Asus M2A-VM motherboard signs of erosion close to 4pin 12V CPU PSU connector?

[eneuro]

Hello,
I've tried powerup this Asus M2A-VM motherboard after realized I've Nvidia CUDA capable old GeForce 9500 GT graphics card which was used with this  MB before, but without success 

Only 32 CUDA cores on this GeForce 9500 GT 1.1 CUDA capable GPU, but I'm interested if GPU random numbers generation even on this 32 cores GPU will be faster than on its pure dual core AMD CPU 

Unfortunatelly, it looks like there is problem in CPU not working at all, while there are solder paste erossion signs close to 4pin 12V CPU power connector and it doesn't sounds great.

Additionally, some unpopulated power mosfets pads coroded too-it is interesting there are 3 such unpopulated pads on this ASUS, but Made in China MB 


I found deplated CMOS battery, but probably MB should beep connected speaker when memory slots are empty, green power led on MB is on when connected PSU and when tried turn on using PWR pins CPU fan turns on BUT ONLY when this bad looking 4pin 12V CPU connector is not plug in-when plugin CPu power, even with empty memory slots no PC speaker beeps, nothing, looks like completly dead MB?
I suspect this CPU power PCB area, but will try tomorow with fresh new CMOS battery.

Do you have some "patents" to help clean such nasty solder paste (green) erosion signs on PCB?
Probably this MB is lost-didn't check caps, but this CPU power connector area looks not so great-maybe not worth to mess with caps before futher examination of this 4pin 12V CPU connector area 

Note: CPU fan was of course connected during power up experiments-it is now disconnected to make photos only...

[Ian.M]

Its probably totally FUBARed.  It only takes a little bit of corrosion down a via on a multi-layer board to make it BER.

[eneuro]

Yep, probably the only hope is that those  N-channel gate level 50A 9.5mOhm ( http://www.cheertech.com.tw/Niko-sem/Data%20sheet%5CP0903BDG.pdf ) mosfets gates are not properly switched while its  RT9618A Synchronous-Rectified Buck MOSFET Driver ( www.richtek.com/download_ds.jsp?s=352 ) control signals are messing on this eroded PCB.
Maybe after removing this erosion garbage and dust things will look better and at least will be able check those mosfets for any shorts, gate signals, etc...
RT9618A typical application circuit from linked datasheet with its pins:

Better than nothing to see voltage levele on those (I believe RT9618A ot RT9619A) on my dead looking MB 

[Guni]

You can try to clean - isopropanol, pcb cleaner, toothbrush etc.
But this kind of damage on multilayer mobo.... you know not so many chance for success. Probably some vias are corroded.
From my experiences with flood laptops, if there is corrosion is to late. Few days or weeks after recovery their die because is it impossible to clean vias, place under ic and so on. It works like galvanization etching.
Is also risk of damage other components like processor or graphics card.
This is quite old mobo so...

[eneuro]

Probably some vias are corroded.

Dificult to find visible vias with erosion, but the worst affected RT9618A mow partialy cleaned using paintbrush and alcohol resign 

A little bit hot air and after final cleanup powerup with CMOS already reset and brand new VARTA CR2032 inserted.
Moment of truth comes 

[eneuro]

Nope, Something switches PSU off-maybe some kind of short protection in RT9618A, when 4pin 12V PSU in pluged into connector.

It looks like there are three similar Synchronous-Rectified Buck MOSFET Drivers which fits into circuit concept above: two in parallel bottom Q2 and Q1 without unpopulated one.
However, this clossest to this 4pin 12V psu has pins 5 6 7 eroded in RT9618A (marked red), so no chance to work properly.

Only gate Q2 pin 8 is fine on top side, checked also pin 4 Vcc is OK (probably pins 1 2 3 4 are fine).

Probably all those RT9618A driven Synchronous-Rectified Buck MOSFETS are designed to work in parallel to provide decent power to CPU.

BTW: While this mobo is dead thinking about tricky thing 
Desolder last good pin 8 in eroded RT9618A and.... bend this chip on probably good pins 1,2,3,4 ... verticaly and try solder 4 wires to bypass PCB 
It looks doable when watching camera photo images, but those RT9618A ICs are very tiny, so probably mission imposible but... probably last chance to wake up this mobo of course for fun because it can't be reliable fix 

[ptolemy9]

I inherited the same board with the same problems and symptoms.. I don't have any corrosion on mine. I have tried blind-flashing the bios with no luck. I'm looking at a possible memory problem/component,..since with no memory in the slots, the CPU will at least warm up and the machine stays on(?) Will post if I succeed with anything.