Hello Guest it is October 14, 2019, 10:15:32 PM

Author Topic: Encoder error accumulation  (Read 1813 times)

0 Members and 1 Guest are viewing this topic.

Encoder error accumulation
« on: August 19, 2016, 08:17:37 PM »
Recently I retrofit our Galil 3-axis CNC mill with a CSMIO-IP/A motion controller. After a few months of wiring, tuning and testing we had the machine almost perfectly set.

The issues we had been struggling with were determined to be related to encoder wiring being unprotected from electromagnetic noise and the CSMIO diagnostic windows was giving us error readings which disappeared after grounding the shieldings. Everything was going smoothly until two days ago when we grounded the Y-axis shielding and completely lost the ability to have correct positioning. For some reason now when I even jog the Y axis a few inches back and fourth, we just add up errors until the red counter maxes out. We reverted back to the unshielded y-axis and tried everything possible to even drop down these errors with no success.

I'm starting to go crazy with this encoder counts error business.

Heres a simple diagram to help understand the layout of the system.


Offline Hood

*
  •  25,855 25,855
  • Carnoustie, Scotland
    • View Profile
Re: Encoder error accumulation
« Reply #1 on: August 20, 2016, 05:59:25 AM »
Have you connected only one end of the shield to Earth?
Are you using twisted Pair for the wiring?

Hood

Offline mc

*
  •  381 381
    • View Profile
Re: Encoder error accumulation
« Reply #2 on: August 21, 2016, 04:15:39 PM »
How long is the cable before the connector/line driver convertor?

I'm going to guess by connecting the shielding, you've introduced a ground loop, which is causing interference on the encoder wires before they reach the line driver. First thing I'd do is move the line driver to as near the encoder as possible, as then you minimise he length of wire where the single ended signals can be affected, and get the maximum benefit from the differential pairs/line driver, which is to greatly reduce interference problems.
Alternatively, you could upgrade the encoders to line driver versions, and remove the need for any converter.
Re: Encoder error accumulation
« Reply #3 on: August 22, 2016, 02:57:12 PM »
Have you connected only one end of the shield to Earth?
Are you using twisted Pair for the wiring?

Hood

Yes, the whole machine has a common ground with the outlet.
The 5 wire cable is not twisted until it hits the 5->10 converter, then its twisted pairs.

How long is the cable before the connector/line driver convertor?

I'm going to guess by connecting the shielding, you've introduced a ground loop, which is causing interference on the encoder wires before they reach the line driver. First thing I'd do is move the line driver to as near the encoder as possible, as then you minimise he length of wire where the single ended signals can be affected, and get the maximum benefit from the differential pairs/line driver, which is to greatly reduce interference problems.
Alternatively, you could upgrade the encoders to line driver versions, and remove the need for any converter.

The cable before the converter is roughly 15-20 feet. I think this is where the issue lies, but what throws me off completely is the fact that this axis was running without error with the exact same setup just weeks ago.

I believe you are right, the twisted 10 cable differential needs to be as long as possible. Were probably going to just buy some brand new encoders and just skip the entire conversion step.

Thanks for the input guys, I will update this post when I come up with some results.