axis creeps in MPG mode - Galil card

When switching to MPG mode (manual pulse generator) the axis has started creeping. machine is two years into a refit and this problem just cropped up.

There is a standard manual pulse generator wheel with just A+ and B+ channels and of course +5V and ground. There MUST be a line noise problem. I see the A+ and B+ wires are NOT on a twisted pair shielded cable of their own, just asking for trouble.

It might work to just change to a twisted pair shielded cable, but maybe not. At this point its going to be quite a job to pull in a new cable (conduit is full - need to add another) so I do NOT want to do it twice. Differential encoders are known for FAR more noise immunity.

i seen this offer for a differential line driver
https://www.usdigital.com/products/a...rfaces/PC4-H10
or
https://cdn.usdigital.com/assets/dat...68589686344445

Has anyone ever used this product or something similar? Or is there a better route?

Karl

I have used those US Digital line drivers before. They do work.

But before you do that, try pulling the A/ and B/ to either 5V or GND on the Galil, I normally use a 1K resistor for this.

If you go into the Galil terminal software you should be able to watch the drift on the counter for the axis you have the MPG connected to.

Thanks for the great idea. i will try this today.

What about a very small capacitor on the signal lines? Pulses are only maybe 4 per second maximum.

Karl mentions just A & B channels as in my experience that is what most MPG's have, I would also decouple the 5v supply with a ~10µf capacitor at the MPG, could be picking up some pulses from noisy source, generally shielded cable is used also.

You could try a capacitor, Al's suggestion is good also.

I was suggesting pulling the Galil end of A/ and B/ high or low rather than the MPG end.

Another option is to get a MPG that has a 4 channel output. That's what I use.

https://www.ebay.com/itm/19371017463...EAAOSwKWJfh-wo

OK tried the resistors, made MPG drift much worse. Almost continuous run or jog.

I had been working on a scale for the knee. This problem came up while doing that. i noticed the +5V and common laying right beside the A+ and B+ signals. moved these wires and problem went away. I now note if in MPG and I move the knee, that axis begins drifting. So, I won't do that.

It would be best to run a shielded twisted pair wire but the conduit to the operator panel is full, no way to pull in another without removing the whole bunch and pulling all in at once. So, I'll just say a quick prayer and hope problem just goes away. My luck on just hoping has not been that good. I'll bet this comes up again right when doing a critical rush job.

Add a conduit maybe? :rolleyes: