Thank you for being a valued part of the CNET community. As of December 1, 2020, the forums are in read-only format. In early 2021, CNET Forums will no longer be available. We are grateful for the participation and advice you have provided to one another over the years.

Thanks,

CNET Support

Question

Computer Crashes Not Overheating

Mar 11, 2016 8:29AM PST

Recently, my computer has started freaking out when I play games that are even semi-demanding(League of Legends on mid-high settings, KOTOR 2). The fan starts running on max and it sounds like it is hitting something, like maybe a piece of dust/plastic is stuck in there. I haven't opened the case yet to see which fan it is. Then, usually after about 30-45 minutes(and getting worse) it just crashes. No error screen, nothing. Just reboots as if I just turned it on. I have downloaded a couple different programs(CoreTemp and OpenHardwareMonitor) to check if something was overheating and nothing is. It tops out at about 60 C when in use. The only things I can think that I've done recently is switched my monitor out for a Samsung Smart TV and upgraded to Windows 10 about a month and a half and two months ago, respectively. I've considered that maybe the PSU is crapping out, especially since I've always had weird issues with USB/other ports not working properly. I want to cover all of my bases before I go off spending money on a better supply though. Any tips for fixing this would be much appreciated.

Discussion is locked

- Collapse -
Clarification Request
Sounds like you have work to do.
Mar 11, 2016 8:42AM PST

While you didn't reveal details, the 60C is a known set point for throttling so it is indeed heat related. Your newer techs will write it's not heat but you proved there is a heat issue in your post.

Also, you are measuring just a few chips and what if there was a motherboard or other chip that didn't like the elevated temps?

As to the PSU, the make, model and the rest of the machine is not revealed. Here's my rule for PSUs. 1. Have 50 to 100% spare AMPERES. Not Watts! 2. Most folk can't balance the +12V rails. Make sure you have a single rail +12V PSU.

Finally, if fans are like you noted, fix that.

- Collapse -
artfh
Mar 11, 2016 8:58AM PST

Here's what I know about the computer
AMD FX-6300
Nvidia GeForce GT 640
16g DDR3 RAM
Gigabyte GA-78LMT-USB3 Motherboard

I really don't think it's a temp issue. The 60C was a ballpark, not like a recorded maximum. The record high for GPU/CPU is 54/51 C, according to OHM. On top of which, the case is open at the top, and I'd be able to feel if the air coming out(which is a lot) is hot. I am going to open the case and take a look at the fans for now. Just for the record, I ran a full scan and hard drive defrag as well as removing a lot of random stuff that I didn't need anymore and making sure I'm running minimal background processes. Again, I really don't think it's cpu/gpu malfunction as my cpu/gpu/RAM never get more than 50% in use and never heat up to abnormal levels. Is it possible that the fan is sticking and using a lot more power and crashing the PSU(sorry, I'm a complete noob at computers). I also have the resolution set to a custom res for the sake of the TV. I don't think that can do much, according to what research I've done, but the internet isn't always reliable.

- Collapse -
Just some info.
Mar 11, 2016 9:10AM PST

The number of PCs like the one you detailed has so many posts about crashes that I worry it's another one.

The 640 was not known to run cool even with it's TDP of 75 Watts.

It looks to me to be a machine from 2012? So my advice is to stop saying what it isn't and get the usual shop work done. When we see an old PC like this we do not say we'll fix it due to the history of such computers. We'll move right to "new motherboard and parts." This may upset a few but most clients want it fixed.

If this was mine I'd do the work.

- Collapse -
Expanding on the work.
Mar 11, 2016 9:18AM PST

Since it's older you:

1. Inspect all boards, even the PSU for "BAD CAPS" (see google)
2. Use the canned air, clean and replace all heatsink compound.
3. Pull memory, cards and reinsert to scrub the contacts.
4. Fans must work like new.
5. Check the BIOS and set the timing, speed to default or safe settings (varies with model.)

Re-test.

- Collapse -
Answer
Crashes
Mar 11, 2016 9:28AM PST

Take the side panel off and give the innards a good cleaning with some compressed air.
Get those heatsinks and fans nice and clean.
Don't forget the psu....a few shots of air from both directions.

Leave the side panel off and test.

- Collapse -
Answer
6 revisions
Mar 11, 2016 11:45AM PST
- Collapse -
In this case, I'd be updating the BIOS.
Mar 11, 2016 11:50AM PST

It's because there are issues now. It's worth trying all the usual PLUS what you added to avoid replacing gear. Some new techs always want to measure temps. The seasoned techs dive in and do the usual first.