4

I want to make a ZVS flyback driver, but all the sites say to use 2 n-channel MOSFETs and I only have one powerful n-channel MOSFET. I have a bunch of p-channel MOSFETs, specifically, 2 of powerful Fairchild SFH9154 MOSFETs. I am wondering if there is a way to make a ZVS driver or at least some type of reliable flyback driver with a p-channel MOSFET and how I would do it. Also, does putting MOSFETs in parallel increase the "Tolerance" or make the Amperage rating higher?

Additional information: If one n-channel could be used in combination with p-channel, I have an IRFP450.

I also have about 15 IRF9450 p-channel transistors and a couple IRF610 n-channel (although I don't think they are great for high current applications).

feetwet
  • 2,322
  • 9
  • 31
  • 55
skyler
  • 10,136
  • 27
  • 80
  • 130

1 Answers1

2

You can make it with one MOSFET (paralleling them increases current capabilities), you just have to make sure you keep within it's ratings. Something like this circuit only uses a single IRFP450:

Flyback Driver

You could drive from the high side using your P-channel FETs, but I'd keep it simple and lowish power/voltage to start with.

Obviously be very careful with the high voltages ;-)

For a P-channel version, you can just swap all the polarities around from a design like this. You should end up with something like this (disclaimer, although it simulates okay, I have not tested this, so check everything carefully. The parts are not recommendations, just what LTSpice had available):

ZVS P-Channel

Oli Glaser
  • 54,990
  • 3
  • 76
  • 147
  • 3
    Note that in the circuit shown the gate drive of the MOSFET MUST NOT exceed Vgsmax for the MOSFET (OK with parts shown). (Not usually a problem). Also - it is a VERY VERY VERY good idea to add a small zener diode physically close to the MOSFET (short leads) connected gate-source so Zener never conducts in usual use. (Cathode to gate). Zener voltage to be >= Vout of regulator (12V here). This zener absorbs inductive spikes coupled from drain / flyback winding to gate via Millar capacitance. In some situations this enhances MOSFET survival vastly. – Russell McMahon Dec 25 '12 at 11:43
  • 10k in pot cct may want to be much lower or 18k much higher to allow lower pulse on % times. – Russell McMahon Dec 25 '12 at 11:44
  • Thats good, but the driver is not ZVS – skyler Dec 26 '12 at 20:12
  • You're right sorry about that - I was just focusing on how you can use your single N-ch MOSFET in a simple first go at it. (as @Russell notes there are a few things to watch out for with this type of circuit) If I get any time later I'll try to add another option or two. – Oli Glaser Dec 26 '12 at 21:41
  • I added a P-channel version of one of the circuits out there just to give an idea of what may be possible. Note - do not start with 40V, try a lot less first (e.g. 5-10V and winp up of nothing overheats) Careful with the polarities (e.g. notice the input voltage is -40V relative to the ground symbol, diodes reversed, etc) – Oli Glaser Dec 26 '12 at 22:58