Help: Freezes "All The Time", No BSOD, Event Viewer: Kernel-Power EvenID:41 (63)

Toggle sidebar Toggle sidebar
Status
Not open for further replies.
M

militar3rd

Honorable
Jan 29, 2013
14
0
10,510
Hello Guys,

I have built my PC about 10 months ago, here are the Specs
CPU: AMD 8350 Vishera @ Stock 4.0Ghz Mobo: Asus Sabertooth 990FX R2.0
Mem,: G.Skill Sniper 1866Mhz 8GB (2x) GPU: GTX 670 Gigabyte OC 2GB (2x) HDD: WD Green 1.5Tb PSU: 750 Watt Rosewill Fortress Plat. Certif.

About a week ago, my system started freezing. Which was weird since the computer ran very flawlessly on intensive applications (3D Cad, Inventor, Film Editing) and ofcourse flawless in gaming. The problem kept on persisting and the frequency started to become closer and closer, until it came up to the point where I can't even last more than 10 minutes in desktop. Even if i take it to Safe Mode, it still kept freezing. Event Viewer read a "Critical" Kernel-Power EventID:41 (63) but there wasn't any BSOD, it just froze and neeeded a hard reset every single time it freezes.

Things I did:
- Format HDD, - Reinstall Windows 7 x64 and x86 prior, - Modified Power Options to never Sleep
- Reapplied Thermal Paste and monitored the Temperature (Everything was perfectly fine)
- Checked PSU w/ Multimeter (Every connection was fine and voltage checked out perfectly fine)
- Disabled redundant Audio Drivers that are conflicting within the system
- Ran chkdsk overnight (even after formatting hdd)
- Ran AVG 2014, CCleaner and RegClean Pro (Everything came out fine, no malware or virus)
- Updated the Bios through the USB EZ Flash
- Tried to run 1 Memory Stick on every single channel
- Pulled out the 3V CMOS Battery and is only putting in 2.15-2.4V (Fluctuating)
!!! Could this be the problem? I mean as far as I know, the battery controls the UEFI of
the motherboard and the basic information, such as the time and date.
Whenever I'm in the bios, it never froze on me. Haven't replaced it yet.

I'm floored here. The HDD seems fine to me. The Memory Sticks were being registered as 16GB by the MOBO with 14.6 Actual Available Memory. Although it looked like the Mobo only ran the RAM in 933 Frequency, when its rated at 1866 and everytime I'd try to up the frequency and voltage at the Bios it would indicate that "Overclocking Failed" once I turn on the PC, so I just left it at 933 almost all of its life. Could this have been the problem too?

The Sabertooth Motherboard has a 7-year Warranty. I think that it might be faulty, but I couldnt see any damage or electrolyte oozing out of the capacitors, nor that it is damaged.

I'm Floored here, any advice would be much much appreciated

THanks,
Santos
 
Sort by date Sort by votes
B

bigpinkdragon286

Splendid
Oct 3, 2012
4,648
35
28,590
933 is the correct bus speed for your RAM to be running at if it's rated for 1866 as the base frequency is doubled.

Even if your CMOS battery failed completely, your system would not suddenly generate a freezing problem such as you describe.

I would start by running a memory test such as Memtest86:

http://www.memtest86.com/

Memtest86+:

http://www.memtest.org/

or Windows Memory Diagnostic (manual instructions for running are at the bottom of the page):

http://windows.microsoft.com/en-us/windows7/diagnosing-memory-problems-on-your-computer

Give one or more of these memory diagnostic tests several full passes and see if you come up with anything, perhaps even leave one running overnight.

I know you say your temperatures are fine, but your CPU is really only good up to approximately 61°C before you're likely to experience erroneous results spitting out. Have you watched the temperature of your CPU to see if it's freezing when it reaches a particular temperature?
 
Upvote 0 Downvote
M

militar3rd

Honorable
Jan 29, 2013
14
0
10,510


Oh ok. I'm pretty sure that I have performed the Windows Memory Diagnostic Tool before I reinstalled the OS and it kept coming up with "No Problems Have Been Detected".

Won't hurt if I do it again though.

I'm going to do the memtest through USB in a few hours after work.

Thanks for reply.

Santos
 
Upvote 0 Downvote
M

militar3rd

Honorable
Jan 29, 2013
14
0
10,510
Alright. I ran memtest86+ for around 12 hours and 3 passes and came up with 0 Errors and Windows Memory Diagnostic was also negative.

I'm really hitting a brickwall here.
 
Upvote 0 Downvote
B

bigpinkdragon286

Splendid
Oct 3, 2012
4,648
35
28,590
Well, there is more we can test out.

One last thing to check as far as the memory is concerned, the settings in BIOS. It's possible the voltage and / or timing is incorrect. The easiest way I can think of to check is to read the SPD (Serial Presence Detect) data from the RAM modules themselves using a utility such as CPU-Z and then verify that is what your RAM is actually running at. Voltage might only be listed on the modules themselves.

After you rule out the RAM as being configured to run out of spec, you should move on to another component, such as the PSU. Do you have access to anybody who might have a backup or spare power supply you could test in your system? I know you tested yours with a voltmeter, but that really tells you only a fraction of the story. There could be excessive ripple, drooping under load, or noise which is affecting one of your components.

Essentially, it's the same for the rest of the components. You need to work your way through each, until you find the one device, not attached, which allows the computer to run in a stable fashion again.

You may also want to unplug unnecessary devices for now, such as any extra add-in boards such as a sound card, tv tuner, add-in network card, wireless card, reduce the RAM to a single stick, unplug any optical drives, all external devices except for your keyboard, mouse, and video. See if you can increase stability by running at a bare minimum, and work up from there.

I have seen weird grounding issues before as well, where an optical drive was too close to the motherboard and it would only freeze, reset, or shut down the system when you bumped the chassis (literally like sneezing at it,) or somebody had failed to remove an old riser / stand-off when installing a new motherboard, and over the course of a year it shorted the motherboard's power to the graphics. You may need to investigate closely if it's a home-built system, and see if each component is fully plugged in, both power and data, that all cables are free of damage, components haven't been damaged, that nothing is close enough to cause an electrical short, that no wires are falling off of their intended pin-headers, etc.
 
Upvote 0 Downvote
J

johnbl

Polypheme
Nov 4, 2012
10,072
185
67,090
A hang like that should be internal to a piece of hardware and its driver. Most often a card with a CPU inside (the graphics card or some fancy raid controllers, fancy network cards, and some solid state driver with firmware bugs)

you might also want to make sure that your BIOS has hotswap enabled for your drives. I have seen drives get disconnected because of various issues and not be able to reconnect.

basically any hardware that windows has direct control of will generate a bugcheck rather that a hang.
- Memory, corruption results in drivers being corrupted, the driver crashes and generates a bugcheck
- CPU can detect is own faults and calls a bugcheck directly
- Network cards, most are prettty simple and use microsoft software componets that will generate a bugcheck on errors
(some fancy cards with their own cpu and custom stack will just hang but the sympton is the device just stop working)
- Raid controllers (same as the network cards, but the OS can hang or bugcheck)
- Errors internal to a graphics card and its drivers can cause a hang
- Errors external to the graphics card, driver interface to windows should bugcheck
- Bugs in Chipset drivers can bugcheck or cause long delays depending on the issue
- Power issues to the motherboard result in various errors (mostly a reboot).
- CPU can call a bugcheck when its memory controller fails
- Power supply can tell the motherboard to reset the CPU (no hang, no bugcheck)
- mother board can reset the cpu if graphics card takes too much power from the BUS (no hang, no bugcheck)
- Cheaper motherboard just let the graphics card fail
-various failures with USB ports(what a mess)

There are more, but that is what I can think of off the top of my head.

Overall, I expect it to be a graphics subsystem issue but I would update your chipset drivers, move your sata cable to different sata port or secondary sata chip, and enable hotswap on your sata ports

force a memory dump and I will look at it to see if I can see any errors/issues in the memory image
http://msdn.microsoft.com/en-us/library/windows/hardware/ff545499(v=vs.85).aspx
 
Upvote 0 Downvote
M

militar3rd

Honorable
Jan 29, 2013
14
0
10,510
Ok finally a break. After running the memtest for 5 runs, which ended up being around 18 hourrs with no error. You suggested that to run bare minimum on the pc.
Already having everything stripped down to its bare necessities (1 ram and 1 gpu and the hdd plugged to a different SATA Socket and cable, the straight instead of the 90 degree, and w/o the optical hard drive) so I did that and the system ran almost flawlessly.
Although there were some times that I thought it was going to freeze while running applications such as chrome, cad, and some games but it was prolly because of the shoddy Internet.

Nextstep is to probably perform a prime95 test to see if its gonna fail.

Thanks guys, ill keep ya updated

-Santos
 
Upvote 0 Downvote
M

militar3rd

Honorable
Jan 29, 2013
14
0
10,510
Bad news. After a few hours of leisure use, didnt get to the testing part. The system kept freezing.

While I was watching a video, the monitor just went blank, nothing. And its staying blank for a while now even after a couple of tries of doing a hard reset, the power button on the case doesnt work anymore sovmy reset consist of turning the system on and off using the PSU.

Now im thinking that its the motherboard?
 
Upvote 0 Downvote
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom