[peter.steele]

Hello, all -

I have a pretty mongrelized setup at this point. It started life as a Vision 2448 combo router / engraver with dual heads and a Series II serial controller. I've pretty well hacked things apart at this point - I've removed the second head (I only need one for most of what I'll do) and I've bypassed the control functions of the controller, using it for nothing more than a power supply - I feed the drive boards (Intelligent Motion IM483's) from Mach3 and an extra parallel port with added +5V from a USB port. I've replaced the engraving spindle with a Bosch Colt.

Now we get to the problems.

1. The X/Y stepper motors appear to be NEMA 34's. The mounting faces are approximately 3 7/16 square, BUT they don't have any data plates or markings. The current limiting resistor on the driver boards has them locked down to 2.6A. This may or may not be even close to the real amperage rating though, because the Z/A boards are set to ~1.9A, and the actual Z steppers - which do have data plates - are rated at 2.3A.

2. I'm losing zero while cutting. Let's say I'm doing a simple profiling operation. I'm cutting out a piece of ~6" x 9" UHMW. (This is based on reality this past weekend.) I've got the cuts set at 100 ipm, .0625"/pass. (When jogging, I can move - usually, but we'll get to that later - the head at 200+ ipm, and there's no bogging down on the spindle when cutting at 100.) The first few passes will be perfect, but the next pass ... the zero seems to have shifted by 3/16 or 1/4 or so, in a seemingly random direction. No idea why. This also happens while jogging from time to time - I can set zero, then jog around some, then tell Mach3 to send my head back to zero ... and I'm off.

3. Stalling. In general, everything moves as expected ... but now and then, the steppers will start to stall, seemingly randomly. They will continue stalling - while still moving the head - until I kill power to the drivers / steppers and then turn it back on again. After this, they'll run for a bit longer without any stalling - anywhere from 1 to 15 minutes or so - but then they'll make that stalling sound again, and nothing is accurate until I kill power and start it again.



Anyone have any ideas?

It's been put to me that the power supply might be inadequate, but this doesn't really make any sense as I'm still using the exact same power supply setup that the machine shipped with from the factory. The only thing I've really changed is that I'm feeding the step/dir and opto power from a different source (LPT1 & USB, respectively). The transformer appears to be an Avel Lindberg D4040, which looks to have 2x 25V/4.6A outputs. This seems a little low ... but 2x 2.6A + 2x1.9A seems to add up to about the same as 2x4.6A.

I'm pretty much lost, here. Help?

[jschmitt]

Are you using screw shafts to move your machine? are the couplings tight? are nuts holding the screw tight, not turning.

Are you using shielded cable for your motors and control? Make sure it isn't picking up RF. must be grounded properly.
Try slowing your speeds and see if that make a difference.

[peter.steele]

Okay, I've had a little time to work on this some more.

1. I'm using ball screws and linear bearings. Couplings are all tight.

2. Cables are the factory installed ones, and are shielded, although the hack-job I've done on the controller has some unshielded wires temporarily because the controller housing is open for the moment. Still some things that need to be finalized.

3. Y-axis is looking like it might be finally working right - I had a test run set up today of 100 moves from Y0 to Y-1 and back, and at the end of it the position error was .001 measured with one caliper and .005 measured with the other. (Measured with the third it was 35.37, though I suspect that we can probably disregard that as it's only a 6" caliper.) X-axis is still not quite right. Same test (except X0 to X-1, obviously) it was off by .018 or so. Haven't tried Z yet, mostly because I haven't quite figured out how to set it up yet.


I think part of the problem may be that the active low / active high might not have been right. The acceleration was already set to 1 in/s^2, so I'm fairly certain that's not the problem. The max speed was set to 50 ipm. Mechanically the system is capable of 300+, although I wouldn't like to try that with a 4-foot-long lead screw.

[jschmitt]

the im483 is opto iso. Is breakout board opto iso? I would try running step and direction from the computer parallel port to one of the drives and see if it works.

[peter.steele]

IM483's are wired directly to LPT1. No breakout board.

Oddly, today I had things running very nicely ... for a while. I had a test pattern set up of 100 moves from x0 to x-1 and back. At the end of the run, it was within <.0005" of actual zero, as measured with my Mitutoyo calipers. I ran this a few times, and got the same results every time. Then I did it with a 3" run instead of a 1" run. Worked great for the first two runs. Same <.0005" accuracy. I was thrilled. Ran it again ... and right at about line 100 or so, the motor just started stalling again. No idea why. Motor wasn't too hot. Drivers weren't too hot.

No idea what the problem is... but it ain't going to be my problem for much longer. I just ordered a G540 and a new power supply today. So, unless it's a problem with my computer, then I should be good to go by the weekend. If it is a problem with my computer, I've got a PCI LPT card I can try, and I've also got an extra hard drive I can format with XP, and I've even got a whole completely separate set of hardware I can build into a whole different machine if I need to. So, I'd say it should be pretty well set. Maybe.