Computer crashing/freezing/video switching off

Associate
Joined
24 Sep 2011
Posts
21
Hi everyone,

So I built my computer about a year or so ago with these components ordered from OCUK:

HIS ATI Radeon HD 6870 1024MB GDDR5 PCI-Express Graphics Card
Intel Core i3-2100 3.10GHz (Sandybridge) Socket LGA1155 Processor - Retail
MSI H67MA-E35 Intel H67 (Socket 1155) DDR3 Micro-ATX Motherboard - (Sandybridge) **B3 REVISION**
Seagate Barracuda 7200.12 1TB SATA 6Gb/s 32MB Cache
Lepa W-Series 500W '80 Plus' Power Supply
Kingston HyperX Blu 4GB (2x2GB) DDR3 PC3-12800C9 1600MHz Dual Channel Kit (KHX1600C9AD3B1K2/4G)
Samsung SH-S222AB/BEBE 22x DVDąRW SATA ReWriter (Black) - OEM

I installed Win7 on it. After around a couple weeks of use it started to randomly freeze, both when gaming and with general desktop use. I looked into it and couldn't find a solution, and basically put up with it (maybe a stupid thing to do, but I'm quite busy with work etc. and didn't have a lot of time to troubleshoot an intermittent problem).

In the last month, however, things have got much worse:
- random shutdowns whilst gaming
- random freezes as before, both during gaming and while using the internet to surf etc.
- the video output cutting out whilst gaming after only a few minutes (the computer doesn't crash, you can still hear sounds etc ingame)

The random freezes aren't guaranteed to happen, I can go for several hours with everything working fine. The random shutdowns and loss of video happen regularly when gaming and are quite easy to reproduce.

I've looked into heat issues. The mobo is ATX in a full tower, so there is good airflow. Similarly, I've had the GPU and CPU temperatures display whilst gaming and they do not seem to be getting very high.

I've tried formatting the HDD and doing a clean install of Windows - no help. Also have most recent drivers for all my hardware.

I thought it might be the memory. I manually entered the RAM timings into the mobo, no change. I also ran memtest86. This finds errors IMMEDIATELY and then crashes. I tried each stick in each RAM slot and the same thing happens each time.

I'm wondering now what my next step should be. I've browsed tech help forums and it would seem the problem could be:

1. Memory. The memtest86 results would support this, except that the crashing is supposedly not typical of RAM errors. The mobo supports 1333mhz RAM but I was told it should run 1600mhz RAM fine.
2. PSU. Maybe a dodgy or underpowered unit?
3. Motherboard
4. Something else?

Is there anyway I can work out what the problem is, or is most likely to be, without ordering new components and swapping them out? Any suggestions? I'm pulling my hair out here. Sorry if rambling, I've tried to include everything that could potentially be relevant.

I'm posting this before commuting home so should be around to post replies in an hour or so if I don't get run over on the way home :D Hope someone can give some insight!
 
i was about to say, if there is no BSOD and nothing so to speak in the systems event log in control panel, id say psu problem, did you say you bought a new psu when you built
 
Bizarrely, I realised that I was running Memtest86+ version 4.1.. I tried 4.2, ran it for 13 passes overnight and came back with zero errors.

I have removed the 6870 and am running the computer from the i3's onboard GPU. So far, no problems whatsoever for a week: no freezes, lock-ups, random crashes etc. Obviously I haven't been able to do any gaming.

Does this mean it's a GPU issue? I'm pretty sure the 6870 wasn't overheating as I was monitoring it whilst gaming when the computer used to freeze. The problems also kept repeating after fresh OS installs so don't think it's the driver etc.

I haven't had the opportunity to test the PSU but the computer is working fine without the GPU - however I'm not sure if this is proof of anything with regard to the reliability of the PSU.
 
Sounds very similar to an issue that I had with my 7850. Any time it would come under load the system would lock up. I've sent this off for RMA as I believe the card was at fault. Tried on 2 different systems, 2 different PSU's etc.
 
Back
Top Bottom