Failure Mechanisms Of Insulated Gate Bipolar Transistors .
Failure Mechanisms of Insulated GateBipolar Transistors (IGBTs)Nathan Valentine, Dr. Diganta Das, and Prof. Michael Pechtwww.calce.umd.eduCenter for Advanced Life Cycle Engineering (CALCE)2015 NREL Photovoltaic Reliability Workshopdiganta@umd.educalce Prognostics and Health Management ConsortiumTM1University of MarylandCopyright 2015 CALCE
CALCE Introduction The Center for Advanced Life Cycle Engineering (CALCE)formally started in 1984, as a NSF Center of Excellence in systemsreliability. One of the world’s most advanced and comprehensive testing andfailure analysis laboratories Funded at 6M by over 150 of the world’s leading companies Supported by over 100 faculty, visiting scientists and researchassistants Received NSF Innovation Award and NDIA Systems EngineeringExcellence Awardin 2009 and IEEEStandards EducationAward in 2013.calce Prognostics and Health Management ConsortiumTM2University of MarylandCopyright 2015 CALCE
IGBT Applications Need for more compact power converters achieved through faster deviceswitchingIGBTs are the ideal choice with switching frequencies of 1kHz-150kHz andcurrent handling of up to 1500A Induction Heating UnitsElectric TrainsPower ConvertersUninterruptiblePower Suppliescalce Prognostics and Health Management ConsortiumTM3Electric CarsWind TurbinesUniversity of MarylandCopyright 2015 CALCE
IGBT TechnologiesSource:Infineoncalce Prognostics and Health Management ConsortiumTM4University of MarylandCopyright 2015 CALCE
Failed Wind Turbine IGBT ModuleFailed IGBT which experienced athermal runaway, burning the moduleUnused IGBTcalce Prognostics and Health Management ConsortiumTM5University of MarylandCopyright 2015 CALCE
Steps in Reliability Evaluation Quantify the life cycle conditions Failure Modes, Mechanisms, and Effects Analysis(FMMEA) reliability analysis, assess designtradeoffs and revise/update design Part, material and supplier selection Virtual qualification (VQ), including stress andthermal analysiscalce Prognostics and Health Management ConsortiumTM6University of MarylandCopyright 2015 CALCE
FMMEA MethodologyDefine system and identifyelements and functions to be analyzedIdentify potential failure modesIdentify potential failure causesIdentify life cycle profileIdentify potential failure mechanismsIdentify failure modelsPrioritize failure mechanismsDocument the processcalce Prognostics and Health Management ConsortiumTM7University of MarylandCopyright 2015 CALCE
IGBT Failure Modes and Mechanisms Failure modes in an IGBT are simple at top level:– Short circuit– Open circuit– Parameter drift Parameter drift occurs as a part degrades and theelectrical characteristics such as VCE(ON) or ICE driftfrom the acceptable operating range due to theaccumulation of damage within a device or modulecalce Prognostics and Health Management ConsortiumTM8University of MarylandCopyright 2015 CALCE
Failure Modes and MechanismsPotential Failure Modes (Sites)Potential Failure CausesPotential FailureMechanisms (Parametersaffected)Short circuit, loss of gate control,increased leakage current (Oxide)High temperature, highelectric field, overvoltageTime dependent dielectricbreakdown (Vth, gm)Loss of gate control, deviceburn-out (Silicon die)High electric field,overvoltage, ionizingradiationHigh leakage currents(Oxide, Oxide/SubstrateInterface)Overvoltage, highcurrent densitiesOpen Circuit (Bond Wire)High temperature, highcurrent densitiesOpen Circuit (Die Attach)calce Prognostics and Health Management ConsortiumTMHigh temperature, highcurrent densities9Latch-up (VCE(ON))Hot electrons (Vth, gm)Bond Wire Cracking,Lift Off (VCE(ON))Voiding,Delamination of DieAttach (VCE(ON))University of MarylandCopyright 2015 CALCE
Examples of Failure ModelsFailure MechanismFailure SitesFailure CausesDie attach, Wirebond/TAB,Solder leads, Bond pads,Traces, Cyclic Deformations(Δ T, Δ H, Δ V)Failure ModelsNonlinear PowerLaw (Coffin-Manson)M, ΔV, T, chemicalEyring (Howard)MetallizationsT, JEyring (Black)Conductive FilamentFormationBetween MetallizationsM, ΛVPower Law (Rudra)Stress DrivenDiffusion VoidingTime DependentDielectric BreakdownMetal Tracesσ, TEyring (Okabayashi)Dielectric layersV, TArrhenius (FowlerNordheim)ElectromigrationΔ:Λ:T:H:Cyclic rangegradientTemperatureHumiditycalce Prognostics and Health Management ConsortiumTMV:M:J:σ:10VoltageMoistureCurrent densityStressUniversity of MarylandCopyright 2015 CALCE
CALCE Simulation Assisted ReliabilityAssessment (SARA ) lcePWACircuit Card AssembliesDevice andPackageFailure AnalysisThermal AnalysisVibrational AnalysisShock AnalysisFailure AnalysisConductor IIWhiskerConductor IcalceFASTSpacing (ls)calceTinWhisker FailureRiskCalculatorcalce Prognostics and Health Management ConsortiumTM11Failure AssessmentSoftware ToolkitUniversity of MarylandCopyright 2015 CALCE
Thermally-Induced Stresses in IGBTMaterial CTE (10-6 K-1) Conductivity(W m-1 u17.5394Mo5.1138Si2.6148AlSiC7.5200calce Prognostics and Health Management ConsortiumTM- Bond Wire Fatigue- Solder Joint Fatigue12University of MarylandCopyright 2015 CALCE
IGBT Power Cycling Experiment IGBT samples were powercycled between specifiedtemperatures TMin and TMax.The devices were switchedat 1 or 5 kHz. Cooling wascarried out passively byexposure to ambienttemperature. This ‘power’ (thermal)cycling was repeated untilfailure occurred by latchupor by failure to “turn on”.Switching at 1 or 5 kHzCoolingTMaxTMinHeatingTimePower cycling illustrationcalce Prognostics and Health Management ConsortiumTM13University of MarylandCopyright 2015 CALCE
Parasitic Thyristor in IGBT StructureInternal PNP Bipolar Transistorcalce Prognostics and Health Management ConsortiumTM14University of MarylandCopyright 2015 CALCE
Parasitic Thyristor in IGBT StructureParasitic NPN Bipolar Transistorcalce Prognostics and Health Management ConsortiumTM15University of MarylandCopyright 2015 CALCE
Die Attach Acoustic Scan ImagesDelaminated surfaceNew IGBT sample.Failure to turn onafter 3126 powercycles, ΔT 75 C.Die attach showsdelamination.Melted die attachFailure by latchup after1010 power cycles, ΔT 100 C. Melting T ofdie attach 233 C*.*Specification sheet for Sn65Ag25Sb10 solder from Indium Corp. Indalloy 209.calce Prognostics and Health Management ConsortiumTM16University of MarylandCopyright 2015 CALCE
Bond Wire FailuresBond Wire Crackingcalce Prognostics and Health Management ConsortiumTMBond Wire Liftoff17University of MarylandCopyright 2015 CALCE
Lifetime Statistics of ExperimentalResults2P-Weibull with95% confidence bounds125-225C Dataβ 2.60η 1191ρ 0.96150-200C Dataβ 2.26η 7134ρ 0.96MTTF 1058MTTF 6320ANOVA p-value 7.6E-6 Different distributionscalce Prognostics and Health Management ConsortiumTM1 kHz5 kHz60% duty cycle18University of MarylandCopyright 2015 CALCE
Prediction of Other Reliability MetricsTemperature RangeMTTF (Cycles)[B5%Life; B95%Life](Cycles)150-200 C6320 cycles[1922; 11,582]125-225 C1058 cycles[381; 1815]MTTF varies with loading conditions and from part to part.Predicting service life of an IGBT based on a population MTTFresults in a high uncertainty.calce Prognostics and Health Management ConsortiumTM19University of MarylandCopyright 2015 CALCE
Physics of Failure Based LifetimePredictionTemperaturePoF Lifetime PredictionExperiment MTTF150-200 C15,300 cycles6320 cycles125-225 C10,800 cycles1058 cycles Thermo-mechanical fatigue due to variations of power dissipation hasbeen identified as a failure mechanism of IGBT. Die attach fatigue failure model was used in the CalceFAST software.The model was based on the Suhir’s interface stress equation coupledwith the Coffin Manson equation.– Model inputs were: T, cycling period, materials, and dimensions.– Failure criteria were based on separation of die attach material. This model does not represent latchup failures and the actualdegradation involves intermetallic growth which changes the crackpropagation due to brittle fracture.calce Prognostics and Health Management ConsortiumTM20University of MarylandCopyright 2015 CALCE
Limitations of the Die Attach Method Die attach area reduction may not be linear as assumed sincethermal stress is highest in the perimeter and reduces as cracksmove toward the center of the die. Crack growth in the brittleintermetallic is not the same as the original material. Power dissipation changes with time as efficiency degrades. The latchup Tj is not always 255C due to difference in currentdensity between operating conditions, metallizationdegradation, and chip manufacturing variations. The developed thermal stack model does not represent theactual thermal resistance network due to unknown spreadingresistance, dissipation through the encapsulant and bond wires,and changing conductivity through the growing intermetallic.calce Prognostics and Health Management ConsortiumTM21University of MarylandCopyright 2015 CALCE
MIL-217 Handbook: ReliabilityPrediction of Electronic Equipment MIL217 Handbook provides formulas to estimate failure rate of militaryelectronic equipment. Constant failure rate is assumed. No formula was provided for IGBT, therefore a MOSFET and BipolarJunction Transistor (BJT) was modeled in series to represent an IGBT. Failure rate is calculated by multiplying a base failure rate with severalconditional factors. For example:λ p λbπ T π Aπ Qπ EwhereλP part failure rateλb base failure rateπT temperature factorλA application factorλQ quality factorπE environment factorcalce Prognostics and Health Management ConsortiumTMfailures/106 hoursTemperature factor does notaccount for temperaturecycling input.22University of MarylandCopyright 2015 CALCE
Comparison of MTTFsTemperatureProfileMILHDBK-217Die AttachFatigue ModelExperimental Data2P-Weibull150-200 C115,843 hours15,300 cycles18.7 hours (6320 cycles)125-225 C96,327 hours10,800 cycles12.2 hours (1058 cycles) MIL-HDBK-217 method does not account for temperaturecycling loading and other relevant loading conditions. Die attach fatigue model provides a better estimate than thehandbook. Improvement to the model includes obtainingmaterial fatigue properties, incorporating intermetallic growthinto the crack propagation, and estimation of junctiontemperature. Predicting lifetime using a population MTTF cannot accountfor variability from part to part.calce Prognostics and Health Management ConsortiumTM23University of MarylandCopyright 2015 CALCE
Motivation for Health MonitoringApproach (for IGBT and System) Using MTTF to predict IGBT lifetime is not sufficient to avoidunexpected failures in the field due to the variability inprediction. Handbook approach ignores relevant loading conditions, devicecharacteristics, and failure mechanisms leading to erroneouslifetime predictions. Physics-based lifetime prediction cannot avoid unexpectedfailures in the field due to variations from part to part and fieldloading conditions. An alternative approach to avoid failures is to monitor IGBThealth individually under operation by using a data-drivenmethod to analyze the operating data and detect for faultyconditions before failure occurs.calce Prognostics and Health Management ConsortiumTM24University of MarylandCopyright 2015 CALCE
What We Need to Do? Relevant material properties for the critical failuremechanisms Ability to update the failure models quickly Modeling platforms for the units and components Life cycle condition information from monitoring Use of data for determination of anomaly at the levelof interest Remaining useful life assessment abilitycalce Prognostics and Health Management ConsortiumTM25University of MarylandCopyright 2015 CALCE
IGBT Prognostics Patil et. al. [9] IGBTs were monitored for VCE and ICE during continuouspower cycling. Proposed a method to predict remaining useful life (RUL)of IGBT under power cycling by extrapolating VCE curve to a failurethreshold using particle filter Sutrisno et. al. [10] generated a K-Nearest Neighbor algorithm for faultdetection of IGBTs under continuous power cycling conditions usingmonitored electrical characteristics such as VCE and ICE .[9] N. Patil, “Prognostics of Insulated Gate Bipolar Transistors,” Ph. D. dissertation, Dept. Mech. Eng., University ofMaryland, College Park, MD, 2011.[10] E. Sutrisno, “Fault Detection and Prognostics of Insulated Gate Bipolar Transistor (IGBT) Using K-Nearest NeighborClassification Algoritihm,” M.S. dissertation, Dept. Mech. Eng., University of Maryland, College Park, MD, 2013.calce Prognostics and Health Management ConsortiumTM26University of MarylandCopyright 2015 CALCE
IGBT Prognostics Xiong et al. [11] proposed an online diagnostic and prognostic system topredict the potential failure of an automotive IGBT power module. Aprognostic check-up routine was implemented that would be activated at apreset frequency and current during vehicle turn-on and turn-off. Ginart et al. [12] developed an online ringing characterization technique todiagnose IGBT faults in power drives. Analysis of the dampingcharacteristic allowed the authors to identify failure mechanisms[11] Y. Xiong, Xu. Cheng, Z. Shen, C. Mi, H. Wu, and V. Garg, ―Prognostic and Warning System for Power-ElectronicModules in Electric, Hybrid Electric, and Fuel-Cell Vehicles,ǁ‖ IEEE Transactions on Industrial Electronics, Vol. 55, No.6, pp. 2268-2276, 2008.[12] A. Ginart, D. Brown, P. Kalgren and M. Roemer, ―Online Ringing Characterization as a Diagnostic Technique forIGBTs in Power Drives,ǁ‖ IEEE Transactions on Instrumentation and Measurement, Vol.58, No.7, pp.2290-2299, 2009.calce Prognostics and Health Management ConsortiumTM27University of MarylandCopyright 2015 CALCE
Unclamped Inductive Switching (UIS)Current Imbalance IGBTs operated with inductive loads can experience voltageswell above their breakdown rating if no voltage clamp isimplemented Voiding and delamination caused by either aging or voidingleads to current imbalance within the IGBT cells, causing localheatingcalce Prognostics and Health Management ConsortiumTM28University of MarylandCopyright 2015 CALCE
Heating within IGBT under UIS ConditionsUnstable behavior observed on die at nominal current localizedheating [13][13] M. Riccio, A. Irace, G. Breglio, P. Spirito, E. Napoli, and Y. Mizuno, “Electro-thermal instability in multicellular Trench-IGBTs in avalanche condition: Experiments and simulations,” in Proc. IEEE 23rd Int. Symp.Power Semiconductor Devices and ICs (ISPSD), May 23–26, 2011, pp. 124–127.calce Prognostics and Health Management ConsortiumTM29University of MarylandCopyright 2015 CALCE
Dynamic Avalanche at Turn-off Similar to UIS conditions, dynamic avalanche can causecurrent imbalance between the cells of the IGBT Dynamic Avalanche can be self-induced if the gate resistanceis too low causing high gate currentsBurned emitter contactpad for discrete IGBTcalce Prognostics and Health Management ConsortiumTM30University of MarylandCopyright 2015 CALCE
Title: Failure Mechanisms of Insulated Gate Bipolar Transistors (IGBTs) Author: Diganta Das Subject
NOTE: In a DUAL GATE INSTALLATION the gate opener on the same side of the driveway as the control box is known as the MASTER GATE OPENER and that gate is refered to as the MASTER GATE. Conversly the gate opener on the other gate is refered to as the SLAVE GATE OPENER and the gate is refered to as the SLAVE GATE. For Mighty Mule FM702, GTO/PRO .
BENCH SEAT SEAT SEAT SEAT TPTP TP TP TP TP SEP TP TP TP PP PP PP TP WV WV WV WV WV WV WV SP SP P GP SEP SP SP SP SP SP SP BBQ PS GATE GATE GATE GATE GATE L B B B B B B B B B B B B B B SEAT B SEAT SEP GATE GATE GATE GATE GATE SEAT L BR. . Raised planter box for grape vine and BBQ her
Gate Drain Gate controls this. Gate can not control below that. So current can leak through there. PDSOI Gate 1V Gate controls this. No leakage path. FDSOI Gate 1V Leak Source Drain FinFET Si Gate 1V Gate Source Drain Better Electrostatics Stronger Gate Control - Lower V t for the same leakage - Shorter channel for the same V t
Insulated, and Pre-Insulated Tubing Installation Instructions These instructions provide the user with information for the installation of Swagelok single-jacketed tubing, multijacketed tubing, insulated tubing, and pre-insulated trace tubing bundles. For additional information, see the Swagelok Multijacketed Tubing, Jacketed Tubing and
the installation of Swagelok single-jacketed tubing, multijacketed tubing, insulated tubing, and pre-insulated trace tubing bundles. For additional information, see the Swagelok Multijacketed Tubing, Jacketed Tubing and Insulated Tubing catalog, MS-02-188 and Swagelok Pre-Insulated Tubing Bundles catalog, MS-02-316. Caution
This GATE study material can be downloaded as PDF so that your GATE preparation is made easy and you can ace your exam. These study notes are important for GATE EC, GATE EE, GATE ME, GATE CE and GATE CS. They are also important for IES, BARC, BSNL, DRDO and the rest. Before you get sta
the unit size output resistance, the unit size gate area capacitance and the gate perimeter capacitance of gate i , respectively. Although the gate shown here is a two-input AND gate, the model can be easily generalized for any gate with any number of input pins. Fig. 4. The model of component i , which is a wire segment, by a -type RC circuit.
classroom teaching to working as a reading specialist, curriculum developer, Title 1 teacher, staff developer, and Title 1 District Coordinator. She is the author of numerous books, articles, and videos and conducts presentations and workshops on literacy throughout the country. Program Advisor: Mary Hawley Mary Hawley is an educational consultant who has worked with teachers, educators, and .
