capacitive level sensor measuring problem

[abfer]

We are using a single probe capacitive sensor to measure the level of a grain. The grain is filled in a box which is installed on a machine. When the machine is running for a while we make the calibrations and there is no problem. But when the machine stops for a while and then restarts the sensor does not measures the level correctly for a while and as time passes it measures correctly again. I'm suspicous of the earth. They try to reorganize the earth connection but no change. When the machine(s) first start there is zero potential difference between the earth and neutral. As time passes it grows up and reaches to 2Vs. I don't know if it's the reason but the change in the sensor follows the change of the potential difference between the earth and neutral. But we did not observe the same issue in another place even the difference changes more than that. What may be the problem? What should we care of? What do you suggest?

 

[itsmoked]

I would think you are having static issues, (Triboelectric), not leakage.

As for the 2V.. Can you tell us more about the machine? Got a picture?


Keith Cress
kcress -

http://www.flaminsystems.com

 

[abfer]

I also thought of static but there should be a protection but it also depends on the earth.

Here is the picture

http://www.dsn2006.org/catalog/pics/Double_Deck_Roller_Mill.jpg

sensor is on the upper part of the machine inside of the box.

 

[controlnovice]

I think it is a static issue as well, but I can't see the picture from work.

Is this a continuous level, or a level switch?

I've used Drexelbrook RF (capacitance) level devices for areas where product buildup or static was an issue. They have worked well for me.

______________________________________________________________________________
This is normally the space where people post something insightful.

 

[abfer]

here is another picture

http://img00.madeinchina.com/products/425/724/9425724.bak.gif

 

[abfer]

It's continous level

 

[IRstuff]

That link didn't work for me either. Please try the recommended approach at the bottom of the dialog box.

TTFN

FAQ731-376

 

[itsmoked]

Guys, it just looks like a giant heavy steel automated teller, with an upper feed window and a lower outlet window. Maybe 6 feet wide, 7 feet high, five feet thick. The picture didn't help much other than show "its a machine with some inner space". Scale dimensions I give are totally a guess as there is nothing, at all, to compare with.

abfer; If you have a capacitance based sensor it is looking at congregated charge. A hopper full of 3kV grain could easily effect that sensor in various ways.

Does the grain fall down a shoot(sp?) to the machine? If it does you could try a few things to test the theory. One would be a very fine water mist being injected. Like steam fine. Or even a humidifier's output. There are some ornamental ones that put out an ultrasonically created water smoke. Alternatively there are static reduction fans you can buy off the shelf. They blow electrically charged air. We had to use them all over a leather tannery to kill the charge created by leather sliding across certain plastics. You could even try damp grain. I offer these suggestion only as a hypothesis test so you could prove what the issue is.



Keith Cress
kcress -

http://www.flaminsystems.com

 

[abfer]

Ok machine is a big one and there are 26 of them. Each have 2 sensors on it. The grain comes from a silo and fills in the feeder on top of the machine where the sensor is fitted in. Moisture of the grain is critical and we cannot change it randomly. The grain comes from the top and moves on through the bottom and the circulation goes on continously. The sensor is helping to adjust the speed of the flow by sensing the level inside the feeder (hopper) and sending that info to the controller.

 

[hacksaw]

this is an old problem in solid handling, we used load cell feeders, you cannot eliminate the charge build up with flowing grain as you have discovered

 

[roydm]

I suspect the static is causing an offset in your amplifier, I would try decoupling the probe from the electronics with a capacitor. Have something like a 1 MegOhm to ground connected to the probe and a 0.1 microfarad from probe to the electronics.
Regards
Roy

 

[sparky970]

Is it feasible to use a different type of sensor. Maybe a bindicator.

 

[itsmoked]

Bindicator... I sat next to the guy who designed their system..

Keith Cress
kcress -

http://www.flaminsystems.com

 

[abfer]

What is it? It seems to be a brand with different kinds of products? What's the difference of their products and which product do you offer?

 

[itsmoked]

It was many years ago and was based on a cable hung in silos. A large cylinder was used to pull the cable up. The grain locks the cable zeroing the weight of any cable still in the grain. A load cell weights the remaining cable that is hanging. Cable weight math does the rest.

Keith Cress
kcress -

http://www.flaminsystems.com