[Cypris]

At random times, all which were moments during operation (not idle), The spindle slams down into the table at great speed (if you've ever tried to operate the z axis with the encoder cable unplugged you know what I'm talking about). Y also strays at random times but does not happen nearly as much. Any help with troubleshooting this would be greatly appreciated.


Thanks!

[gfacer]

Well, to throw out an idea for a similarly intermittent problem I had with a wincnc system (which I think yours would be?). Oops, nevermind, I see it's a gerber, so probably it's own control software but the rest still applies.

I have a CNT motion systems router and the wincnc stuff is in a dust cabinet and the servo drives are on the back of the gantry. The problem with that is that the step and direction signals that command the servo drives (as is typical these days, closed loop is in the drive) run for a good 15 feet plus in close proximity to the power lines for the spindles. The closed loop does NOT extend back to the cabinet, nor does Wincnc hardware look for or read encoder counts. What I had was random gantry movements, small fortunately, that eventually made the motor drives abort due to high persistant torque on one motor or the other on the pair that move the gantry......possibly other axis too but not nearly as noticeable if so.

The cause was an apparently internally broken spindle FAN power wire, that generated some electro-magnetic interference and "added" steps to the stream going into the drives. The fan was 220V but the amperage can not have been much so I could see a larger motor with the same problem having larger effects, and if it was momentary in nature then the momentum of the spindle might mask other signs of such a problem. I am running with a kludge for now but it works and problem is *mostly* gone. Lets just say I am a fan of closed loop to the controller systems or at the very least servo drives well wired in the same cabinet as the control hardware. I only found the problem when one day, hearing one of the motors moving on it's own when it should have been stationary....I tried flicking on the switch that turned the spindle power on and off, on and off and noticed (in the servo tuning software) that that action alone was adding steps to the position.

Good luck, it took me 2 years to find my problem! And, If my router wasn't on the older side now, I would be seriously considering changing to a Kflop/Kanalogue based system that can have encoders feed back to the control software and gives a true closed loop. It's not the easiest hardware to setup from what I can tell, but if I could go back 2 years of a flakey router and do the switch over work then, I sure would have.

Edit: If your system has easy to change wiring, that would be another way of fixing such a problem. Just remember to ground the shielding of the cables on the cabinet side only (and I think only for the data cables is this done, but I'm no electrician)