Next huge milestone reached!


I had a pretty similar problem with two boards I was debugging some time ago, that were connected through an SPI bus. The problem happened only on booting: once booted, if the board started, it didn't fail anymore while powered. Sometimes some of the boards that worked at first started failing. Sometimes some of the boards that didn't work at first, started working. Touching the bus pins with the finger on boards that didn't work, made the problems go away. Also connecting the logic analyzer to the bus improved the situation...

Interesting enough, what caused the problem is not listed on the suspects @EvilDragon wrote on the first post. Do you want to know what was causing it? It was a noob problem: just some pull-up resistors that were missing. Once I activated the pull-up resistors inside the MCU, the problem was gone for good.

It's a bit weird that such a small capacitance variation caused the problem. 20% variation may look pretty high, but commonly used capacitors usually have a tolerance that high or even larger (you can easily find capacitors with +50/-20 % tolerances mounted on a lot consumer electronics). Even though the problem looks solved, I would double-check for missing pull-up/pull-downs just in case...
 
Back
Top