Build Your Own HV Capacitors | Hackaday

2022-06-09 06:55:44 By : Ms. Anna zhu

Finding high voltage capacitors can be tricky. Sure, you can buy these capacitors, but they are often expensive and hard to find exactly what you want. [RachelAnne] needed some low-value variable capacitors that would work at 100 kV. So she made some.

Instead of fabricating the plates directly, these capacitors use laminations from a scrap power transformer. These usually have two types of plates, one of which looks like a letter “E” and the other just like a straight bar. For dielectric, the capacitors use common transparency film.

As you might expect, she had to strip off the insulating varnish where the contact was to be made with the plates. The moving mechanism uses a toy car tire as a handle to make sure it is insulated. The screw moves one set of plates in and out while the other side remains fixed.

For more capacity, it would help to use a thinner dielectric. The transparency film is about 100 microns thick and there are several sheets in between each plate. According to the post, there are 400 microns of dielectric and 600 microns air gap, so that’s a millimeter gap between each plate. Of course, the voltage capacity depends on the gap and the dielectric constant, so too thin and you might need a better dielectric to handle the high voltage.

We’ve certainly seen homemade high voltage caps before. You can also make your own supercapacitors, but they probably aren’t going to take 100 kV.

Back in early 2000’s in Poland we made HV capacitors by taking two rolls of garden foil (the kind that is used for simple greenhouses) and roll of aluminum foil. One unwinds the garden foil roll, covers it with aluminum foil leaving 5cm margins on the sides and ends, adds some wires on one side as output. Then next roll of garden foil is unrolled over it, and another aluminum foil plate on top of that with wires going out on the other side. Then the entire thing is rolled and tied with some zip ties. And pushed into PVC pipe…These were good for 10-30kV, more with thicker foil, and had capacitance of 10-50nF, depending on size…

What’s garden foil made from? I’ve never heard of it.

I’m from Poland, I don’t know its proper english name. It’s a thick, slightly milky foil, usually made of PE. It’s used for making greenhouses instead of glass panes. It can be also used as temporary roof cover…

My first thought was that this sounds like Mylar film, but some googling found that it’s made from PET instead of PE. Both seem like they’d be good insulators though. I did turn up some greenhouse polyethylene film from farm supply shops as well – just one of those things I don’t encounter as a city dweller.

We did it with roofing plastic. Cheap and available in large rolls. Sometimes the have tiny dips in thickness, so for higher voltages we use to use 3 layers, and use the rating for 2. Raid 3. We made a large capacity to test lightning protection circuitry, charging using a neon sign transformer and some high voltage diosed. It sounded like a 9mm handgun when discharged.

This is a great project! Thanks for sharing.

I’ve made high voltage fixed caps in the past with window pane glass and aluminum foil but I have not had a need for a HV variable cap – yet. I’ve used a variable inductor instead of a variable HV capacitor in the past. But now that you’ve shown me the way I might try a variable cap next time.

Charge the cap at maximum capacitance. Turn the knob to minimum capacitance. E = 1/2 C * V^2 (from the charge) is preserved, as you reduce the capacitance, the voltage goes up…

And you just made a mechanical voltage doubler.

It can multiply waaay higher than that. And it’s linear.

Is this a real thing?

See capacitor charge redistribution… It’s the trick used in switched-capacitor SAR ADC (e.g. in microcontroller)

If you want to learn about the physics, here is a lecture (by Walter Lewin) that explains it. https://youtu.be/GAtAG938AQc

See also wimshurst machines It helps to think about it as: the oppositely-charged plates have an electrostatic attraction force, by moving them apart you’re doing work against that force, so converting mechanical energy into electrical. Electricity and mechanical motion are both work, and there’s nowhere else for the energy to go (energy that’s used to overcome the electrostatic attraction – there is of course frictional force on the handle as well), so it goes into the electric field

Might not be doubling as voltage is square root of capacitance ratio. i.e. you’ll need to reduce the capacitance to 1/4 to double the voltage.

I used doubler as a general term. But as you pointed out, it works!

Q=CV. If you keep the charge the same and vary the capacitance (say 10:1) your voltage goes up ten times. Unless something blows up first.

Nope, the energy of the capacitor is *not* preserved. If the terminals of the capacitors are isolated, the charge on the plates is preserved (Q = C * U, measured in Coulomb). So if you reduce the capacitance by a factor 2, the voltage will double and the energy stored in the capacitor will also double. This change in energy will be compensated by the mechanical energy required to move the plates away from each other (against the electrostatic attraction between the plates).

Yes you do work (input energy) moving the knob.

I’ve been making some HV caps as well.

Gorilla glass is pretty thin, and has a very high breakdown voltage. You can get sheets as screen protectors on eBay.

Mica has just about the highest breakdown voltage, and you can get sheets of this on eBay as well (it’s used inside microwave ovens).

Mouser has small thin mica sheets for transistor insulators that are cheap and very uniform, too. T220 size, and a few bigger ones.

There are actually a bunch of good material options, including bigger sheets, in the thermal interface category.

One guy here published tutorial about creating

English translation via Google here: https://danyk-cz.translate.goog/tckond.html?_x_tr_sch=http&_x_tr_sl=cs&_x_tr_tl=en&_x_tr_hl=cs&_x_tr_pto=nui

he used linoleum (floor covering, could be the PVC version) and aluminum foil used for food packing. I also remember he had even a bigger one in a plastic bucket…

FR4: 20kV/mm. The average 0.062″ (1.57mm) copper clad doubled sided PCB should in theory be good for 30kV With dielectric constant ~4, you’ll get 14.5pF/in^2

On the Ham Radio Workbench podcast one guest mentioned getting a CNC mill just to mill plates for an air-core variable capacitor because all the ones with the capacitance he needed on ebay were hundreds of dollars, being very much out of production. Not for kV but a similar concept

Note that this technique will likely not have good high frequency characteristics. Ferrous material have very poor high frequency inductance and resistance. So much so that there was a warning about using electroless-nickel plating on RF boards (RF boards are often built without solder mask because the variation of solder mask dielectric introduces too much variation. Obviously ENIG, and other variations, is acceptable if you don’t plate the entire board).

Coating the laminations with copper or another non-ferrous material will mitigate this if the coating can be made thicker than the skin effect depth.

Pressed the wrong key. Need an editor.

For HV power supplies and other low frequency applications this is perfectly acceptable and the high frequency rejection may actually be desirable.

Be careful! Using that rubber toy tire with the metal shaft running through it (exposed metal) does not seem to me to be safe insulation in the 100kV range quoted in the article.

If you need a lot of thin metal plates, tin caps from your nearest home depot are cheap and about the same thickness as transformer laminations.

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.

By using our website and services, you expressly agree to the placement of our performance, functionality and advertising cookies. Learn more