• 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.

Does this mean my PSU is dying?

K

kevineugenius

[H]ard|Gawd
Joined
Dec 9, 2006
Messages
1,415
When playing Guild Wars 2, sometimes for 10 minutes and sometimes for 10 hours, my monitors will all go black and never recover. Audio continues to play. The system hangs hard and there's nothing in the event logs. I've updated drivers to several different versions on my GTX 670. A quick Google search says the most likely culprit is the power supply. Does that sound reasonable to the experts around here, or should I continue troubleshooting other things?
 
When my PSU was going I was getting full system restarts under heavy load. To me having just the video drop out sounds more like the video card is unstable. I wouldn't rule out the PSU yet though. If your hardware is under warranty it can't hurt to go for a RMA.
 
PSU has been out of warranty for probably 7 years, GPU should be in warranty still. I put a temperature monitor on and the GPU was up around 82*C, so I installed Precision X to crank the fans up. That kept the temps down around 71*C but it still locked up. Now I've turned off two monitors and am playing on only one and the temps are down to 52*C... if it stops locking up it will at least be confirmed in my mind to be either GPU or PSU but I still won't know which. SUPPOSEDLY a GTX 670 can run 85*C without being in any danger but... that's pretty dang hot. I'm on a 700W PSU that's older than some of the members of this board and at load this system probably eats 650W... seems theoretically sound to me but I'd hate to use my college student almost-no-funds on something and have it not be the problem.
 
I dont think its the PSU. PSU symptoms are generally the whole PC shuts down as it runs out of juice. Im thinking its GPU related as well Is the monitor's actually turning off or is just the screen going black?
 
Yeah, that is understandable. It is hard to diagnose these things 100% without other hardware to test. You will probably find as many stories of people fixing a problem like this with a PSU as with a video card. Wait and little while and see what other people have to say.

If you do end up going with a PSU just order at a place with a good return policy.
 
I doubt it's the PSU.

What are your full system specs?
 
kevineugenius said:
MacLeod, the screens "turn off" as in the power LEDs go from green to orange.
Click to expand...
audio continuing isn't out of place given buffers the chips have.
have you checked the fans on the GPU? Friend recently had a similar problem with his (anything heavily graphical would cause a freeze after a period of time) and it turned out to be the GPU fans had stopped. (nearly burned myself on the heat pipe when I pulled it out)
 
I've seen system freezes be a lot of things. We're gonna need full system specs, including OC and voltages.
 
Gigabyte z68-UD4
i7 2600k
2x 4gb G.Skill 1333 1.5v
No overclocking or voltage changes, the GPU is stock clocked MSI

I did get some better info today though. The screen went off again but it was... light black... if that makes sense. Like the backlight of the monitor was still on. So, I waited about 20 seconds and it came back up. I immediately went to event viewer and found:
Event 4101
Display driver nvlddmkm stopped responding and has successfully recovered.

Well, I know what you're thinking. "Oh, just update your drivers!" This is the third iteration of drivers I've installed since this began, and yes I did do a complete uninstall of drivers before installing new ones. At least I have some hard data to send off to the Guild Wars 2 team, see if it's something they know about. I'll hold off on buying a new power supply for a while.
 
I always forget the model number and the sticker is in an awful spot. Hipro 700w, lots of people say it's garbage but I've gotten like 6 years out of it so far, running SLI and Crossfire half the time.

Might be this one... http://www.newegg.com/Product/Product.aspx?Item=N82E16817174028

I can pop the case if entirely necessary but... with an event log showing a driver failure that takes my suspicion off the PSU quite a bit.
 
Use a motherboard monitoring tool to watch the Power supplies 12V rail.
Even better if you can use a multimeter.
Let us know the idle and load voltages.

For load testing, run Prime95 for a minute.
Check to see if the voltage wavers much at idle or load.
 
I'll have to wait till next week, my multimeter is packed (moving this weekend). Let's pretend I don't actually know the layout of power supplies all that well... can I just measure on the pins on, say, an unused 8-pin GPU header?
 
Monitor the 12V plug for the CPU on the motherboard.
It will be a 4pin or 8pin plug.

On a PSU with a single 12V rail , you can measure any yellow and black wires.
 
Software monitoring tools are useless unless calibrated with a multimeter.

The problem may be the GPU, but replace that PSU ASAP. That PSU may actually be causing the GPU to fail by giving it bad power.
 
Tsumi said:
Software monitoring tools are useless unless calibrated with a multimeter.

The problem may be the GPU, but replace that PSU ASAP. That PSU may actually be causing the GPU to fail by giving it bad power.
Click to expand...

They only need to be calibrated if you are after definitive values.
For observing a trend they can be very useful.
 
Observing what trend? It doesn't matter what trend the PSU has, what matters is if the voltages stay within ATX spec.
 
Tsumi said:
Observing what trend? It doesn't matter what trend the PSU has, what matters is if the voltages stay within ATX spec.
Click to expand...
The actual value doesnt matter so much unless its really low or high, ATX spec allows up to 5%.
If the voltage changes much it indicates a problem.
I asked him to provide the voltages so I can see how much it changes, this can show if a PSU is on its way out.
If there is significant change, a meter can be used to verify the voltages.

I suggested using a meter.
If one isnt available, use what tools are at hand.
 
Ten bucks say that it's the PSU being incapable of providing sufficient/clean power to the GPU and mainboard, causing it to display such issues.

Your hunch is right: replace the PSU with one capable of powering this system and your problems should vanish.
 
Nenu said:
The actual value doesnt matter so much unless its really low or high, ATX spec allows up to 5%.
If the voltage changes much it indicates a problem.
I asked him to provide the voltages so I can see how much it changes, this can show if a PSU is on its way out.
If there is significant change, a meter can be used to verify the voltages.

I suggested using a meter.
If one isnt available, use what tools are at hand.
Click to expand...

Except those tools aren't reliable, and his reboots are not necessarily happening while under load, which is when voltage drops would occur.
 
They are useful if you know what you are doing.
I do.
 
I just don't think it's the PSU. If it was giving bad power down the 12V rail, wouldn't it affect more than just the GPU? Wouldn't the CPU have problems too? These symptoms just sound like a GPU issue.
 
That PSU sucks. You really need to get a new one. Something that will provide CLEAN power.
 
MacLeod said:
I just don't think it's the PSU. If it was giving bad power down the 12V rail, wouldn't it affect more than just the GPU? Wouldn't the CPU have problems too? These symptoms just sound like a GPU issue.
Click to expand...

Electronics can behave in very weird ways... some components can be more fault tolerant than others.
 
Tsumi said:
Electronics can behave in very weird ways... some components can be more fault tolerant than others.
Click to expand...

I guess thats true. Wonder if OP could do something like IBT or Prime95 small FFT's to really crank on the CPU. If its a faulty power supply, then a 100% load on the CPU should put enough strain on the thing that it would shut down or at least error out of the test. If it does then I guess it could be the power supply. If not, then I think we're back to looking at the GPU.
 
MacLeod said:
I just don't think it's the PSU. If it was giving bad power down the 12V rail, wouldn't it affect more than just the GPU? Wouldn't the CPU have problems too? These symptoms just sound like a GPU issue.
Click to expand...

The GPU behaves very differently from the CPU on the 12V rail. Its power usage is far more erratic, causing more strain on the PSU. It's also far more demanding in general, just look at the power usage of a GPU compared to a CPU.

Logically this means that loading the GPU is most likely to show the PSU to be lacking.
 
You must log in or register to reply here.
Back
Top