Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

Okay, so here is the deal: We have a PC at home that used to be mine, it was running pretty well (though sometimes blue screen, but most likely due to faulty RAID controller, as now other 'strange' issues have been fixed aswell). It ran on Windows Vista on my own screen and keyboard/mouse.

Now I bought a new one and my dad got that one, installed Windows 7 on it and he is using his own screen and keyboard/mouse (not neccessarily marked as working with W7, but no problems in the actual usage).

PC was working fine, no other errors besides random strange things that were solved since the faulty RAID controller is no longer in use.

Then a few weeks ago the CPU cooler died and PC started to randomly Blue Screen after being on high load for some time.

When the cooler was still dead it would survive max 2 mins on 100% load, and would Blue Screen around 80 degrees celsius cpu temp, but don't think you can draw conclusions from that.

Now a new cooler is alive and kicking, but the PC still Blue Screens after 20 minutes.

Blue Screen errors are pointing towards ntoskrnl.exe and are being caused by DRIVER_IRQL_NOT_LESS_OR_EQUAL mostly and sometimes MEMORY_MANAGEMENT.

Would appreciate any help on this, but I would most likely want to hear a special relation between the fact that the CPU cooler died and that after that it is suddenly giving blue screens.

edit: On low usage the PC runs fine without any errors at all for long periods of time. This was already the case when the CPU cooler was dead, and is now still the case.

Edit 2, system specs: CPU i7-920 (first series), can handle according to Intel up to 105 degrees celsius. Motherboard Gigabyte X58-UD3R RAM 12GB DDR3-1066 I believe 120GB Corsair Forza 3 SSD 2x 1TB HDD GPU GTX285 Corsair TX850 power supply (maybe a little bit overkill) - The pc is about 4 years old -

share|improve this question

4 Answers 4

Then a few weeks ago the CPU cooler died and PC started to randomly Blue Screen after being on high load for some time.

That’s expected; electronics will malfunction when they overheat.

When the cooler was still dead it would survive max 2 mins on 100% load, and would Blue Screen around 80 degrees celsius cpu temp, but don't think you can draw conclusions from that.

Wait a second, so you continued to use the system without cooling‽ ⊙ ▃ ⊙

Surely you mean that the fan stopped spinning, but you left the heatsink on correct? The heatsink alone probably won’t be sufficient, but it’s still better than nothing at all.

edit: On low usage the PC runs fine without any errors at all for long periods of time. This was already the case when the CPU cooler was dead, and is now still the case.

You may have gotten lucky and done no damage, but…

Now a new cooler is alive and kicking, but the PC still Blue Screens after 20 minutes.

…you cannot be sure for certain because even though most of the CPU is still functioning does not mean that some parts did not get burned out. In that case, it would function correctly for a while, but then when the system does something that causes it to use a part that has burned out, *bam!*, BSOD.

Would appreciate any help on this, but I would most likely want to hear a special relation between the fact that the CPU cooler died and that after that it is suddenly giving blue screens.

Like I said, electronics don’t like heat, and it is absolutely no surprise that an overheated CPU causes blue-screens. When you ran it without proper cooling, it may have damaged part(s) of the CPU which now cause intermittent (random) problems.

That said, you mentioned that the system runs fine when there is a low CPU load, so it may not be damaged, but simply still overheating. Have you monitored the temperatures since you installed the new cooler? How hot does it get when the CPU comes under load?

If it get hot, then check to be sure that you installed the new cooler correctly.

Did you add the thermal pad or paste? Is the heat sink firmly attached so that it is tight and doesn’t move? Is the fan spinning at full speed?

Use a system-monitoring program (e.g., SpeedFan) to keep tabs on the temperatures and fan speeds.


Photo of thermal paste found with Google Images using query “thermal paste” Photo of thermal pad found with Google Images using query “cpu pad”

share|improve this answer
    
The CPU has always worked fine on 80 degrees or higher. I maybe should have mentioned the setup, the CPU is a i7-920, they can handle generally more heat than others CPU's, Intel claims it can handle up till 105 degrees but that's not a good idea of course. Cooling, etc. works fine, even better as before. When I had the PC on old cooler it ran Prime95 on 90-95C and didn't have problems back then. How would I check if the CPU died (partially)? Also we were not intending to use it 100% without cooling obviously, but some old DOS program appereantly caused very high load. –  skiwi Oct 5 '13 at 14:47
    
The CPU has always worked fine on 80 degrees or higher. Define “fine”; a car will drive with the parking brake on, but it’s not a good idea and does damage. Just because it can go high doesn’t mean it should stay high for a long time; consumer CPUs are expected to idle most of the time. How would I check if the CPU died (partially)? You can’t. I suppose theoretically a program could exist that could check single transistor, but I doubt such a thing exists. some old DOS program appereantly caused very high load Actual DOS or Windows command-line programs? –  Synetech Oct 5 '13 at 17:35
    
Well I should have said correct instead of fine. The PC did occasionally blue screen in the past (after few hours of BF3 continiously), but it could've been caused by a defect raid controller and BF3 was installed on that drive. And the DOS program is called DBase 3, which runs via DOSBox on W7. –  skiwi Oct 6 '13 at 9:35
    
The PC did occasionally blue screen in the past (after few hours of BF3 continiously) Presumably it would get pretty hot after a while. It really sounds like cooling is and has always been a problem. Is the system in an under-desk cabinet or something? Try putting it up on the desk and out in the open. Try opening the side and point a house fan at it. If you can get the temperatures down so that it runs with a high CPU load without a BSOD, then it’s definitely a heating issue. Also, have you tested your memory? That’s a (too) common cause of BSODs. –  Synetech Oct 6 '13 at 13:27

Use a tool like Core Temp or Real Temp to monitor your CPU temperature under idle, light and heavy loads. If the temperature goes up to unsafe levels before your computer BSODs, it means that your heatsink (cooler) hasn't been mounted properly or isn't receiving adequate ventilation.

share|improve this answer

There is no way to answer this question without it being an opinion...so I will stick to facts and the fact is you can't run a computer without proper cooling and expect it to survive...I am surprised that the computer even booted up. Most computers that I've worked on will try to protect itself by shutting down when the cpu reaches a certain temp and 80 degrees celius is way too high for any machine...I suspect that while you were running the computer without the cooler it damaged the cpu so naturally you will get errors and bluescreens...I would take the cpu out and give it and the heatsink a good cleaning and re-apply the thermal compound and see if that helps... here is a wiki on cpu cooling... http://en.wikipedia.org/wiki/Computer_cooling

share|improve this answer

I'm servicing computers for a long time, and from my personal experience the max temperature i7-920 can go to is ~90C at which it BSOD, but this is under some extreme conditions. According to Intel, its minimum/maximum operating temperature is 5-67.9°C. I wouldn't recommend running this CPU on such high temperatures you are mentioning. With prime 95 and stock cooler it shouldn't pass 80-85°C. With good cooling when overclocked it reaches 70+°C, but to work on 80+ °C regularly is overkill IMHO.

First, take your cooler out and clean the contact surface of the heat-sink and CPU, thoroughly with a medical petrol (best in my opinion), a 90% alcohol, or a nitrocellulose thinner. leaving the residues from previous thermal paste or having dirty surface can seriously affect the heat transfer from the CPU to the heat-sink. Then apply small amount of thermal paste ( I would recommend some high performance thermal paste rather than these cheap ones) and mount the cooler.

Check your temperature after.

PS.

Just to add that you can't use these cheap coolers with i7-920 as your Bloomfields are power thirsty since maximum power dissipation can reach ~230W, while TDP is declared as 130W. So having the cooler which fan running fast, doesn't mean that it "kicking". So google a little and check for a good cooler

share|improve this answer
    
I think the 105 degrees is about the core temperature, and not about the package temp. This goes aswell for the 80 degrees I have mentioned, that is the (average) temperature of the individual cores. –  skiwi Oct 5 '13 at 17:24
    
Yes, it's 100°C AFAIK. But still I've had a lot complaints on 90°C+ BSODS for 920. And we sold a lot of 920 CPUs back then. i7 throttles down speed at 100°C, so if i remember well, CoreTemp has set TjMax for 920 on 100°C (it was long a time ago so I could be wrong) –  Nikola Dimitrijevic Oct 5 '13 at 18:04

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.