TAGAN TG380-U01 PSU - 5v shorted after replacing caps

[kotel studios]

Hi

Recently I got this tagan tg380-u01 PSU with bad caps on 3.3v rail (the line was 2.5v no load!). Replaced them but no go, so I've replaced all of them, apart from some smaller ones and one under the secondary heatsink. Even that didn't help so I've replaced the others. All caps are known good ones.

Powered it on and heard the breaker pop. Opened it up and saw some smaller caps shorted together. Fixed that up but still it trips breakers. Now I get 0ohms from 5v to GND. Secondary heatsink removed and still a short, even with the PWM IC board removed. All diodes on that heatsink look okay.

Now I am quite stumped at what the issue might be, so I ask here for guidance on reparing this fine PSU.

[kotel studios]

Okay I done some testing on the rails and it seems to be okay, so I'll summarise this repair in short:
Low 3.3v (2.4v)
Break one 1n4148 zener
5V transistor cathode and anode shorted together after soldering the primary heatsink after replacing all the 3.3v, 5v and 12v rail caps
Primary 2SC3320's diying due to cascaded damage
Replaced with E170009 transistors^
3.3V low (2.3v)
3.3v PWM IC voltages seem good
See another zener cracked near the 3.3v PWM IC
Replacing
Fix

Many thanks to Momaka for guiding me while reparing this PSU!!

[momaka]

Some people are good at finding various odd schematics, but I'm not... and it's possible that there may not be one for this PSU after all.
So as usual, detailed pictures would be a good start (at least for me.)

All I can tell you so far.. if the PSU made your breaker pop, then very likely something on the primary side has gone short-circuit.
Technically, a shorted 5V rail shouldn't be able to make the primary blow up, as the supervisor/PWM controller should be able to catch that fault as it tries to start up. But I have seen it happen to some PSUs.

That being said, build yourself a "dim bulb tester" with an incandescent light bulb, like so (scroll down to post #70 after opening the link):
https://www.badcaps.net/forum/troubl...684#post563684

[kotel studios]

Yes yes, pictures will be added soon.
The breaker popped once I shorted PS_ON tho... Before it was okay. PS_ON and 5VSB were good. After that now somethings screwed up beyond my knowledge. While it would be nice to get working on this one and get a high quality PSU, we need to finish with the excellent power firstly.
And yes, I already have an dim bulb tester. Got some time to make an cable and now I run both the excellent power and this and future through it. (also, nice drawing skills )

[kotel studios]

Attaching pictures. For some reason it doesn't like the rear picture... I'll try to post some pics with secondary heatsink removed too.

[kotel studios]

Attaching the rest of pictures.

[kotel studios]

Okay, so I tried pulling out all the caps on 5v rail, some resistors but nothing. I've pulled out the L5 (blue coil on pictures) and now there isn't a short on cathodes to gnd on D30 and Q9. There's still a short on 5v rail to gnd tho...
I think this points to the short being lower. Although all the diodes measure fine in circuit....

EDIT: It looks like the stray wires from my solder wick got onto 5v and gnd pins on its diode. Without them it looks okay. Will replace the components and report if I find a short or if it will make my light buld light up like an Christmas tree

[momaka]

Ah, see, this is why pictures are so important - I recognize this PSU!
As if by luck, I also have a very similar PSU in my "collection" and posted about it a long time ago, here (scroll down to post # 2352... and you can see images at the bottom of the post too):
https://www.badcaps.net/forum/troubl...118#post877699

The breaker popped once I shorted PS_ON tho... Before it was okay. PS_ON and 5VSB were good. After that now somethings screwed up beyond my knowledge.

From the first picture in post # 4 above, it looks like your PSU has an add-on board for APFC function, similar to the Tagan EasyCon 480W from this thread:
https://www.badcaps.net/forum/troubl...480w-tg480-u15

So my take as to what may have happened to your PSU: your APFC circuit probably blew up (i.e. the MOSFET shorted on that little add-on board) or is doing something your breaker doesn't like (some APFC circuits can momentarily draw a very large inrush current.).
With more pictures of the primary side (one angled picture, showing the area of the primary caps, and one solder-side picture of the APFC board), I should be able to figure out more stuff and give further guidance.

The good thing about this PSU is that it's a half-bridge -based design, so it uses 2x 200V or 250V primary caps and *probably* can run without the APFC board (with a few modifications) if that comes to be the source of trouble. But before you do that, run it through your dim bulb tester and report back what it does - that is, plug the PSU in the dim bulb tester and see if 5VSB comes up properly. Then try to turn it on via shorting PS_ON to ground. From what I can see, the main fuse has not blown yet... which is strange. In which case, see my last few comments below about the dim bulb tester device (pertains to using it with APFC PSUs.)

While it would be nice to get working on this one and get a high quality PSU, we need to finish with the excellent power firstly.

Agreed.
But figured I'd post here just to get the thread "started" with the pictures, so to speak. That way I can view them and if something comes to mind, write it down here. I'll post more in that one shortly (well, maybe not so "shortly" as I tend to think/write slowly.)

And yes, I already have an dim bulb tester. Got some time to make an cable and now I run both the excellent power and this and future through it.

Good!
Just a heads-up, though: PSUs with APFC might make your dim bulb tester blink rapidly as the APFC tries to come up to full voltage... and in some cases, might even make the bulb stay lit or completely refuse to work after it blinks the bulb. For these, stepping up to 2x or even 3x 100W bulbs in parallel might be necessary. I sometimes use 500 Watt or more heating elements instead, but only after I have thoroughly verified that nothing else is wrong with the incandescent dim bulb device first. Otherwise, 500 Watts of power can often be enough to damage some components again (though still not as bad as plugging the PSU directly in the wall.)

*EDIT*
Also, I have a small personal request for you, if possible: when attaching images, please use the "Upload Attachments" button in the upper right corner of the post editor instead of the Photo icon on the left side. The latter makes images take forever to load on my end... and I can only load one at a time. It's just the new silly forum software here that causes this. When images are uploaded as attachments, I can right-click and load them into a new tab, so I can read/write in the thread while the images load (and also jump between them easier.)

[kotel studios]

Okay bad news. No shorts but the bulb still lights up as a Christmas tree...
But without the APFC the bulb doesn't even light up for a second at the start.
APFC MOSFET's are all good. I have no idea what the hell is even wrong now... There is an missing resistor near the PWM IC board but it looks like it was missing before I got it to pop breakers.

[kotel studios]

Okay, so I did what other suggested in the 480w PSU thread and.... both of the primary drivers are shorted....
What's weird is that one shows 0mV only in forward bias (reverse bias shows nothing) while the other shows 0mV everywhere (cathode and anodes). I will try to pull them out tomorrow and check them with my component tester.
If anybody has any more ideas on what could also died please tell me

[kotel studios]

Yep, both of the primary drivers (2sc3320) are shorted together. Without them the short on PCB is gone. Weird how the smaller diode right next to them on the same heatsink isn't shorted.
Now since I don't have any 2sc3320's at hand and don't know if it will still make my bulb light up I have an idea to just power on the PSU for a quick second without them and see if the bulb lights up. Although since those are primary drivers, damage will be very likely to appear.
Now could somebody tell me if it's okay to do what I wrote above?
Also, are there any other NPN transistors I could replace the 2sc3320's or should I stick with the 2sc3320?

[kotel studios]

Okay so for two 2sc3320's I pay a little less than an mid range 300w PSU from reputable brand.
Are there any equivalents to it which are made to this day? I can't find any equivalent part by el component for sale in Poland.

[momaka]

Okay, so I did what other suggested in the 480w PSU thread and.... both of the primary drivers are shorted....
What's weird is that one shows 0mV only in forward bias (reverse bias shows nothing) while the other shows 0mV everywhere (cathode and anodes). I will try to pull them out tomorrow and check them with my component tester.

Yes, check them with your component tester, just to confirm.
Somehow, I don't see how both of these can be shorted, yet your PSU's fuse still intact. Typically when these two short-circuit, the damage is almost always a blown fuse... or in rare cases, a blown (open) NTC thermistor.

Worth noting that your "primary drivers" are NPN BJT's (Bipolar Junction Transistors) and not MOSFETs. These have 3 terminals: Emitter, Base, and Collector.
If you ever need to test these without a "component tester" and just with a regular meter, you use the diode test function, and there are 6 tests to perform:
1) Multimeter red (+) probe on Base terminal and black (-) probe on Collector terminal should give a reading in the range of 0.5 - 0.75 V (or 500 to 750 mV)
2) Multimeter red (+) probe on Base terminal and black (-) probe on Emitter terminal should give a reading in the range of 0.5 - 0.75 V (or 500 to 750 mV)
3) One MM probe on Collector terminal and the other probe on Emitter terminal (doesn't matter the orientation) should both give open-circuit reading
4) Multimeter black (-) probe on Base terminal and red (+) probe on either Emitter or Collector should both give open-circuit reading.
^ All of these apply only when measured out of circuit.
In circuit, sometimes it may be normal to see a low(er) resistance or mV reading between Base and Emitter due to Emitter pull-down resistors (typically the case of the HOT in CRTs.)

Yep, both of the primary drivers (2sc3320) are shorted together. Without them the short on PCB is gone. Weird how the smaller diode right next to them on the same heatsink isn't shorted.

Can you be more specific here?
- Short on PCB gone... which PCB (main or APFC add-on board) and between what two test points?
- diode on the same heatsink? Are you sure that's not the 5VSB MOSFET? (2sk2645)

Now since I don't have any 2sc3320's at hand and don't know if it will still make my bulb light up I have an idea to just power on the PSU for a quick second without them and see if the bulb lights up. Although since those are primary drivers, damage will be very likely to appear.
Now could somebody tell me if it's okay to do what I wrote above?

Yes, should be OK to power the PSU with the primary BJTs removed. BUT WAIT!
First, remove those big 200V/250V caps and make sure they are fully discharged. Then test them on your component tester to make sure they show proper capacitance and ESR. If you look towards the end of the discussion in that other Tagan 480W thread, member MHTSOS mentioned that on both units he had for repair, it was these two caps that had gone bad. This is not a coincidence! - APFC circuits tend to be very hard on the primary caps and can make them go open-circuit. When that happens, the energy from the APFC circuit inductor toroid has nowhere to go and will reach extremely high voltages - enough to short out the next weakest part. If this was the primary BJTs, then that's why they blew up.

So to summarize here: remove the primary caps, make sure they are discharged (so that you don't damage your component tester) and then test them with your component tester for proper ESR and capacitance. If they are OK, only then install them back in the PSU and try the test where you power up the PSU (but through the dim bulb tester, only) to see if the short is gone. If nothing else is damaged, you should see the 5VSB rail up and running. After all, remember that ATX PSUs are actually made of two separate power supplies inside - the 5VSB one and the main one (3.3V, 5V, 12V, -12... and so on.)

Also, are there any other NPN transistors I could replace the 2sc3320's or should I stick with the 2sc3320?

E13007 and E13009 are typically what most half-bridge PSUs use.
But really, at least for testing purposes, you can take any other ATX PSU that's based on half-bridge topology (that uses BJTs) and use the BJTs from that to test this PSU.

That being said, if your 2sc3320's really are blown, chances are they are not the only parts. You will also need to pull out every part of the Base-driver circuit to check it - that is, all of the components that connect to the Base. Typically, that's one or two diodes, a 1-Ohm resistor, a 2.7-kOhm resistor, and a small 50V electrolytic cap in the range of 1 uF to 10 uF. If any of these are damaged or bad, you will very likely damage your new BJTs again when you try to power up the PSU.
Also, when the primary BJTs blow, sometimes that can take out the secondary-side driver BJTs - the small ones that I had you check in your Excellent ATX PSU... along with their protection diodes and other connected components.
Yeah, it's a lot of stuff... or at least sounds like it. In reality, it's not that bad and these half-bridge PSUs aren't that hard to diagnose once you get the hang of it. If you want a detailed read where I did something like that, see this (very long) troubleshooting thread:
https://www.badcaps.net/forum/troubl...pcb-wf-c-rev-e
.
.

[kotel studios]

Yes, check them with your component tester, just to confirm.
Somehow, I don't see how both of these can be shorted, yet your PSU's fuse still intact. Typically when these two short-circuit, the damage is almost always a blown fuse... or in rare cases, a blown (open) NTC thermistor.

Worth noting that your "primary drivers" are NPN BJT's (Bipolar Junction Transistors) and not MOSFETs. These have 3 terminals: Emitter, Base, and Collector.
If you ever need to test these without a "component tester" and just with a regular meter, you use the diode test function, and there are 6 tests to perform:
1) Multimeter red (+) probe on Base terminal and black (-) probe on Collector terminal should give a reading in the range of 0.5 - 0.75 V (or 500 to 750 mV)
2) Multimeter red (+) probe on Base terminal and black (-) probe on Emitter terminal should give a reading in the range of 0.5 - 0.75 V (or 500 to 750 mV)
3) One MM probe on Collector terminal and the other probe on Emitter terminal (doesn't matter the orientation) should both give open-circuit reading
4) Multimeter black (-) probe on Base terminal and red (+) probe on either Emitter or Collector should both give open-circuit reading.
^ All of these apply only when measured out of circuit.
In circuit, sometimes it may be normal to see a low(er) resistance or mV reading between Base and Emitter due to Emitter pull-down resistors (typically the case of the HOT in CRTs.)


Can you be more specific here?
- Short on PCB gone... which PCB (main or APFC add-on board) and between what two test points?
- diode on the same heatsink? Are you sure that's not the 5VSB MOSFET? (2sk2645)


Yes, should be OK to power the PSU with the primary BJTs removed. BUT WAIT!
First, remove those big 200V/250V caps and make sure they are fully discharged. Then test them on your component tester to make sure they show proper capacitance and ESR. If you look towards the end of the discussion in that other Tagan 480W thread, member MHTSOS mentioned that on both units he had for repair, it was these two caps that had gone bad. This is not a coincidence! - APFC circuits tend to be very hard on the primary caps and can make them go open-circuit. When that happens, the energy from the APFC circuit inductor toroid has nowhere to go and will reach extremely high voltages - enough to short out the next weakest part. If this was the primary BJTs, then that's why they blew up.

So to summarize here: remove the primary caps, make sure they are discharged (so that you don't damage your component tester) and then test them with your component tester for proper ESR and capacitance. If they are OK, only then install them back in the PSU and try the test where you power up the PSU (but through the dim bulb tester, only) to see if the short is gone. If nothing else is damaged, you should see the 5VSB rail up and running. After all, remember that ATX PSUs are actually made of two separate power supplies inside - the 5VSB one and the main one (3.3V, 5V, 12V, -12... and so on.)


E13007 and E13009 are typically what most half-bridge PSUs use.
But really, at least for testing purposes, you can take any other ATX PSU that's based on half-bridge topology (that uses BJTs) and use the BJTs from that to test this PSU.

That being said, if your 2sc3320's really are blown, chances are they are not the only parts. You will also need to pull out every part of the Base-driver circuit to check it - that is, all of the components that connect to the Base. Typically, that's one or two diodes, a 1-Ohm resistor, a 2.7-kOhm resistor, and a small 50V electrolytic cap in the range of 1 uF to 10 uF. If any of these are damaged or bad, you will very likely damage your new BJTs again when you try to power up the PSU.
Also, when the primary BJTs blow, sometimes that can take out the secondary-side driver BJTs - the small ones that I had you check in your Excellent ATX PSU... along with their protection diodes and other connected components.
Yeah, it's a lot of stuff... or at least sounds like it. In reality, it's not that bad and these half-bridge PSUs aren't that hard to diagnose once you get the hang of it. If you want a detailed read where I did something like that, see this (very long) troubleshooting thread:
https://www.badcaps.net/forum/troubl...pcb-wf-c-rev-e
.
.

Yes, I have tested them out of circuit and both of them show 0mV from collector/emmiter to base (no matter MM probe orientation).

What I meant by the short gone was that, all the pads of the BJT's on main PCB appeard shorted together, by that all of the pins showed 0 ohms. APFC board looks to be fine. Without the BJT's they show more than 100ohms each. And no, this isn't the 5VSB circuit. These are 2sc3320 for the power rails generation, at least I think so.
Main caps look to be not shorted together, so low chances that they are internally screwed, but I still need to check the ESR (I hope my cheap component tester will fit them😅).
The E13007 and E13009 ones are in TO-220, so maybe they'd fit the TO-3P?
Well, it looks like it isn't down-hill from here as it looked like for me....

[kotel studios]

Okay I found a pair of E13087's which were probably taken out of the lc psu. Although they are too short I also found a pair of 20n60c3, although not sure if they'll work...

[kotel studios]

Okay applied power without all the transistors on that heatsink (and the 5vsb transistor) and the bulb doesn't light up as a christmas tree! Lights up for a sec but thats normal because it means the APFC circuit initiated (or similar). Now to replace the 5vsb one and then the main transistors and #1 issue is gonna be fixed!

[kotel studios]

Okay with 5vsb transistor in, the bulb doesnt light up as an christmas tree, albeit I hear a loud screech when turning off the psu (I think I even see the apfc board twitch slightly on the desk) and the psu starts to screech after a while on its own (I see the bulb light up) so I cut the power before it explodes..... I am afraid to keep it on for longer.
Any ideas? The primary caps were replaced correctly.

[momaka]

Yes, I have tested them out of circuit and both of them show 0mV from collector/emmiter to base (no matter MM probe orientation).

Ah, OK, they are certainly shorted then.

Main caps look to be not shorted together, so low chances that they are internally screwed, but I still need to check the ESR (I hope my cheap component tester will fit them😅).

It's more dangerous when these caps go open-circuit than when they short out. So that's why check them with your component tester for proper capacity and ESR. I believe all of the component testers can do that.

The E13007 and E13009 ones are in TO-220, so maybe they'd fit the TO-3P?

I guess you didn't read the KDM Power thread I linked to in post
Here's a picture showing what the original TO-3P BJTs looked like (a pair of D13009's).
https://www.badcaps.net/filedata/fetch?id=2226986

And here's my kludge / "fix up" with TO-220 BJTs (13007's from a different PSU.)
https://www.badcaps.net/filedata/fetch?id=2058931
^ Not only were these 13007's a different case size, but even their leads were too short... so I soldered extension leads to make them fit. Not pretty, but it works.
So yes, you can switch between TO-3P, TO-220, and TO-247 as much as you like. Only thing you really have to be careful with is make sure that the tabs stay isolated from the heatsink via their thermal pads and also the plastic washers they have on the screws.

Okay I found a pair of E13087's which were probably taken out of the lc psu. Although they are too short I also found a pair of 20n60c3, although not sure if they'll work...

Too short as in short leads? Or short-circuited?
The latter would be a no-go ideed.
And no, 20n60 won't work, because that's a MOSFET. MOSFET's and BJTs are not directly interchangeable in circuits.

Okay with 5vsb transistor in, the bulb doesnt light up as an christmas tree, albeit I hear a loud screech when turning off the psu (I think I even see the apfc board twitch slightly on the desk) and the psu starts to screech after a while on its own (I see the bulb light up) so I cut the power before it explodes..... I am afraid to keep it on for longer.
Any ideas? The primary caps were replaced correctly.

Can you explain in a bit more detail what's going on?
The loud screeching after cutting power is normal for some 5VSB circuits, provided it's not extremely loud. A quick "zzzzzzzzup" before the 5VSB turns off is somewhat common.
But continuous loud screeching doesn't sound normal.
Does the 5VSB come up at all (i.e. do you get 5V on the purple 5VSB wire?)? Also, why did you replace the original 5VSB MOSFET on the primary? Was it shorted? I find it unlikely to have both the main power supply BJTs shorted and the 5VSB primary MOSFET shorted.
As for the APFC, I don't get what's going on - you say the bulb doesn't light up, but then you say it does?? Define what you mean when you say "light up as a christmas tree" - is it blinking On and Off quickly, or does it just stay lit? If it's blinking quickly, that's probably the APFC trying to turn run. As I said before, APFC circuits don't like dim bulb testers... but I don't suggest running without the dim bulb tester either. So the next thing you have to do is put 2x 100W bulbs in parallel to each other (for a total power limit of 200 Watts) and test again. But before doing that, verify that your 5VSB circuit's caps are all good - not only the two on the output, but also the startup cap near the UC3842b PWM controller (should be a small 10-100 uf cap somewhere on the primary of the 5VSB circuit.)

[kotel studios]

Actually, scratch the issue with the e13007's. They should fit perfectly.
I'll clear out the confusion.
For "lights up like an christmas tree" I mean the bulb lights up fully, not flashing, just continuously on, indicating a high power consumption (or in simpler words a short).
Yes, a quick "zzzzzzzzzzzzzzzup" is normal when cutting power, albeit this one does like an capacitor was boiling mixed with coil while. It's hard to record on video but I'll try my best (note the sound might be uploaded on google drive due to size constraints). Same goes when the bulb lights up after a while and the PSU does the same weird noise on its own (not shorting rails). Not sure what does that mean but nothing promising an successful repair....
About 5VSB, I have no idea. I am afraid to even touch the PSU in fear of getting shocked and having it explode on me (lessons from my compaq psu). The mosfet wasn't replaced, just put back in (yes, I made sure its in its spot and not the 2sc3320's). It is fine otherwise.
About APFC I meant the bulb flashes for a second, which from my knowledge means the APFC initialized or similar, it doesn't stay on.

I don't think the secondary side would go bad after shorting 5v to GND. Especially since this PSU "worked" before I screwed it up due to my stupidity. But everything can happen with old stuff. I'll check that cap and report back.

[kotel studios]

I have accidentally shorted 2 pins on the pwm board... Fixed it and it looks like the screeching is delayed to 2-3 seconds (it does it on its own), although I can't push myself to leave it on when it screeches. While there is still the suspicious screech from killing the power to the PSU, it isn't as strong as before.

[momaka]

For "lights up like an christmas tree" I mean the bulb lights up fully, not flashing, just continuously on, indicating a high power consumption (or in simpler words a short).
Yes, a quick "zzzzzzzzzzzzzzzup" is normal when cutting power, albeit this one does like an capacitor was boiling mixed with coil while. It's hard to record on video but I'll try my best (note the sound might be uploaded on google drive due to size constraints). Same goes when the bulb lights up after a while and the PSU does the same weird noise on its own (not shorting rails). Not sure what does that mean but nothing promising an successful repair....

It sounds like the APFC circuit is trying to run, and I'm assuming you only have one 100W bulb for the dim bulb tester? If so, try something more powerful, like the 2x or even 3x 100W bulbs, all three connected in parallel to basically equal a 300W ballast. At 230V, this gives you P / V = 300 Watts / 230V = 1.3 Amps maximum allowable current to pass onto the PSU, assuming something is shorted. The 5VSB MOSFET is rated typically for 2-4 Amps at least... so even if there is a fault in the 5VSB circuit, nothing should outright exploded. At best, you might smoke a smaller resistor or transistor, *IF* there really is a fault. But I suspect your APFC circuit is just trying to run, and sometimes these just don't like to play nice with a dim bulb tester, so they get "stucK" and keep the bulb lit all the time.
So try that first... and if bad comes to worse, we will just remove the entire APFC circuit and install a regular bridge rectifier on the PSU, so then you would have one less circuit to worry about while troubleshooting this PSU.

About 5VSB, I have no idea. I am afraid to even touch the PSU in fear of getting shocked and having it explode on me (lessons from my compaq psu).

Yes, I remember and understand that anxious feeling back when I started repairing power supplies.
The dim bulb tester really removes the danger of anything exploding violently. Nonetheless, I always wear safety glasses when I'm looking / probing at any electronic device - it's just a good safety measure.
As for the danger of getting shocked - the dim bulb tester does *NOT* make this go away. However, if you just stay focused and always exercise good caution around the high voltage parts of the circuit, you'll be OK. Just follow these safety tips:
1) wear eye protection, always
2) mind where your hands are at all times - don't touch high-voltage and primary-side components with bare hands.
3) make sure your device is placed on a non-conducting (isolating) surface, and also one that is preferably fire-resistant (i.e. don't place a "live"/powered PCB on a piece of cardboard, paper, or other similar surface that can easily catch fire if something overheats)
4) If you have pets or little children around (or anyone unaware of the dangers of electronics), keep them out of the room. With pets, the danger is they can sometimes knock over something or tug on a wire, causing the whole live device to move and become a shock hazard for both you and them.
5) use a power strip with a large "On/Off" button that you can easily toggle Off if/when needed.
6) If you have alligator clips / jumper wires with clips, use those to make all wiring connections prior to applying AC power. Try to set up as much of the equipment and wiring prior to powering up as possible. For measuring voltages with a multimeter for example, I always clip the black MM probe to either ground or primary-side ground (depending on what I'm measuring) and then only use the red MM probe in one hand to measure voltages, while I have the other hand on the On/Off button of my power strip, in case something goes wrong and I need to turn power off quickly.

So for your PSU, you can just set your MM to voltage measurement, then connect the black probe to the PSU's ground and the red prove to the 5VSB. After that, turn the meter on, then power up the PSU and see if the 5VSB comes up or not (though I suspect it probably won't if the bulb is lit all the time.)

The mosfet wasn't replaced, just put back in (yes, I made sure its in its spot and not the 2sc3320's). It is fine otherwise.

OK, good!

About APFC I meant the bulb flashes for a second, which from my knowledge means the APFC initialized or similar, it doesn't stay on.

The flash is just the primary caps getting charged.

But again, what I don't understand is that you say that it flashes (and turns off), yet you also say it stays fully lit all the time "like a christmass tree"... so which one is it? Or does it light up briefly, then go off, then come on again and stay fully lit?
I just want to figure out what is going on when the PSU is given AC power.

I don't think the secondary side would go bad after shorting 5v to GND. Especially since this PSU "worked" before I screwed it up due to my stupidity. But everything can happen with old stuff. I'll check that cap and report back.

Yes, shorting the 5V rail to ground shouldn't do anything to the PSU. In fact, this is a standard test I run on pretty much all of my "gutless" / no-name PSUs to make sure they pass the short-circuit to ground test. I do this also on the 3.3V, 12V, and -12V rail. Even the really crappy PSUs usually don't have a problem with this protection and shut down properly... well most of them anyways.

Now, if you short 5VSB to ground, that is not always a given that the PSU will survive, depending on how the 5VSB circuit was designed. In particular, the 2-transistor designs don't have any protections, other than a very basic power limit on the primary side... which is usually set too high and will overheat and kill the circuit in many cases if the overload on the 5VSB is not cleared quickly. But in the case of this Tagan PSU, I believe it should be very similar to my ePower unit and have a UC3842 IC for the 5VSB... which generally has much better protection from the circuit getting damaged.

That said, I don't know what happened to your PSU... but we will keep digging and hopefully find out... eventually.
.