Drives gone wild-mach2 Gecko 203

[MillingMaiden]

I have installed a new set up on our router consisting of 3 Gecko 203's and a C11g- mulitfunction breakout board from cnc4pc.com and am using Mach 2. I have configured everything (correctly I thought) and have all axis moving. They sound good and travel quietly in all speed ranges and are maintaining distance accuracy. But then all of a sudden one axis will start growling and moving by itself without being told to. It is random.
After I leave mach 2 the motors will still chatter.


Any ideas on where I need to start

[Art Fenerty]

Hi:

It sounds liek your very sensitive to noise for some reason. Id check your printer port is putting out 5 volts, and not 3.5 volts. If it IS 3,5 volts, check with the breakout board maker to see if that one works at 3.5 volts.
That or power supply problems are the only causes I can see for a symptom of that nature..

Art

[pminmo]

is it always the same axis?

[extrapilot]

MM- you might want to consider just going with a dumb breakout board with the 203V- they are rated for TTL level signaling at very low currents, so even modern parallel ports can drive them well. I just did that- with the el-cheapo breakout from CNC4PC, direct wired to the 203Vs, and the thing is amazing. I did have a solder bridge on the breakout though, so you might want to have a look at your current (or replacement dumb) board to verify clean and solid connections...

If you are worried about the whole runaway thing with a rebooting PC, you can always just power down the power supply during normal reboots, and/or put some limit switches in that physically trigger the 203V disable inputs in case of an XP crash. I have never seen any runaway condition on my system, but I guess it only takes one.

Anyway, Im a big believer in keeping it simple, and if you have the 203Vs, there is not much reason to use a buffer/level translator in the circuit IMO.

Regards,
Rob

[plunkg]

Every time I've seen this kind of behavior it's been due to noise induced in the logic end of the system. First thing I would check is that all the grounds are actually connected to one good system grounding point. Grounds loops can sometimes be created through long runs of cable, especially when you have ground points at each end. Sometimes lifting one will solve the problem and make the noise go away. Poor quality signal cables can also cause this, as can running signal cables near sources of noise. I've seen a simple rearranging of cables cure this kind of thing, but it implies a larger problem - usually (again) in the grounding of your system.

If you have access to an oscilloscope, that's the best way to find the source of the noise.