Well, I am not sure what the cause is, but I did find a work around.  I modified the post processor and add the command:

G00 X0.0000 Y0.0000 Z2.0000

immediately after the G43, "Set tool length" command (Inserted between lines 440 & 450), the soft warning disappears.  I can only assume that because the Z is sitting outside the limits when the tool change is requested, and even though the SET TOOL LENGTH OFFSET command puts it back within limits, unless the next command requests a Z move, I will receive a soft limit warning.

I have seen this on machines with no z-axis and where I am several inches inside the boundaries on all sides.

I get this warning also when I run I,J,K codes generated from visualmill. The program extremas go crazy high even tho the code is good.

FWIW, I have noticed this behavior when I have offsets in the code. I does not seem to happen if I use the offset table, only if I offset 'on-the-fly' within the code. That's my recollection anyway.

It seems like perhaps Mach collects the physical extents as it generates the path, but then applies those to the current position at the beginning of the program without regard for subsequent offsets within the body of the code . .  just a guess . . .

Just curious if anyone found any answers concerning this problem.  I am experiencing sporadic Soft Limit warnings when I'm well within the limits.  I'll get the warning on one run and not on the next run.  I'm baffled

Bill, I did report this problem in the new bug reporting link on this forum, so hopefully Brian will eventually work through all the bugs there and put this one to rest.

It would probably help if you reported this bug there also, here's the link:

http://www.machsupport.com/bugs.php

Paul T.

Thanks Paul!

I get this warning also when I run I,J,K codes generated from visualmill. The program extremas go crazy high even tho the code is good.

same here.. i have my softlimits turned off since virtually all my code contains IJK code.