DIY High Voltage transformer winding TROUBLE

[FonziDaytona]

Hi all,

I had a segmented high voltage transformer bobbin 3D printed. I’m having THE WORST time trying to actually wind this thing. The first few layers start off great, but then, somehow/someway, it all just goes to crap. I can’t keep each layer PERFECTLY flat so that the next layer stays flat.

Am I missing something?

Anyone here with some legit skills that I could pay to do this for me? I’ve spent 30 hours and this is as far as I’ve gotten. There’s no way I’m going to succeed in winding each layer flat.



-Matt

[Andrew321]

I made my own segmented bobbin, it worked just fine and I didn't pay any attention to how neatly the coil was wound. You'll have to use your imagination a bit here. My thinking was that if you have 10kv on the secondary and you have 10'000 turns, well that's 1V per a turn, so there is only a 1v difference between turns. Most insulation is rated to about 300v. So if a layer has 5 turns on it, that is 5 volts. You double back as you coil and make a second layer. Now you have 10v total and the first coil is touching the second coil and there is a 10v difference between them. 10v is a lot less than 300v so your good to go In fact, each coil has its own 300v rated insulation so you should be able to get away with 600v between the two, but I don't recommend it. At high frequency the rating drops some and if you have any air bubble they will likely ionize and degrade the insulation. The numbers I used are just random, but the principle is the same. If you keep your volts per a turn low enough even if you sloppily wind it on and one coil ends up half way down, it doesn't matter because the insulation will do its job. If from top to bottom each segment can produce 300v then you should be ok no matter how you coil it. On mine I did a mock up with carboard and oil and tried it out. It worked so I build a bigger bobbin and spun it up in like 3 hours. I still was caution to not mix players too deeply, but it all worked out. Then again I fail 10 times before I succeed once hope this helps! best of luck!

[FonziDaytona]

Hi Andrew,

Thanks for the reply! Maybe I’ll dust this back off and not obsess as much…

-Matt

[FonziDaytona]

Figured I’d update this thread with progress report:

Wound the transformer without worrying about “neatness” and it all worked out fine (so far anyway).

I used the bobbin to create two separate windings that are center tapped and 180deg out of phase with each other. Each winding is made up of 5 sections with 90 turns. The primary is 14awg magnet wire, 18 turns. I put the whole thing under oil for insulation. I won’t go over 6kV peak per coil, so hoping this stands up!

I used a series capacitor to resonate with my leakage inductance and (running around 36khz) this thing really rocks! I consistently get a warning from my amp monitor that I’m exceeding 15Amps from the wall…

[Thor42]

In fact, each coil has its own 300v rated insulation so you should be able to get away with 600v between the two, but I don't recommend it. At high frequency the rating drops some and if you have any air bubble they will likely ionize and degrade the insulation.

The main reason is that the wire enamel insulation can have random scratches. You may be unlucky and these scratches will end up on the turns of adjacent layers. Thus, I'd say that voltage between 2 layers should not exceed 100 volts. Also, it would be nice to cover each layer with varnish IMO.

[Thor42]

BTW, that winding is not only about “neatness”. It's also about parasitic capacitance.
A bit late, I suppose, but you can add insulation after each layer (1-2 turns of Kapton tape or something). That helps the winding stay neat all the way.
Pros:
+ less parasitic capacitance
+ more durability
+ less chance that random enamel insulation scratches meet together.
Cons:
- more difficult and tedious to wind
- a bit fewer turns with the same coil size.