E15 Gen2 Intel i5 - doesnt start & jumping voltage

[danny34]

Hello,

I have an E15 Gen2 with Intel i5 and need your help.

It doesn't start and here's what I've already found out:

All voltages are available.
However, the PU8 (RT3612) does not switch the mosfets for +VCCIN.

On the PU8 I measured 3.3V at pin 23 (VRON), but it drops to 0V every about 3 seconds. This voltage comes from the IT8227E.
VCC from IT8227E has exactly the same symptom. Every 3 seconds the voltage briefly drops to 0V.

In the end I saw that +1.8VALW (VCC from IT8227E), +0.6VS, +2.5V_DDR & +1.2V drop to 0V every about 3 seconds for about 1 second.

I now suspected the PU4 (LV5115) of being defective and replaced it. That wasn't successful.
VCC, PVCC & VSYS on PU4 all have constant voltage.

Anyone have an idea where I can look?

Schematic: HERE Post #9

[mon2]

Go upstream on the power rails. Who is stable?

Is the voltage to ground @ V9B+ stable or also power cycling ever 3 seconds?

This is the primary power rail to power the logic board. It may be that the power cycling is occuring due to an excessive current draw or some regulator believing that there is a higher than normal current draw so the regulator is in chirp mode (on/off/on..). This issue can surface if there is a break in the current sense loop of the power rail being monitored.

Start with the main power rail for the measurement. Perhaps the bios is corrupt and may be worth reflashing but always make a backup of the original before starting. The dedicated bios section should be able to supply a known good bios.

[danny34]

the 9V is stable, as are all other voltages such as +3VALW or +5VAWL. It only affects those described above.

I just swapped BIOS chip UB4 from my other working E15 G2 i7 board with the latest BIOS version. Same game

[mon2]

Is the Vcc to the EC (IT8227E) stable? Is the EC heating up - more than normal? The EC also has its own firmware driven by the SPI bus where the EC is the bus master. If you have the spare working board, consider to swap out the flash device @ U2 (Bono?) used by the EC for testing.

[danny34]

I haven't changed anything, but I no longer have a V9B+. Also 4ohm to ground...

strange

where do the V9B+ come from?

20.7V is still available until after the QB6 and F6101, from there it is called VINT20_IN.

[mon2]

V9B+ is from the charger IC. It is taking in the ~20v from the power adapter and lowering to this voltage. The resistance to ground is very much of concern.

Remove all power.

Meter in resistance mode. Check the resistance from this same V9B+ rail (place one meter lead here) and the other meter lead onto the VCORE cpu power rail. We are checking the resistance between these 2 points. If the resistance is low, then the high side mosfet on the VCORE rail has leaked and injected this very high voltage onto the CPU.

Share the resistance measured between these 2 points of interest.

[danny34]

crap. Resistance between V9B+ and VCCIN (at PL12 and also PL13) are 4ohm. Both highside mosfets (PQ16 and PQ19) have a diode test at 0.0V, no matter in which direction...
Not good news, right?
thanks for your help

[mon2]

Yeah, I am afraid not good news. Respectively, this high voltage must have hit the CPU vcore rail.

[trepachka]

In case this still interest you: I had a similar situation with this board. There was a small spill close to PU8. After cleaning the spill, the motherboard kept resetting every 5 seconds or so.

Check all sense resistors around PU8, make sure they all measure 2.2 ohms. In my case, one of the resistors connecting 5V line (pin 29) was showing mega ohms. So, I figured there was quite amount of current passed thru it to make it fry.
After replacing it, no change but 5V line going into PU8 was restored. I replaced PU8 from a donor board and the motherboard is kicking again.