Instructables

Flyback transformers and how to convert them to AC

I'm in the need of quite a few electronics for my new lightning tube project, on of them is a suitable HV transformer for high frequencies.
Salvaging from plasma balls or similar is pointless as those transformers only get to around 3kV at best.
With BW TV set getting really rare these days I decided to look into the modern version of the flyback transformer - the DST, Diode Split Transformer.
If you ever did some Tesla stuff you will know that everyone says a modern flyback is no good as it only provides DC output.
With basically no info on the net how these things look internally it was down to destorying a few until I got to know what to look out for andhow to remove the HV diodes between the secondary coil packs.

I know not too many people will need a HV AC tranformer that can operate with several kHz so I though I ask first if it is worth to create an Instrucatable on this topic alone, otherwise I will just include a very short version of it with my Instrucable on the lighnig tube project.
Have to say that is a quite messy and stinky procedure involving a lot of heat and the risk of burns to your fingers.

-max-1 month ago

Do not bother with trying this, as the transformer is specifically designed for DC and by converting to AC, capacitive coupling between layers of wires will cause a dielectric breakdown. Even if the dielectric can withstand a lower AC voltage, the transformer will not be as efficient because the capacitive reactance to ground acts like a short-circuit at really high Dv/Dt's (spontaneous changes in voltage over time). Also it would be just as easy as making a new transformer altogether.

Instead, I recommend either finding a very old black and white new-old stock AC flyback or if you are super cheap and have plenty of 35 AWG wire laying around, you can wind your own! I did this and had some success. You should be able to achieve 3KV very easily.

Start be making a paper sleeve on a highlighter near the thickest part, so that way you can slide it down to where is is not as thick. Then carefully wind the wire leaving plenty of it out for your ground. Make sure to only wind about 1cm wide, and leave at least 0.5cm of clearance from the edge. After finishing, cut some electrical tape to 0.5cm width and place it where where the clearance was made. Then wrap at least 2 layers of the tape around the wire and everything, keeping the new surface flat and flush. Continue the same winding process, and adding tape the outside of the coils until you get tired and have something that is reasonably thick! You will get several hundred turns this way. As you build up layers it will get more difficult to add wire as the shape begins to bulge and round out from the center. Once you feel you have done enough, soak the whole coil in molten wax to seal off every part that may cause a dielectric failure. I used a ZVS driver on mode and can struck a very hot arc at 2-3mm, and stretch it to over 1cm with 12V!

-max- -max-1 month ago

If you want a reliable 30kv high frequency transformer, you will just need to find a very ancient flyback or line output transformer, or use a small solid state tesla coil.

You may also be able to find high power RF transformers in high power RF transmitters. I don't think they can produce very high voltages.

Downunder35m (author)  -max-1 month ago

Sadly I had to come to the same conclusion :(
I was able to reove the diodes and to feed it some nicely chopped DC for a test.

Result was not what I expected as compared to an unmodified transformer the output was lower.

Using with AC gave mixed results depending on the frequencies I used.

As I wanted to reach something above the 40kV range this was totally uselss, longest spark I could get was about 2.5cm long, while an unmodified transformer on the same driver and frequency (22kHz) produces 6cm sparks.

I now have 2 hand wound coils connected by a spark gap and capacitor.

First coil charges the cap with about 120kHz DC pulses, when the cap is saturated it discharges over the spark gap and into the second coil.

But for some weir reason the output frequency won't go over about 70Hz.

Seems I have created a stun gun instead of a working HV transformer.

I might have to re-think the whole concept and find a suitable gas mix and pressure for the "lightning tube" first nstead of using a vacuum and extreme high voltages.

There is simply not enough thin magnet wire left here and after weeks of searching I gave up BW TV sets :(

Qcks1 month ago

I'd watch such an instructable simply because induction, as a general electrical phenomena, interests me greatly, yet it doesn't get allot of love.

With regards to the Diode split transformer, are you referring to the Cockcroft-Walton generator?
It's interesting, but I'd question if it's really meant to be used as a transformer at all. There's some interest in using them in High voltage transmission of electricity, but cryogenics and high temperature superconductors need to be better researched. I've seen them used in stun guns, and I know they're used in the Negative ion generators, but just not really as a real transformer for a whole lot of stuff.

Have you considered salvaging a transformer from a microwave? there's several tutorials on how to do it on youtube, and old microwaves are cheap. You might even be able to find one for free, if you're not adverse to digging through dumpsters.

Downunder35m (author)  Qcks1 month ago

I am trying to get a workable high voltage supply that is not limited to the 50/60Hz duty cycle from the mains.

So a microwave tranformer is out of question, both for the frequencies and the low voltage of just around 2000V.
My minimum aim would be 30.000V with a frequency range between 10 and 40kHz.

Ah...
So thinking through this a bit more, don't use a core.
Aircore inductors are preferred in certain applications where eddy currents can be a problem. If you're getting lots of heat and melting, having one winding directly on avoids the heat buildup in the inductive core.
Have the inner, less tall loop be thicker gauge wire with fewer number of turns, and the outer diameter be thinner gauge with higher numbers of turns, and that should step up your voltage.

If you really want to try a core material go with something like aluminum, or chrome (like actually chrome, not stainless steel). You want something that's only paramagnetic, rather then actually ferromagnetic. If you went with aluminum, you could try cooling it as well with mineral oil, or something like that, but at really high frequencies, i think aircore is the way to go.

Downunder35m (author)  Qcks1 month ago

Interesting though with the air core.

Usually I use them only for tesla coil projects.

Did a quick test with the coil of a small fan motor (donated from a microwave) for the high voltage side and about 30 turns of wire for the primary side.

Running it with my small flyback driver produced about 390V and the amps would barely register on my meter set to 20mA - 25kHz frequenzy.
Adding an aluminim core did not really change this although it was a 1x1.5cm hole and I only had 10mm aluminum rod :(
Adding a core in the form of the cut off piece from the motor resulted in total failure due to the high frequency.

Dropping a piece of ferrite in it resulted in close to 500V and almost 10mA.

So I really think using a flyback is the best option to go unless you want to wind a few thausand turns of hair thin wire onto a more suitable core.