Chapter 4 Data Converter Process Tech F~1

DATACONVERTERPROCESSTECHNOLOGYANALOG-DIGITALCONVERSION1.DataConverterHistory2.FundamentalsofSampledDataSystems3.DataConverterArchitectures4.DataConverterProcessTechnology4.1EarlyProcesses4.2ModernProcesses4.3SmartPartitioning5.TestingDataConverters6.InterfacingtoDataConverters7.DataConverterSupportCircuits8.DataConverterApplications9.HardwareDesignTechniquesI.IndexANALOG-DIGITALCONVERSIONDATACONVERTERPROCESSTECHNOLOGY4.1EARLYPROCESSES4.1CHAPTER4DATACONVERTERPROCESSTECHNOLOGYSECTION4.1:EARLYPROCESSESWaltKesterVacuumTubeDataConvertersThevacuumtubewasthefirstenablingtechnologyinthedevelopmentofdataconverters—startinginthe1920sandcontinuingwellintothelate1950s.AsdiscussedinChapter1ofthisbook,thevacuumtubewasinventedbyLeeDeForestin1906(Reference1).AfigurefromthepatentisshowninFigure4.1.Vacuumtubesquicklyfoundtheirwayintoavarietyofelectronicequipment,andtheBellTelephonesystembeganusingvacuumtubeamplifiersintheirtelephoneplantsasearlyas1914.Figure4.1:TheInventionoftheVacuumTube:1906Amplifierdevelopmenthasalwaysbeencriticaltodataconverterdevelopment,startingwiththeseearlyvacuumtubecircuits.Asignificantcontributionwastheinventionofthefeedbackamplifier(opamp)byHaroldS.Blackin1927(References2,3,4).VacuumtubecircuitdevelopmentcontinuedthroughoutWorldWarII,andmanysignificantExtractedfrom:LeeDeForest,DeviceforAmplifyingFeebleElectricalCurrents,U.S.Patent841,387,FiledOctober25,1906,IssuedJanuary15,1907ANALOG-DIGITALCONVERSION4.2contributionscamefromBellLabs.Foradetaileddiscussionofthehistoryofopamps,pleaserefertoWaltJung'sbook,OpAmpApplications(Reference5).Inthe1920s,1930s,1940s,and1950s,vacuumtubeswerethedrivingforcebehindpracticallyallelectroniccircuits.In1953,GeorgeA.PhilbrickResearches,Inc.,introducedtheworld'sfirstcommerciallyavailableopamp,knownastheK2-W(Reference6).AphotoandschematicareshowninFigure4.2.Figure4.2:TheK2-WOpAmpIntroducedin1953(CourtesyofDanSheingold)Pulsecodemodulation(PCM)wasthefirstmajordrivingforceinthedevelopmentofearlydataconverters,andAlecHartleyReevesisgenerallycreditedfortheinventionofPCMin1937.(Reference7).Inhispatent,hedescribesavacuumtubecountingADCandDAC(seeChapter3ofthisbook).DataconverterdevelopmentcontinuedatBellLabsduringthe1940s,notonlyforuseinPCMsystemdevelopment,butalsoinwartimeencryptionsystems.Thedevelopmentofthedigitalcomputerinthelate1940sandearly1950sspurredinterestindataanalysis,digitalprocesscontrol,etc.,andgeneratedmorecommercialinterestindataconverters.In1953BernardM.Gordon,apioneerinthefieldofdataconversion,foundedacompanycalledEpscoEngineering(nowAnalogic,Inc.)inhisbasementinConcord,MA.GordonhadpreviouslyworkedontheUNIVACcomputer,andsawtheneedforcommercialdataconverters.In1954Epscointroducedan11-bit,50-kSPSvacuum-tubebasedSARADCcalledtheDATRAC.Thisconverter,showninFigure4.3,isgenerallycreditedasbeingthefirstcommercialofferingofsuchadevice.TheDATRACwasofferedina19×26×15housing,dissipatedseveralhundredwatts,andsoldforapproximately$8000.00.WhilethevacuumtubeDATRACwascertainlyimpressiveforitstime,solid-statedevicesbegantoemergeduringthe1950swhichwouldeventuallyrevolutionizetheentirefieldofdataconversionandelectronicsingeneral.DATACONVERTERPROCESSTECHNOLOGY4.1EARLYPROCESSES4.3Figure4.3:1954DATRAC11-bit,50-kSPSSARADCDesignedbyBernardM.GordonatEPSCOSolidState,Modular,andHybridDataConvertersAlthoughthetransistorwasinventedin1947byJohnBardeen,WalterBrattain,andWilliamShockleyofBellLabs(References8,9,10,11),ittooknearlyadecadeforthetechnologytofinditswayintocommercialapplications.Theoverallreliabilityofthedeviceswaspartlyresponsibleforthis,asthefirsttransistorsweregermanium,andwerelimitedintermsofleakagecurrents,generalstability,maximumjunctiontemperature,andfrequencyresponse.InMayof1954,GordonTealofTexasInstrumentsdevelopedagrown-junctionsilicontransistor.Thesetransistorscouldoperateupto150°C,farhigherthangermanium.Additionalprocessingrefinementsweretoimproveupontheearlysilicontransistors,andeventuallyleadapathtotheinventionofthefirstintegratedcircuitin1958byJackKilbyofTexasInstruments(Reference12).Kilby'sworkwasparalleledbyRobertNoyceatFairchild,whoalsodevelopedanICconceptin1959(Reference13).Noyceusedinter-connectingmetaltracelayersbetweentransistorsandresistors,whileKilbyusedbondwires.Asmightbeexpectedfromsuchdifferencesbetweentwokeyinventions,socloselytimedintheirorigination,therewasnoinstantconcensusonthetrueICinventor.Subsequentpatentfightsbetweenthetwoinventor'scompaniespersistedintothe1960s.Today,bothmenarerecognizedasICinventors.InparallelwithNoyce'searlyICdevelopments,JeanHoerni(alsoofFairchildSemiconductor)hadbeenworkingonmeanstoprotectandstabilizesilicondiodeand19×15×26150lbs500W$8,500.00Courtesy,AnalogicCorporation8CentennialDrivePeabody,MA01960