Computer Randomly shuts off ingame

[eyeCpc]

And through the process of wiping the drives and starting over fresh it may be that the extra demand for power while seeing the repeat system restarts needed for the Windows installation, all drivers/software for hardwares, and susequent updates along with other things taxed a weak component on the board now causing the problem.

Give the board a good close inspection in case of any bulged out or leaky capacitor and/or possible discoloration seen. At the same time if you can borrow a different power supply and run that if adequate you could see if that ends up seeing the same problem or suddenly sees all going well indicating the supply is where the problem is.

The last item of concern hardware wise would be a cpu about to fail. When placing a good load on it it stalls seeing an instant shutdown. A bad drive seeing heads going after everything was redone won't see the system itself shut totally off but cause problems just seeing Windows load up. A battery if found weak is certainly a far cheaper solution then the possible hardware failure the continued problem is pointing at.

 

[Concore]

My brother is also having this same problem after wiping his computer.

 

[eyeCpc]

The one thing that also comes to mind is not the cpu temp but the settings that may be seen in the bios for cpu and other case fans. Once the rpm drops down to the low limit the system then shuts itself off automatically.

If a fan is loaded up with dust or is wearing out that would see a slowdown and end up with the shutdown being seen as a result. You may want to review those settings in the bios and hardware monitor to see which fans are running at what speed. Once you put a load on the fans may not be seeing the necessary increase in speed but could be binding up instead.

 

[crweaks23]

First off... eyeCpc deserves a huge "thank you" for spending the time to help out. You rock.

Back to the issue...

It's certainly possible that the shutdowns are fan/temp related, but I don't suggest replacing everything until you figure that out (read: don't make my mistake haha). I am having the exact same problem as Concore, and replacing all heat sinks and fans (including the heatsink and fans on the gpu last night, as well as adding an EXTRA fan to try and cool off the gpu) have not helped the situation. After reading some temp charts online, my geforce card was definitely running hot, but no hotter than overclocked cards with decent cooling. So, the updated hardware can't hurt, and has lowered the temp slightly, but it didn't help the shutdown issue in the slightest. :freak:

I do know that at one point my date/time was reset in the BIOS, but it only happened once and it was while back. I never thought to replace the battery but there's some good advice on here that seems to be pointing me in that direction, so I will try replacing the CMOS battery today and will let you know how that turns out.

Note: I also have taken a look at the board, but don't see anything out of the ordinary. Is there anything specific to look for? What exactly does a "bulged out or leaky capacitor" look like? I know how to put parts together, but I'm lost when the electrical **** comes into play haha...

- Chris

 

[eyeCpc]

Any bulged out cap or one leaking won't be that much of a stand out item. You have to look closely at the generally canistar round shape and note any uneven even slightest bump showing an irregular shape or look for a slight discoloration on the board which even looks a little damp.

Remember caps are one of the small items seen on a board where a small pocket magnifying glass and a thin pocket type led even flashlight that can get in close are the things to consider for that. Caps at the base of the transformer in a supply are different there being a flat disk ceramic type while most on a board look like a soda can with two leads going into the board.

Most are short and soldered with the bottom of a cap close to the board while some may see a slight slack on the leads. Yet at times nothing will be clearly seen where some test equipment like an oscilloscope while current is active would be a thought or some type of multimeter for locating a dead cap.

Some boards like one older Socket A board here on one case seem to love batteries seeing a shorter life for even the replacement or a battery itself is simply a reject from mass production. You can buy AAs in a store for your remote and see them quit fast if the charge isn't held to see a normal life.

I usually keep a few onhand usually Duracell while Energizers are sold two in a pack at most stores like Radio Shack, Target, Best Buy, Walmart next to the consumer electronics section on a battery rack with 2032 number seen as DL 2032 for Duracell, CR 2032 for Eveready. The user manual for the board will specify the actual number to see used in one of the last sections there.

 

[crweaks23]

I think I know the source of the problem, and my computer has been running for over 12 hours without a shutdown... but I don't think it's really fixed yet.

First off, I checked over the board as best as I could, and I don't see anything irregular with any caps looking out of the ordinary. If it matters, I'm running a really solid board, an Asus P4P800 Deluxe, which was a great board at the time I bought it about 4 years ago.

I replaced the battery, and while I was in the BIOS resetting the time and date I shut off some of the board features that I wasn't using. I had most of these shut off previously, but I disabled everything that I wasn't using this time around.

After these changes, the computer has been running stable, even with a game (Civ4) running. But here's what I see as the problem:

My card is a PNY GeForce FX 5950 Ultra. Using speedfan, I originally saw that my graphics card was running a little hot, which is what convinced me to invest the time and money in trying to cool it down once the computer started shutting down. While I have been successful in lowering the temp of the card, here's the kicker: The temps for the graphics card sometimes just read ZERO, and it seems like while they are reading zero, the computer stays on. I've noticed them reading zero before, and that the computer was running stable during these periods, but never put 2 and 2 together. So, when the temp gauges decide to "work" again, I'll bet the computer shuts down again.

The only explanations for this that I can think of are that the temp readings are wrong, or that the board is running way hotter than it should, but not hot enough to damage the card (yet), because it runs fine when the temps are reading zero. I reach in and touch the heatsinks on the card periodically, and it's never burning up, so a third theory could be that the heatsinks are not doing their job, possibly because of poor contact with the GPU... but I highly doubt this.

So, the million dollar question is... how do I (1) stop the computer from shutting down if the temperature monitor reads some incorrect value, or (2) permanently disable the temp monitors, with a complete understanding that this may damage the card, but that I don't care.

I'm convinced these are my only options, but am open to suggestions.

 

[C0RR0SIVE]

My brother is also having this same problem after wiping his computer.

Sounds driver related to me honestly, that or crappy updates.

crweaks23, chances are the temp sensor on the gpu is bad and is reporting the wrong temp, try touching the cards heatsink when it locks up, and don't put pressure on it, it should be rather warm to hot, but if it is incredibly, and incredibly hot, then yea, it is overheating and causing the lockups you are having...

 

[eyeCpc]

I didn't know you could read card temps with SpeedFan unless that is now seen in the 4.34 or 4.35 release. The program here that displays card temps is PC Wizard 2008.

The fact that the system can run over 12hrs. without a problem rules out hardware faults right there. That's the best news so far. A temp problem while gaming only? I could see that readily if you were ocing the card, cpu, memory for gaming and then seeing problems come up explaining high temps.

When gaming with or without ocing there's another item namely the larger demand on power for not only the video card that everyone focuses on but also by any expansion type sound card. The old Sound Blasters are noted for being power hungry! If the supply is wimping under load or cpu when loaded as well as fans not keeping pace... protection circuit or lack of power?