842E EtherNet/IP Motion Control Allen Bradley Encoder

I am working with a Allen Bradley Absolute Controller ([FONT=&quot]842E-SIP5BA).[/FONT]
[FONT=&quot]I was wondering if anyone has any experience with them. Pros/Cons? Has anyone experience one going bad?[/FONT]
[FONT=&quot]
[/FONT]
[FONT=&quot]A little background, the one I am working with tends to drift. Basically the encoder is attached to a hydraulic arm. At certain positions, certain events happen. The only position I am have trouble with is the "home" position. Let's say home is at 175 degrees. (Note there are hard stops to help make sure "home" is in the same spot.) I'll visit the machine a couple hours later and the "home" position will be 177.5. This "home" position tends to drift 2-3 degrees. None of the other positions drift. They are always right on.[/FONT]

[FONT=&quot][/FONT]
[FONT=&quot]So I would like to hear if anyone has had a similar experience. [/FONT][FONT=&quot][FONT=&quot]
[/FONT][/FONT]

Is the Home position one that is regularly visited or just the axis setup position? If it is visited every cycle on this motion with disturbance I would teach the position with an offset to see if the problem persists. Also how near the end of stroke is it on the actuator and is it actually resting in the same spot against the hard stops?

If it helps, I recently completed a project with the 842E-M (multi-turn encoder) and it did NOT exhibit this behavior.

In general, an absolute encoder shouldn't be able to exhibit "drift" behavior.

Incremental encoders can; "jitter" on the edge of a code disk slot, or interference in the signal to an encoder input module are notorious and long-standing challenges for distinguishing between shaft motion and signal pulses. The basic quadrature nature of most encoders, and careful low-level signal processing, have largely made that a problem of the past.

But an absolute encoder has an absolute disk. A "noisy" pulse would be negated by no pulses on the other bands on the disk. I won't go into a deep dive on the theory and practice, but in general "that shouldn't happen".

Couplings and mechanisms can slip. Hydraulic valves can leak or bypass. Machine frames can expand and contract and distort. Math used to convert from integer to floating point, or resets of position based on Home sensors, can accumulate error.

I would look at all of those before I realistically thought an absolute encoder was "drifting".

The A-B 842E looks just like the SICK AFS60/AFM60 for a very good reason. They're built by folks who have been in the encoder business for decades, so in general I trust that they work as designed.

Look through the user program to find out if there are any MSG instructions to set the position, or what sort of conversion between counts and inches are being done.