Homemade HV transformer. / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Homemade HV transformer.

first previous 4 5 6 7 next

Move Thread

LAN_403

uzzors2k

Sat Dec 20 2008, 05:21PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Not bad Berni. You should really try submerging the transformer in oil, it's not as bad as it first sounds and really takes the headaches out of insulating. All you have to worry about is sealing your container to prevent leaking, which is much easier than insulating HF HV.

Berni

Sun Dec 21 2008, 10:25AM

Registered Member #1132 Joined: Mon Nov 19 2007, 06:15PM
Location: Slovenia
Posts: 84

Well im using a small PVC drain pipe as the coil former. Meaby the big fat garden hose will fit on the core. But i don't relay want perfect coupling i want a bit of leakage to limit the current so my driver wont catastrophically fail at high powers.

But it seams more like it arced over to the core first and then drawn out to between the two secondary.Its hard to see relay since it all h happens in a bright flash,fire and smoke So i need to make better formers that insulate from the core better or do the messy dunking in oil.I did a quick fix for now, i just unwided the damaged part of the secondary.

But i learned til now that its quite easy to make a reliable HV transformer without any oil or potting, but when you push in too much power it fails quickly.Meaby i will manege to solve the secondary to core acing problem by making better formers.

EDIT: Oh and am i supposed to do something with the secondary "center tap" like ground it? currently im just leaving it float. I guess if one side of the transformer sees a grounded or highly capacitive object it could make the center tap go HV and the other half of the secondary see twice the voltage and arc over to core

Voltwad

Sun Dec 21 2008, 06:21PM

Registered Member #1829 Joined: Sun Nov 30 2008, 01:06AM
Location: Raleigh N.C.
Posts: 74

Hi all, I'm looking to graduate from camera transformers to something bigger. I got a Viewsonic monitor from the free secion of Craigslist and pulled the FB out.

The core window is about 1.5"x0.5" I started with four layers of 30 turns each 22 AWG and on top of that about six turns of some thick wire.

Driving this with the sawtooth function of my little genny through a 3055 with a 5v power source gave about 3Kv peak to peak out. Better than straight turns ratio but still a little disappointing. Also the 3055 got hot. Like really hot. So I sliced open the FB to see how the pros do it.

First I learned, and I think several other sources here back me up, that the primary has to be on the inside? or wound on the other leg. I'd rather keep everything on one leg however. So, for a ZVS driven design, is the small ~5 number of turns/high amperage necessary? Even under oil as I am planning It seems like this would cause some heating issues. Next, I suppose I'll get some overhead transparency (polypropylene?) for the layer insulation as I think this stuff has good thermal and high frequency characteristics, however, the insulation in this FB seems to be PE. It's soft like PE and has about the same color. It also looks like it's been cooked or the wire coating has discolored it, sorry the photo has no more detail. So does anyone know what the manufacturers typically use? Finally, I think I'll switch to 30 AWG for the secondary. I think that'll have the right balance of durability and size. So, in conclusion, any words of warning or advice will be appreciated. Thanks guys.

Edit:
upon further inspection (hacking at it with a boxcutter) the inter layer insulation may in fact be polypropylene.

Dr. Dark Current

Sun Dec 21 2008, 06:44PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Voltwad wrote ...

I'd rather keep everything on one leg however.

You do NOT want to do that if you want to draw arcs. The reason is that with the windings on the same leg, the pri-sec. inductance will be very low and this means it will draw huge currents from your driver (and probably destroy it) if you don't use external inductor in series with the primary. Another problem is that this configuration often causes heating issues because the windings are close together.
It's much better to wind the primary and secondary on separate core "legs" if you want to have the current limited to play with arcs.

Voltwad

Sun Dec 21 2008, 07:17PM

Registered Member #1829 Joined: Sun Nov 30 2008, 01:06AM
Location: Raleigh N.C.
Posts: 74

Well now the Mazilli circuit includes a ballasting inductor in series with the two primaries yes? I was thinking I could engineer a nice variable inductor so I could build a versatile supply capable of a range of voltage/current combos. I guess I could put some spacers between the pri and sec to allow some oil circulation. But if these solutions are just too impractical or ineffective I guess I can use the other leg. It'll just be a pain getting the clip off if I need to take it apart again.

Dr. Dark Current

Sun Dec 21 2008, 07:26PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Voltwad wrote ...

Well now the Mazilli circuit includes a ballasting inductor in series with the two primaries yes?

No. That's just a current stabilising inductor, it does not provide any ballasting.

Voltwad

Sun Dec 21 2008, 07:39PM

Registered Member #1829 Joined: Sun Nov 30 2008, 01:06AM
Location: Raleigh N.C.
Posts: 74

Ah, so I'd need some ballasting between the MOSFETs and the pri's I guess. Hmm. Also I just looked around to see if I could find a ferrite c-core designed to have windings on both legs and came up short. Anyone know of a good source for such a thing. More also, I'm also wondering if casting my own core with powder and resin, with a distributed gap, would be a worthwhile endeavor. Anyone ever tried this?

Dr. Dark Current

Sun Dec 21 2008, 07:53PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Voltwad wrote ...

Ah, so I'd need some ballasting between the MOSFETs and the pri's I guess.

You would actually need a dual-winding choke, not worth the effort IMO. Wind pri and sec opposite on the core and no need to worry about ballasting (well.. to some extent).

Voltwad wrote ...

Hmm. Also I just looked around to see if I could find a ferrite c-core designed to have windings on both legs and came up short.

flyback core?
I believe I also saw some nice cores at

Edit:

Voltwad wrote ...

More also, I'm also wondering if casting my own core with powder and resin, with a distributed gap, would be a worthwhile endeavor. Anyone ever tried this?

I think there was a discussion on this and the conclusion was that the permeability of such core would be way too low to be useful...

rp181

Sun Dec 21 2008, 08:59PM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

ferroxcube cores are nice. I have the U126/91/20 set. Have to make your own bobbins though.

Berni

Sun Dec 21 2008, 09:02PM

Registered Member #1132 Joined: Mon Nov 19 2007, 06:15PM
Location: Slovenia
Posts: 84

Also for a start not layering the coils is better, its easier to insulate without oil.

By tweaking the primary turns i was able to get the arcs to start at 1 to 1,5cm so that suggests about 10 to 15kV. That is pretty impressive for this simple homemade transformer.

I need to make a blasting inductor for it and run this thing from 230V mains.It will aether make spectacular arcs or spectacularly fail.

Oh and the core heats up a awful lot at higher frequencys .Is there a way to reduce the heating?

first previous 4 5 6 7 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob