making a capacitor from a plastic container

[large_ghostman]

sorry if the title seems a bit mad but bear with me on this one, also sorry if the description is poor but i dont have access to the pc that i need to upload pics, i will upload them tomorrow but for now i will describe what i am doing and if i am barking up the wrong tree.
ok i have been reading as i normally do, i just kind of dive in and start reading i dont say pick a topic and read everything in every book about it, i cant do that it dosnt stick so i have to read odd bits here and there then put it all together.
sorry back on track
i was reading about capacitors, and it occurred to me that in its most basic form there isnt much to them, after all even we have capacitance hence cap touch switches. So anyway i was thinking if i had a small plastic container with water and i stuck a pcb board to both sides with the copper touching the plastic then i have a basic capacitor! now i have tried this tonight with my LCR meter with mixed results, for a start i only had one pcb on one side and one clip on the plastic at the other side and one clip on the pcb, i think the main problem was i have kelvin clips on the lcr meter and it made it hard to hold it all so it all made connection at the same time.
But i did get an increase in cap value at 10kH frequency as i added water. tomorrow i will tape 2 bits of pcb copper side to plastic tub on each side and connect clips to the copper, what i am after is to see if i make a accurate water lever sensor for my fogger, mainly because they are meant to shut off with low water but normally dont so wear out as you can only use a depth of about 2.5cm of water with them.
so i am thinking turn the water bath into a cap and use a pic's CTMU to measure the capacitance! RB's cap meter was pretty accurate so i am guessing using the CTMU i should be able to monitor the water level pretty accurately! and have the pic shut the fogger off on low water and start a warning sound! does this sound feasible?? i know its a bit mad but if i can put it on my addon later list to the incubator project that would be cool
i will add pics tomorrow just incase i have babbled, its late and i am tired so i babel a bit
LG

 

[4pyros]

the copper will corrode, and change value all the time untill it will not work any more.
Try stainless steel, maybe pocket rulers or butter knives.

 

[large_ghostman]

will it corrode if it's on the outside of the plastic? thats what i tried, i dont want to put electrodes inside the container,

 

[jpanhalt]

Water can be a good dielectric when it is pure, but with ionic contaminants, it can be a conductor. You then have much greater problems with hydrolysis, corrosion, and electrical leakage between the two plates of the capacitor. Here is a link to some dielectrics. Look for something that has a high value and very low ionic character. Antifreeze (ethylene glycol or propylene glycol) might be a good substitute for water in your experiments.

https://en.wikipedia.org/wiki/Dielectric_constant#Terminology

John

 

[large_ghostman]

has to be water as it is used for the fogger, although you are meant to use deionized water with them to save the ceramic plates from wear

 

[4pyros]

will it corrode if it's on the outside of the plastic? thats what i tried, i dont want to put electrodes inside the container,

Sorry I didn't realize the plates were on the outside.
Thay may tarnish but probably will not corrode.

 

[JimB]

LG, have a look at this old thread.
It relates to measuring oil/water concentration, but the general principle is applicable to what you are trying to do.
https://www.electro-tech-online.com...ill-run-to-an-led-display.107910/#post-884921

I assume that your capacitor plates were on opposite sides of a large container and separated by quite a large distance. This will give very small capaictance values which will be hard to measure.

As a quick and dirty test, try putting the plates on the side of the container, adjacent to each other with a 3 or 5mm gap.
This will not be a particularly large capacitance, but I am guseeing that you should see a large change in capacitance due to the fringing electric field through the wall of the container and into the water.

JimB

 

[dr pepper]

If the dielelectric of the plastic is very different to the water which it probably is I'dve thought the cap change is going to be really small, in the order of pf's.

I've seen a similar thing that used piezo cells, the area above the water made to resonate with an impulse, then the frequency of the resonance measured, resonance goes down with the water level.

 

[MrAl]

Hello,

I assume you have two plates that are insulated and separated by a distance, and that distance is filled with a varying level of water. So the water acts as a conductor, and the two plates make up two capacitors, one on each side of the water.

First, the capacitance is even smaller than it would be with one plate because the two outside caps are in series. That's a no no right off the bat. To get a better value arrange it so that the caps are in parallel, and increase the number of "caps" to as many as possible. This could involved multiple chambers.

The caps on each side are formed by one 'plate' which is conductive metal like copper, and the other 'plate' on that side is water, and the same on the other side. So as the water rises the plate area increases. As the plate area increases the capacitance increases and thus theoretically provides a means for measuring the depth of the water.

Unfortunately pure water is not very conductive. That means you'd have to use water that has something else in it like salt. But that could clog any mechanism that uses the water for some purpose and might also corrode it in time.

So you might want to look for another method.

 

[rumpfy]

A transformer ratio bridge might do the job for you.
Take a transformer which has a single winding and an accurately centre tapped winding.
The single winding goes to the detector amplifier.
The centre tap of the other winding goes to the output of an oscillator (say 1000 Hz). Each end of that winding is connected to an admittance to ground. the capacitive part of the admittance is formed by a capacitor to a water bath. The water is connected to ground and also the other output of the oscillator is ground. The capacitor can be formed for example by a piece of single insulated wire of the underground telephone cable type. For your BPO stuff, this is about 350 to 450 pF per meter from memory.
The real part of the admittance is a simple pot with its centre connected to the centre tap, and each end connected to each end of the winding. With the admittances balanced, the nett current in the detector winding is zero. Ideally the detector is phase sensitive and responds only to capacitance unbalance at each side of the winding. As the water level rises and falls, the bridge will move out of balance; the output voltage can be calibrated, or, you decide the capacitance value at the 'out of water' point. Use a chemists measuring cylinder for example and suspend the insulated wire as a loop from top to bottom.
These kind of capacitance bridges are used extensively in the manufacture of single insulated conductor for making multi-pair telephone cable. Generally, the measuring tube is guarded and the active element is 10 to 50 cm long.
You can guard the bridge by using coaxial cable to connect to the known and unknown admittances. The outer conductor of the coaxial cable is connected the output of the oscillator.
Hope this helps.
pr.

 

[4pyros]

I would think the larger the plate area the easer it will be to detect changes in the water level.

 

[JimB]

Earlier I wrote:

As a quick and dirty test, try putting the plates on the side of the container, adjacent to each other with a 3 or 5mm gap.

So I did, as shown in the picture below.
The silver coloured strip is something made by a company called Chomerics, I think that it is called "Cho-foil" and if I remember correctly it cost a fortune.
The stuff that I have used in the past could be soldered, but this stuff (which I got in a box of junk) does not seem to like taking solder.

I then connected this to the capacitance test jig:

Adding various amounts of water and measuring the capacitance, gave the following results:

Not a big capacitance change, but measurable and repeatable.

JimB

 

[ghostman11]

lol wait till he see's this Jim not sure if he is going to be pleased or annoyed at you doing his experiment lmao
the foil you mention i have never heard of but i think we have somewhere some of that sticky copper foil stuff and that can be soldered to, if i dont have any left i can get a roll as it isnt that expensive and i have seen it in garden centers for slug control
i will get him to post pics of the container etc but the principle is exactly the same. obviously copper tape would be much better than pcb! but the details i will leave to LG i am not allowed to interfere with his work in our competition!
besides the youngster is going to need more than copper foil to beat this old hand

 

[JimB]

not sure if he is going to be pleased or annoyed at you doing his experiment

More a "proof of concept", just making sure that my mouth (fingers on keyboard) was not writing cheques that I could not cash in practice on the workbench. (Metaphorically speaking).

JimB

 

[ghostman11]

he had it kind of working but i think the kelvin clips made it harder as he couldn't just clip them to the pcb he had to hold both sides at once against the copper, cant find that tape so i will order some, but might be worth a bash tonight with thin pcb board. anyway i am not supposed to read anything relating to the competition! (give it up kid your only going to feel when i beat you )

 

[JimB]

MrAl wrote:

So the water acts as a conductor...
Unfortunately pure water is not very conductive.

Yes, pure water is a good insulator.

But the effect which I am exploiting does not have to do with conduction, it is to do with the dielectric constant (Relative Permitivity).
Air has a constant of 1, but water has a constant of about 80 at room temperature (very temperature depandant).

JimB

 

[MrAl]

MrAl wrote:


Yes, pure water is a good insulator.

But the effect which I am exploiting does not have to do with conduction, it is to do with the dielectric constant (Relative Permitivity).
Air has a constant of 1, but water has a constant of about 80 at room temperature (very temperature depandant).

JimB

Hello again,

Oh ok well then that sounds at least a little better then. If the space between the two plates is kept narrow this may stand some chance of working. The air is displaced by the water and that should increase the capacitance. You could run through a few calculations to find out how much capacitance you will be dealing with and see if it seems practical. There may also be some variability due to the varied conductance of water too through unless the water is from a known pure source.

The next step then would be to build up a little test chamber and do some measurements to see if it looks like it could work.

 

[JimB]

The next step then would be to build up a little test chamber and do some measurements to see if it looks like it could work.

I thought that I had just done that! (See post #12 in this thread).

JimB

 

[MrAl]

Hi,

I thought you did too. What happened is that this browser is not showing images again for some reason. What a pain. It's the new software again.

LATER:
Took a look with a different browser.
Cant you make the strips wider? That may help.
Note: if you post images as attachments i can see them easier.

 

[large_ghostman]

i will post my test pics as well, going to use thin pcb until copper tape gets here. as the amount it will vary that matters most to me, i am pretty confidant that the cap touch units in the pic will be able to read very low cap values, my reasoning is, if they can sense a near touch through plastic of a finger then they have to be pretty sensitive.
What i would really like to do is make a microchip serial analyzer because they have software that goes with it to run the cap touch program that would make it super easy to calibrate and tweak