Fun With Power Supplies.....

[SkiFletch]

hey guys, i've been running into a problem with my box since as of late its been crashing as if that were its frikkin job . after investigating the problem, I'm pretty sure that I am left with a power supply issue. unfortunately, I cant read any of the motherboards internal monitoring systems (temp or voltage) so everything I know about the way the system operates comes from my digidoc 5. the digidoc 5 reads a voltage of 5.33 volts for the 5v rail. I've heard of low voltages being common to dead power supplies, but what about high voltages? is this a motherboard issue, or a power issue? furthermore though, is there a way I can relatively easily modify the 5v rail with a resistor or such to just pull down the voltage a bit? call me cheap, but I dont have all that much cash to splurge on a new PS... the system is as folllows:

Tbird [email protected] (not running oced and still crashes)
Abit N7V-133R
2maxtor 40s in raid 0
512 Crucial 2100
Asus V8420 Deluxe
Pow Max 500W PS
winxp

also, one note about the crashing; its VERY specific i've come to notice. it crashes when doing one of 4 things only. Typing responses to internet forums, Typing email, Trying to run Need For Speed Hot Pursuit 2, and Trying to move, use, or access in any fasion a non-corrupt non-viruslaiden image of the NFS HP2 cd. Is this the behavior of a dead PS? or am I on drugs just liuke my computer

 

[Tebore]

Try turning off "Restart automatically on Error" in Xp and wait for a error message. It doesn't seem like a PSU error.

 

[FunkDaMonkMan]

your 5v is high because there is a high 3.3v pull and the powersupply has to bump the voltage on them both becuase they are connected.

I'm pretty sure i'm right, correct me if i'm wrong.

 

[rogerdugans]

I doesn't sound like psu problem to me either.

Sounds more like some corruption in the os somewhere to me.

If you are using the System Restore in XP you may be able to restore to a point that has no troubles.

 

[Tebore]

your 5v is high because there is a high 3.3v pull and the powersupply has to bump the voltage on them both becuase they are connected.

In all the cases I've seen it's true. The PSU usually pumps more power out to all rails if one place needs it more.

 

[SkiFletch]

Tebore, where is this whole restart automatically on error command? That doesnt necessarily sound like my problem. it doesnt restart, it hard locks to a black screen. I have to manually re-set or power down and restart

 

[Tebore]

Right Click my computer> Properties> Advanced tab> startup and Recovery> Settings> System failure.

 

[SkiFletch]

tnx

 

[mbentley]

your 5v is high because there is a high 3.3v pull and the powersupply has to bump the voltage on them both becuase they are connected.

I'm pretty sure i'm right, correct me if i'm wrong.

yep, my system does the same thing for the 5V on mine. i attached a link for my voltages.
http://web.ics.purdue.edu/~mbentley/voltages.jpg

 

[SkiFletch]

yes, but does yours cause crashes?

 

[mbentley]

yes, but does yours cause crashes?

nope, but then again, mine doesn't have so much of a power variation... but i think the idea of the 5V and the other voltages being linked together is right on. my old psu used to crash my system daily... luckily i got my new one... nice, suprisingly reliable, no-name psu

 

[Daemonfly]

How accurate are the Digidocs? I would test the lines with an actual multimeter.

I had one powersupply that under normal operation outputted 5.5v - I spliced a resistor inline to bring that voltage down closer to 5v.

 

[SkiFletch]

ok, so i've got 2 multimeters here, both digital and have measured the 5 and 12 rails. I noticed that the rails, even when under gaming have the following voltages: 5.27 +/-.02 and 11.77 +/-.04. even when I induce something that I know will crash it and watch the readings, the fluctuations are still within those parameters. Not necessarily leaning towards PS anymore, cause i would expect larger fluctuations if it was a failing PS, wouldnt you? either way, its not very far out of spec. i've already ran memtest-86 while I slept twice and that turned up a negative, and checked for bad capacitors, and couldnt find any that were bulging even slightly. all these problems are un-overclocked of course, with normal at spec RAM timings. maybe i'll start trying to investigate my hard drives. Thanks for all the help so far guys

 

[Gnufsh]

It actually sounds a little like an OS or software problem to me.

 

[Daemonfly]

The 5v is a tad bit too high and the 12v is a tad bit too low, but I'd say thats probably not the problem. There is a slight chance that some hardware doesn't like those voltages, but I've seen worse without problems. I had a PSU with a 5.7v line - put a resistor inline to bring that down.



Have you done a full AV scan?