Graphics Display problem- Tough one!

[Stephen D.]

First, thanks for any help on this, it is much appreciated.  I'm experienced with computer troubleshooting, but this has got me stumped, and all the reading of forums can help me thus far.

Recently I upgraded my system, and have started to have a problem I havent seen before.  The problem occurs in any graphics intensive game (DoD: Source, Earth 2160, anything really) at any random stop during regular gameplay, occuring sooner [5-10min] on higher settings and eventually [20min-40min] on lower settings.  The problem, no matter what the game, is always the same: the center point of the screen suddenly has a few dozen lines randomly pointing into with each triangle filled with a color-test style pattern (hot pink, blue with green checkers, complete randomness).  This will go L.S.D. on me for about 5-10 seconds, then the computer will just shut itself off.

First thought: Overheating- probaly not the situation.  The same thing occured if I took the side off my case and put a big fan pointing straight on my CPU/GPU, didnt change a thing.  The temperature of my CPU is fine, and the GPU isnt great, but isnt bad either.  Overall, I have a well-cooled case with 4 X 120mm fans.

Second: memory.  I ran the test from www.memtest.org or whatever overnight, which came back 100% perfect.  If there is anything else to test/investigate there, I am all ears.

Third: Drivers- I installed the new drivers, many, many times.  Different versions, ect.  The only thing I havent tried extensivly with drivers, is attempting to install an old version of them.  

Other things that people have suggested is my PSU, which is a brand new 450w generic.  Dont just tell me to get a new one please, before I start replacing random parts I need to be sure it will fix the problem.  Thats what everyone else has done.

Well thats the status of my problem, I am stuck here.
Platform:
WIN XP x64 Professional SP1
Athlon 64 3700+
Sapphire RADEON 9800 Pro 128mb AGP running x64 Catalyst drivers
K8N Neo2 Motherboard AGPw/ nVidia chipset
1.00 GHz Kingston PC3200 RAM
PSU: 450watt generic

Thanks again,
Stephen
msn: [email protected]

[2k_dummy]

Other things that people have suggested is my PSU, which is a brand new 450w generic.  Dont just tell me to get a new one please, before I start replacing random parts I need to be sure it will fix the problem.  Thats what everyone else has done.

No one can be absolutely certain what will fix the problem, but I suggest you take the advise and replace the PSU with a high quality unit. The problem with many generic PSU's is that they are not well built. Their design allows full power only if the operating temperature stays below a certain point (30C, which is nearly impossible). As the PSU heats up, it will lose approximately 10% of its power for every degree above 30C. A quality, well built PSU will deliver full power up to or above 50C. So heat might still be the culprit, you're just looking for it in the wrong place. The PSU is a computer's heart, so don't try to run it with a pace maker or artificial heart. Give it a real one. The machine you have deserves the best your wallet will allow.

[Frankie]

I would look at the Graphics card.

Is there any compatibility issues with the mainboard and the card.

Have you reseated the graphics card and reinstalled the drivers. Have you run a test in dxdiag and a stress test using 3d mark or higher.

[Dell4700]

Stephen,

I don't think you told us what type of recent upgrade you did.  That is a good starting point for suspicion.

-  you say you have a generic PS.  Did you just change cases?  I've looked at some ads for this MB and recommended PS ratings are 500 watts and above!  http://www.tigerdirect.com/applications/SearchTools/item-details.asp?EdpNo=1508928&CatId=1569  As the previous advice suggests,  you should make sure you have an adequate PS.

-  the graphics card you are using will normally overlap the 1st PCI slot.  Are you trying to use anything in that slot?  

- you don't have to replace anything to test the above possible problems.  Simply remove any unecessary cards, bells & whistles (modem, lan, scsi, etc.) - anything not needed for your game play.  If you use a separate sound card, try removing it and switch back to the onboard system.  If you get your system working without problem, just start putting the cards back one at a time.  If it gets screwy when you do this, then it is probably your PS (likely) or the card you just put back (less likely).

- also previously suggested - test your system under stress

Dell

[-Oscar-]

I've had the exact same problem for a very long time untill recently when I upgraded my chipset to the latest version. After this, my computer started to shut it self of even outside games.

I have a very similar computer as Stephen's.
Radeon 9800 SE and a K8T Neo2 motherboard. My power supply is also at 450w which I highly doubt is the cause of this problem.

I'm currently using a geForce2 MX 400 instead of the Radeon 9800 and it works fine. Could it be some sort of conflict between the motherboard chipset and the radeon graphics drivers?

[Gustaaf]

It seems that the video buffers gets corrupted somehow.

Investigate from that perspective.

1) Static built up. -Earth your system additionally thru the chasis.
2) cooling is insufficient or air-flow outwards is limited...
3) try using newer or even older drivers for your North-bridge CPU on the motherboard.
4) If you have done any bios tweaking, then set it to auto speeds etc.
5) Your 1GHz system memmory might need passive cooling.