CPU 100 celsius degrees at idle - Pictures of my cpu thermal paste state inside (is this the cause?) Please help :/

omonoiatis9

Distinguished
Sep 23, 2013
75
0
18,630
So today for the first time I got a blue screen of death and after powering off the pc and trying to start it up again it wouldn't saying "CPU overheating". I got my pc 3-4 years ago.

So I let it calm down, power it on, and run HWMonitor. Behold, all cores on my i5 3570 at 100 degrees CELSIUS (not Fahrenheit) without doing anything at all. Like, is the pc even supposed to be able to run at these temps? I shut it down quickly just in case. Note that I also took off the pc cover while it was running and put my hand near different parts to see if I was noticing any heat. I couldn't feel any heat but maybe it's because the room temperature was only 18C. I mean, if it was just a faulty cpu temp sensor why would my CPU be struggling at idle.

I looked it up and the most common causes are improperly seated heat sink or thermal paste needing replacement. I took the fan offf to check the thermal paste but I've never done this before and I'm a complete noob so I don't know what's supposed to look right and what looks wrong. Here are the pictures:

Heatsink paste: https://s17.postimg.org/jnsjgj0y7/heatsink_paste.jpg

CPU paste: https://s16.postimg.org/k0nxinst1/cpu_paste.jpg


So based on your experience do you think these look bad enough to make a CPU run at stable 100C at idle state and even give blue screen of death? Or is it something else? Power supply over/undervoltage? (the fan works completely fine btw and there's no dust in the pc or the fan). The funny thing is I took my pc to a technician like 3-4 months ago for the same issue but on my graphics card. Blue screen of death from GPU overheat. Turns out it was the thermal paste. Is this the famous "planned obsolescence"? Or is it just that all thermal pastes need replacement eventually? (I got the pc 3-4 years ago off of ebay and never had any issues other than the GPU thermal paste needing replacement and the issue I'm having now with the CPU)

I hope the thermal paste is the issue and not some harder to spot problem like power supply over/undervoltage or some shit...


Thanks in advance a ton for any help. Really appreciate it.
 
Solution
it's a bit thick, but that shouldn't cause 100C at idle, maybe 45-50 instead of 30-35.

Don't apply new paste for now, reseat the heat sink, and make sure that the legs click into place properly. My bet is that one or more legs had come loose and there was no pressure from the HSF to the CPU.

If the temp is better having done this, then get some iso propyl alcohol and clean off what's currently on, get some more paste and apply a small amount in the middle of the cpu.
it's a bit thick, but that shouldn't cause 100C at idle, maybe 45-50 instead of 30-35.

Don't apply new paste for now, reseat the heat sink, and make sure that the legs click into place properly. My bet is that one or more legs had come loose and there was no pressure from the HSF to the CPU.

If the temp is better having done this, then get some iso propyl alcohol and clean off what's currently on, get some more paste and apply a small amount in the middle of the cpu.
 
Solution




The thing is, it's a cooler master fan, the kind with the push pin installationv(4 pins/legs, with the arrow thingys that you twist). I don't understand how those would simply come off or lose, the fan was a <removed> to take off myself when I was intentionally trying to remove it. But yeah I also thought that paste doesn't look absolutely terrible to be causing 100C idle... I don't know, I'll try reseating it right now and see if there's any difference. If there isn't any does that point to a faulty CPU?


 


The push pins do creep over time, or one was just not pushed through far enough, quite common issue. Refit and see what happens when you are sure it's got high contact pressure.
 
[/quotemsg]

The push pins do creep over time, or one was just not pushed through far enough, quite common issue. Refit and see what happens when you are sure it's got high contact pressure.

[/quotemsg]

It ran with the same issue with the cover off. Anyway, I refit the heat sink and turned it on. The fan was now making more noticeable noise than before but I brushed it off. I entered bios settings for a few seconds and CPU was around 35-40 degrees. So I went ahead to turn the pc on. Eventually the fan stopped making the excessive noise, computer restarted on its own and I got a "Disk error" message. I have 2 hard drives on. After the disk error message my main drive (the one with the windows installed) isn't even appearing in my BIOS settings anymore so I can no longer access it or boot into windows. Did something fry my disk or did I damage my motherboard while trying to fit the sink back in with scratches near where the leg holes were? (I did scratch it a tiny bit to be honest as I was having a bit of trouble fitting it back in unfortunately, but I didn't think too much of it cause I thought it's almost unavoidable and it'd be bad design to put actual board circuits so close to the screw holes..). I actually hope it's the motherboard and my hard drive didn't spontaneously kill itself.

It's funny cause I have a feeling I fixed the CPU issue but in the process somehow fried my HDD or damaged my motherboard or both...
 
The thermal paste in the two linked images looked super dried and crusty, as if it had been worked hard for at least 5 years, 24/7/365. At least, that's how dried and crusty all the thermal compound in my '09 mac pro looked by the time I completely stripped it for a deep cleaning and water cooling overhaul. I had to soak the damn things in 99% isopropyl for about an hour before I could even begin to soften up the crust and scrape it off.

I suspect, with a fresh application of thermal compound, your temps should drop quite a bit.

I'm also probably super late to the party. Carry on, folks...
 
It still looks like overheating, but it may just be taking a little longer now. That paste was applied so thickly, when it dried, it created space between the CPU and sink. The problems detecting disk and so forth are probably still related to heat. Your fan was working harder until it couldn't keep up. And the CPU wouldn't have been generating much heat in BIOS, compared to once it starts loading.

It's definitely time to reapply paste. This is a link to one of the guides here:

http://www.tomshardware.com/faq/id-1824539/thermal-compound-cleaning-surface-spreading.html

The picture links are dead but the thumbnails are still ok. You may need to do like amtseung and dampen the paste with rubbing alcohol for a while before you can remove it. As always, be sure the computer is unplugged while doing this. You should be able to get rubbing alcohol at any pharmacy and most grocery stores.

There may still be other problems, but they'd tough to diagnose until the overheating CPU can be completely ruled out.
 
For reference, and in case you all think i'm mad/foolish, my working theory is that bad paste would cause high temps but not that high, (you could use toothpaste and not get that high), that high can only be reached by no heatsink or by poor contact pressure.

My thought therefore was as a test, refit ensure contact pressure is good and see what happens, i'd expect better temps but not great temps, this would prove the contact pressure theory, then replace the paste.

As to why it has stopped the disk working...

Perhaps mobo power loss has reset bios and you're now pointing at the wrong disk (be aware of the win8/10 system partition that might be on another disk).
Perhaps if you struggled to remove or refit the HSF something has been damaged.
 

I agree that people seem overly concerned with thermal paste on these forums.


Perhaps it's something as simple as the OP knocking a SATA cable loose when messing with the cooler.

 


I used toothpaste when I was a lot younger and couldn't get the computer store for a couple days. It worked great and my P.C smelt minty fresh! Little rough cleaning up though..

As far as the issue goes it's pretty simple. Poor contact between CPU and heat-sink. Either the you don't have the heatsink tightened up properly, the heatsink is bent, or the heat spreader on your CPU is defective.
 
I took my pc to a technician today. It looks like the PC not starting up at all was because of a coincidental corruption of windows on my hard drive (bad sectors). As for the CPU overheating, I'll know the rest when I get the pc back. I appreciate everyone's input, thanks a lot.
 


Thanks for your input Mr. Mod. So here's what happened by order of events: Blue screen of death on cause of bad sectors on hard drive which is slowly reaching the end of its life. I shut it down, try to turn it on, "CPU overheating" error. I reseat the heatsink without changing thermal paste, temps look alright but when I try to turn the pc on, while windows is configuring updates it restarts on its own. Windows corrupt. Reinstalled windows and that's it.