Computer shutting down - not overheating or PSU

[t.scardino]

Hi all,

I play a lot of video games and for the past few months my computer has been shutting down while playing certain ones. The main culprit here, and the game in which this first started happening, is League of Legends. As soon as the game starts loading the computer would shut off, like someone pulled the plug. Arma 2 and any game on the Source engine (CSS, HL2, TF2, etc.) also cause shutdowns. Confusingly, some games like Red Orchestra 2 and Natural Selection 2 give me no problems, allowing me to play for hours.

Suspecting that my power supply was failing (since I know for sure overheating isn't the problem), I put in one of my old ones to see if it would work and bam, everything worked perfectly. It stayed that way for maybe a week or two tops before I started getting shut downs again.

I asked for help on a different forum and they advised me to buy a new power supply. I did and the problem still persists.
The PSU: Newegg.com - COOLER MASTER eXtreme Power Plus RS500-PCARD3-US 500W ATX12V v2.3 Power Supply

Now just recently (the past week or two) I have been getting shut downs while booting up my computer. I have to turn the computer back on multiple times before I can get to my desktop.

Here are my specs:

[Nuclear Krusader]

It'd help the people here help you faster if you posted your system specs.

I'd check the mobo voltages in the BIOS. Are you overclocking either mobo or video card?

[t.scardino]

Thank you for the quick reply,

I posted a screencap from Speccy with my specs in it but I guess it is not viewable.

Here they are in text form:

OS: Windows 8 64bit
CPU: AMD Athlon X2 4800+
RAM: 2gb
Mobo: NF MCP61
GPU: Nvidia Geforce 8800GS

I am not overclocking anything. I looked all throughout BIOS for my motherboard voltages but I couldn't find anything pertaining to that.

[birddog_61]

How do you know for certain that your not overheating? Have you cleaned the heatsink for your cpu and gpu lately?

While I do not like your chosen PSU at all, it should not be the problem. This to me sounds like your gpu overheating, it only happens when your graphics are being taxed.

[t.scardino]

I'm quite sure overheating is not the problem. I have monitored my temps in the seconds leading up to a shut down and they are acceptable (30-40c). I am able to check my temps in the seconds before shut down because, in the case of League of Legends, it happens at the same point every time. I have also cleaned my CPU and GPU heatsinks with no effect.

I don't think its my graphics card overheating because I can play much more graphically intensive games (like the new Hitman: Absolution, for example) for hours with no problems. Whereas League of Legends cause shut downs the second it starts loading, with no gameplay involved.

[Nuclear Krusader]

Try updating the graphics card's drivers. Another thing is that those games might be too much for that old card. I'd check out nVidia's forums: search for the games' titles and see if other people are experiencing the same problems.

Even before you update the drivers, try disabling as much eye candy as possible in the problem game: disable all 3D accelerated features if possible and run the game again: see if it crashes.

Try to shut down as many apps as you might have running in the background. Only the essentials (AV, firewall, Windows services and programs) should be left on. Shut down all E-mail, IM's, system tray resident programs, etc., before firing up the game, so as to minimise interference with other software.

Check the game forums and see if others are having similar issues and what they're doing about it.

[jdeb]

Your video card is more than likely going out on you or that CM PSU is bad, you could check the PSU voltages to verify. or Verify by swapping the video card with a known working one. Additionally, if it is the card and it is an EVGA AR series and it was registered, it has a lifetime warranty.

[t.scardino]

I don't have another video card to check, but I tried running OCCT (link) and I get a shut down when trying to run the GPU test. I guess that enough verification.

[bob]

One thing that is rare.
Overheating somewhere other than monitored. Just use an external fan to test that.
Probably not the answer but one step closer.

[David M]

You need to check your voltages when the graphics card is under a full load. You cannot do this from the BIOS when the card is essentially idling, although you first need to check that, especially your 12 volt rail, which powers graphics cards. Download and install software that monitors and records your voltages on a graph and then run a graphics card stress test.

Voltage monitor...http://www.filehippo.com/download_speedfan/

Graphics card stress tests...
http://www.geeks3d.com/20110407/down...mbustor-2-0-2/
http://www.ozone3d.net/benchmarks/fur/

Your power supply has two 18 amp 12 volt rails rather than a single rail providing all of your 12 volt power. This is a potential problem if your card can draw more than 18 amps on the 12 volt power connector. Running the stress test will tell you if this is the case, if you see your 12 volt rail voltages drop. If you see this try switching your 6 pin PCI-E connector to the other PCI-E connector on the other 12 volt rail and then re-run your stress test.

You have a card that is a number of generations old now. You might be happy with a new card that can play your existing games at higher speeds and at higher setting and provide you with other features like DX11.

A 500 watt PSU is not going to power the more powerful cards and is going to be the minimum recommended for a number of mid grade cards by the card manufacturer. These days 650 watt PSU's for gaming cards are pretty much the minimum that are being recommended. The high quality PSU's are Corsair HX and AX series PSU's and Seasonics. I don't know about the quality of your Coolermaster PSU, glc or jdeb will probably have some insight.

[t.scardino]

Thanks for the info guys. Unfortunately I can't make use of it because my computer won't start up. I get 5 beeps at startup, a pause and then continuous beeps. This happened after I tried reseating my video card and ram sticks (figured it was worth a shot). I've read that these beeps mean a CPU failure? If it is my CPU dying and reseating my GPU and RAM somehow triggered it, the shutdowns I've been having make a lot of sense now. Seeing as they happen when games are loading - afaik which is when the CPU is under a lot of stress. I did used to have some major overheating problems with my processor awhile ago which I fixed with new thermal paste.

Edit: scratch that, seems one stick of ram had some dust on its pins. Ill test my voltages with those stress tests and report back

[t.scardino]

I ran MSI Kombustor while watching my 12V voltages in Speedfan but didn't see any changes. Everything on the chart is flat. I tried doing the same with OCCT's GPU stress test but my computer didn't last 2 seconds before shutting down. After, I tried switching the test to DirectX9 and there were still no changes in voltage, although it did take significantly longer to shut down.

Also, the fan on my graphics card has started screeching. It's to the point where I have to disable it and use an external fan for cooling cause it's so loud.

[glc]

I'd say it's time for a new video card.