I'll start off with my current hardware installed:

Intel Core 2 Duo 6850 stock now, when OC 3,780 @ 1.45v
Asus P5E
1 GB Corsair w/led
2GB Crucial w/led
2x 150GB WD Raptors satas
1x 500Gb WD sata
Nvidia 9800gx2 when OC 716/1829/2106
Seasonic 650w 4 rails @ 18amp each 3.3 at 24a and 5v @ 30a say at 25 degrees full load
Creative x-fi
1 Lg Dvd-rw sata
1 Lg dvd-rom sata
2 usb dongles for mouse and keyboard one is bluetooth
xbox wireless reciever
4x 120mm fans
1 blue neon light with the power box

When the system is overclocked and i'm playing games, I get bsod between 30 min and an hour of play time. Temps on cpu never exceed 60 excpet on hot days and I clock it down. My 5v sits at 4.8 when all is overclocked, the 12v is at 12.10 and the 3v is at 3.30 from 3v at 3.31 and the 12v at 12.15 and the 5v is at 4.94 ish when not overclocked. even at 3.6 is still get bsod

So is my power supply not powerful enough or could the power supply be overheating? Or maybe it's just my overclocks?

Hope you guys can help me out, I have a few powersupplys in mind as I think that is the issue.

http://www.canadacomputers.com/index.php?do=ShowProduct&cm...cid=PS.808

http://www.canadacomputers.com/index.php?do=ShowProduct&cm...cid=PS.808

http://www.canadacomputers.com/index.php?do=ShowProduct&cm...cid=PS.808

And any other recommendations from the store listed above because if I get a new powersupply I wanna be able to get it this weekend.

What are the temps on your video card? I'd be surprised if it was your PSU...4.80V on the 5V isn't great but it is within spec.

You're saying the temp on the CPU is fine, but to be honest, 60c at that OC is enough to cause leakage, shutdown or even damage to the CPU.

I have a feeling its more serious than your PSU.

However, in general, a shutdown after a time of stress would indeed indicate a temperature problem, not PSU.

As Beavis said, check the vid card temps, you might also want to see if your ram or chipset are getting very hot after extended periods of gaming

When I overclock the card I turn up the speed of the fan to 77% and in 3dmark highest was 75 on card 2 and 72 on card 1, chipset is a good call, in everest is says its at 42 right now stock. On a side note when I had my 8800gtx I could have the cpu overclock with no issues, since then I have added the 9800, another hardrive and 2 sticks of led ram, plus being able to overclock the video card now in vista as I didn't do that on the 8800. I might just have to run my OC @ 3.7 and see what the temps are again.

FLO 1337 said:
A few things to keep in mind...

Voltages reported by on-board sensors usually aren't necessarily that accurate and in many cases aren't even consistent.

In addition, just because 4.8V is being reported at one specific point does not mean it is 4.8V at all points. In voltage distribution there are losses. Bad connections, poor design, over-heating, etc. can cause voltage drops to be excessive. So, starting off with a low voltage reading to begin with is definitely not good news. I can pretty much guarantee that at some point your +4.8V is going to dip below the specified minimum of 4.75V (5V+/- 5%). In fact, given the possible error in most of these systems, it may already be under 4.75V even though it is reporting 4.8V.

Some PSU output voltages also aren't all that stable when the input AC mains voltage droops.

Not saying that's your problem, but it's a technical fact to keep in mind when tossing voltage readings about.


Similar thing with temperature readings. A number of CPU temperature reporting programs don't even report the correct temperature, and can be "off" by a substantial offset. And, it also depends on which temperature parameter you're talking about: die, DTS, case temperature, and so on, and whether you're comparing it to the right maximum temp permitted parameter. For example, comparing the max die temp to a published max case temp has no value. To over-use the phrase, everything has to be verified as comparing apples-to-apples.

It also matters whether the value refers to a specific core in a multi-core CPU. 60C applicable to one core may be misleading if that particular core isn't being stressed while another core is.

I was using everest for the voltage readings, which were the same as the bios. I was just playing dirt and in everest core one hit 54 core 2 hit 49, core temp max at 51 @ 3.7 so yeah very unrealible but it's something lol. I'd say for sure i'm hitting 60 though at 3.780 soo what to do lol other than clock it at 3.6ish, no need for new p/s?

And thank you for the quick responses : )