• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Safely tweaking my PSU

N

ncantador

Gawd
Joined
Aug 26, 2003
Messages
714
Thanks to the information found in this forum, and in particular the moderator, I have been able to better understand the workings of a PSU and how to decide which one will fit my needs based on my configuration. I have decided to quicky purchase an APC 500va ups, to tide me over until I can get a suitable PSU.

Working with my current PSU, I notice my 3.3v rail was hovering at 3.1. fluctuating as low as 2.6v. Running sandra bandwidth benchmark would cause the system to reboot. I reset the memory vdimm from 2.8v to 2.6 volts and got the 3.3v rail pretty steady at 3.2, and no more reboots, My memory is stable at this voltage so I see no need to run it higher.

So now my question is, what other components, besides memory, draw power from the 3.3v
rail? The purpose is to see if I can adjust any of my hardware or bios settings to place less of a demand on that rail.
 
well the mobo primarilly employs that rail
alot of videocards heavilly employ it, it sort of depends
keep in mind the +3.3V and the +5V volt rails are interdependent on most supplies*
and the 5V powers most PCI slots (PCI-X employs +3.3V)

*(Antec True Power and PCP&C being 2 exceptions with truely Isolated rails)
 
Thanks again Ice Czar. The 3.3 rail is dropping alot (down to 2.66) under load. This is obviously not a good thing, but everything else is fine. I'm really surprised because this unit was supposed to be good performer. I wanted to get my new unit (Fortron Blue Storm 500W) in a couple of weeks. Do you think I can damage anything by running the current unit in the meantime?
 
Did you measure with a meter? Software-based readings can be way off, not just for absolute readings but even for fluctuations.

C'T magazine measured several XP2400+ systems w/ GeForce3 Ti500 graphics cards and found that they drew about 9-14A @ +3.3V, regardless of whether the CPU power came from the +12V or +5V.
 
If the rail is showing low and he's getting reboots I don't think a multimeter is all that necessary. You can try cutting back on power draw on that rail but I doubt it'll have much of an effect. More than likely it's a bad regulator or something in the supply. Even putting less of a draw on it won't necessarily make matters better. It would only be beneficial if you couldn't get enough power to begin with. Could be a bad cap as well causing the problem but who knows.

If you do want to try cutting back on power take out any extra memory sticks you don't need and remove anything old that's in the case. So if you have any weird PCI or ISA cards in there take em out.

On the other hand try turning up that vdimm voltage. Kind of like oc'ing it without actually increasing the speed. Hoping of course that in the event of a voltage drop it won't be enough to cause a reboot. Just be prepared for the ram to get a little warm although I doubt you'll have to worry about any damage.
 
how would one test the 3.3 rail with a multimeter, i mean where would i plug in the probes ?

i can test the 5 and 12 volts on a spare molex plug, but not sure about the 3.3 rail.
 
Anarchist4000 said:
If the rail is showing low and he's getting reboots I don't think a multimeter is all that necessary. You can try cutting back on power draw on that rail but I doubt it'll have much of an effect. More than likely it's a bad regulator or something in the supply. Even putting less of a draw on it won't necessarily make matters better. It would only be beneficial if you couldn't get enough power to begin with. Could be a bad cap as well causing the problem but who knows.

If you do want to try cutting back on power take out any extra memory sticks you don't need and remove anything old that's in the case. So if you have any weird PCI or ISA cards in there take em out.

On the other hand try turning up that vdimm voltage. Kind of like oc'ing it without actually increasing the speed. Hoping of course that in the event of a voltage drop it won't be enough to cause a reboot. Just be prepared for the ram to get a little warm although I doubt you'll have to worry about any damage.
Click to expand...
I tried increasing the vdimm, but the pc just started rebooting again. It seems the psu just can't provide the power that the memory needs at the higher voltages. My new power supply should be here tommorow, hopefully that will resolve the issue. The only hardware not listed in my sig are 2 optical drives, 6 fans (including the cpu fan) and 2 cold cathode lights. I don't think this is too much for a stable 400w power supply to handle.
 
My new power supply was waiting for me today when I got home, and it in my machine right now. Got another 200mhz from my memory right off the bat, the 3.3 volt rail steady 3.28 through 5 sandra loops. Very pleased. Thanks for all your advice.
 
You must log in or register to reply here.
Back
Top