英文版培訓(xùn)課件：HSPICE –Highlights and Introductions

HSPICE–HighlightsandIntroductionsTechniques

forSI09-17-0312/4/2002FeaturesUseforSI2?

Parameters?

Alters?

Libraries?

Syntaxbased–NOGUI?

SelfdocumentingASCIInodenames?

VoltageControlledResistor?

MonteCarlo?

Nodeequationbasedsource?

PWLlinearbasedsource?

Nodalmeasurementproducemeasurementfiles?

AccurateTransmissionlineswithWelements?

FrequencydependanttransmissionlinesTransient?

IBISbuffersHSPICEforSI12/4/2002GoodPractices3?Modularizewithsub-circuitsand/orlibraries!?Circuittextshouldflowlineadrawing.Don’tputallcaps,resistors,andtransmissionlinesrespectiveseparatesections.?MostSIcircuitsarecomposedofdatagenerator,buffers,transmissionlines,packagemodels,andconnectors.HSPICEforSI12/4/20024Global,Local,andPosition?

Circuitelementswithinasub-circuitorthemainnetlistarepositioninsensitive.Good-news/badnewsItiseasiertofollowelementswhosecodetracesoutthecircuit.?

Ingeneralparametersareglobalunlesspassedintoasub-circuit.?

ParametersarenotpositionsinsensitiveTreatdefinitionofparametersaslastreferencewinsthedefinition.Thiscanbetricktodetermineforcomplexdecks.Deckisoldterminologythatcomesform“punchcarddecks”?

Makethefirst8charactersoflibrarynamesunique?

MostHSPICEiscaseinsensitive.Theexceptionislibrariesand

filenamesthatareenclosedinsinglequotesHSPICEforSI12/4/20025TopLevelProgram?

Firststepistodrawassimulationblockdiagram.?

Thefollowingslidesarealearnbyexamplemethod?

WewillreviewsomecommonHSPICEelementsusedforsignalintegrityPrintedWiringBuffersReceiverBoardHSPICEforSI12/4/20026StructureI–WewillusethisonefornowProducessetoftr0files,etc.ParametersAndinstantiationsLibrariesSubcircuitsSubcircuitsSubcircuitsMainNetlistLibrariesLibrariesAltersWe’lluseaneattrickandkeepallthisinonefile.TransmissionTransmissionTrLainsemissionLineRLCGLfiinleeRLCGfileLargeprogramsmayusemanyfilesRLCGfileHSPICEforSI12/4/20027StructureIIProducestr0,tr1,tr2,…

filesetc.ParametersParametersSetParametersSetParametersSetSubcircuitsSubcircuitsSubcircuitsSetInstantiationInstantiationSetInstantiationSetInstantiationSetMainNetlistSetMainNetlistMainNetlistMainNetlistTransmissionTransmissionTrLainsemissionLineEachSimulationcaseisadifferentcatalogedfileRLCGLfiinleeRLCGfileRLCGfileHSPICEforSI12/4/20028StructureIIIParametersAndinstantiationsLibrariesLibrariesSubcircuitsSubcircuitsSubcircuitsMainNetlistLibrariesSweepparametersProducessingletr0file,etc.butmutlipleTransmissionTransmissionTrLainsemissionLineRLCGLfiinleeRLCGfileRLCGfilewaveformsperfileHSPICEforSI12/4/2002Runningthenetlist9?

ClickingonsimulatewillcreatethefollowingfilesTransientanalysisnodalfile–testckt.tr0ThisiswherethewaveformdataisMeasurementresultsfile–testckt.mt0Listfile–testckt.lisEditthistodebugerrorsInitialconditionsfile–testckt.in0Sub-circuitcrossreferencelist–testckt.pa0Outputstatusfile–testckt.st0HSPICEforSI12/4/200210DataGenerator.LIB.SUBCKT'pulse'DATAS

bit1bit2

datarate=-1V1bit10PULSE0v1v0n0.5n0.5n'datarate-0.5n''2*datarate'V2bit20PULSE0v1v0n0.5n0.5n'datarate-0.5n''2*datarate'.ENDS.ENDL?

Bit1andBit2aredatastreamoutputsforthissub-circuit?

“datarate”ispassedfromthecallsiteNotethatasubcircuitisanalogoustoasoftwaresubroutine”datarate”issetto“-1”toforce

anerroriftheparameterwasnotpassed.?

Thispulsegeneratorexampleproducesa1VAggressorandvictimw/500psrise/falltime.Thepulsewidthis“datarate”

adjustedbytherisetime.Theperiodis2*datarateThisspecialcaseuses0vand1vasabitstreamwhichhasadvantagesthatwewilllearnlaterinbehavioralmodeling.HSPICEforSI12/4/2002ParameterizedGenerator11.LIB.SUBCKT'pulse'DATAS

bit1

bit2

datarate=-1V1bit10PULSE0v1v0ntrtr'datarate-tr''2*datarate'V2bit20PULSE0v1v0ntrtr'datarate-tr''2*datarate'.ENDS.ENDL?Wecanreplacethe0.5nentrieswithaparametercalledtr.(equalrise/fall)?Wecansetthisparameterinthemainnetlistasfollows:.PARAMtr=0.5n?NoticethedifferencebetweenthetwoparameterstranddatarateHSPICEforSI12/4/200212SquareWaveinPreviousThisishowthepulsesourcefunctiondefinespwSource’spulsewidth0.5nsdataratedatarate2*datarate1V0V0.5nsHSPICEforSI12/4/200213PieceWiseLinearSourcebit1,tr*1bit1,UI*1bit3,UI*2-trbit2,UI*2bit2,UI*1-trVol,0S*rise/falltime=trHSPICEforSI12/4/2002Assignment14?CreatesamedriverwithaPWLsourceandwithdatapattern“101100110”.?Assumeallparameterexceptdatarateareglobal?Parameterizebitsasbit0,bit1,bit2…?ParameterforriseandfalltimewithasignalparameterTr?Writeseparatecodeforparameterstatementsinthemainnetlist.HSPICEforSI12/4/2002DriverSub-circuit15.LIB'driver'.SUBCKT

MYBUFEdrive

out1

VssRoutout1out50CoutoutVss1p.ENDSininoutVss0VOL='(Voh-Vol)*V(in)+Vol‘.ENDL?

Thisexamplejustusesthebitsonnode“in”andcreatesanoutputvoltagewithVolandVolonenode“out”.?

VolandVohareglobalinthiscasebecausetheywerenotpassed?

Thisexampleusesaequationcontrolledvoltagesource.Thisaverypowerfulfeature.Teqhueateiqounaitsio.

nisenclosedinquotesmuchthesamewhyaparameter?

ThisentiresubcircuitcanbereplacedatalatertimewithatransistorbasedbuffermodeloranIBISmodel.?

Thesourceimpedanceinthiscaseis50ohmswithapFacrosstheoutputterminal“out”HSPICEforSI12/4/2002ParameterizeDriver16.LIB'driver‘MYBUF

inout1

Vss

VOL='(Voh-Vol)*V(in)+Vol‘out1

outout1

Vss.SUBCKTEdriveRoutoutVssTx_rtermTx_ctermCoutRininvss1G.ENDS.ENDL?

Changethesourceterminationsintoparameters:Tx_rtermandRx_cterm?

AsagoodpracticeplaceahiimpedanceDCpathacrossinput.Thiscanavoidtransienterrors.?

Wecansetthisparameterinthemainnetlistasfollows:.PARAMTx_rterm=50Tx_Cterm=1pFHSPICEforSI12/4/2002PackageSub-circuit17.LIB.SUBCKT'LC_pack'PKG

in1

in2

out1

out2

VssL1

in1

out1

1nL2

in2

out2

1nK1

L1

L2

0.2.ENDS.ENDL?Thisisasimplepackagethatusesacoupledinductorcircuit.?OftenthissubcircuitismorecomplexandderivefromtoolslikeAnsoftHSPICEforSI12/4/200218CoupledInductorsL1in1out1L12KL1?L2WhereL12isthemutualinductancebetweeninductorL1andL2L2in2out2HSPICEforSI12/4/200219Printedwiringboardmodeling.LIB'easy_lines‘BRD

in1in2

VssRLGCFILE=

‘s5_z068.9_z0d108.8'

N=2

L=0.1.SUBCKTWline1+in2

out1

out2

Vssout1

out2

Vssin1.ENDS.ENDL?

BoardetchescanbeaccuratelymodeledwithW-elements.?

Forthatcaseweusecoupledtransmissionlinesfortheboardtraces?

Thefile‘s5_z068.9_z0d108.8.rlc’containsthetransmissionlinecharacteristics.ThisdatamaybecreatedwithinternalHSPICE2-Dfieldsolveroranyother2-DfieldsolversuchasAnsoft.?

Thesymbol“+”isacontinuationline.?

Oftenthissubcircuitcanbecomequitesubstantialcontainingmanytransmissionlinesandboardfeaturesmodeledaspassiveelements.HSPICEforSI12/4/2002W-element:ModelReference20.LIB'easy_lines‘BRD

in1in1

in2

Vss

out1

out2

VssRLGCMODEL=

's5_z068.9_z0d108.8'

N=2

L=.1.SUBCKTWline1in2out1

out2

Vss+.ENDS.ENDL?

Additionallyamodelstatementcanbeused

tospecifyRLGCdata.HSPICEforSI12/4/2002TransmissionLineW-Element21Wline1in1

in2

Vss

out1

out2

Vss+RLGCMODEL=‘s5_z068.9_z0d108.8‘N=2

L=0.1in1in2Vssout1out2Vss?

Thegeneralsyntaxsupportanynumberofinputandequalnumberofoutputport.?

Thislengthinthisexampleis0.1?

TheunitsaretheoftenassumedtobemeterbutactuallyaretheperlengthunitsoftheRLCGmodel.TheinternalfieldsolverproducesunitsinmetersHSPICEforSI12/4/200222Creatingafieldsolution?Createafilethatinvokesthetargettransmissionlines.?Inthisfilealsospecify:FieldsolveroptionsMaterialsStackupThedielectricandpower/groundconductorplanesandwichofaPWBTracegeometriesshapesTheamodelsthatincludetheaboveHSPICEforSI12/4/200223CoupleStripLineExample(twolines.sp)tgσefεr,tanδbσsσtwhσtg.TitleFieldSolverW2+12

0

ab0+Fsmodel=s5_z068.9_z0d108.8N=2L=1DELAYOPT=1*s5w5ef0.5t0.5Tg1*mils*milsconvertedtometers

1.270E-04

1.27E-04

1.27E-051.27E-052.54E-03*b0.5h10ertand

uconduct.4.2E+07*mils3.90

.02

1*milsconvertedtometer5.207E-04

2.54E-04HSPICEforSI12/4/200224UsingtheSolverFirst

line

should

be

comment

or

titleelse

it

gets

ignoredInvoking

the

W

element

will

cause

thefield

solver

to

run,

if

the

FSMODELparameter

is

specified.TitleFieldSolverW2+12

0

ab0+Fsmodel=s5_z068.9_z0d108.8N=2L=1DELAYOPT=1*s5w5ef0.5t0.5Tg1*mils*milsconvertedtometers

1.270E-04

1.27E-04

1.27E-051.27E-052.54E-03*b0.5h10ertand

uconduct.4.2E+07*mils3.90

.02

1*milsconvertedtometer5.207E-04

2.54E-04(cont’donnextpage)Often

this

is

done

outside

the

main

net

list

toinsure

solution

quality.

Then

a

RLGCMODEL

orRLGCFILE

statement

would

be

used

here

insteadHSPICEforSI12/4/2002FieldSolverOptionStatement25.FSoptions

brd_opt2ACCURACY=HIGHGRIDFACTOR=1+ComputeRo=yesComputeRs=yesComputeGo=yesComputeGd=yesPRINTDATA=yes*.MATERIAL.MATERIAL*brd_dielct2brd_cu2DIELECTRICMETALER=3.9,LOSSTANGENT=0.019CONDUCTIVITY=42000000.LAYERSTACKbrd_ssl_stk2+LAYER=(

brd_cu2+LAYER=(

brd_dielct2+LAYER=(

brd_cu2,1.2700E-05),5.2070E-04),1.2700E-05)?

Thisstatementisgoodstarttosetupoptions.?

Inthiscasetheoptionsarecalledbrd_opt2?

Theboardismaterial“brd_dielect2”?

Theconductormaterialis“brd_cu2”?

Noticethatthetracesarenotspecifiedhere.?

Thestackupactuallystartsatthebottomandworksup.Eachlayer

thicknessisspecified.HSPICEforSI12/4/200226MaterialandShapes.MATERIAL

brd_dielct2

DIELECTRIC

ER=3.9,LOSSTANGENT=0.019.MATERIAL

brd_cu2.SHAPE

brd_trap1POLYGON

VERTEX=+(0

0

1.2700E-051.2700E-041.1430E-041.2700E-041.2700E-040

)METALCONDUCTIVITY=420000001.27E-05

1.27E-040

01.143E-04

1.27E-041.27E-04

0origin?

Nextspecifythepropertiesofthedielectricsandmetal?

Thenspecifytheshape.?

Afirstpassguessoftenusesrectanglefortrace.?

InthisexampleweuseatrapezoidHSPICEforSI12/4/200227TheModelstatement.MODEL

s5_z068.9_z0d108.8+W

MODELTYPE=FieldSolver,LAYERSTACK=brd_ssl_stk2

FSoptions=brd_opt2+CONDUCTOR=(SHAPE=brd_trap2

MATERIAL=brd_cu2+ORIGIN=(

6.3500E-05,2.6670E-04)+CONDUCTOR=(SHAPE=brd_trap2

MATERIAL=brd_cu2ORIGIN=(-1.9050E-04,2.6670E-04)++RLGCfile=s5_z068.9_z0d108.8.rlc.END?Herethefieldsolvercallsoutwhatwasspecified.brd_ssl_stk2brd_opt2brd_cu2brd_trap2HSPICEforSI12/4/200228PlacingtheShapestgσ1.27E-03bef5.2070E-041.270E-01.27E-03st1.27E-041.27E-046.35E-05h-1.9050E-042.6670E-04tg2.54E-030,0?

Theoriginforthesolutionisatthebottomofthestackup?

ThepositioningofthetrapezoidswiththestackupareinrelationtotheshapeoriginHSPICEforSI12/4/2002TheRLCGModel29.MODEL

s5_z068.9_z0d108.8

W

MODELTYPE=RLGC,

N=2+

Lo

=

4.460644e-007+9.544025e-008

4.460644e-007+

Co

=

1.019475e-010+-2.181277e-011

1.019475e-010+

Ro

=

1.637366e+001+0.000000e+000

1.637366e+0011.019475×

10?

10

?2.181277×

1011

?

+

Go

=

0.000000e+000+0.000000e+000

0.000000e+000

+9.268906e-005

2.056651e-003?2.181277×

10?

11

1.019475×

10?

10+

Rs

=

2.056598e-003

+

Gd

=

1.217055e-011+-2.604020e-012

1.217055e-011.ENDS?

Onlyhalfofthediagonalandthelowerhalfofthematrixisspecified?

DefaultunitsareH/m,F/m,?/m,S/m,?/(m*srqt(Hz),S/(m*Hz)respectively?

AlternativelyH/in,F/in,?/in,

S/in,?/(in*srqt(Hz),S/(in*Hz)canbeusedifLunitsaretobespecifiedininches.?

AmoredetaileddescriptioncanbefoundintheHSPICEtransmissionlinechapterHSPICEforSI12/4/200230TlineissuesforSIengineers?Validationoftransmissionlinemodels?Comparisontoequations.Mostequationareonlyaccuratetoafewohmsandhavearelimitedtoonlycertainratiosoftracegeometry?Differentialimpedanceequationsarenotreadilyavailable.?ToolstocomparetomeasurementVectorNetworkAnalyzerTimeDomainRefectometryHSPICEforSI12/4/200231Receiver.LIB'receiver'.SUBCKT

RCV

in

VssRin

in

Vss

45Cin

in

Vss

0.5pf.ENDS.ENDL?ThistoocouldbemorecomplicatedtransistororIBIScircuit.?Inthiscasewestartwith45ohmstogroundwitha0.5pFshuntacrosstheload.HSPICEforSI12/4/200232Themainnetlist–TopHalf?

ThelibrarieswillgoattheendforthisexampleInfactalloftheabovestatementsarepositionindependentalthoughparameterusageispositionsensitive.Becarefulifparameteraresetinlibraries.Thiscaneffecttheorderofparameterprocessing.?

Thelibrariesarenormallyintheanotherfile.Thisexampleisnotstandardpracticebutitisconvenientforcollaboratingonissues.?

TheglobalparameterforthebitintervalUIissetto10nanoseconds.?

Twomoreglobalparametersareused

forbuffervoltagecontrol,VolandVoh.?

ThetransientstatementtellsHspiceto

startatransientanalysiswhenthe“.end”statementisprocessed.Inthiscasethetimestepintervalis10psandwillstopat20ns.HSPICEforSI12/4/200233Helpfulhintstoresolvetimesteperrors?

VoltagetransitionsthataretoofastConsiderslowertransitiontime?

Un-initializedreactivecomponentscancaseinstantaneousspikesthatcreateveryfasttransitionsbeforesetting.Considersetting“IC”

(initialcondition.)?

CapacitorsandinductorsthataretoosmallConsidereliminatingorcombiningConsiderputtingshuntresistoracrossdevice?

Floatingreferencesornodescancausetimesteperrors.DCpathcan’tbedeterminedifswitchesorcontrolledsourcesareusedandmaybeconsideredfloatingattimet=0Providehighresistanceshuntstonode0?

Transmissionlinesthataretooshort.ConsiderreplacingwithLC?

Switchescancausespikes.Usevoltagecontrolledresistortosoftenopenandcloseresistanceasfunctionoftime.(moreonthislater)?

Smallmutual“k”elements.Considerelimiatingsamekelements.HSPICEforSI12/4/2002MainNetList

FlowsLikeaCircuit34?

“$”isacommentaftercolumn1?

“*”incolumn1commentsthatline?

Ifan.optionprobecontrolstatementisusedonthenodesdata_vandpkg2_vwillbe

storedinthetr0file.HSPICEforSI12/4/200235Assignment?

TakethepreviousHSPICEexampleanddrawacircuitschematic.?

ProducethelastpictureinAvanWaves(ifavailable)?

LookupandreadallchaptersintheHSPICEmanualon:SubcircuitsLibrariesEsourceCoupledinductorW-elementsNoticethisparttotheassignmentislooserthatmostacademicreadingassignments.Inbusinessdata-miningisarequiredskill.Alsolookupany

elementwecoverthatyou

donotunderstand.HSPICEforSI12/4/2002AlookatresultsinAvanWaves36DoubleclickheretoshowdisplaywaveDoubleclickheretoshowsub-circuithierarchy.Noticetheoutputis~.6v…why?HSPICEforSI12/4/2002Measurement37?

Thereisamanualcontainanextensivelistofmeasurementsthatcanbemade.?

Inthiscasewearemakingameasurementcalled“flight_time_v”and“flight_time_a”?

Thetriggerforthebeginningofthemeasurementisat0.5Vonthefirstrisingonnodedata_v(anddata_a.)?

Thecompletionofthemeasurementiswhenthefirstrisingedgeonnodepkg2_v(andpkg2_a)reaches0.5V.?

TDparametermeanstimedelaybeforethemeasurementstartsandis0sinthisexample.HSPICEforSI12/4/200238MT0file?

ThisresultantMT0file?

Thesecondlineisthetitle?

Thethirdlineandallthelinesthatfollowuptothe“alter#”parametersaretheparametersnames.?

ThefollowinglinesarethecorrespondingmeasurementvaluesForthiscasethemeasurementfortheparameters“flight_time_v”and“flight_time_a”arethe955.9ps.HSPICEforSI12/4/2002MonteCarloAnalysis39.TITLE

Signal

integrity

Training

deck*

parameter

variations.param

rx_rterm1=AGAUSS(50,

10,

3)rx_cterm1=AGAUSS(1pf,.8pf,

3).param

rx_rterm=rx_rterm1

rx_cterm=rx_cterm1.param

tx_rterm1=GAUSS(45,

0.1,

3)tx_cterm1=GAUSS(.5pf,0.1,

3).param

tx_rterm=tx_rterm1

tx_cterm=tx_cterm1.param

pkg_coulping1=GAUSS(.2,.5,

3).param

pkg_coulping=pkg_coulping1.param

tr1=AGAUSS(.5ns,.45ns,

3).param

tr=tr1?

Noticeweuseadummyvariable(suffixedwith1).?

Thisisbecauseeverytime

aforexampleRx_term1isuseditwillgetanewvalue.Byassigningitadummyvariableatthebeginningofasweepthevaluewillbesetforthatentiresweep.Elseeachtimethevariableisusedanewvaluewillbeassigned.HSPICEforSI12/4/200240InvokingaMonteCarloSweep?

The.TRANstatementisnewsyntaxaddedtoit“SWEEPMONTE=5000”Thiswillcause5000sweepstobecreatedinthetr0file.?

Optionprobestatementwasaddedsothatonlynodeannotatedwiththe.probestatementwillbestoredsince5000sweepswillcreateaverylargetr0file.HSPICEforSI12/4/200241Afewchangesaddedtotheend?

The“.PROBE”statementisusedinconjunctionwiththe.OPTIONPROBEstatementsoonlynodedata_vandpkg2_varereported.?

Onlyonemeasurementisusedandthethresholdwasloweredto250mvHSPICEforSI12/4/200242The“Sweep”MTOfile?

Noteeachsweepentrycontains

thevaluesthatwereassignedtotheMonteCarloparameters?

AVBAorperlscriptisnormally

used(andrequired)toconvertintoaspreadsheetformatHSPICEforSI12/4/200243ViewingMonteCarloinaSpreadsheet?

Step1:createspreadsheetwithresultcolumn?

Step2:createcellswiththemin,max,mean(average),andstandarddeviationoftheresults?

Step4:Onanewsheetcreateacolumnthatcontainsanumberofequallyspacedbinswhichatleastboundthemaximumand

minimumreadings.?

Step5:Selectthecellsadjacentto

thebins.?

Step6:Gottothemainmenuandinsertfunctionandselect“FREQUENY”fromthestatisticssection.

Awindowwillpopup.?

Step7:Entertheresultdatacellrangepointandthebincellrangepointsrespectivelybut“DONOTHITRETURNorENTER”!FREQUENCY(B2:B6000,F8:F36)?

Step7:PressCTL-SHIFT-ENTER.Thisistherangeentryterminator.Thefrequencyofeachbinwillappearnexttoeachbincell.?

Step8:Createacolumnnexttothefrequencycolumnthatiseachfrequencycolumnentrydividedbythesumof

allbins.Thisistheprobabilitythataresultwillbeinthatbin.?

Step9:Createacolumnnexttothebinprobabilitythatusedthe‘NORMDIST’function.NORMDIST(F8,MEAN,SIGMA,FALSE)?

Step10:Createacolumnnexttothenormdistcolumnthatisnormalized.K8/SUM(K:K)?

Step11:Selectthenormalizeddistributionandbinprobabilitycolumnandchoosechartfromtheinsertmenu.Selectthe“customtypes”tapandthe“l(fā)ine-column”type.Usethebinnameasxlabels.HSPICEforSI12/4/200244CheckScatterPlotFirstMeasure

me

ntScatter10005000Threshold=0.35V0100200300400500600sweep

number?Theabovescatterplotsuggeststhatthemeasurementsarereasonablewelldistributed.HSPICEforSI12/4/2002ResultsofMonteCarloAnalysis45?BreakForspreadsheetwalkthrough?ResultsbelowPDF

of

Measured

vs.

Normal0.180.160.140.120.10.080.060.040.0208.70

6.56

42

.28

.14

.00

.86

.721.58

9.44

7.30

5.16

3.02

0.88

8.74

6.60

4.465.90

3.76

1.624.2.32

0.18

8.04Measured

DataTop724.9

bottom784.5

bins743.9

SIGMA739.6

MEAN720.6835.9678.767

68

69

702

710

718

725

733

74

74

75

76

77

78

78

79

80

81

82

82

83

84

85ps2021.71741.31546MeasuredEstimated

GaussiancopyofMeasured

Normalized

Gaussian

Curve773BinNumber

Bin

ValuesFrequencybin

values

PDFGaussianforbinvalue0.000286870.00076344708.6123678.7014678.70

0.000205686.56

0.000821694.42

0.0045160.0020.0060.014742.3686.56694.42755.7220.001782097HSPICEforSI12/4/200246WhathappensifthescatterplothasoutliersMe

asure

me

ntScatter400020000Theseare0100200300400500600problematic.sweep

numberWhy??

AGaussiananalysisisnotvalid?

Outlierssuggestthatthereexistsphysicalanomaliesthatmustbedetermined.?

Thenextstepistolookathewaveforms.?

ThefollowingpagewillillustratetheissueswithusingaGaussianfitfortheabovedata.HSPICEforSI12/4/200247DistributionresultsforperviousbadscatterLookslikea“mode”

orlikelyoccurrenceherePDF

of

Measured

vs.

Normal0.60.50.40.30.20.109.809.9102

0.13

0.24

35

46

57

.68

.79

90

01

12

23

34

45

56

67

78

89

00

11

2210

11

12

14

1540.1670.18

19

2060.21

23

24

25

27

28

29

31

32

33

34

3620.5810.00

3091.

21.

51.

81.

11.

41.

71.

01.

31.

62.

92.

22.75

88Measured

DataTop3362759.8ps910.6

bottom1021

bins20Measuredcopy

ofEstimatedGaussian923.1

SIGMA947.5

MEAN218.64941.1475904878.4

4.5π

high

1925.005679Measured

Normalized

GaussianCurveGaussian

forbinvalue0.193

0.0012935831209

topπ11.07256826

BinNumber

BinValuesFrequencybinvalues

PDF831.9

3π

high848.2

3π

low940.7

bottomπ1597.052983285.24219820.82945311612345759.80889.911020.021150.131280.24142949338759.80

0.001004889.91

0.4307231020.02

0.494981150.13

0.0381531280.24

0.0020080.2640.2550.1720.0820.0017752690.0017097420.0011555630.000548092882.1

4.5π

low

-42.71049789799.22HSPICEforSI12/4/200248Identifyasweepwiththeanomalyindexrx_rterm@

rx_cterm@tx_rterm@ttx_cterm@

pkg_coulpi

tr@tr1flight_time_temper0.1647

6.18E-10

9.11E-100.2426

5.37E-10

1.02E-09alter#12345678952.2931

7.18E-13

44.4769

5.10E-1343.9272

1.30E-12

49.1339

4.30E-1353.0351

1.28E-12

45.6194

5.10E-1351.4748

1.13E-12