LinuxCNC at RPI4 with using IO Pins only ?

Thx elovalo
I was always thinking if this wouldn't be working... Eventually with tmc2209.
Have bought those part some while ago but never found time to play with it.
I think this could be a cheepo solution for some first diy small "machines".

Would be nice if there would be a solution to preset the tmc parameters via serial.
I guess this could even be done "offline" before starting lcnc.
When I hear how silent my marlin controls the steppers with the tmc's.. This solution might have its charm.
Mike

Years ago I made some PCBs to connect a Pi to a DB25 port, with level conversion:

photos.app.goo.gl/auwbTXH2wa81cAqy7

I actually found the unused PCBs whilst moving some junk earlier today.
Pretty much worthless now, they are for the original Pi Model B, with the shorter GPIO connector.

Mike_Eitel wrote: Thx elovalo
I was always thinking if this wouldn't be working... Eventually with tmc2209.
Have bought those part some while ago but never found time to play with it.
I think this could be a cheepo solution for some first diy small "machines".

Would be nice if there would be a solution to preset the tmc parameters via serial.
I guess this could even be done "offline" before starting lcnc.
When I hear how silent my marlin controls the steppers with the tmc's.. This solution might have its charm.
Mike

Ideally the RPI would interface to the Trinamic chips over SPI so

1. No base thread would be needed (so the step rate is not limited by latency)
2. All chip setup could be done with hal parameters

I agree Peter, but on the other side I see the parameters as a one time shot. I see no need for recurrent cpu time in precious rt. Or maybe a comp that only runs once at startup.
Can't judge the spi part, just regarded a bit around Marlin and skr-pro bord the parameter settting hw side; read and write via one pin and I believe relatively slow serial.
m5c
Mike

Yes, if its a slow serial setup thing then a separate comp makes sense.
If you can setup chip parameters via SPI (like some chips I am familiar with),
hal parameters should be fine as they are not transmitted so take near 0 CPU
if not changed (and allow real time adjustments if needed)

Are you saying that some Trinamic chips can take position commands over SPI and effectively generate steps internally?

The datasheet that I just glanced at didn't seem to be saying that, for the one that I looked at.