Intermittent crashing

Agent766

Agent766

Member
Joined
Feb 24, 2014
Posts
24
I've been having a lot of trouble for a long time. At least 6 months. My desktop's been crashing on and off. When it crashes, it completely loses power like the cord was unplugged, no BSOD. This afternoon it was crashing every 30 minutes. I checked the temps at first and the only thing getting higher than 40C was the northbridge which was spiking to ~100C. Maybe that's the cause, but I don't know what to do about it. I also tried replacing the PSU which didn't help.

Sometimes when I boot after crashing it says "A disk read error occurred. Press ctrl+alt+del to reboot." As far as I can tell though, my SSD (and other drives) are passing every test I can throw at it.

Speccy: http://speccy.piriform.com/results/624XgpOdtmXzy8y8keGzKx8

I'm not sure if it's stress related. It seems to have a lot lower chances of crashing if I'm not in a game, but it has happened. It does seem pretty likely to crash immediately after rebooting.

I'm at my wit's end with this so sorry if I'm leaving anything important out.
 
It sounds like an overheating issue to me...and according to your speccy report, your cpu temp is at 61*C. Is that an idle temp or load temp? Rest of your temps look good as well as the rest of your info.

First thing I'd check is your CPU cooler and fan. Make sure the cooler is correctly installed, has the right amount of thermal paste (to much or to little can cause overheating), the heat sink is not clogged with dust as well as the heatsink fan.
 
On top of what Bass (hi Bass!) mentioned above, how is the ventilation on your CPU case? How many fans do you have? Insufficient ventilation is also a big factor with overheating. You can try to open up your CPU case and leave it like that for a while and see if your PC will continue crashing. If it turns out okay, you may want to add more fans if you have slot available and close the CPU back.
 
Hello Agent 776,

I do agree with the above that heat may be an issue, but 61 C is not a high temperature for the AMD FX processors, especially the Octa-cores like the one you have.

I am still going through the Speccy report, so bear with me.

I would like you to download Hardware Monitor from the following URL.

HWMonitor CPUID - System & hardware benchmark, monitoring, reporting

The .zip file would be fine, and it doesn't install files on your PC.

Extract the archive and run hwmonitor.exe.

Let it run (an hour or so) while you use your computer normally then take a screenshot and post it in your next reply.

I will get back to you as soon as I get through the Speccy report.
 
I wasn't able to get through your Speccy report fully yet and I have to go for a while. (Work)

One thing I noticed...

Your CPU is running at it's turbo frequency on all cores... Is your system overclocked? If so, remove the overclock.

Check AMDs website for updated drivers for your GPU. This is just a suggestion, not a diagnosis... I have seen stability issues caused by AMD drivers.

And further troubleshooting... Check your EFI/BIOS settings for your RAM and make sure the Voltage and Frequency match the DIMM specifications (if the system is OCd then the RAM may very possibly be OCd as well). When you verify the settings are correct, download Memtest86+ from Memtest.org. Follow the instructions for creating bootable media then shut down your computer. Remove one DIMM and place it in a safe, static-free, environment. Reboot your computer with the remaining DIMM installed and boot to the Memtest media.

If any errors are reported, you may stop the test at any time. If no errors are reported, let the test run for approximately two hours or until errors are reported.

When done, swap DIMMs and re-run the Memtest.

Report any errors in your next post. I will check back as soon as I can.
 
Actually, here's a HWMonitor Pro log that was recorded up to a crash: https://drive.google.com/file/d/0BxWUbsdP3vAzdjlVXzR4QjJ3MDQ/edit?usp=sharing

The case is an Antec LanBoy Air with plenty of fans in a cool environment. I'll check the overclock again, but I believe I restored all of the BIOS settings to default when I started having this issue. One thing I did notice, is the my RAM seems to be running under what it should be. I have this module which looks like it should be running at 1600Mhz. I tried forcing it to 1600 in the BIOS before, but the system wouldn't POST with that setting. But it's been fine for a year and a half running at the slower speed. Could that be a problem?

I'll get a new HWMonitor screenshot and test the RAM again. I ran memtest86+ with both DIMMs in previously with no errors, but I can test it again with one at a time.
 
7BJqU


TMPIN2 is the northbridge. This is only after about half an hour. I started taking periodic screenshots after starting it because I knew it wouldn't last long. The BIOS had Auto set to everything in the way of CPU speed, but I reset them to the defaults anyway.
 
I have taken a particular interest in this case because it mirrors a very similar issue I was having with my own PC about a year ago.

I was having very similar issues. Freezing, crashing and immediate power-offs just as you are describing...

My issue ended up being the RAM, which just happened to be Kingston HyperX.

Your CPU temperature does seem to be getting pretty warm. What were you doing with your PC during the hour that HWMonitor was running?

Also, you mentioned you have an Antec Lanboy and HWmonitor is only showing one fan connected to a motherboard header.

How many fans do you have, what is their configuration (placement and airflow direction)?

Also, are you using the stock CPU cooler or an aftermarket? If aftermarket, which cooler?
 
Stock CPU fan. I've been attempting to play World of Warcraft and watch Netflix or Twitch. Chrome playing video seems to warm up my CPU. I have 5 fans, two on front blowing out, and the two on the side and one in back blowing into the case.
 
Okay... Here is what I want you to do...

Flip the fans in the front of the case to intake. They are there to direct air over the HDDs to keep them cool.

Swap the fan on the rear to exhaust heat. It is higher than the other fans and natural convection will cause the heat to pool at the top.

Your case has a spot for a top mounted fan as well. Install that fan and orient it to exhaust air out of the case.

As for the side fans... They should be situated above the GPU, correct? If so, leave them pulling air into the case.

The lanboy is a mesh case and it is pretty open, so there is a very high chance for dust buildup inside the case. When was the last time you cleaned the dust out of your PC?

Check the CPU fan blades and cooling fins to make sure they are free of dust buildup. Even the finest film of dust will reduce your coolers performance drastically.

What we are attempting to do here is create a better airflow pattern inside the case and try to eliminate "hot spots" (areas where hot air pools and creates heat pockets).

Also, how is your cable management?

EDIT:

One more question... Do your fans have speed settings or other form of fan speed control?
 
Two of the fans have low/high speed settings and are set to high. The case only comes with the 5 fans, but I may have another I can install in the top. I don't know if I still have the hardware for mounting it, but I'll see what I can do. Cable management's pretty bad, but I'll work on it. I had the case apart a month ago when I tried a new psu and made sure to clean out the little amount of dust then.
 
That's the biggest drawback...flaw...with the Lanboy. It's and open case design all right but that for got about cable management.

Have you checked your CPU cooler and heatsink/fans...to make sure they are not clogged and the heatsink is not loose?
 
Okay. I suggest putting the fans with the low high settings as intake fans and setting them to low.

I think it's strange that the case didn't come with the top exhaust fan. I rarely see that.

Most important thing is to get the fans situated and see how much of a difference it makes in your temperatures (especially the chipset temp)
 
aha, looks like one of the two side fans (the one that happens to be right over the nb) looks like it seized up. I did find two 120mm fans, but it doesn't look like I have the screws I need to mount one on top.
 
Yes, it is. Looks like the fan may have been the problem with the crashing. Northbridge is still a bit toasty, but 90C is a lot better than 100C.

7CcJZ.png
 
My apologies for coming in so late.
I have 5 fans, two on front blowing out, and the two on the side and one in back blowing into the case.
Click to expand...
I think that is (was) much of your problem. It is important to note it is the case's responsibility to provide a sufficient supply of cool air flow through the case. And generally, you want a front-to-back flow through the case.

What puzzles me is Antec factory installed case fans ARE installed for front-to-back at the factory (I've been using Antec cases almost exclusively for 20+ years). So how did your rear and front fans get reversed?

For side panel fans, I have found they sometimes help, but sometimes they disrupt the front-to-back flow. Typically I don't use side panel fans UNLESS they connect to a tube that directs the fan's flow directly onto the CPU or GPU.

Some chipsets come with small fans and they typically are lousy quality (noisy and seize up easily). I used to replace the chipset fans but replacements became harder to find (and were typically of same lousy quality) so I just removed the fan, cleaned the heatsink, ensured I had excellent front-to-back flow of cool air and surprisingly, no chipset overheating problems.

While true some CPUs are rated for temps higher than 60°C, I do not let my CPUs sit at or above that temp for more than a few seconds. While no harm to the CPU is expected, stability issues (like system reboots, freezes, and unexpected shutdowns) can occur. So when my CPU temps start touching 60°C, it is usually a sign I have a layer of heat-trapping dust blanketing the innards and it is time for some housecleaning.

Stock AMD and Intel HSF assemblies are excellent! Don't let anyone suggest otherwise. The only time you need to replace the stock fan is if it fails (rare - and they have 3 year warranties), with extreme overclocking (not recommended unless you really know what you are doing and have properly addressed cooling), or when setting up a HTPC (home theater PC) and you need "passive" (no fan) cooling for "silent running" (often done with "underclocking" for even cooler operation).

I note your first Speccy dump shows the motherboard's "System Temperature" is a nice, low 30°C. So I am not sure I am buying HW monitors 90°C readings.

I recommend a couple things. What are your temps in Safe Mode? What are they in your BIOS Setup Menu? Note when checking temps in the BIOS Setup Menu, your temps should be about as low as they can go because running the BIOS Setup Menu is about the least demanding task we can ask of our systems.

Finally, I recommend running (at least while troubleshooting) Gigabyte's own HW monitor, EasyTune6 (as found on your motherboard utilities disk or from your motherboard's download page, then check the HW Monitor tab. If your System temp there is show as high, then that's a problem as you can expect Gigabyte HW monitor properly sets reading to sensor.
 
Back
Top