Dolphin, the GameCube and Wii emulator - Forums

this has never happened to me before.

a precursor to this is that i do alot've stress testing on my pc. my motherboard's northbridge has always ran hot (60'c at idle, and up to 75' at full load, even at the lowest voltage setting available), but the position of it was really bad and there was no way to fit any type of cooler on it with my h70 radiator in the way. so, i have always been weary about this board and wondered when something bad was going to happen. i was pretty much "waiting for the day" so to speak.

the other day i was testing some tighter ram settings and had prime95 running for a couple hours before i decided to take my dog on a walk.
when i got back i immediately smelled toxic fumes all in my house. i knew right away it was my pc. i walked into my room and saw that the light on the case was on, but no fans or anything were running. as if it wasn't already obvious enough that something was fried.

my motherboard had shutdown temperatures set (@ 60'c), so whatever happened somehow managed to bypass that safety feature. only my cpu was overclocked, and i barely had extra voltage going to it (like .15v).

the toxic smell in my house was too strong for me and my dog so i had to open all the windows and let it air out for awhile. 30 minutes later it still stank to high heaven in my room. for shits and grins i plugged the ps back in and hit the power button. in the 2 seconds my pc attempted to start up i heard 15 sparks and black smoke/dust shot out of my upper front exhaust fan. well... i was already running late to get to a show i had to play and didn't have time to mess with it.

so, i took it apart today. after removing everything from the mobo and removing it from the case i was able to access the northbridge heatsink. the northbridge heatsink on my board has a heatpipe that runs to another heatsink that lies over about 12 little voltage regulation chips. as it turns out, i think my chipset itself wasn't running hot. i bet these little bastards were actually sending heat back to my chipset. i guess is i could have glued a waterblock to these little guys, but i always assumed it was my nb running hot.

[attachment=5675]

what mainboard?

that would indeed be transistors
transistors can get hot. however i doubt they would have such a critical failure...EVER
and even if they did i doubt the mobo would still work

that, and it looks more like a small explosion happened (see the smd capacitors & resistors) instead of heat issue
also, i think a few capacitors might have popped as well (you know, those tower like components); or did those for once stay stable? o.o;

EDIT : also, i think its pretty clear what happend. mobo got unstable, it crashed and burned

(03-11-2011, 06:07 PM)DacoTaco Wrote: [ -> ]that would indeed be transistors
transistors can get hot. however i doubt they would have such a critical failure...EVER
and even if they did i doubt the mobo would still work

that, and it looks more like a small explosion happened (see the smd capacitors & resistors) instead of heat issue
also, i think a few capacitors might have popped as well (you know, those tower like components); or did those for once stay stable? o.o;

EDIT : also, i think its pretty clear what happend. mobo got unstable, it crashed and burned

no, thats a VRM (Regulator) the debris across the transistor is just bits of the vrm after it blew up.

it was an am3 dfi board (m3h5). my pc was completely stable. like i said, it always had high heat issues... i tried everything... even ziptied a fan on the nb/icm cooler and it wouldn't lower the heat. the board was rated for 1600 memory, so it shouldn't have overworked it too much with my ram at 1600 and my cpu slightly overvolted. i could get it down to like 70'c at full load if i had everything clocked down, but that wouldn't have been any fun.

before buying this board i only had found good reviews. then, after buying it and having problems i googled it and found just shittons of posts where people were crying about it having overheating issues and as a result, crashes.

the truth is, this mobo could have lasted longer, but i had prime95 running anytime i wasn't playing a game. the fact that the icms and the nb chip were always so hot just kinda pissed me off. i thought for awhile that maybe it was just reporting hotter than it really was since it would be at like 75'c for ~72 hours of prime95 and still working. so, i suppose if it hadn't been at 100% usage for almost a year straight, it might have lasted quite awhile.

if my new board has heat issues i suppose i will rework my entire cooling system. get rid of the h70 in favor of traditional water cooling setup.

if anything it seems like the VRM pads were either not fully seated against the base of the heatsink.

(03-12-2011, 01:37 AM)Squall Leonhart Wrote: [ -> ]if anything it seems like the VRM pads were either not fully seated against the base of the heatsink.

from the looks its a Transistor regulator anyway

and the only reason they would heat up are not grounded properly or just connected wrong

hence why he said what he said

mosfets get really hot, and the major problem with them is the hotter they get, the less clean the power flow becomes, so it strans the mosfets harder making them even hotter.

so its not uncommon for a mosfet to die hard if the heatsink becomes unseated or displaced, or the thermal pad shifts.

(03-12-2011, 08:25 AM)Squall Leonhart Wrote: [ -> ]so its not uncommon for a mosfet to die hard if the heatsink becomes unseated or displaced, or the thermal pad shifts.

agreed but ive never ever seen them blow up before they just die. and ive killed a few from heat by not grounding them at all XDD

I've seen my friend blow them up pretty much like these, he bought a new MoBo and installed like 3 more fans in the sides and the top