My processor is going to esplooooode!!!

Vinny

CAGiversary!
Feedback
216 (100%)
So, a few weeks ago, my processor all of a sudden jumped temps. Big time.

Relavent Specs:
AM2 X2 3800+ Windsor (95watt)
Asus M2N-E
4X HDs, all Seagates
Sapphire X850XT (OCed, with Arctic Cooling Silencer 5 which during this incident I replaced with a Zalman VF700 Alcu b/c the Silencer 5 started making a funny noise).

I used to idle at 32C and load at 55C (using AMD's stock cooler). My OC was light and at stock volts so that's not much of a concern.

Then, one day after turning on my computer (I usually leave it on, but I turned it off for a day), my idle temps jumped to 39C and my load temps were 60C. I thought that it was because it was getting hotter outside but funnily, my graphics card was at the same idle temps it's always been at.

I was concerned.

So I thought that maybe I need a new heatsink and so I get a Arctic Cooling Freezer 64 Pro. And it didn't help. My idle temps were still 39C but my load temps did fall to about 47C.

(*This is also when I install my Zalman VGA cooler to replace my Silencer 5.)

So today I remounted my processor (I always use Arctic Silver 5 of course) and lo and behold, my idle temps now jumped to 48C!! My load temps jumped to 58C.

Not only that, but I noticed these strange marks on my processor. One is this coarse edge around the whole processor (which I think is normal but can't remmeber). The other are these two little dark spots...

Here are some pics, I couldn't get any good pics since I didn't have a decent camera.
Both problems: https://publish.insightbb.com/personalfile/personalsite/fyreboltx/cpu1.jpg
Highlight of problem 1: https://publish.insightbb.com/personalfile/personalsite/fyreboltx/cpu3.jpg
Highlight of problem 2: https://publish.insightbb.com/personalfile/personalsite/fyreboltx/cpu2.jpg

Any ideas about WTF is going on?
 
Those picture really suck. But not becuase of your camera. Your focus is way off, the paper towels and mother borad look fine but the processor is too close to the camera. Just move it back 3 inches and you will have good pictures.
 
As a question, are you using those paper towels to wipe off the processor? I always was told that a lint free cotton cloth is the only way to clean a processor with 99.9% isopropyl alcohol.

Also, i thought that thermal grease needs a set period of two weeks with normal startup and shutdown cycles to correctly set.
just a few thoughts off the top of my head but those marks are iffy.
 
[quote name='dtarasev']Those picture really suck. But not becuase of your camera. Your focus is way off, the paper towels and mother borad look fine but the processor is too close to the camera. Just move it back 3 inches and you will have good pictures.[/QUOTE]

Nah, I tried going back. Or maybe I suck. I took about 35 pictures and those were the best ones sadly... the camera is really old, like 6 years old and it was a cheapo $150 camera back then.

[quote name='TheClassicObserver']As a question, are you using those paper towels to wipe off the processor? I always was told that a lint free cotton cloth is the only way to clean a processor with 99.9% isopropyl alcohol.

Also, i thought that thermal grease needs a set period of two weeks with normal startup and shutdown cycles to correctly set.
just a few thoughts off the top of my head but those marks are iffy.[/QUOTE]

You can use paper towels according to what the guys over at TH told me so that what I use. I also use 91.x% isopropyl alcohol.

I know it takes some time to set, but can the thermal paste really make that much of a difference (like, 15C)?
 
[quote name='Vinny']
You can use paper towels according to what the guys over at TH told me so that what I use. I also use 91.x% isopropyl alcohol.

I know it takes some time to set, but can the thermal paste really make that much of a difference (like, 15C)?[/quote]

I use paper towels and alcohol as well so it is not that. And thermal paste should not make that much difference.
 
I figured as much... most people only report a 1-2C drop in temps with AS5.

I really wish I knew what it was. I would take my processor out again to try and get some better pics but it's such a pain, especially with the Freezer 64 Pro.
 
[quote name='michaema']Maybe its the sensor that reports the temp

http://www.silentpcreview.com/article191-page1.html


*EDIT*
BTW, have you made any changes to your computer software wise, install any new utilities or apps? that may be constantly running. How is your CPU load, is it higher than usual at idle?[/QUOTE]


These are some good questions; CPU load was one of the first things that popped into my head. Granted you probably have nothing intensive running in the background, but we need to know for sure before we move on.

I might've missed this part, but you're not overclocking the CPU right? Which software do you use to monitor the temperature? I don't really have any tips on what to do right now, but to say that you should closely monitor the temp (I'm sure you're doing that already) and see how much does the temp fluctuate. I would probably remount the setup once more and get rid of the excess thermal grease (as you dont need a whole lot per CPU surface).
 
Current temps (idle):
CPU: 48C
Core: 43C
VGA: 38C
VGA environment temp: 35C (temp around the VGA)
HD Temps: 38C, 38C, 38C, 39C
My room temp: 23C

My idle usage is under 2% (with the occasional jump when I start an app).

Usually, I have Firefox, iTunes, and uTorrent running- during which the processor stays under 2% and jump to about 5-15% when a new song starts or go to a new page on a website (but it's usually like 7%). I don't do anything extensive for any long period of time besides gaming.

My CPU was OCed but since first noticing the huge jump, I restored it to stock.

No new hardware or software was installed when I first noticed this. I was off on my timing, it was actually a long time ago, about mid-Feb when it first happened. I turn off my computer because I was going home for the weekend and when I come back, my temps are high.

I'm using SpeedFan to monitor temps (ATITool for VGA temps). I have been trying to get ASUS Probe from ASUS' site but they have to have the worst site ever... it always loses the connections and speeds are terrible.

My computer's been working fine besides two random restarts in the past few days (it seems to be gone now). I think I may need to replace my PS (OCZ PowerStream 600W).
 
can your mobo monitor the fan rpm on your ac64? Ive gotten a few low fan warnings, but for the most part, my ac64 fan runs at about 2200rpm. Ive got the same cpu and idle about 37 degrees and at load it hasnt gone over 49 degrees.
 
[quote name='uli2000']can your mobo monitor the fan rpm on your ac64? Ive gotten a few low fan warnings, but for the most part, my ac64 fan runs at about 2200rpm. Ive got the same cpu and idle about 37 degrees and at load it hasnt gone over 49 degrees.[/QUOTE]

Um... well, SpeedFan currently shows my fan speed as 1350000RPM so it can't be that.:rofl:

I know ASUS Probe II (my mobo's temp detecting software) can but like I said, ASUS' site sucks ass and I can't even get it to load to download and try it. I can control the fan speed and stuff via SpeedFan but I don't get to see my fan's RPM, at least, not the actual RPM.

[quote name='Kayden']Sounds like a sensor issue. Either that, or your heat sink isn't sitting right.[/QUOTE]

That seems like the most logical answer but I don't know of any physical way of measuring my CPU's temp. I guess I could not have the HS on, turn on my computer, stick a thermometer on my CPU and compare it to the reading SpeedFan gives me.

I'm 99% sure the HS is sitting right but I won't leave that option out.
 
[quote name='Vinny']
That seems like the most logical answer but I don't know of any physical way of measuring my CPU's temp. I guess I could not have the HS on, turn on my computer, stick a thermometer on my CPU and compare it to the reading SpeedFan gives me.

I'm 99% sure the HS is sitting right but I won't leave that option out.[/quote]


What gets me, is the temperature of vid card has stayed the same, logically it would seem that temperature would rise somewhat if the cpu temperatures are actually rising. IMO if you arent having any performance issues, and dont physically notice excess heat from your machine, I would just ignore it.. Hell if you wanted to you could aim a portable fan at the opened case and see if the temp drops accordingly
 
[quote name='michaema']What gets me, is the temperature of vid card has stayed the same, logically it would seem that temperature would rise somewhat if the cpu temperatures are actually rising. IMO if you arent having any performance issues, and dont physically notice excess heat from your machine, I would just ignore it.. Hell if you wanted to you could aim a portable fan at the opened case and see if the temp drops accordingly[/QUOTE]

Well, the graphics card has it's own sensor... which is another reason to believe my processor sensor's gone bad. I'm trying to find a way to measure the processor temp via other methods.

And I did try a fan directly on an open case, no difference. My HDs cooled down but that's pretty much it.

But the reason I'm worried is because I can feel the heat. I put my hand on my HS and it feels warm when it should feel cool (at idle temps!).
 
My board has a couple of sensors. One for the CPU, one for the GPU and one for ambient case temp, (some hard drives also have heat sensors).

If your processor were actually running hotter, ambient temp would rise as well. Unless, however, the heatsink isn't seated properly, in which case, your machine is more than likely to overheat and shut down.

If neither happens, then its most likely a sensor issue. Which isn't real good in its own right as it means other things could be wrong or you wouldn't know when you did have a real thermal event. Some motherboards shut down at X temp, even a false increase could cause headaches.

As for putting a thermometer on your CPU... I hope that was a joke. Not only would the reading be higher than it should be due to the lack of a heat sink and the structured air flow, but the CPU would probably overheat and shut down before you even got the thermometer warm.

[quote name='Vinny']Well, the graphics card has it's own sensor... which is another reason to believe my processor sensor's gone bad. I'm trying to find a way to measure the processor temp via other methods.

And I did try a fan directly on an open case, no difference. My HDs cooled down but that's pretty much it.

But the reason I'm worried is because I can feel the heat. I put my hand on my HS and it feels warm when it should feel cool (at idle temps!).[/quote]
 
[quote name='Kayden']My board has a couple of sensors. One for the CPU, one for the GPU and one for ambient case temp, (some hard drives also have heat sensors). [/QUOTE]

I'm pretty sure mine doesn't have an ambient temp sensor.

[quote name='Kayden']My board has a couple of sensors. One for the CPU, one for the GPU and one for ambient case temp, (some hard drives also have heat sensors).

If your processor were actually running hotter, ambient temp would rise as well. Unless, however, the heatsink isn't seated properly, in which case, your machine is more than likely to overheat and shut down.
[/QUOTE]

Thank god for Newegg's 1-year no questions asked replacement policy then. I might RMA it and see what happens. Everything seems to be dying for some reason (I've got a questionable PS, a HD that keeps corrupting my OS, and now this). Luckily, summer is only a month away.

[quote name='Kayden']

As for putting a thermometer on your CPU... I hope that was a joke. Not only would the reading be higher than it should be due to the lack of a heat sink and the structured air flow, but the CPU would probably overheat and shut down before you even got the thermometer warm.[/QUOTE]

Hey man, I'm not an idiot.:p

But a friend of mine recommended that I try one of those front bay fan controllers which have temp sensors. He said to leave the HS on, place the sensor on the heatpipe closest to the processor (one the closet one you can get to). Let the system idle for a while and then shut off all fans. Let the temps read for a while on the fan controller and compare the results.

I can't check ambient case temps though because I never did before.
 
[quote name='SOSTrooper']These are some good questions; CPU load was one of the first things that popped into my head. Granted you probably have nothing intensive running in the background, but we need to know for sure before we move on.

I might've missed this part, but you're not overclocking the CPU right? Which software do you use to monitor the temperature? I don't really have any tips on what to do right now, but to say that you should closely monitor the temp (I'm sure you're doing that already) and see how much does the temp fluctuate. I would probably remount the setup once more and get rid of the excess thermal grease (as you dont need a whole lot per CPU surface).[/quote]
Heed this advice! A dab will do ya! Use a straight-edge to smooth the thermal paste evenly. You prolly have paste caked on the thermal sensor if I was a betting man ;)
 
[quote name='CappyCobra']Heed this advice! A dab will do ya! Use a straight-edge to smooth the thermal paste evenly. You prolly have paste caked on the thermal sensor if I was a betting man ;)[/QUOTE]

I used half the size of a grain of rice.;) I spread it out though, as some guide on Corsair's website once showed (wrapping your finger with a plastic bag).

And... what's this about paste caked on thermal sensor? I thought the CPU temps were sent from the processor to the mobo (and then to me via SpeedFan or whatever app I choose).
 
Hm... a strange drop in idle temps for some reason.

My idle temps are now (differences from temps reported last time in bold):
CPU: 40C -8C
Core: 37C -6C
VGA: 36C -2C
VGA environment temp: 32C (temp around the VGA) -3C
HD Temps: 37C, 37C, 36C, 38C -1C, -1C, -2C, -1C
My room temp: 18C -5C

Strange, maybe the ambient temp does have something to do with it. Or maybe AS5 finally seated properly. Whatever the case, these temps are a bit more reasonable though still higher than normal.
 
Actually, if your PSU really is "questionable" that would explain all your problems. Fluctuating voltage could cause damage to any number of components. As for your HDD corrupting OSes, if it is the HDD, that means that a lot of sectors are dying and its on the way out. If its a new HDD, I'd say bad ram. The HDD writes the info it gets from the CPU and ram, if the ram is bad, the info gets fudged up. I've gotten in a lot of machines where they said they had a bad HDD, but memtest86 showed their ram was crap.

Get ultimate boot CD and run HDD and RAM tests. I'd also look into doing a system burn in after to see if its going to stay stable.

[quote name='Vinny']
Thank god for Newegg's 1-year no questions asked replacement policy then. I might RMA it and see what happens. Everything seems to be dying for some reason (I've got a questionable PS, a HD that keeps corrupting my OS, and now this). Luckily, summer is only a month away. [/quote]
 
I've done 8 hour tests on each of my mem sticks (individually) using memtest. I did a processor burn in using CPU Burn-In for... I think 4-5 hours. But these were when I first built it back in August.

I will most likely RMA both the HD, PS, and mobo this summer.

Now the question is, do I get another OCZ or that fancy modular Corsair?:whistle2:k
 
Are you still overclocking? That would cause corruption errors. With legacy motherboard/CPU's you could overclock by changing the bus speed AND the multiplier (prefered method). Nowadays, the multiplier on CPUs are locked so your only option is the FSB (front-side bus). This unfortunaly has a bad side affect. IDE controllers (esp SATA) are VERY finicky to FSB. If you OC, it throws the standard FSB outta whack which causes the IDE controller to misbehave. This will in turn 'corrupt' the hard drive. The drive itself is fine if it's a light OC, but you may want to run the HD diagnostics from your respective manufactor. I know with Maxtor drives, if the error code is 3-digit, it's fixable with a utility. If it's a 6digit, it's RMA time.
 
After monitoring it for a week, nothing's change... I'm not sure why but the temps fall to more stable levels sometimes for no reason.

Idle is now at 44C but the outside temps are definitely cooler now.

I still won't know for sure what it is until this summer when I can RMA the parts. I'll replace my main HD and PS with something else, and exchange my mobo and see if that helps. If not, then at least AMD's processor prices are falling like crazy.

Shippings going to be a bitch though, especially for that PS.
 
bread's done
Back
Top