Very Very High Power Flyback Driver / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Very Very High Power Flyback Driver

previous 1 2 3 4 next

Move Thread

LAN_403

Antonio

Sat Nov 28 2009, 11:14PM

Registered Member #834 Joined: Tue Jun 12 2007, 10:57PM
Location: Brazil
Posts: 644

The simulation below shows the strange effect of reducing the inductance of the series inductor below a certain value. The circuit enters an irregular oscillatory mode draining a huge current:

The plot shows the input current. With 67 uH of inductance in the series inductor, the circuit works correctly. With 66 uH, it doesn't.
This is not a regular quasi-linear oscillator. There is no feedback when the diodes are not conducting. It is a strongly nonlinear oscillator with a structure quite different from classical structures (that I know, at least). It works correctly with two uncoupled inductors instead of the center-tapped primary coil. Two flybacks (AC type) could then be used with it, doubling the output, and in this case the series inductor is not necessary.

Colin 99

Sun Nov 29 2009, 02:03AM

Registered Member #192 Joined: Fri Feb 17 2006, 03:08AM
Location: Canada
Posts: 44

Even with a DVM it's hard to read the input current, mine gave erronous readings. I end up using an external current shunt and reading the mV output and then converting to amps.

With My zvs I could draw 5 inch long, thick arcs. My power supply was 36 volts DC off of 3 - 12volt batteries. I changed the resonant cap, added bypass caps to the supply ,changed the gate resistors and zeners all to maximize power output. I've measured 13 amps when I drew the arc out to it's full length. That works out to 468 watts max.

Shaun

Proud Mary

Sun Nov 29 2009, 03:23AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Aha! You have obviously discovered an electronic secret which can not be revealed by scientific measurement techniques alone. If we take a very ordinary 21kV LOPT, rated by its manufacturer such as HR Diemen at 90W, shouldn't we at least wonder how the tiny wires of the secondary could really support 1kw or more as has been claimed?

The process of power amplification for the >100W input by the <1kW output seems to me a mystical and magical thinking, where, to put the kindest complexion on experimental results, arc length is wilfully confused with the dielectic breakdown of air in order to inflate air.

But I'm not an HV or Tesla girl, so my observations could well be out of order here.

Stella

Dr. Dark Current

Sun Nov 29 2009, 09:28AM

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

Proud Mary wrote ...

Aha! You have obviously discovered an electronic secret which can not be revealed by scientific measurement techniques alone. If we take a very ordinary 21kV LOPT, rated by its manufacturer such as HR Diemen at 90W, shouldn't we at least wonder how the tiny wires of the secondary could really support 1kw or more as has been claimed?

In the same way as a 500VA transformer will run at 5000VA, for a limited time?
1kW seems a lot to me too, but obviously this depends on the construction of the transformers. Most diode split transformers of mine can take a few hundred watts. Some are overrated more, some less (for their real job in which they need to run tens of thousands hours without a failure).
Another factor is that these huge output powers come from running the transformer in forward mode, where the power is supplied when the diodes are conducting, not in flyback mode like in the TV where the power is extracted just from the energy stored in the core during the time the switching transistor is off. This allows for greater power throughput and less heating of the secondary winding, because it conducts for longer time.
Also we use just the secondary winding, not the original primary and auxiliary windings, so these do not contribute to the heating of the integrated transformer.

Steve Conner

Sun Nov 29 2009, 04:52PM

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

Antonio wrote ...

This is not a regular quasi-linear oscillator. There is no feedback when the diodes are not conducting. It is a strongly nonlinear oscillator with a structure quite different from classical structures

Yes, if you want to keep your sanity, just don't think too hard about how it works!

In my earlier discussion, I don't cover the feedback aspects of the circuit: it might as well be driven from an external fixed oscillator set to the LC resonance frequency.

This is deliberate because any attempt to explain how the feedback really works would just confuse everyone. OK, what I mean is that I don't really understand how it works.

But consider that one or other of the diodes is always conducting, and the circuit resembles the old two-transistor astable multivibrator.

Even if the two inductors were uncoupled magnetically, they'd still be coupled through the tank capacitor.

Dr. Dark Current

Sun Nov 29 2009, 05:20PM

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

Steve McConner wrote ...

This is deliberate because any attempt to explain how the feedback really works would just confuse everyone. OK, what I mean is that I don't really understand how it works.

The feedback is quite simple - when transistor 1 is conducting, it keeps transistor 2 off through the diode. The voltage on tr.2's drain is high enough so tr.1's gate is pulled up with the gate pullup resistor. When the voltage on tr.2 falls below tr.1's gate threshold voltage minus the diode drop, tr. 1 is turned off.
Well, kind of. During the phase the primary voltage is near zero, both transistors are actually in linear mode with the voltage drop on them set by their gate threshold voltages (again minus the diode drops). This is the moment I believe the circuit might enter the "parasitic oscillation" mode, where the transistors oscillate just with their delay times, only possibly damped by the gate resistors. This is probably why higher gate resistor values seem to decrease the probability of these oscillations.
When the circuit gets past this linear phase (the primary voltage has risen enough in the other direction), transistor 2 is now conducting and keeps transistor 1 off until the voltage falls back again.

EDIT: I wanted to add that maybe putting small capacitors across D-S of both transistors could decrease or eliminate the chance of the unwanted oscillations, but I have not tried this.

Firefox

Sun Nov 29 2009, 07:23PM

Registered Member #1389 Joined: Thu Mar 13 2008, 12:50AM
Location: Pittsburgh, PA
Posts: 346

Wow, I apologize for not putting out a basis for my 500-1kW guess for arc power. I am told that this is not the correct way of measuring real power, but for now it is the closest I can get with the tools I have (getting a kill-a-watt soon). My measured DC voltage on the bus caps was 155VDC and measuring the current before the input inductor with a 2m resistor showed a DC current of 11A (22mV).

Steve Conner

Sun Nov 29 2009, 08:04PM

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

The unspoken assumption behind the kill-a-watt (or DC bus volts * amps) method is that most of the power going into the circuit is coming out as arcs.

Given that a flyback is a physically small object that can't handle much heat without burning up, that's probably a reasonable assumption. If the flyback doesn't melt after a few minutes, and you measure 1kW going into it, then maybe 900W must be coming out.

If you have huge heatsinks on your driver transistors and they're still getting stinking hot, that's a different matter.

Firefox

Sun Nov 29 2009, 08:21PM

Registered Member #1389 Joined: Thu Mar 13 2008, 12:50AM
Location: Pittsburgh, PA
Posts: 346

I understand that, Steve. That's why my conservative guess was .5-1kW of arc power with measured 1750VA of measured power in. I can tell you the heat of the arc was more than enough to melt a couple of centimeters of my 14AWG ground point over the course of it's life. My little heatsink only gets lukewarm. Of course, at this point, it's all speculation, as the flyback ultimately did overheat after a 3 minute run with 155V in.

Colin 99

Sun Nov 29 2009, 09:27PM

Registered Member #192 Joined: Fri Feb 17 2006, 03:08AM
Location: Canada
Posts: 44

I have no doubt that your output power was 1.5 to 2 times greater than what I was getting out of my zvs flyback circuit. How the flyback can handle that much punishment is beyond me. I just hope that when I bring it over to someone's place to do a demonstration that it doesn't burst into flames. (I do worry about that!) :-0

previous 1 2 3 4 next

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