How To Repair A Switch Mode Power Supply Without Blowing It Up! Part 3

[@gregwmanning]

Amazing! Figuring out the fault is hard enough but dodgy replacement parts makes repairing much harder. Welldone Richard

[@Chief_Engineer]

Superb tutorial, better than anything I have seen before👍

[@fredfabris7187]

Love your details and drawings. Very helpful

[@nasosst3092]

👍👍👍 Your analysis is really impressive Rick.

[@BrianSmith-ow9gy]

Did you ever complete the finale Rich? I've been with you through all 3 vids and I'd love to see the repair completed and put to bed with pukka, as opposed to Chinesium, mosfets fitted.

[@anthonydenn4345]

I was just thinking the same thing 😄

[@Tims_Projects]

Do you think it would be a good idea to put the scope on the primary and secondary of the small transformer to check those small transistors are sending a good separated signal to the large MOSFETs?

[@alandrury9955]

The only other thing I would have done is put the scope on the output windings of the drive transformer to see the signal from the chip coming all the way through. Other than that, try the proper MOSFETs

[@LearnElectronicsRepair]

I didn't have the PSU connected to the mains, I could have safely done it as I have an isolation transformer on the workbench but I decided not to as it may have confused some viewers into doing something dangerous if they don't have the correct equipment or understand the potential risks involved. The disadvantage of that is that I couldn't test the transistors that drive the (gate drive) transformer or the transformer itself, though I did check them with the PSU powered off and could not find a problem there. Well I say couldn't, I probably could have (safely) using another bench PSU at 60V or similar. Could be a topic for another video...

[@TheSilmarillian]

I grew up diagnosing bread boars and building them , shame these days this talent is being lost by repair by replacement , most of my time was spent on computers back in the day when fixing something was the norm unfortunately.

[@ethanclement9647]

Are there any bad capacitors especially electrolytic capacitors? Perhaps you could try driving the power transformer at 44. Khz to see how it responds in case you have a shorted winding?

[@LearnElectronicsRepair]

I do hope it doesn't have shorted windings. That would quite possibly cause the fault if it is not due to fake parts which is a distinct possibility here. Your suggestion just goes to show it's very hard to test for every eventuality.

[@ethanclement9647]

I have done some repairs on switching supplies with success. Don't think I've worked on one that uses a transformer to trigger the mosfets like the one you're working on. Hopefully ginuwine parts fixes the problem. I always am curious what initially caused the parts to fail in the first place.

[@glennkingdon5129]

You should set up a test rig and see if the "fake" mosfets work at all or at what point they fail.

[@Chief_Engineer]

Could you separate the main switching circuit from the driving and auxiliary bits and then power it with a low voltage DC source from a bench supply with current limiter? That way the Mosfets wouldn‘t blow up in case of a short, but you could measure what‘s happening. The power stage should work with 30 Volts just the same as with 300..

[@ralphj4012]

Perhaps check the soft-start part on its own, just in case. I seem to remember this is the potentially-blow-the-capacitor-up-design (if the switcher doesn't start properly), because of the single balancing resistor, but I may be mistaken (as I am most of the time).

[@LearnElectronicsRepair]

Yes that is the one Ralph

[@thomasachlios8302]

also is practical to test the sound from subsonic and supersonic ammo....video available in you tube.....remind both have frequencies it out of limit in recordings mics......

[@LearnElectronicsRepair]

what???

I would appreciate it if I could hear back from you in the mail, I've been waiting

[@blindkeith]

just a though, why not make a small cct the put the fake transistor on the soak test at thier rated extremes.

[@davidmonk7773]

Hi Richard, have you got any voltage on your drive pins on the SG3525A? I had a similar issue, although the chip was oscillating it was sending voltage out on those pins, this was blowing my mosfets.

[@LearnElectronicsRepair]

I had the scope set to DC so it should have shown any DC offset. The other factor here is the mosfet gates are being driven via a small transformer so would that not prevent the type of fault you describe? It is an interesting one by the way

[@davidmonk7773]

@LearnElectronicsRepair Hi, Not sure, I've basically learned what I know from yourself, I remember though, I measured voltage at those 2 pins and apparently that meant the chip was no good, I changed the chip, now no voltage at those pins and it all worked? I think it also went through a small transformer too but not 100% on that.

[@LearnElectronicsRepair]

@@davidmonk7773 Thanks, I can check that with a multimeter, it didn't look like there was a DC offset on the scope. I've not seen the type of fault you describe but I would have thought that if the PWM IC is connected directly to the mosfet gate (or via a resistor) then if there is a DC offset plus the oscillations and the offset was enough to turn on the gate then the mosfets would go bang. If they are driven by a transformer like in this PSU, and DC offset would be blocked from reaching the gates though the transistors driving the transformer probably would not work if they were just being biased on so the PSU would not work either. This is speculation of course, just trying to imagine what would happen with that type of fault condition. Anyone else like to comment on this?

[@alaincharnier1971]

Rich, do you ever discharge big capacitors before working on this stuff?

[@LearnElectronicsRepair]

Yes, very often and I show it often. This PSU has not been plugged into the mains for 2 weeks since I published part two but you are right in saying I should have mentioned that 😊

[@alaincharnier1971]

@@LearnElectronicsRepair Thank you, I'll look for it in other vids. Full disclosure; I have made the mistake of discharging with a small 1/4 watt resistor and burning my finger 🤣🤣

[@ethanclement9647]

Fascinating! Pen/pencil to paper and you can figure things out.

[@LearnElectronicsRepair]

Yeah old tech but it works