Problems with an i5 750 - Could use a hand..

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510
First off, this is my first time posting in this forum, so thank you very much for your time in reading my post, and apologies if I've posted this in the wrong forum area (I'm hoping I haven't).

My computer has been giving me a huge headache, and while I'd like to consider myself pretty techno-savvy, I've hit a brick wall.

I bought an i5 750 processor, along with an Asus Sabretooth 55i motherboard, 8gb of OCZ DDR3 12800, and a Palit GTX 285 2GB Graphics Card. I built the machine myself with a brand new 1000W Power Supply.

After 6 or so months of working flawlessly, my computer began to blue screen and freeze/crash the moment I opened 3d video games. Thinking immediately that it was a graphics card problem, I sent the card in for RMA, but when I got it back, the problem persisted with a brand new card.

I found out that when I removed 2 of the 4 sticks of RAM in the machine, the freezing problem would go away but I would still get a screen blackout and a nvlddmkm error that would kick me out of most games.

I ran a full MEMtest on the memory to make sure I didn't have a failed stick.

I called OCZ, and they had me try different timings with my memory, but no timing setting would make my computer stable. It was then that they recommended I buy RAM with a lower base voltage, because the i5/i7 has a lower operating voltage. So, I dropped money on low voltage RAM that runs at 1.65v instead of 1.85v. The problem persisted (which also eliminated faulty RAM from the mix).

I thought perhaps that I might be having a motherboard problem, so I flashed the BIOS, and the problem persisted. I figured I was going to build another machine for my girlfriend anyway, so I purchased another Sabretooth 55i motherboard and switched it out for the existing motherboard. The problem... PERSISTED.

So, at this point, I've replaced the RAM, the Video Card, and the Motherboard. I've completely wiped the Operating System with a clean format and re-installed. I've tried the earliest and latest version of my video drivers. For those that are still reading (thank you), I'm using Windows 7 64-bit. I even took the heatsink off of my CPU, checked to make sure it was seated properly, and checked all of the pins on my CPU to ensure there were no bent pins or bad contacts.

I've tried all of these things and somehow, still, my games will black out (with 2 sticks installed instead of 4) anywhere from 30 seconds to 5 minutes into trying to play a game. If I try to run games with all 4 sticks of RAM installed (8gb), my games will instantly freeze when I load them.

When my screen blacks out temporarily, I get the system message "Display driver nvlddmkm stopped responding and has successfully recovered."

If my system freezes, I have to power down and start the computer back up.

Assuming there's nothing wrong with my power supply, I only see two possibilities. Either I have a faulty processor, or my motherboard is somehow incompatible with my memory, even though OCZ is on the vendor list in my motherboard manual and the computer was working fine for 6 months before I had these problems.

Am I missing something? Have I somehow created a problem with the voltage tolerance on my processor? Is there anything I can do to fix this problem short of buying another processor and eating another 300 dollars?

Would changing the voltage on my processor increase or decrease my stability? I haven't tried to overclock, was just running with the base settings.

It runs perfectly when I'm not playing 3d games, but the "Display driver nvlddmkm stopped responding and has successfully recovered." message has started creeping into simple things like opening a YouTube video.

I feel like I've tried just about everything and I don't know what to do. I feel like my expensive computer is a giant paperweight after all of the money I've sunk into it, and that's a horrible feeling.

Anyone able to offer help or insight on this problem? I'd appreciate it more than you know.
 
Welcome! :hello:

Sorry for your problems ... hope we can help.

I don't believe this is a hardware issue. When you tried the "earliest and latest version of my video drivers" were you confident that the previous drivers were completely removed?

Have you tried to disable the audio on the video card? Reinstalled Directx? Have you tried to disable physx?


edit: Sorry ... I forgot to add --- If this happens again (LOL) on another build I would suggest that you not reformat and reinstall your OS unless you are certain that is creating a problem for you. Prior to doing this, did you attempt to roll back to a previous restore point from when the system was functioning correctly?
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510



First, thanks a lot for the response.

The previous drivers were completely removed. I know because I did a full format and reinstall of the OS and was able to recreate the problem with the drivers off the boxed installation CD without allowing windows to perform its first update.

I tried disabling PhysX.

I wasn't aware my video card had audio.. how would I even go about disabling it? If you're referring to my motherboard, I switched out for a brand new motherboard, which should have (hopefully) eliminated that from the list of the usual suspects.
 


It's been a while for me with nVidia --- hopefully some other folks can help you with this. The same issue occasionally pops up with AMD Raedeons ... so hang in there!
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510


Unfortunately, that site points to the error message (which is a VERY broad and general error message) and then offers a bunch of solutions that I've either eliminated as possible causes or don't apply to me (desktop not a laptop, registry changes when I've already reformatted, etc).

I really appreciate the resource, though. Thanks for trying to help.

Any help is appreciated at this point.
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510


Such a change would be done with software though, or be a BIOS setting? Either way, I've both formatted and re-installed the OS and replaced the motherboard.

A BIOS based or OS based solution seems impossible by that right.
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510
Could changing my processor voltages make my system more stable? It seems like my processor and power supply are the only weak links, unless there's some kind of hard drive problem that would persist through a format and a re-install of an OS.
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510
Only a failing psu can cause those problems.Did you try a different hdd?

I didn't try a different one, but I did a clean format and reinstalled the OS. The operating system drive is two 74GB SATA Raptors in RAID 0, so I'd assume if I was having a drive problem, my RAID 0 configuration would be falling apart.

Does this power supply look like a suitable replacement? http://tinyurl.com/2altmmj
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510
As a side note, I've been doing some reading about the NVLDDMKM error, for those of you who are not familiar.

The "Display driver nvlddmkm stopped responding and has successfully recovered." message is from a fail-safe built into later operating systems that stops a system freeze that is caused during the course of the GPU functioning (though it doesn't necessarily indicate a graphics card problem).

This means several things:
1: This mechanism can be disabled (for debugging purposes) by adding a REG_DWORD named TdrLevel to the registry at: HKLM\System\CurrentControlSet\Control\GraphicsDrivers and assigning it a value of 0. Doing this will simply stop the TDR from resetting your display driver (this may simply tell your computer to allow itself to freeze instead, read on).

Another thing you can try (for debugging purposes) is adding a REG_DWORD named TdrDelay instead of the TdrLevel entry. This entry specifies the number of seconds that the TDR will wait before resetting the GPU/Driver. The default value is 2 seconds, so you can up to to 4 or 10 seconds to see if your system regains responsiveness. Don't leave it at the default value of '0'.

Keep in mind that in most cases these registry entries will not be there and you will have to create them. NEVER mess with the registry if you don't know what you are doing.

To remove the keys (after debugging), simply use the Find function and match it to TdrLevel or TdrDelay, respectively.

Source: http://www.microsoft.com/whdc/device/display/wddm_timeout.mspx#E1B


2: What this error really means is that your system is freezing when you're using applications. All the TDR function really does is stop freezes in programs by trying to reset the display driver/graphics card.

This means that this message can come as a result of any problem that can cause a system freeze (and might self-correct with a display reset).

Typical Examples:
Graphics Card Problems
RAM Problems
Power Supply Problems
CPU Problems


3: This error message is just like a fever. It's not a problem in and of itself, rather it is a symptom that can be caused by many different problems. As such, any standard solution that helps one person may not help another.



Common Solutions Include:
Using a Different Display Driver Version
Disabling PhysX
Replacing a Faulty Video Card
Down-clocking Video Card
Cooling / Heat Build-up Issues
Reducing Memory Timing
Replacing Memory
Reducing Number of Devices Connected to Power Supply
Replacing Power Supply
Checking the CPU for bent pins
Checking thermal contact between CPU and Heat-sink
Replacing the CPU
Checking to see if Heat-sink is seated properly


All I have left on the list is the Power Supply and the CPU. I've tried everything else. Hopefully going through this experience will prove useful to someone else later on who is having the same problem. I know, at the very least, that I've learned something from the experience.
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510
Well, got a brand new Power Supply today. Installed it. After about 10 minutes, got my NVLDDMKM error.

At this point the only thing I haven't replaced is the CPU.

New PSU
New Motherboard
New RAM
RMA'd Video Card
Replaced Case Fans

How.. irritating. Any thoughts?
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510
Purchased an i7 870 CPU from New Egg. I'll replace my current processor with it and see what happens.

You would think that since this is the last stone to be turned over, that I would be happy. Honestly, I'm dreading it. If I put it in and I still get the error, I'll be back to square one.

However, if I'm lucky and this fixes the problem, then perhaps I can send the old i5 back to Intel (I bought it less than a year ago) and I can have a second working processor to put into the spare motherboard, along with the spare RAM, Power Supply, and other things I have as a result of this mess. Maybe I can get a working computer up for my girlfriend.

I'll let everyone know how it goes. Maybe my experience will save someone else a headache in the long run.
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510
Well, I installed a brand new i7 870 and I'm still getting a "Display driver nvlddmkm stopped responding and has successfully recovered." message in my games. Sometimes it will go for 5-20 minutes without crashing, other times it will crash the moment it loads.

At this point I've:
Re-installed Drivers after Uninstalling fully
Rolled Back to a Previous Driver Version
Tried Downclocking the Graphics Card
Tried Using TweakForce.com's Alternate Drivers
Re-Installed the OS after a full format
Replaced the Graphics card via RMA. Twice.
Replaced the RAM
Replaced the Power Supply
Replaced the Motherboard
Replaced the CPU

My system was working perfectly for 6-7 months before this happened. At this point, I've pretty much replaced everything I can. I'm at the end of my rope.

Hell, I'm practically using a brand new computer at this point. I'm lost. Does anyone have anything else I can try?
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510
I found a workaround solution.

I installed some alternate 64-bit drivers for Nvidia here:
http://downloads.guru3d.com/Forceware-182.47-Vista-64-bit-download-2226.html

I'm pretty sure that they work because they have absolutely no support for higher versions of DirectX (these drivers are from all the way back in March of two years ago).

I had tried using earlier versions of drivers before, but these drivers go back even further than the oldest in Nvidia's driver archive (not to mention they seem to be a different version type). What still boggles me is why my computer worked perfectly for 7 months and was rock-solid with Nvidia's drivers and then suddenly became so finicky and fickle.

My games now run flawlessly, but I had an issue with getting a blue screen every time I opened up a YouTube video. I solved the issue by disabling hardware acceleration in Flash (right-click, settings).

So, I finally found a light at the end of the tunnel, and I can use all these extra parts I've accumulated from this troubleshooting nightmare to build my woman a new PC.

Thanks for the help, for those that contributed. Hopefully this solution will help someone else at some point in the future.
 

AnsharSeraphim

Distinguished
Nov 17, 2010
19
0
18,510
Update:

The workaround solution does work, but I kept getting nvlddmkm BSODs while streaming video, especially webcam video full screen.

I found another workaround, though it's bizarre and I'd love someone to explain to me why it works. I changed my driver version to 190.62 (so I could use Riva Tuner), installed Riva Tuner, and started changing my GPU's clock settings to see if I could make it more stable.

If I OVERclock my card from 648mhz to 720mhz, open a game that's able to recover from TDR resets, wait for the problem to occur, and THEN put my clock back to its standard settings, my card will remain as stable as a rock until I restart the computer.

After I restart, I can't leave my card at the default settings. I have to over and then underclock again.

It's a bizarre workaround, but it works and also leaves my full screen video stable, so I suppose I can't complain.

It's so weird, though. It points to the possibility that my card is STILL the culprit, even though I've RMA'd it twice.

Don't know what to think.