Weeks of freezes, glitches, and BSODs, no solution yet

sperlingk1

Honorable
Apr 4, 2012
8
0
10,510
I've been having issues for quite some time now and tried everything I can think of....

System specs first.....All components are about 3 years old.

SYSTEM SPECS:

cpu = AMD Phenom II X4 955
mobo = Gigabyte MA790XT-UD4P
RAM = 8GB (4x2GB)
video card = nvidia 9800GT (replaced with GTX 550 ti)
sound card = HT Omega Claro
psu = Corsair 550W G24F264096-4N
secondary psu = FSP Group BoosterX5 450W (powers only the video card)
HDD1 = 160GB WD Caviar Blue
HDD2 = 1TB WD Caviar Black
HDD3 = 1TB Seagate Barracuda ST31000524AS
HDD4 = 1TB Fantom Drives Green Drive (External via eSATA)


Occurrences near beginning of problems, but may be unrelated: auto-updated nvidia drivers, and replaced a failed HDD (160GB WD) with a new one (1TB Seagate)

Graphics freezed initially (after auto-updating), so i rolled back my graphics driver, and everything seemed to work for another week or so until.........


SYMPTOMS:

Monitors would display "glitches" (either green or magenta pixels in wild arrangements) and the screen would freeze, seemingly when under a heavy graphics load. For example, when loading modern warfare 3, everything would work fine through the menus and the lobbies, but when a game round would actually start, the problem stated above would happen almost instantly. Now, after a number of fix attempts, the "gliches" no longer happen, but the freezes do. Audio always continues unaffected.

Alternatively, sometimes the monitors shut off as if there is no input (same circumstances). These two issues (monitors either freezing or shutting off) is the vast bulk of the problem. Either of the two will occur: always when under 3d graphics load; most of the time when playing videos and such online; and occasionally at random times.

Furthermore, these freezes have began occurring at all various points during system startup.

Two or three random BSODs during startup and normal use

One time there was an error I had never seen before, and unfortunately didn't right down the details. It occurred right after I turned my computer on. Something about the BIOS failing to initialize.... scary.... but that only happened once.


FIX ATTEMPTS:

Tried every driver available from nvidia as well as legacy drivers found elsewhere online. UNSUCCESSFUL.

Full antivirus and malware scans. UNSUCCESSFUL.

Verified cool temps with speedfan, but reattached gpu fan anyway to be sure. UNSUCCESSFUL.

Cleaned out case. UNSUCCESSFUL.

Swapped power connections. UNSUCCESSFUL.

Unplugged unnecessary components such as additional hard drives, dvd drives, extra fans, and lights. UNSUCCESSFUL.

Removed 4GB RAM. UNSUCCESSFUL.

CHKDSK on all drives. UNSUCCESSFUL.

Driversweeper and reinstall drivers. UNSUCCESSFUL.

MSCONFIG disable unnecessary items. UNSUCCESSFUL.

Uninstall software I wasn't using. UNSUCCESSFUL.

Reinstall java, flash, directX, Visual C++ redist. UNSUCCESSFUL.

Tried repeating most scans and reinstalls while in safe mode. UNSUCCESSFUL.

Purchase new video card (replace 9800GT with GTX550ti) UNSUCCESSFUL.

Note: I pulled an old video card (8500 GT) out of another computer before I had decided to purchase a new one. This card ran modern warfare 3 fine; there were no freezes. This is why I thought I just needed to buy a new one. (the only other difference this card had was that it only had one DVI port whereas I am running two on both the 9800 and the 550)


"NEW" SYMPTOMS:

I wiped out my sound drivers too while driversweeping. Then, the drivers wouldn't load back in and I had no sound. I finally got it working again at one point, but then after a while I lost sound again and it said that no device was plugged in. Now some generic devices appear in "playback" but they all say "not plugged in"

Now for no apparent reason only one of my two displays starts up... cant get the other to turn on at all.

Drivers sometimes do not load properly, sometimes it will only install a "generic vga adapter" when I try and let windows find drivers itself.


SPECULATIONS:

Power Supply Failure: this i was considering because of the freezes and lockups while under load, but now the other problems suggest this is probably unlikely.

Mobo Failure: the new video card is having basically the same problems (although without the screen "glitches") and the fact that the sound card is not being recognized. Onboard LAN also went out, but that was well over a year ago. I'm not exactly sure what typically would occur during a mobo failure, so i'm just speculating.

Virus/Malware: At one point, both malwarebytes anti-malware and MS security essentials seemed to be uninstalled from my system without my doing so.... but, every scan I do comes up clean.

Driver Problems: It began here, so it probably has a great deal to do with it, even though i think I've tried just about everything.



....I think that's everything. I'd much appreciate any help. I'm totally stuck.



 
Solution
Sounds like a PSU/motherboard issue then. Windows hard locking basically means the entire memory subsystem is completely messed up. So either the PSU isn't putting out enough/consistent power, or the motherboard isn't distributing it well. My 790i based mobo used to have the same issue, and I had to overvolt the Northbridge to 1.4V(!) to keep it stable.

mrmez

Splendid
You're on the right track. Isolate every single component, one at a time.
Unfortunately this means you need pretty much a 2nd entire spare system in parts.

Doing a zero fill on the HDD will ensure no nasties are left alive after a reinstall. The whole process is time consuming and probably a waste as id think this is hardware, i.e. Mobo.

The PSU is probably a little on the low side with everything plugged in, but should be ok with non essentials taken out, which leads back to the mobo...
 
i would start out by checking the outlet where the pc is..that working right and the wires are on tight. if it an older home sometime you get outlets that the power too high or too low.sometimes moving the pc to another outlet in an older home fixes the issue. i think what happing is there a bad cap in the ps or on the mb that controling power to the mb. pci video cards can get up to the max (75v) though the pci slot on and older video card the power needed might not have stressed the power part of your mb or ps to the point where is not holding good dc voltage. it called an ac ripple...unless you have an o-scop or tst tool to see if ps is making this ripple i would see if someone you know has a test ps. other issues that happen is what plug into the mb by ports...you could have to many devices on a usb port and it pulling to much power and locking or hanging the system up.
(know issues with some old mb and iphone/ipods) some of your issues are software and damaged windows...i would look at hardware issuse fix them then nuke and pave your main hd and sart with a fresh install of windows. start by check that the cpu fan is on fine and that the if it a stock cooler that the paste or tape is fine...it a know issue with thearmal tape as it gets older it does not work as well. I run some dos burn in like amidiags from usb or cd. with a bare bone system. main hd one ps and the video card and see if it locks up in dos or any hardware errors show up.
 
I'm with mrmez on this one, don't see you've tried a fresh install, need to completely wipe the HDD. Unless the one you replaced had the OS on it... Then again if there is an obscure virus, is it lurking on one of the other drives? Could try running the machine with just one HDD that has been nuked first...
 

alishathomaz

Honorable
Mar 6, 2012
128
0
10,710
Follow these guidelines::

Remove unused programs
Install and run antispyware software
Free up wasted space with Disk Cleanup tool
Defragment your hard disk drive
Disconnect unused network connections:

Click Start, and then click My Computer.

On the Tools menu, click Disconnect Network Drive.

Select the network drives that you no longer need, and then click OK.

Stop Unwanted startup programs:

Start > Run > write msconfig > Click on StartUp tab
Deselect the programs you want to stop

ALSO

Click Start - Run, type in %temp% and then click OK.

Once that temp folder is open and you can view its contents, select and delete EVERYTHING that's inside it.

It's all junk and is a good place for "nasties" to hide.

If a few files resist being deleted, that's normal. Leave them alone and delete EVERYTHING else.

Repeat the steps with c:\windows\temp

After you're done with both temp folders, restart the computer.
 

sperlingk1

Honorable
Apr 4, 2012
8
0
10,510
So far today....

Deleted the %temp% and C:\windows\temp files. Shut down my system as I was leaving the house. Left before it shut down completely.

When I got back, I had a BSOD. It read:
DRIVER_POWER_STATE_FAILURE
STOP: 0x0000009F (0x0000000000000004, 0x0000000000000258, 0xFFFFFA80087E8220, 0xFFFFF80000BA2740)

Disconnected all HDDs but my C: drive and restarted. Uninstalled and reinstalled drivers. Sound function came back. Video still failed after running ATI Tool (video stress test) - the monitor shut itself off as if there was no longer an input.

MSCONFIG, booted in diagnostic mode - only basic services. Ran ATI Tool again; freeze.


I think my next step will be to clear my C: and do a fresh windows install - that way I can rule out any software issues and start thinking about whether I need to get a new mobo, psu, or both.....

....unless there's more ideas....
 
Bug Check 0x9F: DRIVER_POWER_STATE_FAILURE

Farther, with a Bccode of 4, the cause of the crash is described as "The power state transition timed out waiting to synchronize with the PnP subsystem.".

So its looking like a driver problem right now. Run sfc /scannow and make sure none of the windows drivers are corrupt. Also using a different video driver would be a good next step.

I don't think its a GPU problem per-se, as I'd expect one of the TDP BSOD's [0x124 or 0x125], so I'm suspecting a bad driver rather then hardware right now.
 

sperlingk1

Honorable
Apr 4, 2012
8
0
10,510
Burned a Repair Disc so I could write zeros on my C: drive.

Finished that and I am trying to reinstall windows via USB.

When I got to "Where do you want to install Windows?" I deleted the existing partitions and created a new one, but when I hit next, I get the error: "Setup was unable to create a new system partition or locate an existing system partition."

What do I need to do to fix this? I formatted my drive so why isn't it able to install???
 

sperlingk1

Honorable
Apr 4, 2012
8
0
10,510
Turns out, pulling the CMOS battery out and waiting 10 minutes allowed the installation to continue properly. But after spending about 8 hours going through everything needed to reinstall, I downloaded ATI tool again and.....Same Freezing Problem!!!

So unfortunately I did all that and now I'm still back where I was......

It has to be the motherboard or psu or both????
 

sperlingk1

Honorable
Apr 4, 2012
8
0
10,510
I should mention, since this is certainly hardware related now....

I said I replaced a failing hard drive recently, but prior to that, the hard drive that was going out would only occasionally be recognized by the BIOS. Sometimes in the list it would appear, but mostly it would say either that no device was connected or would simply appear blank.

During this time, I put off buying a new HDD due to the flooding in Thailand doubling prices... I thought I'd wait it out. So in the meantime, I would have to hit the reset button numerous times during startup in order for my HDD to finally be recognized. I would have to hit reset anywhere from 3 to 15 times each day and this went on for probably 6 to 8 weeks... maybe longer. I knew this wasn't good for my hardware, but I just kept putting it off....

I didn't bring this up before because I thought for sure this was a driver related problem and because once I had installed the new HDD everything had been working fine for a little while....

So, if anyone knows what kind of damage constantly resetting would cause, I would really like to know.... What is most likely in need of replacing? Mobo? CPU? PSU? RAM?..... All of the above?

I suppose I could just run another memtest to verify the RAM is OK. But I don't have a proper mulimeter to test a PSU. My supposition is that it is the mobo that needs replacing anyway... but I wonder if the CPU is damaged as well.
 
Frustration! But if it ran smoothly for 8 hours installing everything then failed having loaded the ATI tool...presumably you did all that with basic m'soft drivers for the graphics, what is the ATI tool doing to cause the freezing would be my concern...
 

alien pirate

Honorable
Apr 6, 2012
1
0
10,510


HI: just seen your problem and it looks very similar to the one I had some years ago - do you have a dvi to vga (plug) converter on your video screen lead? if so is it damaged? mine was damaged and it caused the video card to burn out (eventually) which caused all sorts of weird 'software' problems it was only after I smelt the graphics card did I find it had failed replacement of the card cured all the problems
 

Well. the greatest 'strain'would be on anything involving a motor, constantly starting and restarting Hard Drives and fans etc. With that in mind and considering the cost of each item I would probably opt for the PSU as it will have been having to provide starting current to the motors of several HDDs when you reset the system. (If you lived next door I have an Asus M4A78LT you could have borrowed...)
 

sperlingk1

Honorable
Apr 4, 2012
8
0
10,510
It seems I MAY have a fix....

I raised all the voltages that were available to me on my motherboard (Gigabyte GA-MA790XT-UD4P) since I was thinking I was going to need to get a new one anyway, and it couldn't hurt..... My problems seem as though they might be fixed, but I am still going to need to figure out optimal (safe) voltages

With all I went through, if this was all I needed to do...... :heink:
 
Sounds like a PSU/motherboard issue then. Windows hard locking basically means the entire memory subsystem is completely messed up. So either the PSU isn't putting out enough/consistent power, or the motherboard isn't distributing it well. My 790i based mobo used to have the same issue, and I had to overvolt the Northbridge to 1.4V(!) to keep it stable.
 
Solution

mrmez

Splendid
 

mrmez

Splendid





True it could be either.

But from my personal experience, PSU's have always died catastrophically. No warning, no instability just a completely dead system, and a nice smell. Opening the PSU usually reveals a blown cap.
If the PSU was getting old and losing efficiency, you could increase or decrease the crashes by increasing or decreasing the load. Ie. Leaving everything plugged in and playing a game would cause an instant crash. Unplugging everything and just using the OS would result in no, or far fewer crashes.

Surprisingly and a little counter intuitively, motherboards have died in varying degrees. USB ports go, ram slots, PCI slots, but they all seem to share one major characteristic. Variability. Repeating or causing a crash with the same method will have mixed results, and the errors they cause can be just about anything. And thats the biggest problem. Because the errors are so non-specific, they can be caused by just about anything else, including endless software issues.

Again, until you test it with a new one you will never know.
 

sperlingk1

Honorable
Apr 4, 2012
8
0
10,510
I'm selecting gamerk316 as best answer because raising voltages does fix the problem. I do realize however that this could be only a temporary fix as is it likely indicative a larger failure in either the mobo or psu. Once I have the money, I'll probably replace the psu first (just because it will be easier to replace), then if there seems to still be a problem, I'll replace the mobo as well. But for now I'll stick with the voltage increase as it seems to have solved the problem. Thanks everyone for your help and ideas!