Encoder problem?

[pauls]

I have one of Aaron's early machines which has produced some incredible parts. I was ready to machine today when the x axis decided to quit. What's happening is that the ballsrew-encoder knob is moving is short jerking motions and the Gecko fault light goes on then off. The belt from encoder gear to servo is tight and not slipping and there is no slop in the ballscrew to encoder gear connection.If I try to run a program the x axis works very poorly and I'm losing position. If I try to jog position with the keyboard it jerks in short increments. If I turn off the power the x travel moves smoothly via the knob on the ballscrew.The servo motor runs great, I've swapped Geckos with no change. The y and z move just fine. I'm using Mach III. The machine has been cranking out parts for months without a glich until today.
Any ideas or trouble shooting hints would be welcome.
Thanks
Paul

[ViperTX]

WAG.....the encoder disc may have some debris on it....

[Cruiser]

Or ... your getting cold out in the shop and your finding out that the gib adjustment on the x is a bit too tight. I found this to be the case on my z yesterday. otherwise you might need a touch of adj on gecko. the way the encoders are sealed up, i just can't see it to be the problem till the simpler possibility's are tried first.

[Bubba]

I have had a similar problem on my Y axis, but over time, I think it might be Temp/Humidity related. The problem seems to occur when we have had a hard rain and the temps fall below 40. I am begining to wonder if it might not be a condensation problem (I have a bad one in my shop). Lately, when this problem occurs, I have been able to power up the controller and let it set for a few minutes and then things start working properly.

[MAX711]

I had the same problem on my Z axis. After troubleshooting everything from the encoder back, it turned out to be my breakout board. Evidently some of the earlier CNC4PC boards didn't provide enough current to drive the Geckos reliably.

[pauls]

I tried your suggestions and still a problem. I've used a Campbell board for the past two years and it works just fine, it all checks out. I took apart the encoder and the disc is clean. The shop temp and humidity have been relatively constant and even with the gibb screws loose the problem persists.
Any other suggestions would be welcome.
Thanks
Paul

[MAX711]

Try changing to different pins on the breakout board for step and direction. I know that probably doesn't make sense but it worked for me in the end. Even after I upgraded to the latest CNC4PC board I had the same problem. Then just out of desperation, I changed to different pin outs and bingo! it worked. Can't explain it, but it just worked

[ViperTX]

Check the cabling between the encoder and the Gecko.....

[Richards]

Connect an oscillscope to the motor and encoder and watch the signals. When the fault light comes on, the axis is at least 128 pulses off. I've had that same problem from time to time on my test benched servo w/G340. It turned out to be loose connections.

[pauls]

Thanks to all. I don't have an ocilliscope so that's out for now. One of the members suggested it might be a bindig issue with the ball screw. That just may be it, but it move manually fine so I don't know.
I'll take a look and let you all know if I find the answer.
Paul

[Cruiser]

This one died in a hurry, i am curious as to what the deal actually was ?

[pauls]

Sorry I didn't post the solution. I spoke to Aaron and once again he knew just what it was. The brushes in the servo left thick deposits which were causing suboptimal contact. Once the commuter was cleaner with an eraser everything works again like a charm.
Paul