ImprovetheAccuracyofYourReliabilityPredictions!
ThegeneralpurposeofthecalculationmethodsoutlinedintheTelcordia(Bellcore)document,PredictionProcedureforElectronicEquipment,istotakeintoaccountadditionalinformationaboutthedevices,units,andsystemsunderanalysis.Thesecalculationmethodsconsidervariousburn-in,field,andlaboratorytestdata.Theyalsoprovideforcalculatingtheinfantmortalityrate,orfirst-yeardropout,forthesystemsunderanalysis.However,thisparticulartopicisbeyondthescopeofthisbrief.
BeforereviewingthethreeTelcordiacalculationmethods,itisimportantthatyouunderstandthebasicterminologythatisusedtodescribethem.ThefollowingtabledefinestermsusedwithintheTelcordia(Bellcore)document.
[table=98%][tr][td=1,1,140]Term[/td][td=1,1,660]Definition[/td][/tr][tr][td=1,1,140]Device
[/td][td=1,1,660]Anelectricalpartwithwell-definedelectricalcharacteristics.Devicesincludeintegratedcircuits,diodes,resistors,andmore.
[/td][/tr][tr][td=1,1,140]Unit
[/td][td=1,1,660]Anassemblyofdevicestypicallyatthelowestreplaceablelevel.Unitsincludecircuitpacks,modules,powersupplies,plug-indevices,andmore.
[/td][/tr][tr][td=1,1,140]System
[/td][td=1,1,660]Acompleteassemblythatperformsanoperationalfunction.
[/td][/tr][tr][td=1,1,140]Steady-StateFailureRate
[/td][td=1,1,660]Theconstantfailurerateafteroneyearofoperation,providinginformationaboutlong-termproductperformance.
[/td][/tr][tr][td=1,1,140]Burn-in
[/td][td=1,1,660]Theoperationofadeviceunderacceleratedtemperatureorotherstressconditionstostabilizeitsperformance.
[/td][/tr][/table]
YourchoiceofaparticularTelcordiacalculationmethoddependsuponyouranalysisrequirementsandtheamountofdataavailable.Generaldescriptionsofeachmethodanditsdifferentcasesareprovidedintheremainingpagesofthisbrief.Formoredetailedinformation,pleaserefertotheTelcordia(Bellcore)documentitself.
Note:TheterminologyfrompreviousBellcorestandards—MethodI,MethodII,andMethodIII—isreplacedinTelcordiaIssue1withBlackBoxTechnique,BlackBoxTechniqueIntegratedwithLaboratoryData,andBlackBoxTechniqueIntegratedwithFieldData,respectively.Theunderlyingcalculationproceduresarethesame;onlytheterminologyhasbeenupdatedtopromotebetterunderstanding.Althoughthesectionsthatfollownotebothsetsofterminologyintheheading,theMethodI,II,andIIIterminologyisusedwithinthesetopicsforsuccinctness.
MethodI(BlackBox)
MethodIisgenerallyreferredtoasaPartsCountmethodbecausethesteady-statefailurerateforaunitisassumedtobethesumofthesteady-statefailureratesforitsdevices.BecauseMethod1isbasedongenericfailuredataforvariousdevicetypes,itisusedwhenspecificpartdataisunavailable.
InthepreviousBellcorestandards,threedifferentcasesofMethodIaredefined.Case1andCase2,whichareverygeneral,bothassumethatoperatingtemperatureis40degreesCelsiusandratedstressis50percent.Case1assumesaburn-inoflessthanorequalto1hour,whileCase2assumesaburn-inofgreaterthan1hour.Case3providesfortheuseofvariabletemperatureandstressvalues.Case1issimilartotheMIL-HDBK-217PartsCountmethodology,andCase3issimilartotheMIL-HDBK-217PartsStressmethodology.
[table=98%][tr][td=1,1,70]Case[/td][td=1,1,630]Description[/td][/tr][tr][td=1,1,70]1[/td][td=1,1,630]PredictionsarebasedontheBlackBoxoptionwithunit/systemburn-in1hour.Thereisnodeviceburn-in.Alldevicesareassumedtobeoperatingat40degreesCelsiusandat50percentratedelectricalstress.
[/td][/tr][tr][td=1,1,70]3[/td][td=1,1,630]ThisisthegeneralcaseforMethodIreliabilitypredictions.Itcantakeintoaccountdeviceburn-inandvaryingtemperatureandstressdata.
[/td][/tr][/table]
MethodII(BlackBoxIntegratedwithLaboratoryData)
ThepurposeofMethodIIistoadjustthepredictedMTBF(MeanTimeBetweenFailures)ofaunitordevicebasedonavailablelaboratoryortestdata.UsingMethodII,thesteady-statefailurerateiscalculatedasaweightedaverageofthemeasuredlaboratoryfailurerateandtheMethodIgenericfailurerate,withtheweightsdeterminedbythelaboratorydata.
Whenlaboratorytestsareveryinformative,theMethodIIbasefailurerateisheavilyinfluencedbythelaboratorydata.Whenlaboratorytestsarelessinformative,theMethodIIbasefailurerateisheavilyinfluencedbytheMethodIgenericfailurerate.Thefactorstakenintoconsiderationintheweightingoflaboratorydataincludethenumberofdevicefailuresduringlaboratorytest,thenumberofdevicestested,theactualtimedevicesweretested,andthetemperatureaccelerationduringtest.
Whenlaboratorydataisincluded,thecalculationsforpredictingsteady-statefailureratesaredependentuponwhetherdevicesorunitshavehadpreviousburn-in.InthepreviousBellcorestandards,fourdifferentcasesofMethodIIaredefined.Thetablebelowdescribeseachofthesecases.
[table=98%][tr][td=1,1,70]Case[/td][td=1,1,630]Description[/td][/tr][tr][td=1,1,70]L1[/td][td=1,1,630]Devicesarelaboratorytestedandhavenoburn-in.
[/td][/tr][tr][td=1,1,70]L2[/td][td=1,1,630]Unitsarelaboratorytestedandhavenounit/deviceburn-in.
[/td][/tr][tr][td=1,1,70]L3[/td][td=1,1,630]Devicesarelaboratorytestedandhavehadburn-in.
[/td][/tr][tr][td=1,1,70]L4[/td][td=1,1,630]Unitsarelaboratorytestedandhavehadunit/deviceburn-in.
[/td][/tr][/table]
MethodIII(BlackBoxIntegratedwithFieldData)
ThepurposeofMethodIIIistoadjustthepredictedMTBFofaunitordevicebasedonfielddata.MethodIIIiscalculatedasaweightedaverageoftheobservedfieldfailurerateandtheMethodIgenericfailurerate.Thenumberoftotaloperatinghoursduringfieldtestingdeterminestheweights.
Whenthenumberoftotaloperatinghoursislarge,theMethodIIIbasefailurerateisheavilyinfluencedbythefielddata.Whenthenumberoftotaloperatinghoursissmall,theMethodIIIbasefailurerateisheavilyinfluencedbytheMethodIprediction.
InpreviousBellcorestandards,threedifferentcasesofMethodIIIaredefined.Thefollowingtabledescribeseachofthesecases.
[table=98%][tr][td=1,1,70]Case[/td][td=1,1,630]Description[/td][/tr][tr][td=1,1,70]III(a)[/td][td=1,1,630]Providesfailureratepredictionsfordevices,units,orsubsystemsbasedonactualin-serviceperformance.
[/td][/tr][tr][td=1,1,70]III(b)[/td][td=1,1,630]Providesfailureratepredictionsfordevices,units,orsubsystemsbasedonin-serviceperformanceaspartofanothersystem.Adjustmentsaremadetotheseestimatestotakeintoaccountdifferencesinoperatingconditionsandenvironments.
[/td][/tr][tr][td=1,1,70]III(c)[/td][td=1,1,630]Providesfailureratepredictionsfordevices,units,orsubsystemsbasedonthein-serviceperformanceofsimilarequipmentfromthesamemanufacturer.Adjustmentsaremadetotheseestimatestotakeintoaccountdifferencesbetweentheoperatingconditionsandenvironmentsofthetwosystems.
[/td][/tr][/table]
TheTelcordiacalculationmethodsdescribedinthisdocumentaresupportedintheRelexReliabilityPredictionmodule.Inadditiontosupportingawiderangeofpredictionmodels,includingMIL-HDBK-217,Telcordia,RDF,etc.,RelexReliabilityPredictionallowstheTelcordiacalculationmethodstobeappliedtoanylinear-basedmodel.ThismeansthatyoucanadjustthefailureratesforvirtuallyanyreliabilitypredictionmodeltoaccountforlaboratoryandfielddatabyusingtheseTelcordiacalculationmethods.
