I built my first ZVS driver a while ago and had some fun making big arcs before I blew up the MOSFETs trying to make some alterations for induction heating. When it comes to this stuff I dont know exactly whats going on, but I know how to put it together well enough from a diagram.

I'm going to build another ZVS driver and I'm wondering: if I get IGBTs rated at 900 V and about 30 A, increase the capacitance, the inductance of the choke, and the number of turns on the primary, would I be able to rectify 120 VAC to about 170 VDC (with smoothing caps) and run it all with no transformer?

I know IGBTs need lower frequencies, so I figured upping the capacitance and number of turns on the primary will lower the frequency and I'm hoping that more inductance in the choke will limit the current input at 170 VDC.

Everything I see on-line says the voltage in the tank circuit is about 3-4 times the supply, so I figure 900 V for the IGBT should do it. Im just not sure about the current; would 20-30 A be enough? I can't draw more than 15 A with out tripping the breaker.

Am I on the right page here? I just don't want to order all this stuff and blow it up as soon as I turn it on.

(I have a transformer that can put out 48 V and the shorted secondary current is 10 A, but with my first ZVS it would not draw more than about 2 A, not until I made the alterations, and then it was pulling about 8 A and it made some funny sounds and POP!, there went the MOSFETs.)

enter image description here

  • \$\begingroup\$ Words are far less important than a schematic of what you propose. \$\endgroup\$
    – Andy aka
    Jan 17, 2018 at 11:25
  • \$\begingroup\$ Sorry about that, its up there now. Those are the original parts form my first one, but if i shaped out some of the parts, as i said, cold i just run it at a higher input voltage? \$\endgroup\$
    – Andrew
    Jan 17, 2018 at 12:50
  • 1
    \$\begingroup\$ Well, if you blew it up on 48 volts I'd stick with 48 volts until you understood the reason (rather than go to a lethal supply which is even more likely to bust devices). \$\endgroup\$
    – Andy aka
    Jan 17, 2018 at 12:53
  • \$\begingroup\$ When i ran it just like in the schematic it worked like a charm. It was only once i altered it that it broke. So i want to build another one, just like the schematic but with higher rated parts and run it at a higher voltage. Thats really all I want to know. Also, i didn't include details about the alteration that lead to failure because i didn't think it was so relevant to what i wanted to know. I will say i think i know what caused the failure occurred, i just dont know the why. But so long as i stick to the original design i should be alright.... right? \$\endgroup\$
    – Andrew
    Jan 17, 2018 at 13:45
  • 2
    \$\begingroup\$ The real big issue with such circuits is that the are self oscillating. Anything going wrong, like a short-circuit or maybe even an "odd" spark on the output may lead to a steady state operation point in which both MOSFETs are fully on and blow in microseconds. If it works like a charm at 48V I would stick on that, going higher voltage/power without substantially improving drive and protection circuits is indeed a big hazard. \$\endgroup\$
    – carloc
    Aug 19, 2018 at 7:37

1 Answer 1


Saturation in your RF isolation inductor is the likely cause of the midget failure. Saturation results in high currents with high drain voltages present.

  • \$\begingroup\$ What's an "RF isolation inductor"? \$\endgroup\$
    – winny
    Aug 25, 2022 at 9:25
  • \$\begingroup\$ It is the feed coil from the dc supply to the drain circuit of the MOSFETs. It blocks the rf present in the load coil. \$\endgroup\$ Aug 26, 2022 at 10:09
  • \$\begingroup\$ Strange. I would call it an inductor. \$\endgroup\$
    – winny
    Aug 26, 2022 at 10:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.