I am having issues with this one mosfet flyback driver, please help if you can.

[zeal422]

Hey everyone, I hope you're having a good day / night. I was wondering if someone could help me out with this circuit please 😞 ? I have attached a video and a diagram. The sound you hear it's from the bench power supply and it doesn't sound healthy after this, actually I smelled some burnt something, don't really know from where, the psu still works, I tried it with a pc psu and it enters short protection and shuts off. I attached the circuit diagram on the photo. And I also put a snubber cap 100nF across the drain and source and and fast recovery diode as well. The mosfet I used is a irf3205. I can also try with a irfp460 or a irfp260n but I'm afraid of burning them since they're more expensive. I don't know exactly what I'm doing wrong.

Small edit: When the bench power supply was hissing (prolly from the mosfets inside it I am guessing, somehow this circuit was stressing the smps) was drawing all the amps was giving to it, same with the pc psu which can draw 26A on the 12v rail = entering in short circuit protection.

Any help is genuinely appreciated!

Video: https://i.imgur.com/NvT6Ir0.mp4

Diagram:

Thank you very much!

[xvr]

There is not a best way to implement flyback 
There are a lot of specialized DC/DC controllers for this.

This schema do not implement any regulation, so secondary voltage will be whatever it could. I guess that in your case it was very high and just burn out isolation of your flyback transformer.

[PGPG]

L of primary coil?
Isat of primary coil?

[zeal422]

"I guess that in your case it was very high and just burn out isolation of your flyback transformer."

No. That's not it. The flyback is fine.

[zeal422]

You mean what's the inductance of the primary wound coil ? I have no idea, I cannot measure it. It's around 10 turns.

[PGPG]

You mean what's the inductance of the primary wound coil ?

Yes.
U=L*dI/dt.
Assuming constant U we get U=L*ΔI/Δt and ΔI=U*Δt/L.
10nF*2k2 = 22us. NE555 on/off times are something around this.
If at the pulse beginning I=0 than Iend=ΔI.
We can assume Δt being around 22us, U being 12V and we need L to find if current at the pulse end is expected to be 1A or even 100A.

One of ferrite cores parameter is AL. For coil at this core you can assume L=AL*N², where N is number of turns.

[zeal422]

Thanks for the reply... unfortunately I do not have a meter that can measure inductance :/

I used like 1mm stranded copper wire for the windings

[MathWizard]

Was the FET even turning fully off ? I know from just simple DC mosfet circuits I've tried on a breadboard, sometimes they don't turn off for a while, when lowering the gate voltage back down by hand.

In every computer SMPS I've worked on, they have a voltage divider on th gate, usually like 10k to GND, and maybe 10, 20 ohms back to the driver. And usually they put a diode in parallel with the small series resistor, to clamp the max voltage drop to something reasonable.


Are you familar with LTSpice or any other simulator ? I'd try it in there, you may have to guess or calculate the inductance tho if you don;t have a meter. But you could also see the snubber currents easily.

[zeal422]

Hey there, thanks for the reply!

Uhm... I'm not really aware on how pc smps are really built (only partially)

Also uhm... I kinda did it... new vid here: https://i.imgur.com/K50cGV2.mp4

But the fet still is getting warmer and warmer on the small heatsink for only 12v at 4A... Is the fet fully turning off ? I mean, genuine answer from me is no, I don't know...

I have used LTspice for my previous one fet flyback driver with pwm module, like this one but don't really understand the program that good

[Andy Chee]

To minimise transformer core saturation (and the consequent MOSFET overcurrent), try and keep the duty cycle under 50%.

According to my calculations, your 555 component values give a duty cycle of 59%.

If you do not have the ability to measure the inductance, then calculate it using math formulas.

[zeal422]

I don't know how... I'm a weird learner, I usually learn visually or by someone explaining it like Mehdi (electroboom) does, or at least if I know the formulas I will give it a go of course. So for the duty cycle I should change what exactly ? C1 ? the 10nf one ? Leave the voltage divider alone or change that as well ? Thank you

[Andy Chee]

Do you know how to search for things using Google?

If so, then using the search terms "555 duty cycle calculator" will give you useful page results like this:

https://ohmslawcalculator.com/555-astable-calculator

Whatever stage of learning you are at, it is imperative that you do some basic searching on your own.... we are not here to spoon feed you.  For example, what was the original website of your schematic diagram?  What kind of information did the original website supply to help you build this project?

As a beginner, it would also be helpful for you to describe the bigger picture of your project.  For example, are you attempting to power a mains appliance with this device? Or is this intended for a tesla coil? 

These bigger picture details will help us experts narrow down aspects of the project that you, as a beginner, may not be aware of, or have considered.  Unfortunately even experts cannot read your mind and your intentions.