Overheating Problem?

[Guest]

hi

Ive been having problems recently with my system hanging. This only happens when im playing games. They dont even have to be particularly demanding games like Call of Duty but it still happens. The computer stays on the sound stays on a small loop for about a minute and then goes into weird moshy sound. The display remains on, recieving a signal but blank.

Now you may be wondering why on earth i have posted this thread here but i have a few suspicions about my motherboard. I have reinstalled windows twice with no avail and the ram has been tested. There was a burning smell coming from my psu however on closer inspection this was just a massive build up of dust. Just to be safe i tried another power supply however the system still hung.

Windows error reporting just reports it as an application hang so theres nothing there that gives me any information. The cpu temperature does not seem to get very high and when left running for 3 hours just in the bios monitoring temps it didnt go above 36 degrees. I also tried overclocking it however the temperature still didnt rise and it didnt crash. The motherboard temperature is reported to be about 40 degress but doesnt seem to go any higher than that. The cpu cooler is not hot to touch however one of the chipset coolers becomes unbearable to touch after about 5 seconds. I think this is the south bridge however they are all linked together with copper cooling pipes. However because of the reportedly low temperature in the bios for the motherboard i dont think this is a problem. This system was playing games fine since christmas so it did work.

Any suggestions on how to fix this would be greatly appreciated. Ive been tearing my hair out for more than a month now.

thanks

 

[Guest]

hi

Ive been having problems recently with my system hanging. This only happens when im playing games. They dont even have to be particularly demanding games like Call of Duty but it still happens. The computer stays on the sound stays on a small loop for about a minute and then goes into weird moshy sound. The display remains on, recieving a signal but blank.

Now you may be wondering why on earth i have posted this thread here but i have a few suspicions about my motherboard. I have reinstalled windows twice with no avail and the ram has been tested. There was a burning smell coming from my psu however on closer inspection this was just a massive build up of dust. Just to be safe i tried another power supply however the system still hung.

Windows error reporting just reports it as an application hang so theres nothing there that gives me any information. The cpu temperature does not seem to get very high and when left running for 3 hours just in the bios monitoring temps it didnt go above 36 degrees. I also tried overclocking it however the temperature still didnt rise and it didnt crash. The motherboard temperature is reported to be about 40 degress but doesnt seem to go any higher than that. The cpu cooler is not hot to touch however one of the chipset coolers becomes unbearable to touch after about 5 seconds. I think this is the south bridge however they are all linked together with copper cooling pipes. However because of the reportedly low temperature in the bios for the motherboard i dont think this is a problem. This system was playing games fine since christmas so it did work.

Any suggestions on how to fix this would be greatly appreciated. Ive been tearing my hair out for more than a month now.

thanks

p.s. my system specs are amd phenom x4 965, asus M3N72-T deluxe, 6gb ram, XFX 5770, 3xHDD's, 2xDVD's

 

[T_T]

Have you run MemTest86+ yet?
Are your RAM spec set to "Auto" or "Manual" with manufacturers recommendations?

 

[RetiredChief]

Temps @ idle (In bois) although relavent, are not what you need to monitor. Install a temp program such as CPUID HWMonitor and run (1) prime 95 to check CPU/system temps under load. (2) download furmark and run in a window and monitor GPU temp and cpu/system temps.

Also look at min voltages while runing the programs.
Prime 95 is better for finding Mem/CPU errors under Windows (Quicker also at finding errors). I use a bootable memtest CD after a build (or changes to system hardware) prior to loading/entering operating system.

 

[Guest]

hi

i ran CPUDID and prime95 with some ermm.... interesting results. The temperatures for the CPU rise to about 60 degrees during the stress test which seems ok. However a temperature called TMPIN2 appears to move between 101 and 110 degrees. However when prime95 is closed TMPIN2 disappears from CPUID. Why does this happen?; and is this temperature reading real?

The southbridge heatsink is really very hot and i cant keep my fingers on it for more than 4 seconds. Speedfan also shows a temperature around 110 but when prim95 is closed the temperature reading goes up to 125 degrees.

Help!!! This is really quite weird stuff

 

[RetiredChief]

(1) might be a good idea to mount a small Fan on the Southbridge - Not sure how hot it should get.
(2) Not lodical that temp should increase when Prime 95 is closed, unless a fan is cutting back on it's RPMs as cpu cools ????
(3) are your voltages staying within spec at idle, and under load.

 

[Guest]

hi

well my motherboard came with a fan that should b mounted on the northbridge. It all connected together with copper pipes so that heat is transfered to the bigger cooler ( or at least that is what is claimed). However this fan is ment to only be used with passive CPU cooling ie. water cooling. I could try mounting a small fan on the southbridge however how would i mount it and secondly my graphics card is in the way of it which doesnt make life easy.

I could try moving the gfx card however that would block the bottom of the cooler and i dont think motherboards like having the gfx card in any other slot apart from no.1 if there is only one card. As for the voltages well there didnt seem to be any huge variation even under stress. There is probably about .15 volt difference either side of the specified voltage on all the rails. I have also tested this with a voltmeter so its not incorrect information.

 

[BradyT88]

Unless you have some really heat resistant fingers, I would be too worried about your south bridge. If you can actually touch it for that long without getting burnt then I doubt that is your problem. I would download everest (it has a 30 day free trial) and go into it's stability testing and then running P95. Everest will graph your temps and voltages while you run P95. I found that P95 is more heat intensive than Everest is. Also maybe use CPU-Z and make sure that your core clock isn't fluctuating and such.

I find the temp thing a little weird that it disappears. On my mobo speed fan has a 3rd temp that HWmonitor doesn't, but this temp just sits at something ridiculus like 283C as it doesn't exist. My brother's HP desktop, has 3 temps which HWmonitor labels as Systin, CPUtin, and Auxtin. The Sys and Aux temps are ridiculusly high, in the low 100's C. I just assumed they were also faulty readings until they actually increased a few degrees when I ran P95. One of those is most likely the ambient case temp and there is no way that it is a 100C inside that case, especially since it is one of those energy efficient models in a really cool room. I'm fairly certain that those sensors reading are off by 50 to 70C, as I would say yours probably is too. I'm surprised that it disappears though.

Are all of your fans on 100% or auto control? I would use speedfan to set them all to run at max all the time and see what happens and then just run P95 and watch those programs for anything weird.

 

[ChemicalAlex]

Hi

Ive run the everest stress tests and it didnt crash. Ive run all sort of stress tests and i still havent been able to replicate the fault that occurs on games. The temps are really not high at all. Any other ideas? This is really starting to wind me up now.

thx

 

[BradyT88]

How about some graphics card stability tests, I'm beginning to think that either your card is unstable or bad, or maybe your psu lacks the power to push the card.

 

[T_T]

Let's take a fresh look at this problem. Start by posting your system components. Be as detailed as possible when labeling your parts. Next to the part, let us know what driver (if applicable) you are using. And to add on to BradyT88, you can test your GPU by installing the 3DMark tool from www.futuremark.com

The reason I ask for your system configuration is because the possibility of hardware conflicts or incompatability hasn't been addressed yet. Also tell us what OS you're using, including the SP#.

 

[ChemicalAlex]

Mobo: Asus M3N72-T
Processor: AMD Phenom X4 965 3.4 GHZ
Ram: 6gb (2gb Kingston + 4gb Corsair DDR2 800 MHZ)
GFX: XFX 5770 1gb
1x Maxtor 160gb, 1x Western Digital 320gb, 1x Samsung 1TB
2x LITEON DVD/RW
Belkin PCI Wifi Card
600w Arctic Power

Windows 7 Ultimate 64bit

thats my system spec. It did work for about 2 months and then it stopped working.

 

[T_T]

Forgive me, after carefully reading through the posts, I noticed you did list most of your system spec, but nonetheless, thank you for specifying the rest of the system.

Is the PSU listed the one that had the burning smell, or the one used after noticing the burning smell?

I'm actually thinking the same as BradyT88. You do have a large power requirement for your system. How many 12v rails does your PSU have, and what are their output?

I haven't seen you post whether or not you've tried to run your system without the extra HDD; have you done this yet?

What about using a different GPU? I recently had to replace my GPU even though there were no apparent failures. I replaced my GPU because after everything (RAM, PSU, Mobo, HDD, optical drives disconnected) was replaced my system froze w/ audio loop.

 

[ChemicalAlex]

hey

i just ran 3DMark Vantage and nothing happened which is odd. Yes the psu i am using the same as the one with a burning smell but i tested it with another psu and same thing happened. I am planning to replace it but it all adds up you see. the spec for it can be found here http://www.compucase-hec.co.uk/psu_arctic_power_500-600-700.htm [/url] . I agree it is probably underpowered however it was working for a couple on months.

As for the graphics card i originally had a 4890 and when the problems started happening i put it down to the gfx card and sent it back. ebuyer accepted it had a "problem" although i doubt they checked and then they sent me a 5770 because they had run out of 4890's. I highly doubt it is the gfx card because the fault is still here even with a new one.

It can only be the motherboard or power supply. If it is the motherboard how on earth can i prove that?

 

[T_T]

Honestly, I do not know of any quality tests, outside of the manufacturer's factory, to difinitively say your mobo is bad. However, after reviewing the specs on your PSU, I am leaning toward insufficient power.

Although you have seen two months of nominal usage - without any failures, i'm guessing- you may have either:

1. Damaged the PSU because it was over burdened; or
2. Damaged your mobo because it was underpowered.

As likely as it is to say you have a bad board, I am still curious to know how your system performs w/o the unnecessary components. Disconnect the two HDD that you're not booting from. Disconnect your two optical drives. Remove the 2 x 2GB Corsair RAM modules. Though the 64-bit OS can use all of your installed RAM, the voltage to the mobo must also be increased by ~.1-.2v because of the occupancy of the RAM slots. By removing the 2x2Gb modules, you're reducing the amount of power needed. Also adjust the BIOS configuration to just the 2x1GB Kingston RAM.

The above paragraph assumes you are using a dual channel set up. After you've tried the steps above, run GTA again and see what happens.