0
\$\begingroup\$

I am using a STM32F767ZI-Nucleo development board from ST. i'm using it to communicate by SPI with a UM7 orientation sensor board. There a register to read in the FW version. Normally I only receive the right response about 66% of the time to what seems are good inputs (SCK, MOSI and SS lines). I used a wire to measure with the scope the MOSI line. When I accidentally touched the tip of this wire to the RJ45 connector casing of the dev board, all of a sudden 100% of my readings are now OK!

The RJ45 casing is not grounded otherwise there would be no readings, but the metal of it somehow changes my MOSI signal (no apparent change in scope). Since my team s developing an actual board, it would be nice to know what is going on here, so I can maybe replicate it in the board.

Any ideas? Thanks

\$\endgroup\$
1
\$\begingroup\$

When you touch a trace, it adds capacitance. When this makes a system suddenly ‘work’ it usually means that the suspect trace is failing hold time or has some other integrity issue like reflections.

Try adding an AC termination to the end of the MOSI line to improve its signal integrity.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.