ValidationandTestingofDesignHardeningforSingleEventEffectsUsingthe8051MicrocontrollerAbstractWiththedearthofdedicatedradiationhardenedfoundries,newandnoveltechniquesarebeingdevelopedforhardeningdesignsusingnon-dedicatedfoundryservices.Inthispaper,wewilldiscusstheimplicationsofvalidatingthesemethodsforthesingleeventeffects（SEE）inthespaceenvironment.Topicsincludethetypesofteststhatarerequiredandthedesigncoverage（i.e.,designlibraries:dotheyneedvalidatingforeachapplication?）.Finally,an8051microcontrollercorefromNASAInstituteofAdvancedMicroelectronics（IAμE）CMOSUltraLowPowerRadiationTolerant（Culprit）designisevaluatedforSEEmitigativetechniquesagainsttwocommercial8051devices.IndexTermsSingleEventEffects,Hardened-By-Design,microcontroller,radiationeffects.I.INTRODUCTIONNASAconstantlystrivestoprovidethebestcaptureofsciencewhileoperatinginaspaceradiationenvironmentusingaminimumofresources[1,2].Witharelativelylimitedselectionofradiation-hardenedmicroelectronicdevicesthatareoftentwoormoregenerationsofperformancebehindcommercialstate-of-the-arttechnologies,NASA’sperformanceofthistaskisquitechallenging.Onemethodofalleviatingthisisbytheuseofcommercialfoundryalternativeswithnoorminimallyinvasivedesigntechniquesforhardening.Thisisoftencalledhardened-by-design（HBD）.Buildingcustom-typeHBDdevicesusingdesignlibrariesandautomateddesigntoolsmayprovideNASAthesolutionitneedstomeetstringentscienceperformancespecificationsinatimely,cost-effective,andreliablemanner.However,onequestionstillexists:traditionalradiation-hardeneddeviceshavelotand/orwaferradiationqualificationtestsperformed;whattypesoftestsarerequiredforHBDvalidation?II.TESTINGHBDDEVICESCONSIDERATIONSTestmethodologiesintheUnitedStatesexisttoqualifyindividualdevicesthroughstandardsandorganizationssuchasASTM,JEDEC,andMIL-STD-883.Typically,TID（Co-60）andSEE（heavyionand/orproton）arerequiredfordevicevalidation.SowhatisuniquetoHBDdevices?Asopposedtoa―regular‖commercial-off-the-shelf（COTS）deviceorapplicationspecificintegratedcircuit（ASIC）wherenohardeninghasbeenperformed,oneneedstodeterminehowvalidatedisthedesignlibraryasopposedtodeterminingthedevicehardness.Thatis,byusingtestchips,canwe―qualify‖afuturedeviceusingthesamelibrary?ConsiderifVendorAhasdesignedanewHBDlibraryportabletofoundriesBandC.Atestchipisdesigned,tested,anddeemedacceptable.NinemonthslateraNASAflightprojectentersthemixbydesigninganewdeviceusingVendorA’slibrary.Doesthisdevicerequirecompleteradiationqualificationtesting?Toanswerthis,otherquestionsmustbeasked.Howcompletewasthetestchip?Wastheresufficientstatisticalcoverageofalllibraryelementstovalidateeachcell?IfthenewNASAdesignusesapartiallyorinsufficientlycharacterizedportionofthedesignlibrary,fulltestingmightberequired.Ofcourse,ifpartoftheHBDwasrelyingoninherentradiationhardnessofaprocess,someofthetests（likeSELintheearlierexample）maybewaived.Otherconsiderationsincludespeedofoperationandoperatingvoltage.Forexample,ifthetestchipwastestedstaticallyforSEEatapowersupplyvoltageof3.3V,isthedataapplicabletoa100MHzoperatingfrequencyat2.5V?Dynamicconsiderations（i.e.,nonstaticoperation）includethepropagatedeffectsofSingleEventTransients（SETs）.Thesecanbeagreaterconcernathigherfrequencies.Thepointoftheconsiderationsisthatthedesignlibrarymustbeknown,thecoverageusedduringtestingisknown,thetestapplicationmustbethoroug...
