Weird ADC failure STM32F091, any ideas

[Oliver Sedlacek]

We've been building a PCB with four STM32F091s on each PCB and we've built about 50 units by now (over about 3 years). I have now seen three PCBs where the ADC readings are 'bad' on two of the four micros. By 'bad' I mean conversion values are 30% ish low.

With the design having been stable for quite a while, the chances of uncovering a new bug seem slim, but chip damage seems even more unlikely. The fault also 'went away' on one of the PCBs, so more evidence that it's not damage.

The ADC reference is VDD 3.3V on all the chips, and they all run the same code.

Anyone got any thoughts?

[Oliver Sedlacek]

I really appreciate the help I've had on this problem, so thanks to you all.

So what was the problem down to? Why were so many micros working before this problem started to appear?

Mechanical stress! The boards are bolted to a heatsink, so there's a thermal pad under most of it. Recently a thicker pad was being fitted and where the PCB overhangs the pad (not ideal with hindsight), pulling down the bolts was bowing the PCB. The chips don't like it!

[LCE]

Thanks for the feedback!

[waclawek.jan]

Humm, I didn't expect that...

Thanks for coming back with the solution. Please select your post as Best so that thread is marked as resolved.

JW

[raptorhal2]

->The chips don't like it!

Just to complete the investigation, was the package type leadless? LQFPs would probably be the most forgiving of board strain.

Cheers, Hal

[Oliver Sedlacek]

QFN, LQFP was a bit on the big side. Don't know how a BGA would have fared, but at the end of the day the mechies now know not to stress PCBs.