I find it painful to have to remove the end mill, lower the maslow to the lowest position, extend the belts, run the “find anchors” (as my frame isn’t fully square), then retract the belts, install the end mill, set the Z-home, extend the belts… just to be greeted with:
[MSG:INFO: Center point deviation: TL: 0.000 TR: 0.000 BL: 8.681 BR: 13.532]
[MSG:ERR: Center point deviation over 12.000mm, your coordinate system is not accurate, maybe try running Find Anchors again?]"
and having to repeat the process again.
Is it possible to install the end mill, set the Z-stop, set the Z-home (higher than the stop), and run the anchor finding logic from the Z-home? I see that there are already offsets for the spoilboard thickness and the work material thickness; can this be added as just another offset?
I like that when locating the anchors the z-axis zero position is set which would have to some other time frame if we don’t do it there.
If the machine doesn’t know the correct z-axis height then all of the math gets thrown off and we don’t have any other time during normal setup that we ask people to home it (although we could)
Needing to take into account the spoil board thickness while locating the anchor points sounds like a bug for sure! I would merge a PR to fix that right away
I just tested out the AI-generated code, and it worked quite well, including running it at the home position instead of the Z-stop.
That being said, maybe add a UI-selector if you want to run it from the Z-stop (more accurate) or from the Z-home (if you don’t want to remove the end mill)
It’s not ideal (limit switches would be better), but one approach could be fixed height spacers so bottoming out the cage is something like 30mm higher than normal.
It does require the user to have a 3D printer or something, and still to check the bit doesn’t still poke out, and (hopefully) minor software changes though.