PC shuts down / won't start

Hello,

I am having a serious and rather puzzling problem with my home-built PC. I'm not even sure how to go about searching for an answer to my problem, so my apologies if this is already on the forum.

First, here's the basics of my system:
MB: ASUS M4A88TD-V EVO
CPU: AMD Athlon II 955 (4 core)
PSU: OCZ 700MXSP (700W)
Graphics: Sapphire 5670, 512MB
OS: Win7, 64-bit
8GB RAM
Not Overclocked
Built Summer 2010

[NOTE: This problem happened once 4-6 months ago, but the next day it was working fine again. Now, the problem is not going away...]

[NOTE 2: I live in Japan with 100v electricity (2 1/2 years). Don't know if that matters - The PSU says it can handle it. The PC was built in the USA.]

It started a couple days ago. I was using my computer, playing a game (a DS Emulator, DeSmume), when all of a sudden, the computer just shut down. No blue screen, no freeze, it just powered off. I checked all the connections to see if they were secure, then pushed the power button. The power and HDD lights came on, the fans started, then after a few seconds, the power went off again. I tried it a few times, and the same thing happened.

I started thinking maybe it was a heat problem (I'm in a non-air-conditioned room), and the MB sensor shut down the machine. So I checked the fans (they were fine), and waited for about ten minutes before starting again. This time it was promising -- I got the all clear POST beep, got the the screen saying Windows did not shut down properly. I hit "Start Normally", and Windows started booting. Then, before it reached the login screen, it powered down. When I tried again a few seconds later, I got a few seconds of power, then power off.

This has been the constant pattern -- If I leave it for a substantial amount of time, it will stay on for 30 - 40 seconds before shutting down. The longer I wait, the more time I get.

Still thinking it was heat, I thought maybe the thermal paste on the CPU had gone bad, so I took off the HSF and CPU, cleaned them thoroughly, and reattached them with another helping of paste. I got the same 30 seconds or so of power.

Other symptoms:
- One time when I rebooted, it said that Overclocking Failed, even though I am not using OC. I could go into the BIOS and change things, or revert to default. I chose default.

- After I reinstalled the CPU and booted, I was given the message that a new CPU has been installed. I could go into the BIOS and change things, or revert to default. Again, I chose default.

Recent Changes/maintenance to the System:
- A few weeks ago, I added a Blu-Ray burner (LG BH14NS48). It had been working fine, but just in case, I have detached the power and SATA cables while trying to get the system to work again. (BTW, I'm not sure this drive, the -S48, is available outside of Asia)

- The day of (or a day before?) the problem started, I replaced a dead exhaust case fan. The old one had a molex connector, the new one connects directly to the MB. It worked fine at first. Again, for testing, I have detached it entirely.

- A few days before the problem, I cleaned a bunch of caked dust from the HSF. I removed the fan, used a small jeweler's screwdriver to gently pull the clumps of dust from between the HS blades, blew the remainder out with canned air, then reattached the fan. It had been working fine for two or three days before the shut-down, with temps about 2 degrees C lower.

I'm not sure what is going on. The POST, when it gets that far, gives the all-clear beep. The BIOS seems a little wonky, but it won't stay on long enough for me to get in and check it out -- Could a weak/dead CMOS battery do this? I still wonder about the heat, but the HSF is barely even warm to the touch. Could it possibly be the PSU, maybe a heat issue there?

I will be very grateful for any help. Thank you!