New PSU, still getting hard crashes

[udprod]

----------------------------------------------
This is my system:
ATI 4850 1GB
Intel (R) Core(TM)2 Quad CPU Q9300 @ 2.50 GHz (4 CPUs)
Win XP 32 Bit
3 GB RAM
2 Harddrives @ 7200 rpm
-----------------------------------------------

Dear TH community,

I am writing here today because I after 2 months still are getting hard crashes when doing something graphical (such as a video game). I used to have a PoS 350W PSU, which I have replaced three times. And today my Cosair 520HX arrived (which should be waaaay more than I need), but my PC still hard crashes.

The crash
This has never happened outside graphical applications, such as games. When playing, my screen suddenly turns black, or is filled with horizontal lines of color, and the sound played the last ½ second before the crash starts looping like a machine gun. My computer does not respond at all at this point, so I have to reboot it by holding the power button down.

What I've tried
I've went through various PSU's, going from 350W to 450W to 520W (bad brand, only 16 amps on the 12v rail) and now finaly my Corsair 520 HX.

I've increased the fanspeed for my graphic card to 100% via CCC, up from 50%, and my gaming temperatures are between 66 and 78 degrees (the higher it gets the higher "chance" there is of it crashing. Though I've had them happen at 69, 73 degrees etc aswell).

I've tried downgrading to my old nVidia 8600 card, which gave me no errors. But that uses no power compared to my 4850 (at that point I upgraded my PSU).

I've tried about 10 different Catalyst Drivers; both with and without Catalyst Control Center.

Notes
The airflow in my case is pretty bad, but I might be able to transfer my components to a better case.

I get the crash more while playing advanced graphic games (Oblivion on Ultra high as an example) than on older games (WoW).

Every time I get a hard crash, 2 errors spawn in my computer log (Control Panel -> Administration -> Computer Administration - Log - Program) (the second one is translated from Danish, since I was unable to find an English error to copy->paste):

Source: MSDTC Client Event ID: 4386
MS DTC cannot start becuase it failed to initialize the log file for it's
first usage after a cluster install. Please try to restart MS DTC. Error
Specifics: d: \srv03rtm\com\complus\dtc\dtc\adme\uiname.cpp:7825 , CmdLine:
"C:\WINDOWS\cluster\resrcmon.exe" -e 876 -m 880 -p 972, Pid: 344 No
Callstack(O)

Source: COM+ Event ID: 4691
The environment could not initialize transactions, which is required as support for transaction components, while running. Control that MSDTC is running. (DtcGetTransactionManagerEx(): hr = 0x8004d027)

What I think could be the cause

Broken/damaged/defect GPU or CPU
Incompatibility
Bad air flow in computer case
Overheating
Me


If I need to provide additional information, please tell me asap!

-------

Thanks for taking time to read my post, any help, or tips, are greatly appreciated. I also apologise for any spelling or grammatical errors if found.

 

[kufan64]

It shouldn't just flat out crash from overheating at only ~70C, and it doesn't seem likely that it is a problem with your PSU.

The crash where it goes black and loops sounds like that used to be a pretty common error for me, though I haven't seen it in a while. I'd investigate either your video card, or your RAM; though it may also be a broken Windows install.

You said that it works fine with your old video card which should automatically make your new card your prime suspect.

I'm sorry I can't offer you a whole lot of help, but maybe just bumping this back to the top will attract someone who can give you a better answer. I wish you well in finding out the cause of your errors.

 

[swifty_morgan]

Your video card ( 4850 ) can be within temps and still have a bad memory module. Did you try UNDER clocking it using the CCC control panel ? ..... although it might not show a difference if a bad module exists.

 

[shortstuff_mt]

Did you use something like Driver Cleaner to completely remove your old Nvidia drivers before installing the ATI drivers? Not doing so could cause problems like you describe.

 

[Outlander_04]

get a case with good air flow .

 

[musicmanivi]

sounds like a bad video card to me - 70c is hot, but not overheating. see if you can try your card in another machine and duplicate the problem.

 

[udprod]

Hi.
I recently tried replacing my card with my very old 8600 GT, and my pc has ran with no problems for many hours now. Seeing it as I have only adjusted the fan speed, is there still warranty on the card?

Thanks for all of your answers.