0
\$\begingroup\$

Previous thread: Flyback transformer

So, in short, I'm attempting to build a flyback transformer. The schematic I am using is this:

test

And the flyback core looks like this:

test

Upon recommendation from the previous thread, I got an actual ferrite core from a TV and wound it myself, 10x 80 turns of 0.125mm copper wire (as in 800), each layer (composed of 80 turns) separated by insulation tape. The primary coil (5 turns) was wound underneath these. All the turns are over the airgap in the flyback core.

The circuit is running to an extent. There's a high pitched squeal when it turns on, and the MOSFET heats up. But no arc. I haven't tried measuring output voltage because I'm worried about frying the multimeter, but it should produce an arc.

I'm really not sure what's going wrong here. I can confirm that the circuit is assembled correctly. The flyback core was damaged when trying to take it out (one of the bars snapped in two), but I glued it back together, having been told that would fix it, and I don't see why that would not be the case.

My power supply is 12V, provided by a 9V and 2x1.5V in series, as required.

I have been told before that the circuit schematic given is not very efficient (as shown by the MOSFET heating up). I'd like to point out that at this stage, getting the circuit to work at all is a greater priority for me than efficiency. I don't want to shell out on new components or build a whole new circuit until I'm sure that that's the issue here.

Overall view of the circuit build:

test

Okay, got round to testing it. Anda aka were right about the power supply. When running, the PD across the batteries drops to ~9V. I tested it with 15V - dropped down to 10. At 12V the output current was about 0.2A. Makes sense I guess.

Question is: what could I use to fix this? I'd rather still use batteries due to portability.

\$\endgroup\$
  • \$\begingroup\$ You need to add a resistor between the gate and ground - about 2k2 should do it. There's nothing to discharge the gate capacitor so your MOSFET is probably not switching off (latching) and hence getting warm. Have a look at sites.google.com/site/uzzors2k/flybacktransformer \$\endgroup\$ – JIm Dearden Jul 7 '13 at 19:55
  • 2
    \$\begingroup\$ Jim: That comment makes no sense. There is a 330 Ohm pull-up, and a transistor pull-down, on that MOSFET gate. The MOSFET shouldn't get warm, though. Perhaps try changing the 330 Ohm pull-up to 100 Ohms? Make sure the switching frequency out of the 555 isn't too high? \$\endgroup\$ – Jon Watte Jul 7 '13 at 20:24
  • \$\begingroup\$ @JonWatte Your absolutely right Jon (+1)- its getting late and I've been on here too long - my old eyes must be seeing things that aren't there. \$\endgroup\$ – JIm Dearden Jul 7 '13 at 20:36
  • 1
    \$\begingroup\$ A regular 555 can source and sink 200mA, I don't get why th 2N2222 is in there. \$\endgroup\$ – jippie Jul 8 '13 at 16:47
1
\$\begingroup\$

I did debate about this answer being a comment so don't bust a gut voting down (or up) but I had too much to say to fit a comment.

Can you get 3A from those batteries? For now I'll assume you can...

Assuming that you are producing no-more than 80V on the primary (because the FET will break if you are doing a 100v) and given your turns ratio of 1:160, and given the "imperfect coupling factor" (that I pull out of the experienced side of my brain) I'd say you should produce about 5kV.

The FET is getting warm and this isn't a worry but it does rather hint at your 12V supply rail falling to somewhat less when it is powered on. Maybe you can measure it. My suspicion is that it might be falling to about 6V or 8V and this will not efficiently turn the fet on and this could be causing the FET to warm. If this is true then maybe you might only be producing 1kV.

Of course, there is a strong possibility that you damaged your FET with over voltage the very forst time you used it. I note on your earlier threads/questions you were advised to use a 200V device with lower on-resistance.

I'll also point out that a lot of these circuits are from unprofessional sites that might have (with luck) got something to work and easily misconstrued their own circuit diagram thus misleading everyone who tries to build one. Also notable about the website that contains the circuit is the lack of information about the transformer and how to wind it.

The original design came from: -

http://www.geocities.com/CapeCanaveral/Lab/5322/fbt2.htm

But don't bother trying the link because geocities is no more.

Looking at your pictures and knowing how difficult it is to get a step-up transformer working with ratios over 100 I have to point out that there is no hope for the way you have it connected. 555 needs to drive the transistor which provides voltage to the FET and the FET source, emitter of the transistor and 0V of the 555 have to be "rigidly connected". The currents you might be inducing in 0V connections could easily mean you get less than 200V out of this design.

\$\endgroup\$
  • 3
    \$\begingroup\$ Replace geocities by reocities and the link opens. \$\endgroup\$ – Renan Jul 7 '13 at 21:08
  • \$\begingroup\$ I've got a few solutions to test. I'd to clarify a couple of things though. I can test the output current and voltage easily, but how would I be able to test if the FET is working? (I do recall leaving the circuit on for a little too long a while ago and I might have smelt something melting, not sure) Also, what do you mean by rigidly connected? \$\endgroup\$ – Alex Freeman Jul 7 '13 at 21:22
  • \$\begingroup\$ Rigidly connected means on very short wires. 0V power in should first (no compromise) go to the FET then on very short wires go to the emitter then the 555 circuit. Power in should go to the transformer first then to the 330ohm and 555. Lead between drain and transformer minimal. Ditto transformer to 12V in. \$\endgroup\$ – Andy aka Jul 7 '13 at 21:49
  • \$\begingroup\$ I see. But I'm not sure I see why that's important, is hundreds of milliohms really going to affect the performance compared to other factors? Also any ideas for testing the FET? \$\endgroup\$ – Alex Freeman Jul 7 '13 at 22:05
  • \$\begingroup\$ @AlexFreeman it's not just 100 milliohms (which at 3A might wobble all signals by 300mV); it's the loop inductance of wires. If i think a FET is bust I just throw it away and get a new one. \$\endgroup\$ – Andy aka Jul 8 '13 at 7:19
0
\$\begingroup\$

Your MOSFET may heat up because of a slow turn on transition, as it depends completely on the pull up and creates an RC charching circuit with the gate capacitance that are not neglectable for power MOSFETs. Whenever the MOS conducts with a resistance dissipates power. Also, when it's fully on it may have a too high Ron resistance, check it. The old IRF510 is not the best choice I'm thinking... Can you simulate the circuit with LTSpice or any tool? It pays well the time you spend on it, and avoids frustration and empirical searching. The best driver for a MOSFET would be a push-pull architecture giving/drawing enough instantaneous current to charge/discharge the gate in some ns, any greater than that, especially if switching some amps, sounds to me like needing a fan and a chimney.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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