Sign in with
Sign up | Sign in
Your question

Hard Locks with new build, already ruled out some components

Last response: in Systems
Share
March 19, 2013 12:34:58 PM

Hi all,

Looking for some advice here on what to try next to isolate what is causing my random hard locks. Sometimes I can game for hours before they occur and other times the computer will stall when logging into Windows and recover 30-40 seconds later.

I should note that the stalling and hard locks appear to different and the hard locks seem to only come after a long uptime and doesn't recover like the stalls.

So far I have ruled out the following:
- Temps across the board are nice and low even under stress test loads
- It is not the HDD as the problem persisted using a different drive
- It is not the SATA cable or SATA port as they persisted after changing both
- It is not the GPU as the problem persisted using a different GPU
- It is not the Windows 8 install as it persists even after a clean install
- I have reset the UEFI to defaults to return everything to stock and the issue persists with and without any overclocking
- I ensured the timing values were correct for the RAM in the UEFI and gave the RAM a slight voltage boost to 1.46v and still noticed the stalling (not enough time to test for hard lock yet)

My next steps:
- Adjusting timing from stock values
- Replacing RAM

In short it seems to be pointing at the RAM not getting along with my motherboard. Outside of the RAM that really only leaves me thinking it may be a faulty Motherboard or PSU. (I wish I had a multimeter.)

Any and all advice would be greatly appreciated. Thank you!

Specs:
Motherboard: ASRock Z77 Extreme6
CPU: i7-3770K
CPU Cooler: Corsair Hydro Series H100i
Memory: 4x4GB Samsung 3V4G3D
Memory Speed: 1600MHz
Video Card(s): Asus GTX Tittan
Sound Card: HDMI to Pioneer SC-57
PSU: Corsair HX1000w
Hard Drive(s): Samsung 840 250GB SSD (boot) and 3TB HDD (media)
Operating System: Windows 8 Pro
March 19, 2013 12:40:48 PM

I was having similar problems with my Gskill low voltage until I upped the voltage a tad do you have any spare sticks of ram you can try out
March 19, 2013 12:49:56 PM

Pgooch said:
I was having similar problems with my Gskill low voltage until I upped the voltage a tad do you have any spare sticks of ram you can try out


I am going to try running with the voltage bumped to 1.45v a bit longer, then if that doesn't work, adjust the timings. Lastly if that all doesn't help, I will indeed swap the RAM.

It is just a pain to troubleshoot since it happens sometimes after hours of use. :( 
Related resources
March 19, 2013 1:03:22 PM

When I was having the problems I would get a couple minutes into windows and it would stall and hard lock.. Since it happens after hours of use could just be a bad pin on your ram. But I definitely think it's a memory problem have you tried running a memory test?
March 19, 2013 1:32:00 PM

Pgooch said:
When I was having the problems I would get a couple minutes into windows and it would stall and hard lock.. Since it happens after hours of use could just be a bad pin on your ram. But I definitely think it's a memory problem have you tried running a memory test?


I forgot to mention that I did try running with only two of the sticks, it happened after a few hours, then I tried with the other two and it happened again after a couple hours.

Possibly a bad DIMM slot?
March 19, 2013 1:38:07 PM

Have you overclocked?

If swapping RAM out does not improve the situation, I would suspect the CPU before the board. That's where the memory controller is, and it's rather sensitive.
March 19, 2013 3:48:45 PM

Proximon said:
Have you overclocked?

If swapping RAM out does not improve the situation, I would suspect the CPU before the board. That's where the memory controller is, and it's rather sensitive.


I have overclocked to 4.4Ghz without a voltage bump and the temps never exceeded 72c even with Prime95 and the Intel Burn-in test. At that speed the temps stay around 60c under load.
March 19, 2013 6:20:26 PM

It's more tied to the RAM. Overclocked RAM can cause the memory controller, which has its own voltages that are almost always left on auto, to become overworked. It is NOT related to the RAM voltage or the CPU voltage, at least not directly.

One of the voltages is called VCCIO.

It could simply be that your RAM isn't compatible with the board of course. Happens all the time.
March 19, 2013 7:15:59 PM

Proximon said:
It's more tied to the RAM. Overclocked RAM can cause the memory controller, which has its own voltages that are almost always left on auto, to become overworked. It is NOT related to the RAM voltage or the CPU voltage, at least not directly.

One of the voltages is called VCCIO.

It could simply be that your RAM isn't compatible with the board of course. Happens all the time.


I upped the voltage to 1.5v (per standard) to see if that is the issue and so far so good but have only played a couple hours of Far Cry 3.

I did get the random hard locks with the RAM at stock 1600mhz.

When RAM isn't compatible with a board, the issues can manifest with a random hard lock many hours in?
March 20, 2013 1:15:55 AM

Yes, absolutely. We see BIOS updates often to address such instability. Have you checked your actual RAM settings against the SPD tab in CPU-Z? That's not all the setting but you can at least confirm the main timings.
March 20, 2013 7:44:47 AM

Last night I went digging through the Event Viewer and saw errors that happened when I was getting the weird stalling during boot. (The IO operation at logical block address # for Disk # was retried)

I realized that both my drives were on the 3rd party SATA 3 ports. I moved both of the to the Intel ports, booted 5 times without the errors and much faster, without any stalls.

After that I benched the SSD and it was significantly faster.

I feel like an idiot but that very well may have been the cause. I had switched the ports and cables but only back and forth between those two ports.

Hopefully those 3rd party ports were the cause of my freezing during gaming.
March 20, 2013 2:21:55 PM

That's great news. Have you checked the drivers for those ports? If they are up to date you should at least let ASRock know about the issue, even if you choose not to RMA.
!