25 replies to this topic

[prov1]

Hi, I appreciate the time and help.

About 2 weeks ago, my room was feeling warm, more than normal. I had a movie running throughout the night on my pc so i thought nothing of it, and just opened a window to vent. Well the other day i was playing a game and my pc froze, this happened 2 or 3 times. So I decided to check my core temp using SIW and saw it was 70% C, scared me so I did not hesitate to turn off my pc. After I turned it on the next few days after start up the temp immediately began to rise slowly and steadily, and I was just watching You Tube videos, checking email, and surfing. I deciding to open it up and spray some can air around which I do every 4 months or so. It didn't help. Next I removed my heatsink off my cpu, cleaned all the grease off and applied some fresh artic silver 5. Didn't help. I did that 3 times using different methods of application. Last being the dot in the middle. I did notice one of the push pins from the heatsink mount wasn't all of the way through my mother board, I finally fixed that yesterday, but it didn't help temps. Typing this my Core 0 is @ 48% C and Core 1 @ 52% C, which is normal i think, weird...past few days its been mid 60s just from surfing and watching movies. When in the past normal use was 50-55% C and playing Crysis or other graphic stressing games was like 65% C at the highest.

Guess my question is what could be causing the sudden over heat issue? or Could I have damaged it/fried part of it by leaving it running all night with a minimized game, and a movie looping all night?

I started playing Star Wars Revenge of the Sith and immediately my temps are Core 0 @55% and Core 2 @61% and GPU is 56%, is this normal? Definitely lower than the past few days, I think. Yesterday I did download some new drivers for ATI CCC, a HD audio driver, and ATI audio driver, but after restart I still had high temps. But today or now it seems better. But nevermind, seems the longer my pc idles the hotter it gets. Also my clock changed, I wanna say last weekend, it was 30 min behind. Why would that happen?

I rarely turn it off and it has been running smoothly for the past 2 years, I have a great case and ample fans.

Custom Built PC
AMD Athlon 64 X2 6400+
Nvidia MCP61 MOBO
Samsung HDD 250 GB
Maxtor 7 HDD 500 GB
Radeon 5850 HD
CM 690 II Advanced Case 4 intake fans & 2 outtake

can't find my heatsink
and btw i left my pc running idle for 6 hrs with just a browser minimized and I'm up to 75% C for Core 0 and 81% C for Core 1. Is that normal? All temps are read under Value not MIN or MAX readings.

Any help is appreciated!

Edited by prov1, 03 October 2010 - 10:46 AM.

[8210GUY]

I wouldn't be panicking yet, it is common place for a heatsink to need re-seating after x period of time, hardcore users probably do it annually, but for most this would not be required, now heat can certainly damage the parts, but at the temps you are quoting I doubt any harm has happened yet, if the BIOS is configured properly it should auto shut down the system once it hits a pre set level everything else you describe is typical overheating behaviour, so re-seat the heatsink with some new thermal paste and you should be good to go again. Now when doing this it wont hurt to check for dust either, this also adds to the temp's, so have a good clean out taking care not to disturb components or cables, then use some thermal grease remover to clear the old paste (with a lint free cloth) and prepare it for the new stuff, most consider Arctic Silver 5 to be the best, so when it's clean apply the new paste and put it back together again, you're temp's should drop considerably if you have done this right, hope this helps.

[prov1]

I deciding to open it up and spray some can air around which I do every 4 months or so. It didn't help. Next I removed my heatsink off my cpu, cleaned all the grease off and applied some fresh artic silver 5. Didn't help. I did that 3 times using different methods of application. Last being the dot in the middle. I did notice one of the push pins from the heatsink mount wasn't all of the way through my mother board, I finally fixed that yesterday, but it didn't help temps.

I did clean and apply thermal paste 3 times the past few days. No change. There is no dust. Also my time changed, it was 30 min behind. Could something have changed my bios?

The first time I took off my heat sink was in January, in order to put new RAM in, I cleaned off the old grease and applied new, using the credit card scrapping method, it ran flawlessly the past 9 months, until 2 weeks ago.

So 52% to 60% C is ok for normal browsing use?

First time i tried playing a game in 2 weeks.... Tonight CoD MW 2 froze my pc after 5 min of playing. This use to never happen, even after playing for 5-8 hrs straight

Thanks for the quick reply btw. Any other thoughts?

Edited by prov1, 04 October 2010 - 12:40 AM.

[8210GUY]

Apology's for that, I missed you had already said about doing that (brain is fried lol), now the pin you mention is a possible reason, they are complete swines to get through and secure, if this is not sitting correctly it will obviously not do the job right, I appreciate you say it's looking secure, but I mention it given the pin issue you mention, and I am wondering about any warping of the board, with the heat and 1 pin not secure, I have to wonder if it's possible it warped the board, and as a result the heatsink no longer sits flat, hopefully unlikely, but it's worth a look.

One of the easier things to check is the fan, I had a mates system where although the fan was working, it was not running at it's full power, and as a result wasn't moving enough air which increased the temp's, so see if the fan is running correctly.
Now in my mates case the fan itself was fine, turned out to be the PSU, it wasn't pushing enough power to the fan to run correctly, so if you have a PSU tester check that out, or get and run SpeedFan and report back the power readings, probably easier to take a screenshot of it and post it here, then we can see whats going on with that, hopefully something here will give some clues.

[prov1]

Hi 8210GUY,

i like your fan theory, as I don't think its running as fast it should. I could turn it faster with my finger. I remember my heat sink fan rattling/making noise awhile back, i decided to clean it while it was seated. I can almost guarantee that the rattling derived from it no being seated properly with one of the push pins not being locked in. Or the heat could have warped my board pulling the push pin back.. I do have 2 cats, so maybe some hairs got into the ball bearings or whatever kind of mechanism that makes it turn. And it is 4 years old i think.

I also got Speedfan and ran it, but its not showing any fans, or I am unaware of how to get it to recognize them. I hopefully uploaded the image correctly. Is there anything I need to do to get it to read my fans, in particular my heat sink fan? Do I need to take screens of any other tabs?

Instead of purchasing a PSU tester, I will purchase another heat sink fan.

link here

http://cgi.ebay.com/...=item2a0aab1bc5

as it appears to be identical to mine, just with additional mounts. Says AM2 compatible so instead of a $10 psu tester, i'll try this for $15, hopefully.

I guess my mother board could be warped, I don't know how to check that without taking my pc completely apart.

Thanks for helping. Thoughts?

Attached Thumbnails

• 

Edited by prov1, 04 October 2010 - 07:07 PM.

[8210GUY]

I am not convinced that the item you have found will be the cure, given what you have said about the fan speed, and the screen-shot, and my recent experience with a near identical issue, I fear you're PSU is going faulty, do you know anyone who has a volt meter and knows how to use it ? Can you also click on the "Exotics" tab, then click show the magic, and post that screen-shot back here. I have a PSU tester and it showed the PSU on the system that had a similar issue as testing fine, it was only when I put a meter on it I saw that the power reaching the fan was not the same as it should have been, I had tried a new heat-sink and 3 fans before finding that out, put a new PSU in and it was running fine, but not getting enough power made the fan run slow, I hope I'm wrong, but thats what I experienced recently. As a thought, do you have access to another system ? Given the probability of you being unable to sort a meter reading, disconnect the fan from the problem system, then connect it in another system (swapping it for an identical fan), if this fan suddenly seems like it is running at full power (which currently you feel it isn't), then that is as good a test I can think of which is free, and will support my theory, although I would be happier if a meter reading could back this up before suggesting you spend money, but what I have seen and heard so far is identical to a system I had problems here with, so let me know what you are able to do, and don't spend money on a fan yet, and we will see what we can figure for the best.

[prov1]

Hi again,

That sucks if it is a psu problem. I'll have to check the warranty

i spent $139 on my psu in january.
Its SeaSonic M12D SS-850 850W ATX12V V2.3/EPS12V V2.92 SLI Ready CrossFire Ready 80 PLUS SILVER Certified Modular Active PFC Power Supply.

I don't have another system as of yet. My friend may have a voltage meter. If he does it may be tomorrow night b4 I post again.

Well, since I already bid on the heat sink, I can't retract it and its only $23 total. I just can't believe its my psu, although the first one I ordered off of newegg arrived DOA, I returned it for the one I have now. Maybe it is, maybe not. I'll let you know how the volt meter test go.

So What numbers am I looking for in the readings? What voltage?

Attached Images

• 

Edited by prov1, 05 October 2010 - 03:39 AM.

[8210GUY]

I am concerned over the +3.3v reading, it is nearly half what it should be, they say it should be ± 5% (+3.135 VDC) to (+3.465 VDC), Here is a chart that shows what the readings should be, and Here is the ATX 24 pin 12V Power Supply Pinout, needless to say take extreme care when working inside the system, but with these links you have all you need to know what to look for, hope this helps a bit anyway.

[prov1]

Morning...

Ahh, well so I should go ahead and test it anyway with volt meter??

So the orange wires should read between 3.1 and 3.4? Test them all? or just the orange ones which all 4 are +3.3V?

Just so I'm clear, this has nothing to do with the connector between heat sink and mother board. Only with the psu 12v rail, that is supplying power to my heat sink, which you think is not getting enough fan power to draw away heat?

I also have 2 questions to add, sorry. in the attached thumbnail above, exotics tab, what about the +12v isn't it suppose to be reading 11.4 to 12.6 and not 4.93? and if I'm reading the chart correctly the -5v reading at 0.43 seems extremely low. The tolerance for the -5v says -4.5 to -5.5. Am I reading those correctly in being extremely off? now i'm certainly worried.



Thank you for your continuing help 8210guy.

Edited by prov1, 05 October 2010 - 06:48 AM.

[8210GUY]

Yes, check it to be sure, software is not always correct, and is no replacement for taking a meter reading for confirmation, but Yes, you're quite right, the +12v is very low, I was concentrating on the 3.3v reading as thats the one that stood out in my case, so it is certainly starting too look like a bad PSU, given you're earlier statement about this it's possible they had a bad batch, hence getting a 2nd bad one, but be aware, when a PSU goes bad, it can take other components with it, so this needs to be checked, and if needed resolved before it does any damage to the system. I found the easiest access to check the power was via a 4 pin molex plug, the 24 pin plug can be tricky due to the amount of wires in it, but do take extreme care when dealing with electrics of ANY sort, even though it is 12v, it can still carry enough juice to kill you if you touch the wrong thing (mainly inside the PSU).

[prov1]

man, sure is sounding like a psu fail. gonna have to take off some eggs at newegg product review! I got my friend with the volt meter coming over in a few hours, he's a certified electrician so we will be careful. Will report back with the results.

[8210GUY]

TBH Seasonic are pretty good, I always considered them to be anyway, but ANY maker can have a bad run, whether it be a flawed range, or a simple bad batch that got through inspection without being picked up by quality control, the supplier will be able to shed more light here, if it is a bad batch then they will be seeing a lot of returns, and will liaise with the makers to replace the batch if this turns out to be the case, but if you look too changing brands then PCP&C are the tops, although OCZ have bought them out now, or look to Corsair as these are very good choices.

[prov1]

i thought they were good also but this was my 2nd PSU in 9 months. So, i bought it this year, and have a 5 year warranty on it. I am just worried about how this RMA process will go down... like will the 3rd one be bad also? the fees and such plus i don't have the money now for a new/different one.

Edited by prov1, 05 October 2010 - 07:42 AM.

[8210GUY]

I feel that lol, if it turns out to be from a bad batch you will be fine, hopefully if this is the case they will have replaced the batch that was bad with good ones, but as a thought, do you use a surge protector or UPS device ? I am wondering if you may have bad power where you are, and are getting power spikes that is doing this, a good PSU takes the hit before the system, so for peace of mind ask you're mate if he could check out you're power from the wall, if this is where the problem lies then any future system(s) will go the same way, or worse, and fry the whole system, a surge protector helps reduce the risk, but a UPS device is the ideal, but they are not cheap, but you should get a surge p[protector for not too much money, it wont stop everything, but it will reduce the risk, fingers crossed you're mate can confirm these thoughts.

[prov1]

could it be a mother board issue? could an altered bios have anything to do with it? all voltage readings were normal. i will check again later, but psu is fine. the problem with speedfan is where is it taking the voltage readings. could it be through the end devices? b/c its apparently not through the 20 pin connector, which is where we took the readings. actually got 1 fan reading now....yay.

Attached Images

• 

Edited by prov1, 05 October 2010 - 04:27 PM.