Limitations of the Marzzilli Flyback Driver / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Limitations of the Marzzilli Flyback Driver

previous 1 2 3 next

Move Thread

LAN_403

Marko

Mon May 15 2006, 12:21PM

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

Ah, my crap =)

I use inductors with lots of turns and small ferrite bead inside, so I could hardly make it go in saturation.

Wire was relatively long so it suffers more ohmic losses, but I didnt care much for that as flyback would melt much faster than the inductor.

I didn't know that for powdered iron, I tought the core won't fell the ripple from switching transistents too good.

cheers..

Dr. Shark

Mon May 15 2006, 01:17PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Steve Ward wrote ...

To further answer the original question, your voltage vs. current output is going to be strongly dependent on operating frequency. The secondary of the flyback doesnt "know" how many primary turns you have. All it "knows" is the frequency and flux intensity through the core. So, by changing your LC, you are changing up the frequency. Smaller C and bigger L means higher Z, so you need greater input voltage to drive it to the same power levels.

I see the surge impedance, i.e. the inductance/capacitance ratio in the tank circuit will influence the output impedance. Since the primary voltage is fixed at pi*V_in, the circulating current could probably be modualted in this way, i.e. a smaller capacitor would lead to less current. Correct?
Problem is, I cannot get the L below a certain limit, since the minimum number of turns is dictated by core saturation.

I am not so sure about the point with the operating frequency, since in the other thread you convinced me that a flyback transformer has an intrinsic secondary resonant frequency, and driving it close to that frequency would maximize the output. So when changing the surge impedance, I should probably try _not_ to vary the frequency, and keep the L*C product constant.

I am confused! But since my mains halfbridge killes my flybacks

I guess it is time to really understand how this driver works:

blackplasma wrote ...

Joe, I'd have to say you haven't played with the circuit enough.

It's just a matter of finding the right combination of resonant capacitor and series choke. Work with high input voltages and a reasonably small resonant capacitor, a reasonably large choke, and you'll have nice long thin purple arcs...

The great thing about the ZVS is the purple arcs aren't squealy and annoying, they're wispy and flappy instead. :P

Just watch out, as with too much Vin you'll find that after starting an arc the output voltage will promptly double, sending a VERY squealy purple arc down the Ultor, HV cable, across/down the case to the HV return pin, or sometimes it punches through the plastic exit if you're not using proper HV cable (like me).

So, try lots of Vin, perhaps 10+10 turns, and a largeish (1.5" OD, 0.75" ID, 0.5" thick) powdered iron power supply choke, a small resonant capacitor, and you'll have beautiful stable long purple arcs.

Just try! =)

back to the workshop for me!

Oh, just how much is "lots of Vin"?

Marko

Mon May 15 2006, 01:42PM

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

Another option for a high-output voltage driver for with lowest power input would be something like my flyback converter.

I threw a schematic somewhere on the site.

Basically I used a variable duty cycle multivibrator, push-pull, mosfet, a 10:1 base drive transformer and PNP hifi transistor 2SB863, I got some of them dirt cheap (now sold out)

It preformed a lot better than 2N3055 and other crappy transistors.
I guess transistors made originally for lyback operation would also work fine.

Frequency was pretty high, I think about 200kHz (I could go even more but it wasn't needed), duty cycle kept low, with best sparks at about 50%.

From small 28V power supply I got some nice arcs starting at 5cm, with about 2-3mA of current.

I don'gt know the input power because my cheap meters are obivously lying for 2,5A of current, I think I could no way get 70 watts from a small, about 30W transformer.

I didn't try to power this driver from a bigger transformer yet.

Dr. Dark Current

Mon May 15 2006, 03:37PM

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

joe wrote ...

I am confused! But since my mains halfbridge killes my flybacks

flyback cancer

wrote ...

Frequency was pretty high, I think about 200kHz (I could go even more but it wasn't needed), duty cycle kept low, with best sparks at about 50%.

50% is not low, my flyback made the best power/output voltage ratio with about 10% (10W vs. 30kV), maybe it is not too high with so high frequencies... At 200khz you are going to cook the core/transistor/secondary soon, maybe you can try to drive a class-E sstc with bipolar transistor

To the original question: lowering the cap lowers the output power but voltage remains the same, but the fets (and core) usually heat more. That's it.

J.M.

Marko

Mon May 15 2006, 03:50PM

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

Again I messed up the words, I wanted to say that I USUALLY kept the cycle low (about 10% was maybe most efficient) but arc was biggest at 50%.
More than about 50% just lowers the output.
Once I left it like some 70'%, saturated the BDT, blew the mosfet to hell and charred entire breadboard.

Bipolar wants some serious current at base for proper turnon/off, I fed the base with just about 2 volts from BDT (ratio was 10:1 with 20V in) to keep current as high as posible.

And believe me, B863 transistor coped with high frequency operation quite well.

2N3055 wasn't happy at all in it's place.

Core and flyback would heat up a bit over time but it was nothing serious.
Transistor needed only a small heatsink, and everything could operate hours at higher duty cycles.

At 50% only problem was BDT and mosfet driving it, probably a totem pole driving it was too crappy.

Steve Ward

Mon May 15 2006, 04:45PM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

I am not so sure about the point with the operating frequency, since in the other thread you convinced me that a flyback transformer has an intrinsic secondary resonant frequency, and driving it close to that frequency would maximize the output. So when changing the surge impedance, I should probably try _not_ to vary the frequency, and keep the L*C product constant.

Now i have you all confused

. Anyway, what i meant is that the flyback will respond differently to different operating frequencies. The royer oscillator *forces* oscillation at the primary resonant frequency, so the secondary is just along for the ride. If your primary Fr is close to secondary Fr, you get much higher voltage output (and possibly thinner looking arcs, but i think this is due to another mechanism). Meanwhile, if you drive far from the resonant point, you might get less voltage and much more current in the arc. So this is all that was meant about the frequency dependency. You can observe this with other high power flyback drivers as well (i know i certainly have).

Anyway, if you want maximum voltage, you will most likely need to drive it at the secondary Fr (i would say its absolutely necessary, but im sure someone could throw enough power at it to still achieve the flybacks maximum output voltage, ie. insulation failure, while being off resonance).

cbfull

Sat May 20 2006, 06:02PM

Registered Member #187 Joined: Thu Feb 16 2006, 02:54PM
Location: Central Ohio
Posts: 140

Please excuse me for my ignorance, but can someone point me in the right direction to understand what the Marzilli circuit is? I've never heard of it before and google turned up nothing.

Thanks

...

Sat May 20 2006, 06:34PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

The Mazzilli circuit is a self resonant flyback driver that was designed by Vladimiro Mazzilli, here is the old forum thread on it

We all like it because it can force huge amount of power through just about any transformer that has an exposed core, is very efficient, and is even simpler :)

Dr. Shark

Sun Apr 13 2008, 02:28PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

I am bringing this thread from two years ago back up, since I don't want to start a new one one the same topic. (Also it contains this beautiful image of flyback cancer.) I dug out my old drivers and played around with them a bit to evaluate if they could be used for my Marx generator. I am looking for 30kV out at a few mA, and I want it to run off a 12V supply. This is my new shot at an answer to the original question:

The Marzilli circuit driver the flyback like a transformer, not as coupled inductors as a proper "flyback" circuit would. Flybacks work by creating an inductive kick of a few 100V in the primary (otherwise, why would there be a 1600V BJT in the orignial TV circuit), which is futher amplified by the turns ratio. Therefore, relatively few turns are needed on the secondary to get high voltages. The Marzilli driver with its pure sine wave does not create any sort of inductive kick, so it would be necessary to use significantly higher voltage on the primary.

I have experimented with this a little bit, by inserting an impedance matching transformer between the driver circuit and the fiyback. For this I used a flyback core with a 2+2 primary connected to the driver, and a 20 turn secondary which went to the flyback. This way I could supply up to 30V / turn to the flyback. Predictably though, this seemed to saturate the flyback core, i.e. the voltage in the LC tank circuit collapsed without any arcs on the output.

Now I am wondering, do I really need to go back to my half-bridge driver (which is hard-switching and therefore supports real "flyback" action), or is there any way to get this beautifully simple circuit to produce insane output voltage in addition to the insane currents it is known for?

...

Sun Apr 13 2008, 06:03PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

You need to tune the primary cap so that you run at the flybacks resonant frequinecy, I have got well over 30kv out of flybacks with the mazzilli driver

previous 1 2 3 next

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