Manual tool change + tool lengh touch off

orangecat wrote: Actually reading this, something isn't right. Perhaps the sequence for M600 is wrong. I always issue M600 in the MDI before loading the file and running it. At the first M6, it should not pause and prompt for a tool change. If it does, that's because M600 did not work as the first M6 after M600 should only do a touch-off without a tool change pause/prompt.

It actually does always pause and prompt for a tool change at the first M6 of the job file. Even if I would have inserted the correct Tx via MDI before (I also issued the M600 before loading the tool).
With every next M6 it also pauses and prompts for the tool change.

Thanks Orangecat for your reply and help so far!!

I'd check the installation, to ensure there isn't some problem with the way the M600 command has been configured (maybe the filename is incorrect, or the file is missing...) It's almost certainly the case that the g-code that should be associated with M600 is not being run.

This is the sequence that I think you will need to do to keep things in synch.

Put in the tool and manually run the M600 sequence to measure the tool.
(At this point the tool number in the spindle is correct and the tool length has been measured)
Touch-off to the workpiece and offset the G54 coordinate system to set Z = 0.
To test, run M600 again, check that the Z = 0 still matches the work.

Now things should work OK (assuming the M600 sub is correct).

I think that your problem lies with touching-off the top of the work with unknown tool-length. Or possibly touching-off the top of the work into the tool table and not the work-axis offset.

andypugh wrote: This is the sequence that I think you will need to do to keep things in synch.

Put in the tool and manually run the M600 sequence to measure the tool.
(At this point the tool number in the spindle is correct and the tool length has been measured)
Touch-off to the workpiece and offset the G54 coordinate system to set Z = 0.
To test, run M600 again, check that the Z = 0 still matches the work.

Now things should work OK (assuming the M600 sub is correct).

I think that your problem lies with touching-off the top of the work with unknown tool-length. Or possibly touching-off the top of the work into the tool table and not the work-axis offset.

Hello andypugh,

thanks for your help but if I do it that way, the G54 is put into the TLO offset and (of course) it says that the Joint is hit, so the machine stops.

But thanks for the input - it was worth a try!

orangecat wrote: I'd check the installation, to ensure there isn't some problem with the way the M600 command has been configured (maybe the filename is incorrect, or the file is missing...) It's almost certainly the case that the g-code that should be associated with M600 is not being run.

Hey orangecat,

thanks again for your help! I've double checked now and the file is there, the link to the file name is correct and the only error I could find was that I put the "TOOL_CHANGE_AT_G30 = 0" under "EMCMOT" instead of "EMCIO". But that also did not solve the problem.

It seems that the offset is reversed: I tested by putting in the second tool a little bit deeper into the spindle and it mills in the air, just like it would have reversed the offset - like -2mm would be needed to get to Z0 and it uses 2mm.

Really strange because everything else within my tests with MDI commands works like a charm...

hancock wrote: thanks for your help but if I do it that way, the G54 is put into the TLO offset

I don't even know what you mean by that. Can you expand on the statement?

Have you looked at the demo config that automatically probes for tool length?
github.com/LinuxCNC/linuxcnc/tree/master...h-tool-length-switch