Volts per turn question. / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Volts per turn question.

1 2 3 4 next

Move Thread

LAN_403

Ash Small

Sat Dec 03 2011, 04:48PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I've been toying with some ideas for my proposed HV power supply.

It occured to me that if I feed 2 kV into the primary of a H-bridge-driven ferrite cored transformer, I'd only need ~25:1 turns ratio to output 50 kV.

However, I've read that there is a 'minimum volts per turn' limit with these things.

How do I calculate it, etc? any other relevant advice?

Steve Conner

Sat Dec 03 2011, 05:21PM

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

600V is the highest drive voltage that most people will dare to use. 320 or 400V are more usual as they're easily derived from 230V AC mains.

The limit is maximum volts per turn, not minimum, and it depends on frequency. The higher the frequency, the more volts per turn you can have, up to some limit. The ferrite core datasheet gives it as a graph of maximum flux density vs. frequency, and you can get from there to volts per turn at a given frequency using the transformer equation. The one I can never remember except that it has 4.44 in it.

If you don't have a datasheet for your core, you can just put a test primary on it and try different voltages and frequencies to see how many volts per turn it can stand without getting excessively hot.

The TV flyback is a convenient device for generating high voltage with a low turns ratio. The flyback action generates a high voltage on the primary, and a correspondingly higher voltage on the secondary, even with a low supply voltage. Most TVs used about a 150V supply to the line output stage, but a 1200V transistor was needed to stand the flyback spike.

For power supplies above a few hundred watts, the halfbridge or H bridge is used, and the turns ratio is made reasonable by putting a voltage multiplier on the output of the transformer.

Ash Small

Sat Dec 03 2011, 05:34PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

These are the spec sheets for the cores, Steve, Are any of the figures on there relevant to calculating maximum volts per turn @ 20~25 kHz?

Yes, I need at least 50kV to feed into a multiplier

EDIT: Link to N27 data sheet here:

Steve Conner

Sat Dec 03 2011, 06:26PM

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

The N27 datasheet will give you the flux density vs. frequency, and that's what you need to know along with the cross-sectional area from the core datasheet. Maybe someone else can help you with the math on a Saturday night.

Why do you need 50kV to drive your multiplier? It might make more sense to use a lower driving voltage and more stages of multiplication. For instance, a commercial 100kV DC supply from Glassman or Brandenburg might use 5kV of high frequency drive into a 20 stage stack, or 10kV into a 10 stage one.

If you want to achieve some really big output voltage like 250kV or 1MV, that might not even be realistic.

Ash Small

Sat Dec 03 2011, 06:35PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I'll have a look at the figures over the weekend, Steve. I had a quick look, but it'll need some studying.

I have another thread here somewhere about a multiplier using VMI(?) diodes from RS, and mounting the multiplier inside the vacuum chamber, with capacitor plates, to get up to ~650 kV (~15 stages with 50kV input). I'll try and find the link for you.

Steve Conner

Sat Dec 03 2011, 06:43PM

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

If you have your heart set on 50kV input, some sort of Tesla coil-like air cored transformer might make more sense than a ferrite one. It is really hard to design insulation to stand 50kV of high frequency AC, and also hard to get the stray capacitance of the 50kV winding low enough for efficient operation (or even to make it work at all)

You could put it inside the vacuum chamber too and have a vacuum cored transformer!

Marko

Sat Dec 03 2011, 06:45PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Hello,

50kV AC from a ferrite transformer will be rather difficult to obtain unless you intend to drive the transformer at resonance (in which case you might be better off with a tesla coil). If you absolutely need this I would try constructing a number of smaller pancake secondaries and connecting them in series to keep parasitic capacitance as low as possible. Or, rely on resonance and include some sort of protection circuit that will stop your transformer from frying itself when unloaded.

If you're OK with DC, standard approach would be to use split-diode rectifiers across your transformer windings or generate a more reasonable voltage (say 10kV) and multiply it.

Marko

Ash Small

Sat Dec 03 2011, 07:21PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Dr. Pork wrote ...

Hello,

50kV AC from a ferrite transformer will be rather difficult to obtain unless you intend to drive the transformer at resonance (in which case you might be better off with a tesla coil). If you absolutely need this I would try constructing a number of smaller pancake secondaries and connecting them in series to keep parasitic capacitance as low as possible. Or, rely on resonance and include some sort of protection circuit that will stop your transformer from frying itself when unloaded.

If you're OK with DC, standard approach would be to use split-diode rectifiers across your transformer windings or generate a more reasonable voltage (say 10kV) and multiply it.

Marko

I'm planning on using several pancake secondaries in series to get the 50kV.

I need to feed the 50kV into a multiplier, so I need a 50kV AC supply.

I'll do the maths for this before considering an air core (or vacuum core), loosely coupled, transformer, so that I have some sort of a comparison.

Ash Small

Sun Dec 04 2011, 06:45PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Steve McConner wrote ...

The N27 datasheet will give you the flux density vs. frequency, and that's what you need to know along with the cross-sectional area from the core datasheet.

This graph gives Complex Permeability over Frequency.

Does it tell us anything useful?

And this formula has a four in it, and looks vaguely relevant:

If I put some numbers in, I get Npri = 2000V x 10^8 / (4 x 25000Hz x Bmax x 840mm^2)

which is Npri = 240 / Bmax

(assuming my maths is correct)

Steve Conner

Sun Dec 04 2011, 08:08PM

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

Sure, that would be the right formula if only you knew Bmax.

(It's 4.44 for a sine wave, 4 for a square wave.)

Maybe Epcos specify it differently. I've seen datasheets where they gave the core loss in watts per cubic centimetre as a function of Bmax and frequency. It's left to the designer to decide how many W/cc constitutes "too hot", as it depends on the copper losses and cooling method, not to mention the surface-area-to-volume ratio of the particular core.

The complex permeability graph gives you an idea of core losses vs frequency but it doesn't tell the whole story. I don't know how to derive a core loss figure from it, anyway.

I've always done this by applying HF magnetic flux to the core and feeling it to see how hot it got.

1 2 3 4 next

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