Re: Rotary Axis home and/or limit switches

Andy's proposal has merit.
There are several reasons that we may want to take an A-axis home reference during a rotary operation:
* The rotary's mechanical design may have tight spots that can cause loss of steps, perhaps when well-meaning adjustments to minimize backlash turn, snarl and bite.
* Cumulative bad math calculations in sloppy CAM software may lose tiny chunks of position.
* Errors in the CNC drive software (e.g. certain versions of Mach3 and the M30 function) may cause gross errors in the rotary position that need correction so the process does not have to be repeated.
* Rotary drives that alternate between Y and A using the same controller channel may lose both the Y and A stepper phase relationships when switching (probably WILL lose ...), resulting in an unexpected rotary position or Y position.

Those are merely 4 of many possible sources of bad rotary position after a process. Then you will want a trustworthy method to correct the zero datum before the next pass, not to inherit and build on the past position loss.

Depending on the step count per revolution, it may be possible optically to read a well-divided (perhaps vernier) rotation scale and restore the rotary datum, before the next rotary cut process. If not, then a high performance rotary homing switch can be your friend. You don't need to know how wrong you were, you just need to make it right.

Such homing switches should not be used only at 8am sunrise startup, but should be part of a strict between-processes rotary homing regimen. A classic example is 3-channel controller using the Y channel for A by switching it between motors. Not only can you lose rotary datum by sending the stepper phase relationships down the toilet at switch-over time, but also to send the Y-center phase relationships down the same toilet. This causes double the lost Y relationship to show up in a 2-sided flip job. Add the loss of rotary zero knowledge and you can have huge difficulty diagnosing quite what went wrong, when.

It doesn't matter if you go to rotary home thru software commands and then correct the A-axis coordinate register. It may be equally acceptable to brute-force the rotary to a genuine zero, matching the airy-fairy electronic zero the software naively maintained. It only matters that you CAN determine if you're off zero when the software thinks it's on zero ... and do something about it.

Mark B
Fourth Axis
http://fourth-axis.com



Send instant messages to your online friends http://au.messenger.yahoo.com

[Non-text portions of this message have been removed]