• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Power supply no good?

theanorak

Junior Member
EDIT: Summary

Is my power supply too weak for this config? Power supply is Seasonic S12 500W?
--


Hi there

I'm having problems with a new build (the first in some time). Materials:

Asus P5W-DH
Core 2 E6600 (stock)
GEIL 2x1Gb (@800 Mhz, 2.1V, per instruction sticker)
Sapphire 1900xt
Seasonic S12 500W
2 x WD 250Gb drives
DVD-RW
in Antec P180B

The problem: After a period of time (variable, but not long - not long enough to have done much more than install XP, and I did have problems during the XP install too), the display...corrupts. Most of the time, it goes to a single colour (solid red, solid white/grey, solid blue) with occasional "striped" versions of the solid colour. Every now and then I get "proper" display corruption, where the desktop becomes garbled.

The machine is still responsive and active behind the corruption - for example, it happened whilst I was downloading/installing the first batch of updates from windowsupdate. I left it until the hard drive light stopped flashing and rebooted - updates installed successfully.

I'm having difficulty monitoring the temperatures on the 1900xt, but the catalyst control centre doesn't seem to indicate scary temps - about 50C. The same more-or-less goes for the other temps - CPU is pretty cool, and mobo doesn't seem to get much over about 45C.

Is the 500W S12 too weak? Or should I be thinking elsewhere. Because this is my first build in...a while...a don't have any components I can swap in - I doubt that the 500W enermax from my old PC is going to be any better, and I think it only has a 20pin mobo connector.
 
This is almost not helpful, but...

The only other vid card I have which will work in this system is an old 16Mb PCI TNT2. With this in (surprise surprise) it's stable.

I even tried shorting the pins on my old PSU (465 enermax) and powering the 1900xt from that - was even *less* stable than before.

So - new graphics card, or beefier PSU?
 
I run an x1900xt, an Athlon, and a Raptor HD on a 500W Enermax modular which is about equivelant to your unit. I'd say it's probably adequate.

hardly needs to lift a finger to power your rig at full load.

I wouldn't go that far though.

Since it's stable with the other card, maybe it's a driver problem or a buggy x1900 or something? Just a guess.

 
Originally posted by: dBTelos
"I wouldn't go that far though."

His system probably uses 250W at full load.

Well reviews like this one show test systems drawing more than 350W at load, measured with a multimeter. As the ratings on PSUs are generally exaggerated, and you always want more power than the bare minumum, I'd say 450-500 watts is about right.

 
I experienced almost your same problem, & had a terrible time locating the source of the problem. Display would corrupt into a weird multicolored checkerboard pattern.

You'd think the problem would be isolated in the video card, but this is the surprising result in my particular case:

It was excessive chipset temperature.

I replaced the stock CHIPSET cooler with this (ship via 1st class mail, total around $12)., mounted onto the chipset with Arctic Silver 5 thermal paste.

It brought my under-load chipset temps down by about 12 deg C, and instantly cured the "mysterious, hard to troubleshoot" video problem, which never recurred, even under very heavy game-play loads.

So in my particular case, the problem was excessive chipset temps, inexpensively fixed by better cooling. You might see if that also applies to your situation.
 
Thanks all - I didn't *think* the power supply should be the problem but with all the massively high-rated supplies available I wanted to check.

It was excessive chipset temperature.

I replaced the stock CHIPSET cooler with this (ship via 1st class mail, total around $12)., mounted onto the chipset with Arctic Silver 5 thermal paste.

It brought my under-load chipset temps down by about 12 deg C, and instantly cured the "mysterious, hard to troubleshoot" video problem, which never recurred, even under very heavy game-play loads.
This might be worth investigating - I was going to RMA the 1900xt but maybe I'll try this first. Would the 1900xt cause additional strain on the chipset (above that of the crappy PCI TNT2 I've got in there right now) which could tip it over the edge? With the TNT2 in place, I've run simultaneous loops of Sandra & Everest's CPU & Memory benchmarks with no ill effects...but the reported shipset temperature was a toasty 45C, and the headsinks did feel extremely hot to touch. I'll try the "desk fan ventilation" approach this evening and see if that makes a difference.
 
...but the reported shipset temperature was a toasty 45C,
Well 45 C is really not all that bad, if that's your temp while under load.

45 C is a tad high, but not extremely bad.

Therefore I now question if maybe the chipset cooler that fixed the problem in my case may not be the solution for your particular case after all.

Mine's around 38C now while under heavy load. It used to hover around 50 - 52C or a little higher (nForce4 chipset on DFI Lanparty UT Ultra-D mobo).
 
Back
Top