Okay in my parts bin I also found an SVF2N65D, although this one has only 2A continuous drain current.....
Same as CEF02N6 which isn't an MOSFET but an N-chanel enhancement mode field effect transistor... Although the datasheet shows its internal structure as an MOSFET.
Then there's the IRFS730 N-MOSFET, albeit this one has 400v drain-source voltage and +/-20V gate-source voltage. But at least its continuous drain current is 5.5A!
Would any of these work?

kotel wrote on 2025-03-02, 08:23:

Okay in my parts bin I also found an SVF2N65D, although this one has only 2A continuous drain current.....

Should be OK... if nothing else, for a test. I suspect it will be fine, though.

kotel wrote on 2025-03-02, 08:23:

Same as CEF02N6 which isn't an MOSFET but an N-chanel enhancement mode field effect transistor... Although the datasheet shows its internal structure as an MOSFET.

Oh, but that is a MOSFET. 😁
Can use this one too.

kotel wrote on 2025-03-02, 08:23:

Then there's the IRFS730 N-MOSFET, albeit this one has 400v drain-source voltage and +/-20V gate-source voltage. But at least its continuous drain current is 5.5A!

Yeah, that 400V V_ds voltage rating won't cut it, at least for Europe or elsewhere with 220/230/240V AC line voltages. In fact, if you look at IRF's datasheet for this part here, note that right at the end of page 1, it says "for US line input only". This is because US / North American AC is 120V, which rectified comes out to about 160-165V DC. But on ATX PSUs, the primary voltage is double that - typically around 340V, as you noted with the Tagan PSU... and that's without APFC.

So OK for the first two, nay for the IRFS730.

kotel wrote on 2025-03-01, 21:38:

Was hoping the tagan would get fixed firstly, but I take whatever I can fix.

We'll get to it 😉 , I promise I haven't forgotten. Just not always having time to dig in and reply to each thread.

kotel wrote on 2025-03-01, 21:38:

Aha, got it. I'll try to put an 35w 12v bulb on 5v and see if that does anything.

Try 20W first, if you have that. I find the 35W and 50W halogen ones (MR16) will sometimes trip PSU's short-circuit protection, due to very low resistance when cold. Regular auto bulbs seem to behave a bit better... though the higher power ones (40+ Watts) can also exhibit pretty low resistance. So just beware of that.

kotel wrote on 2025-03-01, 21:38:

May I know how to see if they are okay? Since those are generic purpose OST or teapo or whatever FSP used don't give info about normal ESR readings. And now I got told from TRW server that I should just see if the capacitance is in spec. If it is then keep, if it isn't then bin.

I just put them on my component tester and check for proper capacitance + to see if the ESR is at least somewhat in spec / reasonable.

For small 5x11 mm caps for example, sometimes it's OK if the ESR is as high as 1-4 Ohms, especially on old 85C general purpose parts. But anything more and I usually look to replace it. For bigger caps, I expect the ESR to always be less than 1 Ohm, preferably even less than 0.5 Ohms.

Yes, GP caps don't list ESR on datasheets, but that's OK, because I find the cheap component testers aren't really all that accurate to make good comparisons anyways (or at least mine isn't.)

The other thing I (vaguely) look at is the V_loss, as that can *sometimes* indicate a cap with high leakage current (mostly applicable to testing low-voltage caps 16V and under.) Anything under 5% is typically normal (or the tester isn't able to capture it.) But above, I put the cap in series with a 1-kOhm resistor and bring it up to the full operating voltage of the cap. After roughly a minute, the current going to the cap should be very low - less than 0.03CV (uA), as specified in most datasheets.

Now TWR is correct that the capacitance is probably the biggest factor to use for analyzing, at least when it comes to regular GP caps without ESR listed. After all, electrolytic caps are +/-20%, so that gives quite a variance on the capacitance.
That said, when the capacitance starts getting suspiciously close to the +20% spec and if the cap in question is from a manufacturer and series that is known for becoming electrically leaky (e.g. Teapo SEK, OST RLP, Ltec LZG, just to name a few), I usually go in favor of replacing it.

kotel wrote on 2025-03-01, 21:38:

Yep, will do that but pretty much most of them show 0 ohm esr on my lcr t7 and capacitance is within spec on the big secondary ones.

Yeah, i have the same "problem" with my tester. It's not really a problem, but more so a limitation on the accuracy.

kotel wrote on 2025-03-01, 21:38:

As for the transformator are there any ways I can detect an short without disasembling it or testing it on an live circuit? I should have some left-over from dead LC clones and one from an dead delta 500w PSU.

Nope, no way to know really, other than opening it... but at that point, you're bound to create more problems with the transformer (core). So just leave it as is and plug it all in (through the dim bulb) to see what happens.
If the transformer is bad, the bulb will glow / stay lit. Of course, the bulb could be lit due to something else being bad. The worst part is that you do have to re-check a lot of components after powering up the circuit every time. In regards to that, once you've replaced all parts of the 5VSB that you thought were bad, make detailed notes of *all* of the in-circuit reading of various components that are likely to blow (e.g. the small TO-92 transistor, various diodes, and etc.) Once you have that list of in-circuit readings (resistance and diode tests), you can then use it to compare if anything has changed after you've powered up the 5VSB and if you suspect something could have burned out again. This way, you may save some time removing some components from the circuit to test them. (For example, if a 1-KOhm resistor reads 500 Ohms in circuit when everything was checked at the time, then very likely if that resistor still reads 500 Ohms after powering up the circuit, it probably isn't bad.)

momaka wrote on 2025-03-02, 09:06:

Should be OK... if nothing else, for a test. I suspect it will be fine, though. […]
Show full quote

kotel wrote on 2025-03-02, 08:23:

Okay in my parts bin I also found an SVF2N65D, although this one has only 2A continuous drain current.....

Should be OK... if nothing else, for a test. I suspect it will be fine, though.

kotel wrote on 2025-03-02, 08:23:

Same as CEF02N6 which isn't an MOSFET but an N-chanel enhancement mode field effect transistor... Although the datasheet shows its internal structure as an MOSFET.

Oh, but that is a MOSFET. 😁
Can use this one too.

kotel wrote on 2025-03-02, 08:23:

Then there's the IRFS730 N-MOSFET, albeit this one has 400v drain-source voltage and +/-20V gate-source voltage. But at least its continuous drain current is 5.5A!

Yeah, that 400V V_ds voltage rating won't cut it, at least for Europe or elsewhere with 220/230/240V AC line voltages. In fact, if you look at IRF's datasheet for this part here, note that right at the end of page 1, it says "for US line input only". This is because US / North American AC is 120V, which rectified comes out to about 160-165V DC. But on ATX PSUs, the primary voltage is double that - typically around 340V, as you noted with the Tagan PSU... and that's without APFC.

So OK for the first two, nay for the IRFS730.

Okay, at least I have two attempts before I need to buy another MOSFET for this one 😁

kotel wrote on 2025-03-02, 09:26:

Okay, at least I have two attempts before I need to buy another MOSFET for this one 😁

If you use the incandescent bulb in series with the AC and *not* the toaster, you should have an "infinite" number of tries... more or less. 😁
Exceed the MOFET's Gate voltage and it would probably die quickly. If you have something like a 19-22V Zener diode, I'd suggest putting that in the circuit between the MOSFET's Source and Gate terminals (if there isn't one in the circuit already), to avoid the Gate getting destroyed, should the circuit misbehave.

BTW, I also added more info to my post after you made your last one here, so give that a re-read, as I think I answered a few questions there.

Okay with the small TO 96 transistor on the primary side 5vsb rail replaced (it burnt itself so badly one of the legs fell out) and the MOSFET inplace and without Q4/5 and J12 and the two pull up 1k resistors I get absolutely nothing on 5vsb and the sec. aux. rail has -0.27v according to my metex MM.... I have no idea what the hell is going on anymore
Transistor was placed in correctly. Although an small part of an zener diode fell out while I was desoldering that primary side 5vsb transistor. Unsure if it was inside the PSU or was it from the tagan.... Especially since there was an thin leg attached to one of the resistors without any signs of breakage and there's no marking for the diode where it "should" have went.

So the series dim bulb on the AC doesn't light up or stay lit (other than briefly when the caps charge up)??

If yes, make sure the primary caps are discharged first.
Then check these...
- Low-value Source resistor going to the MOSFET still OK?? (not open-cirucit... should be 1-2 Ohms value) *edit* this should be R16:
- high-resistance resistor(s) (typically 500+ KOhms to 2 MegaOhms) that pull the MOSFET's Gate up to one the big primary cap's (+) terminal also OK?? *edit* this should be R14 - 470 KOhms (check out of circuit)
- primary winding of the 5VSB transformer shows good low resistance in circuit (shows as shorted or close)?? (see final edit below)

The fact that the dim bulb on the AC is not going lit tells me one of the three listed cases above will not check out.
To help aid with what a 2-transistor 5VSB circuit should look like more or less (particularly one that uses a MOSFET), here's the one out of a HEC Orion HP585d:
https://www.badcaps.net/filedata/fetch?id=2041942
.

*EDIT*
The main primary winding of the 5VSB transformer is between one of the pins that you have connected with that yellow wire as the trace repair (the primary + DC bus.), and the pin right next to it, which goes to the Drain pin of your MOSFET. See if you have continuity between those two pins... and if yes, what resistance you get. Again, MAKE SURE THE PRIMARY CAPS ARE DISCHARGED... or your multimeter may go sparky sparky. 😁

Okay so the low value resistor R16 (which according to band markings (yellow,purple,gold,gold) is an 4R7) seems to not let anything through it. Even in diode mode I get nothing. I only have an bulky 22W one but it's too big to fit...
High resistance resistor R14 is good (680kOhms).
Primary winding of the 5vsb transformer with the two pins you've mentioned gives me 1.8ohms between them.

Any ideas before I go replace R16 and power it on?

kotel wrote on 2025-03-03, 19:35:

Any ideas before I go replace R16 and power it on?

Yes.
Replace R16 and also check the small TO-92 transistor (Q7)as well as R54 next to it before powering up the PSU.
Seems like R54 is fairly high value resistor (16 KOhm?? please check the color bands for me if possible)... and if so, then Q7 isn't likely to get damaged. On the other hand, on the schematic diagram of the HEC Orion HP585d I linked to above, that resistor is much lower value and can kill the small driver transistor.

Okay the R16 is replaced (somewhat smaller but it should be enough for testing) and we have 5VSB!
...if we skip over the fact that in the beginning it was going up to around 2v (me cutting power due to hearing "sizzling" noise may also be the culprit) and the fact that the whole PSU dies after a while (no more sec. aux. rail and 5vsb) and the bulb lights up like 1/4 to 1/3rd (strong orange color) of the full brightness..... And that "sizzling" noise got quieter but its still veeery suspicious.
Might be some other caps that got reformed? Might be something cooking up again but the bulb limits it? Who knows....
I am starting to be even more skeptical of the stability and safety of my PC parts after I fix this PSU up...

And now there's close to nothing on 5vsb (0.2v) and the bulb stays halfway/ 1/3rd of the full brightness when power is applied.....

kotel wrote on 2025-03-04, 14:53:

Okay the R16 is replaced (somewhat smaller but it should be enough for testing) and we have 5VSB! ...if we skip over the fact th […]
Show full quote

Okay the R16 is replaced (somewhat smaller but it should be enough for testing) and we have 5VSB!
...if we skip over the fact that in the beginning it was going up to around 2v (me cutting power due to hearing "sizzling" noise may also be the culprit) and the fact that the whole PSU dies after a while (no more sec. aux. rail and 5vsb) and the bulb lights up like 1/4 to 1/3rd (strong orange color) of the full brightness..... And that "sizzling" noise got quieter but its still veeery suspicious.
Might be some other caps that got reformed? Might be something cooking up again but the bulb limits it? Who knows....
I am starting to be even more skeptical of the stability and safety of my PC parts after I fix this PSU up...

If you replaced the electrolytic caps on the output (or at least checked that they have good capacitance and normal ESR), then it's not any electrolytic caps.
Since you did seem to get 5VSB, but albeit with the dim bulb glowing 1/2-1/3 of the way, that indicates parasitic power dissipation somewhere most likely. Though I wil note here that in rare cases I have seen 5VSB circuits not want to work with only a 40W or 60W incandescent bulb, but be OK with a 100W bulb. So use the 40W and 60W bulbs you have in parallel for when you do the testing again. DO NOT use the toaster as the series limiter, though, because that has too high of a current rating and will likely cause parts to blow up or burn. That aside, I think the parasitic power dissipation is most likely from either of the following:

1) snubber circuit (consisting of ceramic cap C8, HV diode D5, and resistor R61) might have a bad component. Most likely either C8 is leaky, D5 is leaky, or R61 is open. Check R61 first. Should be 22k according to what @DAVE86 wrote (and what I see from the pictures too.) If that checks out on resistance, replace C8 and D5 outright. Checking these two with a component tester simply won't do them justice. Both have to be tested with high voltage to check leakage current, which I don't think you have the tools & experience for quite yet to tell if they are bad. When replacing C8, make sure to use a ceramic cap with the same ratings - i.e. same capacitance and a voltage rating of 1 KV. D5 is also likely to be rated for 1 KV (or at least 800V) and be FR (fast recovery) type.

2) Transformer has shorted windings. Nothing you can do here, other than use a transformer from another PSU. Basically, just check all of the primary-side components again and replace those stated in 1) above. Then fire up the circuit once more and see if nothing changes in terms of behavior. If that's the case, then the transformer has gone bad likely. You can use one from another parts PSU if you have it (the 350W L&C??)

As for trusting the repaired 5VSB circuit... there are actually two ways to make it safer - one easy and one a lot more involved (but perhaps might be something you want to pursue if you want to learn more about electronics.) So here are these two ways:
1) Slightly reduce the voltage on the 5VSB circuit to something like 4.9V, then add a strong (1W-rated) 5.1V Zener diode across the 5VSB output. That way, if the 5VSB ever tries to go above 5.1V, it won'tkill your motherboard and attached hardware, but simply the Zener will short out and kill the 5VSB.
2) Modify the 2-transistor design into one with a chip. This will involve getting an offline SMPS IC of some sort and making a small PCB for it to drive the 5VSB circuit on the primary side. DM311 and VIPer 22A are IC that were widely used before and isn't too hard to make a board for (you end up reusing most of the components in the 5VSB circuit and simply remove the MOSFET to let the DM311 or VIPer 22A drive the 5VSB transformer directly.) Another one is any IC from the TNY2xx family. These are also fairly straight-forward to add as a modification. NCP1200/NCP1216 is yet another one, and this one reuses the original MOSFET in the 5VSB circuit.
Some relevant threads:
https://www.badcaps.net/forum/troubleshooting … -up-and-running
https://www.badcaps.net/forum/troubleshooting … er-22a-pwm-chip

kotel wrote on 2025-03-04, 18:56:

And now there's close to nothing on 5vsb (0.2v) and the bulb stays halfway/ 1/3rd of the full brightness when power is applied.....

Something's burned again on the primary probably.
Time to re-check everything - particularly the small TO-92 transistor and any small resistors and diodes connected to it. Also check the optocoupler's output side - should appear open-circuit when there is nothing driving the LED side, and show low-ish resistance when you drive the LED on the input side with a 5V source and something like 470 Ohm resistor in series. And while at it, also check the MOSFET, to make sure it doesn't have a partial or complete short between any of its terminals. The secondary sid of the 5VBS circuit should be fine, so nothing to check there.

momaka wrote on 2025-03-05, 11:45:

1) snubber circuit (consisting of ceramic cap C8, HV diode D5, and resistor R61) might have a bad component. Most likely either […]
Show full quote

1) snubber circuit (consisting of ceramic cap C8, HV diode D5, and resistor R61) might have a bad component. Most likely either C8 is leaky, D5 is leaky, or R61 is open. Check R61 first. Should be 22k according to what @DAVE86 wrote (and what I see from the pictures too.) If that checks out on resistance, replace C8 and D5 outright. Checking these two with a component tester simply won't do them justice. Both have to be tested with high voltage to check leakage current, which I don't think you have the tools & experience for quite yet to tell if they are bad. When replacing C8, make sure to use a ceramic cap with the same ratings - i.e. same capacitance and a voltage rating of 1 KV. D5 is also likely to be rated for 1 KV (or at least 800V) and be FR (fast recovery) type.

2) Transformer has shorted windings. Nothing you can do here, other than use a transformer from another PSU. Basically, just check all of the primary-side components again and replace those stated in 1) above. Then fire up the circuit once more and see if nothing changes in terms of behavior. If that's the case, then the transformer has gone bad likely. You can use one from another parts PSU if you have it (the 350W L&C??)

As for trusting the repaired 5VSB circuit... there are actually two ways to make it safer - one easy and one a lot more involved (but perhaps might be something you want to pursue if you want to learn more about electronics.) So here are these two ways:
1) Slightly reduce the voltage on the 5VSB circuit to something like 4.9V, then add a strong (1W-rated) 5.1V Zener diode across the 5VSB output. That way, if the 5VSB ever tries to go above 5.1V, it won'tkill your motherboard and attached hardware, but simply the Zener will short out and kill the 5VSB.
2) Modify the 2-transistor design into one with a chip. This will involve getting an offline SMPS IC of some sort and making a small PCB for it to drive the 5VSB circuit on the primary side. DM311 and VIPer 22A are IC that were widely used before and isn't too hard to make a board for (you end up reusing most of the components in the 5VSB circuit and simply remove the MOSFET to let the DM311 or VIPer 22A drive the 5VSB transformer directly.) Another one is any IC from the TNY2xx family. These are also fairly straight-forward to add as a modification. NCP1200/NCP1216 is yet another one, and this one reuses the original MOSFET in the 5VSB circuit.
Some relevant threads:
https://www.badcaps.net/forum/troubleshooting … -up-and-running
https://www.badcaps.net/forum/troubleshooting … er-22a-pwm-chip

kotel wrote on 2025-03-04, 18:56:

And now there's close to nothing on 5vsb (0.2v) and the bulb stays halfway/ 1/3rd of the full brightness when power is applied.....

Something's burned again on the primary probably.
Time to re-check everything - particularly the small TO-92 transistor and any small resistors and diodes connected to it. Also check the optocoupler's output side - should appear open-circuit when there is nothing driving the LED side, and show low-ish resistance when you drive the LED on the input side with a 5V source and something like 470 Ohm resistor in series. And while at it, also check the MOSFET, to make sure it doesn't have a partial or complete short between any of its terminals. The secondary sid of the 5VBS circuit should be fine, so nothing to check there.

Case 1 shouldn't be the culprit. C8 was replaced and doesn't show a short between the pins, D5 still shows around 500mV in bias and nothing in reverse bias. R61 is the one I couldn't locate using the pictures DAVE86 provided and my own PSU (the pcb got scorched too much). I don't think I have any PS106R's to cannibalize from other PSU's but I'll check....

Case 2 also shouldn't be the culprit. Transformator has no pass-through from secondary side to primary.

As for the diode way of making 5vsb safer it's a veery good tip you told me 😀 I am pretty sure I'll use it on my other DEER/LC clones I had just to be less paranoid about the boards diying.
I think I could also use it on 3.3v/5v/12v too. Although then I'd need to make if one diode blows others disconnect too to prevent further damage to the board.

As for the opto coupler I will need to check once I get more free time (my school is killing me with the tests...). It's from that LC 350w PSU which after recapping the 5vsb had correct voltage. But everything can happen with these PSU's....

kotel wrote on 2025-03-05, 16:09:

Case 1 shouldn't be the culprit. C8 was replaced and doesn't show a short between the pins, D5 still shows around 500mV in bias and nothing in reverse bias. R61 is the one I couldn't locate using the pictures DAVE86 provided and my own PSU (the pcb got scorched too much). I don't think I have any PS106R's to cannibalize from other PSU's but I'll check....

Just outright replace D5, because a leaky one will still measure around 500 mV in forward direction and seemingly nothing in reverse... until you hit it with high voltage. It's a rare failure, but it does happen.
R61 you can see in the picture below, right under small TO-92 tramsistor Q7:
download/file.php?id=212155&mode=view
^ It's a pretty large resistor, 22 KOhms.

kotel wrote on 2025-03-05, 16:09:

Case 2 also shouldn't be the culprit. Transformator has no pass-through from secondary side to primary.

No no, these transformers will *NEVER* have pass-through from primary to secondary or vice-versa. That's the whole idea behind them - to completely isolate the secondary side from the primary.
However, what can happen is there can be a few turns shorted on the primary winding, drawing excess power as those windings heat up inside the transformer. This can definitely cause the half-dim bulb you were seeing.

Of course, the half-dim bulb could also be caused by the MOSFET not getting fully turned On and Off.

kotel wrote on 2025-03-05, 16:09:

As for the diode way of making 5vsb safer it's a veery good tip you told me 😀 I am pretty sure I'll use it on my other DEER/LC clones I had just to be less paranoid about the boards diying.

Well, if you do, just make sure to turn down the voltage on the 5VSB circuit a little, since many circuits will briefly go above 5.1V upon power up (in a really quick spike that you cannot measure with a MM.) So by turning down the 5VSB to 4.9V or less, that should make it safer from the Zener diode to blow during normal power up.

kotel wrote on 2025-03-05, 16:09:

I think I could also use it on 3.3v/5v/12v too. Although then I'd need to make if one diode blows others disconnect too to prevent further damage to the board.

I wouldn't. These rails already have OVP and UVP, so not much point to it.

Anyways, let me know what you find when you have more time and good luck with your exams! 😀

momaka wrote on 2025-03-05, 17:18:

Just outright replace D5, because a leaky one will still measure around 500 mV in forward direction and seemingly nothing in rev […]
Show full quote

kotel wrote on 2025-03-05, 16:09:

Case 1 shouldn't be the culprit. C8 was replaced and doesn't show a short between the pins, D5 still shows around 500mV in bias and nothing in reverse bias. R61 is the one I couldn't locate using the pictures DAVE86 provided and my own PSU (the pcb got scorched too much). I don't think I have any PS106R's to cannibalize from other PSU's but I'll check....

Just outright replace D5, because a leaky one will still measure around 500 mV in forward direction and seemingly nothing in reverse... until you hit it with high voltage. It's a rare failure, but it does happen.
R61 you can see in the picture below, right under small TO-92 tramsistor Q7:
download/file.php?id=212155&mode=view
^ It's a pretty large resistor, 22 KOhms.

Okay so I have no idea why but the R61 was... 33 ohms? What the heck did I even do here.... Replaced it with 22k ohm as you said but same results. 5VSB is minimal (0.3-0.5v range) and the bulb still lights up 1/3rd of the way there.

momaka wrote on 2025-03-05, 17:18:

No no, these transformers will *NEVER* have pass-through from primary to secondary or vice-versa. That's the whole idea behind t […]
Show full quote

kotel wrote on 2025-03-05, 16:09:

Case 2 also shouldn't be the culprit. Transformator has no pass-through from secondary side to primary.

No no, these transformers will *NEVER* have pass-through from primary to secondary or vice-versa. That's the whole idea behind them - to completely isolate the secondary side from the primary.
However, what can happen is there can be a few turns shorted on the primary winding, drawing excess power as those windings heat up inside the transformer. This can definitely cause the half-dim bulb you were seeing.

Of course, the half-dim bulb could also be caused by the MOSFET not getting fully turned On and Off.

Aha. Got it. An thermal cam would be ideal but it's way out of my budget....

momaka wrote on 2025-03-05, 17:18:

Well, if you do, just make sure to turn down the voltage on the 5VSB circuit a little, since many circuits will briefly go above 5.1V upon power up (in a really quick spike that you cannot measure with a MM.) So by turning down the 5VSB to 4.9V or less, that should make it safer from the Zener diode to blow during normal power up.

I think putting like an small 1k/10k resistor before the zener should do the trick, right? Can't think of any other way....

All of the smaller caps on the primary side and in the 5vsb circuit were replaced (one OST even had 10ohm esr!). What's weird is that I don't get any readings on source of the MOSFET to drain out of circuit (reverse and normal bias). Not sure if this is normal. I had made sure there weren't any signs of it being burnt, and there weren't any.
The small TO 96 transistor next to it (C945) seems to measure fine. No shorts between legs in diode mode.
As for the PS106R diode I can only find FR104, although this one, according to my knowledge, would be 1A 400V instead of the 600V needed... I'll try to look on the other PCB's I had laying around and maybe I will find an 600V FR diode.

Okay for the PS106R I found an FR107, which should be 1A 700V. It's legs are a tad bit short but I think soldering some copper wire to one pin should be okay. At least for testing.
As for the opto coupler I'll just replace it with a new one. Have like 20 of them I forgot about😆

Okay I've replaced that diode with the one above and..... 5VSB IS BACK!

...but again without a catch.. It's waaay out of spec (5.8v!).... I have no idea what else now is screwed. Caps were replaced on both the primary and secondary side... At least the sec. aux. rail is okay (10.20v).

You need some big enough load on the output for voltage to get in spec, i.e a few LEDs or something like a 100 ohm resistor. Unloaded SMPS outputs are rarely if ever in spec.

1k ohm (for good measure) resistor on 5vsb to gnd. Got it.

Tiido wrote on 2025-03-06, 14:53:

You need some big enough load on the output for voltage to get in spec, i.e a few LEDs or something like a 100 ohm resistor. Unloaded SMPS outputs are rarely if ever in spec.

Actually, on a second thought this doesn't sit right with me...
The PSU after replacing all the bad parts for the first time had spot on 5vsb (~5.17v). Then something screwed up fully and now its 5.8v after replacing the PS106R (mine started with FR10 and the rest was burnt away) with FR107. I don't think its gonna be the voltage increase by 100v.

It is possible there's a damaged voltage reference component in the feedback loop or maybe there is a need to adjust one of the feedback resistors due to different replacement components etc.