TMC2209 needs to be reinitialized after sensorless homing

[troth530]

Sensorless homing works for a single axis but afterward requires a M122i command before hard limits will work for stall detection when it hits a hard stop. Also, homing X and Y axis for instance in the same cycle ($44=3) will result in sensorless homing not working. What happens is diag pin will trigger on the first axis which hits a hard stop during the seek operation, but diag pin will not trigger on the second axis and the homing operation will not complete. I believe these are part of the same issue. For now my workaround is set $44=1 and $45=2 and after a homing cycle issue M122i. In this case homing completes and hard limits are triggered when it hits a hard stop. This is of course with the correct limit settings as well as the correct thresholds that work for single axis homing.

I am testing on ESP32 S3 on a single pin UART for all 4 drivers.

Thank you!

[terjeio]

I just have commited an update that may work better, more info here.

I still do not have a board to test StallGuard homing with TMC2209 drivers, they are special in that the diag output is not open drain like the other drivers I have so are not possible to wire-or with regular limit switches.

[troth530]

Hi @terjeio, I am unable to test for now as I got stuck on an issue with your latest changes.... seems USB_SERIAL_CDC is now crippled. During linking I get an error, odd that it reports the error a few lines before the function call. Setting USB_SERIAL_CDC to 0 is able to compile, I only get this error when attempting to use USB_SERIAL_CDC = 1 (S3 native USB). I've found UART to USB bridges are unreliable for my application.

Can you please look into this. Thank you!

[terjeio]

It is a regression from the last PR, move this line to line 33 to fix.

[troth530]

It is a regression from the last PR, move this line to line 33 to fix.

Thanks, that fixed USB_SERIAL_CDC issue. I will test and report back later this week.