Messages

1

Mach3 under Vista / Driver stops responding?

« on: March 02, 2011, 11:04:46 AM »

I was having problems with my original PC periodically locking up solid whilst running Mach3, so I have built a new box.

I am using an Intel Atom D525 @ 1.8GHz on a Gigabyte GA-D525TUD board with 4GB RAM and have switched from XP on the old machine to Win 7 32-bit.

I installed a clean copy of the current lockdown version of Mach3 (022).

I have turned off all power saving settings I can find.

The driver test program shows 'excellent' at all pulse rates. I set up Mach3 at 25KHz for soak testing.

I left Mach3 idle on the Diagnosics tab for an hour and then noticed that the 'time in int' value had stuck at +6.9. It was oscillating a couple of points either side of 7 previously.

Mach3 was still responsive to screen input and I could load G-code, but the DROs would not change value in response to jogging or code.

The test environment is a clean install with no extra plugins. It is configured for the parallel port, but nothing is plugged into the port at this stage.

Looks as though something causes communication with the driver to stop?

Terry

2

General Mach Discussion / Re: I seem to be gaining steps...

« on: March 18, 2010, 05:52:34 AM »

In this particular instance, the feedrate was set to 2mm/minute, so each 'clonk' is the sound of a single microstep. In my case, the problem is the fact that I get a nice regular pace of steps, then at a particular angle the motor spins forward several microsteps in quick succession.

Terry

3

General Mach Discussion / Re: I seem to be gaining steps...

« on: March 15, 2010, 04:05:59 PM »

Yes. It's a fast machine that runs with a stable clock up to 65kHz. There may appear to be some pulses missing in these traces, but this is because the width of the pulses is close to the sampling rate at this timebase, so the scope is not catching all of them. Sampling at higher rates shows a certain amount of jitter between pulses, but an average pulse rate that matches the desired feed rate.

What concerns me is the asymmetric profile of one channel of the stepper drive. I was expecting to see two phase shifted sinusoids, but the B channel appears to collapse each time it crosses the peak and at this point you can hear the stepper jump several steps. I am guessing that the motor lets go as the current falls sharply and gets pulled towards the next whole step.

Not sure if this might be due to a wiring issue in the stepper, or a driver fault?

I have checked my wiring against the manuals for the steppers and the drivers and all appears to be correct, but I have found at least one error in the driver docs so anything could be happening.

Terry

4

General Mach Discussion / Re: I seem to be gaining steps...

« on: March 14, 2010, 04:47:23 PM »

Closeup of artefacts in slow speed trace...

5

General Mach Discussion / Re: I seem to be gaining steps...

« on: March 14, 2010, 04:46:24 PM »

And traces for rapid movement...

6

General Mach Discussion / Re: I seem to be gaining steps...

« on: March 14, 2010, 04:44:48 PM »

Traces attached...

7

General Mach Discussion / Re: I seem to be gaining steps...

« on: March 14, 2010, 03:07:14 PM »

Link should read http://www.kill-9.co.uk/tdc/frankenmill/trace1.html

8

General Mach Discussion / Re: I seem to be gaining steps...

« on: March 14, 2010, 03:04:40 PM »

Ok, here are some scope traces...

http://redpool1.trance.co.uk/tdc/frankenmill/trace1.html

The Blue trace is the X axis step output from the BOB.
The Tan trace is a current clamp placed around one of the phase wires from the driver to the stepper.

The first four traces are 4 runs of the code G1 X0.5 F2 (2mm/minute) with the clamp meter on each of the stepper wires in turn.

Listening to this very slow traverse, you hear a steady "clonk-clonk-clonk" until the point shown on the B+ trace where there is a steep asymmetric drop in current. At this point, you hear a sputter of steps, then back to the slow steady pace.

Can anyone tell what the problem is from the shape of the traces?

Terry

9

General Mach Discussion / Re: I seem to be gaining steps...

« on: March 11, 2010, 04:56:19 PM »

Thanks. I might take you up on that if I can't diagnose the problem with things I have to hand here. I will try putting a scope on the pulse train to see if there is obvious noise or jitter and I will try measuring the current through the coils to see if I am getting what I expect to see. I guess I ought to try a different PC to eliminate the parallel ports etc. The other thing that might be worth trying is disabling the microstepping to see if it is a problem in that area.

Terry

10

General Mach Discussion / Re: I seem to be gaining steps...

« on: March 11, 2010, 12:13:46 PM »

Yes, South Wales.