This is more of a mechanical measuring issue. We use polyester felt in our manufacturing process that is cut to length. In the past we used mechanical measuring wheels that were expensive and never lasted long. Not to mention they were not very accurate.
We recently installed a bunch of encoder with digital counters for foot measurement and most of them do a great job. Not super accurate but good enough for rough cuts. They are all the same model and all but one machine has repeatable accuracy.
The machine in question is our Slitter that cuts the material into measured widths.
In this case I calibrated the counter using 4mm plastic coated felt. Since this is the first step in our process we set it to be approx' one foot over @ 100 feet. An issue started popping up where the material from this machine was coming up short. They would cut 510 to 520 feet and the next machine would measure it as say around 485 feet.
We checked the counter on the second machine a few times. It's good to go.
On the first machine, the Slitter, we found out that the thicker the material the shorter it would be at the same foot count. This is what caused our short lengths on the second line. This might be an easy issue for some but I am stumped. It would be a major pain to relocate the encoder on the Slitter so here I am trying to get some insight on this.
The encoder is spring loaded against a guide roller on the slitter, top of picture. All other encoder wheels are running on flat material that runs through our production lines. Similar to the second machine in the picture.
Anyone see what I am missing here?

 

1) are you sure the wheel is not slipping?
2) are you sure you are not missing pulses from the encoder in an noisy environment? Can you test it without noise (in workshop in small scale test setup for example?
3) do you "digitize" the encoder signal after a long length of cable? Or directly and very near to the encoder wheel?
4) how do you read the encoder signal? With a special PLC input card? Are you sure it does not miss pulses because of timing?

 

Can you enter the material Thickness as a parameter in your calculations ?

 

I presume the face of the felt against the guide roller is not distorting/stretching as it passes by, but the face of the felt against the encoder wheel will be temporarily as it takes a longer path. Try dividing the guide roller circumference by the guide roller circumference with felt wrapped around it, then multiply this by your encoder measurement.

 

I think what you are doing with the spring loaded design is that you are driving it faster than the units that Run against the web.
This is because it is on a radius and the radius changes with the thickness of the material.
Think of it as two gears meshed if they both have 50 teeth then you have a 1;1 ratio. how ever when the driving gear has 51 Teeth and the measuring gear has 50 the measuring gear will show a 2% increase in length measured. 51/50 = 1.02
that is what you are doing by changing the thickness. Measure the roller diameter of your encoder , measure the roller diameter of the driving roller , measure the thickness of the material . Double it and add it to the roller diameter. Use that for you calculation.
e.g. Encoder 3 in Dia Wheel , Roller 3 in Dia wheel , Material 5/16 thick your ratio would be 3.625/3 or 1.2083 .
Use that as a multiplier for the length
Example 100 revolutions of the encoder would be 3*3.14157 *100 *1.2083 /12 = 94.902 feet (C = pi*D)

 

your ratio would be 3.625/3 or 1.2083 .
Use that as a multiplier for the length
Example 100 revolutions of the encoder would be 3*3.14157 *100 *1.2083 /12 = 94.902 feet (C = pi*D)

The way you have calculated the ratio is opposite to what I stated so it will be a divisor for the length: 3*3.14157 *100 /1.2083 /12 = 65.016 feet

 

Interestingly, for a 4 mm web (0.1575 in), and 3" roller, the encoder wheel sees a 5.25% increase in measured web length when measured opposite the drive roller compared to the encoder on the same side as the drive roller.

Given your data of 510 FT and 520 Ft on the one machine with the encoder opposite the drive roller, you have a increase of 5.15% to 7.22%.

 

Hey guys, I somehow missed the notification of replies. I appreciate your responses and I actually feel pretty dump for not seeing the issue right off the bat.
It was the bending of the felt that made it longer on the outside while being measured. The thicker felt really made it bad.
I should have looked at this differently.
This machine has been like this for over 20 years, it just had a different measuring device in the same configuration.

I keep forgetting that a lot of the original designs were flawed.
This is one of many.
The operator on this line noticed something that others didn't see or care about

I moved the encoder to a flat spot before the cutting wheels and everything is good.

 

I appreciate your responses and I actually feel pretty dump for not seeing the issue right off the bat.

Glad we could help you solve this problem. The forum is here for people to ask any question. To quote Carl Sagan, "There are naïve questions, tedious questions, ill-phrased questions, questions put after inadequate self-criticism. But every question is a cry to understand the world. There is no such thing as a dumb question".

 

River, a hands up for the operator who caught it!

 

The way you have calculated the ratio is opposite to what I stated so it will be a divisor for the length: 3*3.14157 *100 /1.2083 /12 = 65.016 feet

The measured lengths are increasing not decreasing , 65 is <100 , and the result should be greater than 100.

 

The measured lengths are increasing not decreasing , 65 is <100 , and the result should be greater than 100.

See detail in OP, measured length: 510 to 520 (too high), actual length: 485

They would cut 510 to 520 feet and the next machine would measure it as say around 485 feet.
We checked the counter on the second machine a few times. It's good to go.