A connector passes inspection, then fails in the field

Fictional engineering case. No real company or reader is being quoted.

Previous problem: The approved MOSFET swap adds 18°C. What was missed?.

Last week’s answer

In this fictional case, the replacement passed the headline voltage and current checks but switched more slowly with the existing gate driver. Simultaneous drain-voltage and drain-current measurements, with probe timing checked, showed greater voltage–current overlap during transitions. Comparing switching energy per cycle with conduction loss at the actual gate voltage and junction temperature identified switching loss as the main contributor to the extra heat.

Gate-charge curves helped explain the driver mismatch, but total gate charge alone could not quantify the MOSFET’s heating. Body-diode reverse recovery was another parameter to investigate only if that diode conducted and was then commutated in this circuit. Neither reverse recovery nor switching frequency should be treated as a universal explanation for a hotter substitute.

The team revised the approved-alternative checklist to compare the operating-point values, package thermal path, gate drive and switching energy instead of relying on headline ratings.

Technical background: Infineon’s MOSFET datasheet explanation.

This week’s problem

An industrial controller passes final inspection and cable testing, but several units develop intermittent sensor readings after months beside vibrating equipment. Disconnecting and reconnecting the plug temporarily clears the problem. The contacts carry low current, there is no visible corrosion and continuity testing on returned units is inconclusive.

Next week, Troubleshooter will examine fretting, contact finishes, retention, vibration evidence and how to reproduce the fault without destroying it.

Readers can submit their own problem through the Troubleshooter page.