Use average value from multiple tmc drivers for bed leveling and delta calibrate.
Sensorless homing does not offer the required precision for z-leveling.
1 Like
But I saw sensorless z homing on the picasso printer, and even the calibration of the height of the second nozzle for the tool changer. Can they really do this but the accuracy is not enough for you? Maybe I should make a video of it parking and calibrating for you?
Video would be interesting but what I think you need to provide is the results from probe_accuracy:
That would be along with, of course, the klippy.log for the printer.
I’m curious to see the results. I’m a big fan of sensorless homing but in my informal experiments, I see the home position vary between +/- 0.1mm to +/-0.5mm which is an order of magnitude higher than what’s required for Z axis homing.
Unfortunately, I have to disappoint you right away. This printer works on firmware with closed source code compatible with gcode from reprap firmware. If you tell me how to get this data from such firmware, I will gladly do everything.
So, how do you know that they are using the stall-guard feature from the TMC drivers? Could be some kind of loadcell, piezo sensors etc.
Please forgive me. I was mistaken and thought that this was true, but when studying the service manual I realized that they use their own version of an adjustable pressure sensor on the print platform. I thought that they use Sensorless homing on the z-axis because they use it on the other two axes.
1 Like
are you sure this is even a pressure sensor?
To me this looks like two PCBs that are pressed together by two spring to close a contact. When the nozzle is pushing down the bed, the contact opens and the endstop/probe triggers.
clever and simple design, similar to klicky probe but for the whole bed.