PC SHUTS DOWN BY ITSELF

[lilith]

I got prince of persia yesterday...installed it correctly and shutted down the pc to enable the game to settle down or whatever...anyway!

while the pc was turned off I thought of cleaning it a bit,so I took a rug and cleaned the fans a bit...I took a closer look at the cpu fan and noticed that while the fan was clean enough,the cooler(behind the fan) was full of dust! so I took a couple of those cue tips you use to clean your ears lol and took out most(if not all)of the dust on the cooler.I then took a hair dryer and blew in the case,lots of dust came out lol.

after that,I switched on the pc,played prince of persia for about 30-40 min then the screen went black for just a milisecond(seriously...it wasn't even half a second) the sound stopped immediately after that and then the pc shutted itself off.

WTF?

that's the 1st time it does that!

restarted it immediately after that,all was fine...played prince of persia again,after 5-10 minutes, EXACTLY THE SAME THING HAPPENED...screen went black,sound stopped,pc shutted down by itself.

again,I restarted it,again,everything was fine,played prince of persia again,after 5-10 minutes,again the same thing...

after that I was worried my GPU(radeon 4850)was leaving me so I thought of testing other games...played COD:MW2 for about 30-40 minutes without a problem,quited the game,played venetica for about 1:30-2 hours and everything went well...no shutdowns,no restarts no nothing.

after that my bro was using the pc to surf the net for about 3 hours with no problem.
when he finished, I again played venetica for about 2-3 hours BUT! this time the pc shutted down again! while playing venetica!

WTFOMG?!?!

I then remembered that I installed new ati 10.6 drivers about 4 days ago so I rolled back to the previous version drivers,played prince of persia for about 20-25 minutes and content nothing was happening,exited the game,shutted down the pc and went to bed.

today,I play prince of persia for about 10 min with no problem, then I got up because family was having launch so I pressed pause and left the game running while I ate.

after about 20 minutes,I came back and found out that pc was turned off!
so obviously,rolling back the drivers didn't fix the problem...

WTF is going on here?this is the 1st time my pc does that and from what I've seen only on games...it can stay on the desktop for years without anything happening.
now I'm worried I may have broken something when I was cleaning the pc but if something was broken in the first place, then the pc wouldn't switch on at all,right?

upon searching a bit,people say that because some motherboards have a limit of 60 degree celsius,if they go over that limit,the pc shuts down.but I don't think that's the case...I've had the same m/b for about 3-4 years now with no problems...
PSU?don't think so as well...had it for 2 years and if it really was the PSU's fault then i don't think the problem would only appear in games...

help anyone? I really am at a loss here guys

 

[banthracis]

1. Make sure no virus/malware.
2. Make sure windows and all drivers are up to date and the drivers you used are in
fact for your OS.
3. Dl HW Monitor from CPU ID and check out load temps.
4. Prime 95 12 hrs min to test CPU stability.
5. Memtest 8 passes min to test RAM stability.
6. Reformat and reinstall everything.

That's the easy stuff.

 

[lilith]

while playing prince of persia I noticed the game stuttering for a bit so I immediately at+tabbed out of it and saw a window saying "application generated an exception"or something like and then 2 strings of code(I think)after that...I didn't get a good look at it because THE PC TURNED OFF A SECOND LATER.
I managed to get a glimpse of the gpu temp using catalyst...it was 118C JUST before the pc shutted down...supossedly 120C is the temp that the card switches off...coincidence?

measured temps and fan speed with speedfan and this is what I got:
cpu temp when idle 40-43C
fan speed when idle 4500 RPM

cpu temp when playing prince 58C is the most I've seen
fan speed when playing 5500 RPM

catalyst measurements for GPU
IDLE
temp 80-90C
activity 0%
fan speed 40-50%

PLAYING GAME
temp 100-118C (118C is the maximum I've seen it)
activity 90-100%
fan speed 100%

I'm going to try those programs you suggested banthacis...
should I invest in a GPU cooler?temps have always been like this for almost 3 years now....why should it be a problem all of a sudden?

everything is stock cooling and nothing is overclocked

 

[lilith]

I downloaded prime95 and had it strees the cpu and ram for about an hour...even though it had cpu usage 100% and almost all of my 2gb ram were taken up (the damn thing took 1.9gb ram!),cpu temp BARELY went over 60C...highest I ever saw it was 62C with the fan at 5600RPM till I finally decided that that wasn't the problem...

downloaded 3dmark vantage and kept on running tests with full details,maximum resolution,AA,multisampling,whatever EVERYTHING TURNED ON MAXIMUM in the options for about 40 minutes and still it didn't shut down!
even though most of the time I was seeing slideshows of 5-10 fps it didn't shut down...couldn't check temp and fan speed though as it wouldn't let me minimize 3dmark...

I thought 3dmark pushed GPUs TO THE ABSOLUTE LIMIT? guess not...

I haven't got any thermal paste...but really...gpu cooler...is it really necessary? damn...

 

[apache_lives]

furmark - downlad it

 

[adaman2576]

while playing prince of persia I noticed the game stuttering for a bit so I immediately at+tabbed out of it and saw a window saying "application generated an exception"or something like and then 2 strings of code(I think)after that...I didn't get a good look at it because THE PC TURNED OFF A SECOND LATER.
I managed to get a glimpse of the gpu temp using catalyst...it was 118C JUST before the pc shutted down...supossedly 120C is the temp that the card switches off...coincidence?

measured temps and fan speed with speedfan and this is what I got:
cpu temp when idle 40-43C
fan speed when idle 4500 RPM

cpu temp when playing prince 58C is the most I've seen
fan speed when playing 5500 RPM

catalyst measurements for GPU
IDLE
temp 80-90C
activity 0%
fan speed 40-50%

PLAYING GAME
temp 100-118C (118C is the maximum I've seen it)
activity 90-100%
fan speed 100%

I'm going to try those programs you suggested banthacis...
should I invest in a GPU cooler?temps have always been like this for almost 3 years now....why should it be a problem all of a sudden?

everything is stock cooling and nothing is overclocked

Those temps on your radeon 4850 are way too high!! The max spec temp for that card is 120C and you are not supposed to be getting close to that. A general rule I have for GPUs is that you never want to go over 100C.

The high temps might be a faulty card or not enought cooling in your case. It should idle around 60-70C max with fan speed at 100%. Could you give me more insight as to what case you have and what size and how many case fans you have?

 

[jdbaker82]

As soon as I read your post I knew this is a over heating issue, most likely your video card if you have a cheesy video card to replace it with to test throw it in and see if your computer still shuts down on its own my bet is it wont. You will need to buy a new video card.

 

[lilith]

air flow inside the case? nonexistant.

a couple of years ago the fan blowing air inside the case broke and I never got around to replace it,so I left the case open, and I mean OPEN! I'm looking on the cpu fan spinning around as I type this lol
right now the case is open and I only have 1 fan on the top back of the case blowing air out.
the case I have is older than I can remember...bought maybe 10 years ago? I've been building all my systems in there for all these years with no problems.

adaman,really? too high? I was seeing everyone saying that the 4850 is way too hot and noisy that i didn't really pay attention...but it's always been like this...

could it be that the hair dryer melted or dissolved some of the thermal paste of my hardware?

 

[lilith]

downloaded and run furmark as apache_lives suggested.
I set it at stability mode at 1680x1050 resolution(maximum) and after 20 minutes gpu max temp recorded was 119C but most of the time it ranged around 115C-118C...I didn't even see it go up to 119C but anyway...

what now?lol CPU survived an hour of prime95...GPU survived 40minutes of 3dmark vantage with everything set to maximum and 20 minutes of furmark at max resolution...could the problem have fixed itself overnight? lol

 

[Jonathern]

Make sure you buy thermal paste that isn't liquid based try silicon type paste trust me the first little tug on the cooler the pasr cracks then comp has issues,but 10/1 you evapourated your paste

 

[lilith]

prince of persia still shuts down my pc after about 10 minutes of use...is it really true that prince stresses the gpu more than all those benchmarks I did?!?! and for a longer amount of time? wtf?

if I use rivatuner to have the gpu fan 100% all the time that'll mean it'll go in the game with a lower temp and won't peak at the same temp as before?
or am I dreaming here?

 

[lilith]

if I get a 2nd case fan to close the case will that decrase the temp levels AT ALL? I'm thinking of buying one,closing the case so the hot temp outside won't mess with my damn hardware anymore goddamnit
but then I'm worried that with my old case and the millions of jumbled wires in there,I may be doing more harm than good!

 

[Jonathern]

Ok i have 5 fans in my case wiring is workable just gets some extensions cable tie to the sides o the case as best you can,if you don't have a see through cover panel it won't bother you,wires won't do anything if you neaten them up a bit.

more fans cooler system,less drag on hardware more performance hope that helps

 

[lilith]

cooler system ok but by how much? I don't think 2-3C is gonna save me...

 

[Jonathern]

It depends if if you on a strict budget i would look up the raidmax skyline or cool master haf 922-932,hardware needs to stay cool and its good to clean them up now and then dust build up ain't good

 

[lilith]

installed a new fan and closed the case....temp went up to 92C
that sucked...opened the case again and have the fan in there anyway.

can any of you people here answer me this?
I used prime95 to stress the cpu and ram.after 1 hour of stressing,it worked great,no shutdowns.
I used 3dmark vantage with everything maxed out for 40 minutes to stress the gpu and it didn't shut down
I used furmark for 20 minutes further stressing the gpu,and it didn't shut down.

I play POP for 5 minutes and it shuts down.

can ANYONE explain to me WHY this happens?

 

[Jonathern]

Does your case have any ventalation?,like at the back of your case is there room for a 120mm fan or 80mm?my guess is you don't get enough airflow through the inside even with a fan you need the hot air to exit the case,as for the computer seizing i doubt really anyone has a answer as it could be multiple things going on in there give me a description of what you have in there for cooling?

 

[lilith]

I have already mentioned my pc case is open.that means that if I look to my right over here I can see every little thing that goes on inside my case...it's not a see-though cover or anything...there are no covers...simple as that.
I have one fan down in front putting cool air in the pc and one up in the back to throw out the hot air.
but ventilation is thrown out the window because the case is open...with the case closed,temp goes up to 92C so I have it open all the time

my guess is something's wrong with the fan...at 100% fan speed temp barely went down 10C...it may have moved from its position or something...I'll take out the card and examine it with some help today.

 

[Jonathern]

Ok well let me know what happens,it could be that your cpu fan is not good enough for the cpu itself,you didn't change any jumpers?

 

[lilith]

there's nothing wrong with the card as far as I can see...everything's in its place,fan's working normally...I guess all that wear and tear of the past 2 years have finally shown themselves and it's past the FUBAR point right now...

this should cut down those pesky temps but is it worth it?
http://www.plaisio.gr/Computers/Tuning/Aircooling/VGA-Coolers/Scythe-Vga-Cooler-Musashi-SCVMS-1000.htm
that cooler's good for 4850 and 4870 and when I upgrade I'm gonna get a 5000 series so I'll have to throw it away?

it's either that or this
http://www.multirama.gr/products/product?pid=0429155&catid=020300

I can't afford a 5850 at the moment and a 5770 seems the best option... it has 10+ fps difference,lower power consumption and lower heat levels.

or I could buy a 4850 again

 

[Jonathern]

Dude computer parts like this shouldn't die in such a short space of time,what happens you replace all that and the problem still persists?take out your G-card test your pc with on board graphics and see if it still does it?

 

[lilith]

THERE MAY STILL BE HOPE!

i've searched around and it seems that many people have the EXACT(idles at 80C,shuts down near 120C) same problem and not just with connect3d but all kinds of 4850 cards.
they fixed it by cleaning the dust that was under the cooler!between the card itself and the gpu cooler.
I'm gonna have to take it out again and clean the living daylights out of it.
what's the best way to clean a card?a piece of cloth?I'm afraid it might get snagged on some of the chips and pull them out or something..
how about a huge fan(the ones that cool your room.let alone your pc lol) that blows cool air?
or one of those compressed air cans?

if I have to take out the gpu cooler, do I absolutely have to put thermal paste again on it before I put it back on?
people say it's not necessary but for maximum results you should put thermal paste again on it.

 

[Jonathern]

I study PC Engineering and i handle hardware best use something with alcohol properties like surgical spirits,Isophopal Or your sisters nail polish remover