Single Mosfet Flyback Driver / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Single Mosfet Flyback Driver

1 2 next

Move Thread

LAN_403

Dr. Dark Current

Sun May 14 2006, 01:58PM

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

You all know that flyback driver with one power bipolar transistor that efficiently converts the input power to heat

, but the main point is that it is simple. Today I was trying to adapt it to run with a MOSFET instead of the bipolar transistor with hope for better efficiency while still keeping the circuit simple, but all I got were some weak 5mm sparks.
I searched everywhere, but can't find a flyback driver which uses single mosfet and feedback winding. One reason for this could be that it never worked better than with bipolar transistor. If so, then why? Or have someone actually built this circuit with success?
I don't see a reason why it wouldn't work better than with bipolar transistor.

Thanks for help, J.M.

Dr. Shark

Sun May 14 2006, 02:23PM

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

I think Firkragg can answer this, whereas I can only speculate. In principle the circuit should work much better with a FET than using a Bipolar, you just have to use more feedback windings to get the voltage high enough for the gate. You should also add some Zeners to square up the gate signal a bit, to spend less time in the linear region, and the circuit might acutally become somewhat efficient. The problem - compared to the Marzzilli circuit - is that you drive the flyback with a square wave, so you will get voltage spikes. Use some kind of snubber to limit these, or put a TVS across the FET.

Marko

Sun May 14 2006, 03:07PM

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

I didn't try for flybacks, but I did some SSTC's with mosfets in royer configuration and lots of small inverters based around royer/blocking oscillator, etc.

Joe already explained what should you do, use more windings for feedback (calculate it etc. if your input is 30V and you use 6 primary windings, you should use around 3 feedback windings and 15V zener etc (zener can be from about 12 - 18V, 15 is somewhere in the middle and assures that mosfet is ON and not blowing up)

Good thing with mosfets is that you don't need huge 5 watt tesla resistors to bias the gate, but just small about 5:1 voltage divider.

Gate zener won't actually make transistnts any faster, but just cut the top of the signal making it look neater (but in the base it's still sinewav-ish crappy waveform).

All the windings on the flyback have a lot of leakaghe inductance and you can't help there much.

Maybe you can also try blocking oscillator (thing with bias directly on gate/base nad DC blocking cap on feedback.

Dr. Dark Current

Sun May 14 2006, 03:27PM

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

The Mazzilli driver is good, but I'm aiming for highest output voltage with lowest input power, so it is probably not a way to go :P

Firkragg wrote ...

Joe already explained what should you do, use more windings for feedback (calculate it etc. if your input is 30V and you use 6 primary windings, you should use around 3 feedback windings and 15V zener etc (zener can be from about 12 - 18V, 15 is somewhere in the middle and assures that mosfet is ON and not blowing up)

Good thing with mosfets is that you don't need huge 5 watt tesla resistors to bias the gate, but just small about 5:1 voltage divider.

Thats exactly what I did, but only 2 feedback turns though. The problem is that the feedback winding also spikes. With good flyback driver you got ca. 20V/turn, thats 60V spikes on feedback! (with only 3 turns)
The primary spiking is not a big problem, just use high voltage FET.

Marko

Mon May 15 2006, 11:32AM

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

You can try coupling feedback as strong as possible to primary, that should reduce spikes and make signal more square.

Zener is to kill the spikes but it may die if too much power is fed, dragging the mosfet with it.

Dr. Shark

Mon May 15 2006, 01:26PM

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

Jmartis wrote ...

The Mazzilli driver is good, but I'm aiming for highest output voltage with lowest input power, so it is probably not a way to go :P

In that case, you would probably be best of using some kind of discontinous, pulsed circuit. You could for example adapt the Ignition coil driver, which uses an SCR (replace by IGBT) to pulse a capacitor into the primary at a low rep rate. You would have to modify the circuit by using a smaller capacitor, as you are aiming for about 10 times the resonant frequency.

Hazmatt_(The Underdog)

Mon May 15 2006, 07:06PM

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

I have simple here

It's open loop contorlled, but once you tune it you have some pretty nasty output. I will make it closed loop sometime. If you use and opto-isolator to drive your feedback you don't have to worry about spikes too much because they feed into an led.

What you could do, and I would do this, is take a simple single output switcher, reverse engineer it, get the parts and have at it.

Dr. Dark Current

Mon May 15 2006, 07:19PM

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

Hazmatt_(The Underdog) wrote ...

I have simple here

I need something as simple as STFD, just with Mosfet (no timers/other active parts except the FET) and 2 windings.
Today I made an optimized STFD, the only problem is that the fly back gets hot (not the transistor!)

J.M.

Marko

Mon May 15 2006, 08:53PM

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

Today I made an optimized STFD, the only problem is that the fly back gets hot (not the transistor!)

Are you sure it's not arcing? Some pics of the setup, arcs..?

I have simple here

I think you should remove that 1 ohm resistor, I dn't see any point of it for 12V supply and burns power heavily.

Clamping circuit is completly unnecesarry for the low voltages given, just diode across the transistor can be used but it's not needed too.

Diode across flyback primary eats up the inductive spike and just burns it in a resistor.

Also note that you need a resistor of at least 1kohm at pin 3 of NE555, mine used to undergo very vierd behauvior when hooked directly like this. Totem pole overloads it in gate charging/discharging peaks.

I think you can get a lot better than just few mm arcs

Hazmatt_(The Underdog)

Mon May 15 2006, 10:43PM

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

mine gives about an inch. The resistor (if you understand design) is to limit the instantaneous current from going to infinity. I don't understand why so many of you guys don't see that.

All the clamping is there so it works for 20 yrs!! It is at college and needs to work day in and day out for a very long time after I graduate. Understand now? I hope so.

If you build it you will understand more about it rather then just trying to explain it and dismiss certian aspects.

1 2 next

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