Here what I've learned when I fixed few mobos while ago.
The obvious & most popular problems, sick caps with low ESR. This should be the 1st one to suspect as you mentioned above. I assume you got an ESR meter handy.
If I were you, I would check 1st all those recap-ed job done by the previous owner such as the solder jobs and other tiny "accidents" that probably happened like cutted/grooved trace and etc.
Focus and watch out for those power caps that supplies the main memory, usually they used cheaper one compared to other caps like for CPU's VR. I have observed on few mobos, that they still can even beep or lighted out the mobo's led even with bad caps in it, although no POST.
In your case, too bad you never seen that mobo when it was working fine.
Two cases I've founded on hissing mobos when working, that when died, I revived them perfectly just by re-soldering or re-heating all those inductor's lead.
Dunno what exactly the theory behind it, I can only speculate that those vibrating choke was loosening it's own soldering joints and slowly degradring the joints, but this makes sense for me though !
Or prolly I was lucky !
Dead power mosfet from CPU VR, solved 1 mobo that had it, and it was really a "minor" or really small buldge on it's plastic body but not obvious just by taking a glance on it, I was lucky to spot it when inspecting the mobo under Luxo light with magnifier. Without it, I could have missed that. Solved it by cannibalized mosfet from other same broken mobo.
Now bout your symptons :
Super Nade said:
Zeners are o.k. Zeners have a forward drop voltage of 0.7~0.5V. All the Zeners checked out fine.
Zener works on their "reverse" bias voltage, while their "forward" voltage should behaved like normal silicon diodes, so it abit difficult to test them but most of the times they're quite reliable.
Also most mobo now are using 3.3 Volts or 5 Volts zener, except from some strange serial RS-232 port area. Also some of them are schotky diode with lower forward bias voltage like 0.3 to 0.4 volts.
Super Nade said:
No PS_ON low signal. This is a big problem.
It is supposed to drop to 0 momentarily for the PSU to turn on. Now I have to figure out what is causing this.
What did you use on measuring those PS_ON signal ? Scope ?
I doubt when using DVM you could catch that "momentary" state / spike from logic low and back to high that could happened in mili seconds.
These sequence is normal when the mobo logic detects some error, it tries to switch on (PS_ON=low) but because of some errors, it switches off back (PS_ON=high). That PS_ON signal is mostly controlled by the IO chips, popular one like Winbond or ITE.
Imo, most of the problems with the mobo are the analog parts, but when it comes to digital part, I just give up !
You should take these with grain of salt, cause I'm not mobo repair expert, its just I got all of this from learning by doing in the past.
Hope this contributes to your mobo reviving mission !
For those who are wondering, PS_ON is a low logic signal telling the PSU to turn on. The measurement points are between the green wire in the 24 pin connector and any black wire. Mine stays at a constant of 4.7 V. It is supposed to drop to 0 momentarily for the PSU to turn on. Now I have to figure out what is causing this.
