Erratic, but steady degeneration

[UndeniablyRexer]

My Computer:
M4A89GTD PRO/USB3
AMD Phenom II X4 965 BE
Nvidia GTX 260 core 216
OCZ Plat 4GB DDR3 2000
Antec TruePower 430W
Crucial C300 RealSSD 64GB (Windows 7 x64)
Maxtor 250GB (Ubuntu 11.04 32bit/Program Files/Games)
Hitachi 2TB(Storage)

How It Fails:
It will blue screen or freeze or restart(and not post afterward). It usually will not turn back on immediately, no post. If I wait a few minutes, then it will turn on, but soon after will fail again. How long before it fails seems to be proportional to how long I leave it off and inversely proportional to how far along the failure process it is. By this I mean that, after the first BSOD/freeze/restart, if I let it sit for 12 hours, it will stay on for ~7 hours; a few days later, though, if I let it sit for 12 hours, it will last ~1 hour.
Freezes and restarts will happen at any point in the boot sequence, even idling in BIOS.


Timeline:
Went through the few-day failure routine.
RMA'd motherboard.
Worked for ~3 weeks, then went through failure routine.
Advance RMA'd motherboard, so I got a completely different one.
Worked for ~ 2 weeks, then went through failure routine.
RMA'd and received new CPU.
worked for ~1 week.
After the start of the failure routine, I replaced the PSU with an Antec BP550W; this didn't seem to effect anything.

Current Status
I left the computer off for ~ 13 hours last night; it restarted(with no sequential post) after ~30 min.

Possibly Helpful Info:

• The original motherboard was purchased open-box(yes, I've learned my lesson), but was replaced during the advanced RMA after the second failure.
• Each time it's failed, I've broken it down to just onboard video, CPU, RAM, mobo; it still would not turn on. Also tried different RAM with no luck.
• After it fails, It won't produce any beep codes, even with no RAM.
• Even when my computer is working fine with no BSOD/hard locks/restarts, if I tell the computer to shutdown, it will restart. The only way to turn it off is to hold the power button.
• The situation degenerates steadily over a few days from normal operation to no post.

Something has to be breaking; I think it's the mobo and/or CPU.

What's confusing me is that after I RMA'd either the motherboard or the CPU, I had a working computer for a couple weeks. So by RMA'ing the motherboard or CPU, something gets fixed, but, then by using the computer, this thing breaks.

I considered the PSU to be the cause, but even after replacing that with one that should be able to supply enough power, the degeneration continues.



Thank you very much for any help. This has been a 6 month problem; I desperately want to fix it.

 

[Route44]

If you can please do the following...

How to find and post your Minidump Files:

My Computer > C Drive > Windows Folder > Minidump Folder > Minidump Files.

It is these files that we need (not the folder). Attach to your next post the five most recent dumps. Notice the Manage Attachments button at the bottom when you go to post the next time. You can Zip up to five files per Zip; if you only have one or two you don’t need to zip them, just attach as is. Please do us a favor and don’t Zip each one individually.

 

[UndeniablyRexer]

Thanks for the response. Sorry for taking so long, the computer won't stay on for more than 30 seconds so I had to pull the hdd.


There were only 2 dumps, those are uploaded.

 

[Route44]

The Nvidia driver nvlddmkm.sys was cited as the cause of your system crashes. I suggest doing the following...

1. Download Driver Sweeper free version to your desktop screen and install.

2. Download the latest diver(s) for your video card but don't install them.

3. Uninstall your video card drivers and reboot your PC into Safe Mode. Run Driver Sweeper and remove ONLY the video card drivers. I had someone use it on their chipset drivers! If it doesn't find any video card drivers that is quite okay; just leave all other drivers alone.

4. Reboot and install new video card drivers.

* You might have to boot into Safe Mode to do this.

 

[UndeniablyRexer]

Thanks again for the response, but I don't think that would do anything.
The windows install on it is less than 3 weeks old, and I installed the most recent drivers.
Also, the computer won't even get to windows, it's lucky to begin booting windows.

My system caused one or more of the following to break: CPU, RAM, motherboard.
I know this because a week ago everything was peachy; now, it is lucky to post.
Right now I have the mobo sitting on the mobo box with the RAM, CPU, and speaker installed, PSU connected, and monitor connected to onboard video. It will not post, not even a beep code.

 

[Route44]

Have you tested the psu with a digital multimeter? Does the psu even start up?

 

[UndeniablyRexer]

I just tested the PSU with a multimeter. Tested the 5VSB with it off, got 5V. Tested everything else while it was "on" and they were all correct. Not only were the voltages within a good range, but they were very close. The 12V measured at 12.14V, 3.3V at 3.33V, 5V at 4.99V. Switched my multimeter to AC to measure the ripple voltage and everything ranged from 3mV to 7mV.

However, this was not done under load
Is it possible that when the PSU is stressed, the voltages change to dangerous levels?

 

[Route44]

Yes, I believe it needs to be tested under a load. Can you get your hands on another power supply that provides more wattage to test it out on your system?

By the way, in order for your motherboard to run your memory at 2000 it has to be overclocked. Have you done so?

Check the voltage of your memory as set in the BIOS and compare it to OCZ's designed voltage specs. Do they match?

 

[UndeniablyRexer]

After it showed the first sign of failure, I replaced the PSU with a 550W Antec; it continued to degrade.

I manually set the timings and voltage for the RAM. I believe the motherboard ran the RAM at 1067, though.

I can't test the system under load anymore, though, because it won't stay on long enough.

 

[Route44]

"Right now I have the mobo sitting on the mobo box with the RAM, CPU, and speaker installed, PSU connected, and monitor connected to onboard video. It will not post, not even a beep code."

This one of the most tried and true as well effective forms of testing/diagnosing that there is and the fact that it won't even beep let alone post is very disconcerting. Is this with your Antec 550?

 

[UndeniablyRexer]

This one of the most tried and true as well effective forms of testing/diagnosing that there is and the fact that it won't even beep let alone post is very disconcerting. Is this with your Antec 550?

Yes.

I should mention that if I leave it alone for a couple hours, it will boot up and possibly login, but very quickly will BSOD/hard lock/restart, and subsequently not post. If I restart it--hold the power button, ctrl-alt-del, etc...-- before it fails, it will post again. But if it fails, then it will not post until I wait a few hours.

 

[Route44]

If the original psu became faulty a bad psu can cause damage to other hardware. I wonder if the motherboard is damaged.

 

[UndeniablyRexer]

I'm pretty sure the CPU and/or mobo is damaged.

What's confusing me is that after RMA'ing either the mobo or CPU, it'll work for a couple weeks. It sounds stupid, but it appears like the CPU and mobo are half broken. After RMA'ing one, one is fixed and the other is half broken, so it works until the PSU half breaks the other one.

Another possibility is that the PSU is just frying one component and I got lucky 3 times by picking the right one to RMA. Although, Asus told me they didn't do anything to the mobo the first time I sent it in.
Last December: Computer won't post, I RMA mobo, they don't do anything and ship it back, computer works for ~3 weeks. Definitely a WTF moment.

 

[Route44]

Then it looks like it is pointing to something else.