Solved PC sudden shutdowns...

August 3, 2011 at 03:34:59
Specs: Windows XP, 2.7GHz/2GB
Ok so I'm just looking for some opinions here.

Ive recently put together a small gaming rig, running on cardboard box/table top configuration.

I had it running for 1 week before problems started to arise. I started playing Steam - Call of Duty MW2, and after 30 minutes, immediately shut-down. then it worked fine for 1 day, next day had 2 shutdowns in 1x 5 hour gaming session. Then yesterday I had 2 shutdowns, one after about 2 hours, and one after literally 2 matches on the game, the problem never happened until i played MW2, i played BF2 for hours on end, no problems, but now I do get the occasional shutdowns when playing BF2!

Ive ran a virus check, nothing, just tracking cookies. My PSU meets the requirements for my system, and currently run an AMD Athlon x64 2.7GHz (Dual Core), MSI K9 Neo v2 mobo, XFX GeForce 9600GT OC Edition, 300GB Seagate @ 5200rpm, DVD/RW optical drive, 2x 1GB Kingston RAM, Windows XP Professional SP3, 400watt PSU w/ 22amps on +12v rail, 1x PCI 802.11n NIC.

I have a fan blowing directly at the CPU block, and used RivaTuner v2 to force GPU fan speed to 90%, GPU temp @ 44-48 when playing BF2 on max graphics, 100% view scale and 4x AA. There are no stuttering/lag issues, and no warning before a shutdown. There arent even any indications i.e. always shutting down before a graphic intensive scene, or whenever I press certain keystrokes, or at a certain temp....

The only theory I can think of is either, the power supply is slowly dying and as a result shuts off to prevent any hardware damage, or I have a virus hidden somewhere, or faulty drivers. I will add now that I have re-installed video drivers twice and only started getting the problems.

Also max resolution I play is 1280x780 on my 17inch vga monitor, not too taxing on memory.

I have heard of FurMark, to stress test the PSU and GPU, can anybody tell me if this is a safe benchmark to use, when ive already encountered 5+ sudden shutdowns. Or if it is the PSU, will it just shutdown like usual when things get intense?

If you need any more information let me know, im going to write up a theory of possible causes and a maintenance audit tonight, could help me for a couple of my microsoft tech assignments anyway, if you would provide me with some trusted, and good benchmarking utilities that would be great. Ive used GPU-Z, CPU-Z and RivaTuner, would like to know some more. Just to add I dont have a multimeter either, so would need software monitoring solutions.


See More: PC sudden shutdowns...

Report •

✔ Best Answer
August 3, 2011 at 13:27:45
Alpine is a cheap brand(not good), i see u live in UK try to search this PSU in your local stores Antec Basiq BP430

We can not fight new wars with old weapons, let he who desires peace prepare for war - PROPHET.

August 3, 2011 at 05:13:08
Did u check CPU temp?

Remove automatic restart
If u don't see any error then the PSU is bad.

400watt PSU w/ 22amps on +12v rail
This is capable of running the whole system but next time get the one with +12v@30A

BTW i have galaxy 9600gt i oc with evga precision(core 745mhz, mem 990mhz, shader 1862mhz gpu fan 100% all the time, idle temp 51c at night 48c).

We can not fight new wars with old weapons, let he who desires peace prepare for war - PROPHET.

Report •

August 3, 2011 at 05:54:56
Thanks mate,

Im buying this PSU tommorow from my local PC store:

Its a 500watt with 34amps on the +12v, i think that should be more than sufficient with extra leeway. I'll try monitoring the CPU tonight and let you all know the results :)

Report •

August 3, 2011 at 07:50:36
24 bucks for 500w PSU doesn't sound good.

We can not fight new wars with old weapons, let he who desires peace prepare for war - PROPHET.

Report •

Related Solutions

August 3, 2011 at 08:06:02
Well, im not a hardcore gamer or anything so it should suffice. I dont want to be a fool paying 100+ for a PSU with similar wattage, just because it has a reputable name such as "Corsair" or "Thermaltake" with the added gimmicky cable management solutions. I mean the PSU at the moment is generic, 400watts, cost just £15, 2 and a half years ago, ive ran my 9600gt overclocked and CPU w/ 2x 500GB HDDs, no problems up until now.

I know you do tend to get what you pay for, but I just cant comprehend the difference, this one does have fail-safe mechanisms (or at least it claims to have) so if anything did go wrong in the future, I would just have more shutdowns and out to buy a new PSU. I would only really spend good money on RAM/CPU/GPU nothing else really.

Report •

August 3, 2011 at 08:37:41
Better read this thread

We can not fight new wars with old weapons, let he who desires peace prepare for war - PROPHET.

Report •

August 3, 2011 at 12:19:36
OH I see, just when I thought EZcool sounded pretty decent!

What do you think to this one in comparison?

Still cheap, but if its a temp solution for 2-3 months then ill get that one instead.

Just to add by the way. I ran FurMark on Xtremeburn mode, 1280x780 fullscreen, 16x MSAA, didnt adjust fans, i left the drivers to control it. Ran this for 15 minutes and no shutdown, it was so intense my min FPS was 4 and max 9, so im starting to wonder if it really is down to graphics or something else? possibly the CPU?

Report •

August 3, 2011 at 13:27:45
✔ Best Answer
Alpine is a cheap brand(not good), i see u live in UK try to search this PSU in your local stores Antec Basiq BP430

We can not fight new wars with old weapons, let he who desires peace prepare for war - PROPHET.

Report •

August 3, 2011 at 23:36:49
Thats not a bad PSU for the price I suppose, just a shame that store doesnt sell them. Reason I buy from this store is I can reserve and collect on my lunch break as it is next to the business centre I work at. But I've noticed most of the PSU's they sell are actually EZcool, Alpine and a couple of more expensive corsairs and thermaltakes which are really for SLI configs, so that would probably be too much power for what i'd need.

Report •

August 3, 2011 at 23:38:54
I've got a very funny feeling that these shutdowns have something to do with the new 802.11 NIC i installed last week, bought from the same store, cost only £20, its an Addon 300mbps dual antenna, I looked on wireless status last night, the signal bar was going up and down rapidly, which is very odd, almost like theres a fluctuation in the power being sent to its PCI slot or something? Im going to remove this tonight and retest

Report •

August 4, 2011 at 14:14:48
Update: Tried a new PSU, this was not the reason for the shutdowns, thank god I didnt buy one online after-all. I did however still go for the cheap EZcool, but same results. I decided to dig deeper (physically) on the bench testing, I noticed that I had the harddrive placed on top of the optical drive, face down, seperated by paper....

Now although this is an insulator, it also absorbs heat pretty good.....

I powered down and thought I'd feel the hardrives, both were pretty hot, almost the feel of a kettle thats been boiled and cooled down only 5 minutes, so probably running at 80+ degrees. Not so good and a very amateur mistake when I put this rig together (without case)

What ive now done is flipped the samsung HD (Win 7) on its side, place a fan 4 inches away from it at an angle so it can dissipate heat as well as cool the underside of the HDD.

Before this I also installed additional hardrive, and setup windows 7 on it, tested, same results, so it isnt a virus, nothing to do with compatibility, nothing to do with operating systems, nothing to do with power supply, nothing to do with GPU (as a result of the FurMark conclusion) all I got left is the motherboard, RAM (I also swapped slots, still the same...) and HDD

I will do a follow up on my results for anyone else that is interested, for now im going to absolutely go mental with 3D applications and intensive games, and I really do hope this is the solution, logically my mind is saying ive already found the problem, but lets just see whats going to be posted by me next!


Report •

August 4, 2011 at 22:14:38
I do not prefer to Alpine. Alpine is a cheap brand(not good), i see u live in UK try to search this PSU in your local stores.

Dell Technical Help

Report •

August 5, 2011 at 01:13:32
Nope it was not the hard drive after all! even though it was very cool when shutdown occured.

Now ive eliminated most things, im going to try removing the 802.11 NIC, as problems only started once i installed it.

Report •

August 6, 2011 at 09:43:01
Found the culprit, after running an intensive series of tests on individual components I found that it was the CPU overheating. There was quite a buildup of dust between the fan on the heatsink and the top of the heatsink. So the heat would only have escaped through the sides of the heatsink.

Now on the tests, I was shutting down at 94 degrees celsius on the CPU, at maximum load using AMD stress tester, it only took 14 minutes to reach this temp from an idle temp of approximately 34 degrees (first boot of the day).

Im running an AMD Athlon 64 2.9GHz x2 (I believe its the brisbane?) and with a quick google search apparently 80 degrees is its critical temperature.

Now im going to give it one more shot, the only thing i have done is cleaned out the heatsink with some compressed air, ideally would have liked to apply fresh thermal paste, but I will have to get that on monday as im all out of it. Have also pointed a 120mm fan pointing to the CPU block, will check temps again, doubt it will make much of a difference though, I think it definately needs the new paste, as when I bought this CPU, im pretty sure it had the thermal adhesive pad that all I had to do was peel back and apply the heatsink.

I don't think there is any damage to the processor, so hopefully this whole problem could be down to the basic items, ill probably refrain from using that PC until I have the paste, dont want to wipe my whole system out :)

Will keep you all posted,


Report •

August 6, 2011 at 10:16:52
Good god!!!

I just knocked off 54 degrees from a simple cleaning job, im ecstatic!

Before Dust Removed: 94 degrees - never finished the total 30 minutes stress test
After Dust Removed: 39 degrees max - finished the total test, 0 errors, brilliant!

And there wasnt alot of dust, which is surprising how a small amount of it can insulate so much heat!

Hope this helps someone else with the same problem!!

Well this has truly been an experience for me, and will help solve other similar problems in the future.

Report •

Ask Question