[romtim]

Hello, i will now describe my problem.
Around a week ago, i had my pc - core i7 920@2,66ghz, my 6gb 1333mhz ocz ram, my 600 watt ocz power supply (cost alot), and my asus p6t motherboard. I was on windows 7. What happened was, i have had my pc freeze while watching a video (the video continued, the mouse and keyboard unresponsive), i had my pc freeze while playing a NOT DEMANDING game, and i have had several freezes while sitting in msn messenger, icq, and facebook. The last time i had a glitch, was me getting strange artifacts all around my screen, and the blue screen of death while typing in icq messenger to a friend. So i just restarted and boom, i got glitches on my boot screen, i got dots all around bios. my video card was geforce 8800gts 512mb. Generally, thats the problem of the video card, i have tried to clean it (it was very dirty), but it didnt work. So i thought ok... i cooked my card, so i put another one. I took an older one - 7800gt i think, not sure - factory overclocked. It has no fan, just a heatsink, and is pretty lazy to do its job (compared to 8800 gts). It was fine for 1 day. The next day(yesterday) i have had 2 glitches with it - for half a second i had red vertical and horizontal lines, and they were gone straight after, with no impact on performance whatsoever ( i was sitting in facebook at that moment. The next time was while i was again in ICQ messenger, in the evening). I was like... umm.. ok... thats not so bad... But today i had my pc freeze once again in the morning (coinsidence i have said to myself), and then just while sitting in forums again, i have had this terrible color pixel line glitch it f---ed up my desktop and went into reboot. I was like OMG, i poiwered down my pc, my card was hot enough (even tho i didnt use it.) TO NOTICE - i have installed software to monitor my cards temperature, it was always fine like 50-60c idle, 70-80 load. So there was no general problem with it. I was like stunned - after all it seems its not the video card... My processor was checked at prime95, it was running it stable for an hour. I have started a ram test on 1 piece of 2gb, it ran sucessfully. Then i inserted all the pieces, ran ANOTHER test( a completely different one, from a cd REANIMATOR). And it like froze (unresponsive keyboard and mouse), the seconds changed once in 10 seconds, and nothing happened. It had no errors before this bs---. To mention - my vcard was very hot during that moment. I have rearranged my ram sticks, i have defaulted bios, i have checked my voltage on my psu - it was around 12.25v for 12, and just normal for every else. I was like.. ok maybe its fine.. maybe i should go and put a voltage checking program to shutdown the pc if something goes wrong. I loaded windows, and at the login screen it just froze. unresponsive keyboard and mouse. No video glitches however. So i just turned off my pc and now i write this post from another pc.

If its the ram, why does my video card cook (the other one is toasted, finish, caput), this one delivers glitches sometimes in windows.
If its the voltage, why does it show good parameters with multimeter, and in bios? Could be still possible that it delivers high voltages instantly for few seconds - enough to overheat the card and cause video glitches.
If its the video card, why do 2 video cards in a row start glitching and cooking?
If its the motherboard... ? I must be sure on this one to try and claim that its defective.

I have had all this equipment (except for video card) for about 2 months, its all new and shiny. But i got no idea what is wrong and what to do.

Ill however mention some other strange stuff happening since i bought this pc. Some 'click' sound and my pc starting to act more silent (hard drive parking probably), happened during gaming and not only. Some strange windows freeze, and at boot it shows only 2 sticks of ram = 4gb. While i have 3 inserted. Happened 2-3 times. Fixed by restart. If i turn on my volume in windows to max, i can hear some strange noise in my headset while resizing windows and doing some other actions.

SO, i would like to know a possible fix, or your opinions of what's defective. I dont want to cook another video card, and i want to find the defective part of my pc and replace it.

Thanks

[win2kpro]

It doesn't sound like a voltage problem to me it sounds like your chassis has poor cooling design.

[Tanis]

I agree, it does sound like a temperature problem. The temps you mentioned (50-60 idle / 70-80 load) is that for the 7600GT? If so, that is little higher than I would expect for that card, I would expect a 7600GT with stock cooler and average airflow through case to idle at around 50 and maybe go to around 70 tops under load.

What fans do you have in place in the system (CPU, GPU, case fans etc?).

You say you have had freezes running a RAM test with all sticks in? Try using memtest86 (www.memtest86), test 1 stick at a time and test each stick in the same slot on the motherboard. If they all pass in the first slot then move one to the 2nd slot and test, then test 1 in the 3rd slot etc. See if you get any errors during those test.

If they all pass in the first slot you can be fairly sure the RAM is OK, if they then subsequently fail in another slot then it may be that slot on the motherboard is defective. Just because parts are new doesnt't mean they can't be defective, it does happen.

[romtim]

hello, well first of all, this gpu does not have a cooler at all. it has a heatsink, thats all. It is Gigabyte gv-nx85t512hp. i get 50-65 idle, 70-85 under full tension.

i have a case fan, a stock cpu fan (its okay, no problems with cpu). I did mem test 86 for all 3 at the same time (where they are) and it finished with 0 errors, and its all fine.

Could it have been an error that i put these fairly old video cards in port PCI Express 2.0 16x (16x)? While i have a port PCI Express 2.0 16x (4x - in which i situated it now).

[win2kpro]

If you just have 1 chassis fan (most likely exhaust) and a processor cooler fan you just have a chassis with poor thermal design.

Preferably you should have an intake fan or fans at 1.5 to 2 times the exhaust fan in CFM. You should also have a large air vent on the removable side near the graphics cards. It is also preferable to have a processor air duct to funnel ambient air to the processor cooler.

Here is a chassis I use quite often although I have to install a larger power supply than than the one that comes pre-installed in the chassis.

http://www.newegg.com/Product/ImageG...Power%20Supply

I set up these chassis's with two Antec 92mm variable speed front fans running on medium speed and set the 120mm exhaust fan that is pre-installed also on medium speed.

Click on the thumbnail pics on the side of the page and you can get a good idea of how it is setup.

Look for a chassis with a similar thermal design.

[romtim]

Like i meantioned, there is no problem with temperatures. my motherboard is fine, my cpu is fine, and my video card temperature is normal. There is no reason for me to lower the temps even more down. The card cant break at 80c gpu load and get artifacts. So the problem isn't in this. If i would have extra cash i would buy myself a Good new gtx460, a new cpu fan, few other fans for my system in whole. but i dont have the money to do all this, all i need is to find out the cause. To exchange it for a new non defective part, or buy a new one.

[jiml8]

I agree that it sounds like heat.

Take the cover off the box, set a fan next to the case blowing air into the open case, run the system. Then see how hot your video card is getting.

As for your insistence that it can't be heat, I have this to say.

None of us are there. None of us see your system. None of us can know - for sure - what your problem is.

But you describe a set of symptoms, and you get responses from people. It is up to you to decide if those responses are from knowledgeable people, and if they make sense - totally your choice to make.

But you now have three people saying the same thing.

[Tanis]

OK, so the card you are using is an 8500GT, not 7800. With that in mind then around 50s idle seems to be fairly common. Having said that, personally I would still be looking at aftermarket cooling if I was running the card and it was regularly reaching around 80 under load. Higher temps will reduce the life of the card (even if those temps are within the design parameters) and will also increase the likelyhood of component failure.

Originally Posted by romtim

The card cant break at 80c gpu load and get artifacts

Yes it can, it could die and artifact at lower temps than that if there is a problem with it!

As an example, my old XFX 8800GT (factory overclocked) used to lock my PC totally and at random if the GPU went over 75 degrees, that is well within the operating temperatures of that card, but this only happened with games! I did numerous tests on the entire system and passed them all, it also never crashed when stress testing any components either individually or together.

As jiml8 has said, we can't say with 100% certainty what the cause is, we can only advise what we think could be causing it based no the information you have given us.

Have you tested the voltage outputs of the PSU to make sure there is nothing funny going on there?

[romtim]

hello, yes, i have tested, and everything is ok (within range).

[win2kpro]

Can you make the machine duplicate your problems laying on the side with the side panel removed?

If so, the only way to accurately diagnose your problem is first with a calibrated multimeter with needle probes take readings at the mainboard plug with the machine under load and record the readings on the +12v, +5v and +3.3v rails. Don't neglect to take separate readings on pins #11 (+12v), #12 (+3.3v) and #23 (+5v). This is the +4 connector on a 20 +4mainboard connector. If all the readings are within + or - 5% of the rated output then you can consider the power supply good.

If all the electrical output readings are within the specified range then you need to take an infrared thermometer and start taking readings on the motherboard and components looking for hot spots.

It's hard to correctly diagnose a problem in a forum. You really need the machine on a bench with the proper diagnostic tools to find out what is actually going on with the machine.

[romtim]

thanks, actually, i did test with multimeter, its all fine, yes. i will try to find an infrared thermal diagnostic tool

[sickcars]

Have you done what somebody else has said above?

Open the side of the computer and point a desk fan at it on high. then run the computer and see if it lasts longer without crashing... also take note of the temps and post them here.


As for your temps they are on the high side... sure the parts are apparently made to run at higher temps but regardless you want your parts to run as cool as possible so they have a better chance of not messing up. Now I realize you think its not a heat issue but this is what you have to do to troubleshoot the issue. You must test & try everything and in this case you need to try the fan suggestion above.

Because if your not going to try it and listen to us and just tell us no temps are fine then we will just stop trying to help you and go and help somebody else that will try our suggestions.

[romtim]

ok, thanks, i have attached an air cooler with a thread works fine, at load 74C instead of 84, cools down very fast to 55-60c if load dissapears.
i hope that was the cause, if not..

[sickcars]

Originally Posted by romtim

ok, thanks, i have attached an air cooler with a thread works fine, at load 74C instead of 84, cools down very fast to 55-60c if load dissapears.
i hope that was the cause, if not..

Ok so has your computer had an issues since? any freezing?

[romtim]

not yet.. ill say if it appears again