PC freezes randomly

[DJ 24]

I've searched both, the net and this forum for similar problems but none have had the same symptoms.

My problem is that my PC crashes randomly. Before crash i have been using it just like normal, in either gaming or anything else like browsing the net, and all has been fine. Suddenly it just freezes. There are a few facts i should point out.
- Just as it freezes, i hear a zap/chhhh type of sound VERY fast from my speakers.
- Picture stays on screen(s) but doesent move at all.
- USB devices lose power.

When it crashes, the only way to get back up is manual reboot. After this everything works just as nothing has happened, and this might continue even for a month or two. And then it crashes again. Now it has crashed allready a few times, probaply two, but the latest one was weird, not like the others. I lost power from my main screen (i have two graphics cards NOT connected together, working independently.), but the second screen stayed on just normal, i could even move my mouse over to it and use it. Also all the USB devices stayed on, BUT, the sound coming from the speakers in the crashes before occured, just a bit smaller, shorter, and not so aggressive. I shut my pc down (pressed the power button and waited for windows to shut down normally), and powered it up again, but there was no picture on my main screen. I shut it down again and unplugged the power plug, waited a few seconds for the leds on my motherboard to fade and powered it up again. This time the picture came back.

The first time my PC crashed about 2-3 months ago, second time 1 month ago and the latest screen crash occured a few days ago.

A few more facts
- I've had issues with my main graphics card. These started before the crashing. Sometimes this Asus smart doctor screen pops up and warns me about a fan error, saying "fan replacement may be necessary". Takes a few seconds and the text changes back to where it was, "Everything is OK..." Wait a few minutes or hours, and the error pops up again. A few times it has popped up i've taken a look of my graphics card, trying to see problems with the fan but it seems to be running just normal. Normally the error lasts only a few seconds, but once it lasted propably about 20 seconds, until i opened the temperature page of the smart doctor program and noticed very high temperatures still rising, (68 core, 62 memory etc. (notice, i was NOT gaming!)) and i took a look at the fan again, and it had stopped. I opened settings in the smart doctor, and choose "manual fan speed" instead of "automatic fan control", and set the value to 100%. After this i clicked apply and the fan turned on again, with temperatures dropping. A day after i chose "automatic fan control" again and it worked just fine eccept the error pop ups.
- I bought a power meter to see im not running too many watts through my Antec 500w power supply. It has overload protection option cutting power when consumption rising over given value. This i set to 500w. It also displays the power consumption in various formats, naturally. I have checked it a few times and the Max Watt value ever passed through is 372, so i guess im not overheating my supply.

My setup:

Hardware...
Asus engtx 285 - graphics card
MSI 8500gt - graphics card (only for my third display, it couldn't be connected to the primary card due to this having only two outputs...)
MSI P45 Platinum - motherboard
Intel Q6600 - processor
OCZ reaper HPC DDR2 1066mhz - memory modules
Antec NeoHE 500 - Power Supply
Seagate Barracuda 500gb - HDD
Western Digital Caviar Black 500gb - HDD
Samsung 400gb - HDD
Samsung 160gb - HDD

And software....
Windows 7 Ultimate 64bit
Windows XP Home Edition 32bit
Ubuntu Linux 32bit

All operating systems are located on different HDD:s, Windows 7 being the most used located on the Seagate barracuda.

As maby you can guess from my setup, it has been "improved over time". So i think its wise to write down the ages of different components. The list is based on memory and isn't totally accurate:
(listed from newest to oldest)

Western Digital - HDD = ½yr
Asus engtx285 - Graphics card = 1yr
Seagate Barracuda - HDD = 1yr
MSI P45 Platinum - motherboard = 1½yr
OCZ Reaper - memory modules = 1½yr
Intel Q6600 - CPU = 2½yr
Antec - Power supply = 2½yr
Samsung 400gb - HDD = 3yrs
Samsung 160gb - HDD = No age confirmation (sata 2)
MSI 8500gt - graphics card = No age confirmation

I hope this information is enought, im really worried of this freezing/crashing, mainly about the fact that you never know when it starts braking your hardware. At least in my knowledge.

Yeah, and some of you propably will suggest i just build a brand new PC with no old components, but in my current situation this isn't possible and now that i got my PC running just smooth, and the new Asus engtx 285 really kicks the pixel a**, im not very willing to just throw this all away.

_______________________
"Edit: Added a bit info on the screen crash and crash dates"

 

[shortstuff_mt]

Freezing problems are often RAM related. Your RAM requires a lot of voltage to work correctly. It's rated at 1066MHz with 5-5-5-18 timings and 2.2v. Have you manually set all those values in the BIOS? After you get the settings correct then I would perform an overnight run of Memtest86+ to test for RAM errors.

 

[DJ 24]

When i open my BIOS and check the values it gives me 5-5-5-14-24-6-3-3-3 (tcl-trcd-trp-tras-trfc-twr-twtr-trrd-trtp). I really have never set any voltage or timing for memory before so im not sure if i got you right or if i read the values from the bios right. The Bios voltage option was grey and read [AUTO] and it didn't let me change it.

I also noticed that when i open Everest Ultimate Edition, click the Motherboard tab and from there SPD, it gives me values about my memory and voltage. Here it reads they are set 5-5-5-15 with a voltage of 2.1v. To me this makes no sence, how can Bios and everest give different timing values..?

One more clue to give, when i shutdown my computer, if i start it right up again, it freezes to Bios loading screen to the part where its checking my memory (according to the Debug leds on the motherboard and my motherboard operating manual). At this point i have to shut it down and unplug for it to work again. This occured just now as i was trying to get a hang of the Bios values and had to restart a few times.
_________________
Edit: forget the are you sure part, i checked the http://www.ocztechnology.com/products/memory/ocz_ddr2_pc2_8500_reaper_hpc_4gb_edition page and found the same values, sorry about that. Now i also remember this one thing that occured just as i had built the new system 1½ yrs ago. I had some memory issues, i think they were running too slow (800mhz), and i handed out my PC to my friend. He said my motherboard read the values from the cards wrong and the voltage was set wrong, and he set it to 2.1v, but i dont remember him saying anything about the timing.
_________________
Edit: one more thing i forgot to say, the values i wrote Bios gave were from Memory-Z (button f5 in bios)

 

[DJ 24]

I think i found the right setting in the bios, i changed all values from auto to the ones suggested:
Cas latency (cl) 5
tRCD 5
tRP 5
tRAS 18

I saved and exitted, opened Bios again to check the values where still there and exited. Still when i open everest it says i have 5-5-5-15 values. Yes, i refreshed it.