When a board fails certification
An industrial electronics product failed EMC testing at an external laboratory — and the first redesign failed too.
The problem
An industrial electronics product went to an external EMC laboratory for compliance testing and did not pass. The failure was an emission at approximately 100 MHz, above the permitted limit line.
What I did
I investigated the emission and selected an appropriate filtering solution. The board was then redesigned around it.
What happened
The first redesign did not clear the laboratory either. Further iterations were required before the board passed.
That is worth stating plainly rather than compressing into a single successful fix. Converging on a real root cause took more than one attempt, and the intermediate failures were part of the work.
How it was verified
By an external certification laboratory, not by me. The result is independent of my own judgement about whether the problem was solved — which is the only kind of EMC result worth reporting.