Delete Delete Delete

[The_Almighty_Bacon_Lord]

Delete Delete Delete

[technogeeky]

The problem with transformers is that they are usually under-specified in datasheets. One has to rely on intuition and out-of-circuit LCR measurement to determine if it is damaged or not. It's normal in some cases for transformer windings to have very low resistance (though something like 0.01 ohms is probably too low). And a multimeter will beep SHORT even with a few ohms of resistance, which can be normal.

To me, the most interesting thing you said is: removing Q601 causes the fuse to stop blowing.

I would check the devices around it:

C606 (probably OK given that removing Q601 keeps fuse intact)
C607
D607A

I would also check the things on the high side:
C605
C611
C612


I'm at a loss to how any of these would result in the problem you're seeing.

I think you might need an oscilllscope for this. Or at least, I would need one. And it would need to be isolated (or diff probe, or isolate the TV, or both, blah blah).

[xavier60]

You should not be checking for the presence of shorts by seeing if it blows a fuse, especially a home made fuse. It causes more damage and can fry the transformer primary winding if it hasn't already.  Do you have the necessary equipment to do a "Ring Test" on the transformer?
This sort of self oscillating power supply circuit is very unforgiving. Miss a single damaged component and it will destroy itself again.
And there is no guaranteed way of safely powering up for the first time. A 100W lamp in place of the fuse might minimize the damage if something goes wrong.
"What stopped blowing the fuse: Removing Q601", wouldn't this mean that the MOSFET is shorted?
Take notice of the startup bias circuitry on the Gate of the MOSFET and figure out what happens next if for some reason the power supply circuit fails to self oscillate, assuming that there is no short circuit from Gate to ground.
Q621 doesn't make any sense at all. Does its Collector really go to the Base of Q602.

[Andy Watson]

What facilities do you have for powering this thing? Do you have a controlled voltage source - either AC or DC - that can be used to limit the fault current?
As your schematic says "HOT CIRCUIT, BE CAREFUL".  If you do not have an isolated AC/DC supply then it might be time to make a "dim bulb" current limiter, and proceed with extreme  caution!

... ~4A home made fuse

I'll not ask!
But make a mental note that the fault is mostly likely to be in the primary side. "Popping" fuses suggests a reasonably short circuit on the main high-voltage side - either a direct short-circuit, faulty component short (C605, C611, etc) or the switching transistor (Q601) is turned-on, permanently. Anything other than a direct short would probably be sending smoke signals. 

What I've replaced: Primary side: D608 (replaced with 1N4747),

Why? What led you to this?
Also, a quick google suggests that D608, a 1ZB36 is a 36V zener, whereas a 1N4747 keeps returning "20V Zener".

Q601 (replaced with IRF842)

Again, why? Have you determined for certain that the original Q601 was defective?

Ordinarily, targeting the switching transistor and the control IC would be my first line of attack - together with a-n-other faulty component. However, this circuit doe not have a control IC - it's self-oscillatory - so it becomes a bit more critical to replace like with like. I've not checked IRF824 vs TK10A50D - are they identical? Are they close ? I would certainly replace D608 with something closer to 36V operating voltage.

I'm starting to feel like the transformer has gone bad, because when I probe 6 and 2 in continuity mode, it indicates a short circuit

In or out of circuit? If it is still in-circuit there could be something else causing the fault. If it's out-of-circuit - there's your problem.

I'm sort of tempted to just completely remove the transformer, and see if the fuse will continue to blow. Will that yield any important results if I do that?

I suppose it will remove suspicion from some of the other components (like C605/11 etc) - but you could realistically (and more importantly, easily) test these components individually.

My focus of attention would start with Q601 - is it short circuit? - Does shorting gate and source allow power to be applied without the fuse blowing? No - you've got a short that should be traceable with an ohmmeter. Yes - move on to Q602. It is open? - remove and perform diode-test between base-emitter, base collector, etc. Not a conclusive test but if Q602 appears to be good, look for things that are holding its base down - like Q621.

[Rog520]

At this point I would pull the transformer and just do a basic shorts test for peace of mind. My guess is that the transformer is fine. With the secondary disconnected you're not getting proper feedback to the PWM control, however, so perhaps Q601 being turned on for much too long, causing the transformer to saturate and blow the fuse. I would leave the secondary connected. As others have suggested, use a 100w incandescent lightbulb in series with the hot lead of the supply. This will limit the fault current and allow you to troubleshoot without blowing the fuse.

[xavier60]

Maybe it is possible that Q601 is being permanently turned on and surviving while blowing the fuse because of the low mains voltage.  With the 250VAC here, chopper transistors never survive a fuse blow because the higher voltage cause fuses to arc internally and pass high fault current for longer.
Some sense needs to be made of the circuit. What turns the MOSFET on and how it's turned off before the primary current ramps up too high. Then how voltage regulation is eventually achieved.
I can't make sense of Q621, there must be a mistake.

[xavier60]

The function of Q621 is supposed to be to protect the power supply in case an over voltage of the mains power.
 It actually guarantees failure if an over voltage ever occurs.
Q621 should be omitted.

[xavier60]

The MOSFET in the kit appears to be a fully encapsulated type.