I wrote some here about my initial experiments trying to get ultra-precision readings with a few cents worth of hardware I had on hand (not quite good enough, but @runger saved the day sending me some proper diametric magnets and sensors to use), but I don’t think I documented the rest:
- Struggled getting grblHAL to compile for RP2350
- Was tormented by a tin whisker that eventually fried the SCL pins on both STM32’s
- Gave up after sporadic triggering of step interrupts
I had a rat’s nest of tiny wires soldered to the Gooser due to not enough GPIO accessible, so it would be more likely to work with the latest version, which could have ground wires together with each signal wire to hopefully minimize EMI pickup, and use STM32’s hardware step/dir counting which has some spike filtering.
I’m pretty happy as it is since I can still use the sensors and RP2350 to display the axis positions on a little OLED, and use the buttons I added to zero the positions and do probing operations more quickly than I could before. But I probably will try closed-loop with SimpleFOC Stepstick since it should be relatively easy just plugging it into the GRBL board and swapping back if it doesn’t work out.