“This board arced between two unconnected output terminals under 15V DC, and in severe cases, it caught fire.”

This customer comment marked the start of the incident. A batch of PCBAs suffered recurring arcing in field operation, with individual boards catching fire — this is a top-tier quality incident for any electronic product. Today we fully recap this failure analysis, explaining how we investigated, pinpointed the root cause, and implemented preventive measures.

1. Observation: Not an isolated failure, but a batch issue

The customer did not report a random single failure. Instead, a batch of boards exhibited arcing between two unconnected output terminals at 15V DC; some boards even ignited. There is no electrical connection designed between these two terminals, so no current path should exist under normal operation. Arcing means the insulation between terminals broke down, creating an unintended discharge path.

2. Step 1: Review design margin

Upon receiving the report, we immediately reviewed design documents to answer one critical question: what voltage rating was this clearance designed for? Our review showed the gap between these two output terminals was originally sized for a higher voltage class. Theoretically, 15V DC should not cause breakdown. In short, the root cause was not the rated capability of the design. So why did breakdown occur in the field?

3. Step 2: Check changes and identify variances

When product failures occur, the fastest investigative path is always identifying where changes were introduced. After line-by-line verification, we found the only change for this batch was in a Gerber section unrelated to this function. This area was optically isolated by optocouplers and had no electrical linkage to the two arcing terminals. Conclusion from change review: the change itself was not the direct trigger. Still, this investigation taught us a key lesson. Any change — even seemingly irrelevant ones — deserves a spot at the top of the failure analysis checklist, because field anomalies often correlate closely with change windows.

4. Step 3: Review batch records and operating environment

Having ruled out design flaws and engineering changes, we moved to two further areas:

• Batch consistency: We compared process parameters, component lots and storage conditions of this batch against other production runs, confirming no abnormal deviations on the manufacturing side.
• Operating environment: Arcing is strongly tied to creepage and clearance. Temperature, humidity and contamination at the customer site can drastically alter the breakdown threshold of insulating media. The same physical gap delivers vastly different safety margins between dry, clean conditions and humid, polluted environments.

This is the true value of failure analysis: it does not predefine conclusions. Instead, it examines all four dimensions — design, changes, manufacturing process and environment — eliminating hypotheses one by one while gathering evidence. Our FA conclusions are backed by test records and evidence, with no vague assumptions.

5. Final findings & recurrence prevention mechanisms

Three process improvements came out of this case:

1. Enhanced change management: All Gerber / design revisions, regardless of whether they affect electrical functions, go through formal ECN review. Creepage and clearance are rechecked before and after every change.
2. Voltage gap verification: For terminals operating at seemingly low voltages for customer applications, creepage distances are revalidated under worst-case environmental conditions, not just nominal voltage.
3. FA knowledge archive: Every failure analysis case is documented, including symptoms, investigation workflow, evidence records and corrective actions, serving as design input for future projects.

We shared the full analysis workflow and conclusions with our customer. No evasion, no concealment; every investigation step and supporting evidence was laid out clearly. What customers truly care about is rarely “zero failures forever”. It is whether the manufacturer can deliver a responsible response within days once something goes wrong.

PCBA failure analysis does not rely on luck. It hinges on a robust system: traceable changes, verified clearances, and environmental considerations. Many “unexpected” failures can be avoided if these three items are properly managed.

If your products are facing similar failure issues, feel free to send us your boards and failure descriptions. We will conduct a complete FA following the design-change-process-environment framework and deliver a traceable investigation report.