Sensorless Homing - Z Axis (Non-Lead Screw)

Howdy howdy,

I am working on a pen plotter type device that uses a rack and pinion for my Z Axis (that moves the pen on and off the surface). I've been trying to use the sensorless homing to touch off the surface with the pen, but having issues. I have set up the X and Y axis properly, adjusted their sensitivity and validated that they work on my BTT skr mini E3, but when I try to use the Z Axis, it simply moves ~5-6mm and then stops. The travel seems to stay the same regardless of there it starts, so it doesn't seem to be a spike in current. I've tried the bare motor in both the Y Axis pins and the Z axis pins, and it seems to work just fine on the Y Axis.

My question: Is there something in the Marlin firmware that would be causing the behaviour to be different with the Z Axis compared to the Y and X Axis? I've combed through the config and adv config, but am worried I am missing something that has implications for how the homing sequence is run for the Z Axis.

Thanks much for any leads!

I think you're running in unknown territory. Sensorless homing on belt driven systems works well. There's a caution about damaging machines with Z lead screws. Rack and pinion is unusual. I haven't found any mention of how well the two work together.

You may have to add a mechanical Z MAX endstop and then experiment with Z_MAX_POS to get acceptable results.

Certainly unproven whether the rack and pinion will ultimately work. My current issue was that I was using the same motor (with no mechanism/transmission attached) and it was having different behaviour on the Y pins than the Z pins.

I believe I may have gotten something working though for the moment. I adjusted a couple of the Z Axis to make sure it matched the Y Axis completely (max acceleration, max feedrate, homing feedrate) and it seems now that I can replicate the Y Axis behaviour on the z axis. I think especially the homing feedrate was like 10x slower for the Z axis (which makes sense in other configurations) which means the baseline current is probably lower making a spike much harder to detect. I read that it works better at higher speeds, so this makes sense as to why the homing speed was relevant.

Thanks for the thoughts, Bob.