Remap M6 with "toolchange with length-probe"

Hello,

I'm quite new to this, but I'm trying to setup my machine with the "toolchange with length-probe" from sim/remap. Run into a few problems. Linuxcnc is on 2.6.0-pre0-3255-gce55978. Fixed a physical probe and it works. Made a simple program:

Workflow is:
1. Home machine
2. Touchoff all axes to workpiece in G54
3. Start program.

Problem with the original manual_change.ngc is that the M73 when exiting the sub reverts the tool length offset (G43) to the state before the sub was executed, effectivly canceling out the tool length offset calculated in the sub for anything other than the reference tool.

Fixed this by changing to M70 in the beginning of the sub, and M72 in the end of the sub. But after M72, I added the g43.1 z[#5063-#1000] . This works good with the above program.

Problem now is that after the first run of the program, I want to change tool, touchoff on workpiece and run it a second time. This time there is problem with the offsets, Z0 after the sub is not on workpiece. I suspect it's because after the first run of the program G43.1 is still in use, and I touch off with this, then sub changes g43.1 and messes up my Z0 somehow. If I MDI a G43 before touching off Z inbetween programs, everything is working. But I do not want to do this G43 manually before touch off every time, sometime I will forget it, be sure about it....

Any ideas on how to solve this I would appreciate.

Best regards
Henrik

Thanks for the replies, been a while, and now I have got the time to work on this and found a solution. The problem was that G49 was still in use when the probing started. And that screwed up things. I added just before the probing:

And now it works If it is a good solution I do not know, but works for me right now. But I think it would be great if there was a named parameter for the dynamic tool length offset.

andypugh wrote:

h_munktell wrote:

Problem with the original manual_change.ngc is that the M73 when exiting the sub reverts the tool length offset (G43) to the state before the sub was executed, effectivly canceling out the tool length offset calculated in the sub for anything other than the reference tool.

I think that the original application of that demo was for a machine with no repeatable tool-length, and the idea is that the machine re-probes on every tool change.
If you want the tool length to be persistent then you need to store the length to the tool table using G10 (which has a number of options which I can't recall, but are documented)

well I rather think the demo was written by somebody who didnt proper check results
-m

mhaberler wrote:

well I rather think the demo was written by somebody who didnt proper check results
-m

Well' I'm glad "somebody" did not, have learnt a great deal of linuxcnc because this Keep up the good work!