Distance sensor for diameter compensation

[lukin1977]

We have in operation one re-spooler machine to wind welding mig wire on 15 kg plastic spools

The machine consists on a un-winder, a dancer unit and a winder

As the unwinder gives wire to the winder the diameter of the wire coil decrease. To compensate the diameter reduction there is installed a sensor (see picture attached)

The sensor is very old and it was custom made for it. I am not sure what kind of sensor is, but I am guessing is an ultrasonic sensor

The signal from the sensor is unstable (it varies +/- 1V ) while the coil is spinning and this is making the speed syncronism between winder and unwinder to become unstable too
My first thought was that the problem was a signal interference through the cables so I made a few measurements and the signal is coming unstable from the sensor it self

Is an ultrasonic distance sensor good for this application?
Is there another type of sensor that will not be affected by the spin of the wire coil?

Any help will be apreciated

 

[Skogsgurra]

Ultrasonic is very common for diameter/acceleration compensation. It works well for papr and foil where the sensed surface is smooth. I have never used ultrasonic on wire spools, the dancer is usually able to do a decent job without the compensation.
If your spooling process takes long time in comparison to the drum speed, you should be able to filter out the variations with a simple first order lowpas. I would try a 10 k resistor and a 100 uF capcitor. You may need to experiment with other time constants or something else if your signal is a 4-20 mA. But still - a low pass filter usually works qute well.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[lukin1977]

Thanks Skogsgurra
I will buy those components and make a Low Pass Filter and see what happends

 

[Strong]

If the sensor you are referring to in the center of picture is using ultrasound, then I would expect it to be measuring distance to copper wire on reel and not rotation speed. There may be a sensor on the left side of the photo that measures speed.

Walt

 

[Skogsgurra]

We are talking diamater already. No need to confuse. It only attracts nebulous comments.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[Strong]

"It only attracts nebulous comments."
My apology!
Walt

 

[lukin1977]

The sensor is for distance only. The speed of the unwinder is sense by a tachogenerator on motor´s shaft

 

[Skogsgurra]

No probs, Walt.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[lukin1977]

@Skosgurra. If the sensor if faulty and sending unestable signals.
The filter will stabilize those signals?

 

[Skogsgurra]

Probably not. A faulty sensor seldom outputs a correct mean value with superimposed noise.

The filter removes fast variations that the uneven coil surface causes. Nothing else.
It is very easy to find out if the sensor works correctly. Direct it towards a smooth surface and measure output as a function of distance. If the sensor is healthy, you will get a linear characteristic.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[lukin1977]

Thanks. I did the test and the sensor is working correctly. Is the uneven surface that is causing unstable readings

 

[Skogsgurra]

Did you also test with a filter?

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[lukin1977]

Is this the filter your are suggesting?

 

[Skogsgurra]

Couldn't open that file. Try the "Upload image" in the tool bar.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[lukin1977]

 

[itsmoked]

Yep, that's what he means.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Skogsgurra]

Thanks, Keith!

Was AFK for a while.

A 100 uF capacitor is almost certainly an electrolytic one. Make sure that you get the polarity right.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[lukin1977]

I have confirmed that the sensor is working good. The problem is the uneven surface of the wire reel
I put an empty reel. Made the unwinder spin. The output of the sensor stays on same value. No more unestable reading

I could not try the low pass filter yet because we just run out of wire. We will receive it next week

The components I have to build the filter are:
resistor: 10k ohm - 1/4 watt
capacitor: 100uF - 25 V

picture attached

 

[itsmoked]

That should mush your signal into about a 1/2 to 1 second average that should swamp the dithering out completely. Hopefully you have a voltage out sensor and not a 4-20mA one.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[lukin1977]

Yes, The sensor output is 0 - 10V but the voltage range that it is used is:

Full reel: 2 V
Empty reel: 7,2 V