I've been using Mach3 Turn. I was having problems with the G83 cycle when the cycle completed, just before the last retract; there was an audible "thud" with no Z axis movement while the Z DRO indicated Z movement as if stuck at the bottom of the hole (though this was in air), then the X axis would move with the drill still inside the virtual workpiece. (Well ... the first time this ever happened, the workpiece was aluminum; the 1/4" cobalt drill bit snapped off.) Installing Version 1.90.015 seemed to solve this problem, and resulted in much faster rapid speeds. However, I noticed that there was now a "thud" at the beginning of each threading pass of a G76 cycle, with accumulating loss of Z motion. Slowing down the Z acceleration solved this problem.
The PC I was using then went belly up, so I've temporarily replaced it with a PC that I'm not free to optimize. The "thud"/Z stuck problem with the G83 cycle returned, but finally went away when I slowed down the X and Z velocity and acceleration to a relative crawl. However, I now have a problem with a G76 cycle. There's now a "thud"/Z stuck problem at the end of this cycle. Perhaps a very, very slow velocity and acceleration would solve this problem, but this is unreasonable for the production I must undertake.
It seems an odd coincidence that these problems are occurring at the very last rapid movement at the end of these canned cycles, as if there is some fundamental problem with returning from a canned cycle that happens to be clearly exhibited by a poorly tuned PC. Is this possible?
Guest
