Blue screens only during gaming!

[romelako]

My specifications:

OS: Windows 7 64-bit (OEM)
Age of System: Was purchased refurbished; been using for 1.5 months
Age of OS: Haven't re-installed
CPU: AMD Athlon II x3 435
GPU: GeForce 9500 GT/ATI Radeon HD 5670
Motherboard: MCP61
PSU: Sunbeam 480W (12v@18a)

Past Issues:
Before I purchased a new CPU, GPU, and RAM, I originally had an AMD Sempron 140 with the power supply listed above along with the GeForce 9500 GT. My system was in perfect condition. I experienced no issues with BSOD's. I decided to purchase A-DATA (2 x 2GB) RAM and popped them in. I experienced BSOD's right at startup (when I see the desktop) and I confirmed that the memory was causing the problem by popping in my old memory stick. I assumed that it was DOA and I purchased new G.Skill (2 x 2GB) memory. I no longer crash at startup.

Current Issues:
As it stands, I am unable to play Starcraft II or Heroes of Newerth without crashing mid-game and getting a BSOD. I have tried to diagnose it in many ways. This is what I have done:

* Update all drivers
* Uninstall any conflicting drivers (mainly ATI and NVIDIA drivers)
* Test the memory using memtest86+
* Use the system file checker (the one at startup)
* Use my old GPU (GeForce 9500 GT) instead of the new one (ATI Radeon HD 5670)

I had originally assumed that my BSOD's were caused by the under-powering of my ATI card. It requires at least 30a on the 12v rail and my current PSU only supplies 18a. It also requires at least 400W, while my CPU requires 95W so 400W + 95W = 495W which was a little more than what my PSU could handle. So I figured, "okay, I only crash when I play video games which is probably drawing the maximum amount of power from my PSU, causing me to crash because my PSU can't support it." So, I popped in my old GPU which requires 18a on the 12v rail and 300W of power and tried playing Heroes of Newerth. Sure enough, I crashed again.

Maybe it's the overheating of my CPU. Here are my temperatures:

Idle: 41C
Load: 61C - 62C
GPU (load): 70'ish C

Keep in mind that all of these components are at most 1 week old and have not accumulated that much dust (which rules out that possibility). I haven't been able to pinpoint the cause of these blue screens. Here are some of the minidumps from my crashes:

Some minidumps: http://www.mediafire.com/?0580gk0meuifei9

The main culprit for this entire problem is "ntskrnl.exe." I've read up on it and people say it relates to some memory issue but I have no clue where to even begin. I've tested my memory and everything and it showed no problems. Right now, I'm using my old stick of memory (that came with the computer) and I have yet to crash, but I doubt that I would get two DOA RAM in a row. I'm thinking that it's a power supply problem. When I crash, I hear a whirring noise from my computer (which is probably my power supply). Maybe it's working too hard and shutting off?

Help please!

 

[pjmelect]

I think that your problem is a memory issue, run memtest86 overnight to eliminate this.

 

[romelako]

I think that your problem is a memory issue, run memtest86 overnight to eliminate this.

I did run memtest86 and it came clean.

 

[jimmysmitty]

Check the BIOS and make sure the memory is set to the correct timings and voltage per the spec on the RAM. A lot of BIOS tend to mess that up.

I had sort of the same problem with my current build. Everything was fine but I would load a game and it would crash after a few minutes. Set my memory to spec and all works great.

If that doesn't work it is possible that the old RAM corrupted some files when in use and you might need to do a Windows reinstall.

 

[romelako]

Check the BIOS and make sure the memory is set to the correct timings and voltage per the spec on the RAM. A lot of BIOS tend to mess that up.

I had sort of the same problem with my current build. Everything was fine but I would load a game and it would crash after a few minutes. Set my memory to spec and all works great.

If that doesn't work it is possible that the old RAM corrupted some files when in use and you might need to do a Windows reinstall.

My BIOS is really old so I cannot adjust timings or voltage. I made sure the timings and voltages matched that of my motherboard before I purchased them. I know they're compatible.

 

[jimmysmitty]

My BIOS is really old so I cannot adjust timings or voltage. I made sure the timings and voltages matched that of my motherboard before I purchased them. I know they're compatible.

Its not about compatability. My mobo is compatable with my memory. Its the fact that some BIOS tend to put the wrong voltages or timings which can screw things up. If your memory needs 2.1v, it needs to be set at 2.1v and most BIOS will set it at 1.8v due to JSPEC.

Whats your memory specifics?

 

[romelako]

Its not about compatability. My mobo is compatable with my memory. Its the fact that some BIOS tend to put the wrong voltages or timings which can screw things up. If your memory needs 2.1v, it needs to be set at 2.1v and most BIOS will set it at 1.8v due to JSPEC.

Whats your memory specifics?

1.8v with a 5-5-5-18 timing (which were the specifications of the older memory).

 

[jimmysmitty]

Does the BIOS at least show the memory voltage so you can verify that its set properly?

 

[romelako]

Does the BIOS at least show the memory voltage so you can verify that its set properly?

Unfortunately not, but CPU-Z shows that it's operating @ 1.8v.

It's funny, because after removing the memory, I haven't crashed at all. It seems as if no memory is compatible with my mobo. The last memory, as I said, would crash me right when I saw the desktop. What could be the problem with this memory? It was shown to be clean using memtest86+...

 

[daship]

If you haven't run memtest86+ for atleast 8 hours straight, you havent ran it enough.

I would try a fresh install since you have mixed ati and nvidia, or at bare minimum use driver sweeper to get rid of all nvidia stuff.

 

[romelako]

If you haven't run memtest86+ for atleast 8 hours straight, you havent ran it enough.

I would try a fresh install since you have mixed ati and nvidia, or at bare minimum use driver sweeper to get rid of all nvidia stuff.

I did leave it on for 8 hours straight and it showed no errors!

I don't think a fresh install would help me as this is a hardware issue and as it stands, the new memory is causing me to crash even though it's clean. I'm so confused.

 

[romelako]

and what are the temps with all this going on, your reports seem like it's running slightly warm to hot..

The temperatures listed above in my original post. The memory is definitely the culprit but I don't understand why. Again, memtest86+ showed no errors. This is pretty irritating.