Reliability laboratory
What we test, how often, and what happens when a lot fails.

Sampling, not inspection
Final test proves the lot works today. Reliability sampling proves it will still work in year five.
Every lot gets 100% electrical test for forward voltage, reverse leakage and breakdown voltage. That catches the parts that are wrong now. It does not catch a batch that passes at 25 °C and fails at 125 °C after 1,000 hours.
The laboratory exists for the second problem. Sampling runs per family, per quarter, and on every lot where a process input changes.
- Sampling per family each quarter, plus change-triggered runs
- Failure triggers a hold on the lot and a root-cause report
- Reports available to customers under NDA
Test matrix
What runs, on what, and to which standard.
| Test | Condition | Duration | Acceptance |
|---|---|---|---|
| High-temperature reverse bias | 125 °C, rated VRRM | 1,000 h | Leakage within datasheet limit |
| Thermal cycling | -55 to +150 °C | 500 cycles | No electrical drift beyond 10% |
| Damp heat | 85 °C / 85% RH, biased | 1,000 h | No leakage increase, no corrosion |
| Surge current | 8.3 ms half sine, rated IFSM | 10 pulses | No failure, VF within 5% |
| Solderability | 245 °C, 5 s | Per batch | 95% wetting coverage |
| Temperature cycle on solder joint | -40 to +125 °C | 1,000 cycles | No crack initiation |
What we will not claim
Four honest limits.
- We are not an AEC-Q101 qualified line. Automotive-grade parts are quoted per project with third-party test reports.
- We do not run long-term field data collection. Our reliability evidence is accelerated testing, not ten years of field returns.
- We do not publish MTBF figures. For discrete diodes they are marketing, not engineering.
- We do not have a certified calibration lab on site. Instruments are calibrated by an external accredited body annually.
Send the part number and the application.
We will confirm the equivalent, the price at your quantity and the lead time — usually within 48 hours.