After setting with the probe the z axis does not go to zero

[RJSavell]

I found this in 0.4.7 rc12 and then retested in the release version of 0.4.7 and can still reproduce it.

Using the following g-code with a probe to set the axis height:

G90 G21
M70
G21
M5 S0
F100
G53 G38.3 Z-10
G92 Z20.00
G0 Z21
F25
G53 G38.3 Z-10
G92 Z20.00
G0 Z20.5
M0 (MSG, Remove probe)
M72
M30

and then using the MDI to move the Z Axis to 0 "G0 Z0" the Z Axis does not go to 0 and is somewhat higher then 0. After testing some more I found it is actually at the offset height. For example when I tried to set the height using the probe the screen shows:

and after the G0 Z0 command the screen shows:

but the actual height is just over 4mm (seems to be off by the offset value) instead of being at 0. Of course if you start cutting from here the Z Axis is too high and does not cut properly.

This is the same probe gcode I have been using with previous versions but hopefully this is just an issue with my probe code.

Thanks,

Randy

[jcoffland]

I'm looking into this.

[jcoffland]

There are some problems with your probing code. A previous bug may have masked the problem in eariler versions. See my notes below:

G90 G21
M70                      (Save modal state does not make sense here.)
G21                      (Redundant, already set above.)
M5 S0
F100
G53 G38.3 Z-10           (G53 Move in Machine Coordinates only affects G0 and G1, see docs)
G92 Z20.00               (Set global offsets such that Z = 20.00)
G0 Z21
F25
G53 G38.3 Z-10           (Again G53 has no effect here)
G92 Z20.00               (Set global offsets such that Z = 20.00)
G0 Z20.5
M0 (MSG, Remove probe)
M72                      (Restore modal state is pointless in this context.)
M30                      (End program, resets all global offsets to zero.  See docs.)

The M70 and M72 commands are unnecessary here. See M70.

G53 does not work on probe moves. There is a warning about this but the message was wrong. It said something like 0x76856b7c used without G0 or G1. It should say G53 used without G0 or G1. See G53.

The real problem is that M30 sets the global offsets back to zero. There was a bug, which will be fixed in the next release, that caused it not to show that the global offsets were cleared immediately after M30. See M30.

[RJSavell]

Joe, thanks for going through the probe code. It has evolved over time but edited it based on your comments and it does work. I guess this may be a bit OCD on my part but once I zero the Z-axis it would be nice if there was a way to zero the offset and absolute as well.

Are you saying that I really do not need to use M30 - post processors just output this but if it is unnecessary I can remove it

[jcoffland]

It sounds like you want to set the absolute position, not the offset. The absolute position is only set during homing. You could use G28.3 instead of G92.

You can use M30 but you should understand that it means the program is done and lots of things, including global offsets, will be reset.

[RJSavell]

Yes, that is exactly what I was looking to do. I replaced G92 with G28.3 and now the position and the absolute values are set correctly while the offset remains at 0.

Thank you!

[jcoffland]

Great.