Dell model: H235P-00 - after repair and computer does not start

I have a problem with the power supply for my Dell Optiplex 780 SFF.
I repaired the power supply because the capacitor on the original one flew, for the sake of principle I replaced all the capacitors and the fan, because it rattled a little. I put it together - everything works on the table, when I put it in the computer, it does not start, the LEDs on the housing indicate error 1-3- and the Power button flashes - i.e. according to information in the computer manual is an error in the power supply, motherboard or peripheral device. The computer is operational, it wakes up on a different power supply.

Voltage measurements on the power supply:
connected to the network:
PS_ON 4.97V; SB 5.05 V
Start without load:
3.37 V; 5.22V; 11.54V; -11.91V; SB 5.05 V; PG 4.99 V
with a load of approx. 50%
3.38 V; 5.11V; 11.85V; -11.91V; SB 5.04 V; PG 4.98 V

So everything is normal, even exemplary. By holding down the test button on the power supply, the computer starts up. I assume that the motherboard is doing some kind of self test of the power supply, because when it is turned on, the fans start for a moment when it is turned on, and on the fixed one, it is not. I don't understand any of this, what doesn't suit him? Could I have missed something?

Edit:
I was digging a bit on the net and found this information:

Quote:

"Indicates system has power, but the POWER_GOOD signal is not yet active".

This is a clue, if the PG appears after more than 500ms, it may not allow the power supply to turn on. The problem is that I do not have how to measure this time, I will not connect the power supply tester, because this model has a mini ATX connector

So I will answer myself later, because I had a great idea to connect 2 meters, one to 12V, the other to PG and record a movie, then play it frame by frame and see what is happening. I know that multimeters have a long delay, but what came out gives a certain view on the matter. So it looks like this (seconds: hundredths)
00:00 - Power ON, 12V -> 10.25V, PG 0.0V
00:25 - PG 4.76 V (there is a response of 250ms, it's not bad, but ...)
00:56 - 12V -> 11.53 V (voltage stabilizes)
00:75 - PG suddenly drops to 0.47V
01:12 - PG goes down 0.09 V
01:51 - PG starts to rise by 1.47 V
01:92 - PG stabilizes at 4.95V
It seems to me that this explains why the power supply does not pass the self-test in the computer, return to the workshop, I had to skip something.

I would start by measuring PG in a good power supply. See how he behaves.

What are your layouts? integrated into this PSU.
Such as in this topic from this forum Link

What IC as a watch dog?

Supervisor: WT7507
H&L side driver: CL2181
AC / DC: TNY280PN

And now a curiosity. I soldered an adapter and connected the power supply to the tester.
First, a working power supply that works in the computer:

PG time 0 ms ???
PG voltage measurement 4.68 V

And here's a power supply that won't work:

The only thing that puzzles me is a little too low voltage of 12V, although it is within the norm and under load it grows as I mentioned earlier.

Read on Link

I read it, but it clearly went up in smoke there. To be sure, I checked all the elements in question - they are ok.

Maybe the capacitors you replaced? Do you have options to put them back?

Damage to either the power supply or the WT7507 SOP16 (himself or the RC parts in his vicinity)

Quote:

PS_ON # Signal Characteristics
VIL, Input Low Voltage min. 0.0V, max. 0.8V
VIH, Input High Voltage min. 2.0 V, max. 5.25V

* Your signal is OK (usually it is 4.3V) below 3.7 >> problem

Quote:

one of the most important protections (the OTP or Over-Temperature Protection) is only supported by two monitoring ICs, the SITI PS223 and the Weltrend WT7507. However, there are other ways to implement this protection and trigger one of the other protections in a monitoring IC, in order to shut down the PSU.

About WT itself briefly on Link
and Link
Comparison of available systems "watchdog = supervisor" under Link

Read the datasheet of that other system Link__ SITI PS223
- compare pin "y
- * pin.9 PS223 OTP Over temperature protection input

* between IC PS223 can be opto. (1..or 2) controlling the operation of the transistors of the system start (impulse transformer on input)

Quote:

The Fairchild SG6520 is a direct replacement (pin2pin) for the PS223 so have a couple on order from Mouser

SG6520 Link worked at Enermax PSU
- there is a lot of info about him.




And on the PSU ATX 250W supervisor system CCG8010DX
- CG8010DX (= WT7520).