Another encoder problem..

[scouringpad]

Ensure again that the signaling type from the encoder is quadrature and that it is set up that way in the 1500. Secondly review the filter settings for the inputs. It is possible that the sudden movement is generating pulses faster than the filter setting will allow. You should be able to get correct readings even with very high pulse rates.

The filter setting for that channel is 25 uS. I believe I needed to adjust that when starting this press line a couple years ago as HSC was not getting all the counts on the default setting.

 

[rpoet]

It sounds to me like the vibration described is causing the encoder to dither (not sure if that's the correct term) back and forth between two settings. As in, it's unpredictably toggling between it's two quadrature states even though it's not moving. The PLC is seeing that as an up/down counting signal, thus giving you your random position errors only when stopped.

Since it seems as though you have a pretty good idea as to how much the matting shrinks in between cuts, could you just ignore the encoder when stopped and add/subtract that value as a constant to the position when you start the next move?

-rpoet

 

[scouringpad]

The basic program should be attached as a zip file.

 

[Steve Bailey]

Scouringpad,

Each time Bernie has asked if you're using a quadrature signal you haven't specifically answered that question. Please confirm that's what you have and that the 1500 is configured for quadrature counts.

If you have a quadrature encoder the wires from it should be labeled A and B. If you have differential signals, you should also have NOT A and NOT B.

If you 're not using quadrature counting, how do you swap between counting up and counting down?

 

[scouringpad]

It sounds to me like the vibration described is causing the encoder to dither (not sure if that's the correct term) back and forth between two settings. As in, it's unpredictably toggling between it's two quadrature states even though it's not moving. The PLC is seeing that as an up/down counting signal, thus giving you your random position errors only when stopped.

Since it seems as though you have a pretty good idea as to how much the matting shrinks in between cuts, could you just ignore the encoder when stopped and add/subtract that value as a constant to the position when you start the next move?

-rpoet

Yes, it does seem like dithering, and not being that up on encoders, I was wondering if this is even possible for this type of encoder..? If so, is there a type that would work better?

The problem with subtracting the shrinkage is that we run different thicknesses (3/8" to 3/4"), and the cooling rates from 330 degrees F. are much faster for the thinner material. We also sometimes run a water mist during hot weather to precool before the shear rough cuts. I don't think I could be close enough with an estimate, considering all the possible variables, to be any better off than I am now.

 

[scouringpad]

Scouringpad,

Each time Bernie has asked if you're using a quadrature signal you haven't specifically answered that question. Please confirm that's what you have and that the 1500 is configured for quadrature counts.

If you have a quadrature encoder the wires from it should be labeled A and B. If you have differential signals, you should also have NOT A and NOT B.

If you 're not using quadrature counting, how do you swap between counting up and counting down?

I'm using an A-B 845T-DZ62EEH encoder which is 11-24vdc in/11-24vdc dld out. It has a ten pin connector. I'm using inputs 4 and five on the Micrologix 1500 and it is set up for count with external direction. It counts fine forward or backward when moving, but slight vibration while it sits on the matting is causing counts in either direction. If I lift it off the matting it does not count.

 

[Pego]

and it is set up for count with external direction. It counts fine forward or backward when moving, but slight vibration while it sits on the matting is causing counts in either direction. If I lift it off the matting it does not count.

You are right, this will never work.
Answer the questions and you will get the right answer
If you don't understand the question, just let us know.

 

[Steve Bailey]

Try a value of 6 in the MOD parameter of the HSC1 function file.

 

[scouringpad]

Try a value of 6 in the MOD parameter of the HSC1 function file.

I meant to repost, but have been busy. The MOD value did the trick. Thanks for the info, and thanks to all who posted with suggestions!

 

[cjd1965]

A workaround maybe..
At startup move the material to a new prox just past the shear and cut it. Then use the encoder until the next time bubba stops the line. You will get 1 longer length at start up which maybe a issue but hey, your 1st length is no good now anyway.

Edit: See its fixed. What does the mod do? is this a modulus division thing?

 

[scouringpad]

A workaround maybe..
At startup move the material to a new prox just past the shear and cut it. Then use the encoder until the next time bubba stops the line. You will get 1 longer length at start up which maybe a issue but hey, your 1st length is no good now anyway.

Edit: See its fixed. What does the mod do? is this a modulus division thing?

MOD value is the mode the HSC runs in. I had it set for 2 (count with external direction) rather than 6 (quadrature counting).