Most efficient High Voltage method / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Most efficient High Voltage method

1 2 next

Move Thread

LAN_403

Turkey9

Tue Jan 26 2010, 01:26AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

So here's the deal: I need to go from 120 to around 10 kV DC. I don't need it to be very high power, preferably below 10 watts. And I want it to be as efficient as possible. It doesn't even have to be a constant source. A pulse of DC of at least 20 mS is needed however.

I'm thinking that a multiplier circuit driven by a small inverter would be pretty good but I have no ideas about efficiency. Maybe just a simple flyback like design would work. And there's no issues about making my own transformer.

The input voltage and power is pretty flexible but I'd like to keep it below mains rectified and smoothed (120 were I live).

Thanks!

dmg

Tue Jan 26 2010, 05:06AM

Registered Member #2628 Joined: Fri Jan 15 2010, 12:23AM
Location:
Posts: 627

here's an idea: you could have a high frequency push pull driver (like ones in disposable cameras), then either use a prewound flyback, or wind yourself a ferrite transformer with a high voltage secondary (about 1-2 KV output) and then you can feed that output to a half wave cockroft walton multiplier to get that DC pulses, using this method I think you can even go bellow a 120V input,
properlly designed, you can step up faily low VDC input to the 10KV you need,
may I ask what do you need this supply for?

Turkey9

Tue Jan 26 2010, 05:22AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

It's for ionizing the fuel of a propulsion system that I'm building. It has to be pretty close to DC, but needs only supply a DC current for at least 20 mS. Constant DC is fine but I know that multipliers need insanely large capacitors to get constant DC at any current worth mentioning.

dmg

Tue Jan 26 2010, 06:21AM

Registered Member #2628 Joined: Fri Jan 15 2010, 12:23AM
Location:
Posts: 627

In that case, I think you should build a flyback driver for an AC flyback then put a fullwave rectifier (bridge) on the output of the flyback transformer, the newer ones have an inbuild half wave rectifier so If you got an AC flyback its easier to smoothen the output, this setup however may eat more then 10 watts,
pehaps a smoothening capacitor if you have one rated for the 10KV at a decent capacitance
as for the driver, you can build a low power ZVS or a transistor based driver.

another method is too use some kind of a vacuum rectifier like a thyratron, although I dont know if there are any that can handle 10KV.

Dr. Dark Current

Tue Jan 26 2010, 09:40AM

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

For this I'd go with the IR2153 (or similar) half-bridge driver chip supplied from a little mains voltage transformer (12-15Vac 0.5VA) and a 78L12 regulator. This chip draws almost no current by itself, but of course it needs a little power to drive the gates.
This would drive a half-bridge of IRF630/640 MOSFETs and a diode-split flyback transformer (can be small) with a lot of turns on the primary winding to keep core losses down and still get 10kV.
I think a >=90% efficiency should be achievable

Turkey9

Wed Jan 27 2010, 04:38AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

Do you think that I should supply the bridge with 12 v? I have a 12 vDC supply for other parts of my project but I imagine that the bridge would have to be isolated from the rest of the circuit. Am I correct in thinking that?

As for the flyback, I have a couple small ones and they are internally rectified, but only by a single series diode on the output (discovered by dissecting a few). This isn't a problem, is it? It seems like I heard the diode split ones use some sort of multiplication scheme inside them.

And when you talk about a lot of turns on the primary, are you talking about something like 30-40? Or more than 50?

EDIT* Wait I see from the data sheet that the normal application uses the chip supply as the bridge supply too. Wow, I love that little chip!

Dr. Dark Current

Wed Jan 27 2010, 08:05AM

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

I meant to run the bridge off-line (rectified mains) and supply the chip from a little transformer. A dropper resistor from main DC supply can be also used, but it might waste a watt or more and drop your efficiency.

If run from 170V DC, Id try something like 40 turns at first and then remove turns if the power is not enough, frequency ca. 40kHz

Turkey9

Wed Jan 27 2010, 08:41PM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

Ok thanks.

What size capacitor do you think that I would need to get at least a semi filtered DC output?

Dr. Dark Current

Wed Jan 27 2010, 08:55PM

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

I guess two ~100-220uF 100V caps in series (center tap for transformer primary) should be sufficient for 10 watts...

Turkey9

Thu Jan 28 2010, 12:05AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

i meant for the high voltage side.

1 2 next

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