Variable capacitor

[Tomata]

Iam thinking on how to build a hv variable capacitor for 100kv. I found the patent of tesla 464,667 and 567,818 but i was thinking to take an air variable capacitor and immerse it in oil instead.
What is your thoughts? Did you tried to build one? Any ideas?

[OrionLightShip]

Iam thinking on how to build a hv variable capacitor for 100kv. I found the patent of tesla 464,667 and 567,818 but i was thinking to take an air variable capacitor and immerse it in oil instead.
What is your thoughts? Did you tried to build one? Any ideas?

This is something I always wanted to do and may need soon, so I will be following this thread. My only thought is that you need to research the breakdown distance and dielectric constant of the oil you intend to use to know how big this will be for the values you need. Utility companies of course used to use PCB's but I don't know what they use these days.

A table on wiki shows the dielectric breakdown strength of a diverse range of materials. Distilled water looks really good if there is no corrosion expected with the materials you use.

Orion

[skaght]

Distilled water...

Distilled water probably won't work. The tiniest solute dissolved in the distilled water will make it way too conductive. You would need deionized water and a very clever processing technique to stop any possible contamination. Standard distilled water purchased from the store, typically will still carry a current.

[OrionLightShip]

yup, you are right. You'd have to run it through de io resin filters every day. Whatever one chooses needs to be balanced keeping in mind dielectric constant along with dielectric strength. Decisions, decisions

Orion

Distilled water probably won't work. The tiniest solute dissolved in the distilled water will make it way too conductive. You would need deionized water and a very clever processing technique to stop any possible contamination. Standard distilled water purchased from the store, typically will still carry a current.

[Kokomoj0]

yup, you are right. You'd have to run it through de io resin filters every day. Whatever one chooses needs to be balanced keeping in mind dielectric constant along with dielectric strength. Decisions, decisions

Orion

yup maybe score one from a scrapped edm or something, but they often save them for spares, they have the resin deionizers on them.

[OrionLightShip]

After seeing the table, I will probably just stick with polyethylene or just goe with off the shelf products. I didn't realize oils have such low strength and water is too big of a hassle.

yup maybe score one from a scrapped edm or something, but they often save them for spares, they have the resin deionizers on them.

[mahlonritt]

I hadn't heard about de io resin filters before.So can you provide me some details about it..


Usana new Zealand

[OrionLightShip]

Mahlonritt - I just sent you a private message

I hadn't heard about de io resin filters before.So can you provide me some details about it..

Here's a link from Wiki that has a paragraph on it. In the lab, at least, a source of pure water is needed so it is normally put through a reverse osmosis membrane filter and then through a couple of resin beds. The resin beds have to be regenerated occasionally. They take out cations and release H+, another resin takes out anions and release OH-

You end of with really pure water of low conductivity which is usually read from a meter calibrated in mho's. which is the reverse of ohm's or resistance. mhos = 1/ohms

Now I think they measure conductivity in Siemens/meter. I'll stick with mhos its easier.

Orion

[Tomata]

If to use oil as a dielectric which is the most probable the breakdown voltage
will be around 15kv per mm.That is 6.6mm minimum distance between the plates.

Is the frequency plays any important role on the breakdown of dielectric? On the current project will be 500khz.

[OrionLightShip]

I'd say that frequency has nothing to do with it, but then, I am NOT an expert.

That said, you need to look at one more thing. You need to find the dielectric constant for the oil you use, and using the standard formula for calculating capacitance, determine the amount of plate surface area per picofarad, nanofarad, microfarad, etc....

Without the formula in front of me; I'd be willing to guess that a 6.6mm gap will not give you much capacitance at all unless you use multiple, large plates. It just depends on what the maximum capacitance you need will be.

Also, you are saying the oil you use will be 15kv per mm but the table I saw said it could be as low as 10kv per mm so you need to be careful there or you will under design and with high voltage you should always overdesign.

Orion

If to use oil as a dielectric which is the most probable the breakdown voltage
will be around 15kv per mm.That is 6.6mm minimum distance between the plates.

Is the frequency plays any important role on the breakdown of dielectric? On the current project will be 500khz.

[Kokomoj0]

why not buy super thin single sided pc board material, polish the copper and use mylar between them?

only use as many plates as you needed for whatever capacitance you wanted?

High-Voltage Capacitors, make your own! HV

[OrionLightShip]

You could but if the breakdown voltage is the same as polyethylene, say 20 kv/mm then you would need 7.5 mm thick mylar if you include a 50% recommended safety factor. That's almost a 1/4 in thick!

Also, he wants to make it variable. I can't even think about how to make a big chunk of metal charged to 100,000 volts variable without turning off the power and connecting both sides of that monster to ground while I do it.

why not buy super thin single sided pc board material, polish the copper and use mylar between them?

only use as many plates as you needed for whatever capacitance you wanted?

High-Voltage Capacitors, make your own! HV

[Kokomoj0]

You could but if the breakdown voltage is the same as polyethylene, say 20 kv/mm then you would need 7.5 mm thick mylar if you include a 50% recommended safety factor. That's almost a 1/4 in thick!

Also, he wants to make it variable. I can't even think about how to make a big chunk of metal charged to 100,000 volts variable without turning off the power and connecting both sides of that monster to ground while I do it.

yeh plus arc around space

would need to have like 2 slider sections on a rail

not sure why anyone would want such a thing but mylar seems like the best material for the job?

long story short this thing would be huge

[Farmhand]

I think Tomata has a good idea, just buy a HV variable capacitor and immerse it
in oil vertically so that an insulated extension can attached to the shaft which
extends through the top/lid. the shaft extension could slide through a tube
already inserted through the top so that a slightly raised tube with the extension
passing through it would form a leak resistant arrangement without needing to
be sealed, which would also serve to release pressure.

Good thinking Tomata, I can make a drawing to better explain what I mean but
that's the general gist of what I think could work, and be practical as well as accurate.

Cheers

P.S. The variable capacitor could be mounted (screwed or otherwise) to a
plastic frame so that it is positioned in the container without movement and
the lid just put on over the shaft extension to cover the top, some kind of
plastic container with a sealed lid should work fine, that way the whole frame
and capacitor could be inserted into the container then the oil added.

[Farmhand]

I think I would try it kinda like this.



Uploaded with ImageShack.us

..