Out of Pod Experience

Hi guys,

I have a computer related problem, and this forum caters to the largest collection of computer types that I know, so I am hoping you can help me.

I have an issue with my PC that I am trying to identify, I believe it to either be a hardware failure or limitation due to its tendencies to occour only during specific activities.

Basically what I get is that when playing certain games, not all games, I will get a total loss of video signal. On the affected titles after about 20-60 minutes my screen will go black and display a "no signal" message, yet the games sound will continue for another 5-10 seconds before it too crashes and I have to reset the PC.

This will happen on World of Tanks, the original Crysis, Saints row 3 (in dx 9 or 10 versions), as well as How to Survive (a zombie apocalypse). Wheras it has never happened on EvE, Empire/Napoleon Total War, any of the 3 Mass Effects no matter how long they are played.

My PC is getting on for 4 years old now and I had assumed it was just lack of performance, however Saints still failed on the old dx9 mode so that shouldn't have been pushing the graphics capability.

My system;
AMD Phenom II 550 black dual core 3.1GHz
4 Gb RAM
Radeon 4870 1Gb (I know this is only dx 10.1 capable)
All of the titles are running off of an SSD, most via Steam which I have never had problems with.
Win 7

I know that I am due an updated system, however I would like to know if any of my current hardware is failing or that I am simply asking too much of it. I initially thought it was an over temperature thing due to the time it took to kick in, but why would that not apply in Empire after many hours?
Also the direct X issue, I understand that the most recent games will require dx 11, and I would fully expect to fail at running them, but why should my system fail when running a game on dx9?

Any help and or advice will be much appreciated.
Evil.

I agree with this diagnosis too

If it is heat, why is it only occouring on specific titles? surely if one title can overheat in 30 minutes then another would get there after 3-4 hours?

It could depend on how a particular title taxes the GPU. Other factors could be where the computer is sitting (on the floor, in a cabinet, up high, etc) as well as the ambient temperature in the room at the time. These are probably just secondary factors, but could contribute.

I had this same issue before on similar hardware. I would first check cooling. Next, I would check drivers and power supply. This is most likely cooling though.

Not necessarily. Could be something so simple as loose cables.

[Alpheias]

My graphics card only has a pair of DVI, I am using one of those to connect.
Current temperatures, PC has been on for about an hour, nothing taxing, internet etc,
CPU - 45C
GPU - 62C

[Commissar Kate]

Ohh yes Captains Quarters... I forgot about that, I disabled it long ago, yes that kills it faster than a shot to the head.

Seems a bit hot on idle, What's your ambient temp?

I have about 30c on the GPU at idle with 24.5c ambient.
My CPU die temp is broken so I can't help with a comparison there.

I think you should try two things; remove the graphics card and clean it from dust with a can of compressed air and secondly, try the second DVI port.

Oh while cleaning out the entire computer don't forget to blow out the power supply as those can also overheat.

Also try to hold the fans in place while blowing out the heatsinks. Just a personal thing I do to avoid bearing damage.

Hmm.. apart from updating drivers etc i am not sure what else it might be.