The Hackery at KlugeWorks:

Variable-Output High Voltage Power Supply

Caution and Disclaimer: The voltages and currents delivered by this design are potentially lethal. You must use adequate safety precautions when you build, test, and use this device or any similar device.

How-to: Find an older microwave that still works. Newer ovens tend to use switching power supplies, which are probably viable for use at 2000-2200 V, but are somewhat unlikely to be adjustable; ovens that don’t work may have bad magnetrons or blown fuses (either which is fine, as we aren’t going to use the magnetron or the fuse here), or bad diodes (which is not so fine, as we need the diode). It isn’t always easy to tell from the outside what the problem is, so if you can start with a working oven you can avoid a certain amount of hassle.

Open up the oven and take it apart. Reserve anything that seems like it might be useful. I got 4 microswitches from the one in these photos, for example. Remove the transformer, the capacitor, and the HV diode...

                       

I actually paid about $12 for this oven, including tax, because they were having a holiday special and gave me 25% off. The transformer is at bottom center on the third photo, and the capacitor (which has the diode attached to it) is at top right. In the next photo, I show just the transformer, capacitor, and diode:

Note that some of the nuts are coated to keep them from coming loose from the vibration of the oven; you will probably require a nut-driver of the correct size in order to complete the disassembly.

Here is the schematic diagram that was printed on a piece of paper and glued inside the top of the box:

The section that is relevant to us starts at the transformer. Here is a revised schematic, showing what you will actually end up using:

The resistor, which should be a power resistor that can withstand high voltage, limits the amount of power that the supply can provide. I used 3,000 ohms, but that may be slightly low. (In certain applications, particularly those in which you are performing pulse-charging, you may want an inductor as well, after the resistor.) In some applications you can leave the capacitor across the supply for suppression of ripple. Connect one side to ground, the other between the diode and the resistor.

[Note that the wire colors I’ve indicated on the schematic are the colors of the wires that were plugged onto the transformer when it was in the oven. Most of the actual connections are just metal tabs except for the orange wires, which were used to drive the filament of the magnetron, and which are in fact orange wires.]

—> For the full rated voltage (in most cases actually the peak voltage, which is about 1.4X the rated voltage), you can plug the transformer directly into the line. Be sure you use appropriate grounding, and see the next paragraph.

Crucial considerations:

  1. You should put a fuse or circuit breaker into the primary circuit, and the fuse should probably be about 12 amperes. (The actual value depends on the power of the original microwave oven. The one I took apart for this supply had a 15-amp fuse in it.) Even if you run it on variable input voltage, it’s a good idea to have a fuse or circuit breaker. (In this world, safety lies partly in having as many backup systems as possible. We did not get the “perfection” option, and things fail, usually unexpectedly and at inopportune times.)
  2. If you decide to include the capacitor, you should check it very carefully to be sure it can handle at least twice the rated output voltage. (Remember, in ordinary service in the oven, it will never see more than the rated voltage of the transformer, and possibly not even that. This service is different.)

—> For variable voltage, you can plug the supply into a variable autotransformer. If you are very careful or if the oven transformer is rated for only 750 watts or so, you can probably use an 8-Ampere unit, but a 10-Ampere unit is less likely to get hot if you are operating at or near the full rated voltage. (My application requires at most half of the rated voltage, so I’m not going to worry about this issue.)

Crucial consideration: if you use a variable transformer, you should carefully avoid turning the output voltage up beyond 120 VAC. (Many units offer outputs as high as 140 VAC, and some are switchable.) The lifetime of the transformer will be shortened at higher output voltages, especially those that are higher than it is rated to handle. (If you want higher voltages, see below.)

Crucial design feature: you may wonder why anybody would use a 12,000 V diode in a 2,000 V power supply. Here is the reasoning:

  1. The peak voltage is 1.4X the rated voltage, which is generally 2,000 to 2,200 VAC for these transformers. For simplicity, let’s call it 1.5X. That’s as much as 3.3 kV.
  2. The transformer puts out AC, so when you have fully charged whatever device you are powering your diode will see 2X the peak voltage, up to about 6.6 kV.
  3. A further factor of 2 to provide a wee bit of safety margin is reasonable, so we use a diode rated for 12 or perhaps 15 kV, those being cheaply available.

—> If, for some reason, you need voltages higher than the oven transformer is rated for, you have several choices. One is to build a multiplier chain for this supply. (In fact, if you just leave the capacitor and diode in their original configuration, you already have one doubler stage.) Another is to use the ignition transformer from an old oil burner. Many of them are rated for 10 kV but have centertapped secondary windings, with the centertap grounded, which means you’ll need one or two HV rectifiers, and you will get up to 5 kV (~7.5 kV peak) from the supply. I have also seen single-ended units rated for 7.3 kV, which will require only a single rectifier, and will put out proportionally higher voltages. Remember to use rectifiers that are rated for a lot more voltage than you expect to get from your supply. 10X or 12X may be too expensive, but you should certainly get as close as you can. It is possible to put several HV rectifiers in series, but you should research this, because unless they are designed to stack it is not as simple as just putting one after the other. (If you just connect them up the weakest one will blow out, followed by the next-weakest, and so on until they are all trashed. I trust that this is not what you had in mind.)

In either case, oil-burner transformers are lower-power devices, typically rated for only a few milliamperes. Be sure you limit the output current so you avoid damaging the transformer in the event of a short circuit, though it is possible that a given transformer may be engineered to withstand a short circuit. My understanding is that microwave-oven transformers are, indeed, engineered this way, though of course you should never count on it.

Again, please be extremely careful if you build or use anything even remotely resembling this design.



Home

For Jon Singer’s personal homepage, use this link.



This work is supported by
the Joss Research Institute
312 Laurel Ave.
Laurel MD 20707-4320 USA



Email: a@b.com, where you can replace the b with joss, and the a with my first name (just jon, only 3 letters, no “h”.)

Phone: +1 301 SAPWOOD (That’s 7279663, if your phone doesn’t have any letters on it..)

Last modified: Thu Jan 12 13:57:22 EST 2005