E-field Hotspots In Transformer Windings / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

E-field Hotspots In Transformer Windings

1 2 next

Move Thread

LAN_403

jpsmith123

Wed Nov 05 2008, 01:50PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Well I think I've convinced myself that my original idea of using teflon coated wire for a high voltage secondary winding (in order to hopefully obviate the need for encapsulation), probably won't work very well. It seems that unless I run the core at impractically low volts/turn, I'll end up with efield hot spots that exceed 30kv/cm, thereby causing corona. The problem areas seem to be the wedge-shaped air spaces adjacent to the contact point between the teflon jacket and the (in this simulated case) polypropylene layer insulation.

So I guess I just have to order some magnet wire and figure out how to encapsulate the winding.

Hazmatt_(The Underdog)

Thu Nov 06 2008, 09:49AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

Are you sure your volts/turn and volts/adjacent layer are accurate?

Dr. Dark Current

Thu Nov 06 2008, 12:41PM

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

Hazmatt_(The Underdog) wrote ...

Are you sure your volts/turn and volts/adjacent layer are accurate?

I was wondering the same thing. What is your winding scheme? standard is left-to-right, next layer right-to-left etc. In this case there would be a very low potential difference on one side of each interlayer and high on the other side.

Steve Conner

Thu Nov 06 2008, 02:02PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Now you know why flybacks are all impregnated in epoxy. Steve Ward made some nice single-layer ferrite transformers that work fine, though: he used Tesla coil secondaries cut up and threaded onto a giant ferrite core.

Being limited to one layer may mean that you need a bigger core, or fewer turns, or thinner wire, than you'd like, but it works without impregnation.

jpsmith123

Thu Nov 06 2008, 02:25PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Here is a simplified case of two 26 gauge teflon insulated wires, separated by 0.010 inch of polypropylene "layer insulation". In this case the simulated potential difference between the two wires was 2 kv. The brownish-reddish areas are where the field is greater than 30 kv/cm.

You can see that there is a small area within the wire where the field is high, but this is inside the teflon, so it shouldn't be an issue. It looks to me like the air space near where the wire touches the layer insulation is a problem area...somewhere in that vicinity the field has a peak of 46.5 kv/cm supposedly.

It seems that in order to get the field below 30 kv/cm with some safety margin, I'll have to run the core at no more than 20v peak per turn, which almost defeats the purpose of using a large core in the first place.

Hazmatt_(The Underdog)

Thu Nov 06 2008, 07:51PM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

What's the application of this transformer though?

Unless you're dealing with extreme amounts of power you're not going to see 1kv/turn.

Have you also considered your source impedance? Sometimes your design requires a load to reduce a potential stress like your dielectric problem here. What I'm saying is that once your source is loaded, the potential drops from open circuit value to a loaded value.

baleworker

Thu Nov 06 2008, 08:19PM

Registered Member #1596 Joined: Fri Jul 18 2008, 08:43PM
Location: USA
Posts: 65

Alright, here is a stupid question:
What software do you guys use for simulation?

jpsmith123

Thu Nov 06 2008, 09:29PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

I think I remember Steve Ward stating that he had some kind of problem with a single layer coil, too, didn't he? After a while it flashed over or something?

At this point I'd like to find a custom coil winding shop that will do a small run of HV coils for a U100/57/25 (double "U") core. Being that a double U100/57/25 core can be bought for about $26, maybe a group buy of HV secondary coils would be of interest? Maybe in quantities of 10 or so, an epoxy encapsulated coil with, say, 600 to 800 turns of 28 gauge wire, or something like that, could be had for less than $50.

Anyone know of a good custom coil winding shop?

Edit: That S/W is CST EM Studio (and I don't like the plots it does).

PanosB_GR

Fri Nov 07 2008, 05:05PM

Registered Member #1680 Joined: Fri Sept 05 2008, 04:19PM
Location: Greece
Posts: 43

This effect is called the triple joint, when air (e_r=1), metal and dielectric (e_r>1) contact together. If the angle of the air side is less than 90 degrees, a mathematical solution of the field gives infinitively high values to that little corner of airspace. As they say, this gives rise to surface flashovers for adequately high voltages (i guess near the limits of air strength).
Stupid idea: If you put the coated wire into plain distilled water and simulate it, this effect will probably vanish!

Hazmatt_(The Underdog) wrote ...

Unless you're dealing with extreme amounts of power you're not going to see 1kv/turn.

pulsed power transformers suffer such high values of electric field between turns. i mean, coin shrinkers do put LOTS of kV/kAmps from discharge capacitors into a tiny coil and it totally explodes :) (due to the magnetic force though)

Dr. Dark Current

Fri Nov 07 2008, 05:49PM

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

Hazmatt_(The Underdog) wrote ...

Unless you're dealing with extreme amounts of power you're not going to see 1kv/turn.

Not 1kv/turn, but 1kv/layer, and a layer is indeed a lot of turns

If you have 50 turns per layer and use 20V/turn, you get 2kV between layers.

1 2 next

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