Friday 20 March 2009

I could have killed my gaming PC!

Its been quite some time, a matter of several months in fact, since my gaming PC started having problems; it crashing to a blank screen whenever I played a 3D intensive game.

Not really the sort of problem you would expect from a PC running an Nvidia GTX280 graphics card!

I haven’t been too fussed about it though, since I haven’t had a great deal of time to spend on gaming lately, and as such I put it on my “things to investigate later” list.

Today I found the answer to the problem, here is how I went about investigating it.

orthos Initially I thought, and secretly hoped the problem was down to the CPU, as I have a replacement part waiting to be fitted if need be, so I tested the CPU with Orthos which puts a 100% load on all the CPU cores.  If the CPU was on it’s way to the big bin in the sky, then Orthos would surely let me know.

After Running Orthos for a couple of hours, I determined that there was little wrong, so moved on to the next, and more likely culprit, the graphics card.

As previously mentioned, I have an Nvidia GTX280 installed in my PC, a pretty powerful graphics card, and it’s pretty new, having only purchased it a few months ago.

rthdribl I tested it initially with rthdribl 1.2 which is a pretty intense program for testing any graphics card’s stability.  Sure enough, within 10 minutes the machine blank screened on me.

I had a little think about why it may be causing a problem.  My PC is cooled very well, with a 120mm fan pulling cool air from the front, while two more 120mm fan’s exhausted hot air from the rear and top.

To exclude temperature as the fault, I downloaded and installed Nvidia’s GPU monitoring application, Ntune.  I was very disappointed though when I read what the program had to say.

My GPU was running at a massive 105°C!!!  Why though?  I then read that the speed of the graphic card’s fan was running at just 145rpm.  How can that be possible?

I opened the case up and took a look, what I saw annoyed me.  It didn’t surprise me, or shock me.  It just annoyed me.  A single cable, from my memory card reader to the USB header of the motherboard, was touching the fan of the graphics card, preventing it from moving, therefore allowing the very expensive card to overheat.

I was annoyed because it was a fundamental, amateur error to make when putting the PC together, by not securing the cable it had been the cause of my computer crashing, and could have potentially cost me a great deal of money.  Oh well, you live and learn.

So, now that I have secured the cable, and ran some more tests, and played a few hours of Far Cry 2, I am happy the my PC is once again in good running order.

Ntune records the idle temperature of the card at 43°C, which is far more acceptable than the previous 105°C, and peaks at around 78°C after a couple of hours playing Far Cry 2.

I hope my mistake will help you if you are having similar problems, and to learn a little lesson – keep those cables tidy!!!!

3 comments:

  1. Thanks for the comment Magre!
    Indeed, "D'oh!" is suitable here!

    ReplyDelete
  2. Its always such a relief when its something little like that. Dead/Slow fans can cause a lot of damage if they are not caught in time.

    ReplyDelete

Related Posts Plugin for WordPress, Blogger...