Professional Documents
Culture Documents
Measurement Techniques
PIM
D
Tx
D
PIM
DUT
Tx
PIM
D
D
PIM
Reflected
Transmitted
| INTRODUCTION
Through interaction with engineers and technicians in many countries throughout the
world,Kaelushashadtheopportunitytodiscussandevaluatemeasurementtechniquesas
they apply to a wide variety of items. We have discussed the problem of IM with
component manufacturers, infrastructure providers, site managers and service providers.
This perspective combined with the knowledge gained through the development and
manufactureofourPassiveIntermodulationAnalyzers,formsthebasisofourapproachto
the measurement of passive intermodulation, and the test capabilities built into our
analyzers.
Ourfindingssuggestthefollowing:
DynamicmeasurementsshouldbeperformedonallPIMcriticaldevices
Fixedfrequencymeasurementsmaynotbeadequateduetothefrequencydependent
characteristicsexhibitedbymanydevicesandsubsystems
PassiveIMisanincreasingconcerninthewirelesscommunity,becauseofthedebilitating
affect it can have on the performance of the telecommunications network. Passive IM is
generatedwhenevermore thanone frequencyencounters a nonlinearelectrical junction
or material. The resulting creation of undesired signals, mathematically related to the
originalfrequencies,canresultindecreasedsystemcapacityand/ordegradedcallquality.
Reduced capacity and call quality at a cell site leads to reduced income for the wireless
serviceproviders.Theeconomiclossiscompoundedwhentheaffectedcustomerbecomes
frustratedandswitchesovertoacompetingprovider.
www.kaelus.com
| Q&A
Question: IsPassiveIMADesignIssue,AManufacturingIssue,orAMaintenanceIssue?
Answer: AlloftheAbove.
Agooddesignisanecessary,butnotasufficientconditionforsuccess
Atonetime,manycompaniesconsideredIMsomethingthatcouldbecontrolledsimplyby
followingafewwelldocumenteddesignrules.
Avoidtheuseofferrousmetals
Solderorcoldweldalljunctionswherepossible
Minimizethenumberofcontactjunctions
Designallcontactjunctionssuchthattheyarepreciseandundersufficientpressure
tomaintaingoodcontact
Avoiddissimilarmetalsindirectcontact
Plateallsurfacestopreventoxidation
Makecertainplatingisuniformlyappliedandofsufficientthickness
Thebiggerchallengeisbuildingit
Although, the rules seem straightforward, manufacturing excellence isthe key to success.
MinordeviationsfromtheidealprocesscancreateunacceptablelevelsofPassiveIM.
Someoftherealworlddeficienciesthatoccur:
Pooralignmentofparts
Inadequatelytorquedscrewsandfasteners
Badsolderjoints
Insufficientorincompletecleaningofpartspriortoplating
Contaminatedplatingtanks
Platingmaterialbuildup
Usingwrongmaterials
Poorplatingadhesion
Thenyouhavetointegrateandinstallit
Once you have good designs and wellmanufactured components, the next challenge is
integrationandsysteminstallation.RecallthatalljunctionsarepotentialPIMgenerators.
2000 Kaelus Inc. All Rights Reserved.
www.kaelus.com
Therefore,everyinterconnectisapotentialsourceoftrouble.Jointsthattestoutfineina
benign environment can become problematic when mechanically stressed (bending of
cables,torsionalloadsonconnectorinterfaces,over/undertorquedconnectors).
In summary, it is not a given that interconnecting components that meet their IM
specificationassuresasubsystemthatalsomeetstheIMspecification.
AndthenMotherNaturetakesover.
Thenaturalelementsaretheenemyofthoseportionsofthenetworkthatareexposedto
them.
Windinducedvibration
dailytemperaturevariations
moistureinitsvariousforms
thermalloadingbythesun
airbornedirt
Each of these works to break down and erode the quality of the components in the
network,andultimatelythecommunicationchannel.Theweakeningofjoints,separationof
junctions,invasionofmoisture,oxidationofmaterials,andcontaminationbydirtanddust
allgiverisetopassiveintermodulationanddesensitizationofthereceiver.
Theresultisacellsiteperformingbelowitsdesignpotential
| TEST METHODS
QuantifyingpassiveIMperformance
Good communication quality requires maintaining an acceptable CarriertoInterference
(C/I)ratio.Therefore,thegoalistokeepIaslowaspossible.Intheidealcase,Iwould
always be below the receiver noise floor. One source of undesired interference is passive
intermodulation.
A typical specification requires a passive IM level no greater than 110dBm when two
+43dBm carriers are injected into the device under test, i.e., 153dBc. To put this into
perspective, this is a ratio of 1:2,000,000,000,000,000, or the equivalent of trying to
measurethedistancetothesuntoanaccuracyofonetenthofamillimeter.
www.kaelus.com
Theoldway
A commonly used approach to measuring passive IM is to inject two fixed frequency CW
tones of 20W per tone into the device under test and measure the IM power levels
generated. This test scenario is consistent with that proposed by the international
standardscommitteetaskedwithdefiningtestmethodsforthemeasurementofpassiveIM
(IECTC46WG6).
To realize the necessary test setup, the classic approach has been to rack and stack two
synthesizersfeedingtwohighpoweramplifiersthatconnecttoanarrayofRFcomponents,
whichcombine,filterandduplexthesignals,andthenroutethesignalofinterestthrougha
lownoiseamplifiertoaspectrumanalyzerfordetectionanddisplay.Apowermeterisused
tosetthepropertransmitpowersandthenreadjustonaperiodicbasistocompensatefor
drift.
TherackandstackapproachtobuildingapassiveIMtestbenchisclearlyadifficultprocess.
Furthermore, because there are so many discrete instruments, components and
interconnectingcables,themeasurementresultsaresometimesdifficulttorepeat,andthe
setups are often unstable and susceptible to damage. As if that werent enough to
discourage all but the stouthearted, the measurement is timeconsuming, and as we are
beginningtorealize,theresultscanbeinconclusiveandmisleading.
TheKaelusWay
WithanunderstandingofhowpassiveIMtestingwasperformed(andtheassociated
deficiencies),Kaelussetforththefollowingdesigngoals:
Buildatestinstrumentthatproduceshighquality,repeatablemeasurements
Staycurrentwithemergingtechnologytrendsandteststandards
Maketheinstrumenteasytosetupanduse
Advancethestateoftheartbyprovidingnewandusefultesttools
Bringstandardizationtothisdifficultmeasurementinordertoestablishaclearbasis
ofperformancecomparison
Listentothecustomerandincorporatetheirrecommendationsasappropriateto
providethemostcomprehensiveandeasytousetestinstrumentpossible
This led to the three significant design features incorporated into Kaelus Passive IM
Analyzers:
1. Ahighlyintegrateddesign:Minimizethenumberofexternalcomponentsbypackaging
all of the RF components in a single enclosure. This results in improved reliability,
stabilityandrepeatability.
www.kaelus.com
2. Highspeeddigitalreceivertechnology.ThespeedandthefrequencyagilityoftheKaelus
analyzersmakeitpossibleto:
MeasuremultipleIMproductssimultaneously
MeasureandrecordtransienteventssuchasPIMburstandchangingIMdueto
mechanicalandenvironmentalstress
Measuresweptfrequencyintermodulationcharacteristicstoverifyperformanceover
theentirespecifiedfrequencyrange
3. Innovativenewtestcapabilities:Theresultinginstrumentiseasytouse,timeefficientand
thorough, so that it is practical to make the measurements necessary to assure product
qualityandnetworkintegrity.
| REVEALING REALITY
Most measurements are made with a single pair of carrier frequencies, usually set at the
transmitbandedgesforthewirelessstandardyouweretesting.
For example, consider the measurement of a device used in a PCS1900 Band network.
Typically,thistestisperformedbysettingCarrier1to1930MHzandCarrier2to1990MHz.
(Note:TherearesomePIMtestsetupsthat,duetotheirdesign,mustuseonecarriersetto
afrequencyinthereceivebandwhiletheothercarrierissettoafrequencyinthetransmit
band.Thevalidityofthisapproachopentodebate.)
After the transmit frequencies and powers are established and the spectrum analyzer
configuredtomeasuretheIMresponse,thetestisinitiated.Duringthemeasurement,care
is often taken not to disturb the device under test or the test setup in case instabilities
mightproduceadataaspikeattheprecisemomentthespectrumanalyzerhappenstobe
sweepingpasttheIMfrequency.
ExperiencesuggeststhatthisapproachtomeasuringPIMismuchtoolimitedandleavestoo
manypotentialproblemshidden.
DynamicMeasurements
Despitethebesteffortstodesignandmanufactureproductsthatarestableregardlessof
their environment, the passive IM response of components and subsystems can vary
dramatically when subjected to stress. In a dynamic measurement, IM performance is
monitoredduringtheapplicationofanappropriatestimulus.
Theissueofdynamicmeasurementshasbeengivenagreatdealofattentionwithregardto
cable assemblies. The concern has to do with the vulnerability of the connector/cable
interfaceaswellasIMcreatedinthecable(bymicrocracksinsolidconductorcablesand
2000 Kaelus Inc. All Rights Reserved.
www.kaelus.com
discontinuouscontactinbraidedcables).TestinginvolvesmeasuringtheIMasthecableis
flexedand/orabendingmomentisappliedattheconnector/cableinterface.
In our view, all components on the base station transmit side of the
wireless network should be subjected to this scrutiny. Two
measurements routinely performed by Kaelus on all PIM critical
componentsusedinouranalyzersarethetaptestandthebending
momenttest.Thetaptestissimplyamatteroftappingthedeviceand
watchingtheIMresponse.Forexample,tappingthetuningscrewson
filtersfrequentlygenerateshighlevelsofpassiveIM.Afterthetapping
is stopped, the IM sometimes returns to its low IM condition and
sometimesitremainshigh.Thetaptesthasbeendemonstratedtobe
highly useful in screening devices and cables that will fail at some
future(andinconvenient)time.
Twomeasurements
routinelyperformed
byKaelusonallPIM
criticalcomponents
usedinouranalyzers
arethetaptest
and,thebending
momenttest.
Thisisillustratedinthefollowingexample.Inthiscase,weusedtheStripChartModeofthe
Kaelus analyzer to record the IM response over time. The device under test is a PCS1900
bandpassfilter.WhatweareobservingisthechangingIMduetotapping.Latertestingalso
showedthisdevicetohavemuchdegradedIMperformancewithtemperaturechanges.
TM
-60
-70
-80
-90
-100
-110
-120
-130
0
10
www.kaelus.com
15
20
25
30
Time, Seconds
Figure 1: Dynamic IM Measurement
35
40
45
Thebendingmomenttestisperformedbyapplyingamodestsideforcetotheconnectoron
the device under test. If the connector is not adequately attached to the body of the
component, or if the launching mechanism within the device is not solid, IM will be
generated.
SweptFrequencyMeasurements
Until now, the accepted method for measuring passive IM was to set the carrier to two
fixed frequencies, usually, but not always in the transmit band, and measure the IM
generated.Whathasbeendiscoveredisthatonmanydevices,thisisnotadequate,because
thesedevicesexhibitIMcharacteristicsthatvaryasafunctionoffrequency.KaelusPassive
IMAnalyzersmakethismeasurementatmultiplefrequenciespracticalandeasy.
An example clearly demonstrates the virtue of the swept frequency IM measurement. In
thiscase,aPCS1900bandduplexerwithabadconnectorismeasured.Itshouldbenoted
thattheconnectordamagewascausedbyovertorquingtheconnectorandisundetected
by visual inspection and swept return loss measurements. The comparison between the
goodandthebadduplexerreturnlossisshowninthefollowingfigure.
TM
-5.00
Return Loss, dB
-10.00
-15.00
-20.00
-25.00
-30.00
-35.00
-40.00
1850
1860
1870
1880
1890
1900
1910
1920
1930
Frequency, MHz
www.kaelus.com
1940
1950
1960
1970
1980
1990
The good duplexer meets the 115dBm specification at all frequencies, but the bad
duplexerhasaseverelydegradedIMasthetranmsitcarriersaremovedfromthebandedges.
BothduplexerswouldpassQAscreeningifonlybandedgecarrierswereusedforthetesting.
Tofullycharacterizethedeviceundertest,thesweptfrequencycapabilityoftheKaelusPassiveIM
AnalyzerrecordstheIMperformanceaseachofthetwocarriersissweptinfrequencyacrossthe
transmit(downlink)band.Onlyonecarrieratatimeisswept,whiletheremainingcarrierisheld
fixedatthebandedge.Intheplotbelow,Carrier1isfixedat1930MHzwhileCarrier2issweptfrom
1950to1990MHzproducingthirdorderintermodulationproducts(IM3)thatoccuratfrequencies
from1870to1910MHz.
Themeasureddatacanbeloggedtoafile,copiedtothecomputerclipboardwhereitcanbepasted
intoanotherapplication,oritcanbedirectlyhardcopiedtoaprinter.
| SUMMARY
Withthewirelessserviceprovidersworkinghardtowinandmaintaincustomers,controllingthe
generationofpassiveintermodulationdistortioncanimpactthefinancialsuccessoftheproviders.
ThepresenceofpassiveIMatacellsiteimpactsbasestationcapacityandservicequality.Therefore,
itmustbeconsidered,measuredandcontrolledfromdesignthroughmanufacturing,andatthebase
stationinordertorealizeoptimumperformance.
TM
Good Duplexer
-95
-100
-105
-110
-115
-120
-125
-130
1870
1875
1880
1885
1890
1895
www.kaelus.com
1900
1905
1910