that
= 0. 'I'hen
j is the ratio o f the Laplace transform o f the response
under
to that
o f the input K,, i.e., a reverse voltage gain when port 2 is open-circuited. This situation causes a
predicament: an independent voltage source for Vj causes a current
to flow. To circumvent this
predicament, we use the slightly modified formula

1
^11
/7 = 0
The quantit)'
forward voltage gain when port 2 is open-circuited) suggests that
we excite port 1 by Kj with port 2 open-circuited, which forces /-, = 0. It is then straightforward
to calculate /j j as the inverse o f
Similar interpretations must be made with regard to
the other defining formulas in 18.53.

A simple example illustrates r-parameter calculations.
EXA M PLE 18 .1 7 . Consider again an ideal transformer circuit, shown in Figure 18.32. Here V-,
and /j =
This leads to the /-parameter matrix
0 \ V2
a
0
-/o
1 :a
O
+
-o
+
V.
FICiURt 18.32 Simple cransformer circuit for Example 18.17.
Input and output impcdance calculations for terminated two-ports modeled by f-parameters do
not follow the usual pattern. Nevertheless, the t\s'o-port of Figure 18.33 has input impedance
7
_ h\'^L +^12
(18.54)
and output impedajice

7
ôut
= ^22^v *^^12
^21^5 +^11
(18.55)
The derivation o f these results is left as a homework problem.
+
V,
FIGURE 18.33 Terminated r\vo-port modeled by /-parameters.
Exercise. The rwo-port o f Figure 18.33 has r-parameters r, j = 0 .0025, ^p= 500
^ ,^ 21= 3 .1 2 5 x 1 0 -8 S, and
= 0.00625. IfZ^ = 200 kQ. and
= 20 kH, find
AN SW ER: 80 kL X 200 k L l
One o f the most important characteristics o f /-parameters is the ease with which one can use them
to determine the overall /-parameters o f cascaded rwo-ports, as illustrated in the next example.
994
EXA M PLE 18 .1 8 . Compute the f-parameters o f the cascaded two-port o f Figure 18.34 in terms
of
and Tj, the t^vo-port parameter matrices o f the first and second sections, respectively.
FIGURE 18.34 Cascade of two two-ports modeled by f-parameters.

S o l u t io n
From the definition o f the f-parameters for each two-port,

1
/ l.
But since V2 =
r ^21
and
= 7-,
[ K;*T 1
.-^ 2 .
.-^ 4 .
and -1-, - ly it follows that
r^ ii
=
V3-
= T\1
7 i1
-h
= T\Tl
1 L
h
.-^ 4 .
implying that the f-parameter matrix o f the cascade o f Figure 18.35 is simply
Tj.
Exercise. The f-parameters o f the two-ports in Figure 18.34 are (in standard units)
0.1
T\1 =
0
10
and Ti =
2
0.1
10'
1
Find the f-parameters o f the cascade (T j comes first, followed on the right by Tj ) Then find Vj
when a voltage source o f 4 V is applied to port 1 o f the cascade and port 2 o f the cascade is opencircuited.
A N S W I- R : T =
0.2
10
9. GENERAL RELATIONS AMONG TWO-PORT PARAMETERS

The h-.y-, t-, and z-parameters are interrelated. This subsection derives several relationships, with
others left to the homework problems. The complete set o f results is summarized in Table 18.1 for
reference purposes.
To express the /7-parameters in terms of the z-parameters, first note that
995
Vl-
^11 hi2
h'
.^2.
.^1
.'^2.
and
Vl'
Zii
Z\2
.'^2.
Z21
^22. h .
The trick is to rewrite the jzf-parameter equations so that Vj and I2 are on the left with /j and Vj
on the right:
~ ^ \ lh - ^\\h
Z2 2 I 2 - ^ \ h ~ ^2
Writing these two equations in matrix form yields

O' /r
-Zi2
y ,-
Zii
Z22
h.
-Z 21
(18.56)
Solving equation 18.56 for the vector [V^ /j]ûnder the proviso that ^^2
T ,-
1 -Zl2
0 Z22
-1
Zii
O'
-Z2\
ZIIZ22 212^21
Z22
0 yields
h'
V2
(18.57)
Z12 r/i
-Z 2 1
V2
Thus, we have used matrix methods to direcdy compute the A-parameters in terms of the 2^-parameters under the condition that 222 ^
hu
h i'
211^22 Zl2^21
Z\2
h\
^22,
Z22
-Z 2I
(18.58)
All other relationships are derived in a similar manner. For example, to express j^-parameters in
terms of /^-parameters, one must rewrite the A-parameter equations so that and /j appear on the
left-hand side with
and
on the right. Then using matrix form and inverting the appropri
ate matrix under the condition of a nonzero determinant produces the desired result.
Vw>
Table 18.1 specifies the interrelationships among all the parameters studied thus far.
Chapter 18 * Two-Ports
FABLE 18.1. Interrelations among Two-Port Parameters, Where y Can

Stand for z,y, /;, or t in A/ = y ,, ^22 " /12 >'21
2-Paranietcrs
z-Parameters
y-Parameter
^11
^12
Z21
Z2 2
^22
A-^21
/;-Parameter
/-Parameter
j'-Paramcters
>'22
Aj
-> 21
Av
~>'12
Ay
>11
Av
A/;
>11
>21
/!21
hn.
/222
>12
>22
!h i
I1 2 2
J_
12
Ac
/-Parameters
//2i
_- h 12
^21
^21
ti2
/21
h\
[^22
-A t
^12
^12
-1
^11
/12
^12
^12
A/
^11
Az
A'
^12
^22
Z2 2
>I1
>11
^21
>21
A>-
Z22
Z2 2
>l 1
>^l 1
^11
A zl
-> 22
-1
-A/2
- h 11
^21
^21
>21
V21
/?2i
^22
-^ >
-> 11
-/?22
_^21
^21
>21
J 21
/2.21
/Ml
Ar
/Ml
A::
fh 1
[^11
/M2
^22
/22
-1
^21
^22
^22
Il2\
^11
^12
-1
^21
^22
h')')
Chapter 18 * Two-Ports
Exercises. 1. Use the code below to create an m-file in MATLAB for conversion o f 2-parameters
to /^-parameters. Verify that Z = [1 2;3 4] produces H = [-0 .5 0 .5 ;-0 .7 5 0.25].
% convert z parameters to h parameters
function [h, hi I,h l2 ,h 2 1 ,h 2 2 ] = ztoh(z)
zl 1 = z ( l,l) ; zl2= z(l,2);z21= z(2,l);z22= z(2,2);
deltaz = z ir z 2 2 -z l2 " z 2 1 ;
h 11 = deltaz/z22;
hi 2 =zl2/z22;
h21 = -z21/z22;
h22 = l/z22;
h = [ h l l h l2 ; h21 h22];
2. Use the code below to create an m-file in MATLAB for conversion o f z-parameters to ^-parameters. Verify that Z = [1 2;3 4] produces T = [1/3 -2/ 3;l/ 3 4/3].
%converting z to t paramters
function [t,tl I,tl2 ,t2 1 ,t2 2 ] = ztot(z)
zl 1 = z (l,l); z l2 = z (l,2 ); z21=z(2,l); z22=z(2,2);
deltaz = z ll * z 2 2 - z l2 * z 2 1 :
tl 1 = zll/z21;
t l2 = deltaz/z21;
t21 = l/z21;
t22 = z22/z21;
t = [ t i l tl2 ; t21 t22];
3. Write m-files for the remaining items in the conversion table for your own future use.
10. RECIPROCITY
Writing node equations for an ordinar)' linear circuit leads to a matrix equation having the form
A'
>11
y\n
Vi'
vî
=Y
y'n1 y'nn
Often the node admittance matrix Y = [y^^] is symmetric, i-e-.7,y = yjj for ; ^ J. Such networks are
termed reciprocal.
998
Chapter 18 Two-lorts
RECIPROCAL NETWORKS
Any circuit that has a symmetric coefficient matrix either in a nodal equation or loop equa
tion representation is said to be reciprocal. Further, a two-port represented by either ^-param
eters or ^-parameters is said to be reciprocal if Z j 2 = Z21
~^2r
From Chapter 3 we know that circuits without dependent sources have symmetric coefficient
matrices in both the nodal and loop equation representations. On the other hand, the symmetr)'
o f the z-parameters and j'-parameters is typically lost when dependent sources are present. In gen
eral, we can prove that a circuit containing R s, L's, C s, and transformers, but no dependent sources or
independent sources, is a reciprocal network. From the definition o f a reciprocal two-port, we can
conclude further that if the hybrid and/or f-parameters exist, then from Table 18.1
= ^21
and/or
Conversely, any two-port that has parameters satisfying these conditions is said to be reciprocal.
Exercise. Recall that the relationship between the ^-parameters and z-parameters is
-11
-12
'II
A/
^21
^21
ho
^21
^21
^21
Show that if 2 j 2 = ^21

if A/ = 1, then Zp = z^j.
Ar = 1, where A denotes determinant. Then show the converse, i.e.
Proving that any two-port created from a reciprocal network has symmetric z-parameters is
straightforward. We write the loop equations in matrix form with
/p
and /, being the volt
ages and currents o f ports 1 and 2. Since the underlying netw'ork is reciprocal by assumption, its
loop equation has a symmetric coefficient matrix partitioned as shown:
Vi-
^11
-12
-1 3
vs
-12
-22
^23
-In
^13
^ 23
^33
-3/j
'1/j
'-2/1
-3/?
VV
999
Solving for [Kj ^-5]^ by the method o f matrix partitioning yields

V,-
^^2
-li
\V,| - VV,2VV2VVV2i
where the W-j matrices are defined in the obvious way.
^11
^12
/l1
C21
Z2 2
j is symmetric. Since the inverse o f the
symmetric matrix Wj-y is symmetric, and since the sum o f two symmetric matrices is symmetric,
the resulting z-parameters are also symmetric. The symmetry o f the ^-parameters follows by the
symmetry o f the inverse o f a symmetric matrix.
We now set forth physical interpretations o f reciprocit)^
Reciprocity Interpretation 1: Consider a reciprocal two-port, N. If the voltage inputs,
are
the same as in Figures 18.35a and b, then by reciprocit)' the zero-state short-circuit responses, I 2 a and l-^y, coincide.
FIG URE 18.35 Equivalence o f short-circuit zero-state responses induced by voltage sources for
reciprocal netw'orks.
What reciprocity interpretation 1 says is that if one applies a voltage at port 1 and measures the
short-circuit current at port 2 with an ideal ammeter (zero meter resistance), then applying the
same voltage at port 2 would result in measurement o f the same short-circuit current at port 1.
Conversely, we can show that if reciprocit}' interpretation 1 is true, then the j'-parameters are sym
metric. To see this, observe that the configuration o f Figure 18.35a implies
1()()()
_
^<>1
Vo I ----
ha
and the configuration o f Figure 18.35b implies
_ Ôllt
2 -
în r,=o
Thus one must havej/p = 72 i>
^2h
symmetric j^-parameters.
Reciprocity Interpretation 2: Consider a reciprocal rvs'o-port, N. As illustrated in Figure 18.36,

is the same in Figures 18.36a and 18.36b, then the open-circuit zero-state responses V-^
and
coincide.
FIGURE 18.36 Equivalence oF open-circuit zero-state responses for a reciprocal network.
Reciprocit)' interpretation 2 says that if one injects a current at port 1 and measures the voltage at
port 2 with an ideal voltmeter (infinite input resistance), injecting the same current at port 2
would result in measurement o f the same voltage at port 1.
Exercise. Assuming reciprocit)' interpretation 2 is true, show that it follows that
2-parameters are symmetric.
^2V
1001
Reciprocity Interpretation 3 : Consider a reciprocal two-port, N. As illustrated in Figure 18.37,

leads to a zero-state response denoted ~ hii as shown in Figure 18.37a, and
leads to a
zero-state response denoted
as in Figure 18.37b. For a reciprocal two-port, the short-cir
cuit current ratio in Figure 18.37a is equal to the open-circuit voltage ratio in Figure 18.37b.
Exercise. Show that if reciprocity interpretation 3 is assumed true, it follows that A|, =
the forw ard short-circuit current gain,
în
, equals the reverse open-circuit voltage gain,
\i.e.,
.
V;
The three reciprocity interpretations have rigorous proofs. They however, are beyond the scope ol
the text.
Finally, a two-port that is reciprocal has an equivalent circuit representation with no dependent
sources. For example, suppose a reciprocal two-port has the z-parameter representation
V\ - 2|I /j + 2|2 A
and
~22
^2
1002
in which case
(18.59a)
and
(18.59b)
liquations 18.59 have the so-called T-equivalent circuit interpretation given by Figure 18.38.
I.
^22 ^12
-o
+
V.
V,
oFIGURE 18.38 2-parametcr T-equivalent circuit of a reciprocal 2-port.

The resistors represent general impedances.
Exercise. Suppose the r\vo-port o f Figure 18.39 is reciprocal and modeled by ^param eter equa
tions. Compute K, ^ Yj, and
in terms o f the ^/-parameters, |, j/p,
AN SW ER:
1 .J'l2 -^22
O-
-'12
-o
-I-
-I-
V.
V,
oFIGURE 18.39 ^parameter jt-equivalent circuit of a reciprocal rwo-port.

The resistors represent general impedances.
11. PARALLEL, SERIES, AND CASCADED CONNECTIONS OF

TWO-PORTS
A general linear two-port has four external terminals for connection to other networks as illus
trated in Figure 18.40a. When the input and output ports have a common terminal, only three
external terminals are available, as indicated in Figure 18.40b. Such a rwo-port is typically called
a common-ground tivo-port, although the common terminal is not necessarily grounded in the
sense o f being connected to earth.
1003
V.
General
two-port
V,
V.
'^Common'^
ground
two-port
V,
W
1.
L
(b)
(a)
FIG URE 18.40 (a) A general rwo-port. (b) A cornmon-ground r\vo-port.

Interconnecting two common-ground rvvo-ports
and
forms a new two-port N. Figure
18.41 shows three typical interconnection structures: parallel, series, and cascade. To avoid over
crowding o f symbols, Figure 18.41 omits all voltage and current reference labels. These labels are
understood to conform to those in Figure 18.40a. In particular, note that at each port the current
entering one terminal must equal the current leaving the other terminal for the two-port param
eters to be valid or meaningful.
An interconnection o f rwo-ports has a new set o f z-, y-, /;-, or ^-parameters obtained very simply
from the individual t^vo-port parameters. The interconnected rwo-ports o f Figure 18.41 have new
parameters computed from those o f
1.
and
as follows:
For the parallel connection o f Figure 18.41 a.

(18.60)
2.
For the series connection o f Figure 1 8 .4 lb ,

(18.61)
3.
For the cascade connection o f Figure 1 8 .4 lc,

(18.62)
(a)
(b)
1004
Chapter 18 Two-Porrs
-
Na
-#
#
N
(c)
FIGURE 18.41 Three typical interconnections of rvvo-ports. (a) Parallel connection,
(b) Series connection, (c) Cascade connection.
where Y, Z, and T denote the admittance, impedance, and transmission parameter matrices,
respectively, and the subscripts a and b refer to the networks
and N^, respectively.
A derivation o f these formulas is straightforward. For example, from the definitions o f
and N^,
where
Vi.
r/i.
ha
and similarly for the voltage and current vectors o f N^. From Figure 1 8 .4 la,
K ^ h - ^ substitution yields I - (K^+ Y^Vy w'hich verifies equation 18.60.

To verify equation 18.61, consider Figure 1 8 .4 lb. Here
= /. By direct substitution V = (Z^+
and
and
= Z^/^. But V=
verifying equation 18.61.
Equation 18.62 w'as derived earlier in this chapter.

The derivation o f equations 18.60 and 18.62 is easily extended to the case o f more than two rwoports;
1.
If two or more common-ground t\vo-ports are connected in parallel, then

K=
2.
k. h-
(18.63)
If two or more two-ports, common-ground or not, are connected in cascade, then
r = r ,r ,r ,...
(18.64)
Equations 18.62 and 18.64 for the cascade connection hold whether or not the component rwoports are o f the common-ground t)'pe. However, equations 18.60 and 18.63 for parallel connec
tions in general \\o\(\ only for common-ground two-port connections, as shown in Figure 18.4 la.
Similarly, the series connection equation 18.61 holds only for the case illustrated in Figure 1 8 .4 lb.
Series connection o f two general rwo-ports (Figure 18.40a) or series connection o f more than t\vo
common-ground two-ports (Figure 18.40b) requires an ideal transformer for coupling, as demon
strated in the homework problems. Examples 18.19 and 18.20 explain why equations 18.60 and
18.61 fail when two non-common-ground two-ports are connected together.
100=)
E X A M PLE 18 .1 9 . This example illustrates the difficult)^ with a non-common-ground series con
nection. Consider the two-port shown in Figure 18.42, which is a series connection o f two com
ponent two-ports. The z-parameters o f the individual two-ports are given by
2
and
Zf,u -
Show that Z 7^:
+ Zy when
(b)
Show that Z =
+ Zy if
(c)
Justify the statements o f parts (a) and (b).
(a)
=6
Rx+2
/?3 + 2
and R-^ = 3 Q..
= Rj = 0.
FIGURE 18.42 A series connection that causes difficulty.

S o l u t io n
(a) Observe that the parallel connection o f the 6 H and 3 ^ resistors is 2 Q. Thus, by direct cal
culation, the z-parameter matrix o f the interconnected two-port is
Z=
6
4
4'
6
2
^ Zn
i + Z;,
I) =
r
2
-1-
10
2-
(b) With R ,= R . = 0,
Z,, =
On the other hand, by direct calculation, the z-parameter matrix o f the interconnected two-port is
4
Z=
24
+ Ay =
(c) In part (a), Z ^ Z^ + Z^ because, after the interconnection, neither
nor
acts as a two-
port, as defined in Figure 18.40. This can be understood by inspecting Figure 18.43, with the
indicated the loop currents.
1006
------- O-
-oV
V,
-------o-
o ---------^ s / s A FIGURE 18.43 A nonzero

W ith y?j and Rj nonzero, the mesh current
/|. Hence, for the left terminal pair o f
leads to /j ^ /*.
is in general nonzero. Observe that 1*^ = 1^ the current entering the top terminal does not equal
the current leaving the bottom terminal. With unequal terminal currents,
no longer has a z-
parameter characterization, because the 2-parameter definition requires equal currents entering
and leaving the terminal pair, as per Figure 18.40. On the other hand, if
= /?, = 0 in Figure
18.43, then the third mesh equation is satisfied for arbitrary values o f the mesh currents /j, Ij, and
7^, as there is no resistance at all in the third mesh. In particular, let
and the 2-parameter characterization o f
= 0, in which case /|j = 7j,
is valid. Similar arguments hold for N^, If
= R^ =
0, equation 18.61 holds because Figure 18.43 now has the same interconnection as depicted in
Figure 18.4 lb.
E X A M PLE 18.20. This example illustrates the problem o f a non-common-ground parallel con
nection. Figure 18.44 shows rwo two-ports connected in parallel. Before the connection, each
two-port has ^-parameters
0.7
- 0 .2
- 0 .2
0.7
S.
After the connection, by direct calculation, the new two-port has ^-parameter matrix
Y=
Clearly,
Y^+ Yy = 2Y^\n this case.
r 1.625
-0 .6 2 5
-0 .6 2 5 1.625
S.
lOO'
FIG URE 18.44 New rvvo-port N subject to invalid application of equation 18.60 for non-commonground rwo-ports in parallel.
The reason for the failure o f equation 18.60 under these circumstances is the same as for the cir
cuit o f Example 18.19. If a voltage source is applied to port 1 o f N, we find that the currents
and /j'^^ are not equal. Thus,
eters in forming N.
cannot continue to be characterized by a set o f rwo-port^-param
There are, however, some special cases o f non-common-ground two-ports for which equation
18.60 holds for a parallel connection. The following is one example.
E X A M PLE 18.21. fhis example illustrates how to achieve a parallel interconnection o f two gen
eral rwo-ports o f Figure 18.40a so that equation 18.63 remains valid. Reconsider the tw'o-ports
and
o f Example 18.20, which have y-parameter matrices
a 1:1 ideal transformer is placed at the front end o f
and Y^, respectively. Suppose
in Figure 18.45; call this new two-port
(a)
Show that the ^-parameter matrix o f N^* is Y^.
(b)
Show that the j-param eter matrix o f the interconnection o f

even though N^* and
are not common-ground two-ports.
and
is Y = Y^ + Y^,
1 008
FIG URE 18.45 Equation 18.63 holds for these non-common-ground rwo-ports. (a) Parallel connec
tion of rwo general rvvo-ports. (b) Justification of equal currents at rw'o terminals of each port.
S o l u t io n
(a) The )'-paramcrers o f N j arc rhe same as for

ed dor positions forces N * and
if N * and
because rhe ideal 1:1 transformer with indicat
to have the same port currents and voltages at port 1. In fact,
were enclosed in a box with only the leads observable, the two-port properties would
be identical.
(b) To show that the ^/-parameters of N (the interconnection o f N * and
must first show that
(ii)
and 12 1 , - / ,j
are Y =
+ Y^, we
l()())
To confirm these equalities, consider Figure 18.45b, which represents Figure 18.45a with all the
nonessential components removed to avoid overcrowding figure. Four Gaussian surfaces are
drawn,
5,^,
and 5,^^. Note that all Gaussian surfaces go through the core o f the trans
former. Recall that KCL holds for a Gaussian surface: the algebraic sum o f the currents entering
(leaving) the surface is zero. This immediately asserts the validit)' o f statements (i) and (ii) above.
Hence, within the interconnection, the two-ports N * and
continue to act as individual twoports. Therefore, equation 18.60 remains valid, i.e.,
n .4
Y = Y a -fY b =
- 0 .4
- 0 .4
1.4
This example extends directly to multiple parallel interconnections. Although the precise condi
tions for the applicability o f equations 18.60 and 18.61 to the parallel and series connections o f
non-common-ground two-ports are known, they are not practical enough to be included here.
Our emphasis is on interconnections o f common-ground two-ports, which occur most often in
practice.
12. SUMMARY
This chapter presented a unified setting for one-port analysis while providing a comprehensive
extension to two-ports. Two-ports are common to numerous real-world systems such as the utili
ty power grid that delivers power to a home from a generating facility through a transmission net
work. Another representative two-port is a telephone system that delivers a speakers voice to a lis
tener by sending a converted electrical signal through a transmission network. The characteriza
tion o f a two-port for such systems is done through their input-output properties. Four sets o f
characterizing parameters were developed: impedance or z-parameters, adm ittance or ^-parame
ters, hybrid or //-parameters, and transmission or f-parameters. In order to analyze various
aspects o f a system characterized by a two-port, formulas for computing the input
impedance/admittance, the output impedance/admittance, the voltage gain, etc. were derived.
Quantities such as voltage and power gain are very important aspects o f amplifier analysis and
design, as illustrated in Example 18.12, which depicts a two-stage transistor amplifier configura
tion. Although Example 18.12 utilized the medium o f z-parameters, the more customar)' medi
um for transistor amplifier design is //-parameters.
Conditions and formulas for parallel connection o f two-ports were presented in terms of^-parameters while series connections were studied using 2-parameters. Formulas for determining the
transmission parameters o f cascades o f rwo-ports were also developed. In addition, the chapter
introduced and interpreted the notion o f reciprocit)' in terms o f the different two-port parame
ters. Reciprocal circuits generally contain only R's, Z.s, Cs and transformers. Under certain restrict
ed conditions a reciprocal network may contain a dependent source, as the homework problems
will investigate.
101 U

Admittance or j-parameters: descriptive two-port parameters in which the port currents are
functions o f the port voltages.
Hybrid parameters or /^-parameters: descriptive two-port parameters in which Kj and
are
expressed as functions o f /j and K,.
Impedance or z-parameters: descriptive two-port parameters in which the port voltages are func
tions o f the port currents.
Input admittance; the admittance seen at port 1 o f a possibly terminated two-port.

Input impedance; the impedance seen at port 1 o f a possibly terminated two-port.
Matrix partitioning: the partitioning o f a matrix set o f equations into groups to obtain a simpli
fied solution in terms o f the partitioned submatrices.
Norton equivalent of one-port: a current source in parallel with the Thevenin impedance.
Open-circuit impedance parameters; the impedance or z-parameters.
Open-circuit output admittance: the hybrid parameter hj-,Output admittance: the admittance seen at port 2 o f a rvvo-port possibly terminated by a source
impedancc.
Output impedance: the impedance seen at port 2 o f a two-port possibly terminated by a source
impedance.
Partitioned matrix; a matrix that is partitioned into submatrices for easier solution o f sets o f
equations.
TU-equivalent circuit: equivalent circuit o f a reciprocal two-port containing three general imped
ances in the form o f k, as in Figure 18.39.
Reciprocal network: a network whose node equations or loop equations have a symmetric coef
ficient matrix.
Reciprocal two-port; Z]2 = ^21

~J2V
Reverse open-circuit voltage gain; the hybrid parameter A jj.
Short-circuit admittance parameters; the admittance or ^'-parameters.
Short-circuit forward current gain; the hybrid parameter
Short-circuit input impedance; the hybrid parameter
T-equivalent circuit; equivalent circuit o f a reciprocal two-port having three general impedances
in a T shape as in Figure 18.38.
Terminated two-port: a two-port attached to a load impedance and a source with, in general, a
nonzero impedance.
Thevenin equivalent of a one-port: Voltage source in series with the Thevenin impedance.
Transmission or ^-parameters: parameters where
and
are expressed as functions o f Vj and
-h -
Two-dependent source equivalent circuit: equivalent circuit tor a two-port containing imped
ances/admittances and two dependent sources.
P r o b le m s
+ V -
ONE-PORTS
1. In Figure PI 8.1, suppose
n , /?^ = 1 k n , and (3 = 49. Find
port. Then find
100 V,Z^ = 20
o f rhe one-
and the power to the one-
port. Does Z j have any effect on the answers?
Figure P I8.3
C H EC K : (a) 1 1 k n
= \ 0 j2 cos(lOOOf) A, a = \0, R
Suppose
= i o n , and C = 1 îF.
Figure P I8.1
(a)
Find the Thevenin equivalent seen by
(b)
Find the load impedance for maxi

mum average power transfer and com
the load.
C H E C K : 50 V
2. In Figure P I 8.2, suppose
2500
= 100 V, Z j =
pute the resulting average power.
= 50 Q, and (3 = 49. Find Z-^ o f the
one-port. Then find
and the power deliv
(c)
Find a series RL load that achieves

maximum power transfer.
ered to the one-port. Does Z^ have any effect

on rhe answers?
1 :a
'(t)
Figure P i8.4
C H EC K ; (c) 1 kQ resistor in series with 1 H
inductor
Figure P I8.2
C H EC K : 50 W
5. T h e loop equations in standard units

describing the one-port o f Figure P I 8.5 are
3. Consider the circuit o f Figure P I 8 .3 .
given by
Suppose C = 0.1 mF, y(;;(0) = 10 V, Z j = 1 kH,

Z2 = 2
Z 3 = 1 kl^, and
vr
= 4 mS,
(a)
Find the input impedance Zy^^.
(b)
Find v^r) for t > 0 .
0
(a)
10
-a
0.5
-1
0.5
Find the input impedance o f the oneport as a function o f a. Hint: Consider
(b)
Cramers rule.
If = 2 and a voltage source is con
nected so that V|(/)= 12V 2 cos( 2 /)
V, find the average power consumed
1012
8. Consider the circuit o f Figure Pi 8.8. In solv
by the one-port.
(c)
Repeat part (b) for a = (1+ ;4/3) at
CO
ing this problem, fully utilize the properties o f

an ideal transformer.
=2 rad/sec.
(a)

as a function o f the turns ratio, b, and
the resistance, R.
i-Port
(b)
If
= cos(10r + 45) A, find the
Norton equivalent circuit with the

Norton source represented as a phasor.
Figure P I8.5
AN SW ER: (a) Ga U, (b) 12 W. (c) 8.64 \V
(c)
If /? = 25
^ = 2, and the output is
terminated in a parallel LC, compute
L and C so that the bandwidth is 10
6. Consider the circuit in Figure P I8.6.

(a)
(b)
and the resonant frequency CO^ = 5
using the method o f matrix partitioning.
rad/sec.
(d)
If the circuit is terminated in a 1.2 Q

resistor, find
Compute the zero-state response with

the above input and element values.
for r > 0 assuming
Identify the steady-state part. I f the
the initial inductor current is zero.
input cosine frequency were changed

to 100 rad/sec, what would happen to
the steady-state magnitude?
R
-------syx/'-----b :1
Figure P i8.6
ANSWKR; (a) Z,;, = .s- + 0.8. I
Figure P I8.8
= -2 /
hR
2h~ - 2 h + \
7. Consider the one-port circuit o f Figure P I8.7.

(a) Find the Norton equivalent using the
methods of matrbc partitioning and node
analysis.
(b)
Making use of the information obtained

in part (a), compute the impulse and step
responses o f the circuit.
40
^tUII
2h~ - 2 h + \
C= 0.02 F and
/. = 2 H
/-P A R A M ET ER S
9. Find the v-parameters o f the two-port in
Figure P i 8.9.
40
in
R
^lll
I "\/V^
I
20
'2 0
J_ '
16
Figure P I 8.7
CH ECK:
= 0.5
Figure P i 8.9
10.
1013
Consider the circuit o f Figure P i 8.10,

and G 2 denote conductances.
where
(a)
Compute the short-circuit admittance

parameters.
(b)
Suppose port 1 is short-circuited and a

voltage Vjit) is applied to port 2. Find
Figure P I8.12.
I^{s) and I~,{s), in terms o f the literals

and
AN.SWl-RS: in random order: ()',| - Vp). (V| |

0'22 ^.>12^
+
1: a
V.
13 . Consider the circuit o f Figure PI 8.13.

(a)

parameters.
O------------(b)
Figure P I8.10
>11
>12
>'21
>22
Reverse the process and compute K,,
Yj, Yy and g^^^ in terms o f the yparameters.

which
and
(a)
= 2, /?, = 2 Q,
= 16 Q,
= 320 Q,
(c)
Suppose K| = K2 = K3 = (^ + 1),
= 80 Q.
is+\r
parameters.
(b)
V/, is) =
(c)
and port 2 is short-circuited. Find
If port 2 is short-circuited and
/,W, AW , /jCr), and /^C/).
64
------- , find /, is) and I,{s).
5 + 16
If port 2 is terminated in a 240 H

resistor, /?j = 6 Q, R2 ~
64
V] ( 5 ) = ------- , find /, W and
.v + 16
Figure P I 8.13
14. In the circuit o f Figure P 18.14, /?, = 2 0 .,
= 2 Q, /?3 = 2 Q, C, = 0.5 F, Cj = 0.25 F,
and C3 = 0.5 F.
(a)
Compute the v-parameters using the

method o f matrix partitioning
Figure P I8.11
(b)
If port 2 is short-circuited, find /,(j),

/2W, i]U), and ;-,(/) assuming v^{t) =
12. The two-port o f Figure P i 8.12 has yparameter matrix
yu^=
".V li
.Vi 2
>21
}'22
Compute Kj, V-,, Yy and^^^^ in terms o f th e jparameters.
\Ou{t)V.
1014
17. Reconsider the circuit o f Figure Pi 8.15, in
1n
which
He
He
2F
1F
10
2F
in
.'^11
.^12'
>21
.'22.
60
= 1 k n , and
- o .r
1
m S,
= 2 kn .
(a)
Compute K,,,. Z^. Y^,, and Z,.
(b)
Find the voltage gain
Figure P I8.14
G,. = ^ .
K,
15. The terminated two-port configuration in

(c)
Figure PI 8.15 is characterized by ^-parameters.
If v^{t) = 9u{t) V, find the power

absorbed by R^.
Suppose
V/ -
ANSW FRSM a) ); = 8 m S ,Z ,,= 1 2 5 a

= 2.2 mS; (b) - 4 .4 4 ; (c) 0.8 wart
(a)
Find^, j i f = 10 ^^,721 = - 1 S .^,2
(b)
= 0.03 S ,y 2 2 = 0.2 S, and R ^ = \ O n .

If v^{t) = 3 0 i4 {t) V, find the power,
pÛ), absorbed by the 10 Q load at

port 2.
18. Reconsider the two-port o f Figure P I 8.15,

in which R^= 10 H ,^ 2i = 2 S ,^ p = 0.02 S ,^22
= 0.2 S, R^ = i o n , and the voltage gain is
= 0 .6 .
(a)
Findj/j,.
(b)
If v^{t) = 10(r) V, find the power

absorbed by Rj^.
C H EC K : (b) 160 watts

Figure P i8.15
19. Consider the circuit o f Figure P i 8.19, in

which R^ = 2 Q , Rj = 8 Q, R^ = 52 Ol, P = 3,
and /U = 4.
(a)
Com pute the ^-parameters o f the
(a)
Com pute the >'-parameters o f the
(b)
boxed rwo-port.
If port 2 is terminated in a 2 S resistor,
compute
and the voltage gain
(b)
Compute the input admittance,
(c)
Compute
(d)
If i^{t) = 5u{t) A, compute y,(r).
_ V jis)
G .,=
Vds)
boxed two-port.
o f the complete circuit.
.(s)
Figure P i 8.16
20. In a laboratory, you are asked to determine

the admittance parameters o f a circuit. You
1015
decide to shon-circuit port 2, place a unit step

current source at port 1, and measure the port
voltage, t/j = (1 V, and the pon 2
current, /2 (^) =
A. Knowing that this is
sufficient to determine at most two of the
parameters, you then break the short circuit
and terminate port 2 with a 1
resistor and
measure the new step responses as = (1 +
V and ij{t) =
A.
(a) Compute the/-parameters of the twoport.
(b) If port 2 is terminated in a 1 Q resis
tor, find the input impedance seen at
port 1.
(c) If port 2 is terminated in a 1 resis
tor and driven at port 1 by a current
source i^{t) = cos{t)u{t) A, compute
the steady-state magnitude of the gain,
network assuming y2 (/) is the output.

If vj^t) =10 s'm{2t)u{t) V, compute the
steady-state and transient responses
assuming V2(t) is the output. Use the
residue command in MATLAB to
compute the partial fi^ction expansion.
(d)
Figure PI8.22
CHECK: (b) 0.25 / (s + 1), (d) transient
response e~û{t) V, steady-state response
(-cos(2f) + 0.5 sin(2/))(f) V
Yz
23. Figure P I8.23 represents a two-stage ampli
fier. Suppose
21. This problem shows that one can simulate

an inductor using an active two-port terminat
ed by a capacitor. For the circuit of Figure
P18.21, Cj = 125 juF, C2 = 0.8 F,/u =>^22 = 0
-JVi2 = 4 S.
(a) Compute the input impedance Z-JJ).
What is the equivalent L seen at port 1?
(b) Determine the resonant frequency CO^
(c) If /? = 2 5 is placed in parallel with Cj,
find the new resonant fi^juency 0)^
o-n eY=
CHECK: (a)
240 rad/sec
y.. y.j
Figure PI 8.21
= 50 mH, (b) 400 rad/sec, (c)
22. For the circuit of Figure PI 8.22, R^= I Q,

=1 Q, and q = 1 E
(a) Find the ^parameters.
^2 (5 )
(b) Find the voltage gain, Gy2 = y
(c) Compute the impulse

^
and step responses of the terminated
-0.64'
r, =
25
mS, Y2 =
0.4
- 0 .0 0 r
7.5
0.025
= 150 Q,
= 2 kn, and /?! = 2 kQ. Find
the voltê gain
z-' _ VOM/
Although the solution may be obtained by solv

ing a set of six simultaneous equations, a much
better method that also gives more insight into
the performance of the amplifier, and works for
any number of stages, is to proceed as follows:
(a) Find the input admittances y.nl and
iî-
(b)
Find the voltage gains of the stages

successively, starting from the source
end. Use this information to find the
overall voltage gain,
G, = ^
în
101(>
z-PARAMETERS
Stage 1
25. (a)
Y.
Compute the z-parameters o f the

mutually coupled inductors in Figure
P I 8.25 when (i) the dots are in posi
tions A-B and (ii) the dots are in posi
tions A-C,
Figure P I8.23
(b)
AN SW ERS: (a)
(b )6 V = S 0 0
port by inverting the z-parameters
= 500 LlS:
= 10 inS,
Compute the y-parameters o f the twomatrix. Do the y-parameters exist if

the coupling coefficient ^ = 1?
24. Consider the switched circuit o f Figure

P i 8.24, in which C = 0.25 F and R^= \ Q.. For
t < 0.25 sec the switch is in position A, and for

t>
0.25 sec the switch is in position B. Suppose
the y-parameter matrix is

S.
K(.v) =
(a)
Figure P I8.25
With the switch in position A, find

(a)
V2(-v)
V'|(5)
Compute the open-circuit impedance

parameters.
(b)
(b)
If
(c)
0.25 sec.
Is rhe circuit stable? Explain your rea
= -4u{t) V, find
0 < r<
If port 2 is terminated in Z-,(s), find

the input impedance,
(c)
If port 1 is open-circuited, Zj(s) is a 1

H inductor, Z-,(s) is a parallel combi
soning.
nation o f a I Q resistor and a 0.5 F
id)
Find i/-)(0.25~).
capacitor, and
(e)
For t > 0.25, after the switch has

moved to position B, draw and label
2K
the frequency domain equivalent cir
S-+4
cuit that will allow one to compute
(0
y^(s).
Compute Zj(s) if ^ = 2.
(g)
Compute
I, A
-^ s/ V -
for /> 0.25.

B
find f,(r) and v>y(r) in steady state.

1 :a
z.
b;1
Y=
Q v ,
Vm y,2
Figure P I8.26
C H EC K : (c) i^2 sss(^) = 0.707A>/lsin(2/-45) V
F'igure P i8.24
C H EC K : (d) i^M 25) = 3.463 V, (g)
3.436 f-('-'^ -5 ),; ( / - 0 . 2 5 ) V
27. Consider the circuit o f Figure P I 8.27, in

which /?, = 35 Q, /?7 = 50 Q, 7?^ = 800 Q,
= 100 Q, and a = 4.
1017
(a)
(b)
Compute the z-parameters.

Find the ^parameters by matrix inver
sion using, for example, MATLAB,
If port 1 is open-circuited and
(c)
------ ^
r
11
It
0.5 F
-------
\f
1
< 1
_ 0.25 F
>
*
r\
0.5 F
----- 0
D-----
Figure P i8.29
1:4
-> / W -
R.
60
??1
-100
??
10
+ Zîs) = , Z^ = 10 Q, and
V,
tS^
(a)
(b)
(c)
Figure PI8.27
CHECK: (a) Z =
30. Consider the circuit of Figure PI8.30, in

which Zj(s) = 5 0 , Z2(s) = 10 n ,
Compute the z-parameters.

Compute Z-J,s).
If v-^{t) = 10tt(/) V, find /j(/) and /Û).
28. Consider the two-port of Figure PI8.28.

(a) Compute the open-circuit impedance
parameters in terms of the Z- and r^.
(b) If the two-part has z-parameter matrix
'^11
m '
^21
^22
find Z j, Z2 , Z3 , and
the ^-parameters.
(c)
3 1 . The
in terms of
port 2 is short-circuited, and v^{t) =

10W V, find /j(?) and
z,
circuit of Figure PI 8.31 hasZ^= 10 Q,

and 2 -parameter matrix
Uij] =
V,
V.
-o
Figure PI8.28
CHECK:
(a)
(b)
(c)
5^ + 100 1
205
79
Use the result of Problem 28 part (b)

to draw an equivalent circuit having
these z-parameters.
The circuit is resonant at what value of
Compute the Q of the circuit.
Plot the magnitude and phase
response of Zy^(/co), using a program
such as MATLAB.
Z] H- Z3 Z3
99
29. Use the method of matrix partitioning to

find the z-parameter matrix of the circuit in
Figure P I8.29.
O '
Figure PI8.30
If Z iW = Z 2 W = ^,Z3W = l . r ^ = - 5 ,
-------
= 10 .
Figure P I8.31
1018
output o f the two-port to the load.
32. Consider rhe circuit o f Figure P i 8.32.

Suppose Vy(0 = 20 V 2 co.s(w/);/(0 V,
= 1
Q,
(b)
= 1 Q, and the z-parameter matrix is
s+
Z=
(a)
(b)
0.5
0.5
0.5
,v +
What is the maximum average power

absorbed by
= 2
when
/^(/) = 2>/2 cos(2/) A, a = 2 , and/?Q
= 50
0.5
Find Z-W and

Find Vjis) and the average power
b :l
aR
2-Port
absorbed by Z^.
For CO = 0 and cu = 1 rad/sec, find the
(c)
Figure P I8.33
average powers absorbed by Z^.

(d)
Using the equivalent circuit o f Figure
CHECK: (a) (??),
79
18.38, construct a circuit having the

given 2-parameters.
5a
'V R,
; (b) 2 watts
34 . T he two-port shown in Figure P i 8.34 is

called a g)Tator. A g)'rator terminated in a
capacitor beiiaves like an inductor at its input
terminals. This is one o f the many ways to sim
''O
ulate inductors in the design o f an active fil
2-Port
ter. Suppose the active two-port in Figure

P I 8 .34 , built with resistors and ideal opera
tional amplifiers, has z-parameters
Figure P I8.32
ANSW ERS:
2.v' + 2.V- + 2.V + I
1000
-1000
(a)
0.25
1 + uy
a surprise to you?
(b)
V + 2.V" + 2.V +
and
: (c) 100 w and .50W
If C-, = 10 nF, find Z (j), the imped

ance looking into port 1. Is the result
(?).v- + (?).v-(-(?)
0.5
Q and C, = 100 pF.
If
= R j= 100 kQ and rhe output is
Kj(/), find OJ,;; and the bandwidth o f

the circuit.
(c)
If
=/?-,= 10 kQ and the output is

find the impulse response, h{t),
o f the circuit.
(c) 100 W an d SOW
(d)
Find thej'-parameters o f the two-port,

using the property that theZ-matrix is
3 3 . Consider the two-port configuration of
the inverse o f the K-matrix (and vice
Figure P18.33 having z-parameter matrix
Z=
aRn
a~Rv
2a/?o
2 6 a /e
versa).
(e)
If /^] = /?2 = 1^^ kD, and the output is
I'lU), what is the impulse response?

Hint: Make use o f the results o f parts
(c) and (d), and the two-dependent
(a)
Determine the value o f b that yields

maximum power transfer from the
source equivalent circuit in terms o f

the v-parameters.
1019
The two-port
hasj^parameters and the twoport Ny has ^-parameters as given below (with
standard port labeling and units):
2-Port
'O
0.1
2 (s)
Figure PI8.34
- 0.2
35. The stages in the circuit of Figure P I8.35

have i^parameter matrices
Z, =
2
-1 0 ^
01
20
62.582
1.20751
63.75
1.25
respectively.
(a) Compute the input impedances, Z -^2
andZ;,.
(b) Compute the voltage gain, VgJVj(c) Compute the power gain,
p
^ o u t^ o u t
(a)
(b)
(c)
kQ
(d)
0 . 1'
36
0.1
40 4
FindZ^,
FindZ,..
Find the gains
V,
G .= ^ ,G 2 = ^ ,G 3 = ^ ,a n d
Ha
Hfe
Find the power gain of the circuit, i.e.,

the ratio
p
_
S o in -
^ o u t^ o u t
..
vi/,
37.
^ equivalent circuits for a pair of coupled
put impedance of the amplifier circuit.
inductors using z-parameters. Each equivalent
circuit consists of one ideal transformer and
I.
I,
L
two inductances. The analysis of a coupled
circuit with the use of such equivalent circuits
'len is very often more illuminating than writing
and solving simultaneous equations.
(a) Find the z-parameters of the two-port
Nj of Figure PI 8,37a.
(b) Show that two-port N j of Figure
Figure P18.35
P I 8 .3 7 b has the same z-parameters as
N j.
CHECK: m = 2 kii and ^y = -9.8147
(c) Show that two-port N3 of Figure
''s
P18.37C has the same z-parameters as
36. Consider the cascaded two-port in Figure
N,.
P I8.36, in which = 10 Q, /?| = 20 Q, and Rj^
(d) Use the equivalent circuit N2 of pan (b)
=4Q .
and the properties of an ideal trans
former to find 0)^, the bandwidth, and
' ^
Vi max
approximate values for the half-power

frequencies for the coupled tuned cir
cuit of Figure P18.37d.
Figure P I8.36
This problem shows
10 2 0
1 :n
(d)
(a)Two-port N,
Figure P I8.38
m v_o
39. Find the /;-parameters for each two-port
(1-k^) L,
shown in Figure P I 8.39, assuming standard

port voltage and port current labeling, with R =
2 kf2, C = 0.1 mF, and = 0.1.

ideal transformer
1 :n
(b) Two-port Nj
k y ~ L ,: >/l .
o -fY Y \
(1-k^) L,
NT
N2
Ideal Transformer
Figure P I8.39
ideal transformer
(c) Two-port Nj
40. Find the /^-parameters for each two-port
k= 1
shown in Figure P i 8.40, assuming standard

port voltage and port current labeling, with R =
2 k n , C = 0.1 mF, and w = 10.
(d) Coupled tuned circuit
Figure P I8.37
(a)
/i-PARAMETERS
38. Find the /^-parameters for each two-port
shown in Figure P I 8.38 assuming the standard
labeling and units.
----- o
o----- r^ s/ V '------ o
Figure P I8.40
41. (a)
-o
o-
(b)
(a)
1: n
Find the /^-parameters for the two-port

of Figure PI 8 .4 la assuming that the hparameters o f the two-port N, are
Z,
A
111I
(0
(b)
1021
Hint: What does the transformer do

to the output variables of the boxed
circuit? Break the problem up into
two separate parts.
Apply the result of part (a) to the cir
cuit of Figure Pl8.4lb.
rv / W -o z
+
2-Port
Figure PI8.42
43. Consider the amplifier network of Figure
P I8.43. Stage 1 is a common-emitter stage that
drives stage 2 , the common-collector stage.
Such an amplifier combination might be used
to drive a low-impedance load. Suppose Z^ = 2
Idi, Z^ = 3 kn, Z^ = 64 2, and the A-parameters in standard units of the two stages are
O '
(a)
H ,=
Ho =
Figure P18.41
42. The /f-parameters of the two-port of Figure
P18.32 are
= 250 SI, h -^2 - 2-5 x 10^, />2 i
= 125 , h22 = 2.25 mS, -2^ = 1 kH, and
=
500 a
(a) FindZ. andZ^^,.
Find the gain
(c)
Find the power gain of the circuit, i.e.,

the ratio
Suppose a capacitance of 5 |AF is con

nected across Dort 2, Suppose
V j(/ ) = 10V2cos(400f) V. FindKjW
in steady state and the average power
absorbed by Z^ = 500 Q. Hint: Obtain
the Thevenin equivalent circuit seen by
the Z^ and C combination.
50
0.05x10-3
nooo
0.966
-51
0.8x10-3
Find the input impedances of

stages successively, starting from
load end.
Find the output impedances of
stages successively, staning firom
source end.
Find the overall voltage gain,
^
''out
C .-
(d)
V,/,
(d)
(b)
(c)
(b)
(a)
2000
the
the
the
the
A 1 pF capacitor is inserted in series

with Z^ to prevent dc voltage in the
power supply (not shown in the dia
gram) from entering the signal source.
Because of this capacitance, low-fre
quency signals will be amplified less.
Determine the frequency (in Hz) at
which the magnitude
ôut - 0.707 X Max Value
Hint: Analyze the simple circuit con

sisting of Vp Zp C, and Z- .
1022
' 6
Common
Common
emitter
collector
stage-1
stage-2
Figure PI8.43
44. Reconsider Problem 43. If-2^ is adjustable,
find the value of
so that
= Z^ = 64 Q.
Then find Z^- and
G. = ^
(a) Compute ^22 i*' terms of

and pos
sibly other /^-parameters. Explain your
reasoning.
(b) Derive a formula for the current gain,
Gp'm terms of the A-parameters and Y^.
(c) Find />2 i(d) Suppose now that the source, - Z^,
at the front end of the two-port is
briefly disconnected and a voltage,
V2{t) = 10(^) V is applied to port 2.
The measurement v^(t) = -5u{t) V is
made. Compute h^2'
(e) Suppose the source is reconnected to
the two-port. If the current gain
45. For Figure P I8.45, the and ^-parameter

matrices (in standard units) of two-ports Nj
and N j, respectively, are
0.02
0.001
2.5
0.2
H;V2 =
10
0.051
-125
0.5
^ = 0.8
and if YXt#= 0.25 S, determine Z-in and

A .
If Zjr = 8 Q and Z^ = 25 12, find the input

impedance and the voltage gains,
for the cascaded two-pon.

p y s/ V -o -
2-port
N1
CHECK:
02-port
N2
Figure PI8.45
= 0.8 and G^2 ~ ~ 1 0 0
46. Consider the terminated two-port configu

ration in Figure P I8.46, with h-- indicating the
two-port /^-parameters. Suppose that (i) the
current through the admittance, Yj^ equals the
current through ^22 ?
the current gain
Figure PI8.46
CHECK: />2i = 200,
= 1800
L Q ,Z } = 2 k Q
4 7 . In this problem you are to design an ampli
fier circuit represented by the doubly terminated
equivalent circuit shown in Figure PI8.47. This
means you will be given certain amplifier speci
fications that will allow you to determine the
parameters of the amplifier circuit.
Amplifier specifications: R^ = 2 Q and
(i)
= 40 Q.
When /j is zero, the ratio
G, = -^ = 100;
and (iii) the source resistance is Z^ = 8 Idi.
(ii)
when a source is applied to port 2.

There must be maximum power trans
fer from the amplifier output to the
1023
load under the condition that
(b)
800 Q..
.....
(ill)
(iv)
Verify that the /-parameters o f the cir

cuit o f Figure PI 8.49b arc
V.
25
V,
26
_L =
O'
>2
T =
The voltage gain
Vo
= -1 0 0 .
(c)
Compute the /-parameters o f Figures

P I8 .4 9 c and d.
Given these specifications;

(a)
(b)
Compute Z/j 2Compute
-o
o-
-O
O-
^îd the turns ratio
a.
(c)
Compute//jj.
Compute the input impedance
(d)
(e)
(a)
Compue *2 ,
Compute the ratio o f the power deliv
(b)
-o
o-
-o
o-
ered to the load to the power delivered

to Rjj
(d)
(C )
Figure P i8.49
ANSWHRS:
1
2i
: (d)
1
>2
5 0 .(a)
Figure P I8.47
Z.
Show that the /-parameters o f the ideal

transformer o f Figure P I8 .5 0 a are
48. Repeat Problem 47 with the amplifier spec

T =
ification
(b)
= 0.01
//
\ h i
Given the answer to part (a), compute

the /-parameters o f the circuits o f
Figure P I8 .5 0 b and c.
when /j = 0.
2 xlO -
0.01
n: 1
AN SW ER: l/?,/l =
4W .3S
3.6538 x lO " '
(in standard units)

(a)
n: 1
t-PARAMETERS
4 9 .(a)
Verify that the /-parameters o f the cir

cuit o f Figure P i 8.49a are
T =
(b)
1024
n: 1
c^
JY Y Y -6
4H
(0
Figure P i8.50
Figure P I8.53
51. Find the /-parameters o f the ner\vork in
Figure P I 8.51.
54. The rwo-port o f Figure PI 8.54 is described

by r-parameters.
4; 1
(a)
2H
Derive the input impedance relationship

^ll^L +^12
ti\^L
Figure P I8.51
-4
ANSWlUl:
0
1
1
0
4_
(b)
2.v^
-4
Derive the output impedance relationship
-S.v
ôut ~
t~>2^s
network in Figure P I 8.52.

If a 14 volt source is applied at port 1,
(c)
V,
Derive the voltage gain Gj.] = .
Kv
(d)
Derive the voltage gain
Figure P I8.52
Figure P I8.54
7
12
53. Use the results o f previous problems to

obtain the r-parameters o f two-ports A, B, C,
and D in Figure P I 8.53. Then use matrix mul
55. Suppose a t\vo-port has both 2-parameters

and /-parameters. Compute the r-parameters in
terms of the z-parameters. Hint: Rewrite the zparameter equations in the form
Ml
= Mo
tiplication to obtain the r-parameters o f the

overall two-port.
= Ir.
^1
compute Vjit)-
ANSW l-R: (a)
hi
-0 .2 5
5 2 .(a) Find the r-parameters o f tiie cascaded

(b)
^22
-h
Vi
and invert the appropriate matrix to obtain
<21
-21
o.
'21
-21
102S
57.
- ZpZ-,j.
where Az =
A certain high-voltage transmission line
operating at 60 Hz is represented by a two-port

56. For the circuits of Figure P i8.56, suppose the r-
with the following r-parameters:
parameters in standard units o f the two-port N j are
y0.022, r ,2 = 40
0.1
-0 .0 0 1 5
-0 .0 0 5
current -I-, = 3 6 1 Z 0 A at V-, =

1 1 5 ,2 0 0 Z 0 V, find the sending end
Find the /-parameters o f the cascaded
(port 1) voltage Vj^ the current Ip the
two-port for each configuration.
power delivered by the source, and the
Find the steady-state /j(/) under the
(b)
j = 0.895 +
= - 2 .6 1 7 x 105
+ y i.l0 2 X 10-3
=o.895 + y0.022.
(a)
If the receiving end (port 2) draws a
0.1
R = 200 Q.,L = - 0 .5 H, and ;/ = 5.

(a)
y l8 0 iX
conditions that
= 10 cos(lOOr) V
power loss in the transmission line.

(b)
II V j = 1 3 4 .0 0 0 Z 0 V, and a resistive
and each rwo-port is terminated in a
load represented by a 500 Q resistor is
parallel RC circuit with ^? = 10 Q and
connected to the receiving end (port
C = 1 mF.
2), find the power delivered by the

source and the magnitude o f the load
voltage.
1 :n
2-port
V.I
NT
ANSW l-K; (a) V, = 135,550Z29..S7<' V, I, =

3 4 7 .3 6 Z 2 2 .8 6 A. 4 6 .733 megawatts. 5.146
megawatts; (b) 35.331 megawatts, 12".96 k\.
(a)
PARAMETER CONVERSION
AND INTERCONNECTION OF
TWO-PORTS
2-port
58. A two-port N has known 2-, y-, and h-
N1
parameters. A new two-port
is formed by
adding one single impedance Z or admittance
Y to N in various ways. Prove the following
(b)
I.
relationships bervveen the old and new -rwo-
'
port parameters.
+
(a)
2-port
V,
If Z is connected in series with port 1

o f N, then Zn
N1
^^\\,new =
= 2Ml
,, + Z and
+ Z. Other z- and h-
parameters remain the same.

(0
(b)
Figure P I8.56
- 0 . 0 0 l.v-0 .1
\i\ =
- 0 .0 0 2
If Z is connected in series with port 2

o f N, then 2^2 n rtv = -2=22 +
parameters remain the same.
0.02
0.02
0.0075
0.025
-0.1
-0.1
-0 .0 0 2 -0 .0 0 5 5
0.(K )275.s -0. 1

-0 .( )() 5 5
; (b )
= 2 . 3 0 5 c o s ( l 0 0 ^ - 1 2 . 5 3 " ) A.
1026
(c)
(d)
If K is connected in parallel with port

2 of N, then
=^22+^
^21,new = ^22 + ^ ^ther y- and hparameters remain the same.
If Y is connected in parallel with port
1 of N, then;
+ K Other
^parameters remain the same.
59. A common-ground two-port N has known

is
formed by adding one single impedance Z or
admittance Kto N in two different ways. Prove
the following relationships between the old and
new -two-port parameters.
(a) If Kis conneaed across the top termi
nals of N, then
= jVn + Y,
61. Compute the^-parameters of the two-port

of Figure Pi 8.61.
z- and ^parameters. A new two-port
Figure PI8.61
62. Compute the j/-parameters of the two-port
of Figure PI 8.62.
yil,new " yil'* ^ y\2 ,new = >^12

^ ^ y 2 h n c w = y 2 l-^ '
(b)
If Z is connected in series with the

common terminal of N, then
=
2 ] 1 + Z, Z22,new ~ ^ 2 '^ ^ ^12,t
12
r )
60. Assume standard labeling in Figure P I8.60,

in which the inner two-port labeled N has parameter matrix
(a)
(b)
63. Assume standard port labeling for the cir

cuits in Figure P I8.63. Find the z-parameters
of each of these circuits. For the interconnec
tions shown in Figures PI8.63c and d, when do
T4 5'
Q.
the overall ^-parameters equal the sum of the
3 4
individual two-port parameters? When this is
Find the /-parameters of the overall not the case, explain why not.
two-port N* and then find the z- o-^ys/ V
o o------------ 1------------o
1o
1n
parameter matrix.
Now suppose the ports are connected
1o
to current sources, in which case /j(/)
= 12(f) = I5u(t) A. Assuming zero ini
(a)
(b)
tial conditions, find
and
o o
Is/X/vo
What is the response if ij(t) =
= o \/\y\/I
10
)
10
5u(t) A? Hint: Use linearity.
2F
10
port 1 port 2
1O
------- -------
10
On / \ / V
(0
Figure P I8.60
<
^
10
(d)
Figure PI 8.63
r ^
102'
64. Compute any one set o f the z-, y-, and iparameters for each of the circuits o f Figure
P I 8.64. Then obtain the remaining two sets by
the use of the conversion table (Table 18.1).
yYY\
160
1 H Ij
0.5 F
'
67. Find the z-parameters of the interconnect

ed circuit shown in Figure P i 8.67, in which L
= 1 H , assuming standard units for the yparameters.
1#
m \
Figure P I 8.64
Y=
"4
-1
.-2
65. For the circuit o f Figure P 18.65, find the zparameters o f each rvvo-port,
and N^, and
2-parameters
of the intercon
then the overall
Figure P I 8.67
nected two-port.
AN SW ER:
\-ij
v-l-0.5
5 4-0.5
.v+l.O
.v-h2.()
Q.
68. Repeat Problem 67 with the inductor

changed to a 0.5 F capacitor.
69. Find the z- and ^'-parameters of the inter
connected network shown in Figure P I 8.69, in
which ;?, = 2 n ,
Z =
Figure P I 8.65
66. For the circuit o f Figure P I 8.66, find the^parameters o f each two-port,
and N^, and
2-parameters
o f the intercon
then the overall

nected two-port.
= 2 iX
-2
= 1 iX and
a.
102.S
Figure P I 8.69
A N SW ER:
[3.2
!' //Im'ii- ~ (^)
l.S
Q.
- 0 .8
72. Two rwo-ports
70. The rwo-port
is connected to a 1:1 ideal
transformer to form N a *. Then N^*

a and N 0/ are
connected in series to form a new rwo-port N
and N^âre connected in
series to form a new two-port N , as shown in

Figure P I 8.72. Find the new z-parameters.
as shown in Figure P I 8.70, in which /?j = 2
/?2 = 2 Q , /?3 = 3
and R^ = 1Q .. Find the z-
parameters of N .
Figure P I 8.72
73. Consider the two-port in Figure P i 8.73,
which depicts a transistor amplifier stage for a
microphone. Suppose that the transistor has h71. Two two-ports
and
are connected in
series to form a new two-port N , as shown in

Figure P I 8.71, in which = 2, /?j = 4 Q, R j =
4 i i , /?3 = 2 n ,
^7 = 1 Q, and
= 8 n . /?5 = 2 n ,
parameters given as Aj j = 4.2
^h\
= 150, and h ^2 = 0-1 mS.

(a)
Assume the input and output capaci
=2h.
tors are short circuits at the frequency
= 4 Q, Find the new z-param-
of interest and that the capacitor
eters.
across the 470 Q resistor is also a short

circuit. Determine the /;-parametcrs of
the overall two-port.
(b)
Given your answer to part (a), deter

mine the value of n for maximum
power transfer.
(c)
Given your answer to part (a), deter-
1029
(cl)
mine rhe voltage and power gain of
A N SW ER:
rlie overall c\vo-port.
30.15) V
(h)
/'| ^^{[) = 2.505 a)s(5()0r -
Plot the frequency response o f the
ampliher as a function ot / = 271(0 for

^
^
, ,,
. ^ .
0 < /< 20,000 kHz using Spice.
^
which u = 0.2, /?, = 1 Q, and R . = 4 Q.

/\ c
l
i
c
l
(a)
For what value or
is the rwo-port
r r
..
,
.
r tn,o-rc
Consider the circuit or Figure P I S . 75, m
jn
reciprocal?
(b)
For the value of
found in part (a),
compute tlie circuits z- and />parameters.
Figure P i 8.73
(c)
If port 2 is terminated in a 2 Q-1 mF

parallel
RC
combination,
and
RECIPROCITY
\l(/) = 20>/2 cos(20000 V ,
74. The two-port in Figure P I 8.74 has port 1
v-,{t) in steady state. Use the value of
attached to an ideal current source. It is known
find
found in part (b).
that the two-port consists only o f R's, L's, Cs,

and transformers. If /|(r) = 5(^), your oscillo
scope shows a waveform having the analytic
expression V2U) =
cos{t - 30) V for
f > 0. Figure PI8.74b shows the same t\vo-port

with different terminations.
(a)
If
= u{t) A, find the zero-state
response,
(b)
If the input is changed so that ;-,(/) =

10 cos(500/)//(f) A, find the steadystate response, i^j(f).
Figure P I 8.75
A N S W FR; (a) r
i,(t)
vjt)
(a)
v,W
7.5
5.0
5.0
4.0
2-Port
v/t)
2-Port
(b)
Figure P i 8.74
(b)
= 1 Lh
1.25
;c' r,: r) = 10 co.s(200(); - n ) \.
-1 .2 5
1.25
0.25
1030
76. Consider the circuit of Fieure PI 8.76, in

which
R^ = I Q, L - y/l H and C = ^^2 F.
The transfer function
H (s) =
njs) ^ ____
d(s)
+ ypis + l
Use all the theorems you have learned in this

course (including reciprocity) to construa an
alternative circuit that has the same d{s).
r>
yyy\
R.
Figure P I8.76
77. Again consider the circuit o f Figure PI 8.76,
in which 7?^ = 1 Q,
= 0.5 2, Z = 1.673 H
and C = 0 .8 9 6 6 F, or Z = 0.4 4 8 3 H and C =

3 .3 4 6 6 F. The transfer function is
în (^)
5^ + >/2 J + 1
Use the theorems you learned in this course

(source transformations, reciprocity, scaling) to
construct two new circuits in which
= 1 Q,
= 2 Q, and the new transfer function is

HngJ^^ = KHiJ) for an appropriate K. W hat is
the value o f K that is obtained? The result of
r\
this problem shows how one circuit with a
r s
resistance ratio
r\
R.
can be derived from another with
Rc
a
n
C H EC K : L = 3 .3 4 6 6 H or I = 0.8 9 6 6 H in
series R^ with AT= 2
Principles of Basic Filtering
LOUDSPEAKERS AND CROSSOVER NETWORK

In a stereo system, a power amplifier is connected to a pair of speakers. The most common type of speak
er system consists of one or more drivers enclosed in a wooden box. The amplifier feeds a signal to each
drivers voice coil. The voice coil sits within a field produced by a permanent magnet and is attached to
a heavy paper or plastic cone. When excited by a current from the power amplifier, the coil interacts with
the magnetic field of the permanent magnet and vibrates. The coil pushes and pulls the speakers cone,
which, in turn, proportionately moves the air, making sound waves.
The high and low frequencies that make up music place opposite requirements on a loudspeaker.
A good low-frequency speaker should be large, in order to push a lot of air. A good high-frequency speaker should be light, in order to move back and forth rapidly.
A rwo-way speaker system consists of a small, light tweeter to handle the treble signals and a large
woofer to handle the bass. A better system is a three-way system, with a third, midrange speaker
to handle the f-requencies in the middle. The magnitude response o f a typical rwo-way system is
shown in the following figure.
lÔ ltoa;
U
II
OI
CT
fO 'O
:=
e
E
k.
O C
OJ
1032
Chapter 19 Principles of Basic Filtering
A crossover netiuork (low-pass or high-pass filter) separates the frequencies so that the woofer
receives the low-frequency content of the music and the tweeter the higher-frequency content. In
the magnitude response plot illustrated in the figure, both curves have the same 3 dB frequency
at 2000 Hz. This frequenq is called the crossover frequency. This chapter explores the basic design
principles and realizations o f low-pass and high-pass filters. Some band-pass filtering is also dis
cussed. Some simple crossover circuits are considered in the problems section.
CHAPTER OBJECTIVES
1.
Introduce the meaning and (brickwall) specification of low-pass filters in terms of dB

loss.
2.
Set forth the maximally flat Butterworth magnitude response and associated Butterworth
transfer function.
3.
Present a step-by-step design algorithm for finding the filter order and associated
4.
Present basic passive and active circuits that realize a Butterworth transfer function.
Butterworth transfer function.

5.
Set forth the properties of the Butterworth transfer function.
6.
hitroduce high-pass filter design through frequency transformation.
7.
Present basic passive and active circuits that realize a Butterworth high-pass filter.
8.
Introduce an algorithm for the design of a band-pass filter using frequency transformations.
SECTION HEADINGS
1.
Introduction and Basic Terminology
2.
Low-Pass Filter Basics
3.
4.
5.
6.
7.
8.
Butterworth Solution to the Approximation Problem

Butterworth Passive Realization
Active Realization o f Low-Pass Butterworth Filters
Input Attenuation for Active Circuit Design
Properties of the Butterworth Loss Function
Basic High-Pass Filter Design with Passive Realization
9.
10.
11.
Active Realization o f High-Pass Filters

Band-Pass Filter Design
An Algorithm for Singly Terminated Butterworth Low-Pass Networks
12.
Summary
13.
14.
Terms and Concepts

Problems
1. INTRODUCTION AND BASIC TERMINOLOGY

Types o f Filtering
Filtering plays an important role in circuit theory, as well as in communication theory, image pro
cessing, and control. There are three fundamental types of filters: analog, digital, and switched
capacitor. Digital filter analysis and design requires knowledge of the Nyquist sampling theorem.
Chapter 19 * Principles o f Basic Filtering
1033
Switched capacitor networks/filters, an idea introduced in Chapter 13, are something of a hybrid
between analog and digital filters. Both are beyond the scope o f this chapter, which takes a cir
cuits viewpoint on some basic analog filtering concepts and techniques.
Analog filters process the actual input waveform with circuits composed of discrete components
such as resistors, capacitors, inductors, and op amps. Analog filters are of two types, passive and
active. Passive analog filters are composed only of resistors, capacitors, and inductors. Active ana
log filters consist o f resistors, capacitors, and op amps or other types of active elements.
B a s ic T erm in olog y
A filter is a device (often an electrical circuit) that shapes or modifies the fi^quency content (spearum)
of a signal or waveform. We represent a filter by a transfer ftmction H{s) whose frequency response is
The gain, gain magnitude, or frequency response magnitude is |//(^)|. If
is normal
ized so that its maximum gain is 1, then the gain in dB (decibels) is Gjg{vi) = 20 log]Q|//(/ci))|. In this
chapter we will assume that the maximum value of |//(/o))| has been normalized to 1.
An important frequency called the cutofF frequency (or frequencies), also called the half-power
point (or points), is that frequency, denoted O)^ for which
2
H (j(Oc)
'
= 2
Since power is proportional to voltage squared or current squared (gain squared) for a fixed load
resistance, we arrive at the terminology of half-power. This is also called the 3d B down point
O '
because
= 10 logiol//(/copP = 10 logjo(0.5) = - 3 dB.
Often, design specifications for a filter are expressed in terms of attenuation or loss rather than
gain. This results in several definitions that are dual to the gain-related definitions above. The loss
fiinction, denoted H (/w), is the reciprocal o f the transfer ftmction. Thus, the attenuation, or loss
magnitude, is
H (jw )
'
HiJco)
It follows that the filter loss or attenuation in dB is
/o55 = -201ogio |//(yw)| = 201ogio
(19.1)
With these definitions established, the chapter will examine low-pass (LP), high-pass (HP), andband pass (BP) filters, leaving the study o f band-reject filters to higher-level texts. A low-pass fil
ter is a device (typically a circuit) offering very little attenuation to the low-frequency content
(low-frequency spectrum) of signals while significandy attenuating (blocking) the high-frequency
content o f those signals. High-pass filters do the opposite: they block the low-frequency content
and allow the high-frequency content of a signal to pass through. Finally, band-pass filters, as
described in Chapter 16, allow a band o f frequencies to pass while significantly attenuating those
outside the band. Interestingly, the general practical design of all such filters, especially for the
high-order case, is generally based on a low-pass prototype design; the low-pass prototype is trans
formed into a high-pass or band-pass type using a frequency transformation.
H )34
Chapter 19 Principles o f Basic Filtering
As a firsr step in exploring the design o f such filters, vve describe Butterworth (low-pass) transfer
functions. Butterworth was the name o f the English engineer who first developed this special class
o f transfer functions in his paper O n the Theory of Filter Amplifiers. ' The next step is presen
tation of an algorithm that adapts these basic Butterworth transfer functions into ones that meet
a given set of L P filter design specifications. Once the transfer function is known, an engineer
must implement this transfer function as a passive or active circuit. This chapter will also outline
some techniques for generating passive and active reaHzations.
2. LOW-PASS FILTER BASICS

As mentioned, a low-pass filter allows the low-frequency content o f a signal to pass with little
attenuation while significantly blocking the high-frequency content. The immediate question is,
why use a low-pass filter? One possible answer is that the noise in a noisy signal often has most of
its energy in the high-frequency range. For example, so-called white noise has a constant frequenq- spectrum. Hence, low-pass filtering a sinusoidal signal corrupted by white noise will generally
result in a cleaned-up information signal, as illustrated in Figure 19.1. In Figure 19.1, the thickness o f the curves represents the infiltration of noise. By using the low-pass filter, the curve is
sharpened by reducing the high-frequency noise content.
The suspension system o f a car is a low-pass filter: slow, rhythmic (low-frequency) road variations
are permitted, while the effects o f chuckholes and bumps (high frequencies) are filtered out.
Lowpass Filter
F IG U R E 19.1 The effect of low-pass filtering on a noisy signal.

Whenever LP filters are needed, an engineer must provide design specifications. Historically this
was done using a low-pass filter (brickwaJl) specification, as illustrated in Figure 19.2. Two pairs
of numbers (d)^,,
1.
2.
and (to^,
characterize this brickwall specification, where
equals the pass-band edge frequency,

0 < 0) <
3.
is called the pass-band,
niaximum dB attenuation permitted in the pass-band,
4.
to^ is the stop band edge frequency.
5.
0)^
<
6.
7.
CD d e f i n e s t h e s t o p b a n d ,
is the minimum allowable dB attenuation in the stop band, and

co^ <
(0 <
defines the transition band.
103s
A{(d) = Loss (db)

A
F IG U R E
19.2 Brickwall specification of low-pass filter.
The shaded region in Figure 19.2 represents a brickwall. The attenuation of the filter in dB, i.e.,
/l(co) =-20 log,Q|//(/w)| = 20 log,o|/^(/w)|, must reside outside the shaded region. Finding a (nor
malized) filter transfer function that meets the brickwall specs is callcd the approximation prob
lem. Once the approximation problem is solved by identification o f the proper normalized trans
fer function, the next step is to construct a circuit realization of the normalized transfer function
and then to frequency-scale to obtain the proper pass-band edge frequency, and finally to magnitude-scale to obtain the proper impedance levels.
The simplest technique for solving the approximation problem is with 3dB
normalized
Butterworth transfer functions whose squared magnitude responses arc given by
H ( jc o )
( (ti )
(19.2)
1+
U c/
where
(tip
is the 3 dB down point, or cutoff frequency, o f the filter, and e is a to-be
specified constant. As mentioned earlier, the term 3 dB down point arises here because, for all w,
10
lO g lQ
1+
( coY
= 10 1og,ol2j = 3 dB
i.e., there is 3 dB of loss at (O = co^ W hen e = 1 in equation 19.2, the pass-band edge fre
quency
and cutoff frequency
coincide. If we now define the normalizedfrequency as Q =
CO,.
the magnitude response of equation 19.2 becomes
^3dBNLP ( J ^ ) ~ =
(0
1 + Q^
( 1 9 .3 )
1036
Equation 19.3 denotes the wtli-order 3 dB normalized Butterworth magnitude response. The words
3 dB normalized refer to the hict that at Q = 1, the loss is 3 dB (the gain is -3 dB). Remember
that the actual filter transfer function depends on a proper choice of
or f , which we will clar
ify shortly.
Before proceeding further, we ask a critical question: Does this kind o f representation make sensed To
answer this question, note that the dB loss of equation 19.3 is
A{Q) = lossidB) =
2 0
logiQ H^dBNLpU^) =
1 0
logSlO 1 + Q^
( 1 9 .4 )
Plotting this function for various ;/s as a function of normali/xd frequency, Q, for a normalized
stop band edge frequency,
= 3.5,
= 3 dB, and
= 20 dB yields the polynomial curves
in Figure 19.3. Clearly, for h i g h e r w e can have higher values o f
These cur\'es can be made
to lie outside the brickwalls o f Figure 19.2 and hence can validly be used to meet a L P filter spec
ification.
-XD
Normalized Frequency, Q
FIGURE 19.3 Plot of the normalized Butterworth magnitude responses for n = 2, 3,
1()3
3. BUTTERWORTH SOLUTION TO THE APPROXIMATION

PROBLEM
From Figure 19.3, vve see that as n increases, the pass-band magnitude response becomes flatter
and the transition to stop band becomes steeper. This property suggests that there is an appropri
ate value of w to meet a given set of brick\vall specifications (ci3^,
and (o)^,
The idea
is to use equation 19.2 and these specs to determine a proper value for the filter order , a prop
er cutoff frequency (o^., and the normalized transfer function.
Computation o f Filter Order n an d CO^

To compute the filter order, observe that n must satisfy the pass-band and stop band edge fre
quency constraints. From equations 19.1 and 19.4, at the stop band edge frequency,
1 -I-
(^ s\
After some algebra, we obtain

(19.5a)
2: 1 0 ^ - 1
Similarly, at the pass-band edge frequency.
Again, after some algebra we obtain
/'" p V
(19.5b)
Dividing the left and right sides of equation 19.5a by the left and right sides o f equation 19.5b
(this maintains the inequalit)' o f equation 19.5a) yields
(w X
(19.6)
Solving for n produces the order form u la
gio
(19.7)
na
( COs^
lo g lO
(Or, J
\ ^ P
Thus n can be any integer satisfying inequalit}' 19.7. Usually one takes the smallest such n.
103S
Once n is picked, there is a permissible range of co^ given by
(M,
LOcjnm
CO.
2 ,^ A A min
= COc,max
(19.8)
__ j
To derive the range given in equation 19.8, we reconsider equations 19.5. Taking the 2th root of
both sides of equation 19.5a and solving for
yields the right side of equation 19.8. On the other
hand, taking the 2th root of equation 19.5b and solving for to^ yields the left side of equation 19.8.
E xercise. Verify the mathematical details in the above paragraph on the derivation of equation
19.8.
EXA M PLE 19.1. Suppose we are given the brickwall specs
= (500 Hz,2 d B ) and
- (2000 Hz,3 dB). Find (a) the minimum filter order , (b)
and co^
(c) and
the normalized L P squared magnitude function, and then (d) plot the magnitude responses for
the cases where O)^ =
and a)^ =
over the frequency range 0 :s
I n x 2500.
Step 1. Find the minimum filter order n from equation 19.7. Here
/
1 0 ^ ^ - _1 '
iogio
_ I
(
log 10
l O - '- l ^
10 - 1
- = 2.68
ogio(4)
1
ogio
\
P/
implying that the minimum filter order is 3. This can also be accomplished in M A T L A B as follows:
n=buttord(wp,ws,Amax,Amin,V)
n =3
Step 2. Using equation 19.8 with w = 3,

I n X 500
CD,
COc \ m
= 2;r X 547 = 3435 rad/s
2 ^ 1 A "max _ I
and
2jt X 2000
CO,
COc , m
ax
min _
= 2;r X 633 = 3975 rad/s
^ 1 0 ^ -1
Step 3. Since the order is 3, we have from equation 19.2

^ (> )
1
1+
(O
co^
1039
Step 4. Plotting this function over the frequenc)' range 0 ^ 2jvf^ 2 ttx 2500 can be achieved with
the following M A T L A B code. The resulting plot is displayed in Figure 19.4.
f= 0:4:2500;
hl = sqrt(l. ./(I + (2*pi*f ./wcmin).^6));
h2 = sqrt(l. ./(I + (2*pi*f ./wcmiix),^6));
plot(f,-20*logl 0(h 1),f,-20*logl 0(h2))
grid
Frequency in Hz
F IG U R E 19.4. Plot of third-order LP filter response using
Observe that with the choice o f
= (jo^
the magnitude response curve passes through the
pass-band edge frequency with loss exactly equal to

O n the other hand, with the choice of co^ = co^
leaving more than adequate loss at (O^.

the magnitude response curve passes through
the stop band edge frequency with loss exactly equal to

loss at
and
with less than the maximal allowable
. In practice, to allow for element tolerances, one would choose O)^ somewhere in
between.
Exercises. 1. Verify that if co^ =

2. Verify that if co^ =
the loss
and ^(co^) >
and A { lo^) s
1 04
-7.071 le-Ol + 7.071 le-Oli

-7.071 le-01 -7.071 le-Oli
7.071 le-Ol + 7.071 le-Oli
7.071 le-Ol -7.071 le-Oli
Note that rvvo zeros are in the right lialf complex plane and two in the left. The two in the left
half plane determine the Butterworth loss function. For illustration we provide the zero plot of
Figure 19.5. Again, note the symmetry o f the zeros with respect to the imaginary axis.
1 .......... i.............. i ..............
.......................
;
0.8
I /'
i/
0.6
0.4
0.2
I**
-0.2
-0.4
-0.6
-0.8
i \
:
i
/
X
-1
-1
-0.5
1
0.5
1
1
Real part
F IG U R E 19.5 Plot of the zeros of H (s) H {-s) = 1 +
As per our earlier statement, from stability considerations we pick the left half plane zeros to form
i-e.,
Hhat = poly(zeros(l :2))
Hhat =
l.OOOOe+00 1.4l42e+00 l.OOOOe+00
W e conclude that
r + y[2s+\
For higher-order cases, the procedure of Example 9.3 can again be used. The results are given in
Table 19.1, which presents the 3 dB normalized Buttenvorth loss functions. In practice, one never
computes the actual transfer function o f a filter that meets a set of non-normalized specs. Rather,
one realizes the normalized loss or transfer function and then magnitude- and frequency-scales to
obtain the proper circuit. This process is illustrated in the next section.
1042
C h ap te r 19 Principles o f Basic Filtering
T A B L E 1 9 .1 N orm alized Bu tterw o rth

Lo s s F u n c t i o n s , = 1,
5.
^3 d B N L p (^)
S+ 1
S~ + y / l s + 1
+ 2 r + 2s + 1
{s + l ) ( j ^ + s + \ ) =
{ r + 0 . 7 6 5 3 7 s + 1)(^2 + 1 . 8 4 7 7 6 i + 1 )
= / + 2 .6 1 3 1 P + 3 .4 1 4 2 j2 ^ 2 . 6 1 3 b + 1
{s +
5
= P
l)(j^
+ 0 . 6 1 8 0 3 ^ + l ) ( j ^ + 1 . 6 1 8 0 3 j + I)
+ 3 .2 3 6 1 /
+ 5 .2 3 6 1 P
+ 5.2 3 6 1 ^ 2 + 3 . 2 3 6 b
+ 1
4. BUTTERWORTH PASSIVE REALIZATION

W e begin this section with an example of a third-order passive realization.
E X A M P L E 19.4
The circuit of Figure 19.6 must realize a Butter\vorth filter meeting the low-pass brickwall speci
fication of Figure 19.7. In the final design we require R^= 100 Q.
R.
LC
out
Network
-o
F IG U R K 19.6 Structure of a filter driven by a practical source.
40 db
2db
> f(Hz)
100 Hz
1000 Hz
F I G U R E 19.7 Low-pass brickw all specification o f Exam ple 19.4.
So
1043
lu t io n
From the brickwall specification of Figure 19. 7 , we have (oj^ = 2 0 0 k ,

20007T,
= 2 d B) and (o)^ =
= 40 dB).
Step 1. Compute the filter order. Here

_ 1
loglO
1 0 - _ 1
2 0 0 0 .T
o g lO
- =
2.1164
200 jt
This implies that the minimum filter order is n = 3.

Step 2. Compute
Recall equation 19.8;
COn
CO.
By historical convention, we will use
0j^=(0 c.min
in which case
(O,
2 0 0 JZ
= 687 = 21 SJjT rad/s ec
Step 3. Looking up the third-order normalized Butterworth loss fiinction and inverting to obtain
the transfer function yields
H MBNLP
(19.12)
Step 4. Choose a candidate passive circuit and obtain its transfer function,
W e wind up with the circuit o f Figure 19.8. Note that we will adjust the
is)
. Vouti^)
Vin(s)
source resistance later with magnitude scaling.
/VY\
in
V.
F IG U R E 19.8 Third-order filter circuit for Example 19.4.

W e obtain the transfer function using nodal analysis. At the top node o f C j,
Vc - - ! - K u, =
Ls
A t the o u tp u t node.
v,
(19.13a)
1044
--V /C +
Lv
C-iS H-----
(1 9 .1 3 b )
Ls
Solving equation 19.13b for Kq and eliminating K q from equation 19.13a yields
" - ( 'J - i r M
'//il-V
+-
C|
Step 4. Equate the coefficieuts o f
C 1 + C 2------------ i ~
------LC jC t
LC^C2
s ~ + -^ -------^ s +
Equating the
an d
coefficients in equations 19.14 and 19.12 yields
.
1
,
. Cl +
.
= 2 , ------ = 1, and ---- = 2
C|
LC^C')
LC\C~)
2
Clearly, Cj = 0.5 F and
LCi
= 1 implies LC^ = 2. Then
C 1+ C 2
0 .5 + C 9
LC ,C 2
1c
1r
r-
---- = 2 => C ? = 1.5 F and L = H
In summary, C, = 0.5 F,
^11
3
= 1.5 F, and L = 4/3 FI.
Step 5. Frequency-scale the circuit to obtain the desired, a n d magnitude-scale to obtain the specified
source resistance o/lOO Q. W e frequency-scale by Kj-= co^ =

scale by
= 100. Thus,
Q
= 687 = 218.7TT and magnitude-
= 100 /?^ = 100 Q
..........
^ ^
= 7.2773 HF.
C.
______
^
.
K ,L
= 2 1.832 ^F, and L ,,,, = ^ = 0.19406 H.
This completes the design.
E xercise. Reconsider Example 19.4. I f
is to be 50 Q and O)^ =
recompute the final ele
ment values.
.\ N S W F R S : 7.38~2 iiK 22.162 uF. and 0.0492 H
In the above example, there is no load attached to the filter. A situation in which there is no source
resistance but a load resistance is given in the problems at the end of the chapter. In addition, there
is a problem containing both a source and a load resistance.
Here we have used a coefficient matching technique that is manageable for orders 1, 2, and 3. For
higher orders the method is unwieldy and we must use the methods o f nerwork synthesis studied
in other courses or resort to normalized filter tabulations given in filter handbooks.
1045
5. ACTIVE REALIZATION OF LOW-PASS BUTTERWORTH

FILTERS
For active realization of Buttervvorth filters we use the factored form of the transfer functions given
in Table 19.1. For example, we would represent the third-order normalized Butterworth transfer
function as the product
K.
H?iclBNLp{^) = --
S~ + .V + 1
where
= 1. The gains
= H{s)Hi,{s)
(19.15)
i + 1
and /iQ, are present in equation 19.15 to allow the individual stages
to have dc gains different from 1, as is sometimes necessary. The overall transfer function is real
ized as a cascade of
= -------- a n d H ij { s ) =
s~ + .s -I" 1
with the constraint
.v+1
= 1. All transfer functions beyond first order require one or more sec-
ond-order stages in cascade for their realization. The use of a basic second-order active circuit is
key to the realization of any of these second-order sections.
The Sallen and Key Second-Order Section

One of the common single-op amp second-order circuits is the Sallen and Key (S & K ) configura
tion illustrated in Figure 19.9. As part of the realization process we must first compute its secondorder transfer function. Denote the circuit transfer function by
F IG U llE 19.9 Sallen and Key low-pass second-order section.

To derive
note that the properties of an ideal op amp force the voltage across 7?^ to be
Using voltage division,
Vo, u t
o r equivalently,
1046
This observation permits us to simplify the diagram of Figure 19.9 to the equivalent circuit of
Figure 19.10.
F IG U R E 19.10 Equivalent circuit of the Sallen and Key low-pass

filter of Figure 19.9;
identifies node b and
identifies node a.
Construction o f H { s ) now proceeds by nodal analysis. At node a the node between C , and
Ri,
1
( 1
(19.16)
Similarly at node b,
(19.17)
Eliminating
and solving for the ratio
( 5)
^
yields the transfer function of the Sallen and
Key low-pass filter:
K
R\R2C\C2
K ,A ^ )
ViU)
S + 1 ------------ h
]-K \
H-------
-I
R 2 C 2 )
U ^ lC ,
(19.18)
H-----------------------R\Ri C iC 2
w h e r e / r = l / , , / l / = 1 + /? ^ //? ^
Butterworth LP Design Algorithm Based on Sallefi and Key Circuit

For practical reasons, design engineers have developed S & K circuit design templates based only
on the circuit Q and dc gain K for a transfer function of the form
H [s) =
K
s^ + s+ \
(19.19)
The design template derives from equating the denominator coefficients of equations 19.18 and
19. 19. The dc gain can be adjusted using other methods such as input attenuation (to be described
later). In particular, equating the denominator coefficients o f equations 19.18 and 19.19 requires
that
104'
1- K
/?,C,
/?2Ci
R 2C 2
(19.20a)
and
(19.20b)
R\R2C\C2
The solutions to equations 19.20a and 19.20b are not unique because there are t%vo equations in
five unknowns. This means that we can impose up to three additional constraints to produce dif
ferent solutions. Different solutions produce the proper filtering action, but have different behav
iors in terms o f the sensitivit)' of the frequenq' response to variations nominal resistance and
capacitor values. Also, different designs have different ratios of clement values that may have prac
tical significance.
One popular and robust design is the Saraga design. For the (normali/ed) Saraga design, the three addi-
R.
R]
v3
tional constraints imposed on the solution of equations 19.20 are C, = 1, Cj = >/3Q, and =
which were chosen to minimize certain sensitivities in the circuit performance. Solving equations
19.20a
and
19.20b
for
R ,,
C 2 = IC \ = S q . R] = Q~\ R2 =
/?2,
and
I/V ,
using
these
additional
K = 4 / 3 , Rs = R^/3.
constraints
yields
The final circuit realization of
equation 1 9 . 1 9 is given in Figure 1 9 . 1 1 . Again we note that the dc gain K = 413 will need to be
adjusted by other means, to be described. Frequency and magnitude scaling are necessary to
achieve proper cutoff frequencies and impedance levels.
F IG U R F. 19.11 Normalized Saraga design of Sallen and Key circuit for realizing the
transfer function of equation 19.19.
W ith this design in hand, we now redo Example 19.4 using an active realization.
1048
E X A M P L E 19.5. This example presents an active realization of a third-order Butterworth filter

with 0)^ = 687 = 218.7;r rad/s ec. Recall from equation 19.15 that we factor the third-order
Butterworth transfer function as a product:
^7,clBNLp{^) = -------------- ^ ------- - H^,{s )H ij{ s )

S~ + .V -I- I
where
5 + 1
= 1. Later, we will frequency- and magnitude-scale to obtain proper cutoffs and
impedance levels. For
S~ + .S + I
S~ +
.V +
we have Q = 1. The Saraga design element values are C-, = 1 F, Cj = >/3 F, /?, = I Q, 7?2
Q., K = K j = 4/3, and
These element values realize the transfer function
.V + .V+ 1
To obtain the correct
realize the transfer function
K,, _ 3/4
s -t-1
(19.21)
.V+
Observe that the choice o f /Q, makes A'/Q, = U the correct overall dc gain. The transfer function
of equation 19.21 is realized with the leaky integrator circuit o f Figure 19.12, whose transfer func
tion
set equal to H(^{s) is
Hl,âr(s) =
3/4
R ,C
.y-l-
5+1
R'yC
To complete the normalized design, choose C = \ V, Rj - \
and /^| = 4/3 Q.
R,
F I G U R E 19.12. Leaky integrator circuit.
1049
Combining the second-order and first-order sections o f the normalized design produces the cir
cuit of Figure 19.13.
F IG U R E 19.13. Realization of third-order 3 dB N L P Butterworth transfer function.

In the last step o f our design we frequenc}'-scale by /y = 218.7;r. If the smallest capacitor is to be
100 nF, then
K,n =
c o ld
K jC ,,,,,
= 14.555 x 1 0 ^
2 I 8 . 7 ; r l O ^
Hence the final parameter values for the second-order section are
Q = 100 tiF, q = 173.2 uF,
= 14.56 kQ, Rj = 8.4 kQ, R^ = 30 kQ, ami Rf^ = 10 K2
Similarly, for the first-order section C = 100 nF,
/ ?2
= 14.55
and R^ = 19.4 kQ. This leads to
the final circuit design given in Figure 19.14.

173.2nF
14 .5 5 kO
1050
This design assumes the fiirer is driven by a volrage source with a very small resistance; otherwise
the source impedance must be considered part o f the front end series resistance. Also, we could
have magnitude-scaled each section separately (but chose not to).
Other Sallen and Key Designs

in addition to the Saraga design, there are two other S & K normalized designs that we will set forth
here. The first is termed Design A: /?^ = oo,
To obtain design A we normalize K =
1 and
= 0, /T = 1, /^, = /?2 = 1
= /?2 = 1
Cj = 2Q, and C j = ---
Using these three constraints
and equating the coefficients o f equations 19.19 and 19.18 leads to
K
s~ +
.V
C\C2
+1
+ --
C,
.V +
(19.22)
--------
C ,C 2
Matching the denominator coefficients produces Cj = 2Q and C j = 1/2Q. Since the dc gain with
K = 1 is 1, there is no need to modify the overall circuit gain. Figure 19.15 illustrates design A.
C, = 2Q
F IG U R E 19.15 Design A for realizing the transfer function of equation 19.19.

Notice, however, that for Design A the ratio C^/C, = 4Q^ is 400 for a circuit with Q = 10. Such
a large variation may be undesirable in a practical circuit.
2, C, = 1 F, C 2 = 1/Q F,
A third design we term Design B: /?^ = /?^ = 1 Q,

= Q L2. This design is computed using the constraints
and equal time constant,
and /^2 -
Q- Thus,
/?jCj =
= ^2^2'
= I Q (yielding /f= 2), C, =
these choices,--------- =
1 = RjC2 implies that C ,
= 1 Q, and
1implies that /?|
1 F,
1
= . Here the maximum parameter ratio is
= C j/ G ) = Q, w h ic h is better than the 4Q^ ratio of D esign A . Figure 19.16 shows the result
in g circu it.
lO r
C =1
F IG U R E 19.16 Design B for realizing the normalized transfer function of equation 19.19,
assuming standard units.
6. INPUT ATTENUATION FOR ACTIVE CIRCUIT DESIGN

In design A of the previous section, the gain o f the filter is A"= 1. In design B, the filter gain is K
= 2. In Example 19.5, the gain of the Saraga design {K = 4/3) was corrected by the subsequent
first-order section. I f the filter were second order with a dc gain of 1, then the circuit gain would
need to be modified for both Design B and the Saraga design. In order to reduce a high gain,
maintain the filtering properties, and keep as many parameter values as possible at their original
design values, we use a technique known as input attenuation. In this technique, the front-end
resistor is replaced with a voltage divider circuit, as illustrated in Figure 19.17.
(b)
F IG U R E 19.17 Illustration of input attenuation concept, (a) Original circuit,
(b) Circuit with input attenuation.
10 ^ 2
Figure 19.17a represents the original active network, while Figure 19.17b represents the modified
network. For the filtering characteristics to remain invariant, the impedance looking into the
dashed boxes must remain at /?, for both circuits. Thus, the parallel combination of
and
must equal /?j, i.e.,

'
(19.23)
In addition, if the new gain is to be a K , a < 1, instead o f K, then
a =
^4
/?3 + /?4
(19.24)
Solving equations 19.23 and 19.24 for R^ and R^ yields
=^
a
and
(19.25)
\ -a
Thus, one can reduce the gain o f the Sallen and Key low-pass circuit via the simple technique of
input attenuation.
7. PROPERTIES OF THE BUTTERWORTH LOSS FUNCTION

As mentioned earlier, the Butterworth loss function has a maximally flat pass-band response in the
sense that as many derivatives as possible o f the loss magnitude response are zero at Q = 0. This is
often termed m axim ally flat. In the case o f the ;zth-order Butterworth loss function, it is possible
to show that
H ( j Q ) = 0,
for A: = 1,2,...,
cIQ'^
2/2-1
(19.26)
at Q = 0 and
cf
dQ^
H (jQ )\ = 0 .5 { 2 n ) \ * 0
( , 9.27)
at Q = 0. This is consistent with the notion of being maximally flat. To verify equations 19.26 and
19.27, observe that if
1, then
Equations 19.26 and 19.27 follow after differentiation of equation 19.28 2u times and evaluation
of each derivative at Q = 0.
Several closing remarks are now in order. First, the cutoff frequency is the half-power point, or the
3 dB down point. The terminology follows because |MyO)| = 1 (the gain at dc is unity) and
10^ 3
H (j\ ) = - ^ = 0.707,
V2
yielding half power for normalized Butterworth gain functions H{s). Second, for large Q,
Hence as Q becomes large, the gain G^g{Q) decreases according to
-n
[20 logjQ(Q)]
indicating that the gain rolls off with a slope proportional to the number of poles, n, o f the gain
function. Specifically, the slope equals -20 dB per decade. These statements are also valid if to
replaces Q.
8. BASIC HIGH-PASS FILTER DESIGN WITH PASSIVE

REALIZATION
High-pass filters invert the frequency characteristics of low-pass filters. A high-pass filter is a
device usually a circuit that significantly attenuates the low-frequency content o f a signal while
passing the high-frequency content with minimal attenuation. Figure 19.18 illustrates typical
brickwall specifications for a high-pass filter. Here frequencies above the pass-band edge frequen-
cy co^ have little attenuation, while frequencies below
are significantly attenuated precisely
the inverse function o f a low-pass filter. In fact, low-pass and high-pass specifications are related
by a simple inversion of O). In particular, we define a H P (high-pass) to L P (low-pass) frequen
cy transformation as
^ =
(19.29)
This frequency transformation applied to the briclc\vall specifications o f Figure 19.18 yields the
set of normalized low-pass specifications given by Figure 19.19.
10 V i
A(co)
min
3dB
A_..> (0
0
CO.
F IG U R E 19.18 Typical brickwall characteristic of a high-pass filter.
A(co)
F IG U R E 19.19 Normalized low-pass equivalent specifications derived from high-pass specifications.

Given the normalized L P (N L P ) equivalent specs, one finds the Buttenvorth filter order accord
ing to the usual formula, equation 19. 7, reproduced as equation 19.30:
1''VniH _ I
log
clO
10
o g lO
_
lo2 10
(19.30)
( Q ,\
By convention we take the filter order to be the smallest integer n satisfying equation 19.30. For
a passive realization, once n is known, one would realize the resulting N L P Butterworth transfer
I05S
function. For example, suppose = 2; the second-order 3 dB N L P Buttervvorth transfer function

could be realized by (among others) the circuit of Figure 19.20, whose transfer function is
J_
J_
Cs
Ls + /?< +
_
1
Cs
LC
S
+ S+ -
LC
F IG U R F 19.20 Circuit to realize second-order N L P Butterworth filter.

As with the development o f equation 19.8, once n is picked, there is a permissible range of nor
malized
given by
Q.
Qc ju i n
where
(19.31)
Once
is chosen, the Butterworth transfer function that meets the N L P
equivalent specs is found as
^NLP i^) = ^3dBNLP | ^ j

This is, of course, equivalent to frequency scaling a realization of the Butterworth filter by Ay =
Q^, as illustrated in Figure 19.2L
/YY\
u o .C
V,
C/Q,
F IG U R E 19.21 Circuit that realizes

The actual H P transfer function is related to the other normalized transfer functions as follows:
10
H H p{s) = H N Lpi \ = H
This is equivalent to doing a special frequenq' transformation on the circuit elements o f Figure
19.21. Specifically
1056
(O.
(19.32)
is called the L P to H P frequency transformation and has a natural interpretation in terms of
the inductors and capacitors of a passive network. In particular, a capacitor C with impedance
Cs
becomes an inductor o f value
, I.e.,
C(0:
Cs
(0),
C oj,
A similar substitution shows that an Inductor L with impedance Ls becomes a capacitor of value
Figure 19.22 illustrates the transformation.
/YY\
-O
1
L=
Cco
o- fY Y \
-o
-o
1
c = Leo.
F IG U R E 19.22 Circuit element change under the LP to H P transformation.
Hence the L P circuit of Figure 19.21 becomes the H P circuit o f Figure 19.23.
Cnew=
L CO
F IG U R H 19.23. H P circuit derived from the N L P reali/aiion of L'igure 19.21. Observe that
to.
It remains only to magnitude-scale this circuit to obtain an acceptable passive H P filter. W e illus
trate the above ideas in the following example.
lOS"
E X A M P L E 19.6. Design a minimum-order Buttenvorth filter meeting the high-pass brickwall

specifications of Figure 19.24, in which the loss magnitude curve is to pass through the pass-band
corner.
>0)
2000n
200n
F IG U R E 19.24 High-pass filter specifications tor Example 19.6.
Solution
9000 t
Step 1. Compute the equivalent N LP specificatiom . Using
Q =
the H P to N L P trans
formation, the N L P specifications are easily computed as
(H/. ' '
= 2 dB) and (Q^ = 10, A,. =40 dB)
Step 2. Detennine the Butterworth transfer function. As these specifications correspond to those of
Example 19.3, one computes the filter order n = 3. From Tiible 19.1, the third-order Butterworth
s~ -I- .V -I- 1 ](.v + 1)
v-' + 2 .v + 2.V + 1
Step 3. Realize the third-order 3 dB N LP transfer fin ctio n . According to Example 19.4, the 3 dB
N L P Butterworth transfer function can be reali7.ed by the passive RLC circuit o f Figure 19.8, with
values assigned as in Figure 19.25.
/YY\
1n
V,
C^=1.5F
-O
+
V
-y
C =0.5 F
-o
---- c
F I G U R E 19.25 Third-order 3 d B N L P filter circuit o f Example 19.4 having 3 d B down point at 1 rad/sec.
1058
Chapter 19 * Principles of Basic Filtering
Step 4. Compute the norm alized 3 dB down frequency. The
3 dB
frequency
of the actual N L P
equivalent filter is calculated from equation 19.31, using the left inequality:
1/6
Step 5. Frequency-scale by
= 1.0935
to obtain the N L P circuit. The N L P circuit o f Figure 19.25 is fre-
quency-scaled by the factor Kj-= 1.0935 to produce the network o f Figure 19.26.
/YY\
V.
-O
+
1.219H
1o
- J
'' N
0.4573 F
1.3717 F
t
F I G U R t 19.26 Normalized low-pass circuit of Figure 19.25 frequency-scaled by /y-= L0935.

Step
6. Apply the LP to H P circuit element transformation detailed in Figure 19.22. Converting the
circuit of Figure 19.26 to the required H P circuit yields the network of Figure 19.27.
0.1305 mF
-O
+
1Q
V.
V
0.3481 mH
0.116mH
F IG U R E 19.27 High-pass circuit meeting the specifications of Figure 19.24.

Step 7. Impedance-scale. To obtain a 10
source resistance, we impedance-scale by the factor
= 10 to obtain the final realization shown in Figure 19.28.
13.1 mF
-O
+
lOO
V.
1.16 mH
3.481 mH
-o
F IG U R I: 19.28 Passive circuit realizing the high-pass specifications of Figure 19.24 with 10
source impedance.
E x e r c is e . Use SPIC E to confirm the high-pass characteristic o f the circuit o f Figure 19.28.
10^)
The preceding example completes our basic discussion o f passive high-pass filtering.
9. ACTIVE REALIZATION OF HIGH-PASS FILTERS

Active realization o f high-pass filters ordinarily proceeds with the cascading o f first- and secondorder sections, as done earlier. There is a Sallen and Key second-order high-pass section obtained
with an interchange of resistors and capacitors in the S & K low-pass configuration. As with all
active circuits, inductors are not used. W e illustrate this interchange of resistors and capacitors as
an approach to high-pass design with the following example.
E X A M P L E 19.7. Let us reconsider Example 19.6, in which
= (2000;r, 2 dB) and
= (200;r, 40 dB). The filter design proceeded by converting to equivalent N L P specs,

which led to a 3 dB N L P third-order transfer function,
H 3dBNLP {s) =
( i - +l ) |. v
ij
+ 2 a + 2.S + 1
One approach to H P design first realizes this 3 dB N L P transfer function as a cascade o f a secondorder active circuit with a first-order circuitin Example 19.5, this resulted in the circuit o f Figure
19.13, redrawn as Figure 19.29.
1O
F IG U R E 19.29. Realization of third-order 3 dB N L P Butterworth transfer function.

The second step in the completion o f a H P design is to replace each resistor with a capacitor whose
value is the reciprocal of the resistance; likewise, each capacitor is replaced by a resistor whose value
is the reciprocal o f the resistance value:
R
(19.33)
C
1060
Chapter 19 Principles of Basic I*iltering
Applying the transformation o f equation 19.33 to the circuit of Figure 19.29 produces the circuit
o f Figure 19.30, with one exception to the rule: the feedback resistors on the second-order section
that connect to the inverting input node are not changed to capacitors; these resistors only set up
a voltage division, so it is unnecessar)' to replace them.
F IG U R E 19.30. 3 dB normalized third-order high-pass filter in which
1.
In terms of transfer functions, the operation o f equation 19.33 changes s in the 3 dB N L P trans
fer function to \/s. Specifically,
H 7.CIBNHP
HM BNLP W =
s
M BNLP
(^ + 0
^ -f-5-f-l
Notice that each section of the transfer Rinction has a (normalized) H P characteristic. For example,
l i m ------ = 1 and lim

5 *o c
s+ 1
------- = I whereas li m ---- = 0 and lim
The final design is achieved by frequency-scaling by Kj-=

tor K
= 0.
and impedance-scaling by a fac
. The counterpart of equation 19.8 for the H P case is given by
In
-1 5
- 1
(19.34)
(The derivation o f equation 19.34 is assigned as a homework problem.) By convention, frequen

cy scaling places the magnitude response through the pass-band edge, in which case
(o, Hp = w "^ 1 0 '^ " - - 1 = 5 7 45.9 rad/sec

and K j- = 5745.9 .
1061
Let us suppose that the smallest capacitor is to be 10 nF. Then
= 13.05
t<fCne.
Hence the final circuit design is given by Figure 19.31.
13.3 nF
7.54 kO
M G L IR E 19.31. Final design of third-order high-pass filter.

The circuit o f Figure 19.31 presumes that an ideal voltage source is input to the filter. I f the source
has an internal impedance, this impedance will affect the pole locations and the overall perform
ance o f the high-pass circuit. To circumvent this problem, one may insert a voltage follower cir
cuit between the practical source and the input to the filtering circuit o f Figure 19.31, as illus
trated in Figure 19.32.
Practical
Voltage
Follower
Circuit
F I G U R E 19.32. Practical source in series w ith voltage follower circuit.
1062
It is possible to justify the RC to CR transformation through the use o f well-known networks

theorems. However, such justification, which is found in more advanced texts on filter design, is
beyond the scope o f our current endeavors. There are other H P design algorithms that do not uti
lize the transformation o f 19.33, but these are left to texts on filter design.
10. BAND-PASS FILTER DESIGN

Band-pass circuits were first considered in Chapter 16, mostly from an analysis perspective. In this
section we will briefly oudine the design o f a B P filter to meet a given set o f brickwall specifica
tions as shown in Figure 19.33.
A Loss(dB)
> CO
0)
0).
0)
P2
FIGURK 19.33. Typical brickwall B P filter specifications.

In Figure 19.33 O)^, and
10^2
pass-band edge frequencies while
and 0)^2
band edge frequencies. The center of the B P filter is defined as the geometric mean of pass-band
edge frequencies, i.e.,
= yjojp\(0p2
Given the B P brickwall specs, the design procedure begins with a frequency transformation to
convert these specs to equivalent N L P specs, i.e., (Q^ = 1,
and (Q^.,
To illustrate how
to generate these N L P specs we use a frequency transformation as follow's: define a bandwidth

w ith respect to
as Z? = 0)^2
define two potential N L P stop band edge frequencies
as
^.v/ =
Iko,i
(19.35)
for / = 1,2. Given these two numbers, the N L P equivalent specs are
= '
and (2, = m in[Q ,,, Q,^], A J
(19.36)
To illustrate the need for the magnitude sign in equation 19.35 and the need for the minimum
function in equation 19.36, let us do a simple example by breaking apart the formula o f equation
19.35 and
putting it back together. Suppose
= 2 rad/sec and 0)^2 = ^ rad/sec with to^j = 1
Chapter 19 * Principles of Basic Filrcring
rad/sec and
1063
= ^0 rad/sec. Then
rad/sec. Let us evaluate a part
- yj(0p^(0^y2 = ^
of equation 19.35 in which B is not present and the absolute value signs are gone. Thus
_ [-6
i=\
I 6
i=l
a ; ,,
[19.2
i= 2
One can view this partial transformation as the generation of two distinct L P filters: one defined
on negative frequency (Q^
6,
= - 6,
=-15) and one defined on positive frequency
= 19.2). O f course, we cannot really have two distinct L P filters, one for negative fre
quency and another for positive frequency. O u r mathematics requires that the magnitude respons
es of each filter be symmetric because magnitude is an even function of (o. Thus the only way to
properly interpret the above transformation is that it implicitly generates two distinct low-pass fil
ters, one of which is more stringent than the other. Specifically, the negative filter is really a LP
filter with edge frequencies (Q^, = 6, Q^, = 15), and the positive filter is really a L P filter with
edge frequencies
19.2). Observe that the ratio
^,v/
Q^,
|3.2
/= 2
differs for the rwo filters. The first filter is more stringent than the second, prompting the need
for a minimum fiinction in equation 19.36. Since our edge frequencies are always specified on the
positive axis and since the magnitude response is symmetric, w'e insert the absolute value signs in
equation 19.35. Finally, in order to have Q^- =
1 we divide by
- to^j so that the edge fre
quencies correspond to a N L P filter.
EXA M PLE 19.8. A B P filter has the specs

krad/sec, 0)^2 = 32.6 krad/sec,
= 26 krad/sec,
= 3 dB, and
= ^7 krad/sec,
= 30.6
= 14 d k Design a minimum-order
Butterworth passive B P filter meeting these specs, assuming the filter is driven by a voltage source
in series with a
10 kQ
source resistance.
S o l u t io n
Step 1. F in d the equivalent N L P specs fo r this filter. Using M A T L A B , we have

wsl = 26e3; ws2 = 37e3; w'pl = 30.6e3;
wp2 = 32.6e3; Amax = 3; Amin = 14;
K = w pl*w p2
K=
997560000
B = w'p2 - w pl
B =
2000
ws = [wsl ws2];
Wsi = abs((%vs M - K) ./(B*ws))
1064
W si =
6.1838e+00 5.0195e+00
W s = m in(W si)
W s =
5.0195e+00
Step 2. Find the filter order.
n = buttord(l,W s,Am ax,Am in,V
n= 1
Step 3. Find a realiztition o f the NLP filter. Since
= 3 dB, the first-order Butterworth N L P
has transfer fûnction Hf^j^p{s) = ----,whose passive circuit realization is given in Figure 19.34.
.V+ 1
-O
+
1Q
1F
F IG U R E 19.34. Rcalizjuion of the N L P transfer function.

Step 4. Realize the given BP filter. The algorithm to generate the B P filter from the N L P filter is
based on the frequency transformation
.v +
(19.37)
Bs
This frequency transformation causes a change in the capacitive and inductive circuit elements o f
the N L P circuit. From equation 19.37, the impedance o f an inductor changes as follows:
Ls
( s - + K\
Bs
L
= s +
B
1
(19.38)
\LK I
rhe expression on the right of equation 19.38 is a sum of impedances. This sum, then, is a cir
cuit composed o f an inductor in series with a capacitor as shown in Figure 19.35.
/YY\
L
L
B
3.
LK
F IG U R E 19.35. LP to BP transformation on an inductor.

Similarly, from equation 19.37, the admittance of a capacitor changes as follows:
106S
Chapccr 19 Principles o f Basic Filtering
= .V+
Bs
(19.39)
XK)
The expression on rhe right is a sum of admittances. Hence under the frequency transformation,
a capacitor becomes a parallel L C combination as illustrated in Figure 19.36.
C
CK
F IG U R E 19.36. LP to B P transformation on a capacitor.

Applying the transformation of Figure 19.36 to the N L P circuit of Figure 19.34 yields the circuit
o f Figure 19.37, where
- = lO""^ and = 2.005 x 10'^\
K
------- O
+
1O
1
'ôut
------- O
F IG U R E 19.37. BP circuit without magnitude scaling.
Step 5. Miignitudc-scale to obtain the correct source resistance. W ith
= 50 nF and
10"^, we conclude that
= 20 m H . The final circuit realization is illustrated in Figure 19.38.

Km
-o
I
_ BK
bp
k"
-o
F IG U R E 19.38. Final BP design.
The final transfer function o f the circuit is
/Y gp (.v)-
Bs
This transfer function and circuit show a clear similarity to the material developed in Chapter 16.
In contrast to the above procedure, the development o f Chapter 16 works only for w =
1 in
the
equivalent N L P circuit, which results in a B P circuit o f order 2. I f the order o f the N L P equiva-
'0 6 6
lent were 2 or higher, rhe B P order would be at least 4 and the background from Chapter 16
would prove inadequate. The next example illustrates how design is done for higher-order circuits.
E X A M P L E 19.9. A B P filter has the specs to^j = 1500 rad/sec,

rad/sec, to^^ = 4000 rad/sec,
= 2 dB, and
= ^^00 rad/sec,
these specs that is driven by a voltage source having an internal resistance of

So
= 3000
= 20 dB. Design a passive B P filter meeting
100 Q.
lu t io n
Part
1. F in d equivaletn N L P specs.
Define K =
12 x 10^ as the square o f the center fre

respect to
as 5 = co^2 ~
~ 1000 rad/sec.
=
quency of the filter. Define the bandwidth with
W e use the following M A T L A B code to obtain the necessary numbers.

wsl = 1500; ws2 = 6000;wpl = 3000; wp 2 = 4000;
B = wp2 - wpl
K = w p l*w p 2
ws = [wsl ws2];
W si = abs((ws .^2 - K ) ./(B*ws))
W s = m in(W si)
The M A T L A B output is
B = 1000
K = 12000000
W si = 6.5000e+00 4.0000e+00
Ws =4
Therefore, the equivalent N L P specs arc
= 2) and (Q^ = 4,
= 20)
These N L P specs are then used to design a N L P transfer function.

Part 2. F ind an equivalent N L P Butterioortb transfer fin c tio n an d its realization. In M A T L A B , we
have
W p = 1; W s = 4; Amax = 2; Amin = 20;
n = buttord(W p,W s,Am ax,Am in,V)
n=2
From Table 19.1, the second-order Butterworth filter is given by
" T T T ir r T
(19.40)
1067
Chapter 19 Principles of Ba.sic Filtering
at Q = 1. W e will adjust this by frequency-scaling the circuit
This docs not have the correct

that realizes
^ circuit that realizes the second-order Butterworth transfer function is
given in Figure 19.39.
yvYV
-o
1n
'- 0
F IG U R E 19.39. Second-order circuit.
The circuit o f Figure 19.39 has transfer function
LC
1
(19.41)
H----- A + ------
LC
By equating the denominators of equations 19.40 and 19.41, we obtain L = ~
H and C = V2 F.
V2
spec at Q = 1. For this we use
Next we frequency-scale so that we obtain the correct
1
c,w/
in which case
- W - \A^mat _ I
= 1.1435 rad/sec
= 0.61838 H and
= 1.2368 F.
Thus the circuit that realizes the N L P characteristic with a loss of
at Q =
1 or, equivaiendy,
3 dB of loss at Q^= 1.1435 is given in Figure 19.40.
IY Y \
IQ
-O
0 .6 1 8 3 H
+
V out
1 .2 3 6 8 F
F IG U R E 19.40. Second-order circuit.

Note that the circuit of Figure 19.40 has transfer function
o
, - H},dBNLP
u
n^LPkS)
1.3076
U .1 4 3 5 J
-1-
(19.42)
\ \ A 435}
+1
+ 1.617 Ly-f-1.3076
Part 3. Realize the passive band-pass filter. To compute the desired B P transfer function, one would
replace s in
as follows:
1068
+Wp\COp2
'
Bs
One could then attempt to realize the B P transfer function directly. This is numerically unwise
and, fortunately, unnecessary. A simple, numerically sound procedure is to replace each inductor
and capacitor in the N L P realization by an equivalent circuit representing the above frequency
transformation. These equivalent circuits were developed in equations 19.38 and 19.39, and illus
trated in Figures 19.35 and 19.36, respectively. The substitution scheme is repeated in Figure
19.41.
/YTV
L
L
B
_B
LK
(a)
(b)
F IG U R E 19.41. Illustration of the LP to B P transformation in terms of inductors and capacitors.
Given the transformations of Figure 19.41, the following M A F L A B code produces the desired B P
circuit parameter values, assuming the final source resistance is
% Insert L, C, B, K, and Km
L = 0.61838; C = 1.2368;
B = 1000; K = 12000000; Km = 100;
% Compute K1
K f = sqrt(K)
K f= 3.464 le+03
% Compute B P circuit parameters associated with L
Llb p = Km *L/B
L I bp = 6.1838e-02
C lb p = B/(L*K )/Km
C lb p = 1.3476e-06
% Compute B P circuit parameters associated with C
>)C2bp = C /(B*K m )
C2bp= 1.2368e-05
L2bp = K m *B /(C K)
L2bp = 6.7378e-03
100 O..
1069
The final circuit is given in Figure 19.42, where R^= 100 ^2,
= 61.8 m H ,
- 1.35 |.iF,
Îbp ~
LI bp
Cl bp
R.
-O
+
C2bp
V.
ôut
L2bp ^
F IG U R K 19.42. Final BP filter realization.

This concludes our illustration of the passive B P filter design procedure.
Active B P design is beyond the scope of this text, but coverage can be found in more advanced
texts.
11. AN ALGORITHM FOR SINGLY TERMINATED

BUTTERWORTH LOW-PASS NETWORKS
In 1937 E. L. Norton published a paper in the Bell Systems Technical Jou rn a l? The paper contains
explicit formulas for the L C element values in a singly terminated 3 dB normalized Butterworth
L P filter where the load or source resistance is 1 ohm. O u r goal in this section is to present these
formulas without derivation. Similar formulas have been developed by other researchers in the
context of network synthesis, a subject for which many texts are available for further reference.
For an th order 3 dB normalized Butterworth filter with a single
1H
termination, the sequen
tial formulas for the L C elements are

' 7T \
= s in
(19.43a)
\I n )
/3 > t ^
s in
S in
a-) =
\ 2nj
I In ]
co s*
sm
(19.43b)
sm
2/1
( 1 9 .4 3 c )
In
1070
a = ;jsin
JT
(19.43d)
\In }
It turns out that equation 19.43d is a special case o f 19.43c with m = n.

These formulas can be easily programmed into a M A T L A B m-file as follows:
function buttLC (n)
nn = 2*n;
a = zeros(l,n);
a(l)=sin(pi/nn);
for m = 2:l:n ;
X = a(m -l)'(cos((m -l)'pi/nn))^ 2;
a(m) = sin(( 2*m-l)*pi/nn)*sin((2*m-3)*pi/nn)/x;
end
Elvalues = a
To illustrate the use, suppose the Butterworth filter order is 4. Then
buttLC(4)
Elvalues =
3.8268e-01
1.0824e+00
1.5772e+00
1.5307e+00
Given these values, it remains to interpret them as inductances or capacitances. Two rules govern
the realization:
(i)
(ii)
There is never a shunt element in parallel with a voltage source.

There is never a dangling element at the load end.
Hence for the fourth-order filter above there are two possible circuit realizations, given in Figure
19.43.
0.3827 H
1.577 H
/YYY.
/YY\
1Q
V.
1.5307 F
1.0824 F
(a)
-O
+
1071
0.3827 H
1.577 H
/YYY
/YY\
1Q
1.0824 F
-O
+
1.5307 F
-O
(b)
F IG U R E 19.43. (a) R ^ = \ Q . (b) 7?/ = 1.
All that remains in the design procedure is frequency and magnitude scaHng, which are covered in
earher sections o f the chapter.
12. SUMMARY
This chapter has covered the basics of Butterworth LP, HP, and B P fiher design. Such design tech
niques build on a set of filter specifications requiring that the desired filter magnitude response lie
outside certain briclcvvall regions. Finding transfer functions that meet a set o f briclcwall con
straints is called the approxim ation problem . This chapter developed algorithms for finding
Butterworth transfer functions for the LP, HP, and B P cases. In addition, basic passive realizations
were presented as well as active circuit realizations, except in the B P case. In the active case, the
focus was the Sallen and Key low-pass and high-pass topologies. Both the passive and active t)^pes
of realization are built around the coefficient-matching technique, which associates the coeffi
cients in the circuit transfer function with the coefficients of the desired transfer function; one
then solves the resulting equations for appropriate circuit parameter values.
There are, of course, many other types of filter transfer functions; Chebyshev, inverse Chebyshev,
and elliptic are other well-known t)'^pes. Also, in addition to low- and high-pass filters, there are
band-pass, band-reject, and magnitude and phase equalizers. To add to the richness of the area of
filtering, there are analog passive, analog active, recursive digital, non-recursive digital, and
switched capacitor implementations o f all o f these filter types. The preceding exposition is mere
ly a drop in a very large and fascinating bucket of filter design challenges.

3 d B bandwidth o f B P filter: the difference between the two frequencies of a B P filter at which
the gain is 3 dB down from the maximum gain; if the maximum gain is 1, these fre
quencies correspond to a gain of M -Jl this is the most common reference for the mean
ing o f bandwidth.
3 d B bandwidth o f L P filter: the 3 dB down frequency, to
Active realization: a realization consisting of op amps, /?s, and C s.
Approxim ation problem: the problem of finding a transfer function having a magnitude
response that meets a given set o f brickwall filter specifications.
1072
Attenuation (dB); the loss magnitude expressed in dB, i.e., y4((o) = - 2 0 logjQ|//(/(D)|.
BP to LP frequency transfonnation: the generation of equivalent normalized LP specs from a
given set o f BP specs.
Bandwidth with respect to
the difference between the two frequencies o f a BP filter at
which the gain is down by a value of
from the maximum gain.
Band-pass filter: a filter that passes any frequency within the band 0)^j s O) s
while signifi-
cantiy attenuating frequency content outside this band.
Coefficient-matching technique: a method o f determining circuit parameter values by matching

the coefficients of the transfer funaion of the circuit to those of the desired transfer func
tion and solving the resulting equations for the circuit parameters.
Cutoff frequency: the frequency at which the magnitude response o f the filter is 3 dB down from
'
its maximum value.
Filten a circuit or device that significandy attenuates the frequency content o f signals in certain
frequency bands and passes the frequency content within certain other, user-specified fre
quency bands in the sinusoidal steady state.
Frequency response magnitude: magnitude of the transfer function as a function of j(a, i.e.,
^
Gain in dB: 2 0 logjo|//(/w)| or 10 log,o|//(/co)p.

Gain magnitude: frequency response magnitude.
Half-power point: the point at which the magnitude response curve is 3 dB down.
High-pass fUten a filter that significantly attenuates the low-frequency content of signals and
passes the high-frequency content.
HP to LP frequency transformation: a transformation that converts a given set o f high-pass
'
brickwall specifications to an equivalent set o f low-pass specifications.

1
Loss magnitude: the inverse of the gain magnitude, i.e., ^ ^ ^ .^ y

Low-pass (brickwall) filter specification: a filter specification requiring that the desired filter
magnitude response lie outside certain brickwall regions.
Low-pass filter: a filter that passes the low-fi-equency content o f signals and significantly attenu
ates the high-frequency content.
LP to BP fi%quency transformation: a technique for converting a passive LP filter to a BP filter

by converting Vs to series LC circuits and Cs to parallel LC circuits.
LP to HP frequency transformation: a technique for converting a passive LP filter to a HP fil
ter by converting Vs to Cs and Cs to Vs.
Maximally flat: property o f a filter at a point (0 wherein the magnitude response has a maximum
number o f zero derivatives.
Normalized Butterworth loss functions: a set o f Butterworth transfer functions, ordered by

degree, having 3 dB loss at the normalized frequency s= jQ = j\ .
Passive analog filter: a filter composed only of resistors, capacitors, inductors, and transformers.
R C to CR transformation: a technique for translating a low-pass active filter to a high-pass active
'
filter in which resistors become capacitors and capacitors become resistors.
* Wireless En^neer, vol. 7, 1930, pp. 536-541.

^ E. L. Nonon, Constant Resistance Networks with Applications to Filter Groups, Bell Systems TechnicalJournal,
vol. 16, 1937, pp. 178-196.
'
K r3
'Problems
2. Figures P I 9.2 (a) and (c) show an approxi
LOW-PASS BASICS
1. Filters can have magnitude
different transfer functions H{s). Figures P i 9.2
mate gain magnitude response associated with
responses quite
different from the maximally flat Butterworth

magnitude response. For each of the plots in
Figure P i 9.1, identif)'^ whether it is HP, LP, or
(b) and (d) show new gain magnitude respons

es whose transfer functions are related to the
H{s) o f (a) and (c) by scaling. Properly label the

horizontal axis of each graph in (b) and (d).
BP
Radian Frequency
(a)
Radian Frequency
(b)
Figure P i 9.2
3. A N L P filter transfer function
(0
Figure P I 9.1
0.65378
. r -h 0.803816435 + 0.82306043
10 7 4
has a pass-band edge frequency

at
= 2 dB)
1 rad/sec.
(a)
Use M A T L A B to plot the magnitude

response o f the transfer function and
verify that the magnitude response is 2
dB down from its maximum of
1.
= > / i # lA.
= e.
-1
(c o A
(OPJ
7. Suppose //jW and
third-order N L P
are second- and
Butterworth transfer func
(b)
Use the roots command in M A T

L A B to compute the poles of
tions, respectively, with co. = 03^ = 1 rad/sec and
(c)
The transfer function is to be scaled so
//(0) =
1.
that the new pass-band edge frequen-
(a)
Findand//.W .
cy is 750 Hz. Compute the new trans
(b)
Using M A T L A B and the freqs com

mand, plot the magnitude frequency
fer function. W hat are the poles of this

new transfer function? Where are the
response of H^{s) and //2W for
zeros of this transfer function?
< 5 rad/sec on
0 < O)
the same graph.
Properly label your plots.

4. Suppose H{s) is the transfer function of a sec-
(c)
Find the step responses for both sys

tems using the residue command in
ond-order Butterworth low-pass filter, H{Qi) =
M A T L A B to compute the partial frac
1, with 3 dB frequency at 500 Hz.

(a)
Determine H{s).
tion expansion, and then use M A T
(b)
Using M A T L A B , plot the magnitude
L A B to sketch the response curves.
and phase responses.

(c)
Determine the magnitude of the trans

fer function at 250 Hz and
(d)
1 kHz.
LOW-PASS APPROXIMATION
8. Find a minimum-order Butterworth transfer
Realize the filter using the circuit of
function meeting the brickwall specs in Figure
Problem 14 as follows:
P i 9.8 as follows:
(i)
First realize
(ii) Frequency-scale the circuit
to
(a)
Find the filter order.
(b)
Find the range of allowable
(c)
Find the poles, zeros, and gain con
range o f allowable^-
obtain the proper cutoff frequen

cy for the filter.
and the
stant of the 3 dB N L P transfer func

tion and verify that they lie on the
5. Fill in the details to the derivation of equa
unit circle.
tion 19.8, i.e., of the relationship

(d)
to,
0),
AAn
sections.
(e)
6. Recall equation
Find the 3 dB N L P transfer function

as a product o f second- and first-order
Using co^
find the zeros, poles, and
gain constant o f the non-normalized
19.2, i.e.,
transfer function. Do the poles lie on
H iJ c o )
2n
+ '
a circle about the origin?
.2n
1+
(0
' CO '
Accurately plot (using M A T L A B ) the

magnitude response of your filter in
terms of gain and gain in dB for 0
<f
This equation specifies a relationship between
< 1.2^. Does your filter meets the
the coefficient
e , (O^, and to^.. Show that for any
coeff
given brickwall specs?
L P brickwall
specification the range o f e
brick
given by
is
1075
and
Loss (db)
= (45 dB, 450 Hz) as detailed
below. Hint: M odify the M A T L A B code of

Problem
(a)
(b)
8.
Find the filter order.

Find the range o f allowable co^ and the
range of allowable/^
(c)

stant of the N L P transfer function and
0.3 dB
verify that they lie on the unit circle.

> f ( H z ) (d )
Find the N L P transfer function as a

product o f the second- and first-order
sections.
Figure P I 9.8
(e)
Using (JL)^
gain constant of the non-normalized

A N S W E R : Enter the following M A T L A B code
transfer function. Do they lie on the
to obtain the solution.
unit circle;
fp = 100; Fs = 1200;Amax = 0.3; Amin = .35:
(0
n = buttord(lp,fs,Ama.\,Amin. V)

magnitude response o f your filter in
fcmin = fp/((10A(0.1*Amax)-l)^(l/(2*n)))
terms of gain and gain in d B for 0 < f
Fcmax = fs/((1 0^ (0.]*A m in)-l)''(l/(2 n)))
< 1.2^. Does your filter meets the
wcmin = 2 pi Fcmin
given brickwall specs?
wcmax = 2*pi fcmax

wc = wcmin:
11. Repeat parts (e) and (0 of Problem 10
Ic = fcmin;
using
[z,p,k] = buttap(n)
M A T L A B code of Problem 10 as necessary.
instead o f
M odify the
zplane(p)
grid
12. The specs
pause
znew = z'w c
pnew = p \vc
knew = k wc'^n
f=
f^) = (2 dB, 100 Hz),
= (20 dB, 500 H i) , and

= (40 dB, 1000 Hz) are associated with the
brick%vall specifications o f Figure P i 9.12.
0:fc/50:l. 2 * f s ;
Loss (db)
h=
freqs(knew* poiy(znew),poIy(pne\v),2' p i' t ):
plot(f,abs(h))
grid
pause
plot{l',20 lo g l 0(abs(h)))
srid
2db
9. Repeat parts (e) and (0 of Problem
^cnuix instead o f
L A B code of Problem
8 using
2000n
M odify the M A T
8 as necessary.
10. Find a minimum-order Butterworth filter

meeting the specs
> 0)
lOOOn
= (1 dB, 75 Hz)
Figure P I 9.12
(a)
Find the minimum filter order meet

ing these specs.
1076
(b)
Find the range of allowable co^ and the
range o f allowable
(c)

stant of the N L P transfer function.
(d)
' " O
Find the N L P transfer function as a

product of second- and first-order sec
Figure P I 9.13
tions.
(e)
Using
gain constant of the non-normalized

transfer fijnction.
(0
14. The circuit of Figure P I 9.14 is to realize a

maximally flat second-order low-pass filter with
3 dB frequency
Now find the transfer fijnction as a prod
(a)
= 500 Hz.
uct of first- and second-order sections.

(g)

magnitude response of your filter in
(.v) =
LC
terms o f gain and gain in dB for 0 < f
<
1.2f^. Does your filter meet the given
briclcwall specs?
LC
(b)
PASSIVE LOW-PASS
REALIZATION
W ith /?^ = 1 Q, compute I in H and C

in F to realize the normalized secondorder Butterworth transfer function.
is to be i o n in your final design,
(c)
13. The circuit o f Figure P i 9.13 is to realize a
choose a magnitude scale factor,
maximally flat second-order low-pass filter with
and the proper frequency scale factor,
3 dB frequency
Kf, to obtain the given filter specifica
(a)
= 1000 Hz.
Show that the circuit transfer function is
tions. Then compute the new values of
L and C.
= s~ +
(b)
LC
,
,
-s+R^C
LC
(d)
As an alternative design, suppose C is

to be
1 |.iF. Find
the new values of
and Kj- to achieve the original filter

specifications. W h at are the new val
W ith Rj = \ Q., compute L in H and
ues of L and R jj
C in F to realize the normalized sec

ond-order Butterworth transfer func
tion.
(c)
(D
I f Ri is to be 1 kQ in your final design,

choose a magnitude scale factor,
and the proper frequency scale factor,
Figure P i 9.14
K p to obtain the given filter specifica

tions. Then compute the new values of
L and C
(d)
As an alternative design, suppose C is

to be
1 uF.
Find the new values of
(a)

circuit is
and Kj- to achieve the original filter

specifications. W h a t are the new val
ues o f L and /?^?
în
LC
1
R,
1
1 + ^sl^L
s~ -f- ---- - H----- 5 +
CR,
L
LC
1077
If R = 2 Q and Rj = 8 Q, find two solu
(b)
tions {L, Q so that H{s) has a second-
17. Consider the circuit of Figure P I 9.17.

(a)
Show tiiat the transfer function is
order Butterworth (maximally flat)
response with 3 dB down (from its

maximum gain) frequency equal to
R ,L C
H{s) =
T
.V +
rad/sec.
s + + ! / ,
LC
Magnitude- and frequency-scale both
(c)
net\vorks resulting from part (b) so
(b)
If Rj= ^
find two sets of parameter
that in the new networks the smaller
values that realize the second-order
resistance is 2 kQ, and the 3 dB fre
normalized Butterworth gain func
quency is 5 kHz.
tion.
(c)
Magnitude- and frequency-scale both

networks resulting from part (b) so
that in the new nervvorks the smaller
resistance is 2 k il, and the 3 dB fre
quency is 5 kHz.
R,> 1
Figure P I 9.15
16. W ith
1Q
R^= Rj = \ Q .m the circuit o f Figure
P19.15, the transfer function becomes

Figure P I 9.17
LC
H .A s) =
s~ +
2
---- f- -- y -I-----\C
U
LC
.A N S W E R S : (b) (C, L) = (0.4483, 3.3460) or

(1.6730, 0.8966) in F and H; (c) ( C L) = (26.6
which can be used to realize the normalized sec
nF, 57.1 m H ) or (7.1 nF, 0.213 H ).
ond-order Butterworth characteristic
4 [d l
H i s ) = T.V
K
-I- J 2
:s
18. Consider the second-order low-pass trans
fer function
for K = 0.5. Suppose such a filter is to have cut
H(s) =
off f r e q u e n c y = 3500 Hz. (In this case, the fil
V'^p
Find values of L and C to realize the

where (0^^= 10^ rad/sec.
normalized transfer function.

I f the final value o f the resistors is to
be
(a)
1 kt^, compute the new values o f L
Alternatively, suppose the final value

o f the capacitor is to be
this case?
that realize the desired filter.
(b)
Find the impulse response and step
(c)
Using the RLC topolog)' shown in
response.
10 nF.
Compute the new values of L and C
Find |Myw)| and co^ (the 3 dB down

frequency). Is it true that co^ = O)^ in
and C that realize the desired filter.

(c)
( s
maximum gain at f^.)
(b)
^
în
ter is to have magnitude 3 dB down from its

(a)
becomes
Figure P i 9 .18a, realize H{s) given that

/? = 10 k n in your final design. W hat
is the value of K in vour solution?
1078
Hint: First set O)^ =

quency(d)
and
1 and
then fre
magnitude-scale
gain is down 3 dB from its maximum
to
value), find the new element values.
obtain the final answer.
A N S W E K . C, = C , = 0 .106 u F and
Using the R L C topology shown in
= 0.2122 H.
Figure P i 9.18b, realize H{s) given that
R^= 100
in your final design. W hat
20. Consider Figure P i 9.20.
is the value o f A" in your solution?
(a)
lY T v
L^LjC
p .
I \
n
----- 1
^ ^1
^2/
( L\ + Lo + C\
2
+ ------ =------ j + ---------L 1L2C
}
L^LjC
(a)
yvYY
(b)
Determine values so that the circuit

realizes a third-order Butterworth gain
characteristic.
(c)
(b)
Find parameter values o f a third-order

low-pass Butterworth
filter having
cutoff 20 kHz and resistor values o f

Figure P I 9.18
kn.
(d)
19. Consider the doubly terminated filter cir
Use S P IC E to verify the frequency

response.
cuit o f Figure P I 9.19
10
/Y Y \
+
Vo,
/Y Y \
L,
K
p
10
I
>
>
Figure P i 9.20
Figure P I 9.19
A N S W F R S : (b) A, = L , = ] H, C = 2 1-: (c)
(a)
Show that the transfer function of the
15.9 nF and 15.9 niH.
circuit is
L C .C .
c > ( .0 = -
c ,l c'-2
(b)
L C .C t
5 4-
By equating the coefficients o f the
2I.
denominator o f
with the
denominator o f the third-order 3 dB
shown in Fig ure P I 9.21.
Consider the two-port configurations
(a)
N L P Butterworth loss function, show

= C2 =
1F
Prove that for Figure P I 9.2la
ôiit _
that the only possible realization is

(c)
L C .C
l '-2
^21
and Z. = 2 H.
If the source and load resistance are to

be 1 k n an d^ = 1500 Hz (meaning the
(b)
Prove that for Figure P I 9 .2 lb
1079
N LP
-V21
V22 +
Vin
(c)
transfer function obtained in
part (a).
Making use of the results o f part (a),
(c)
Magnitude- and frequency-scale the
design a second-order Butterworth
circuit so that the gain is 3 dB down at
low-pass filter having to^ = co^ = 1
^'Kntin
rad/sec,
=1
and Rj = so. Hint:
Divide the numerator and denomina
(d)
source resistance is 100
'
Plot the magnitude response in terms

of dB gain vs./ (in Hz).
tor o f His) by s, and then equate the

result with the gain expression to
obtain two z-parameters. Next, review
ing of one series element and one
ACTIVE LOW-PASS DESIGN

23. Consider the circuit of Figure P i 9.23,
shunt element (a special case o f a T- or
which has a practical source driving a leaky
7i-network).
integrator circuit.
the z-parameters of a two-port consist
Finally, put all results
together to design the two-port.

(d)
(a)
u s i n g R^^^ = R^ + R ^.
Making use o f the results o f part (b),

associated with Figure 19.21b, design
Compute the circuit transfer function
(b)
Realize
the
first-order
normalized
Butterworth transfer function
a second-order Butterworth low-pass

filter having co^= 1 rad/sec, R = 0 Q
^ 3dBNLP^^'^) - ~rr~r7 - r
and R^= \ Q.
(e)
V in is )
Frequency- and magnitude-scale the
i.e., compute values for R , R-,, and C.
circuits o f parts (c) and (d) so that the
Hint: Let C= I E
new filters have to^ = 5000 rad/sec,

and the single resistance in the circuit
(c)
I f the 3 dB down frequency is to b e ^

= 3500 Hz, find
is 1 k n .
S+\
so that the capac
itor is 1 nF in your design. Compute

the new resistor values and then deter
mine /?j if the source resistance is R^ =
500 Q.
C H E C K : /?, = 45 kQ.
22. (a)
Find the 3 dB
N LP
Butterworth
transfer function meeting the specs

= (2;o00, 2 dB) and (o)^,
Figure P i 9.23
= (2.T2000, 20 dB). Then find
24. Consider the circuit o f Figure P i 9.24 as a

(b)
Using one o f the circuits with a non

zero source resistance from the earlier
homework problems, realize the 3 dB
candidate for realizing the second-order lowpass transfer function
1080

o
m s ) =_
*out _
( s
0).
(a)
(b)
(c)
For a normalized design, O)^ = 1 rad/sec

and A" = 10, choose element values to
realize the given transfer function; i.e.,
find all circuit parameter values.
+1
(c)

of the circuit of Figure P I9.24.
(b)
For a normahzed design, co
and
compute the final circuit parameter
rad/sec and K = - 1 0 , choose element
values. Simulate using SPICE to verify
values to realize the given transfer
your design. (The solution is not
fijnction; i.e., find all circuit parame
unique, vy^hich is often the case for
ter values.
practical design problems.)
If (0^= 10^ rad/sec and the capacitors

are to be 1 nF, choose AV-and
K m and
j
compute the final circuit parameter
values. Simulate using SPICE to ver
(d)
terms of (O^.
C H EC K : 0)^ = 0.5098co^
r>
R.
R.
= 1 kn
the
down
Find the 3 dB down frequency, o)^ in
ify your design.

Find
If 0)^= 10^ rad/sec and the capacitors

are to be 1 nF, choose iSând
= 1
C H EC K :
dB
frequency,
0
+
0)^ in terms of (O .
C H EC K :
(0^ = 0 .6 4 3 6 co^
r^
n
Figure P I9.25
26. A second-order normalized Butterworth fil
ter can be realized by the Sallen and Key circuit
described in section 5. In the final design, the
filter is to have a dc gain o f 1, a 3 dB down fre
quency
= 1000 Hz, and largest capacitance
of 10 nF.
(a)
Figure P I9.24
Using design A and input attenuation
n
n
if necessary, determine the appropriate

25. Consider the circuit of Figure P 19.25 as a
candidate for realizing the third-order low-pass
(b)
transfer function
Sallen and Key circuit.

Using design B and input attenuation
if necessary, determine the appropriate
K
(c)
Using the Saraga design and input
attenuation if necessary, determine the

appropriate Sallen and Key circuit.
Hint: It is often useful to generate an Excel
(a)
spreadsheet to do the relevant calculations,

=
circuit of Figure P I 9.25.
of
especially for multiple designs such as these;
n
o
this allows one to set forth the normalized and

final designs along with
and
n
1081
Chapter 19 Principles oi Basic Filtering
27. Realize rhe filter o f Problem
8 as a cascade
(b)
M odify the circuit to achieve the
o f a Sallen and Key circuit (Saraga design) and
desired gain of the transfer function.
a first-order active circuit such as the leaky inte

grator. The 3 dB down point should be at
(c)
determine the new clement values.
(H int: It is often useful to generate an Excel

spreadsheet to do the relevant calculations,
If the actual pass-band edge frequency

is to be 7 kHz and C j is to be 0.05 uF,
largest capacitor should be 50 nF.
^'^cmhr
W lia t is the dc gain, K, o f the circuit?
(d)
Plot rhe loss magnitude response of
/V(s) in part (a) in dB. Determine
especially for multiple designs.) Verify the fre

quency response o f your design using S P IC E .
C H E C K :^ ,= 2 d B
28. Repeat Problem 27, except use design A
instead ol rhe Saraga design for the Sallen and
32. Repeat Problem 31 using design A of the
Key portion.
29. Realize the filter o f Problem 10 as a cascade
33. Repeat Problem 31 using design B of the
of two Sallen and Key circuits. The 3 dB down
Sallen and Kev circuit.
point should be at
The largest capacitor
should be 0.1 uF.
HIGH-PASS PASSIVE DESIGN

30. Repeat Problem 29, except use design B
34. A second-order Buttenvorth H P filter has 3
instead of the Saraga design for the Sallen and
dB down point at^. = 2000 Hz. 'Fhe second-
Key portion.
order Butterworth N L P prototype is given in

Figure P I 9.34 and has transfer function
31. In addition to Butterworth filter transfer

functions, there are Chebyshev filters, which
LC
1
1
s~ -I- -------.V-f"
R jC
LC
have a faster transition from pass-band to stop
~>
band. This problem investigates the implemen

tation o f a Chebyshev LP filter transfer function.
Recall that a second-order transfer function
(a)
W ith R^= ] Q., compute L in H and

C in F to realize the normalized second-order prototype.
C H ECK:
(b)
C = l/ V 2
Using the results o f part (a), construct
can be realized by the Saraga design o f a Sallen
a N H P circuit with
and Key circuit.
^^cHP~ ^ rad/sec. This is the so-called
(a)
Determine the new values o f
/?,,
3 dB
normalized
second-order
(c)
(d)
------------
0.803825+ 0.82306
Hint: Frequenc)--scale, s
(DqJ where
ojq= 0.82306. W h at is the resulting

transfer function?
Now construct the final H P design

with R j =
Chebyshev L P transfer fiinction
filter (3 dB
N H P ).
Cj, and C j that will realize the poles

of
normalized H P
\ Q. and
100
Do a S P IC E simulation to verif}- your

design.
1082
(c)
Using the results of part (b), construa

a N H P circuit with the source and
load resistances equal to 1 Q and
^c,HP ~ ^ ^<i/sec. This is the so-called

3 dB normalized H P filter (3 dB
N HP).
Figure P I9.34
(d)
35. A third-order Butterworth HP filter has 3

dB down point at
Given your answer to pan (c), con

struct an appropriate high-pass filter
= 5000 Hz. The third-
circuit whose resistor values are 1 ki2.
order Butterworth NLP prototype is given in

Figure P I 9.35 and has transfer function
r -W --/Y Y \
10
-
Hr
3
S
\ 2
Cj
LC jC j
+ Co
------
-o
+
1o
vJt)(
1
LC\C2
+ ----------
LC\C2
Figure P I9.36
Compute Z in H and Cj and C2 in F
(a)
to realize the 3 dB normalized third-
37. The circuit of Figure P I 9 .3 7 realizes a
a
CH ECK : q = 1.5 F and L = - H
a third-order high-pass filter circuit, with
Using the results of part
= 2 dB az L = 5 kHz so that the largest capaci
third-order NLP Butterworth filter. Determine
order prototype.
(b)
tor equals 100 nF, as follows:
(a), construCT a NHP circuit with

= 1 rad/sec. This is the so-called 3 dB
(a)
normalized HP filter (3 dB NHP).
(d)
max
and
for the
(b)
Compute O)^ for the desired HP filter.
the source resistance is to be 100 Q.
Does your answer make sense? Think
Do a SPICE simulation to verify your

(c)
about this carefully.

Realize the 3 dB NLP transfer func
Now construct the final HP design if
(c)
Determine
NLP filter.
design.
tion
using the circuit of Figure
P 19.37. Set I = 2 H and make Cj =

C2. Recall from Problem 19 that the
transfer function of the circuit is
1
VquA^)
Vin(s)
36.
second-order
high-pass
LC1C2
c, +c
c,c
1^2
LC 1C 2
LC 1C2
Butterworth loss function has stop band edge fre

quency fs= i kHz, pass-band edge frequency
7 kHz, and cutofi^fijequencyf(. = 5500 Hz.

(a)
Determine the attenuation in dB 2x fp
(b)
and^.
Realize the second-order 3 dB NLP
Butterworth filter using the circuit of
Figure P I9.36; i.e., compute values for L
and C.
(d)
Construct the appropriate high-pass

circuit with the largest capacitor equal
to 100 nF.
(e)
Verify the frequency response o f your

design using SPICE.
1083
41. Reuse the specs of Problem 40, except that

there is to be negligible source resistance and a
X
IQ
c.
load resistance of 75
(a)
Compute the transfer fianction, say

o f the circuit in Figure P i 9.41
in terms o f Z j, L 2, and C
Figure P I 9.37
Determine values so that the circuit
(b)
realizes a third-order Butterworth 3
38. The brickwall specs of a H P filter are shown

in Figure P i 9.38, in which
2500 Hz,
= 2 dB, and
dB N L P Butterworth gain function.
= 500 Hz, ^ =
Determine the H P filter.
(c)
= 25 dB. The
filter is to be driven by a voltage source having

a source resistance o f 100
jy y
There is no load
6\
resistance. At the pass-band edge frequency the

attenuation should be exactly 2 dB. Develop a
passive H P filter circuit to solve this problem.
A A(w)
L,
1o
Figure P I 9.41
HIGH-PASS ACTIVE
REALIZATION
42. Consider the set of H P specs for which
- (2400 Hz, 3dB) and
(600 Hz, 30dB). Following the procedure o f
- 0)
Example 19.7, design an active H P filter meet

ing these specs so that the smallest capacitor in
the final design is 10 nF.
Figure P I 9.38
43.
Consider the set of H P specs for which
39. Repeat Problem 38 assuming that the filter
= (2400 Hz. 2dB) and

=
Hz, 30dB). Following the procedure of
is driven by a voltage source with negligible
(600
source resistance and that there is a load resist
Example 19.7, design an active H P filter meet
ance that in the final design is to be 75 Q.
ing these specs so that the smallest capacitor in
Consider using the N L P prototype o f Problem
the final design is 10 nF. The magnitude
34.
response should go through the pass-band

edge. Hint: See equation 19.34.
40. The brickwall specs o f the H P filter in

Figure P i 9.38 now apply to/^' = 500 H z ,/ , =
44. The Sallen and Key Saraga design H P filter
2000 Hz,
o f Figure P i 9.44 realizes the normalized H P
= 2 dB, and
= 30 dB. The
filter is to be driven by a voltage source having
filter transfer function
a source resistance o f 250 Q. There is no load

resistance. At the pass-band edge frequency the
H {s) =
K s-
attenuation should be exactly 2 dB. Develop a

passive H P filter circuit to meet these specifica
tions. Hint: Use the circuit o f Figure 19.8, hav
where A" = 1 +
ing transfer function of equation 19.14.
Butterworth filter is to have f ^ = 3
A second-order H P
kHz.
1o s^
Compute the parameter values o f the H P filter

of Figure P I 9.44. The smallest resistor should
/Y Y l
be 10 kD. Verify the frequenc)^ response o f your
L
8Q woofer
design using S P IC E .
80 tweeter
Figure P i 9.47
AN.SWl-.R: L = 63(> u H , C = 9.95 uF'
48. The crossover nersvork of the previous
problem provides first-order Butterworth low-
Figure P 19.44
pass and high-pass response curves. Better45. Again consider the circuit o f Figure P i 9.44.
(a)
Cascading two sections together, real

ize a fourth-order normalized H P
upgrading the responses to the second-order

Butterworth t}'pe. Design such a crossover net
work having the same crossover frequency and
Butterworth transfer function.

(b)
quality sound reproduction is achieved by
If the smallest resistor is to be 10 kH
loads as in Problem 47.
a n d ^ = 3 kHz, find the new parame

A X S W l'R : For the woofer. L = 0.90032 mH,
ter values for your design.
C = ".033*' uF. For the tweeter, /. = 0.)0032

46. You have a set of rweeters that feature great
sound reproduction for frequencies above
kHz. As a lark, you decide to build an active

high-pass Butterworth filter that will isolate
highs from a particular audio signal. The specs
you decide on are (^ = 5 kHz,
(^ = 1.5 kHz,
= 3 d B) and
= 40 dB). Determine the fil
ter as a product of second-order active Sallen

and Key circuits. M inim um
resistor values
should be 20 k H in a Saraga design.
MISCELLANEOUS
47. A certain audio amplifier has a very low
internal resistance. It is therefore approximate
ly represented by an ideal voltage sourcc. The
woofer and t\veeter each may be approxi
mately represented by a resistance of
8 ohms.
Design the simple crossover network shown in

Figure P i 9.47 such that the crossover frequen
cy is 2000 Hz.
m H. ( = ".0337 itl-.
Brief Introduction to
Fourier Series
CHAPTER OBJECTIVES
1.
Introduce the concept o f and calculation procedure for the Fourier series of a periodic
signal.
2.
Describe the relationship between the complex and the real Fourier series representations.
3.
Set forth and discuss basic properties of the Fourier series.
4.
Show how the basic Fourier series properties can be used to compute the Fourier series
of a wide range of signals from a few basic ones.
SECTION HEADINGS
1.
2.
Introduction
The Fourier Series: Trigonometric and Exponential Forms
3.
4.
Additional Properties and Com putational Shortcuts for the Fourier Series
Representation
Summary
5.
Terms and Concepts
6.
Problems
1. INTRODUCTION
Non-sinusoidal periodic waveforms are an important class of signals in electrical systems. Some
prominent examples are the square waveform used to clock a digital computer and the sawtooth
waveform used to control the horizontal motion of the electron beam of a cathode ray T V pic
ture tube. Non-sinusoidal periodic functions also have importance for non-electrical systems. In
fact, the study of heat flow in a metal rod led the French mathematician J. B. J. Fourier to invent
the trigonometric series representation o f a periodic function. Today the series bears his name.
The Fourier series o f a periodic signal exciting a linear circuit or system leads to a simplified
understanding o f the effect o f the system on the original periodic signal. This idea is briefly
explained in the next few paragraphs and illustrated in Example 20.1.
1086
Chapter 20 B rief Introduction to Fourier Scries
W hen a periodic input excites a linear circuit, there are many ways to determine the steady-state
output. Using a Fourier series method of analysis, the input is first resolved into the sum o f a dc
component and infinitely many ac components at harmonically related frequencies. For example,
a
1 kHz
square wave voltage with zero mean and a 0.5n V peak-to-peak value, w ithy{f) = {u{t)
-2u{t -0 .57) + u {t- 7)), T = 1 msec,

Vy(/) = 0 .2 5 jt ^
/(/ - n T )
( 20. 1a)
H=0
has the Fourier series representation
2/2 + 1
/i=0
( 20. 1b)
s in (w r) - ^sin(3cu^/) + ^ sin(5w ^/) - ... u ( t )
w h e re ^Q =
2jtx
1000 rad/sec. In the Fourier series representation (which we will later develop), we
observe that the signal has a zero average dc value and harmonically related frequency components
(-l)"sin ((2 + 1)0;^/)

2/ ; +l
By linearity and superposition, the effect of the linear system on v^{t) can be determined by sum
ming the time domain effects o f the linear system on each component in the Fourier series.
For steady-state calculations, let
Fourier series representation of
is given by
be a phasor representing the (2;/ + l)tli harmonic o f the

Then, in steady state, the effect o f the system on this term
= H {j{2n + 1)Wq)V^ , assuming that the transfer function H{s) is stable. Again by
linearity, the actual time domain output is then computed for each
The resulting time func
tions are then summed to obtain the steady-state part o f the complete response. Example 20.1
illustrates the particulars.
E X A M P L E 20.1. Figure 20.1b shows a simple R C circuit (/? =
C = 1 F) and a square wave
input voltage (Figure 20.1a) with E = 3 0 jt V and T = 4 sec. Find the first four components o f the
output voltage v^{t) in the steady state.
>
r
b
(a)
v Jt)
v jt)
(b)
F IG U R E 20.1 Series RC circuit excited by square wave,

used to demonstrate calculation of the steady-state response.
1087
Chapter 20 B rief Introduction to Fourier Scries
S o l u t io n
Step 1. Determine the Fourier series representation o f v-^ît). The fundamental frequency o f v-^^{t) is
Jq = 1/7 = 0.25 Hz or C0q = I jifo = 0.5jt rad/sec. As shown in die next section, u-^^{t) has the
infinite (Fourier) series representation
= 30;r ^ [u{t - n T ) - n{t - {/? + 0 .5 )7 )

/)=()
15;r + 6 0sin(a;o/) + 20sin(3a)oO + 1~ sin(5wQ/) +... u { t )
s in ((2 + ^
=0n
For this signal the average dc value is
15 ti. The
nil)
2// + 1
remaining terms in the Fourier series are harmon
ically related sinusoids of decreasing magnitude.

Step
2. F in d the circuit transfer fim ction.
The stable transfer function o f the circuit is
^ s) =
Vi,
. v+1
Step 3. D eterm ine the magnitude an d phase o f H{s) at s = j( 0 . For sinusoidal steady-state analysis,
set s =jo ) to obtain
Jd
H(jco) = -
JOJ + 1
=H/
in which case
, 6 = - tan ^(w )
yjco^ + 1
Step 4. Find steady-state responses to a ll components o f Vj^^{t). Table
20.1
lists the steady-state
responses to several components of Vj^Jit).

TA BLE
20. 1. Several Fourier Series Components of Vj{.t) and the
Corresponding Magnitude and Phase of the Response
0)
Wo
Input magnitude
157:
60
5o
20
12
0.5370
0.2075
0.1263
0 (degrees)
-57.52
-78.02
-82.74
Output magnitude
1571
32.22
4.150
1.516
Output angle
-57.52
-78.02
-82.74
Input angle'
'Angles are in reference to sine functions in this example.
C'hapicr 20 Brief Introduction to I\>iiricr Series
Step 5. Apply superposition to obtain the steady-state portion o f the complete response. Neglecting har
monics o f scventii order and higher, the approximate steady-state solution is
Vg{t) = 15j*r + 3 2 .22sin((/jyr- 57.52") + 4.15 sin(3w Q /-7 8 .0 2 ) + 1.516 sin(5wQ - 8 2 ,7 4 ) + ...
This response shows the effect of the system on each component of the input signal and how it in
turn affects the overall steady-state output response. Since the time constant o f the circuit is 1 sec,
this steady-state response more or less constitutes the actual response for r > 5 sec.
Exercise. Use MATLAB to plot the approximate waveform for vj^t) for 0 < r < 8 sec, based on
Tiible 20.1. From the plot identif)'
Alternatively, you can use the max
and min commands in MATLAB.

AN SW l-R: 82.013 V and 12.235 \'
In Example 20.1, the first step o f the solution was to represent a periodic waveform as a sum o f
sinusoidal components, called the Fourier series. Section 2 covers the definition and basic proper
ties o f Fourier series. Section 3 describes several shortcuts for computing the Fourier coefficients
and identifies other important properties. Since, in practice, only a finite number o f terms can be
considered, the Fourier series method yields only an approximate solution.
Because many mathematical and engineering handbooks have extensive tables o f Fourier series o f
different waveforms, it is convenient to use these tables in much the same way as one uses a table
o f integrals or a table o f Liplace transforms. The Fourier series o f some basic signals are provid
ed later, in Table 20.3. The use o f this table, together with some properties and shortcuts dis
cussed in section 3, make the study o f the Fourier series method much more palatable to begin
ning students o f circuit analysis.
2. THE FOURIER SERIES: TRIGONOM ETRIC AND

EXPONENTIAL FORMS
B a s ic s
A signaly(f) is periodic if, for some T> 0 and all t,

/ r + 7) = P )
(20.2)
r i s the period o f the signal. The fundamental period is the smallest positive real number Tq for
which equation 20.2 holds; ^ = l/7j, is called the fundamental frequency (in hertz) o f the sig
nal; cOq = I k/ q = 271/7*0 is the fundamental angular frequenc)^ (in rad/sec). T he sinusoidal wave
form o f an ac power source and the square wave form used to clock a digital computer are com
mon periodic signals. Figure 20.2 shows a portion o f a hypothetical periodic signal.
1089
Chapter 20 Brief Introduction to Fourier Series
THE REAL TRIGONOM ETRIC FOURIER SERIES

Under conditions tiiat are ordinarily satisfied by signals encountered in engineering practice,
a periodic function y(/^) has a decomposition as a sum o f sinusoidal functions
00
/ (0 = Y +
2 ! ( co s(m v )f/ 7sin (/ ;fo / ))

//= 1
(20.3)
Equivalently,
/ (O =
coa{mor + d)
(20.4)
where
(20.5a)
and
(20.5b)
Both infinite series, equations 20.3 and 20.4, are called the trigonometric Fourier series repre
sentations
We note that equations 20.5a and 20.5b follow fi-om the trigonometric identit)'
A cos(.v) + fisin(.v) =
with due regard to quadrant.

+ B~ cos .Y + tan ^ ----I Aj f
\
From equation 20.4, observe that

is the average value ofy(r). In electric circuit analysis,
refers to the dc component ofJ{r) . The first term under the summation sign, <^jCos(tOQt + ^^,), is
called the fundamentaJ com ponent (or first harmonic) ofy(r), with amplitude
and phase angle
0 j. The term d-, cos(2o;(jt + 6-,) is called the second harmonic o^J{t), with amplitude d-y and phase
angle
and similarly for the other terms d^^cos(;/WQt + 0^^), which are the ;/th harmonics as indi
cated by the term nô^y
Chapter 20 Brief Introduction to Fourier Scries
As illustrated in Example 20.1, given any periodic function/r), it is important to determine the
coefficients in equation 20.3 or equation 20.4. For the purpose o f easier calculation, it is advan
tageous to introduce the equivalent complex Fourier series representation o f the periodic func
tion y(f).
COM PLEX EXPONENTIAL FOURIER SERIES

For a periodic s i g n a l t h e so-called complex exponential Fourier series is
/ (0 =
(20.6a)
n = -ao
where it can be shown that
In + T,
c =
(20.6b)
Since we have two (allegedly) equivalent forms, let us now develop the relationship between the
real and complex forms o f the Fourier series. Recall the Euler identities:
cos( a-)
= --------------- and sm(A-) = --------------2

2j
Then equating the real and complex forms, we have
/ (')=
/I = - oc
/I = I
= -y + ^
n = \\
= ^
+ I
^----------- + h
(0.5(.; - jh
2j
+ 0.5( + jb
/! = 1
Equating coefficients yields
=
(20.7a)
and
c_ = 0.5
+ jh ) = c*
(20.7b)
where c*, is the complex conjugate o f f,,. Equivalently,

= 2 R e (0
b = -21m(^-)
(20.8a)
(20.8b)
1091
(20.8c)
^0 = ^0 = ^
d = 2\cln= 1,2, 3,...
(20.8d)
= ^ c ,n = 1, 2, 3 ,...
(20.8e)
In the real trigonometric Fourier series equations 20.3 and 20.4, the summation is over positive
intêr values o f, whereas in the complex exponential Fourier series the summation extends over
intêrs n such that -oo < < oo. While each term in equation 20.3 or equation 20.4 has a wave
form displayable on an oscilloscope, each individual term in the complex exponential Fourier
series lacks such a clear physical picture. However, two conjugate terms in the complex exponen
tial Fourier series always combine to yield a real-time signal d^cos{no)Qt + 6^.
To develop equation 20.6b for the coefficient
r > " 0 'to obtain
we multiply both sides of equation 20.6a by
* = -00
Integrating over one entire period, [/q,/q + T^y produces
Jto
k = -oo
(20.9)
* = -00
Because (OQrQ = 2ti and
it = -00
= 1, the integral in equation 20.9 becomes

cê
'v ^
'
'
0
cTq
for k ^ n
for k = n
(20.10)
Substituting equation 20.10 into equation 20.9 and dividing by Tq yields

c =
tn + T
Tn
(20.11)
The lower limit of integration, ?q, can be any real number, but is usually chosen to be 0 or - T qH,
whichever is more convenient. In addition, Tq will sometimes be written simply as T.
A hand computation of coefficients would proceed by first computing for = 0, 1,..., using equa
tion 20.11. One would then obtain and using equations 20.8a and b, and and 0^ by equa
tions 20.8c, d, and e. Other formulas are available for obtaining the Fourier coefficients by integrals
involving sine and cosuie functions. However, equation 20.11 is preferred because an integration
involving exponential fimctions is often simpler than an int^ration involving sinusoidal flmctions.
y
w
EXA M PLE 2 0 .2 . Find the trigonometric Fourier series for the square wave signal o f Figure 20.3.
S olution
The fundamental period o f/ f) is Tq = T. By inspection, the average (dc) value
is
To calculate other Fourier coefficients, choose Tq = -7 7 2 . Equation 20.11 yields, for n ^ 0,
(20.12)
A .(
JT^
sin n , / ; = l , 2 , . . .
2}
Jtn
Thus our signal has the complex Fourier series of equation 20.6 with
given by equation 20.12 and
To obtain the real Fourier series, from equation 20.8,

a = 2 R e ( c , J = sm n ,h , = 0
jTn
Substituting these coefficients into equation 20.3 yields the following trigonometric Fourier series
for the square wave o f Figure 20.3:
A 2A /
I
I
/ ( 0 = T *' c o s ( o ; o / ) - - c o s ( 3 a ;o O + 7 C o s ( 5 a ;o / ) - .. .
2
7T \
3
5
(20.13)
T W O PRO PERTIES OF THE FO URIER SERIES

After computing the Fourier series of a periodic signal y(^), it is straightforward to find the Fourier series
o f a related periodic signal g(t) whose plot is a translation o f the plot of/f). A translation o f a plot is
1093
a horizontal and/or vertical movement o f the plot without any rotation. A translation o f the waveform
in the vertical direction causes a change in the dc level and affects only the coefficients
d^y and Cq.
A translation o f the waveform in the horizontal direction causes a time shift that changes only the
angles 0,^ and has no effect on the amplitude
We state the relanonship formally as follows.
Translation property o f Fourier series: Let
be the coefficients o f the exponential
Fourier series o f a periodic functiony(f), and c be those for another periodic function g{t).
If^ f) is a translation o(J{t) consisting o f a dc-level increase
right by
and a time shift (delay) to the
then
g it)= p -r J)^ K
(20.14a)
^ 0= do = CQ + K = dQ + K = + K
/; = 1 , 2 , . . .
=1
= Z.C,,
, /? = l 2 , . . .
=[d -ncjQt^), n = 1 ,2 ,3 ,...
(2 0 .l4 b )
(20.14c)
(2 0 .l4 d )
(2 0 .l4 e )
The proof o f this property is straightforward and is left as an exercise. Note that equation 2 0 .1 4c
indicates that a time shift o f the signal affects only the phase angles o f the harmonics; the ampli
tudes o f all harmonics remain unchanged.
Exercise. Supposey(f) in Example 20.2 has A = 30 and suppose ^ f) =J{t) - 10 . Find the coeffi
c i e n t s a n d
o f the Fourier series o f ^ r).
AN SW I'R:
= 0.5^/,, = 3; all other coefficients arc unLh.mgcd, i.e.,
= sin n
because h.. = 0.
We will now use the translation property to obtain the Fourier series o f a square wave that is a
translation o f Figure 20.3.
EXA M PLE 2 0 .3
The square wave g{t) shown in Figure 20.4 is anti-symmetrical with respect to the origin. Find
the complex and real Fourier series for^^) given the Fourier series (equation 20.12) ofJ{t) depict
ed in Figure 20.3.
u rn
g(t)
>
'
A/2
w
-0.5T
-I
0.5T
-A/2
FIG URE 20.4 A square wave anti-symmetrical with respcct to the origin.
S olution
The curve
is a translation o f the waveform y(f) o f Figure 20.3. Specifically,
S ( 0 = / ((-/ ,/ ) + a: = /
By equations 20.14 and 20.13,
COS ncoQt
(
= cos ncoQf - n
\
2}
resulting in the desired Fourier series,
jr 1
2A
7Z
-----CO.S 3W(,/ 3
2}
?>7Z
+ - COS
5
5 jt
(20.15)
sin(wQ/) + sin(3wQ/)-i- jsln (5 w ()/ ) + ...
For the case o f a square wave signal, by choosing the time origin and dc level properly, the result
ant plot displays a special kind of symmetry that results in the disappearance o f all sine terms or
all cosine terms. The square waves o f Figures 20.3 and 20.4 are special cases o f the periodic func
tions amenable to such simplifications. The general case is given by the following statement.
Symmetry properties o f the Fourier series
(1) If a periodic functiony(/) is an even function, i.e.,
has only cosine terms and possibly a constant term.
then its Fourier series
(2) If a periodic functiony{r) is an odd function,
then its Fourier series
has only sine terms.

The plot o f an even function is symmetrical about the vertical axis. Examples o f even functions
include
cos(cl)/)
and the square wave o f Figure 20.3. The plot o f an odd function is anti-sym-
metrical about the vertical axis. Examples o f odd functions are sin(o)/) and the square wave o f
Figure 20.4. The proofs o f the symmetry properties are left as homework problems.
1095
Chapter 20 * Brief Introduction to Fourier Scries
Exercise. Suppose
in Figure 20.4 From Example 20.3 is shifted into q{t) = g { t - 0 .2 5 7 ). Find
the coefficients
and
AN SW ER:
o f the Fourier series o f q{t).
lA
(h) = (i{) = 0. and hir n ^ 0. d = a = -------sin
.111
-T\
n .h = 0 .
~)
To simplify the calculation o f the Fourier coefficients, we should attempt to relocate the time ori
gin or change the dc level so that the new function ^t) displays even or odd function symmetry.
This may not be possible for an arbitrary periodic signal. When it is possible, we will calculate
the Fourier coefficients o f the new function ^t), which has only cosine terms or sine terms, and
then use the translation propert)' to obtain the Fourier coefficients for the original function//) .
A waveform o f particular importance in signal analysis is the periodic rectangular signal shown in
Figure 20.5. The fundamental period is T, and the pulse width is fiT. The constant (5 is called
the duty cycle, usually expressed as a percentage o f T. The square wave o f Figure 20.3 is a spe
cial ca.se o f the rectangular wave o f Figure 20.5 with a 50% duty cycle.
EXA M PLE 2 0 .4
Find the trigonometric Fourier series for the rectangular waveform o f Figure 20.5.
S olution
T he procedure is almost identical to that used in Example 20,2 for a square wave. By inspection,
the average value
is
(20.16a)
To calculate the other Fourier coefficients, choose fg = -7 7 2 . Equation 2 0 .10 then yields
pT
fW
~E.
= sin(/7/3.T), /? = 1 ,2 ,...
Tin
( 2 0 . 16b)
10 9 6
From equation 20.7, rhe coefficients o f tiie trigonometric series are

^ /\
= 2 R e (c - ) = sin(/?^.T)
(20.16c)
jzn
b.. = 0
(2 0 .l6 d )
Specifying the coefficients completes tine determination ot the Fourier series, i.e.,
2A ^ (sin(nl^jr)\
) = p/\ 1------- y
(1,1
n= \
COS(nwt)
/I = l^
A very important conclusion about the rectangular wave can be drawn by examining equations
20.16: as the ratio o f the pulse width to the period becomes very small, the magnitudes o f the
fundamental and all harmonic components converge to twice the average (dc) values. To see this,
recall that sin(.v)/x approaches 1 as .v approaches 0. From equation 20.16c, we may rewrite
a
'n = sin(///i.;r)= 2ftA
jTn
It follows that
2fM as ft
as
nftji
0.
To give some concrete feel to this property, lable 20.2 gives the ratios o f
^'(ly
= and
âv
for w = 1, ..., 9 for the case o f ft = 0 .0 1 . Answers are rounded o ff to three digits after the decimal
point. Note that when the periodic rectangular signal is shifted vertically, the ratio
is affected, but the ratio
remains unchanged.
TABLE 20.2
A m p lit u d e s o f t h e F i r s t N in e H a r m o n i c s f o r t h e C a s e P = 0 .0 1
F
av
F
av
The constant
2.000
1.999
1.998
1.995
1.992
1.989
1.984
1.979
1.974
1.000
1.000
0.999
0.998
0 .996
0.994
0.992
0 .990
0 .987
property holds approximately true when the pulse width is a very small fraction
o f the period 7q. Even i f a waveform is not rectangular, if the pulse width is very small compared
109:
Chapter 20 * Brief Introduction to Fourier Series
to its period, then the nearly constant propcrrv' o f ~ iind ^ continues to hold, as long as the
I
pulse is o f a single polarit)'. For example, consider the periodic short pulse shown in Figure
In calculating the Fourier coefficients
gral, originally (/q,
+ 7 ),
f;;
o f this pulse using equation
are changed to
(-(^ 7 7 2 ,
^ 7 7 2 ).
As
2 0 .1 1 ,
2 0 .6 .
the limits o f the inte
approaches zero, the factor
in the integrand has a value very close to 1 in the new time interval, as long as n, the
harmonic order being considered, is not very high. Therefore, for pulses o f narrow width we have
(xT
aT
(20.17a)
rhus in terms of the
coefficients, from equation 20.8, again for pulses o f narrow width we
have
(20.17b)
This result is pertinent to the approximate analysis o f a rectifier circuit covered in other texts or
in the second edition o f this text.
To this point, we have calculated the Fourier coefficients only for some very simple periodic sig
nals. The evaluation o f the integral in equation 20.11 becomes much more involved when the
signal j{t) is not rectangular. Fortunately, many mathematical and engineering handbooks now
include comprehensive tables o f Fourier scries. From a utility point o f view, one may use these
tables much the same as one uses a table o f integrals or a table o f Laplace transforms. In effect,
the need to carrj' out the integration in equation 20.11 is not compelling in practice.
In many applications, it is important to know the average power o f a (periodic) signal and the
magnitude o f its various harmonics. From equation 20.5, the effective value o f the dc compo
nent is |<^q|, and those for the fundamental and various harmonics are d^j! yfl , /; = 1 , 2 , . . . . It is
easy to show that the effective value, or the rms value, o f/ r) is
(20.18a)
where the d- coefficients are from the Fourier series o f equation 20.4. If fit) (current or voltage) drives
a 1 ohm resistor, then the average power absorbed by the resistor is
êff Hence we
say that the average power o f a periodic signal y(/) represented by a Fourier series is given by
1098
r~\
1
P av eJ{t)= F }ff= dl+ -di + - d l - ^ ..=

2
k|
(20.18b)
n=-oo
The relationships indicated in equations 20.18 are often termed Parsevals theorem.
The information on the phase angle is important when one wishes to construct the time domain
response in steady state. For the time domain problem, the Fourier series method yields only an
approximate solution, because one can only sum a finite number of terms in the series.
C on v erg en ce o f th e F o u r ie r S eries
Convergence of the Fourier series is an intricate mathematical problem, the details of which are
beyond the scope of this text. On the other hand, it is important to be aware of the ways in which
the Fourier series may or may not converge to a giveny(/). Our discussion is not comprehensive,
but is adequate for our present purposes.
To begin, we define a partial sum of terms of the complex Fourier series of a function ^(/) as
sn )=
k= -N
From our experience thus far, 5yy(f) must in some way approximatey(r). The difference between
j{t) and its approximation, 5yy(r), is defined as the error
ât( 0 = / ( 0 - 5 ^ ( 0
One way to get a handle on how well Sj^t) approximates

is to use the so-called integral
squared-error magnitude over one period, [/gi
T\, defined as
r\
e ^ (l) dt
This is often called the energy in the error signal, as energy is proportional to the integral of the
squared magnitude of a function. It turns out that for ftmaions having a Fourier series, the choice
of the Fourier coefficients minimizes E[^ for each N. Further, for such fiinctions,
0 as
00, i.e., the energy in the error goes to zero as JV becomes large. This does not mean that at each r,
fij) and its Fourier series are equal; it merely means that the energy in the error goes to zero.
Continuous and piecewise continuous periodic functions have Fourier series representations. A
piecewise continuous fiinction, such as a square wave, is a function that (1) has a finite number
of discontinuities over each period but is otherwise continuous, and (2) has well-defined rightand left-hand limits as the function approaches a point of discontinuity. For piecewise continu
ous functions, it turns out that the Fourier series converges to a value halfway between the values
of the left- and right-hand limits of the function around the point of discontinuity. Even so,
-* 0 as TV- 00 for piecewise continuous functions.
There are many functions that are not piecewise continuous and yet have a Fourier series. A set
of conditions that is sufficient, but not necessary, for a function to have a Fourier series repre
sentation is called the Dirichlet conditions.
r\
1099
D irichlet conditions
Condition 1. Over any period, [fg, tQ+T\,J{t) must have the property that
In the language o f matliematics, this means that J{t) is absolutely integrable. The conse
quence o f this property is that each o f the Fourier coefficients
is finite, i.e., the
exist.
Condition 2. Over any period o f the signal, there must be only a finite number o f min
ima and maxima. In other vvôrds, functions like sin(l/^) are excluded. In the language o f
mathematics, a function that has only a finite number o f maxima and minima over any
finite interval is said to be o f hounded variation.
Condition 3. Over any period, y(r) can have only a finite number o f discontinuities.
As mentioned, at points o f discontinuity, the Fourier series will converge to a value midway between
the left- and right-hand values o f the function next to the discontinuity. There may be other differ
ences as well. Despite these differences, the energy between the function
representation is zero; i.e.,
and its Fourier series
with N approaching oo, goes to zero. Thus, for all practical purpos
es, the functions are identical. This practical equivalence allows us to analyze how a circuit responds
to a s i g n a l b y analyzing how the circuit responds to each o f its Fourier series components.
3. ADDITIONAL PROPERTIES AND COMPUTATIONAL

SHORTCUTS FOR THE FOURIER SERIES REPRESENTATION
If a periodic function y(f) is known only at some sampled points, e.g., by measurements, then its
Fourier coefficients must be calculated by the use o f numerical methods. On the other hand, if
j{t) has an analytic expression, then its Fourier coefficients can often be calculated from equation
20.11, The properties discussed below are o f great value in simplifying the calculation o f Fourier
coefficients. Their proofs are fairly straightforward and arc left as homework problems.
The linearity property: Letyj(/) and f^{t) be periodic w'ith fundamental period T. \^j{t)
= KJ'^{t) + K^fyU), then the Fourier coefficients o^j{t) may be expressed in terms o f those
of/j(f) and ^ (r) according to the following formulas:
(20.19a)
(20.19b)
and
(20.19c)
In general,
unless the angle 0^ is the same for all n.
1 100
n s
This property allows us to easily obtain the Fourier series of a sum of periodic signals when the
Fourier series of the individual signals is already known.
DEFINITION
A periodic function y(/) is said to be half-wave symmetric if
-0.57) = -j{t) for all t
In words, y(r) is half-wave symmetric if a half-period shift of the plot combined with a flip about
the horizontal axis results in the identical function y(/). Some simple examples of half-wave sym
metric functions are sin(ft)r), cos{(Ot), and the square wave of Figure 20.4,
The half-wave symmetry property: A half-wave symmetric periodic function f j ) con
tains only odd harmonics.
The waveform of Figure 20.4 illustrates this property.
The dertvadve/integration proper^.
denote the ^
Ik)
functiony(f). Then the Fourier coefficients, Q s o f
ci*> = (jmoo f c for all n
c =
derivative of a periodic
(k)
(t) of satisfy
20a)
c<*)
Except for the constant term, all terms

derive from those ofy(r) by differentiating k times;
conversely, all terms oij{t) derive from those of/^^(/) by ^-fold (indefinite) integration. The exclu
sion of the constant term in the relationships poses no difficulty, since the constant is simply the
average value of the periodic function.
Again, these properties help to simplify the calculation of Fourier coefficients. In fact, these prop
erties make it possible to compute the Fourier series for the waveforms given later in Table 20.3
without carrying out the integration of equation 20.11. Achieving this, however, depends on first
finding the Fourier series of a waveform for which computing by equation 20.11 is extremely
easy. We will illustrate several such calculations in Examples 20.5 through 20.8. During this devel
opment, the following trigonometric identities will prove useful;
/ ^ y\
/ _ y'
cos(jc)-cos(y) = - 2 s i n ( ^ ^ - j sin (^ -Y -j
(20.21a)
cos(x)cos()') = 0.5 [ cos(x + j ) + co s(;f-j)]
(20.21b)
!01
E X A M PLE 2 0 . 5 . Find rhe Fourier series for rhe periodic impulse trainy(^) shown in Figure 20.7.
jk
A
(A)
-21
A
(A)
k.
(A)
A
(A)
(A)
-T
2T
FIG U RE 20.7 A periodic impulse train.

S olution
Using the sifting property' o f an impulse function together with equation 20 .1 0 yields
(20.22a)
T
for all n. Hence
Equation 20.22b states that, for a periodic impulse train, all harmonic components have magni
tude equal to twice the average value. This is the limiting case o f the short pulse propert)- stated
in section 2. The next example uses the derivative and integral properties o f the Fourier series to
develop an alternative derivation o f the Fourier series for a periodic rectangular pulse train, derived
earlier by the use o f equation 20.11.
E X A M PLE 2 0 .6 . Find the Fourier series for the periodic rectangular pulses^(f) shown in Figure
20.8a. (This corresponds to item 2 in Table 20.3.)
S olution
Figures 20.8a and b show ^(r) and its derivative,^ (t). The latter may be written as the sum o f
two shifted impulse trains:
{
f n( t ) = f6 t +
\
^ r
-fa
t-
/(5T
T \\
\ ~ r;
(2 0 .2 3 )
102
(A)
(A)
-T
T
2
i(-A )
(A)
-f
^
2
T
,,(-A )
>
(-A)
(b)
FIG U RE 20.8 (a) Periodic rectangular pulse crain and (b) its derivative. The parenthetical values,
{A) and {-A), next to each impulse in part (b) denote the weight of that impulse, i.e., its area.
Using the time shift property (translation in the horizontal direction), together with equations
20.22b and 20.21a,
f'p ( 0 = Y,
"
= I
2A
- cos{na)Qt - n^jv))
(20.24)
4/\
(sin(/7o;oOsin(/3;r))
^
Applying the derivative/integration property to equation 20.24 yields

00
(2 0 .2 5 )
T h e result agrees, o f course, with equation 2 0 .1 6 .
1103
E XA M PLE 2 0 .7 . Find the Fourier series for the half-wave rectified sine wave fhsi^) shown in
Figure 20.9a (item 10 in Table 20.3).
A f.(t)
(0
FIGURE 20.9 (a) Half-wave rectified sine wave as the product of two functions: (b) a cosine wave
and (c) a square wave.
S olution
The periodic fiinction j{t) o f Figure 20.9a may be viewed as the product o f the sinusoidal wave
y4cos((0Qr) and the square wave^(r), shown in Figures 20.9b and c, respectively. Using the Fourier
series for^(^) given by equation 20.13, we have
//i.v(0= [ ^ c o s ( w o O lA ( 0
M
= y4cos(o;o0
9 /
1
1
\1
~ cos(ojo^) cos(3w()/) + cos(5o;o/) - ...
;r ^ 3
5
(20.26)
104
Applying equation 20.21b to each product in equation 20.26 yields

fh s
( 0 = cos( wq/) + |cos(2o;()r) + cos(Oo;o/) - -cos(4w()/') + cos(2woO
7z I
A
lA
= - + - cos(w()/) +
IT
- c o s ( 2 ft ;,)0 - cos(4coqi)
Id
JT \ J
+ cos(6w()/) + ... + - ^------ cos(2/?W()/) + ...

35
4/;"- _ 1
A
= - + - co.s((OoO + 2
^ 2
JT
Note that the fundamental component is present in
T T -co.s( 2 mu/
4 ,r _ 1
(20.27)
and the remaining terms are all even
harmonics.
E X A M PLE 2 0 ,8 . Find the Fourier series for the full-wave rectified sine wavej^(r) shown in Figure
20.10a (item 9 in 'I'able 20.3).
lOS
Chapter 20 Ekicf Introduction to Fourier Series
> t
FIGURK 20.10 (a) Full-wavc rcctificd sine wave as the sum of the two signals in (b) and (c).
S olution
One approach is to apply the same technique as in Example 20.7. Specifically,
f^ {t)= A cos{io^ t)[lfi,t)-\ ]

where f^{t) is the square wave o f Figure 20.9c. Alternatively, to avoid repeating all the arithmetic,
note that
is the sum o f the two waveforms shown in Figures 20.10b and c; i.e.,
(20.28)
Substituting equation 20.27, the Fourier series

Fourier series:
lA
into equation 20.28 yields the desired
4/4
(20.29)
4 ;r -I
The derivative propert}' is particularly useful for tackling periodic piecewise linear waveforms.
Piecewise linear waveforms consist o f straight-line segments. From differentiating once, or at most
twice, impulses appear. The integration given by equation 20.11 is trivial if the integrand contains
a shifted impulse function. This fact, together with the derivative and integral properties, reduce
the task o f calculating the Fourier coefficients for piecewise linear waveforms to some complex
number arithmetic. These examples and several other commonly encountered periodic waveforms
have Fourier series as given in Table 20.3. Engineering and mathematical handbooks contain
much more comprehensive tables. O f course, when a waveform does not appear in a table, the
Fourier coefficients must be computed manually or numerically.
106
TABLE 20.3
107
6. Clipped sawtooth wave

^0 =
= 0-5a/\
d n = 2~^ sin ^ (nojr) + naji[ncm cm n~
sin(2//o;r))
0.5
: too complicated for inclusion
7. Asymmetrical triangular wave

^/o = F ,= 0 .5 A
d,i = ---------------^ -:rsin ( a jr)

a {\ -a )T r n ~
: too complicated for inclusion
8. Symmetrical trapezoidal wave

0.5ao = Fay =^{a + P)A
n (a + P)ji
b= 0
9. Full-wave rectified sine wave
Tq = 0.5T,
f(t) =
(Oq =
2(0
^ ^ cos(2(o?)
3
4A
K
cos(4w0 ^
15
cos(6(o/)
35
(-l)"^'cos(2wO
7
4 fr-i
"*
10. Half-wave rectified sine wave
Jl
/ (/ ) = - + - c o s (w o / ) +
n+1
v H L _
^ k
ave=A/n
rms=A/2
4n^ - 1
COS (inojQt)
108
4. SUMMARY
Given that many mathematical and engineering handbooks have extensive lists o f the Fourier
series o f common signals, this chapter has taken a practical approach to the calculation o f the
Fourier series and its application to circuit analysis. The idea is to use tables such as Table 20.3 in
the same way engineers have come to use integral tables. The keys to using such tables for the com
putation o f the Fourier series o f a waveform are the various properties that allow one to convert a
known series into one that fits a new waveform. The idea here is to express the new waveform as
a translation of, a linear combination of, a ^-fold derivative of, or a ^-fold integral o f signals with
known Fourier series as in Table 20.3, or any mixture o f these operations. The Fourier coefficients
o f the new signal can then be expressed in terms o f the Fourier coefficients o f signals with known
Fourier series.
Knowledge o f the Fourier coefficients o f a signal such as the output o f an audio amplifier allows
one to investigate phenomena including the distortion introduced by the amplifier. In the case o f
a dc power supply, such knowledge allows us to characterize the degree o f unwanted ripple in the
output o f a rectifier circuit. In addition, the application o f Fourier series plays an important role
in the computation o f steady-state circuit responses to periodic input signals.

Average value o f periodic/(^): the term
in equation 20.4; also referred to as the dc compo
nent ofy(r).
Derivative/integration property: Let/^^H^) denote the ^ h derivative o f a periodic function y(^).
Then the Fourier coefficients o f the ^ h derivative,
and conversely,
satisf}'
= (/>zoJq)^c^^ for all
J{t) is the /th integral of/^^^(^), then
for all n except
0.
D uty cycle: for a rectangular signal having fundamental period T as illustrated in Figure 20.5, the
duty cycle is the constant (3 that determines the pulse width ftT.
Effective value (rms value);
|/l=-cc
coefficients are from the Fourier series o f equation 20.4 and the
are from equation
20.6b.
Even function: y(/) =
The plot o f an even function is symmetrical about the vertical axis.
Exponential (complex) Fourier series; decomposition

as given in equations 20.6.
into a sum of complex exponentials
Fundamental com ponent (first harm onic): the first term under the summation sign in equa
tion 20.4,
cos(o)q^ + 0 j), having amplitude d^ and phase angle 0 j.
Fundamental frequency (in hertz): /o = . where T^^ is the fundamental period. Note:
_
2:7r .
^0
Wq = 2;r/() = is the fundamental angular frequency in rad/sec.
Fundamental period: the smallest positive real number
for w h i c h + Tq) =J{t).
1109
Half-wave symmetria refers to a periodic function/(/) that satisfies

0.57) = -fJ)Half-wave symmetry property: A half-wave symmetric periodic function//) contains only odd
harmonics.
Lineari^ property: let^(z) and^(/) be periodic with fundamental period T.
= ATj^(/) +
A!^(r), then the Fourier coefficients o^j{t) may be expressed in terms of those
and
fj{t) according to the formulas c^ =
+ K2C2,
= ^\^\n
In general,
the angle 0^ is the same for all n.
Odd function:/r) =
Periodic signal, y(f): A signal whose waveform repeats every T seconds. Mathematically, for some
7> 0, and all
+ 7)
where Tis the period of the signal.
Symmetry properties of Fourier series: (1) If a periodic function ^(/) is an even function, then
its Fourier series has only cosine terms and possibly a constant term. (2) If a periodic
function y(/) is an odd function, then its Fourier series has only sine terms.
Translation of a plot: horizontal and/or vertical movement of the plot of a function without any
rotation.
Translation proper^ of Fourier series: if fit) is a translation of J{t) consisting of a dc level
increase K and a time shift (delay) to the right by then ^/) = j { t +K. See equa
tion 20.14.
Trigonometric Fourier series: representation of a periodic signal f j ) in terms of sines and
cosines, as given in equations 20.3 or 20.4.
10
ANSWF^R: (a)
^ ^ r o b le m s
= - 0 .7 2 1 3
0.0795
FOURIER SERIES
COEFFICIENTS BY
INTEGRATION
cos(4.t /- 8 6 .8 4 ')
and write the Fourier series in the form o f

equations 20.3.
-0.5 T
f(t)
0.5 T
(1)
4. Using trig identities, compute the Fourier

series coefficients
-2
(b)
j{t) = cos(4r) sin(2f)

j{t) = sin^(4/^) cos^(8/)
(c)
/ f) = [2 + 1 .5 sin (5 0 0 r)
(a)
A^ f(t)
(-1)
>r
and
for the func
tions
(a)
(-1)
>r
Figure P20.3
(1)
+ ...
>^ f(t)
out the integration given by equation 20.10
-2 -1
1
+ / 2 ; / . t ) ; (b) fit)
P20.3.
impulse trains shown in Figure P20.1. Carry
-3
3. Repeat Problem 2 for the signal o f Figure
1. Find the Fourier series for the periodic
(1)
= 0 .5 /(Io g ^ .
+ 0 . 1 5 8 c o s ( 2 . t / - 8 3 . 7 ' ) +
1 2
(-1)
>r
>
(b)
Figure P20.1
- 2 cos(2000r)]cos(10^V)
FOURIER SERIES USING

PROPERTIES W ITHOUT
INTEGRATION
AN SW ER: (a) 0.5 + cos(.7r) + cos(2rr/) +
5. Construct the Fourier series o f item 5 in
cos(3-t/) + cos (4.t;) + ...: (b) - 0 .5 + cos(.Tr) -
co^{2m) + coa{5m) - cos (4m) + cos(5-t/) ...
Table 20.3;, i.e., find the Fourier series o f the

sawtooth waveform in Figure P20.5. Flint:
2. For the periodic function J{t) shown in
20.22b for/^(r).
Note that f{ t ) = A -
Use equation
Figure P20.2, T = 1 second and a - log^ =

0.693.
(a)
Find the coefficient
by carrying out
the integration given by equation

20.6b.
(b)
Write the first three terms o f the

Fourier scries (see equation 20.4).
f(t)
e**
6. Find the Fourier coefficients
and 6^ o f the
periodic function shown in Figure P20.6. Flint:

Make use o f the result o f Problem 5.
Figure P20.2
1111
Chapter 20 Brief Introduction to F-'ouricr Series
form o f item 2 with proper height and (2) the

waveform o f item 5 shifted down by a suitable
amount.
-1
Figure P20.6
-t
ANSWF.R: /(/) = 0.75 + 0.159 sin(2Tut)

0.0795sin(47rt) + Figure P20.9
7. Construct item 3 of Table 20.3; i.e., com

pute the Fourier scries for the triangular wave
shown in Figure P20.7. Hint: f'{t) is a square
10.
w'ave whose Fourier series was calculated
shown in Figure P 20.10. Hint: Make use o f the
Find the Fourier series for the waveform
before.
result o f Problem 9.
>
Figure P20.10
8. Consider the isosceles triangular wave shown
in Figure P20.8, which is item 4 ofTable 20.3.
11. Consider the asymmetrical triangular wave
Compute the Fourier series o f this waveform
shown in Figure P 2 0 .l l, which is item 7 o f
utilizing the results for items 2 and 3 ofTable
Fable 20.3. Let 7 = 1 and u = 0.25. Find the
20.3. Hint:y(^) is the product o f (1) the wave
Fourier coefficients
form o f item 2 with proper height and (2) the
Differentiating J{t) twice results in periodic
waveform o f item 3 shifted down by a suitable
impulse trains.
amount.
and d-^. Hint:
f(t)
> t
Figure P20.8
12. Find the Fourier series coefficients Tq, cp
9. Consider the clipped sawtooth wave shown
in Figure P20.9, which is item 6 ofTable 20.3.
and C-, for the periodic waveform shown in

Figure P20.12. Hint: The period T = 4, and
Compute the Fourier scries o f this waveform
you should express J{t) as the sum o f several
utilizing the results o f items 2 and 5 ofT able
shifted square waves.
20.3. Hint:y(r) is the product o f (1) the wave
1 112
oj= 10^ rad/secand

10-^ rad/scc. For
the frequency range990 x 10-^ < w< 1010
X 10-^ rad/sec, assume that the band-p;iss

amplifier has the following ideal magni
tude and phase characteristics;
\H{joj)\ = 10
and
-2
Figure P 20.12
/LH ( jco) = -
5000
Express
(a )
SINUSOIDAL STEADY-STATE
ANALYSIS WITH MULITFREQUENCY INPUTS
X 45^'
as the sum o f all o f its
frequency components.
13. The input to the low-pass filter circuit
(b)
Express
(c )
frequency components.
If
is expressed in the form
shown in Figure P 20.13 is
as the sum o f all o f its
^ cosCwq/) V, show that
g{t) = \Oj{t - t j , where

v,(/) = 200 -h 200n/^cos(377/) + 6 0 V 2 cos(3 x 377/ + 30'^ ;
sec. The interpretation o f this result is

that the envelope o f
+ 8 0 V 2 co s(5 x 377/ + 5 0 ^ )V
= 78.54 a
is an ampli
fied and delayed version ol the enve

(a)
Find
(b)
Find the rms value o f
at steady state.
lope o f
and the
the amplification is 10
and time delay is 78 .5 4 [.isec.
average power absorbed by the 10 kQ

resistor.
jy r \
v(t)
10H
4 pF
Fig P20.14
Figure P20.13
ANS\X'ER; (a)
ANSW ER; (a)
0.1 cos((<o
+
r.... (M = 200 + >/2
42.6cos((o/ - 1 7 5 .4 ") +
+ 0.1 cos(((/^ +
2 cos((oj. -
1.192cos(3n;/ - 1 4 S .7 " )+ 0 .5 6 COM5o>/ - 129.2^)
= 0.2 cos(((/;. -
10
Jt) +
+ 2 cos(a;^./) + 0 .2 c o s ((o ;.
V; (b)
+ 4.5) + cos((;. -
y 9) + 20 cos(oj^j') + 2 cos((o>. +
=
+
-
4.5") + cos((o;,+ 2 r o J / - 9 " ) V
with u) - 377 rad/sec; lb) 204.5 V. 4.18 W
15 . Consider the circuit o f Figure P20.15, in

14. Consider the circuit o f Figure P20.14. The
which C = I F (initially relaxed), R = 1.443 =
input to this ideal band-pass amplifier is an
\Un{l) n , and
amplitude-modulated waveform given by
currents Q b[tnT), T - 1 sec, Q = 1 coulomb,
=J{t)x cos(cu^r) V, where
and ;; = 0 , 1 , 2 ,
(a)
Show that the response due to the first
J{t)
2[1 - 0.2 cos(w,/) + 0.1 cos(2(t;,/)]
is a sequence o f impulse
hnpitlse current alone is
13
^-0.693r ^ (0.5)^ V for 0 < ^ < 00; v ^ Jt)
(b)
'w '
(c)
(d)
(e)
starts with 1 V at ^ = 0 and then is

halved for every second of elapsed
time.
Making use of linearity and the result
of part (a), show that for 1 < r < 2,
v^Jt) = 1.5(0.5)<'- '> V. Note that
only the first and second impulses
have an effect on this time interval.
Show that for < /< + 1,
= [2
-(0.5 )"](0 .5 )('-^V.
For very large < r < + 1, show that
v^Jt) = 2(0.5)(' V and that the
waveforms for the time intervals (, n
+ 1) and ( + 1, + 2) are identical
except for a time delay of n. This
means
has reached the steady
state.
Show that for arbitrary values of /?, C,
Q, and T the output in steady state is
'ôut ( 0 =
Ôllt ( 0 =
(b)
(c)
----- u
(a)
,I
ijt )
(Q)
(Q)
-T
1r
C = 1 F, Q = 1 coulomb, and T = 1 sec.

Show that the steady-state response for
the first cycle is
=
2(0.5)^ V, for 0 < r < 1. Sketch the
waveform for the first cycle.
Using the transfer function
1
hnis)
2T
Figure P20.15
Consider again the circuit of Figure
P20.15. If the impulse current train has been
and the Fourier series for

equa
tion 20.22, find the Fourier series for
Write the answer to include up
to the third harmonic.
Find the average power (approximate
value) dissipated the resistor. Since
only a finite number of the terms in
the infinite series are included in the
calculation, the answer is only approx
imate. What is the error when only
harmonics up to the third are consid
ered? To determine this error, compute
the exact average power using the
result of part (b).
CHECK: (c) v^Jt) = 1.4427 + 0.3164 cos(2;rf

- 87.7047) + 0.1589 cos(4;rf - 86.8428) +
0.1060 cos(6;w - 87.88) V; (d) 2.109 watts,
exact answer is 2.164 watts, error = 2.53%
17. The LC resonant circuit in Figure P20.17a
is initially relaxed and the impulse train
applied for a long time (theoretically since t =

- oo), that is, = - 00,..., -2, -1,0, 1,2, ...
(a)
C5 H----
(Q)
(b)
16.
For this and the remaining parts, use

the special set of values
H {s) =
(d)
VJt)
RC
_Z_
= ! = I.4 4 2 Q
M 2)
\ - e RC
O
+
C
\ -e
RC
for n T<t <{ n - \)T.
Show that for the time interval (0, T)

/i=0
14
shown in Figure P20.17b is applied. The natu

ral frequency o f the tank circuit is f.. = ------\-----
2 h /lC
Tq - = 2vtVZc .
and the period is

(a)
shown in P 2 0 .I8 b is applied. This problem

(and circuit) differs from the previous one only
in the presence o f the resistance R. When the
value o f R is large enough, the impulse response
Jo
is a damped sinusoid. The purpose o f this prob
Show that if the frequency o f the
lem is to show that if the damping effect o f i s
impulse train, / = , is the same as

7'
,
the resonant frequency / q =
then
slight and the input impulse train synchro

nizes with the damped sinusoids o f the
impulse response, then the output voltage mag
In y fL C ' nitude, although finite, can reach a ver\' large
/y(r) grows without bound.
value.
Specifically, show that the inductor
(a)
current in the time interval {n - \ ) T <
o f poles at
t < 7i T \s given by i^{t) = n x

Show that iff=fQlm for any integer m,
CJand
the inductor current again grows

unbounded.
(c)
interval o f length
Show that in general if T = (2w +
Wq =
2ji
Tj = ---- ) differs
(X)j
from its adjacent cycle by only a mul

tiplicative constant a =
2m + 1
then /^(/) is periodic. Sketch t\vo
then the
(ii) the waveform o f every cycle (time
two cycles o f /^(r).

1)(0.57'q) or, equivalently,
2/o
CJ
in terms o f R, L, and C.
Note that (i)w j
Show that if/ = 2/q, then i^{t) is peri

odic and remains bounded. Sketch
(d)
^= -
impulse response has the form h{t) =

(,-ot cos{ioj) + 5sin(w^)]//(f). Find
sin(w^r).
(b)
\n6H{s) = - ^ ---- .If//(^) has a pair
(b)
Show that if the input impulse train syn

chronizes with the damped oscillations,
meaning that T = T j = 2nUo^ then
cycles of
the output due to the impulse train

o f Figure P 20.18b is
for (;/ - 1)
(c)
(a)
if the impulse current has been applied

for a long time (theoretically from t =
>I
>
(Q)
(Q)
-T
< r < ;/
Deduce from the result o f part (b) that
-oc), then the steady-state response
during the first cycle is given by
(Q)
2T
= 7 ^ / K O for 0 < r < 7;y

1- a
which shows that as a comes very
(b)
close to 1 (meaning very little damp
Figure P 20.17
18. The LC resonant circuit in Figure P 2 0 .l8 a
is initially relaxed and the impulse train
iin(0 = Q ^ H r - n T )
/i=(i
ing), the magnitude o f the output can

reach a very large value.
(d)
If ^ = 0 .7213 a
C = 1 F, and L =
0.025 H, in which case a = 0.5, find
H{s) and the impulse response h{t).
1115
Further, if Q = 1 coulomb and T = 1

sec, show that v^JJ) in steady state is
given by (consider item 18 of Table
(a)
12.1)
'.,W = 2<'-3I<[cos(2 w ) -
1/
I V
''max - V*mm
ôul,mm ~ *min
o .5 7
0.1103 sin(2;)] V for 0 < r < 1
L,
\+ e
Plot this waveform using MATLAB or

its equivalent. Justify that this is indeed
the steady-state value by showing
v<.,(0-^)-v,(D = | = l V.
(e)
I f /? = 9.747 a , C= 1 F, I = 0.02533
H, Q = 1 coulomb and T = \ sec, in
which case a = 0.95, find the steadystate response for 0 < ^< 1 sec.
If the input has been applied for a long

time, then the circuit reaches steady
state and
as shown in Figure
P20.19c. Using this fact and equation
8.19a, show that in steady state.
V
.
= V^max ôut,max
(b)
(c)
u (t)
V
f
' nt/TT - V' tff
0.5T
\+ g RC
In Example 20.1, find the exact values

of
' '"g
results of part (a). Then plot
for
one complete cycle using MATLAB or
its equivalent.
In the exercise following Example
20.1, four frequency components of
the Fourier series are used to approxi
mate the input, and it is found that
ôut.min ~ 12-235 V and ôuumax ~
82.013 V. What are the percentage
errors of these approximate answers?
(a)
U t)
ik
(C8
(Q)
-T
(Q)
2T
(a)
(b)
.v jt)
V
Figure P20.18
CHECK: (d) //(5) =
's-^
(e)/ / (5 ) = _ ----------
fflin
r + 1.3863^ + 39.96
s
r +0.102586j + 34.481049
and
= 20tf*5^'[cos(2;) 0.0082 sin(2;r^)] V for 0 < ^< 1
19. Consider the first-order RC circuit of Figure
P20.19a, where the input v-JJ) is a square wave
with period T, as shown in Figure P20.19b.
From the analysis methods discussed in Chapter
8, the output
consists of exponential rises
and decays, as illustrated in Figure P20.19c.
...
-0.5T
0.5T
0.5T
(b)
(c)
Figure P20.19
1.5T
16
ANS\X^:RS:(h) 1 1.235V. 83.013 V, (c) H.9 % .

-1 .2 %

which C = I F, 7? = \Un{2) = 1.443
T=A
sec, I/, = 15 V.
= 0.25, v^Q-) =
20. The method o f Problem 19 can be extend
V, and
ed to computation o f the steady-state response
as shown in Figure P20.22b.
o f any first-order linear network to a square
(a)
= 1
is a sequence o f rectangular pulses

Show that for the first cycle, 0 < r < 4
wave input. Let the transfer function o f the
sec, the complete response is V(^t) =
first-order linear network be
[15 - 1 5 (0 .5 )1 + (0 .5 )r = 15 14(0.5)^ V for 0 < / < 1 sec and V(^t) =
H(s) =
/ [Output(/)]
H{0)
/ [InputC/)]
1+T5
8(0.5)^ V, for 1 < t < A sec. Sketch the
v^t) waveform for the first c)'cle 0 <

r < 4 sec.
and the input be the square wave of Figure

P 20.19b. Show that in steady state,
(b)
Show that v^JA') =

= 1 V.
Output(/),, = H{0)
V
* max - V^min
(c)
cycle, n T < t < {n + U T is identical to
l+ e ^
that o f the first cycle except for a time
and
0
Use the result of part (b) to verify that

the V(\t) waveform for any subsequent
0.57
shift, i.e., the first-cycle waveform
''max - V* mm
utput(0 ,av =
found above is the steady-state wave

form
(d)
Using the transfer function
21. Consider the leaky integrator circuit o f

Figure P20.21. The input
is a 1 kHz
H{s) =
square wave with zero mean and 2 V peak to
RC
s+
peak. The circuit has reached steady state.
RC
(a)
Find the transfer function H(s).
and the Fourier series for
(b)
Find the maximum and minimum
o f la b le 20.3 , find the magnitudes o f

the harmonics o f the Fourier series for
values o f
using the results o f
Vo,^,{t), up to the third harmonic.
Problem 20.
(c)
item 2
Use MATLAB to plot the waveform o f

for one complete cycle.
(e)
Find the average power (approximate

value) dissipated in the resistor. Since
only a finite number o f the infinite
50 kO
series is included in the calculation,

the answer is only approximate. What
is the error when only harmonics up
to the third are considered? To deter
mine this error, compute the exact
average power using the result o f part
(a).
Figure P20.21
A.\S\X ER,S: (a)
H (s) =
-S
i()()0.v +
: ( b ) 0 .2 4 5 V
R.
'. W
R.
v jt )
' ' 6
O '
Figure P20.23
(a)
Avjt)
pT
> t
T+pT
(b)
Figure P20.22
23. Consider the second-order circuit of Figure
P20.23, in which = Ifl, /?2 = 8 Q, /?3 = 6 2,
Cj = 0.25 F, and
= 0.125 E
(a) Derive the transfer function
16
3
r + 7J + 10
If the input voltage is a square wave
shown in Figure 20.23 with peak-topeak voltage 18 V and T = \ sec, find
the Fourier series for
and list the
first four components.
(c) Find the approximate average power
delivered to R2 considering only the
first four components.
(d) Assuming the input has been applied
for a long time so that the circuit is in
steady state,
find the expression for
during
the time interval (0, 7). Use MATLAB to plot
the waveform. Hint: Expand H{s) into par
tial fractions. Represent the system as the paral
lel connection of two first-order subsystems.
Then use the result of Problem 17 to solve each
subsystem. Finally, combine the solutions in
time domain to obtain the expression for
(b)
Vw/
------ o
//(double slash) notation, 75, 90

ABCD parameters.
Transmission (/-)
parameters
Abscissa of (absolute) convergence, 556, 591
AC (alternating current)
adaptors, 883
defined, 40
Active ban-pass circuit, 827-828, 861
Active element, 82, 177
Active networks, Thevenin and Norton
equivalent circuits for, 241-245
Active realization, 1045-1051, 1071
Active region, 171
Admittance, 481
of capacitor, 608, 658
defined, 606, 607, 658
of inductor, 658
of the inductor, 608
Laplace transform analysis circuit appli
cations, 605-609, 659-662
Laplace transform analysis circuit appli
cations, manipulation, 609-617
phasor admittance, 454, 455^61
^'-parameters, 1010
Admittance (y-) parameters, 961, 968-973
Algebra, 781, 797
Ammeter, 10
.\mpere, 34, 29, 40
Amplifier. See Operational amplifier (op amp)
Amplifier circuit, 117-118, 120, 121, 136
small-signal low-frcquenc)' equivalent, 117
Amplifiers, 959-960
Analytic continuation, 557, 591
Angular frequency, 384, 417, 433
Apparent power, 510-515, 531
Approximation problem, 1035, 1037-1042,
1071
Armature winding, 736
Associativit)', 766, 797
As)-mptotes, 732, 740
Attenuation (dB), 1033, 1072
Automobile batter)', 14-15, 16, 18, 27
Auxiliar)-voltage variables, 136
Available power, 528
Average pow-er, 531

complex power and SSS, 510-515
SSS power calculation and, 500-505,
533, 534-536
Average value, 10, 1089
Average value of p e r i o d i c 1089, 1108
Balanced terminal, 157
Band-pass circuit, 470, 481, 811-882
defined, 861
general structure of, with one pair of
complex poles, 851-860
ideal band-pass filter, 814-816
with practical components, 830-838,
868-872
resonance phenomenon and resonant
circuits, 839-851
resonant circuits with applications,
872-876
touch-tone phones and, 811-812
transfer function problems, 863-868
transfer functions with no zeros/single
zero off the origin, 876-879
Band-pass filter design, 1062-1069, 1072
Band-reject circuit, 471, 481
Bandwidth
with respect
1062, 1072
3 dB bandwidth, 815-816, 861
Batter)', 12, 14-15, 16, 40
automobile, 14-15, 16, 18, 27
ideal, 26, 30, 84
Bell Systems TechnicalJournal, 1069
Bode plots, 730-736, 760-761
bode magnitude plot, 731-733, 740
bode phase plot, 731, 734-735, 740
Bounded current/voltage, 278, 306, 555,
556, 690, 740
BP to LP frequency transformation (pr?),
1072
Branch, 53-54, 89
Branch current, 107, 110, 128
Branch voltage, 107
Breakdown voltage, 376
Breakpoint, 733, 740
120
Index
Buffer, 164, 177

Bulbs, 19, 30, 34-35, 37
Butterworth, Stephen, 1034
Buctenvorth filter, 634-635, 856-858,
1045-1051. See also Filtering
active reaIi/;uion ol low-pass Butterworth
filters, 1045-1051
Butterworth LP design algorithm based
on S&K circuit, 1046-1051
loss function properties, 1052-1053
normalized Buttenvorth loss functions,
1072
On the Theory of Filter Amplifiers
(Butterworth), 1034
passive realization, 1042-1044
singly terminated low-pass networks,
1069-1071
solution to approximation problem,
1036-1042
Capacitance, 284
ol
a pair of conductors, 306
Capacitors, 269-319. See also Inductors
admittance of, 608, 658
capacitive smoothing in power supplies,
269-270, 303-305
definitions and examples, 274-283,
284-293
discharging, through an inductor,
382-385
impedance of, 452, 608, 658
inductor and, 271-273
linear, 306
in parallel, 301
phasor relationships for resistors,
inductors, and, 449-455
in series, 298
series and parallel inductors and
capacitors, 293-302
v-i relationship and Laplace transform
method, 580-582
Car heater fan, 51-52
Cascaded connections, two-ports and,
1002-1009
Cathode ray tube, 2
Characteristic equation, 388, 417, 591
Characteristic roots, 417
Charge, 3-6, 11-12, 16, 22, 27, 38, 40
Circuit model, 34, 36
Circuit response computation
graphical convolution and, 774-781
Circuits, 2. Sec also Magnetically coupled
circuits; RL/RC circuits
connected, 62
critically damped, 390, 392, 417

difierentiator, 358, 364
elements, 2, 14-15, 53, 58, 60
integrator, 359, 364
model, 34, 36
open, 30, 37
oscillator, 385
overdamped, 389, 392, 417
parallel, 53
passive, 339
planar, 128
series, 53. 67
series-parallel, 73-78
short, 27, 30, 64
stable, 433, 482
tuned, 844
undamped, 385, 417
underdamped, 390, 392, 417
Classical difference amplifier, 164
Closed node sequence, 62, 89
Closed path, 61, 62, 63, 89
Coefficient-matching technique, 1072
Coefficient of coupling, 913, 937
Coils, 306. See also Inductors
coupled inductors, 937
mutual inductance (M), 885, 886-893
primary/secondary, 898, 917, 938
qualit)' factor of, 861
Common-ground rwo-port, 1002-1009
Commutative propert), of convolution, 766
Comparator, 174-175, 177
Complementary function, 352
Complete response, 352, 364, 693, 740
Complex exponential forcing functions, 481
capacitor, 452
complex sinusoid, 441
inductor, 451
in SSS computation, 444-447
Complex numbers, 435-441
Complex power, 510-515, 531
Complex power calculations, 536-537
Computation of response for linear timeinvariant systems, 795-797
Computing equivalent inductances,
940-943
Conductance, 30, 40, 454, 482
equivalent, 70
in parallel, 70, 72
transfer, 79
Conductors, 3, 40
ideal, 3, 4, 41
Connected circuit, 62, 63, 89, 140
Conservation of charge, 288
Laplace transform analysis circuit
121
Index
applications, 6 4 5 - 6 5 2
principle of, 3 0 6 , 6 4 5
Conservation o f complex power in AC
Current, 2 , 4 0
alternating, 5, 8 - 1 0
branch, 107, 110, 128
circuits, 515
Conservation o f power (energy), 19, 2 0 , 3 6 ,
4 0 ,5 1 5 . 5 3 1 ,5 3 7
Constant inputs, second-order linear
networks with, 4 0 0 411
Constant-resistance network, 6 4 0
Continuity property o f the capacitor, 2 8 7 ,
306
direct, 8 - 9 , 3 0 , 33
loop, 110, 128, 141
mesh, 128, 141
Current division, 8 9 , 4 8 2 , 6 0 4 , 6 5 8
formula, 6 8 - 7 3 , 4 5 8
Current source, 4 0 , 5 6
nonideal, 8 5 - 8 8
Continuity property o f the inductor, 2 7 8 ,
v-i characteristics, 56
Current transformation property, 9 1 5
306
's ^
Critically damped circuit, 3 9 0 , 3 9 2 , 4 1 7
Controlled source, 2 8 , 2 9 , 7 8 , 80
C u to ff frequency, 1033, 1072
current-controlled current (C C C S), 2 8 , 39

current-controlled voltage (C C V S), 2 8 , 39
voltage-controlled current (V C C S ), 2 8 ,
7 8 , 80
Damped oscillation frequency, 4 1 7 , 8 5 2 ,

861
Dam ping, 3 9 2
voltage-controlled voltage (V C V S ), 28,

7 8 , 80
D C (d irea current)
defined, 41
Convolution, 7 6 3 - 8 0 9
averaging by finite time integrator
first order R L and R C circuits, 3 3 6 - 3 4 7

motors, transfer function analysis of,
circuit, 7 6 3
7 3 6 -7 3 9
basic properties, examples, 7 6 6 - 7 7 2 ,

7 9 0 -7 9 3
257
circuit response com putation and

graphical convolution, 7 7 4 -7 8 1
O '
circuit response com putation using,

7 8 6 -7 9 0
defined, 7 6 6 , 7 9 7
Dependent (controlled) current source, 4 0 , 89
graphical convolution, 7 7 4 7 8 1 ,
Dependent (controlled) voltage source, 4 1 , 89

Dependent source, 78
by integral, 7 9 9 - 8 0 2
Derivative/integration property, 1100, 1108
interconnection o f active circuits and,
Dielectric, 2 8 4 , 3 0 6
Difference amplifier, 1 6 3 -1 6 4 , 167, 177
by inverse Laplace transform, 8 0 6 - 8 0 7
Differential equations, 3 6 4
Laplace transformation and, 7 7 2 7 7 4
com puting SSS and, 4 4 2 444
tim e domain derivation of, for linear
general second-order circuit, 3 8 7 - 3 8 8
time-invariant circuits, 7 9 3 - 7 9 7
ideal transformers and, 914
Convolution algebra, 7 8 1 , 7 9 7
ordinary, 591
Copper wire resistance, 3 3
series-parallel R L C circuit, 3 8 6 - 3 8 7
C o m er (break) frequency, 7 4 0
SSS analysis and, 4 8 3

Differentiator circuit, 3 5 8 , 3 6 4
Coulom b, 3 , 3 0 6
Coupled circuits.
See Magnetically coupled
circuits
Coupled inductors, models for pair of, 9 3 0 ,
937
Coupled inductors (coils), 8 8 6 - 8 8 7 ,
9 3 0 -9 3 6 , 937
Coupling network, SSS calculation and, 52 9 ,
530
Cram ers rule, 140
gain in dB, 1033

D elta function, 5 5 2 , 591
8 0 7 -8 0 9
'Sw/
257
Decade, 7 3 3 , 7 4 0
convolution theorem, 7 7 3 , 7 9 8
8 0 2 -8 0 4
Deactivating an independent voltage source,
Decibel (dB ), 7 3 3 ,7 4 0
convolution algebra, 7 8 1 - 7 8 6 , 8 0 4 - 8 0 6
Deactivating an independent current source,
Diodes, 3 0 3
Dirac delta function, 591
Dirichlet conditions, 1 0 9 8 -1 0 9 9
Distributivity, 7 6 6 , 7 9 8
D ot notation, 5 4 6 , 5 9 1 , 9 0 5 , 9 3 7
mutual inductance and, 8 8 6 - 8 9 3
Driving point admittance, 6 2 6
Driving point impedance, 6 2 6
D uty cycle, 1095, 1108
122
Index
Effective value, 1 0 ,4 1 , 505, 5 3 2 ,1 0 9 7 ,1 1 0 8

o f nonsinusoidal signals, 5 3 3 - 5 3 4
Butterworth solution to approximation

problem, 1 0 3 6 -1 0 4 2
o f signal and average power, 5 0 6 - 5 1 0
filter, defined, 1033, 1072
single-frequency analysis widi, 5 0 9 - 5 1 0
high-pass active realization, 1 0 8 3 -1 0 8 4
Elapsed tim e formula, 3 5 5

E le a ric field, 11
Electric m otor analysis, Laplace transform
analysis and, 6 8 3
Electron, 2 , 3
Elements
active, 8 2 , 177
high-pass filter design w ith passive

realization, 1 0 5 3 - 1 0 5 9
high-pass passive design, 1 0 8 1 -1 0 8 3
input attenuation for active circuit
design, 1 0 5 1 -1 0 5 2
loudspeakers and crossover network,
1 0 3 1 -1 0 3 2
circuit, 2 , 1 4 -1 5
low-pass approximation, 1 0 7 4 - 1 0 7 6
linear resistive element, 215
low-pass filter basics, 1 0 3 4 -1 0 3 6 ,
passive elements, 178

two-terminal, 16, 2 0 , 4 0 , 58
Energy, 16, 2 3 , 32
energy storage in capacitors, 2 9 0 - 2 9 1
kinetic, 12, 32
1 0 7 3 -1 0 7 4
passive low-pass realization, 1 0 7 6 -1 0 7 9
singly terminated Bunerw orth low-pass
networks, 1 0 6 9 -1 0 7 1
types of, 1 0 3 2 -1 0 3 4
net e n e i^ , 2 7 9
Final-value theorem, 7 2 6 - 7 2 9 , 7 4 0
n o n -D C power and, 2 2 - 2 5
Finite pole, 6 8 5
potential, 16
Finite tim e integrator circuit, 7 6 3
stored, 9 1 0 - 9 1 4
Finite zero, 6 8 5
stored in pair o f coupled inductors, 9 3 7
First order circuits, 3 2 2 - 3 7 8 .
Equilibrium state, 3 3 9 , 3 6 4
Equilibrium value, 3 5 4
Sef also RL/RC
D C or step response of, 3 3 6 - 3 4 7

first-order differential equaiton o f a
circuit, 3 6 4
first-order R C op amp circuits, 3 5 7 - 3 6 3
Equivalent n-terminal networks, 2 5 7
m ath used for, 3 2 4 - 3 2 8
Equivalent resistance, 6 5 , 6 9 , 70
oscillator responses in second-order linear
Equivalent two-terminal networks, 2 1 5
r>
networks and, 4 1 2
Euler identity, 4 3 6 , 4 8 2
response classifications, 3 5 2
Eulers formula, 3 8 9
sawtooth, 3 2 2 - 3 2 3 , 3 6 5
Even function, 1108
source-fiee or zero-input response, 3 2 8 -3 3 6
Exponential (complex) Fourier series, 1108
r>
analysis and theory, 3 5 2 - 3 5 7
Equivalent circuits, 6 1 8 - 6 2 5
Equivalent networks, 2 1 2 - 2 1 3
r\
circuits
Equivalent, defined, 2 5 7
Equivalent conductance, 6 9
superposition and linearity, 3 4 7 - 3 5 2
Exponential decay, 3 9 2
First-order tim e differentiation formula,
Fan speed control, 51
Floating source, 1 2 3 -1 2 8 , 140
Faradays law o f induction, 3 0 6 - 3 0 7
Fluorescence, 5 1 8 , 6 0 3 - 6 0 4 , 6 2 3 - 6 2 5
Farad (F), 2 8 4 , 3 0 6
Forced response, 3 5 2 , 3 6 4 , 7 0 0 , 7 4 0
Femto, 2 9
Forcing function, 3 3 8
Fields, 11, 12
Fourier series, 1 0 8 5 -1 1 1 7
5 7 6 -5 7 9
Filtering, 1 0 3 1 -1 0 8 4
active low-pass design, 1 0 7 9 -1 0 8 1
active realization o f high-pass filters,
1 0 5 9 -1 0 6 2
a a iv e realization o f low-pass Butterworth
filters, 1 0 4 5 -1 0 5 1
band-pass filter design, 1 0 6 2 -1 0 6 9
Butterworth loss function properties,
1 0 5 2 -1 0 5 3
Butterworth passive realization,
1 0 4 2 -1 0 4 4
additional properties and computational

shortcuts for, 1 0 9 9 -1 1 0 7
coefficients by integration, 1110
complex exponential Fourier series, 1090
SSS analysis with multi-fi%quency inputs,

1 1 1 2 -1 1 1 7
trigonom etric and exponential forms,
1 0 8 8 -1 0 9 9
trigpnometric Fourier series, 1089
using properties w ithout integration,
1110-1112
1123
Index
Four-terminal device, 80
Ideal constant current source, 86
Frequenqr, 4 3 3 ,4 8 2
Ideal constant voltage, 86
dom ain analysis, 5 4 7 , 5 5 4 , 591

response, 4 6 7 , 4 8 2
Frequency response, 7 0 7 - 7 0 8 , 7 4 0
Ideal current source, 2 7 , 3 8

Ideal delay o f
T seconds, 7 8 9
Laplace transform analysis and, 7 0 7 - 7 1 4
Ideal insulator, 41
magnitude, 1 0 3 3 , 1072
Ideal integrator, 3 6 0
SSS analysis and, 4 9 4 495
Ideal (linear) resistor, 15, 3 7
Frequency s ^ i n g , 5 8 2 - 5 8 3 , 7 4 0
Laplace transform m ethod and, 5 8 2 - 5 8 3
m agnitude scaling and, 7 1 4 - 7 2 6
Frequency shift property, 5 7 5 - 5 7 6 , 591
Fundamental com ponent (first harm onic),
10 8 9 , 1108
Fundamental frequency, 1088, 1108
O '
Ideal constant voltage source, 8 6
Ideal operational amplifier (op amp), 1 5 8 ,1 7 7

Ideal transformer, 9 3 7
defined, 9 1 7
Ideal transformers, 9 1 4 - 9 2 4 , 9 3 0 - 9 3 6 ,
9 5 0 -9 5 6
Imaginary part, 4 3 5 , 4 8 2
Impedance, 4 8 2
Fundamental period, 2 8 0 , 6 9 5 , 1 0 8 8 , 1108
o f the capacitor, 4 5 2 , 6 0 8 , 6 5 8
Fundamental period o f p e r i o d i c 7 4 0
defined, 6 0 6 , 6 5 8
Fuse, 3 2
impedance transformation property, 9 2 0

o f inductor, 6 5 8
Gain
in d B , 10 3 3 . 1072
power, 8 0
voltage, 8 0
G ain magnitude, 10 3 3 , 1072
phasor impedance, 4 5 4 , 4 5 5 - 4 6 1
Generalized conductance, 6 0 6
o f resistor, 6 0 7 - 6 0 8 , 6 5 8
General summ ing amplifier, 1 6 7 - 1 7 0 , 177
two-port impedance parameters, 9 7 6 -9 8 1
General voltage division formula, 6 6

Generator, 12, 2 6
Giga, 2 9
Graphical convolution, 7 7 4 - 7 8 1 , 8 0 2 - 8 0 4
Ground, 6 0
virtual, 160, 178
Grounded voltage sources

examples, 1 1 1 -1 2 2
w
w
o

Impulse response theorem, 7 9 7
Independent (ideal) current source, 2 7 ,3 8 ,4 1
Independent (ideal) voltage source, 2 5 - 2 7 ,
3 7 ,4 1
Induced voltage, 8 8 4 - 8 8 5
Inductance, 2 7 4
Half-wave symmetric, 1100, 1109
H enry (H ), 2 7 3 , 2 7 4 , 3 0 7
Half-wave symmetry property, 1100, 1109
linear inductor and, 2 7 4
H enry (H ), 2 7 3 , 2 7 4 , 3 0 7
Inductor impedance, 451
H i^ -p a s s circuit, 4 6 9 , 4 8 2
Inductors, 15, 2 6 9 - 3 1 9 .
also Filtering
H P (high-pass) to LP (low-pass) f l u e n c y
transform ation, 1 0 5 3 , 1072
Hybrid (A-) parameters, 9 6 1 , 9 8 5 - 9 9 1 ,
1010
Impulse response, 6 2 5 , 6 3 1 - 6 3 4 , 7 1 9 , 7 4 0
Half-power frequencies, 8 1 5
Hom ogenous differential equation, 4 0 0 , 4 1 7
two-ports and, 9 7 6 - 9 8 1 , 9 8 1 - 9 8 4
Impedance (z-) parameters, 9 6 1 , 9 7 6 - 9 8 1 ,
Half-power points, 1 0 3 3 ,1 0 7 2
High-pass filters, 10 3 3 , 1 0 5 9 -1 0 6 2 , 1072.
applications, manipulation, 6 0 9 - 6 1 7
Laplace transform m ethod and, 5 4 4
Gaussian curves (surfaces), 5 8 , 8 9 , 123, 141
General summ ing integrators, 6 5 2 - 6 5 7
O '
applications, 6 0 5 - 6 0 9 , 6 5 9 - 6 6 2
1 0 1 0 , 1 0 2 0 -1 0 2 3
See also Capacitors
adm ittance of, 6 0 8 , 6 5 8

capacitive sm oothing in power supplies
and, 2 6 9 - 2 7 0 , 3 0 3 - 3 0 5
capacitor, dischaiging through, 3 8 2 - 3 8 5
capacitor and, 2 8 4 - 2 9 3
defined, 2 7 1 - 2 7 3
examples, 2 7 4 - 2 8 3
impedance of, 6 5 8
Ideal band-pass filter, 8 1 4 - 8 1 6
linear inductor, defined, 2 7 4 , 3 0 7
Ideal banery, 84
in parallel, 2 9 5 - 2 9 7
Ideal co nd u aor, 3 , 4 , 41
passivity principle for ind uaors, 9 0 9
24
Index
n
capacitors, and, 4 4 9 - 4 5 5
series and parallel inductors and
nodal and loop analysis in i^-domain, 6 3 4

SSS analysis with phasors, 4 8 3 - 4 8 8
statements of, 6 1 , 6 3
capacitors, 2 9 3 - 3 0 2
Initial-value theorem, 7 2 6 - 7 2 9 , 7 4 0
Ladder network, 2 0 6 - 2 0 8
Input admittance, 9 7 4 , 10 1 0
Lagging, 4 5 1 , 5 1 8 , 5 2 6
Input attenuation, 1 0 5 1 - 1 0 5 2
Laplace, Pierre Sim on, 543
Input impedance, 9 6 2 , 9 9 0 , 1010
Laplace transform analysis
Instability, 3 6 2
basic signals, 5 4 8 - 5 5 4
5 0 0 - 5 0 5 , 5 3 2 , 533
basic signals and signal
Insulators, 3 , 41
In t^ ra tin g factor m ethod, 3 2 7 , 3 6 4
Integration, Laplace transform m ethod and,
5 7 9 -5 8 0 , 5 7 9 -5 8 2
Integrator, 3 5 9 , 3 6 4 , 6 2 9 , 6 5 8
basics, 5 4 3 - 6 0 2
Instantaneous power, 2 2 - 2 5 , 4 1 , 2 7 9 , 3 0 7 ,
Instantaneous stored e n e ^ , 2 8 0 , 291
/n
representation, 5 9 2 - 5 9 5
circuit responses, applied to
r>
o
differential equations, 6 0 0 - 6 0 2
frequency shift property, 5 7 5 - 5 7 6
integration, 5 7 9 - 5 8 0
inverse Laplace transform, 5 6 5 - 5 7 5
ideal, 3 6 0
Laplace transform pairs, 564
leaky, 3 5 9 , 3 6 4
one-sided, 5 5 4 - 5 6 3
Int^ro-difFerential equations, 5 8 5 - 5 9 0
overview, 5 4 7 - 5 4 8
Internal resistance, 6 4
by partial fraction expansion, 5 9 8 -6 0 0
Inverse Laplace transform, 5 6 5 - 5 7 5 , 591
properties, 5 8 4
Inverting amplifier, 159, 160, 173, 177
properties, finding, 5 9 5 - 5 9 8
Isolation amplifier, 164
second order tim e domain methods,
Joule, 12
5 4 4 -5 4 7
solution o f integro-difierential
KUo, 2 9
time dif&rendadon formula, 5 7 6 -5 7 8
Kilowatt-hour (kW h) meter, 5 2 6
time-/frequency-scaling property,
equations, 5 8 5 - 5 9 0
KirchhofTs current law (K C L ), 5 1 - 1 0 6 , 53,

5 5 - 6 0 , 89
definitions, 5 3 - 5 4
explained, 5 5 - 6 0
first order R L and R C circuits, 3 4 8
for Gaussian curves/surfaces, 59
applications, 6 0 9
Laplace transform m ethod and, 5 5 7 , 591
linearity, superposition, source
transformation and, 191
nodal and loop analysis and, 1 0 7 -1 0 8
5 8 2 -5 8 3
Vi relationship o f capacitor, 5 8 0 - 5 8 2
circuit applications, 6 0 3 - 6 8 1
design o f general summ ing
in t^ rators,
652-^57
inductors and capacitors, 6 1 8 - 6 2 5

impedance, admittance, voltage
division, source transformations,
6 5 9 -6 6 2
impedance and admittance,
capacitors and, 2 8 7
manipulation, 6 0 9 - 6 1 7
nodal and loop analysis in x-domain,
6 3 4 -6 4 0
op amp integrator design and, 6 8 0
response calculation with initial

applications, 6 0 9
Laplace transform m ethod and, 5 5 7 , 591
impulse and step responses, 6 3 1 - 6 3 4
defm itions, 5 3 - 5 4
r\
Thevenin, N orton equivalents,
SSS analysis with phasors, 4 8 3 - 4 8 8
5 3 - 5 5 , 6 0 - 6 4 . 90
fluorescence, 6 0 3 - 6 0 4 , 6 2 3 - 6 2 5
impedance and adm ittance, 6 0 5 - 6 0 9
KirchhoflPs voltage law (K V L ), 5 1 - 1 0 6 ,
equivalent circuits for initialized
nodal and loop analysis in ^-domain, 6 3 4

statements of, 55
r\
conditions, 6 6 8 - 6 7 6
sawtooth waveform generadon and,
6 8 0 -6 8 1
switched capacitor circuits and

conservation o f charge, 6 4 5 - 6 5 2
transform ation and, 191
112^
Index
switched capacitor networks, 6 7 8 - 6 8 0
Linear resistive circuit/networks, 2 1 5
switching problems, 6 7 6 - 6 7 8
Linear resistive element, 2 1 5
transfer functions, 6 2 5 -6 3 0 , 6 6 2 - 6 6 8
Linear voltage sweep, 3 2 2 - 3 2 3
convolution and, 7 7 2 - 7 7 4
Loading effect, 7 6
inverse, and convolution, 8 0 6 - 8 0 7
Loop analysis, 1 0 7 -1 5 4
Laplace transform, defmed, 5 5 4 , 591
concepts of, 1 1 0 -1 1 1
Pierre Laplace and, 543
explained, 1 2 8 -1 3 9
second-order linear networks with
floating voltage sources, 1 2 3 -1 2 8 ,
constant inputs, 4 0 0
transfer function applications, 6 8 3 -7 6 1
applications and bode techniques.
7 6 0 -7 6 1
classification o f responses, 6 9 3 -7 0 1
D C m otors and, 7 3 6 - 7 3 9
electric m otor analysis and, 6 8 3

nodal analysis and, 110, 141
terminology, 1 0 9 - 1 1 0
Loop (closed path), 110, 141
frequency scaling and magnitude
Loop current, 110, 128, 141
initial- and Bnal-value theorems.

7 2 6 -7 2 9 . 7 5 9 -7 6 0
Loop equations, 128, 136, 140

Lossless device, 2 8 0 , 3 0 7
Loss magnitude, 1033, 1072
poles, zeros, 6 8 5 - 6 9 3 , 7 4 2 - 7 4 4
Lossy device, 3 0 7
responses and classifications, 7 4 6 - 7 4 9
Lower half-power frequency, 8 1 6 , 861
SSS for stable networks and systems,
Low-pass (brickwall) filter specification, 1072
7 0 1 -7 0 7
Low-pass Butterworth filter, 6 3 4 .
stability problems, 7 4 5 - 7 4 6
steady-state calculation, 7 4 9 - 7 5 3
L C circuit, 3 7 9 , 3 8 0
See also
Bunerw orth filter

Low-pass filters, 10 0 3 , 1072.
See also
Filtering
angular frequency o f oscillation and.
V J
grounded voltage sources, 1 1 1 -1 2 2
frequency response, 7 0 7 - 7 1 4 ,7 5 3 - 7 5 9
scaling, 7 1 4 - 7 2 6
1 4 6 -1 4 8
general SSS analysis, 4 9 2 - 4 9 4
history of, 1 0 7 -1 0 8
bode plots, 7 3 0 - 7 3 6
Linear m atrix equation, 140, 141
switching in R L C circuits, 6 4 0 - 6 4 5
3 8 4 ,4 1 7
basics of, 1 0 3 4 -1 0 3 6
low-pass filter (brickwall) specification.
differential equation of, 3 8 2
1034
parallel, 385
LP to B P frequency transformation, 1072
undriven, 4 1 8
LP to H P frequency transformation, 1056,
L C resonance frequency, 861
1072
Leading, 451
Leaky integrator circuit, 3 5 9 , 3 6 4 , 5 8 8 - 5 9 0
A C adaptors and, 8 8 3
Linear active r ^ o n , 177
analysis of, with open-circuited
Linear circuit, 7 4 0 .
See also First order
circuits; Second order linear circuits

Linear differential equations, 3 2 6
Linearity, 1 9 1 -2 2 5 .
See also Source
transformation; Superposition
coefficient o f coupling and energy
linearity property, 2 1 5 , 4 4 0 , 1099, 1109

linearity theorem, 194
source transformations and, 2 0 8 - 2 1 2
superposition and, 3 4 7 - 3 5 2 , 3 6 5
superposition and proportionality, 2 0 1 -2 0 8
W
transformers, 9 5 0 - 9 5 6
equivalent networks and, 2 1 2 - 2 1 3
Laplace transform analysis and, 5 5 7 , 6 9 3
9 0 1 -9 0 9
analysis o f circuits containing ideal
inverse Laplace transform and, 5 6 5
secondary, 8 9 5 - 9 0 0
analysis of, with terminated secondary.
convolution method and, 7 9 3 , 7 9 8

explained, 1 9 3 - 2 0 0
M agnetically coupled circuits, 8 8 3 - 9 5 7
Light-em itting diode (L E D ), 189
calculations, 9 0 9 - 9 1 4
com puting equivalent inductances.
Z-Js), responses for simple circuits,

9 4 0 -9 4 3
coupled inductors modeled with ideal
transformer, 9 3 0 - 9 3 6
coupling coefficient problems and energy
calculations, 9 4 7 - 9 4 9
1126
Index
r\
diflferential equation, Laplace transform,
phasor models of, 8 9 3 - 8 9 5
dot placement,
M, and basic equations,
9 3 9 -9 4 0
Modulus, 4 3 5 , 4 8 2
M ultiplication-by-r property, 5 5 9 - 5 6 0
Mutual inductance
r^
(M), 8 8 5 , 9 3 7 . See also
M agnetically coupled circuits
general analysis o f circuits with coupled

inductors, 9 4 3 - 9 4 7
dot convention and, 8 8 6 - 8 9 3

rule for induced voltage drop due to, 8 8 7
ideal transformers, 9 1 4 - 9 2 4
magnetically coupled, defined, 883
Nano, 2 9
modek for practical transformers, 9 2 4 -9 3 0
Natural fi-equency o f a circuit, 3 6 5 , 3 8 8 ,
mutual inductance and dot convention,
417, 687
Natural modes o f vibration, 3 2 6
8 8 6 -8 9 3
M agnitude (modulus), 4 3 3 , 4 8 2
Natural response, 3 5 2 , 3 6 5 , 7 0 0 , 741
M agnitude response, 7 0 8 - 7 0 9 , 7 4 0
Network equivalency, 3 7 3
M agnitude scale factor, 6 3 0 , 7 1 4 - 7 2 6 ,
Network fim ction, 6 2 6
7 4 0 -7 4 1
Networks
r s
M atching networks, 8 4 8 , 861
equivalent networks, 2 1 2 - 2 1 3
M A TLA B
two-terminal networks, 2 0 9 , 2 1 2 , 2 1 5 ,
first order R L and R C circuits, 3 2 4 , 3 3 2 ,

341
2 2 8 ,2 5 7
Laplace transform and partial fraction

expansion, 572
Nodal analysis, 1 0 7 -1 5 4 , 110, 141

concepts of, 1 1 0 -1 1 1
defined, 110, l 4 l

floating voltage sources, 1 2 3 -1 2 8 ,1 4 6 -1 4 8
residue comm and, 671
grounded voltage sources, 1 1 1 -1 2 2

history of, 1 0 7 -1 0 8
M atrix inverse, 141
general SSS analysis, 4 9 2 - 4 9 4
series-parallel resistive circuits and, 1 0 4 -1 0 6
r>
M atrix notation, 6 3 6
M atrix partitioning, 9 6 6 , 101 0
linearity and, 195
M axim dly flat, 1 0 5 2 -1 0 5 3 , 1072
loop analysis and, 1 2 8 - 1 3 9
M axim um Power Transfer theorem,
o f pressure-sensing device, 4 7 7 481
See also N orton
theorem; Thevenin theorem

for A C circuits, 5 2 6
defined, 5 3 2
and, 461
r>
terminology, 1 0 9 -1 1 0
Node, 54, 90
Node voltage, 6 0 , 9 0 , 109, I 4 l
Maxwells equations, 2 7 3 , 3 0 7
Non-ideal constant current source, 87
Mega, 2 9
Non-ideal constant voltage source, 87
M em ory
Non-ideal dependent current source, 8 7
capacitors and, 2 8 6
inductors and, 2 7 5
Memoryless device, 41
Non-inverting amplifier, 162, 163, 177

Nonlinear controlled source, 80
Normalized Butterworth loss functions, 1072
Mesh current, 128, 141
Normalized circuits, 7 1 4
Metastable, 6 9 0
Normalized fi^quency, 1 0 3 5 - 1 0 3 6
m ho, 3 0 , 4 0 , 41
N orton, E . L , 2 2 7 , 1069
M icro, 29
N orton theorem, 2 2 7 - 2 6 8 .
MilU, 2 9
Modified loop equations, 136
Modified n o ^ analysis (M D A ), 110, 123,
Nonlinear, 8 0
Mesh analysis, 128, 141, 461
3 7 9 -3 8 0
r\
Non-ideal dependent voltage source, 87
Mesh, I 4 l
Microwaves, second order linear circuits and,
steady-state circuit analysis using phasors
Maxwell, Jam es Clerk, 107
140, 141
r\
with V/I division, SSS analysis, 4 9 0 - 4 9 1
for fiquency response plots, 4 7 2 4 7 6
2 4 9 -2 5 6 , 257, 5 3 9 -5 4 2 .
r>
See also
M axim um Power Transfer theorem;

Thevenin theorem
corollary to Thevenin and N ortons

theorems for passive networks, 2 3 6
equivalent o f one-port, 1010
1127
Index
general approach to finding Thcvenin
O utput impedance, 9 8 1 , 9 9 3 , 1010
and N orton equivalents, 2 3 6 -2 4 1
Overdamped circuit, 3 8 9 , 3 9 2 , 4 1 7

maximum power transfer theorem and,
2 4 9 -2 5 6
N orton equivalent circuit, 2 2 9 , 2 4 2 , 2 5 7
for passive networks, 231
Thevenin theorem and
active networks, 2 4 1 - 2 4 5
Parallel, 3 5 , 37
Parallel circuit, 90
Parallel connection, 53
two-ports and, 1 0 0 2 -1 0 0 9
Parallel resistance, 6 8 - 7 3
Parallel resonant circuit, 8 4 2 - 8 4 6
Parallel R L C circuit, 8 3 3 - 8 3 8
linear passive networks, 2 2 9 - 2 3 6
Parsevals theorem, 1098
op amp circuits, 2 4 6 - 2 4 9
Partial fi^ction expansions, 5 6 5 , 591
SSS analysis, 4 9 1 - 4 9 2
two-ports and, 9 6 4 - 9 6 8
distinct complex poles, 5 7 3 - 5 7 5

repeated poles, 5 6 9 - 5 7 2
Particular integral, 3 5 2
Octave, 7 3 3 , 741
Panitioned matrix, 9 6 6 , 1010
Odd (u n a io n , 1094, 1109
Passive analog filters, 1072
OfF-set terminal, 157
Passive circuit, 3 3 9
O hm , 2 9 , 41
Passive elements, 178
O hm s law, 3 0 - 3 7 , 41
inductors, capacitors, 4 4 9 - 4 5 5
O ne-port networks, 9 6 1 , 9 6 2 - 9 6 8 ,
1011-1012
O n the T heory o f Filter Amplifiers
(Butterw orth), 1034
theorems for, 2 3 6
N ortons theorem for, 231
Thevenins theorem for, 231
Passive R L C circuit, 3 6 5
Passive sign convention, 1 5 - 1 7 , 2 2 , 2 3 , 27,
3 0 ,4 1 , 54
Passivity principle for inductors, 9 0 9
Open-circuited secondary, magnetically
Peak frequency (center frequency), 8 1 5 , 861

Peak-to-peak value, 10, 41
O pen-circuit impedance parameters, 1010
Peak value, 10, 41
O pen-circuit output admittance, 9 8 7 , 1010
Periodic current waveform, 2 7 9 - 2 8 0
O pen-circuit voltage, 231
Periodic signal, 6 9 5 , 7 4 1 , 1088, 1109
O pen-loop gain, 172, 177
Period o f waveform, 4 3 3
Operational amplifier (op amp), 121, 1 5 5 -1 8 9
Phase response, 741
design o f general summ ing amplifiers,

1 6 7 -1 7 0
Phasors.
See abo Sinusoidal steady state (SSS)
defined, 4 8 2
first-order R C op amp circuit, 3 5 7 - 3 6 3
phasor analysis, 4 3 5
idealized, 1 5 7 -1 6 6
relationships for resistors, ind uaors,
op amp circuits and SSS analysis, 4 9 5 -4 9 7

op amp open-loop gain, 741
phasor m ethod for analyzing, 4 6 0 - 4 6 1 ,
4 7 5 -4 7 6
saturation and active region of, 1 7 1 -1 7 6
Thevenin and N orton equivalent circuits
for op amp circuits, 2 4 6 - 2 4 9
capacitors, 4 4 9 455
representations o f sinusoidal signals,
4 4 7 -4 4 9
steady-state circuit analysis using,
4 6 1 -4 6 6
Photo timer, 3 7 7
Pico, 29
Ordinary differential equation model, 591
Piecewise linear curve, 7 3 4 , 741
Oscillation
Piecewise linear relationship, 1 7 1 -1 7 2
natural frequency o f a circuit, 3 6 5
Jt-equivalent circuit, 9 0 8 , 9 3 8 , 1010
oscillator circuit, 3 8 5 , 4 1 7
Planar circuit, 128
sinusoidal, 378
Polar coordinates, 4 3 6 , 4 8 2
W ien bridge, 4 1 2 , 4 1 4 , 4 7 5 - 4 7 6
Poles, 5 5 7
O utput admittance, 9 7 4 , 1010
O pen circuit, 3 0 , 3 7 , 41
coupled circuits and, 8 9 5 - 9 0 0
'< J
op amp and, 156

Passive networks
band-pass transfer funcdon vwth, 8 5 1 -8 6 0
1128
Index
See RL/RC circuits
distinct complex poles, 5 7 3 - 5 7 5
R C circuit.
Laplace transform analysis transfer
R C to C R transformation, 1062, 1072
function applications, 6 8 5 - 6 9 3 ,
Reactance, 4 5 6 , 4 8 2 , 8 4 3
7 4 2 -7 4 4
Reactive power, 5 1 0 - 5 1 5 , 5 3 2
pole frequency, 8 1 9
Real part, 4 3 5 , 4 8 2
pole Q. 8 5 2 , 861
Real power, 5 3 2
pole-zero, 8 1 6
Reciprocity
o f rational fiinction
fmite, 5 9 1 ,6 8 5 ,7 4 1
simple, 5 9 1 , 741
repeated, 5 6 9 - 5 7 2 , 6 8 5
Potential difference, 12, 16
Power, 1 6 - 2 5 , 4 1 .
See abo Sinusoidal steady
state (SSS)
reciprocal networks, 9 9 8 , 1010

two-ports and, 9 9 7 - 1 0 0 2 , 1 0 2 9 -1 0 3 0
Region o f convergence (R O C ), 5 56, 591

Residue theorem, 5 65, 5 6 6
average power and SSS power
Resistance, 3, 3 0 , 4 1 , 4 8 2
converter, 2 4 7
complex, 5 1 0 - 5 1 5 , 5 3 1 , 5 3 6 - 5 3 7
equivalent, 6 5 , 6 9 , 70
conservation o f power, 19, 2 0 , 3 6 , 4 0 ,
internal, 64
negative, 6 8 , 2 4 6
half-power frequencies, 815
output, 2 5 0
half-power points, 1033, 1072
parallel, 6 8 - 7 3
instantaneous, 2 2 - 2 5 , 4 1 , 2 7 9 , 3 0 7 ,
phasor admittance/impedance, 4 5 6
Resistivity, 3 3 , 41
M axim um Power Transfer theorem,
Resistors. 15, 16, 2 2 , 3 0 , 3 1 , 4 1

ideal (linear), 15, 37
power gain, 80
impedance of, 6 0 7 - 6 0 8 , 6 5 8
reactive, 5 1 0 - 5 1 5 , 5 3 2
nonlinear, 41
real, 5 3 2
phasor relationships for inductors,
upper half-power frequency, 8 1 6 , 8 6 2

Power factor (pO. 5 1 7 - 5 2 6 , 5 3 2 . 5 3 7 - 5 3 9
Resonance.
See abo Band-pass circuit
phenomenon, 8 3 9 - 8 4 2
Practical inductors, 8 3 0
resonance frequency, 8 6 2
Practical source, 80
series and parallel resonant circuits,
Primary coils, 8 9 8 , 9 1 7 , 9 3 8
Product over sum rule, 71
Proportionality, 2 0 1 - 2 0 8
8 4 2 -8 4 6
r\
r\
r\
n
classifications, 3 5 2
natural frequency o f a circuit and, 215
complete, 3 5 2 , 3 6 4
forced, 3 5 2 , 3 6 4
Pulse function, 549
r\
series-parallel resonant circuits, 8 4 6 -8 5 1

Response
property, 215
proportionality property, 2 0 1 - 2 0 8
r\
capacitors, and, 4 4 9 - 4 5 5
Power gain, 80
Practical transformers, 9 2 4 - 9 3 0
r>
transfer, 79
lower half-power frequency, 8 1 6 , 861

2 4 9 -2 5 6 , 257, 526, 532, 5 3 9 -5 4 2
7 9 4 -7 9 5
Rectangular coordinates, 4 3 6 , 4 8 2
available, 528
5 0 0 - 5 0 5 , 53 2 , 533
r>
Rectangular approximations to signals,
Repeated pole, 685
5 1 5 . 5 3 1 ,5 3 7
r\
reciprocal two-port, 1010
apparent, 5 1 0 - 5 1 5 , 531
calculation, 5 0 0 - 5 0 5 , 5 33, 5 3 4 - 5 3 6
natural, 3 5 2 , 3 6 5
source-free, 3 2 8 - 3 3 6 , 3 6 5
step, 3 3 6 - 3 4 7 , 365
Quality factor, 8 1 6
o f a band-pass circuit, 861
o f a capacitor, 861
o f a coil, 861
o f L and C components, 8 3 0 - 8 3 3
o f a reactive com ponent, 861
unstable, 3 3 5 , 3 6 5
zero-input, 3 2 8 - 3 3 6 , 3 5 2 , 3 6 5
zero-state, 3 5 2 , 365
Reverse open-circuit voltage gain, 9 8 6 - 9 8 7 ,
1010
Reverse voltage gain, 9 9 2
R L C circuit.
Ramp function, 55 0 , 5 6 0 , 5 9 1 , 741

Rational function, 5 6 5 , 6 8 5 , 741
r\
See abo Second order linear
circuits
computing SSS response and, 4 4 2 444
1129
Index
damping and, 3 9 2
Series connections, 1 0 0 2 -1 0 0 9
Series inductors, 2 9 3 - 2 9 5
Series-parallel circuit, 73
Series-parallel com binations
parallel, 8 3 3 - 8 3 8
passive R C or R L circuits
vs., 3 9 7 - 4 0 0
source-free, 3 8 3 - 4 0 0
Series-parallel impedance
steady-state analysis of, 6 0 4

switching in, 6 4 0 - 6 4 5
RL/RC circuits, 3 2 2 - 3 7 8 .
with V/I division, SSS analysis, 4 8 8 490

Series-parallel resistive circuit
See also First order
circuits
O '
capacitors, 301
inductors, 2 9 7 - 2 9 8
resonance and, 8 3 9
definitions, 5 1 - 1 0 6 , 5 3 - 5 4
dependent sources and, 7 8 - 8 4 , 1 0 3 -1 0 4
analysis and theory, 3 5 2 - 3 5 7
com puting SSS response and, 4 4 2 - 4 4 4
M A TLA B and, 1 0 4 - 1 0 6
D C or step response of, 3 3 6 - 3 4 7
model for non-ideal battery and, 8 4 - 8 5
first-order R C op amp circuits, 3 5 7 - 3 6 3
non-ideal sources and, 8 5 - 8 8
m ath used for, 3 2 4 - 3 2 8
parallel resistance and current division,
passive R L C circuit, 3 6 5
sawtooth, 3 2 2 - 3 2 3 , 3 6 5
source-free or zero-input response, 3 2 8 -3 3 6
steady-state circuit analysis using phasors,
4 6 1 -4 6 6
6 8 -7 3 , 9 6 -9 8
RgO and related calculations of, 9 9 - 1 0 3
series-parallel interconnections, 7 3 - 7 8
Series-parallel resonant circuit, 8 4 6 -8 5 1
Series-parallel R LC circuit, 3 8 6 - 3 8 7
superposition and linearity, 3 4 7 - 3 5 2

Root-mean-square (rms), 10, 4 1 , 50 6 , 53 2 .
See also Effective value

(Thevenins equivalent resistance), 2 5 7 .
See also Thevenin theorem

Running average, 7 6 3 , 7 8 8 , 7 9 0 , 7 9 8
Series resonant circuit, 8 4 2 - 8 4 6

Shifted step, 5 4 9 , 591
Short circuit, 2 7 , 3 0 , 4 1 , 64
Short-circuit admittance parameters, 9 69, 1010
Short-circuit forward current gain, 9 86, 1010
Short-circuit input impedance, 9 8 6 , 1010
Siemens, 2 9 , 4 0 , 41
Sallen and Key circuit, 1046
Sifting property, 55 3 , 5 9 1 , 7 6 7 , 7 9 8
Sallen and Key normalized low-pass
Simultaneous equations, 9 3 0
Butterworth filter, 6 3 4 - 6 3 5 , 8 5 6 - 8 5 8 ,
1 0 4 5 -1 0 5 1 .
See also Butterworth filter
Saraga design, 1 0 4 7 -1 0 5 0
Saturation, 164, 171
Saturation regions, 178
Sawtoodi waveform, 3 2 2 - 3 2 3 , 3 65, 6 8 0 -6 8 1
Single-frequency analysis, 5 0 9 - 5 1 0
Sinusoidal oscillation, 3 7 8 , 4 1 2 - 4 1 5
Sinusoidal steady state (SSS)
analysis, 4 3 1 497
complex exponential forcing
functions in, 4 4 4 - 4 4 7
Scaled sum o f waveforms, 4 1 7
complex numbers and, 4 3 5 -4 4 1
5-domain
defined. 4 3 3 , 4 8 2

elementary impedance concepts,

4 4 9 -4 5 5
nodal and loop analysis in, 6 3 4 - 6 4 0
frequency response and, 4 6 7 - 4 7 6
poles, zeros, and, 6 8 5 - 6 9 3 , 7 4 2 - 7 4 4
as high-accuracy pressure sensor
Secondary coils, 8 9 8 , 9 1 7 , 9 3 8
Second order linear circuits, 3 7 9 430
with constant inputs, 4 0 0 -4 1 1
defined, 4 1 7
discharging a capacitor through an
inductor, 3 8 2 - 3 8 5
oscillator application, 4 1 2 - 4 1 5
source-free, 3 8 5 - 4 0 0
Second order time domain methods, 5 4 4 -5 4 7
Seleaivity, 8 1 6 , 8 6 2
Semiconductors, 4 , 5
Series circuit, 90
application, 4 3 1 4 3 2
Laplace transform analysis and,
7 0 1 -7 0 7
naive technique for computing,
442^44
nodal analysis o f pressure-sensing
device, 4 7 7 -4 8 1
phasor impedance and admittance,
4 5 5 -4 6 1
phasor representations o f sinusoidal
signals, 4 4 7 449
o f R L C circuit, 6 0 4
1130
Index
r>
using phasors, 4 6 1 4 6 6
Square wave, 3 4 5 , 3 4 6
power calculation, 4 9 9 - 5 4 2
Stable circuit, 4 3 3
complex power and, 5 1 0 - 5 1 5

complex power conservation, 5 1 3 -5 1 7
effective value o f signal and average
defined, 4 8 2
stability and boimdedness, 6 9 0
Stable transfer function, 741
Steady state, 4 3 3 .
power, 5 0 6 - 5 0 9
instantaneous and average powers,
See also Fourier series
Steady-state analysis, 6 0 4 , 6 5 8
Steady-state calculation, 7 4 9 - 7 5 3
5 0 0 -5 0 5
m aximum power transfer in SSS,
5 2 6 -5 3 0
Steady-state circuit output response, 701

Steady-state response, 6 9 6 , 741
power factor/power (actor correction,

5 1 7 -5 2 6
Step function, 4 1 7
Step response, 3 6 5 , 4 1 7 , 6 3 1 - 6 3 4 , 741
single-frequency analysis with

effective values, 5 0 9 - 5 1 0
Stored energy, 3 7 6 , 9 1 0 - 9 1 4
Stray capacitance, 3 6 1 , 3 6 2 , 3 6 5
Sm oothing, 2 6 9 - 2 7 0 , 3 0 3 - 3 0 5
Supermesh, 136
Solar cell, 3 7 6
Supernode, 123
Source, 16
Superposition, 1 9 1 -2 2 5 , 3 6 5 .
n
See also
Linearity; Source transformation
controlled, 2 8 , 2 9 , 7 8 , 80
current, 56
o f average power, 5 0 3 - 5 0 5
current-controlled ciurent (C C C S), 28, 39
linearity and, 1 9 3 -2 0 0 , 3 4 7 - 3 5 2
current-controlled voltagp (C C Y S), 28, 39
proportionality and, 2 0 1 - 2 0 8
floating, 1 2 3 -1 2 8 , 140, 1 8 8 -1 8 9
source transformations and, 2 0 8 - 2 1 2
ideal current, 2 7 , 38
source transformations and equivalent
n
n
networks, 2 1 2 - 2 1 3
independent current, 2 7 , 38
independent (ideal) current, 2 7 , 38
superposition property, 2 0 1 , 215
r\
independent (ideal) voltage, 2 5 - 2 7 , 3 7
Susceptance, 4 8 2 , 8 6 2
nonideal, 8 5 - 8 8
Switched capacitor circuit, 6 4 5 - 6 5 2 , 6 5 8
nonlinear, 80
Switching
first order R L and R C circuits, 3 2 3 , 3 4 5
nonlinear controlled, 80
in R L C circuit, 6 4 0 - 6 4 5
practical, 2 6 , 7 9 , 8 0 , 81
v-i characteristics, 8 0 , 86
Sym m etric matrbc, 141
voltage, 6 4
Sym m etry properties o f Fourier series, 1094,
1109
Source free, 4 1 7
Source-free response, 3 2 8 - 3 3 6 , 365
System identification, 7 6 5
r>
Tank circuit, 8 4 3 , 8 6 2
r\
Source-free second order linear circuits,

3 8 5 -4 0 0
Source transformation, 1 9 1 -2 2 5 .
See also
Linearity; Superposition
defmed, 2 1 5
Tank frequency, 8 4 3 , 8 6 2
Temperature measurement, nodal and loop
r\
analysis, 143
explained, 2 0 8 - 2 1 2
T-equivalent circuit, 9 0 8 , 9 3 8 , 1010
Tera, 29
linearity and, 1 9 3 -2 0 0
Term inated secondary, magnetically coupled

circuits and, 9 0 1 - 9 0 9
linearity and superposition, 2 0 1 - 2 0 8
Term inated two-ports, 9 7 3 , 1010
source transformation theorem and
Thevenin, M . L ., 2 2 7
equivalent networks, 2 1 2 - 2 1 3
source transformation theorem for
independent sources, 2 0 9
Source transformation property, 6 1 5 , 6 5 8
S P IC E (Simulation Program with Integrated
Circuit Emphasis), 178
for frequency response plots, 4 7 2 - 4 7 6
W ien bridge oscillator and, 4 1 4
r\
Thevenin theorem, 2 2 7 - 2 6 8 .
See also
M axim iun Power Transfer theorem;

N orton theorem
theorem s for passive networks, 2 3 6
equivalent o f a one-port, 1010
general approach to finding Thevenin
and N orton equivalents, 2 3 6 -2 4 1
1131
Index
O
M axim um Power Transfer theorem and,
2 4 9 -2 5 6
N orton theorem and
Transfer resistance, 7 9
Transformers.
(Thevenins equivalent resistance), 2 5 7

SSS power calculations and, 4 9 9
steady-state circuit analysis using phasors
and, 4 6 4 - 4 6 6
Transient analysis, 5 4 4 , 6 5 8
Transient response, 4 3 3 , 6 9 6 , 741
Translation o f a plot, 1 0 9 2 - 1 0 9 3 , 1109
Thevenin equivalent resistance, 231
Translation property o f Fourier series,
3 dB bandwidth, 8 1 5 - 8 1 6 , 861
1 0 9 3 -1 0 9 4 , 1109
Transmission
{t-) parameters, 9 6 1 , 9 9 1 - 9 9 4 ,
1010, 1 0 2 3 -1 0 2 5
Triangular waveform, 3 5 5 - 3 5 7
Trigonom etric Fourier series, 1089, 1109
o f B P filter, 1071
Tuned circuit, 8 4 4 , 8 6 2
o f LP filter, 1071
Two-dependent source equivalent circuit,
3 dB down point, 1033

3 dB firequency, 7 3 6 , 7 4 0
9 7 2 - 9 7 3 , 1010
Two-ports, 9 5 9 - 1 0 3 0
Three-term ind device, 8 0 , 9 0 8
admittance parameters, 9 6 8 - 9 7 3
T im e constant, 3 2 9 , 3 6 5
amplifiers, 9 5 9 - 9 6 0
T im e differentiadon formula, 5 7 6 - 5 7 8
general relations am ong two-port
T im e dom ain circuit response com putation.
See Convolution
T im e invariance, 6 5 8 , 7 9 3 , 7 9 8
parameters, 9 9 4 - 9 9 7
A-parameters, 9 6 1 , 9 8 5 - 9 9 1 , 1 0 2 0 -1 0 2 3
impedance and gain calculations o f
Tim e-scaling property, 5 8 2 - 5 8 3
terminated two-ports modeled by
T im e shift property, 5 5 8 - 5 5 9 , 5 9 1 , 7 6 7
z-parameters, 9 8 1 - 9 8 4
Touch-tone phones, 8 1 1 - 8 1 2
impedance parameters, 9 7 6 -9 8 1
Transconductance, 7 9
one-port networks, 9 6 1 , 9 6 2 - 9 6 8 ,
Transfer admittance, 6 2 6
1 0 1 1 -1 0 1 2
Transfer conductance, 7 9
parallel, series, cascaded co n n eaio n s of,
Transfer function, 6 2 5 - 6 3 0 , 6 5 8 , 6 8 3 - 7 6 1
1 0 0 2 -1 0 0 9
parameter conversion and inter
applicadons and bode techniques, 7 6 0 -7 6 1

band-pass circuits, 8 6 3 - 8 6 8 , 8 7 6 - 8 7 9
connection o f two-ports, 1 0 2 5 -1 0 2 9
bode plots, 7 3 0 - 7 3 6
reciprocity, 9 9 7 - 1 0 0 2 , 1 0 2 9 -1 0 3 0
classification o f responses, 6 9 3 -7 0 1
^-parameters, 9 6 1 , 9 9 1 - 9 9 4 , 1 0 2 3 -1 0 2 5
D C m otors and, 7 3 6 - 7 3 9
transmission parameters, 9 9 1 - 9 9 4
electric m otor analysis and, 6 8 3
^ p aram eter analysis o f terminated
frequency response, 7 0 7 - 7 1 4 , 7 5 3 - 7 5 9
frequency scaling and magnitude scaling,
7 1 4 -7 2 6
initial- and finai-value theorems,
7 2 6 - 7 2 9 ,7 5 9 - 7 6 0
-V
ideal, 9 1 4 - 9 2 4 , 9 3 0 - 9 3 6 , 9 5 0 - 9 5 6
practical, 9 2 4 - 9 3 0
Transistor photo timer, 3 7 7
231
two-ports and, 9 6 4 - 9 6 8
defined, 885
Thevenin equivalent, deBned, 2 2 9 , 2 4 2
Thevenin theorem for passive networks,
See also M agnetically coupled
circuits
superposition and, 3 4 9
Thevenins equivalent circuit, 2 5 7
Vw^
steady-state calculation, 7 4 9 - 7 5 3
Transfer impedance, 6 2 6
linear passive networks, 2 2 9 - 2 3 6

SSS analysis, 4 9 1 - 4 9 2
'O
7 0 1 -7 0 7
stability problems, 7 4 5 - 7 4 6
active networks, 2 4 1 - 2 4 5
op amp circuits, 2 4 6 - 2 4 9
L >
SSS for stable networks and systems,
two-ports, 9 7 3 - 9 7 6 , 1 0 1 2 -1 0 1 6
z-parameters, 1 0 1 6 - 1 0 2 0
Two-terminal circuit clem ent, 58
Two-terminal device, 15, 3 0
Tw o-terminal networks, 2 0 9 , 2 1 2 , 2 1 5
Thevenin and Nortons theorems, 228, 257
V 257
Laplace transform m ethod and, 5 4 4

poles, zeros, 6 8 5 - 6 9 3 , 7 4 2 - 7 4 4
Unbounded voltage or current, 3 0 7
responses and classifications, 7 4 6 - 7 4 9
Undamped circuit, 3 8 5 , 4 1 7 , 8 6 2
1132
Index
n
Undamped natural frequency, 8 4 3 , 8 6 2
Wattage, 19, 41
Underdamped circuit, 3 9 0 , 3 9 2 , 4 1 7
Waveform
U nit impulse function, 5 5 2 , 591
effective value of, 505
U nit step fu n a io n , 3 2 4 - 3 2 5 , 3 6 5 , 5 4 8 , 591
periodic current, 2 7 9 - 2 8 0
r\
n
U nity coupling, 9 3 1 , 9 3 8
period of, 4 3 3
Universal resonance curves, 8 2 6 , 8 6 2
sawtooth, 3 2 2 - 3 2 3 , 3 6 5
Unstable response, 3 3 5 , 3 6 5
scaled sum o f waveforms, 4 1 7
Upper half-power frequency, 8 1 6 , 8 6 2
second order linear circuit and, 3 8 3 - 3 8 5
sinusoidal voltage, 4 1 2
v-i characteristic, 2 6 , 3 7 - 3 8 , 41
v-i relationship o f capacitor
Laplace transform m ethod, 5 8 0 - 5 8 2
Virtual ground, 160, 178
triangular, 3 5 5 - 3 5 7 , 6 8 0 -6 8 1
W ien bridge oscillator, 4 1 2 , 4 1 4 , 4 7 5 476
W ire, resistance o f copper wire, 33.
See also
Resistivity
V inually grounded, 160

Virtual short circuit model, 1 5 8 ,1 6 0 , 1 7 7 ,1 7 8
V 257
j^-parameters
9 7 3 - 9 7 6 ,1 0 1 2 - 1 0 1 6
Voltage, 2, 16
branch, 107
analysis o f terminated two-ports,

z-parameters and, 9 7 6 - 9 8 1
gain, 117
node, 6 0
regulators, 3 0 3
Voltagp division, 9 0 ,2 9 5 ,2 9 9 ,3 0 0 ,4 8 2 ,6 5 8
Zero
finite zero, 6 8 5
Laplace transform analysis transfer
formula, 6 6 , 4 5 6
function applicadons, 6 8 5 - 6 9 3 ,
7 4 2 -7 4 4
Voltage drop, 10
Voltage follower, 1 64, 178
Zero-input response, 3 2 8 - 3 3 6 , 3 5 2 , 3 6 5 ,
Voltage gain, 80
Zero-state response, 3 5 2 , 3 6 5 , 6 9 3 , 7 4 1 , 7 9 8
Voltage (potential difference), 41
Z Js). 9 4 0 - 9 4 3
2-parameters, 9 7 6 - 9 8 1 , 1 0 1 6 -1 0 2 0
Voltage regulator, 3 0 7
4 8 2 , 6 9 3 , 741
Zeros o f rational function, 5 6 6 , 5 91, 741
Voltage source, 4 1 , 6 4
nonideal, 8 6
v-i charaaeristics, 8 0 , 8 6
Voltage-to-current converter, 188

Voltage transformation property, 9 1 4 , 9 3 1 ,
933, 937, 938
Voltmeter, 14, 7 6
sensitivity, 7 7
n
o
n
r s

Linear Circuit Analysis - DeCarlo-Lin 3e

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Linear Circuit Analysis - DeCarlo-Lin 3e

Uploaded by

Copyright:

Available Formats

'

TIME DOMAIN, PHASOR, AND LAPLACE

Cover image (^^J^ikiaui

^ Used under license from Shutterstock, Inc.

Printed in the United States of America

Charge, Current, Voltage

Role and Importance o f Circuits in Engineering

CHAPTER O B jEC TIV ES

Define a two-terminal circuit element.

that these quantities are conserved.

nal generators in a circuit.

ch ap ter 1 Charge, Current, Voltage and O hm s Law

1. ROLE AND IM PORTANCE OF CIRCUITS IN ENGINEERING

Chapter 1 Charge, Current, Voltage and O hm s Law

2. CH A RGE AND CU RREN T

Exercise. How many electrons have a combined charge o f -5 3 .4 0 6 x 10

c h ap ter 1 Charge, Current, Voltage and O hm s Law

Chapter 1 Charge, Current, Voltage and O hm s Law

Describe the charge movement on the wire at the boundaries A and B.

F IG U R E 1.5 Material through which positive and negative charges move.

T he wire is a metallic conductor in which only electrons move. Therefore, at boundaries

mate current flow is

Chapter 1 Charge, Current, Voltage and O hm s Law

Chapter 1 Charge, Current, Voltage and O hm s Law

Exercise. Repeat the preceding exercise if q{t) = 5e

Chapter 1 Charge, Current, Voltage and O hm s Law

From Equation 1.3, for t>Q,

1; q{t) = IC for r > I

ch a p ter 1 Charge, Current, Voltage and O h m s Law

i{t) = 1 2 0-^ sin (1 2 0 ?) A; (c) neither ac nor dc.

Chapter 1 Charge, Current, Voltage and O hm s Law

X peak-value = Q .lO llK

i.e., 0 .636 X peak value.

Chapter 1 Charge, Current, Voltage and O hm s Law

Chapter 1 Charge, Current, Voltage and Ohms Law

the distance traveled. Therefore,

energy converted = kinetic energy gained oc q

The voltage drop from B to A Is - 2 V.

Chapter 1 Charge, Current, Voltage and O hm s Law

sense. A second, more-common convention uses + and - markings

stipulates which terminal is at a higher potential; neither alone can do

= 9 V. Find the values o f

Chapter 1 Charge, Current, Voltage and O hm s Law

Exercise. In Figure 1.12, find

Chapter 1 Charge, Current, Voltage and O h m s Law

FIGURE 1.14 (a) General circuit element (connected to a battery) as an energy

Chapter 1 Charge, Current, Voltage and O hm s Law

5. VO LTAG E, CURRENT, POW ER, ENERGY, RELATIONSHIPS

Also assume that a constant current

flows from terminal A to terminal B through

circuit element 1, as shown.

FIG U RE 1.15 A general circuit in which a two-element circuit element is extracted

> 0. During a time interval o f T s, (V^g x T) C o f

Chapter 1 Charge, Current, Voltage and O h m s Law

Exercise. In Figure 1.15, the current ^AB - 5 niA, and

= 400 V. W hat is the energy absorbed

J), absorbed during the time interval Tis

-1 ^ ^ ^ ^ < 0 . This means that the remain

der o f the circuit absorbs negative power or equivalently delivers |^ 5 (

Exercise. In Figure 1.15, -Pg =

watts, i.e., the remainder o f the circuit absorbs - 1 0 watts o f

power. How much power does circuit element 1 absorb?

> 0 means the element absorbs (positive) power,