蒙特卡洛分析

MonteCarlosimulation……forbetteryieldandperformance--AtutorialstartSystemrequirementStatisticalanalysisincludeprocess,mismatcheffectsInitialdesignDesignmeetsthegoal?endNOYESMonteCarlosimulation……forbetteryieldandperformanceSomedesignmaydegradeinperformanceOveralldesignyieldcouldbeunexpectedlylowIffabricationprocessparameteranddevicemismatcheffectonsamediearenottakenintoaccountthenHencestatisticalanalysismustfindahighplaceindesigncycleWewillperformMonteCarloanalysisonanRF-frontendLNAandcomparetheresultifnostatisticalanalysisisdone.WewillalsoseehowtoanalyzeyieldandscalardatainMonteCarlowiththehelpofLowpassfilterexample.MonteCarlosimulationMonteCarlosimulation(example)LinearityInputmatchingBiasN/WOutputmatchingCascodearch.toreducefeedbackcapacitanceRF-frontend(LNA)KnowingSystemrequirementInitialdesignbasedonrequirementlikenoise,gain,narroworwideband.MonteCarlosimulationCadencesimulationsetup(Normal)1.Choosingaffirmaanalogartist2.ChoosingSpectresimulatorChoosingmodelfile,whichcontainsallMOS,reg.,capmodelparameters.MonteCarlosimulation1.Choosesetupmodellibraries2.BrowseandchoosemodelfileinthedirectoryCadencesimulationsetup(Normal)Setupanalysis(dc,ac,spetc.),createnetlistandrunsimulatorMonteCarlosimulation1.Chooseanalysistorun2.Chooseoutputtoplot3.CreatenetlistandrunCadencesimulationsetup(Normal)PlottingresultsMonteCarlosimulation1.Choosedirectplotforanalysis2.Clicktoviewthedesiredresult3.AnalyzewaveformCadencesimulationsetup(Normal)MonteCarlomodelinginCadencespectresimulatorProcessSection-describesmanufacturingparameter,theirstatisticalvariationandamodelfordevicethatcalculatesits(width,length,cap,res.Etc.)accordingtoprocessparameter.Design-SpecificSection–designeraccordingtohisneedcanspecifyMonteCarloanalysis.Forexampleinacurrentmirrorcircuit,matchedtransistorsareusedanddesignercangivesomecorrelationfactorbetweenthesematchedtransistor.MonteCarlosimulationCadencesimulationsetup(MonteCarlo)TypicalModelFileProcessSection1.Allparametersetstotheirnominalvalue,nostatisticalvariationdefined2.Model(NMOS’sRg)iscalculatedusingnominalparametervalue21MonteCarlosimulationCadencesimulationsetup(MonteCarlo)Definingprocess,mismatchparameterasstatisticallyassignedvalueAssessesthedevicemismatchondifferentdie,whichcouldhavegonethroughsomedifferentprocessparametersduringfabrication.Assessesthedevicemismatchonsamedie,whichcouldhavegonethroughsomedifferentprocessparameter.VariationdefinedasadistributedfunctionMonteCarlosimulationProcessSectionCadencesimulationsetup(MonteCarlo)DesignSpecificSectionThisincludesthecircuitconnectivity(tworesistors,andcorrespondingcurrentsourcesthatfeedthem)Definingcorrelationbetweentwodevices(R1,R2)††Note:AlternativelythisinformationcanalsobeinsertedthroughArtistMonteCarloTool.MonteCarlosimulationCadencesimulationsetup(MonteCarlo)MonteCarlosimulationModelfileusedforLNAexampleNoteThisisnotbasedonfoundrydatabutmodeledforillustrativepurposes.Cadencesimulationsetup(MonteCarlo)MonteCarlosimulationAfterInitialdesignthatmeetsthesystemrequirement,statisticalanalysismusthavetobecarriedout.1.Makesuretheadditionofprocessandmismatchparametersectioninmodelfile.2.Makecertaintoincludetheparticularsection(forexa.Statsinspectre)insimulationmodellibrary3.GototoolMonteCarloinaffirmaanalogartistCadencesimulationsetup(MonteCarlo)MonteCarlosimulationChoosenoofiteration(default=100)1.ChoosewhichvariationtoincludeProcessdevicemismatcheffectontwodiff.dieMismatchdevicemismatcheffectonsamedie2.Clickifyouwanttoseethefamilyofcurvei.e.curvefromeachiteration3.Definetheexpressions/signalsonwhichMonteCarloanalysiswillbeperformed.Note:calculatorcanalsobeusedtogettheseexpressionFinallyruntheanalysisCadencesimulationsetup(MonteCarlo)MonteCarlosimulation(Analyzingwaveform)MatchingNormalsimulation(withoutstatisticalvariation)MonteCarloSimulation(withstatisticalvariation)ProcessparameterandmismatcheffectInput&OutputmatchingN/WDEGRADESOveralldesignperformance(noise,gainetc.)DEGRADESS11S22MonteCarlosimulation(Analyzingwaveform)Matching(VSWR):IttellshowwellinputandoutputN/Warematched.MonteCarlosimulationNormalsimulationVSWR1VSWR2VariationsinVSWRMonteCarlosimulation(Analyzingwaveform)Matching(forwardandreversetransmissiongain)Ithasdeterioratedtheperformancesignificantly,asaminimumS12andmaximumS21valueisdesirable.NormalsimulationMonteCarlosimulationS12S21MonteCarlosimulation(Analyzingwaveform)NormalsimulationMonteCarlosimulationStability:AKfvalue1,isdesiredforanstableamplifierKfvaluehasbecome1,andconsequentlycreatingapotentialunstability,hencealargemarginisrequiredatinitialdesignphase.StabilityfactorMonteCarlosimulation(Analyzingwaveform)NormalsimulationMonteCarlosimulationNoisePerformanceAsvisible,designhasarobustnoiseperformanceatdesiredband(2.4-2.5GHz)BUT….NoisefigurePeriodicOutputnoiseMonteCarlosimulation(Analyzingwaveform)But…LNAasanRF-frontendhastoprovideenoughgainwithmaximumnoisesuppressiontomaintainanallowableSNRatdemodulator'sinput.ItfailstomeetthegainrequirementGainMonteCarlosimulationWewillquicklygooveranotherexampleofl