Overheating issues.

[Algroda]

I've recently had some pretty big overheating issues. I've had to replace the thermal paste 3-4 times in the last year, and after replacing it last week it already needs it again (core temperatures of ~100 C >_<). I'm running a P55H-A (http://www.newegg.com/Product/Produc...82E16813135244) motherboard and a Core i5 750 (http://ark.intel.com/Product.aspx?id=42915). I do not have an aftermarket fan / heatsink, though I likely should. Is that the issue, or is there something else going on do you think?

[Deleted]

Is it 100c from when you login or does it start when you launching wow, other games or some cpu heavy applikations (video encoding or the like)?

[DarkBlade6]

Tell me which thermal paste you use and how much do you put on ur CPU and what you do when you apply it?! No offense but I suspect a bad installation because you put too much thermal paste .

[Drunkenvalley]

Tell me which thermal paste you use and how much do you put on ur CPU and what you do when you apply it?! No offense but I suspect a bad installation because you put too much thermal paste .

The stock fan may have come with it pre-applied. Suffice to say it's not good at conducting heat.

[Algroda]

Is it 100c from when you login or does it start when you launching wow, other games or some cpu heavy applikations (video encoding or the like)?

Boots up ~50, stays ~70 until I get into something CPU intensive.

Tell me which thermal paste you use and how much do you put on ur CPU and what you do when you apply it?! No offense but I suspect a bad installation because you put too much thermal paste .

Arctic Silver, ~a grain of rice sized, spread over the heatsink contact

---------- Post added 2011-05-31 at 07:58 PM ----------

The stock fan may have come with it pre-applied. Suffice to say it's not good at conducting heat.

I know better than to trust the stock paste. I removed that before I installed it

[DarkBlade6]

Yea did you remove the old pre-applied thermal paste on the heatsink ?!

Did you OC ? -_-

[Deleted]

way too hot,i use a x4 955 be and max temp. on highest use is 39° C-installed "alpenföhn brocken" fan,and its most quiet and coolest fan i ever used

[VersedalX]

Thermal paste shouldnt have to be re-applied that much, what makes you feel that you need to replace it?

Also, What are your ambient temperatures like? (temp in the room) and do you have adequate airflow in your pc case? I.e. Air either being pulled in by a fan or sucked in from the front, and blown out the back via fan? (Example - http://www.build-your-own-computer.n...r-cases-01.jpg )

[Algroda]

Yea did you remove the old pre-applied thermal paste on the heatsink ?!

Did you OC ? -_-

Yes, I did remove the old thermal paste. No, no overclocking.

Thermal paste shouldnt have to be re-applied that much, what makes you feel that you need to replace it?

Also, What are your ambient temperatures like? (temp in the room) and do you have adequate airflow in your pc case? I.e. Air either being pulled in by a fan or sucked in from the front, and blown out the back via fan? (Example - http://www.build-your-own-computer.n...r-cases-01.jpg )

Well, I normally get 60 FPS in WoW with no issues. So when it drops to 5-10 FPS and I open my nVidia system monitor and it says very high temperature issues...it's a problem. Once I let it cool off and replace the paste it fixes it, so I'm pretty positive it's the issue.

Room temperature is unfortunately higher than I'd like. I'm upstairs, but I don't have a thermometer here. I'd say ~80-90 degrees in general. My case has a small fan (~3-4 inches) in the front being sucked in, a large fan on top (~9 inches), and another ~3-4 inch fan in the back

It's possible it's a faulty temp sensor. It's not an uncommon issue with chips. Could you get us a reading from the following if possible;
BIOS (under temperature monitoring)
CoreTemp - http://www.alcpu.com/CoreTemp/
RealTemp - http://www.techpowerup.com/realtemp/
CPUID HWMonitor - http://www.cpuid.com/softwares/hwmonitor.html

This might help us narrow it down. Also, since you're using a stock cooler, can you do me a favour (be very careful), and touch the metal on the heatsink (ideally, if you can ground yourself to protect against any sort of static, that'd be better, touching the case with uncovered skin while doing so is usually enough) when you see these temps just to see whether or not it's actually getting hot to the touch. Do be careful just in case it is actually that hot.

Certainly. NVidia system monitor says between 70-74, with core1 at 75-77 at the moment. Core Temp says 69-73, with core0 between 75-79. Real Temp says 75-79 for core 1, and 70-71 for the rest. CPUID says 75-78 for Core 0, and 70-73 for the rest. In the middle of working on a project, or I'd check my bios too.

No, the heat sink was not getting hot the other times I checked it, which makes me think there's an issue with either the contact plate / heatsink itself, or that I'm just messing up something this simple....I will try removing the thermal paste and reapplying, just to be safe, and make absolutely 100% certain I'm only using a rice sized piece, unless that is incorrect.

[DarkBlade6]

Go buy a coolermaster hyper 212+ please =) cheapo 25$ heatsink but a ton better than the stock heatsink!

Also make sure you are exhausting the hot air out of ur case!

[overclockedamd]

never spread thermal paste always used a dot, line, or x style method. Refer to this video for why http://www.youtube.com/watch?v=EyXLu1Ms-q4 at around 2:00 you can see the massive amounts of air bubbles creating using the spread method.

[Nellah]

CoreTemp confirming those readings isn't good, I'm pretty sure that reads directly from the thermal diode on the CPU, and those things aren't exactly known for being unreliable. I'd ask you to run the IntelBurnTest to see if you can crash it, but what that would do to your CPU, neither of us want to find out.

Even the stock heat sink cools better than that, my 2500K only got into the low 90s with a max overclock in a case with one fan; I'd not recommend anyone run those temps 24/7, and my usual load temps are right at 60. The heat sink may not be making good contact with the CPU for other reasons; if the heat sink got dinged or if the pins aren't pushed in all the way it may be pulling away from the CPU. It pays to make sure they're in all the way (with my old E8400 stock heat sink I had to take the motherboard out and use both hands to push the pins in).

In any case, it's worth getting an aftermarket CPU cooler. The Hyper 212 is what, sub-$40? A worthwhile investment in any modern Intel computer. (Not sure if it supports AMD HSF connectors)

Good airflow through the case is paramount as well, if hot air is being circulated through the heat sink then it won't cool anything. Front to back, bottom to top is the rule of thumb for good airflow; intakes in the bottom and front, exhausts in the top and back of the case.

[Algroda]

Redoing the thermal paste with a different method. I'll let you know how it goes after I finish.

EDIT: This time the heat sink was warm. I'm blowing out the case to remove any dust to make sure it isn't that. And going to get an aftermarket heatsink / fan to be safe

[Dizey]

With those temps at stock settings, I'm having a difficult time believing that your thermal paste application is the culprit.

Make sure that your HSF is seated properly and securely, and that the fan is regulating properly. You may have worn down the prongs on your HSF, or the mounting holes on your motherboard, through excessive removal and installation while reapplying your paste. That could result in compromising the contact between the heatsink and CPU. You may want to invest in a better aftermarket cooler, preferably something with a back plate and screws for mounting.

Make sure all of the fans in your case are running, and are free from dust. Where is your case kept? Is there plenty of air space surrounding it? A lot of people have a designated computer cabinet area build into their desks, but that's a horrible place to store a gaming system.

This may be a stretch, but make sure your bios is up-to-date.

Lastly, check any motherboard related software installed that may be tweaking or auto clocking your system to enhance performance.

[Algroda]

No, the issue isn't the heatsink. As i said, it seems like it's a faulty temp sensor. If it was reaching 100C, the safety shutdown would kick in and power the system off.

I don't think it was the temp sensor. Going down to ~5 FPS is a pretty huge sign of overheating.

I cleaned out my entire case...apparently I had not in a while. Power on temp is 50 C, we shall see what it's at here shortly. There was a *lot* of dust, which I think was the issue. I think it was blocking the fans from working properly, including on the heat sink.

EDIT:
After a few hours they're steadily between 70 and 75. That includes playing WoW, so I think it is fixed. Still hotter than I would like, but an aftermarket fan will hopefully help with that.