Go Down

Topic: How to measure the capacitance in shunt mode? (Read 5015 times)previous topic - next topic

dotnet

Jan 06, 2014, 09:24 pm

First off, I have built a basic LC tank with a comparator to identify an unknown capacitance added to the tank in parallel using resonant frequency. This is working very well with an unknown physical capacitor in the tank.

Now I would like to take it further to measure the unknown capacitance between two copper plates with a certain distance between them (I know this will be in the ~0.001pF range) but I do not know how setup the plates and how to replace the physical "unknown" capacitor in my LC tank circuit with the plates...

MarkT

#1
Jan 06, 2014, 11:04 pm
0.001pF?  Really?  Stray capacitances are larger than a femtofarad (except for very tiny geometries).

A 1cm diameter sphere has a self-capacitance of 0.556pF for instance.
[ I DO NOT respond to personal messages, I WILL delete them unread, use the forum please ]

dotnet

#2
Jan 06, 2014, 11:23 pm
I am trying to measure the capacitance between two plates at a distance of 500mm, where each plate is 100mm2, which theoretically would be around 0.00177pf...

UnoDueTre

#3
Jan 06, 2014, 11:55 pmLast Edit: Jan 06, 2014, 11:57 pm by UnoDueTre Reason: 1

I am trying to measure the capacitance between two plates at a distance of 500mm, where each plate is 100mm2, which theoretically would be around 0.00177pf...

Assuming the dielectric constant of the material between the plates is air, then yes you should have around 0.0018pF, however accurately measuring that is another story.
Even if you use a very large inductance to bring the resonant frequency down to the Arduino's capabilities, any humidity in the air or even other metallic objects (and hands) near the capacitor will have an effect on it leading to erroneous results/readings.

TomGeorge

#4
Jan 07, 2014, 02:53 am
Hi, I agree with other comments posted, that value of capacitance if measurable with a suitable equipment could only be accomplished in a laboratory.
Why do you need a capacitance so low?
If you are trying to make a tank that will resonate at microwave frequencies then you are barking, the wrong tree, at those frequencies the technology is completely different.

Tom....
Everything runs on smoke, let the smoke out, it stops running....

jackrae

#5
Jan 07, 2014, 11:23 am
Sounds a bit like a university thesis "challenge" where the "correct" answer might well be "there is no simple answer"

dotnet

#6
Jan 07, 2014, 11:26 am
Thank you for all the responses so far...

Yes, it is air between the plates and yes, I will have aprox 1-2H to bring down the frequency to a measureable state. PWM will be used to distinguish the signal at the receiver end.

I did left out an important part, I am only interested in the relative values. For example, first I´d like to measure the frequency when both plates are separated by 500mm, then in my "world", frequency would increase/decrease relative to changes in distance?

I do not have any problem with any surrounding objects, e g. capacitance from body etc.., in fact, it is welcome in my cirquit, since I also want to measure any external disturbance...

I do know there might be a lot of things I have not accounted for my little crazy idea, but I do have a theoretical idea I would like to test and I cannot throw it away until I have proven it not to work.

UnoDueTre

#7
Jan 07, 2014, 12:04 pmLast Edit: Jan 07, 2014, 12:27 pm by UnoDueTre Reason: 1
So you have a 0.0017pF cap with around 2nH of inductance and you want to resonate them.
Have you worked out the frequency?
It works out to 86GHz!!!! (that is waveguide territory)
Perhaps you meant 2uH which will give you around 3GHz which is still extremely high and certainly not something anywhere close to what the Arduino is capable of, infact the frequency will be so high that any oscillating/amplifying/detecting circuit will call for microwave techniques probably using alumina substrate board and microstrip.

The other problem you will have is the distance between the capacitors plates.
They are very far from each other which will result in a very big path loss at the resonant frequency and if you decrease that frequency, then the plates will make for very inefficient radiators which means having to use lots of power to overcome these losses.

If you did mean 2H of inductance (giving 3MHz) then have you considered how little capacitive coupling you will get between the plates?
My guess is not very much.
Your L/C ratio is also totally off and not to mention that with an inductor of 2H, there will be interwinding capacitance which will be far far greater than your 0.0017pF cap, that it will totally swamp it.
Although there are many combinations of inductance and capacitance that will (theoretically) give the same resonant frequency, in practice if the ratio is too "skewed" then the Q of the resonant circuit will be too low and you will get very unexpected results.

I certainly don't want to put you off experimenting, but have you considered other approaches?

dotnet

#8
Jan 07, 2014, 01:08 pm

Yes, 1000-2000mH was the plan..

Hmm, I guess you have a point there and to increase the capactive coupling, I need to increase the voltage a lot (which is not an option). I do have another plan B which involve a larger plate with a smaller distance between, back to the drawing board I guess.

Btw, would it be a more viable solution to exchange capacitance for inductance instead? I have basically no experience in either areas, as a senior software programmer with just a few years in electronics I try to solve some odd ideas I have with some unconventional ways and/or the lack of understanding of how it really works and by that way maybe find a way around where an educated wouldn´t go since the books say no. Anyway, this is one of them, and if it doesn´t work, I will at least have learned a lot during the journey... =)

UnoDueTre

#9
Jan 07, 2014, 01:24 pm
Nothing wrong with unconventional approaches but one still has to be realistic.
Changing inductance might solve some of the problems but would create others.
Have a look at the attached equivalent circuit (simplified) to see what I mean.

Perhaps if you gave more details as to what you are trying to achieve, other people could give pointers as to other ways of trying to achieve your goals.

polymorph

#10
Jan 08, 2014, 12:31 am
Voltage does not affect capacitance.

You don't have to tune the resonant frequency of the circuit down so far. Let it be high, and use a mixer to convert it down. Or an external counter.

Make it part of an oscillator. Then mix it with another very stable oscillator running at only 1MHz higher or lower in frequency. Bandpass filter the result.

You'll need a high Q air core coil. But temperature stability is going to be a huge problem, most likely swamping small changes in the distance between plates.

So you -want- it to shift due to nearby objects?
Steve Greenfield AE7HD
Drawing Schematics: tinyurl.com/23mo9pf - tinyurl.com/o97ysyx - https://tinyurl.com/Technote8

dotnet

#11
Jan 08, 2014, 09:42 am

My idea was to investigate if you could build a multitouch frame using an array of electricfield sensors, e g. multiplexing a 50´ish sensors creating thousands of intersections where I not only have a "beam broken", but also have the Z dimension for 3D interaction. If it would work, it would require less components compared to a regular infrared array compared to the one I have already built and could be made cheaper and easier to assemble, and not to forget, easier to hide. I have seen some companies not long ago experimenting with large plates with small distances laid out in a grid pattern (perfect to embed in cellular) and I thought I could take the same idea and mix with the concept of an infrared array. The main problem lies in the distance between the plates, so the concept might work better if they would be arranged in another pattern that I originally thought of, along the sides of the 25" screen (or hidden behind if possible).

TomGeorge

#12
Jan 08, 2014, 09:54 am
Hi, are you trying to make a touch sensitive switch, capacitive sensor?
If so put   "capacitive sensor" in the Search Arduino tag at the top of this page.

I hope you are not trying to re-invent the wheel?

Tom.......
Everything runs on smoke, let the smoke out, it stops running....

dotnet

#13
Jan 08, 2014, 10:27 am
Not exactly, I cannot find the url for the mobile version I saw some weeks ago, I´ll post it when I find it. In the meantime, this was my inspiration for the project, http://web.media.mit.edu/~jrs/efs.html

polymorph

#14
Jan 10, 2014, 06:04 am
That's a capacitive sensor.
Steve Greenfield AE7HD
Drawing Schematics: tinyurl.com/23mo9pf - tinyurl.com/o97ysyx - https://tinyurl.com/Technote8