Hardware problem: CPU Over Voltage!

Factual Questions
#1

I restarted a computer today (an Asus P5KPL-CM, with an Intel Dual Core E2180 Allendale, no overclocking) and heard a beep code (always a scary thing). The code was three…four. The latter four may have been a bit longer or slower. On another restart (it had been running headless at first), a message came up during POST, “CPU Over Voltage!” The exclamation point is in the POST message – which is kind of alarming.

In going to the BIOS setup screens, I took a look at the CPU voltages. The readings went from white text (normal) to red text as they fluctuated. Here is what I saw (first number is the white number, the reading that was mostly there, the second number was red, and only flashed randomly and sporadically):


V Core Voltage:			1.320v/.296 

3.3V Voltage:			3.12v/1.072

5V Voltage:			4.864v/1.587

12V Voltage:			11.880/5.12

Is this like having the oil light on a car go on, in that I’d better shut the system down and leave it off until I figure out what’s wrong?

As for figuring out what’s wrong … is there anything to do short of replacing parts until the message goes away? In what order — is this more likely a board issue or a CPU? Could RAM be affecting it? The PSU?

Could this be OS-related? It’s running Linux (Ubuntu), and I’m awfully new at it. I’m also having some problems, which will likely be another thread (when it rains, it pours). I was getting the “User’s $home/.dmrc file is being ignored…should be owned by user and have 644 permissions” error, which in following some fixes on the net that worked for others, led me to start getting a “The Gnome session manager was unable to lock the file /home/Rhythmdvl/.ICEauthority; Please report this as a Gnome bug…” When I try and boot the system, I now get the beep codes, then it starts loading the OS, then I get the first error message, then the second, then I enter the name/password, then it pauses, and starts the cycle over again with the first message.

I have no idea if the two are related in any way, or if the first problem was there for a while but I didn’t notice it until I started restarting the machine.

Not being an overclocker, I have no idea what these mean or where to start. Any help?

Thanks,

Rhythm

#2

It’s not an OS problem if it happens at the BIOS. I’d suspect either a bad power supply or a bad motherboard.

ETA: As to whether this problem is related to your problems in Ubuntu: hardware problems can cause pretty much any software issue. Fix the hardware problem first, and then see if the software problems are still there.

#3

I had the same problem and it turned out to be the power supply had fritzed. Quick replacement and I was all set.

#4

Definitely the power supply. Your motherboard is using very (very!) simple circuits to read the voltage off of each of the four “rails” from your power supply: +5V, +12V, +3.3V, and Vcore. It would be really improbable for all of the voltmeters to go bad at the same time.

Under heavy load, or with BIOS tweaks, you might see as much as 10% variation in the voltage from the nominal, and generally I would prefer to see more voltage than less, but this much variation tells me that your PSU is dumping power somewhere else, or not generating as much power as it ought to.

Go ahead and check out Silent PC Review for some recommendations on high-quality power supplies. The focus of the site is on quiet components, but users who prefer quiet components often go fanless, which means that their parts get hotter and are generally under more load. A recommendation from that site means that the part you buy will be a quiet PSU that generates the amps, voltage, and watts as advertised, and can survive overheating if one of your fans breaks. You’ll pay a little more but you’ll get a high-quality part.

#5

WWwwwwwwwwwwwwwwwhhhhheeeeeeeeeewwwwwww!!!

Wow, thanks Jurph, that sounds like the end of diagnostics. Of things to replace, a PSU is fairly innocuous and much easier than, say, replacing the board. Plus, if things get dire it’s easier to find a PSU locally.

Your link isn’t working today (they must be having hamster problems), but the unit I have in there is an Antec EA380 — I’ve been going to Antec for years for cases and PSUs, so hopefully this should be an isolated incident; I really like their products and service. It’s also just over a year old, so should be covered by their warranty. I’ll find out on Monday.
Thanks!