encoder calibration on a 1756-M02AE

Hi guys,

I need to calibrate two encoders on a plank edger. One is on an infeed belt, and the second is on the rolls and outfeed belt. This is on a L61 v20.18 through a 1756-M02AE card.

I'm stumped because I can't seem to find the tag that holds the encoder pulses. In the axis tag, I can only find the .ActualPosition tag but thats AFTER the conversion and is in feet, so I can't use it to calculate the ratio with (PosB-PosA/traveled distance) method.

Or is there is another method to do this with a M02AE?

The tag reports around 450'/mins for the infeed and I mesured 439 with a taco. The oufeed tag reports 454 when the taco reads 425. So thats why I want to calibrate, we have issues with the press rolls lowering a little bit early on thicker planks and I figured I'd start by confirming the encoders were calibrated.

I did go trough the "MOTION-UM001F-EN-P" manual, I does explain every tab, but there doesn't seem to be an actual calibration procedure that I could follow.

Thanks guys

thebuj said:

... I'm stumped because I can't seem to find the tag that holds the encoder pulses...

Click to expand...

Can you post the program?

No I can't post the program itself, sorry

thebuj said:

I need to calibrate two encoders on a plank edger. One is on an infeed belt, and the second is on the rolls and outfeed belt. This is on a L61 v20.18 through a 1756-M02AE card.

Click to expand...

It's configured as a SERVO with the axis tag data type AXIS_SERVO with a 0-10v VFD, position units are inches.

thebuj said:

I'm stumped because I can't seem to find the tag that holds the encoder pulses. In the axis tag, I can only find the .ActualPosition tag but thats AFTER the conversion and is in feet, so I can't use it to calculate the ratio with (PosB-PosA/traveled distance) method.

Click to expand...

Actually it's in inches. It's seems that since its configured as a servo on a M02AE and not through a high speed counter card, I don't have direct accces to the encoder pulses.

I'll post screenshots soon

I tried using the velocity offset to correct the speed reading. Since I have a 11'/min difference, I inputted 2.2"/sec.

Inputting values did not change the reported speed. I tried many ways: in online running mode, program mode, even offline & re-downloaded the program. Doesn't do anything. I even tried with much larger values but reported speed still didn't change

I then changed the output offset to see if changing any value in this page would work and it did speedup the VFD. So I reverted it back to zero since it won't help me.



So after that, I figured that my only other option would be to fiddle with the K constant.

I decided to try and recalculate it. So I got my encoder PPT which is 7500, measured the belt travel for one rotation, which is 36 inch, aaaand.... i'm not too sure how to calculate that... The formula I have doesn't give me anything near 835, playing around with it, I did end up with 833 but that wouldn't give me the right speed reading anyways...

I did try with trial and error and having a K constant of around 863 seemed to be pretty accurate. The reported speed dropped near to what I'm reading on the tachometer.

I did not leave it at the guessed value though, since I did not calculate it mathematically, I did not feel comfortable leaving it. The position unwind also bugged me, even after reading the help file for that page, I'm still not sure what to put there, and since the way it's configured now, (it looks like it the K constant with thousands and no decimal), I wasn't sure if I have to also modify it.

So if anyone knows how to calculate the K value and position unwind, or has anything I could read to get me on the right track, that would be greatly appreciated.

I did read the manual MOTION-UM001F-EN-P, I didn't find anything that helped me, but maybe I missed something...

Thanks

thebuj said:

No I can't post the program itself, sorry

Click to expand...

No problem. I created my own test program, and it wouldn't have helped.

Page 209 of the manual covers how you would calculate the number of counts per unit (inches, in this case).

I think that in the tuning section you could get it to travel x amount of inches and see what that equates to in actual distance.

Scaling

You have a AB quadrature encoder which gives you 4 count per pulse
You have stated that the encoder is 7500 ppt (Pulses per turn or 360 Deg)

You are trying to determine the encoder counts per inch

Using the base of 36" per rev you get (7500 [pulses] * 4 [counts/Pluse]) / 36 [Inches] = 833.33333 [counts / pulse]

This is your scaling starting point - It is only accurate if the 36" is correct

I verify the scaling by the following
1) Set the position to zero using a MRP or MAH that does not move that axis
2) Mark the belt to a reference mark - we are going to move the belt a long distance that can be easily measured with a tape
3) Jog (or manually) move the belt forward so the mark has moved a long distance
4) Read the current Position [P1]
5) Measure the physical distance the belt moved [D1]

Scaling is recalculated by using
C1 = R1*P1 (Encoder Counts moved = Start Ratio * position)
R2 = C1/D1 (Final Ratio = encoder counts / Distance moved)

typically the longer the distance moved the more accurate the count (assuming nothing slips)

Unwind
this is how many encoder counts the belt must move before the axis position is reset to 0
in your picture the belt will move forward 1000" before the position "rolled over" to 0
Rollover position = Position Unwind / Conversion Constant

Thanks guys, this is great info, this should get me on the right track.

The thing that throws me off is that I don't have the actual encoder count, the M02AE converts it with the K constant before spitting out a value. I tried with a GSV and there didn't seem to have anything in there either.

I can calibrate encoders when I have encoder counts without any problems, it's the first time I had to do it through a M02AE.

Thanks again