Hoping someone here can help me with an issue I'm having with my VMC 20.
I'm having some issues with the motor speed fluctuating at speeds above 6500 rpms, I know that the easy answer is the encoder is bad, however I've put 2 new encoders on the machine same issue. The machine has a glentek spindle drive in it that can be set to bypass the encoder and run off of basically a VFD. If I run in this mode the rpms do not fluctuate.
So I'm back to a problem with the encoder, but I've tried 2 already as mentioned, about the only thing left is the wiring harness. Has anyone out got some thoughts on this issue?
Don't discount the two new encoders as the problem, it wouldn't be the first time I have seen that happen. But, it really sounds like noise, a poor ground maybe? Or it's possible that the encoder input module on the drive is flaky. Or as you suggest it could be a bad connection in the cable.
If you have access to an oscilloscope that is the best way to see what's going on.
This is best solved with an oscilloscope - and someone who knows how to use it.
I suspect that what could be happening is that the encoder signal is trying to transition too fast - I have had that myself. Let's walk through the figures:
Spindle speed: 6600 rpm or 220 Hz - or 4.54 mS per rev
Encoder resolution: say 500 lines per rev?
So apparent pulse width will be 4.54 mS / 500 = 9 uS.
But a pulse consists of two edges, so each edge has only 4.5 uS to transition.
If your encoder is a common one with a simple transistor output into a pull-up resistor somewhere, it may not be able to create a transition this fast at the receiver. There could be too much capacitance on the line - maybe as a noise filter.
Also, some controllers have software filtering enabled at the input, and if this is set to 10 uS then the pulse will simply disappear.
I assume that the machine was running correctly at one time. Be careful with grounding, all grounds should come back to a single point. Ground loops can be introduced if this is not done, and can actually make the problem worse. Also, the cable shielding should only be connected on one end, normally at the machine end rather than the encoder end. In this case, I would expect that the shield is terminated to the drive case.
The original cable should be shielded, so if that is intact then I would expect that it's adequate. If the encoder output is differential quadrature (A, /A, B, /B) then it is almost noise immune anyway.
It is unlikely that phasing would be an issue. The encoder is well isolated from the power.
The only real way to diagnose this problem is with an oscilloscope, you need to see what the drive is seeing from the encoder. These are the kind of problems that drive techs nuts.
Spindle encoder issue!
Reply with Quote
