I would measure these two regulator output voltages. You can do it without CPU. Just check that Vcore jumper is set to something, like 2.1V.
Orange = Vcore
Yellow = VIO
Green = 5V reg input

Thank you for the tip, I can do it straight away I just have 2 dumb questions:
Orange arrow measured v1.42 (it started from v0.8 and slooooooowly went up).
Yellow arrow measured v1.47.

That with v2.1 voltage setting, same values if I use an higher vcore setting.

Where could be the problem? Capacitors?

Orange is Vcore and 2.1V is OK. But I was wrong with the CPU, for VIO (yellow arrow) measurement CPU must be inserted. Otherwise they are about the same value.
Also try to change GND point if value looks strange.
What CPU are you using?

fool wrote on 2023-02-04, 13:39:

Orange is Vcore and 2.1V is OK. But I was wrong with the CPU, for VIO (yellow arrow) measurement CPU must be inserted. Otherwise they are about the same value.
Also try to change GND point if value looks strange.
What CPU are you using?

What gnd do you advice? I'm using the ps2 ports metal case.
I use a Pentium 133 v3.5 with the cpu installed they both measure v0.82x
So apparently the voltage goes down as the load goes up?

Yes, I got that. USB/PS2 port cover is fine.

That pentium is common rail and interconnects Vcore and VIO.
For what it sounds, Vcore supply is possibly missing and draws VIO down with it.
It would tell more if you have dual rail CPU f.ex. MMX or K6(-2).

fool wrote on 2023-02-04, 14:01:

Yes, I got that. USB/PS2 port cover is fine. […]
Show full quote

Yes, I got that. USB/PS2 port cover is fine.

That pentium is common rail and interconnects Vcore and VIO.
For what it sounds, Vcore supply is possibly missing and draws VIO down with it.
It would tell more if you have dual rail CPU f.ex. MMX or K6(-2).

I've tested a k6-2 350 (v2.2)
Orange arrow measured v0.359
Yellow arrow measured v1.432

Also asus p5a as +3 voltage selection (for dram, chipset, agp i\o cpu buffer) which is actually jumpered for v3.5 (but it can be higher).

Regulator inputs should be 5V, measured from transistor tabs. Green arrows in edited picture.

Yes it is v5.04x for both... I think the problem it's that the cpu doesn't get any power?

Those two PWM controllers HIP6013 near both transistor pairs switches transistor to generate selected output voltage. Clearly something is wrong with the control. Look for missing or weird looking components around them.

I haven't heard of poor capacitors in P5A? Of course doesn't mean it's impossible....

Uhm it seem everything is in order

I would check these next, just to be sure they have supply voltage etc.

HIP6008
-pin 14 (VCC) 5 or 12V
-pin 10 (PGOOD), may not be connected

HIP6013
-pin 14 (VCC) 5 or 12V
-pin 6 (enable), should be above 1V regulator to be enabled

I missed your answer about the capacitors, I can confirm that noone of the socket7\ss7 boards I had ever need a recap but they are usually the first thing which is suggested to change.

Thank you very much for your help here are the values, i'm still using the k6-2 cpu:

HIP6008
-pin 14 (VCC): 12,17
-pin 10 (PGOOD): zero (probably not connected as you suggested)

HIP6013
-pin 14 (VCC): 12,17
-pin 6 (enable): 5,027

If I were to take a wild guess, I'd say that you're probably dealing with a shorted transistor and / or a shorted voltage controller.

bloodem wrote on 2023-02-04, 17:22:

If I were to take a wild guess, I'd say that you're probably dealing with a shorted transistor and / or a shorted voltage controller.

That was my first thought when I noticed the cpu didn't get warm...
Any suggestion how to find it? I tried to keep the motherboard on for some minutes and then touching components to check if there was anything hot but I found nothing.

Well, you need to check resistances. Look at the voltage controller's datasheet and check all pins for potential shorts. If one pin turns out to be shorted, trace the connection and see where it leads. If it's connected to one of the MOSFETs, then... don't even bother with identifying if the MOSFET or the controller is to blame: just replace both (I would replace both MOSFETs and the controller if this turned out to be the case).

Once you remove them, you can then see which one was shorted. They usually tend to fail simultaneously (either the MOSFET kills the controller or vice-versa). 😀 Either way, I would not risk it and would replace them all with new parts (after making sure that the short is gone, of course).

Uhm i'm sorry but i'm not so savy when I have to use the multimeter, so I need some clarifications.
Should I touch every pin with the red probe while the black is on the ps2 ports case and check resistance?
That with the motherboard turned off?

Those are taken from the datasheets:

Nemo1985 wrote on 2023-02-04, 17:41:

Uhm i'm sorry but i'm not so savy when I have to use the multimeter, so I need some clarifications.
Should I touch every pin with the red probe while the black is on the ps2 ports case and check resistance?
That with the motherboard turned off?

Pretty much, yes! 😀
The motherboard shouldn't be connected to power when you do these tests.

Shorted transistor can do nasty things like lead 5V on the Vcore rail. You are lucky it's not that in this case.

Good luck hunting.

Values for HIP6008 values from pin 1 to 16:

1>3k
210.78m
323.48k
40.3
57.53k
64.881k
710.85m
811.90k and then goes up
9
100.2
119.70m
1255.06k
137.91m
1438,06k
1538,06k
169.81m
17First time Goes up to 55k and then down, remeasured gave 2m and goind down

Crazy means that the multimeter start with a low value and then it goes up up up up, while the other are like stable value (they may change of 0.01)

HIP6013:

110.30m
26.13km
311.26m
40.L
53.605k
613k
70.3
8453
97.87m
1038,05k
110.L
120.L
130.L
1438,05k