very strange one this...

T

The Birddog

New Member
Reaction score
0
Hi,

I am the fourth person to have looked at this PC and I really, really want to get to the bottom of it, please help.

Customers PC works fine, she can go on the internet, email, write letters, usual stuff. But when her son plays a game the computer just turns off and reboots, every time, anything from between ten and 30 minutes of gaming. The pc has enough spec to run the game, direct x is up to date.

1st repair, local computer shop - new higher spec graphics card, no difference, as soon as the customer uses it, it crashes.

2nd repair, a different local computer shop - unable to find a problem.

3rd repair, reinstall of XP. Fixed the problem for a day of nearly continuous use. Second day the problem returned.

4th repair - me - and I really want to get this thing working. Customer seems to have lots of contact with local business and you all know, word of mouth is the best advertising.

I have used diagnostic software to do a stress test independant of the operating system. First time computer crashed, second time it completed. The first time was straight after it crashed during the game.

I am suspecting the power supply but if it was, the second stress test would have failed wouldnt it? Very strange this one but I want/need to get it sorted.

Hope you can help.:):)
 
Without knowing more this sounds like heat related issues with the CPU. We have seen this on all kinds of machines that suddenly shut down when doing CPU intensive work.

I would try to really work up the box so it does the problem and then check the CPU heat. Could be the CPU fan is not turning fast enough or the heatsink needs to be reseated with new silver paste.
 
Last edited:
I would say RAM, power supply or overheating. Is the machine overclocked at all? What's the spec of the machine. Latest drivers? What temps in BIOS? We need more details.
 
The Birddog said:
the stress test would have done this though wouldn't it?
Click to expand...

Depends on what it was stressing. I wouldnt trust anything automatic in this case. Really boot the machine, go into the bios or use some tool to see the cpu temp. Next play some stuff for a few minutes or do anything that should really work it up and then check the temp again.

We have seen this on servers that will run flawlessly but when they are used to recompile php or a kernel (freebsd) they will crash. Found the cpu to be running very hot (70c) and even though the cpu fan looked fine it was not turning quick enough. New fan/heatsink with silver paste brought the temps down and no more crashes.

I would really recommend investigating that since it seems you can reproduce the error after a period of time gaming. Is this machine overclocked as well ?
 
i must agree with the general consensus that it is probably overheating,i had a similar thing happen a few weeks back and the thermal paste that was supposed to have been fitted when the pc was updated hadn't been and as a result when anything intensive was done it would crash,new thermal paste and its now as stable as it should have been, prior to the local PC world "techs" getting their grubby mitts on it.
 
CPU overheating or power supply not up to the task. I've seen this when someone bought a bigger video card for their machine and the +-12V amperage of the PSU wasn't high enough.
 
Ok, I have had the PC in my workshop now for a couple of hours. Played the most boring game but I cannot get the thing to crash!

Is it possible, or has anyone had experience, of it being a problem with the electrics at someones premises?

If it is the PSU, prehaps the electrics at the cutomers house are not up to scratch?
 
The Birddog said:
Ok, I have had the PC in my workshop now for a couple of hours. Played the most boring game but I cannot get the thing to crash!

Is it possible, or has anyone had experience, of it being a problem with the electrics at someones premises?

If it is the PSU, prehaps the electrics at the cutomers house are not up to scratch?
Click to expand...

Did you check the temps?. What were they?. How big is this PSU (watts)?

I dont think customer premises electrical would cause this, it would likely be seen with incandescent light flicker or something else in the room cycling or showing fluctuations.
 
Another stress testing tool.

I just found this... Prime 95.

"For overclockers, Prime95 has a feature called "Torture Test" that allows maximum stress testing on the CPU and RAM. There are several options allowing the stress test to focus on the memory, processor, or a balance of both. Usually Prime95 will detect an error within a matter of minutes if an overclock is not stable, however many people like to let the system "burn-in" overnight to ensure long-term stability."

Thanks to geekhelp4u for the find at http://www.technibble.com/forums/showthread.php?t=3016 for the info!
 
Definitely swap out the power supply once you can get the shut-down predictable. Run memtest86+ yet? If it's the cpu heat then try Prime95, or CPU Burn-in. http://users.bigpond.net.au/CPUburn/ I used it when I was overclocking my last computer and it will tell you right away if it's stable or not, damn thing will lock up or reboot all of a sudden. Usually takes a few minutes to reach maximum temperature, sometimes up to 10-15. I'd let it run an hour on max setting and see what happens. If it's the graphics card drawing too much power maybe try 3dMark, seems to stress quite well, again used during overclocking my video card and works very well. Find a temp monitor for your board such as HWMonitor and monitor to see how high those temps get. http://www.hmonitor.net/ Maybe they monkeyed with some cable that are slightly loose such as the video card (the beefier ones need separate power) and it can't get the power it needs when playing. Check if the bios was messed with and possible reset it to defaults. I had a friend of the family who I helped and a week or two later I came back only to find new windows user accounts and a bios password that the parents claimed to not put on there. No virus's or spyware, clean machine... but three kids in the house. Hmmm...
 
Go to the Event Viewer in Administrative Tools, look until you see a yellow or red warning listed around the time it crashed most recently. See if you can find any listings that give you a hardware malfunction. Also, you can check the "phantom devices" in device manager, sometimes they can cause some random hell.
 
You must log in or register to reply here.

Similar threads

GTP
Making Conky's!
Replies
1
Views
1K
GTP
GTP
HCHTech
Windows 11 folder sharing problem
Replies
19
Views
2K
HCHTech
HCHTech
HCHTech
GoDaddy again - Outlook error 7ita9
Replies
3
Views
730
callthatgirl
callthatgirl
E
[SOLVED] 24H2 will not reconnect to mapped drives even with passwords
2
Replies
31
Views
2K
John Kennedy
John Kennedy
britechguy
How in the world does one boot from USB these days?!!
2 3
Replies
48
Views
5K
lan101
lan101
Back
Top