Measuring the angle of a dancer arm

[SteinCollector]

Hi guys,

We have a customer that is using a Yaskawa MPiec Motion Controller with some remote I/O to do a bagging application. They are using a dancer arm to keep the stream (or web, I guess, I'm not a printing guy)of uncut bags taut. They are looking for a way to measure the angle of the dancer from a preset point so that they can move it back to the right angle.

I can add I/O to do analog or some other input signal. But I haven't seen any sort of "angular displacement sensor" that I could use.

Any suggestions for a company that makes something like this, or maybe some other suggestion for detecing angle?

 

[daba]

Hi guys,

We have a customer that is using a Yaskawa MPiec Motion Controller with some remote I/O to do a bagging application. They are using a dancer arm to keep the stream (or web, I guess, I'm not a printing guy)of uncut bags taut. They are looking for a way to measure the angle of the dancer from a preset point so that they can move it back to the right angle.

I can add I/O to do analog or some other input signal. But I haven't seen any sort of "angular displacement sensor" that I could use.

Any suggestions for a company that makes something like this, or maybe some other suggestion for detecing angle?

Loads of them... try this

 

[jdbrandt]

Hi guys,

We have a customer that is using a Yaskawa MPiec Motion Controller with some remote I/O to do a bagging application. They are using a dancer arm to keep the stream (or web, I guess, I'm not a printing guy)of uncut bags taut. They are looking for a way to measure the angle of the dancer from a preset point so that they can move it back to the right angle.

I can add I/O to do analog or some other input signal. But I haven't seen any sort of "angular displacement sensor" that I could use.

Any suggestions for a company that makes something like this, or maybe some other suggestion for detecing angle?

Printing or not, a dancer is used to detect changes in web position caused by any variety of things, but, mostly, coil diameter tracking.
BARDAC (www.bardac.com) makes a dancer controller which would simplify you're life, if this is your first foray into dancer-feedback unwind tension (speed trim) control.

 

[rkukl]

http://www.montalvo.com/products/sensors/dancer-position-sensor-inclinometer/

I've used several of these Inclinometers from Montalvo. They try to sell them as part of their tension control systems but will sell them separately. No moving parts pretty easy to setup.

 

[bkottaras]

http://www.montalvo.com/products/sensors/dancer-position-sensor-inclinometer/

I've used several of these Inclinometers from Montalvo. They try to sell them as part of their tension control systems but will sell them separately. No moving parts pretty easy to setup.

We switched from rotary encoders to inclinometers.
Turck
B2N85H-Q20L60-2LU3-H1151
2 analog outputs (axis depended, M12 Quick Disconnect, easy to wire and use, no calibration needed).
Just an analog module and you're done.
Never going back to encoders as it seems that these sensors do the job without any drawbacks.
We are running web by the way (not printing).

 

[kekrahulik]

I usually keep my dancer feedback devices as simple as I can. A simple potentiometer type setup with a 0-10Vdc or 4-20ma signal wired back to an analog input does well. You can directly mount the rotational device to the pivoting "axle" of the dancer or connect it through a timing belt.

I know we have used "Durapot" devices before. I think they were a little salty, but there are cheaper options out there like it.

The first thing you want to look at is how much can the dancer physically rotate - 45 degrees, 90 degrees, 180 degrees? For the best resolution, you want the span of your feedback device to closely match the maximum rotation. If you mount the device with a timing belt, you can use sprocket ratios to better match rotation.

For example, if you would use a simple potentiometer coupled to the pivoting member of the dancer with a 0-10Vdc signal, a 10 turn pot would provide horrible resolution for a dancer that can only move 90 degrees and that you want to precisely control to a 45 degree setpoint.

 

[rkukl]

HI bkottaras. Thanks for your reply. Checked out the Turck inclinometers and lo and behold they are the same ones Montalvo sells for 2x the price! What a suprise

 

[bkottaras]

HI bkottaras. Thanks for your reply. Checked out the Turck inclinometers and lo and behold they are the same ones Montalvo sells for 2x the price! What a suprise

We have used pots, encoders and such but ever since we switched to the inclinometers, we haven't had an issue with the dancer position control.
Those things seem to be solid and as you have already mentioned no moving parts or further adjustmemts need to be made to get the system to work.

 

[Manglemender]

The only thing I would say against the inclinometers is that the response time is 100mS. This is a long time in control loop terms and won't be suitable for all applications. Don't run your control loops faster than the update rate of the feedback or the loop will over integrate continuously.

In years gone by, an LVDT was the device of choice but for a long lime now a potentiometer, either rotary or linear has been more popular.

Nick