Sign in with
Sign up | Sign in
Your question

PC shutdown after some games run 1-2 minutes

Last response: in CPUs
Share
July 17, 2012 6:31:04 PM

I got a new CPU yesterday and slapped it in my system (Phenom X4 965 125w). Once I had done that, I played TF2 for about 4-5 hours without issues. After getting bored I decided to test the new hardware out by playing a game that I could only play on low with good 40-60 FPS: The Secret World. I'm running it DX9 with settings set to medium now and easily after 1-3 minutes the entire system freezes for 20-30 seconds then shuts down. Alarmed, I ran core temp to check my CPU temps because it was the newest component, no problem 36-40 idle and only reaching around 45-48 while playing for those few minutes (stock heatsink/fan). Another thing I notice is that there are slight clicking/switching noises (not constant and not like a buzzing, just every 30-45 seconds at random) going off in the case from the game rendering my character to the final shutdown. From the "jist" of it, I'm leaning towards possibly the PSU going bad.

Here's my list of things to check in order:

1. GPU temps (I honestly don't believe its this, but can't be 100%)
2. update BIOS if an update is available (I've seen in a couple instances where this has been the solution)
3. replace PSU
4. replace mobo

My system:

Mobo: Asus M4A87TD EVO (installed about 6 months ago)
CPU: 3.40 gigahertz AMD Phenom II X4 965 (got yesterday)
GPU: NVIDIA GeForce GTX 560 Ti (got a month or two ago)
PSU: Can't remember brand, but its a 650w. (installed about 6 months ago)
RAM: 8 gigs, can't remember brand, will open case to check if necessary. (installed about 6 months ago)

Let me know if you need anything else.
July 17, 2012 7:38:41 PM

I think its the PSU... theres some energy failure or I dont know... or a cable is loose
Maybe a virus/malware...

Oh yea... nice TF2 playing!
July 17, 2012 10:14:14 PM

updates: GPU temp is idle at 31C, when playing secret world it goes to about 57-60 before shutdown. I don't think it was even a stable temp at the point but not for sure.

Checked BIOS version, I'm up to date.

update: ran 3dmark11. ran fine for about 3 minutes then shut down again. Therefore it is not client related nor CPU related. It is more than likely GPU problem or GPU overheating or PSU.

SECOND UPDATE: attempted to log GPU temps using Everest Ultimate. However, as I started the 3dMark11, it didn't fully load the first benchmark before shutting down. upon reviewing logs, the few second before and during the loading sequence my GPU temp was 33-35C (about the same for mobo and CPU temps at the same time). This tells me it may very well not be the GPU overheating. So far my gut is telling me (and hoping) that it's the PSU. BUT, thats a gut feeling. I'm terrified as to what it could be and still having no luck finding the culprit.
Related resources
a c 190 à CPUs
July 17, 2012 10:22:00 PM

Can you find the make and model of the PSU, that would help narrow things down, especially since you say you got the 965 yesterday which would have increased your power consumption.


Also, 3dMark is not good for narrowing down suspects as it tests the whole system which doesn't allow for ruling things out. Get furmark and use that to load up just the GPU and see what temperature it gets to, then use prime95 blend test to load up the CPU only and see if that crashes. If both work fine alone, then run them at the same time to create a worst case load for the PSU, if it trips off then it is quite likely the PSU since the individual runs determined that both the CPU and GPU were not the culprit.
July 17, 2012 10:24:11 PM

try checking for new graphics driver,
while you run 3dmark11 open speedfan at the side and watch it. (checks for motherboard, case temp and CPU temp)
also use msi afterburner to see the GPU temps, use it like speedfan above.
was the CPU brand new???
July 17, 2012 11:16:45 PM

hunter315 said:
Can you find the make and model of the PSU, that would help narrow things down, especially since you say you got the 965 yesterday which would have increased your power consumption.


Also, 3dMark is not good for narrowing down suspects as it tests the whole system which doesn't allow for ruling things out. Get furmark and use that to load up just the GPU and see what temperature it gets to, then use prime95 blend test to load up the CPU only and see if that crashes. If both work fine alone, then run them at the same time to create a worst case load for the PSU, if it trips off then it is quite likely the PSU since the individual runs determined that both the CPU and GPU were not the culprit.


Ill do that in a moment. Ran a benchmark using prime95 earlier with no issue and temp maxing at 45C. Ill try out furmark then run them both simultaneously afterwards. Ill let you know the results in little bit. I'm gonna take a stress relieving nap first. Oh yeah. And ill take out the psu and check make and model. I know its a crap brand I never heard of that my store locally had and I bought it in a pinch (mistake).
July 17, 2012 11:21:58 PM

rajaawad23 said:
try checking for new graphics driver,
while you run 3dmark11 open speedfan at the side and watch it. (checks for motherboard, case temp and CPU temp)
also use msi afterburner to see the GPU temps, use it like speedfan above.
was the CPU brand new???


Brand new CPU from new egg.

When I ran 3dmark11 I had a couple temp checkers (speccy, afterburner, and everest running (no custom fan profile set)). While I wasn't able to see speccy or afterburner before initial 3dmark shut down, I ran logging from Everest on my second go. However it didn't make it past loading the first benchmark before shutdown.

I have the latest release drivers (non beta).
July 17, 2012 11:46:42 PM

it seems like it is either your PSU, CPU or motherboard, most probably your PSU, possibly your motherboard (search online for issues with this CPU and your motherboard), or your CPU (prime95 should identfy that.)

good luck though.
July 17, 2012 11:49:15 PM

hunter315 said:
Can you find the make and model of the PSU, that would help narrow things down, especially since you say you got the 965 yesterday which would have increased your power consumption.


Also, 3dMark is not good for narrowing down suspects as it tests the whole system which doesn't allow for ruling things out. Get furmark and use that to load up just the GPU and see what temperature it gets to, then use prime95 blend test to load up the CPU only and see if that crashes. If both work fine alone, then run them at the same time to create a worst case load for the PSU, if it trips off then it is quite likely the PSU since the individual runs determined that both the CPU and GPU were not the culprit.


Alright, I have some bad results. I have yet to open case, but let me update you before I do that. (yes, i skipped power nap, determined to get to the bottom of this). I downloaded furmark, started a 1280x720 res benchmark without AA. as soon as I clicked go the system shut down. The system loaded the 1280x720 window with absolutely nothing in it and instantly as it came up the system went out. So its still down to GPU or PSU. Gah.
July 17, 2012 11:59:17 PM

rajaawad23 said:
it seems like it is either your PSU, CPU or motherboard, most probably your PSU, possibly your motherboard (search online for issues with this CPU and your motherboard), or your CPU (prime95 should identfy that.)

good luck though.


checked for issues on mobo/cpu combo and didn't find any problems similar to this. and again, a short prime95 benchmark showed no issues. I'm hoping its not the mobo or GPU, please god be the PSU.
July 18, 2012 12:13:06 AM

go for you nap, start prime95 and leave it for at least 6 hours and see results, if all good, get memtest86 and run that for a while, if all is good then it is most definitly the PSU.
July 18, 2012 12:30:10 AM

rajaawad23 said:
go for you nap, start prime95 and leave it for at least 6 hours and see results, if all good, get memtest86 and run that for a while, if all is good then it is most definitly the PSU.



did quick blend test. nothing bad other than pretty high temps (heard to be expected with phenom, will get aftermarket heatsink/fan), stabilizing 57-59C after a few minutes and I let it run for quite a bit longer. Since the issue was pretty immediate with shut down (not after long periods of game loads on the processor), I think I can rule out CPU. I'm going to run memtest. If you suggest running a blend test with prime95 for long time, let me know. I just don't think its necessary as I'm not checking stability of an OC and the issue was not a long period load problem.
a c 116 à CPUs
July 18, 2012 12:52:10 AM

My bet would be on the currently-still-unknown PSU as well. Systems shutting down the moment the GPU kicks in would be consistent with a PSU failing to gracefully handle the load change.
July 18, 2012 1:27:34 AM

If PSU 12 Volts trail drops from 12.10 volts to 9 Volts anything begins to suffer on board slowing down, freezing, shut down.

You have 210 W from the card and 125 from the processor is 335 watts make sure the power to your cpu connectors are in place, make sure the 2 6 pin connectors from the PSU are placed in the card correctly if everything is in place try to find a 2 Molex to 6 pin pci-e like you see here:

http://img685.imageshack.us/img685/5007/molexpciepower....

it may solve your problem.
July 18, 2012 9:10:29 AM

I would recommend getting your self a new decent PSU anyway wether the problem is form it or not, you dont want to risk your hardware with an half decent PSU, get about 650W 80 pplus rated and make sure it comes with 2 6 pin GPU cables to avoid using molex to 6 pin as sometimes they cause issues.
July 20, 2012 9:02:07 PM

Alright, i got a new PSU (750W Bronze rated, http://www.newegg.com/Product/Product.aspx?Item=N82E168... ) and I ran furmark and the 3dmark11 and had no crashes. So its safe to say the issue did involve a bad PSU. After taking it out the old one, it was a Blue Star (never even heard of). I also got a new side fan (200mm) for my case and an afterstock CPU cooler (Cooler Master 212 Hyper), now I just need to figure out how to get my GPU temps down and everything is perfect for me.

Thanks all!
July 22, 2012 3:46:28 AM

Yeah, not necessarily a BAD PSU, just overloaded...when it ramped up to display more, watts went up just enough or VDroop kicked in...as most cards don't use maxed voltage until they need all 210W..
Bluestar is probably chinese-made for OEMs....
And don't really worry about GPU temps until they start to hit 65 or higher..and stay there. 35C at full load would be a great temp...even 55C is nice.
July 22, 2012 4:13:31 AM

+1 georgekn3mp

but the problem can also be related with some old capacitos on board malfunctioning the 2 molex to pci-e can solve the problem is good to have one handy, you never know.
April 12, 2013 3:31:07 PM

NOT ANSWER!


I have the same problem!
New CPU FX 8150 125Watt 8 core's 3,6Ghz
No problem what so ever on my (old CPU <- which I'm still using because of my problem)
Power supply 750Watt Corsair
Graphics card Radeon HD 6770 1 GB
Motherboard 990FXA-UD5
SSD 120 GB
HDD 500 GB
2*4 GB Kingston RAM 1600

Going between 50-41'C
Only difference is that mine shuts off even while not playing anything
Btw, old cpu = Phenom II X4 955 - 125 Watt
April 12, 2013 5:31:24 PM

bocajls said:
NOT ANSWER!


I have the same problem!
New CPU FX 8150 125Watt 8 core's 3,6Ghz
No problem what so ever on my (old CPU <- which I'm still using because of my problem)
Power supply 750Watt Corsair
Graphics card Radeon HD 6770 1 GB
Motherboard 990FXA-UD5
SSD 120 GB
HDD 500 GB
2*4 GB Kingston RAM 1600

Going between 50-41'C
Only difference is that mine shuts off even while not playing anything
Btw, old cpu = Phenom II X4 955 - 125 Watt


In answer to other posters, determine what components you want in a computer and then purchase a high quality power supply. Most people overlook the PSU, however, it is the most important component; without it problems will eventually occur.

To the last poster. Some AMD components do not mix well. I suggest you go to the forum for both the CPU and motherboard manufacturer and post your problem.

Although I was a computer expert and understood and did just about everything; I have not worked with AMD computers for a long time and I am an old man now with a poor memory. In fact, I use a great deal of references to answer questions, even posts I have made here and on other web sites.

Even though I know there have been AMD problems, I am sorry that I cannot remember the answer to yours.

Dan

!