PC "randomly" crashes

[cluelessgenius]

ok so a buddy of mine reported his pc crashes randomly and i thought id ask the geniuses on here:

 

he got an amd fx 8120 (old) on some old gigabyte board (very old) with 8gb ram (almost brand new) and a gtx 970 asus turbo (well kinda newish) and an older wd blue 500gb drive all run by an intertech crap psu.

 

he reports it just shuts off with no bluescreen or anything just off. also not just the monitor but the entire system.

 

first thought : psu. so he ordered a new one now (be quiet pure power 10 500W) and tested it today. still crashes.

 

he also tried a different gpu in there still crashes. and he also checked all cable connections and ram fitment.

 

the crashes are seemingly random as we cant figure out what they are tied to. its not load related as it also happens in idle.

 

windows event manager only reports kernel power as a fatal error so it looks to me theres a hardware issue somewhere. but since we already tried a brand spanking new psu idk.

 

also i thought it might be processor instability like you were overclocking but usually that only gets you a bluescreen not complete shutoff and also he didnt touch the bios or any settings related to that

 

i told him to clear cmos now and he will try that later today. 

 

but does anyone here have an idea what this could be related to? 

[Zando_]

I'd guess mobo failure if the PSU has been ruled out. Though if your CPU hits over 110C, doesn't the whole PC shut off? I'd check CPU temps as well.

[Guest]

Did he overclock? Is his cooler seated properly?

[cluelessgenius]

5 hours ago, Zando Bob said:

I'd guess mobo failure if the PSU has been ruled out. Though if your CPU hits over 110C, doesn't the whole PC shut off? I'd check CPU temps as well.

temps are fine. mobo failure not ruled out.

5 hours ago, Konrad Kwasniewski said:

Did he overclock? Is his cooler seated properly?

nope as mentioned he didnt touch the bios or anything oc related. cooler and therefore temps are fine.

 

so now cmos clear didnt work but he did notice that running directly of a dvd and unhooking the hdd completely  ran for 2.5 hours straight hell test how long it runs like that.

i suggested he try a different sata port on the board but doesnt seem to work. well see..

[ElSeniorTaco]

I would not be surprised if the ram is bad

Despite it being new, it may have come defective and finally gave up.

It's actually fairly common in my personal experience and does cause random failures similar to what you are experiencing.

 

Bad ram can cause all types of weird failures under all different types of conditions

It can run great for a full day, or last only 10 minutes before a failure, it can also fail under load or at idle

its a  tricky one to pin point but is simple to diagnose if you target that type of failure directly

 

Normally I would always start with a full diagnosis, however a few things have been ruled out by the description.

So I would proceed with the following

Start with just the motherboard, cpu, hard drive, and one stick of ram

Since It crashes with a different gpu, I would say it might be safe to leave the gpu in it.

 

Boot it up, see if you can make it crash,

If it does, shut it down and try a different stick of ram

If you get through all the ram, and its still crashing let us know

Edited February 14, 2018 by ElSeniorTaco
added more detail

[cluelessgenius]

alright guys i had a chance to visit him and have a look myself on saturday and its now fixed :D.

i went over there and was still confused at first because the new psu did the same thing and my psu tester didnt find anything wrong with his old psu. then we tried memtest and that didnt result in anything either. 

then finaly i noticed his cpu cooler was a little skewed to one side. so after uninstalling the mobo and all that to get access to the cooler mount i took it of and found that te spread of the thermal paste was uneven and for some reason it also looked like it was getting more liquid. like there was condensation or something. if you know what i mean. it didnt have normal toothpaste consistency but reminded one rather of the water ontop of your yoghurt in the fridge. 

so i cleaned it all up put new mx-5 on and remounted the cooler properly and e viola it runs. i havent heared from him since so i assume there have been no issue in the days since then.

but ive never seen paste where it looks like theres water seperating from it. weird.

[cluelessgenius]

oh yeah and temps obviously went down too but they didnt seem high enough before to cause shutdown.