WTH is wrong with this SMPS?

[Eliminateur]

so, like i said in the other thread, i'm with this SMPS from hell that's driving me nuts....
output oscillates with a ~1.5v~ amplitude and a semiperiod of 80mS, here's the vid:

(with background music and annoying phone!)
80mS/div, .5V/div, AC coupled

i've changed main caps, output caps, bridge(fried the original hehe), output is the same...., the vid quality sucks, but after the output "peaks" it lowers in a triangular stairstep way.

i'm attaching a solder side of this smps from hell.

any ideas?

[Zad]

The YouTube link doesnt work if you use the forum's fancy link button thing. Just paste the URL and it will automatically handle it ok.

Do you have the PSU loaded at all? SMPS can be very funny creatures if you don't load the output, and oscillations like that are a typical symptom.

[Eliminateur]

fixed link
output is unloaded, this kind of smps bricks handle no-load without problems(when they work ok).
in fact, this problem appears regardless of load and that makes it worse/untraceable...., which is why it's in the shop to begin with

worse is that the SMD controllers are unreadable, only the output side one i could barely discern it it's a STmicro 103... something, searching the part all i could come up with is an TSM103AI dual opamp with reference, and i'm pretty sure that can't be right...
it also features 2 stacked optos(white silkscreen on top).
the PWM controller is unreadable

[Psi]

Maybe check out the components involved in the feedback.

Is the top or bottom of this waveform the correct output voltage?  ie, is this waveform above or below the correct output voltage?

I do wonder if an opto with a damaged led might result in something like this.
You know how normal leds that have been damaged tend to illuminate and then start flickering. If a feedback opto led did that you might get that kind of peak and then stepdown as it flickers.

[Eliminateur]

Psi, hmm i haven't checked both optos, and since it's all smd it's a pain to map
i haven't checked it DC coupled to see if it's mid or top down, but i venture to say that the peak is the nominal (19.5v), because if you measure by DMM, the value never changes (and the pilot led that is attached to it doesn't show any change as well).
but when you plug it to a notebook, it starts doing weird shit

[Eliminateur]

nope, they aren't the optos.
I lifted the opamp (TSM103/A) and switched to one i scavenged from another dead brick but now it works worse(i'm pretty sure that replacement is fried).
Now i have to find how to get one of these opamps...
oh and if anyone knows a site that has Dell ac adapter schematics, send it, the usual ones i use for laptop schematics don't have the ac adapters...

[Zad]

I think I would be having a go at drawing my own schematic for this. Looks like it is probably only double sided and not too bad. From what you have said so far, it sounds like half the PSU has been cooked.