PSU Help

stuie

stuie

stuie

Soldato
Joined
29 May 2012
Posts
3,364
Location
Dorset
Setup the new rig yesterday havnt had much time to play but i tried to run a few benchies last night and on heaven im having an issue where about 1 min into in the pc just turns off for a few seconds then boots back up again

Im thinking it might be pcu related but the system runs fine with non benchies for example i had a half hour diablo 3 run without issues

Just a recap im running

Asus Sabertooth z77 with stock settings
i7-3770k
2 670gtx
1 samsung ssd
1 green 1tb wd drive
4 ddr3 sticks
a recon fan controller and 7 140 fans attached

this is running off a admittedly 5 year old 720watt ocz fully modular psu ive done a quick calc based on tdps and the cpu and cards pull around 420 watts so 300 watts for a mobo, couple of drives and a few sticks of memory should be plenty and its not like the psu is a cheap crappy brand

Ive read in the manual this morning though that be default the sabertooth attempts to auto o/c could this be related or is it time for a new psu?
 
720 watts should be fine for that setup i reckon. 5 years old does put it in question but does not mean it defiantly is the PSU.

Does windows create a .dmp file in the c:\windows\minidump directory?

if it does post that here and ill examine for you. Also if you mboard is overclocking the cards (cant see it) but if it is, this could cause some issues as its going to draw more power from PSU but if this was the issue i would think the issue would occur more often.
 
dunn2010 said:
720 watts should be fine for that setup i reckon. 5 years old does put it in question but does not mean it defiantly is the PSU.

Does windows create a .dmp file in the c:\windows\minidump directory?

if it does post that here and ill examine for you. Also if you mboard is overclocking the cards (cant see it) but if it is, this could cause some issues as its going to draw more power from PSU but if this was the issue i would think the issue would occur more often.
Click to expand...

Im at work at the moment but will check for a dump file later

The 5 year old bit was what made me worry but doing a side by side comparison to my old q6600 b3 and 8800gtx although the gtx was a single card it was a beast for drawing power as was the q6600 and I never had any issues running on full tilt before

TBH im not too worried i very rarely run benchies anyway im just concerned that it might lead to future issues
 
It may not be the Psu. Check your voltages and temperatures when benchmarking, these are good indicators to protential problems. Did you set your ram up using XMP?
 
i have to be honest ive not had much of a play with it and the bios it is whatever the defaults are set

Temperatures appear ok sat at a desktop with just the task manager running its idling at around 32-33c ive not ran any load tests as yet as im concerned about running them with a possibly faulty psu but ive dropped straight into the bios after the reboot and it was showing just 38c on the cpu i would have thought if there was a thermal shutdown issue it would show a lot higher
 
Just been doing some reading my psu

http://www.ocztechnology.com/ocz-evostream-power-supply-eol.html

has seperate 12v rails running at 15amp each ive got one card ran off the dedicated 2*6pin pci sockets and the other one using the supplied adapters off the 4pin molex now if for example the card off the molex was drawing off the same rail would that not cause the problem? and if so would a swap about say running each card off a dedicated pci cable and one each off the molex system work better? lastly I dont know the internal workings but would it be safe to assume that as there are 4 rails and 4 molex outputs off the psu that one is attached to each rail? Or does it not work like that?
 
You must log in or register to reply here.
Back
Top Bottom