We design and manufacture semiconductor chips, discrete power devices, and power modules for your products.
We design and manufacture semiconductor chips, discrete power devices, and power modules for your products.

Less Energy

More Efficiency

  • Home
  • Products 
    • IPM by Package
    • IGBT Modules
    • IGBT Discretes
    • IGBT Chips
    • SiC
    • FRD(MUR)
    • Bridge Rectifier
  • Application 
    • Electric Vehicle
    • Home Appliance
    • Renewable Energy
    • Industrial Control
    • Electric Power Supplys
  • Technology 
    • Latest IPM Technology
    • High Voltage (HV) Die Technolog
    • Reliability & Qualification
  • About Us 
    • Our Company
    • Technical Team
    • Custom Solutions
  • Contact Us
  • Blog
  • …  
    • Home
    • Products 
      • IPM by Package
      • IGBT Modules
      • IGBT Discretes
      • IGBT Chips
      • SiC
      • FRD(MUR)
      • Bridge Rectifier
    • Application 
      • Electric Vehicle
      • Home Appliance
      • Renewable Energy
      • Industrial Control
      • Electric Power Supplys
    • Technology 
      • Latest IPM Technology
      • High Voltage (HV) Die Technolog
      • Reliability & Qualification
    • About Us 
      • Our Company
      • Technical Team
      • Custom Solutions
    • Contact Us
    • Blog
Free Sample
We design and manufacture semiconductor chips, discrete power devices, and power modules for your products.
We design and manufacture semiconductor chips, discrete power devices, and power modules for your products.

Less Energy

More Efficiency

  • Home
  • Products 
    • IPM by Package
    • IGBT Modules
    • IGBT Discretes
    • IGBT Chips
    • SiC
    • FRD(MUR)
    • Bridge Rectifier
  • Application 
    • Electric Vehicle
    • Home Appliance
    • Renewable Energy
    • Industrial Control
    • Electric Power Supplys
  • Technology 
    • Latest IPM Technology
    • High Voltage (HV) Die Technolog
    • Reliability & Qualification
  • About Us 
    • Our Company
    • Technical Team
    • Custom Solutions
  • Contact Us
  • Blog
  • …  
    • Home
    • Products 
      • IPM by Package
      • IGBT Modules
      • IGBT Discretes
      • IGBT Chips
      • SiC
      • FRD(MUR)
      • Bridge Rectifier
    • Application 
      • Electric Vehicle
      • Home Appliance
      • Renewable Energy
      • Industrial Control
      • Electric Power Supplys
    • Technology 
      • Latest IPM Technology
      • High Voltage (HV) Die Technolog
      • Reliability & Qualification
    • About Us 
      • Our Company
      • Technical Team
      • Custom Solutions
    • Contact Us
    • Blog
Free Sample
We design and manufacture semiconductor chips, discrete power devices, and power modules for your products.

What Is AEC-Q101? Eight Tests, 77 Samples, and Three Misconceptions

Introduction: the Tier 1 reply was one sentence

“Our 650V IGBT discrete has run three years in industrial designs. Parameters match the big brands. Why won’t the carmaker even accept samples?”

“Because you’re missing one document,” Dave said. “Automotive isn’t a performance gate, it’s an evidence gate. They don’t want to hear that your part works. They want a signature that says it still works after a thousand hours of abuse.”

That document is AEC-Q101. It does not define how good a device must be — it defines how you must prove it is reliable enough.

This article covers what it actually tests, what the number 77 really means, and the three places suppliers most often get it wrong.

Scope: what AEC-Q101 covers and what it does not

Issued by the Automotive Electronics Council, AEC-Q101 applies to automotive discrete semiconductors: power MOSFETs, IGBT discretes, diodes, bipolar transistors, and SiC/GaN discretes.

StandardScopeTypical examplesAEC-Q101Discrete semiconductorsTO-247 IGBT discretes, DPAK MOSFETs, SiC diodesAEC-Q100Integrated circuitsMCUs, gate drivers, PMICsAEC-Q104Multi-chip modules (MCM)Multi-die power modulesAEC-Q102 / Q103Optoelectronics / MEMSLEDs, image sensors, accelerometersAQG 324 (ECPE)Automotive power modulesPower cycling and thermal testing guidance for IGBT and SiC modules

So when a customer asks for “AEC-Q101” on an IGBT module, the correct answer is: discretes go through Q101, modules go through Q104 or AQG 324. Handing in a Q101 report for a module is the classic beginner mistake.

What it tests: eight families

The list is short, but every item targets a specific failure mechanism. Conditions below are typical — always defer to the revision your customer accepts.

TestTypical conditionsFailure mechanism targetedHTRBTj_max (150/175°C), 80–100% rated V, 1,000 hTermination breakdown drift, passivation defects, surface ionsHTGBTj_max, ±VGS(max), 1,000 hGate oxide TDDB, Vth shift, gate leakage growthTC-55 to +150°C, 1,000 cycles, controlled transfer timeCTE-mismatch fatigue in solder and wire bonds, delaminationIOLPower-pulse self-heating, agreed ΔTj (often 100 K), thousands of cyclesSolder cracks, aluminum wire lift-off, die fractureH3TRB85°C/85%RH, 80% rated V, 1,000 hMoisture uptake, electrochemical migration, aluminum corrosionHAST130°C/85%RH, biased, 96 hSeal defects, fast moisture ingress (accelerated H3TRB)PreconditioningSimulated reflow thermal shock after moisture soakPopcorning, internal delaminationMechanical & packageShock, vibration, constant acceleration, lead integrity, solderability, solder heatInterconnect fracture, lead damage, solderability

Notice what is absent: efficiency, switching loss, thermal performance. AEC-Q101 does not judge whether your device is good — only whether it degrades under stress. Performance belongs in the datasheet; this report belongs to reliability.

What 77 samples actually mean

Most tests require 77 parts per lot with zero failures allowed. The usual reactions are “that seems few” and “77 passes means zero defects.” Both are wrong.

Seventy-seven corresponds to an LTPD = 10 sampling plan: if the true defect rate were 10%, the chance of drawing 77 good parts in a row is negligible, so such a lot is rejected. Conversely, 77 passes with zero failures supports roughly “90% confidence that the true defect rate is below 3%.”

That conclusion has three preconditions:

  • Lot homogeneity. One solder furnace, one wire bonder, one molding compound batch. If the lot is stratified, sampling is a coin flip.
  • Process control. IATF 16949, SPC and change management, so lot 100 is the same product as lot 1.
  • Screening. 100% burn-in or dynamic screening in production, removing the infant mortality sampling cannot catch.

“77 with zero failures” proves capability, not zero defects. Automotive zero-PPM targets are reached by stacking capability, process control and screening. Remove one layer and the number on the report becomes a promise nobody can keep.

1,000 hours is not just “run the clock out”

  1. Intermediate read points. Measuring only at 0 h and 1,000 h is watching two frames of a film. A proper report reads at 168 h, 500 h and 1,000 h and plots the drift. Flat-then-vertical and flat-all-the-way are very different devices.
  2. Failure criteria. AEC-Q101 uses relative drift plus absolute limits — leakage beyond 2x initial, for example. A report that only says “functionally OK” has concluded nothing.
  3. Closed-loop failure analysis. When failures occur, were cross-sections, acoustic scans and hotspot localization performed, and was a mechanism identified? In automotive audits this often outweighs the test itself.

Three misconceptions

One: passing AEC-Q101 means you can build automotive parts

No. Q101 is product qualification — it proves the design survives lab stress. Manufacturing capability is a system certification built on IATF 16949, PPAP, MSA and SPC. Plenty of companies hold a Q101 report and still fail a customer audit because change control is chaos.

Two: qualification is valid forever

AEC-Q101 is tied to a specific combination of design, wafer process, package materials and critical process parameters. A new fab, a new solder alloy, a new molding compound, a layout change, a different wire material or diameter — any of these triggers re-evaluation. Automotive customers additionally require a PCN process: approve first, then change.

Three: AEC-Q101 tells you how many years the part will last

It does not. It is a gate, not a lifetime model. Predicting life requires extrapolating degradation data against your actual junction temperature, temperature swing and start-stop frequency. That is why the same Q101 report means completely different things in a 105°C cabinet and a 150°C engine bay.

How it relates to AQG 324 and IATF 16949

NameLayerQuestion answeredAEC-Q101Product qualification (discretes)Can this device survive automotive stress?AEC-Q104 / AQG 324Product qualification (modules)How are power cycling, Rth and isolation verified for multi-die modules?IATF 16949 + PPAPSystem and productionCan you build the same thing, repeatably and traceably?

The correct order is: build the system and process first, qualify the product second, submit PPAP third. Do it backwards and the report will not save the audit.

What to actually prepare

  • Samples: at least three lots, 77 parts per lot where required, with lot and wafer traceability retained.
  • Documents: datasheet, DPA plan, test plan, criteria table, customer-specific requirements list.
  • Lab: confirm accreditation and capability for each item — especially ΔTj calibration for IOL and bias application in H3TRB.
  • Failure analysis: agree the FA flow and turnaround up front, so a failure does not stall for three months.
  • Change management: write the PCN process into the quality agreement before the customer asks.

FAQ

Does AEC-Q101 cover IGBT modules and IPMs?

No. AEC-Q101 is for discretes. Multi-chip modules reference AEC-Q104, and automotive power modules typically follow ECPE AQG 324 for power cycling and thermal testing. The three are not interchangeable.

What changed between AEC-Q101 revisions?

Revisions adjust test items, conditions and criteria incrementally, and clarify applicability to wide-bandgap devices. You do not need to memorize the differences, but you must print the revision on the report cover and confirm which revision the customer requires.

77 samples with zero failures — what is the failure rate?

Roughly “90% confidence that the true defect rate is below 3%,” consistent with an LTPD = 10 plan. It is not proof of zero defects; 100% screening and process control are what drive PPM down in production.

How does AEC-Q101 differ from AEC-Q100?

Q100 covers integrated circuits, Q101 covers discrete semiconductors. Test names look similar, but conditions, sample sizes and criteria are specified per device type, and the reports are not mutually accepted.

Subscribe
Previous
Consumer, Industrial, Automotive, Military: What...
 Return to site
feed icon
Profile picture
Cancel
Cookie Use
We use cookies to improve browsing experience, security, and data collection. By accepting, you agree to the use of cookies for advertising and analytics. You can change your cookie settings at any time. Learn More
Accept all
Settings
Decline All
Cookie Settings
These cookies enable core functionality such as security, network management, and accessibility. These cookies can’t be switched off.
These cookies help us better understand how visitors interact with our website and help us discover errors.
These cookies allow the website to remember choices you've made to provide enhanced functionality and personalization.
Save