I'm pretty much at my wits end. My computer keeps crashing randomly when I'm playing video games, and occasionally while watching movies or ripping. What happens is I'm in the middle of a game and my screen flickers to all black, the monitor gets a orange light indicating that the computer has stopped sending a signal, and yet my computer fans (and harddrives i believe) are still running. I then have to power off by holding the button down for 5 seconds because it won't restart with the reset button. Heres my specs to help:

Windows 2000 sp3
AMD T-bird 1.2
MSI K7T Turbo2 (MS-6330)
256 pc133 ram
Geforce 3 (forget model)
Desktar 30 gig HD
WD 120 gig HD
Sound blaster Live value
HP Cd writer

I replaced the power supply with a new one (300 watts, same as original) and I've replaced my geforce 2 with a geforce 3 and I"ve gotten the same results. I really want to know if theres some way I can test what exactly happens, like failing components, when my system is crashing. I've looked into the system monitor, but can anyone give me some counters to put in a log to help me rule out what could be going wrong? From what people tell me it SHOULD be powersupply or video card, but I've also heard frying an agp slot, or maybe my proc or my mobo is just going bad... Can any of those last three be legit reasons for this to happen? Do i need more than 300 watts and thats why this is happening? I do know my cpu gets to about 57 celsius when it crashes, and my old geforce 2 got super hot, so could it be viable that my agp slot is just fried? I guess the most important question is can teh performance monitor help me solve this and/or is there anyway to know for sure whats wrong?

THanks for listening to my rant
Could be a PSU or temp problem. For what you have the 300 should be OK. What's the brand of your PSU? If it's a generic PSU there's a good chance that the values spec'ed are not the correct ones. Generics tend to list better ratings than what it actually is.

Personally I don't use anything less than a 350 watt PSU even for econo builds. And stick with name brands such as Enermax or Antec.

How many case fans do you have? Minimum should be one intake and one exhaust. 57C is a tad high for the proc. T-birds tend to run on the warm side so good cooling is a must. I don't like to see procs go over 50C. What HS/F do you have? If it's the stock AMD HS/F, junk it. Almost any aftermarket HS/F will cool better. CoolerMaster's Aero7 is one of the top performers. The "lite" version will do fine since you don't overclock. Be sure to invest in a good thermal grease like Artic Silver 3.

I would put a chipset fan on the HS of the vid card. A 40mm should be fine. It never hurts to have better cooling.

Check your RAM. Be sure it's seated correctly. Run memtest and see if you come up with any errors. I've seen bad RAM cause crashing as you described.

It could be a bad vid card. Easy to check, pop it into another system and see if it crashes. It could also be the AGP slot crapping out. If you have access to a PCI vid card, pop it in and see what happens. This is after you've ruled out the GeF3 is either good or bad.

When you swapped vid cards, did you completely delete all the GeF2 drivers before installing the GeF3? If not, delete all drivers and reboot. Then reload the GeF3 drivers. The crashing could be caused by driver issues.

There's a few things to check. Good luck with your trouble shooting. Let us know how you made out.
theres a link that says for what I use I should have a 320 watt PSU. i have a 300 watt Heralchi... for a 1.2 t-bird, is that really even worth looking at dust and heating problems on cpu etc?

Originally posted by M3thod_Man
theres a link that says for what I use I should have a 320 watt PSU. i have a 300 watt Heralchi... for a 1.2 t-bird, is that really even worth looking at dust and heating problems on cpu etc?

I disagree with power supply theory here, all failures are during high CPU utilizations. I would suspect CPU failure, not power supply problems. Take heat sink off and make sure it is clean and free of dust and the fan blades are clean too and it is working properly. Also make sure that there is some kind of thermal grease between the cpu and the heatsink and it is making good contact. I actually knew a individual years ago that had a OEM system (HP brand) with a 1.2 Tbird that had simular problems from time to time and it was traced to the CPU getting too warm. They were a hot running chip in their day.