528
« on: September 16, 2011, 06:03:03 PM »
I've been working the last couple of months with Brian and Greg, tracking down and fixing a number of problems I've been having. We're down to the last couple of issues, which, apparently, I'm the only one who's having the problem. I'm just curious if anyone else has seen anything similar. For me, these are hard, repeatable failures. My machine is a mill, using four axes (X, Y, Z/Quill, Z/Knee). These problems are unique to systems using the SmoothStepper, and appear to be caused by communications failures between the SmoothStepper and Mach3. Here are the problems:
1) This one is mostly a nuisance - I sometimes lose the ability to jog one or more axes. This occurs, seemingly, at random. At some point, one or more axes will stop moving when I jog. However, the DROs DO move. There is NO servo fault on any axis when this occurs, and the state will persist for a long time. I can pretty much always cure the problem by jogging the OTHER axes, which seems to get the thing out of whatever bad state it's in. It is, in theory, impossible for this to happen, as the DROs are always showing the position the SS reports, NOT some Mach3 internally generated position. So, this suggests somehow the SS is doing everything it's supposed to be doing, but somehow the step pulses are not making it out to the real world. This one has both Brian and Greg scratching their heads. Of course, when this happens, I have to re-reference the machine, since it is no longer where it should be.
2) This one is a real PITA - One or more of my work offsets get spontaneously whacked. This virtually always happens at the end of a program run - it happened to me twice today. I load and run a program, and it runs perfectly start-to-finish. The very last move of the program is a rapid to X0.000, Y0.000, Z1.500, which it does correctly. BUT, after making that move, one or more DROs show the wrong position. In the case of the last run I just completed, the Z axis DRO ended up showing -4.287! Note, the MACHINE was in the exact, correct position. There is absolutely nothing in the G-code that could possibly explain this. When I started the program, I also set the Machine Coordinates zeros to the same position as the work coordinates zeros, so I could see which one was getting changed. The Machine Coordinates DROs were all showing the correct values. So, the Z axis work offset got spontaneously changed when the program ended! This, too, is quite repeatable. It doesn't seem to happen ALL the time, but it's far from rare.
Anyone else seeing anything similar to these problems?
Regards,
Ray L.