capacitor sensor

[franky]

good morning at all!,

I need to create a capacitor sensor whit ecio40, to measure the distance between a wire and a metal surface...

do you know how can I do??

a lot of thank's!

[Benj]

Hello

I think this should work.

Could the metal surface be grounded. Then all you have to do is output a 1 to an analogue pin with a delay to allow the cap to charge, then convert the pin to an input using a input icon and then start sampling the said analogue pin to allow it to read back the charge remaining on the capacitor.

Peforming the conversion to an input means that the pin is not switching back to outputting 5v in between analogue samples.

When your samples read 0 then its time to display the measurement eg number of samples, and start again.

[franky]

thank you Benj for the answer,
but i'm not sure to have understand, between the "sensor" and the surface there isn't a connection,
is it possible for you to write an example?
best regards!

[Benj]

Hello

The attached example should give you an idea if this is going to work. Really it depends how much capactiance there is. A very low capacitance value will not be measurable by the PIC alone. Your probably talking uF or at least nF before you will be able to measure accuratly with a PIC.

Eg if you are getting the number 0 then there is not enough capacitance to measure. You should also ensure that there is no noise present on the system eg by screening the system in a faraday cage etc. This will help to limit the noise from having basically a open source floating input.

If you are getting values greater then 0 then it should be fairly straightforward to convert this number into a time and then from there into a capacitance.

[franky]

thank benj for your answer
but i can't open your attached file, i don't know why, i'm using V3.6.11.53
can you tell me something?
regards

[Benj]

Here is the V3 version.

[franky]

thank you very much Benj,
I'm going to try something... but i didn't understand how i must to connect the capacitor... the positive lead to port A and the negative? to the ground?

at the end I need a system that give me an analog output that must depend from the distance between the "sensor (a single wire or something else)" and the metal surface..
a lot of thanks in advance

best regards!

[Benj]

Hello Franky

Detecting capacitance is a tricky business and would be hard to do simply using the analogue component alone. Though it may be possible for distances of micrometers etc. If by distance you are talking mm, cm or more then I think you will be better off using another approach such as sonar or IR. This way the surface doesnt have to be metal.

Maybe you could find some similar designs online and see what approach they have used.

[franky]

hi Benj, for distance I'm speaking of minimum of 2 or 3mm to maximum 20mm...
to recognize this distance is it possible for you by a wire that send a wave and reads the time that the wave needs to come back??..

[Benj]

Hello Franky

Sorry I dont think this is possible without going into RF compatible hardware. Even then you are facing many problems. Are there any other example projects out there that use this method?

I think a IR photoreflector similar to the device on our Formula Flowcode would be much easier to acheive in practise.

If the metal surface is magnetised then you could maybe use a hall effect sensor to give you the distance from the surface.

[franky]

Hi Benj,

I can't, i would like to make this system because a lot of company are using it to move up and down a vertical axe in oxycutting machines and it's not possible to take the distance whit other methods.. because there is the oxy flame ...


thank you for your help!
best regards!

[Benj]

Sorry its not something I've ever come across.

Im not saying it cant be done with a microcontroller but I dont know how to acheive it without seeing an example project etc that uses a similar setup. Why wouldn't the hall effect sensor work with the oxy cutting machines?

With my CNC machine it uses motor encoder feedback to maintain its own position so it always knows how far away from the surface it is based on its starting position.

[medelec35]

This is a stab in the dark, and may not be a solution.
What about connecting wire to a LVDT?
That is used in industries to measure movement hence distances.
Although normally used with AC, It could work with two pins of microcontroller working in push pull arrangement or you may need a DAC.
But that is only my theory and could be proven wrong.
Or what about wire connected to a gray scale encoder? so as wire moved up and down it connected to a wheel, which is part of an encoder?
Sorry if not much help.

[franky]

hi at all!
and thank you for the answeres, my problem it's that the sensor will be a ring that must stay around of the oxy torch and it's not enough that the sensor stay near to the torch... so my problem it's not to read the torch's movement but to read the metal surface that can goes up or down, but the distance between the torch and the surface must stay always the same...
thank you for your help
best regards!