STM32H743 UART Communication Errors Appearing After Several Days of Continuous Operation
edit test
The protocol currently uses CRC validation and timeouts, and corrupted packets are discarded, but I suspect my recovery strategy may not be as thoroughly tested as it should be.
At the moment, a timeout triggers a retransmission request, and after a configurable number of failures the communication channel is reinitialized. What I haven't done yet is collect detailed statistics over long-duration runs.
Right now I only know that communication errors occasionally occur, but I don't have enough data to determine whether they're clustered around specific events such as heavy system activity, relay switching, power fluctuations, or particular times after startup.
I think adding counters for CRC failures, timeouts, retransmissions, and interface resets would probably be a good next step. Have you found it useful to log timestamps and system-state information as well, or do simple error counters usually provide enough insight initially?
