Random BSOD Problem

[lwstory]

UPDATED 4/14/11 12:07 PM

Resolution:
1.) Fixing why the computer would freeze: Apparently this was due to the qpi and dram voltage differential. I have to keep the voltages under a 0.45ish Volt difference.
2.) Fixing the BSOD problem: For some reason Zone Alarm caused my BSOD's. Even after reinstalling it, it caused problems. So no more ZA for me.

Problems: (Chronologically sorted)
• I have received various blue screens running windows 7 x64 bit with code 7f and had various random freezes during operation that did not result in a blue screen.
• Sometimes when freezes occured, my speakers would scream at me until powered off.
• Sometimes when freezes occured, my keyboard and mouse lost power.
• It would crash or freeze whether watching a video, playing a game, surfing the web, or doing absolutely nothing.
• One time it crashed and either the motherboard speaker or the video card speaker blared until the computer was shut off.
• It crashed twice when trying to install a fresh copy of windows 7 from cd.
• After the crash, BIOS runs very slow after post.
• My DVD-RW drive that is connected via PATA started giving me could not access errors in windows and slowed my system down whenever it had a cd. It eventually stopped working all together. The Eject button wouldn't respond nor would the led come on. I disconnected it because it was no longer working. Then I decided to give it another try and plugged it back in. Oddly it worked perfectly with no problems.
• I tried to boot my computer once and it didn't post. It gave me this beep sequence once: 1 long then 3 short beeps. I reseated the Video Card then the RAM and it wouldn't even post; the RAM led would light and and nothing else happened. Then I reseated the RAM again and it booted like nothing happened.
• Still had more crashes here is the blue screen code: 0x0000007f (0x0000000000000008, 0x0000000080050031, 0x00000000000006f8, 0xfffff80002e46eb8).
• During a game of MW2: The monitor shut off, sound stopped, and computer appeared to be frozen. I restarted the computer and it worked...
• Still more BSOD's

Troubleshooting:
• Running Memory at 800Mhz and 1.5V with 9-9-9-24 timings (still crashed)
• 5-6 Hours of MemTest86 (No Errors)
• 12 Hours of Prime95 (No Errors)
• 2 Hours of GPU Stress Test (No Errors)
• Monitoring CPU and Motherboard temperatures
• Assuring that there is no overclocking
• Reinstalling the operating system
• Remounting all devices: RAM, CPU, Graphics Card
• Updated BIOS
• RMA'ing the motherboard
• Checked to see if Windows Update (KB973879) was installed (it wasn't)
• Tested each stick of ram individually and received a crash on one and the comp froze then restarted and had a blue screen using the other.
• I ran WhoCrashed to see if anything came up for the problems I'm having. Zone Alarm and Avast were determined culprits. I uninstalled them and things appear to be working fine now.

My System Information:
• P7P55D-E Pro Motherboard
• Intel i7 860 CPU
• 2x2 GB Kingston KHX1600C9AD3K2 DDR3 RAM
• NVIDIA 8800 GTS 640MB Graphics Card
• 3 SATA Hard Drives, 2 of which in Raid 0
• 1 DVD-RW Drive via PATA
• 5 Fans (including cpu fan) every fan connection on the board is used
• PC Power and Cooling S75QB 750W PSU
• 3 connected USB Devices to USB 2.0
• Windows 7 64bit

Resolution:
1.) Fixing why the computer would freeze: Apparently this was due to the qpi and dram voltage differential. I have to keep the voltages under a 0.45ish Volt difference.
2.) Fixing the BSOD problem: For some reason Zone Alarm caused my BSOD's. Even after reinstalling it, it caused problems. So no more ZA for me.

 

[mlc]

Usually the BSOD is caused by memory failure, or some type of corrupt operating system component (bad driver, etc..)..

I would try the following:

1) see if the problem can be recreated in Safe Mode..... If not, it could be a graphics driver issue .. or some type of software interferring with windows ? (firewall, spyware detector, etc..)
2) Could also be a bad stick of ram, etc., s the Mem tests don't always expose these problems.. Try running with only one stick, and then the other ...
3) doublecheck that you have your ram running at the specs shown on their label.. (voltage and latency settings)
4) check to see if you're running with the latest chipset drivers for that motherboard, and check for the latest graphic card drivers. as well...

 

[lwstory]

1) see if the problem can be recreated in Safe Mode..... If not, it could be a graphics driver issue .. or some type of software interferring with windows ? (firewall, spyware detector, etc..)

I tried installing a fresh copy of windows and it crashed during the installation of windows as well... So I would imagine that it would independent of software.

2) Could also be a bad stick of ram, etc., s the Mem tests don't always expose these problems.. Try running with only one stick, and then the other ...

This could be a possibility, maybe memtest is not catching the problem.

3) doublecheck that you have your ram running at the specs shown on their label.. (voltage and latency settings)

I just made an edit above about this. I prolly posted it while you were viewing it and prolly should have just replied.

I made sure the ram is running under the right specs. Thanks!

4) check to see if you're running with the latest chipset drivers for that motherboard, and check for the latest graphic card drivers. as well...
Today 03:40 PM

Thanks again, I have updated all of my drivers including the chipset to no avail.

Thank you for your response, Any other suggestions?

 

[mlc]

.. sorry didn't catch the piece about it happening during installation... So you're right, that would suggest a hardware issue.....

I think intel recommends 1.5v memory with the i7 processors, and 1.65v is the absolute max limit.. But the rule of thumb is that you should be able to go up to 1.65 if the difference between your ram voltage and qpi/DRAM voltage is 0.5v or less..... confused? lol...

Bottom line.. I would try running that memory at around 800mhz or so, at the 1.5 volts.. just to rule this out.. then if that works, you can look to tweak the settings to allow it to run at its max speed....

 

[lwstory]

Thanks I will give that a try!

 

[lwstory]

I decided to look into this ram voltage and qpi/dram differential and decided to increase the qpi/dram voltage to make the difference 0.475v if I remember correctly; before the difference was 0.55v. The Ram is now running at its recommended settings and so far I have not had any blue screens or lock ups. So this may have been the problem. I will give an update in a few days if I don't have any problems. Thanks mlc!

 

[mlc]

good deal.. hopefully it remains stable...

 

[lwstory]

I did recieve one blue screen since I changed to the above differential, but realized the differential was 0.5v not 0.475. So I modified the voltage slightly so it is a bit under 0.5V. I would say that this has solved my problem. Thanks mlc.

 

[mlc]

glad to help..

 

[lwstory]

I thought the DRAM - qpi/DRAM differential was the resolution... but apparently not. I have had 2 random blue screens since the last post. I put my ram settings at 800Mhz and 1.5V as you suggested with the qpi/DRAM set to auto which was 1.1-1.2 volts. So now I think that isnt the solution. It seems as though it just makes the blue screen problem less frequent.

There are also two new problems that have arose:

1.) My PATA DVD-RW drive was acting up. I recieved some inaccessible messages when trying to access it in windows before and then shortly after having those problems, the device stopped working all together. The light does not come on for it or anything. I have yet to test to see if it has died on me or if it is something else.

2.) I tried to boot my computer today and it didn't post. I havent looked up the beep sequence yet but it did this sequence once: 1 long then 3 short beeps. I reseated the Video Card then the RAM and it wouldn't even post; the RAM led would light and and nothing else happened. Then I reseated the RAM again and it booted.

Ugh! What a pain!

 

[dbailey]

Any chance you can swap the power supply to see if it causing the issues?

 

[lwstory]

I would love to switch the power supply but I don't have a spare and I don't know anybody that would be able to loan me one for a week or two. At least its still under warranty...

I'm considering returning the PSU under warranty. Before I do are there any other suggestions?

 

[lwstory]

More interesting news, I looked into my DVD drive not working... I plugged it in and it magically decided to work again... I don't have any problems in windows like I had before. Its like it fixed itself! It was completely dead before! Whats going on?

 

[mlc]

.. sorry to hear you're still having problems.....

based on your latest symptoms, it sounds like it is some type of power related issue, either the board itself, or the PSU... Since you said in your original post that you RMA'd the board already, it's a good bet your psu is faulty...

.. the fact that you reseated your ram, and it was able to post.. could suggest a bad slot or seat.. Was the RMA'd board they sent back to you a new board or refurbished? In any case, try swapping the 2 ram sticks over to the other 2 slots, if you haven't tried that already....

... are you sure that you don't have any extra motherboard standoff's in your case not being used? Make sure there aren't any extra's mounted in your case, that aren't matched up with the motherboard mounting holes.. otherwise it could be causing a short....

 

[Atheus]

... are you sure that you don't have any extra motherboard standoff's in your case not being used? Make sure there aren't any extra's mounted in your case, that aren't matched up with the motherboard mounting holes.. otherwise it could be causing a short....

Good call but surely something like that would be picked up by stress testing...

OP - when you say CPU stress what do you mean? Have you run 12 hours of Prime95 on all cores? Have you run Memtest for 12 hours/overnight too? If you're bluescreening and neither of these pick up any problem it has to be software doesn't it? Maybe your Windows install is somehow becoming corrupted even though you've tried reinstalling. Have you tried another OS? Maybe XP or a Linux live CD with some testing tools on it?

/edit: sorry didn't read far enough down before replying - if it does turn out to be the PSU don't go swap it for another rubbish one - go buy a nice one. It'll last you years and you can be safe in the knowlege your machine won't explode while you're asleep.

 

[lwstory]

Atheus:

Ok, I ran prime95 overnight to have a more reliable test duration. Still no errors.

As for running under a different operating system, the computer froze during a fresh install of windows directly from cd so would I really need to test that?

Also, its not always a blue screen its mixed between blue screens and lockups. It froze again after my post yesterday during a game of MW2. It has also froze when not doing anything computationally intense.


mlc:

As for switching the memory to a different slot, I really can't. From my understanding, I have to have the DIMMs in either slot A1 or slot B1 to boot... When I place the DIMM in A2 or B2, the computer does not post and the RAM led remains lit. The manual says at least one of A1 or B1 has to have a DIMM.

As for the board they sent back from the RMA I have no clue whether it is a refurbished board or the same one repaired. I wish they would have provided some detail of what they did, but I just received a board back and no comments at all. It has the same serial number as the board I returned, but I cant be sure that that means its the same board or not.

As for the standoffs I'm sure they are correct.


Thoughts:

I would rather not buy a new PSU as my current one is still under warranty. So instead I'm thinking I should RMA it. I will hold off until Friday to see if anyone suggests otherwise.

 

[mlc]

If you're certain about the mount of the board being ok.. and you've pull out all the extraneous components (hard drive, usb connector into board, dvd player, etc)... and still get no post with only the 1 stick of ram, cpu/hsf, graphics card.... then it would seem to be down to the board, graphics card or the psu....

The freezing while you're in windows could be many things, but since you're not able to post, that rules out any software related issue, as well as the hard drive, etc.. but disconnect all of those peripherals ensure a bad cable/connection is not causing you the issue.....

Anyway you can get your hands on a spare or friends graphics card to rule that out.. You have an 8800 which is kind of old.. so that could be part of the equation as well.. I think we veered away from that due to your orignal posts being concentrated on the memory settings...

My inclination is that there is some sort of PSU or MOBO issue here... components randomly failing, then working again would point to an erratic power supply or a short in your setup (mobo itself or things connecting into it)

 

[lwstory]

Ok, I know there is a lot of things to keep straight in this post... I will update my first post with all of the issues and try to make it easier to read.

I mentioned that I was able to boot after I reseated my DIMMs and my video card. It was just odd that I had the error out of the blue and then it decided to work again.

As far as the cables/connections are concerned, I have ran with just the hard drives and fans plugged in. I could try removing the fans but they don't really consume much power any ways. This still resulted in at least a blue screen.

I am going to RMA my PSU and see if that is the problem. If that isn't, I suppose I will try switching my graphics card out with a different one.

 

[denis280]

Windows Update (KB973879) remove this and reboot.

 

[lwstory]

Thanks Denis280, I checked for the update and it is not on my computer. Even if a new windows update corrected that, I still had a blue screen since the last windows update. Thanks though!

 

[denis280]

Well do like Mlc. go with one ram. and see.myself im pretty sure it's a ram problem.

 

[lwstory]

denis280: I tested the DIMMs and received a blue screen on one and the computer froze using the other... I think I tested this once before too and I forgot to post about it. But yeah no luck there. Also, I would think the strange problem with the DVD drive and the strange non-post startup may suggest that its not the Ram. Who knows? Thanks for your response.

I updated the 1st post. I am still working on RMA'ing the PSU.

 

[Matt1970]

1 long then 3 short beeps is a Video error.

 

[lwstory]

So I think I have come to a solution on this. I never had the chance to return my power supply as I needed software that was installed on the computer. So, I dealt with all the BSOD's and pulled out a bit of hair...

All of the recent BSOD's were when I had a graphical animation (whether a website or a game like MW2). I no longer had computer freezes nor problems with my dvd drive. Lately, everything revolved around BSOD's up until the following discovery.

After some time I came across the software WhoCrashed and decided for the heck of it, I would run it. Interestingly, it mentioned Avast and Zone Alarm were possible culprits. I thought for the heck of it I would uninstall them. The computer went a while and had no BSOD's so I decided to reinstall them. Again, I had more BSOD's. I uninstalled ZA and and left Avast installed and have had no problems. The BSOD's before were getting to be at least 1 per day and now I have gone a week without ZA and no BSOD.

All my problems seem to have "magically vanished" for now anyways. I will follow up in a few weeks to post if this has been resolved. Thoughts?

 

[lwstory]

Indeed my post above solved my problem. I have gone with no BSOD's for a week and I am pretty confident that this is resolved. Thanks for everyone's help!