Solved Computer Freezes and BSOD after a system upgrade

[BelGarion]

Right so a couple of weeks ago I get the parts needed to upgrade my AMD Phenom II X6 1045T system. I got a new motherboard, an MSI 990FXA-GD80 Rev 2, and 16gb of DDR 2 1600 GSkill memory. Also got a new Fractal Design Define R5 case. I also got a Western Digital WD2003FYPS-27W980 2tb Enterprise Grade hdd as a new primary HDD. Now, this gets a bit convoluted so please bear with. Originally I had a 1.5tb primary, a 2tb and a 640gb data drives. The idea was to make the new 2 tb drive the primary, and replace the 640gb with the old 1.5tb primary. So I get it all together, swapped over the CPU, etc and get it working. I reinstall Windows 7, get it updated, all running relatively well, partition the new primary as a 500gb C: and 1.5tb D:. 2 days later Win 7 *loses* the new primary, literally, Windows Explorer, which I had up copying files off the 1.5tb doesn't show C: and D: not there. Programs crash left and right, complete chaos. I reboot into the bios and bios shows the HDD is there. I reboot and it boots fine if you don't mind the data corruption out of the wazoo due to the drive going missing. So I limp along for 2 days until I have a day off of work and then reinstall Win 7 again. With in the first day it BSOD twice, both Kernel Inpage Data Error. A week later it does it again while attempting to play a video game. And now and again it'll freeze, I do mean whole system freeze, not even the clock ticks off, for several minutes. So I run Passmark on the HDD to see if it's failing and it passes with flying colors. I boot into a USB thumbdrive version of Memtest86+ and the memory passes no errors. So by this last Monday I take out the two 2tb drives, put the old 1.5tb drive in on SATA port 1, and try to install Win 7 onto it. No go. It *seems* to be failing as it's going chick chick chick chick chick and it took a lot longer to install Win 7, lot longer to boot into Win 7, too shut down Win 7, etc. I unplug all the other drive power connectors so it's the only drive on that rail, same problem. My Raidmax 850w power supply has 4 drive rails so I move it to one of the other drive rails, same problem. I swap it to port 2, same problem. I swap out the data cable, same problem. I give up and put the new 2tb in as port 1, the other 2tb as drive two, and left the 1.5tb out. It started off seeming ok, played the same game for several hours with out problems. But tonight I've had 4 hard freezes and one BSOD, same error. I'm at a loss, everything tests out fine yet I am having these issues, the only thing I can't do is test the controller, there is only one 6 port controller on the motherboard. I would have to, quite literally, tear out this motherboard and swap the CPU and heat sink over to the other motherboard, and I don't think I have enough thermal compound to do it. And if that doesn't fix it, swap it all back again with a lot of wasted work. Though it is strange my old primary to start to fail like that after years of flawless service. Never had issues with Seagate or Western Digital.

 

[TVeblen]

There is no reason to believe the old hard drive can't be dying. There are 2 times when it is most common for hard drives to fail: when they are new and when they are old.

The "It *seems* to be failing as it's going chick chick chick chick chick " is the dead giveaway however. Classic symptom of the heads not parking. That drive is on it's way out. Start retrieving and backing up data now.

How does the PC run if you remove the failing hard drive?

 

[BelGarion]

There is no reason to believe the old hard drive can't be dying. There are 2 times when it is most common for hard drives to fail: when they are new and when they are old.

The "It *seems* to be failing as it's going chick chick chick chick chick " is the dead giveaway however. Classic symptom of the heads not parking. That drive is on it's way out. Start retrieving and backing up data now.

How does the PC run if you remove the failing hard drive?

Same, 4 hard freezes and one Kernel InPage Data Error BSOD after.

 

[TVeblen]

This raises the possibility that it is your new hard drive.

You should run CHKDSK on it

Open Windows Explorer > Expand Computer > Right click on the C: drive > click Properties > go to the Tools Tab > under Error Checking, choose Check Now > for a detailed test, check both the boxes: "Automatically Check For File System Errors" and "Scan for and attempt recovery of bad sectors" > since this is the system drive you will get a message saying "Windows Can't Check The Drive While It Is In Use" > click Schedule Disk Check > Make sure all programs are closed and then restart the computer
DiskCheck takes a while. Don't interrupt it.

 

[BelGarion]

This raises the possibility that it is your new hard drive.

You should run CHKDSK on it

Open Windows Explorer > Expand Computer > Right click on the C: drive > click Properties > go to the Tools Tab > under Error Checking, choose Check Now > for a detailed test, check both the boxes: "Automatically Check For File System Errors" and "Scan for and attempt recovery of bad sectors" > since this is the system drive you will get a message saying "Windows Can't Check The Drive While It Is In Use" > click Schedule Disk Check > Make sure all programs are closed and then restart the computer
DiskCheck takes a while. Don't interrupt it.

No errors found, though it did, when scanning the files, pause on occasion and not progress for several seconds or move forward slowly in 1s and 10s instead of 100s and 1000s it was doing previously.

Chkdsk Log File

 

[TVeblen]

Mmmm. We are running out of parts that could cause your symptoms.

Kernel InPage Data Error indicates a problem with the hard drive or the RAM. But you have tested the RAM, the hard drives, the cables. We are assuming the power supply is not the problem, but you could try lessening the load by disconnecting everything except the OS hard drive to see if that makes any difference.

So you really only have the new motherboard left. It is a lot of work but it may be time to do the switch. If everything works with the old motherboard and it's old RAM then it is a cinch.

I assume the new RAM is matched to and compatible with the new motherboard? Or did you just move the old RAM over?

 

[BelGarion]

Mmmm. We are running out of parts that could cause your symptoms.

Kernel InPage Data Error indicates a problem with the hard drive or the RAM. But you have tested the RAM, the hard drives, the cables. We are assuming the power supply is not the problem, but you could try lessening the load by disconnecting everything except the OS hard drive to see if that makes any difference.

So you really only have the new motherboard left. It is a lot of work but it may be time to do the switch. If everything works with the old motherboard and it's old RAM then it is a cinch.

I assume the new RAM is matched to and compatible with the new motherboard? Or did you just move the old RAM over?

Nope, bought some Gskill DDR 3 1600 8gbx2 16gb kit. The specs say it will take up to 1600 with out OC and two more beyond that with OC. Let me see. To quote Newegg 'Memory Standard: DDR3 800/1066/1333/1600/1800*/2133* (OC)' . Kinda had to, the old is DDR 2, which is one of the major reasons for the update, slower memory and limited to 8gb.

 

[TVeblen]

There are more important specs to match in memory than the speed. The voltage range must be correct, and the timings are important too.
Did you check to be sure the RAM is compatible with that motherboard?

 

[BelGarion]

There are more important specs to match in memory than the speed. The voltage range must be correct, and the timings are important too.
Did you check to be sure the RAM is compatible with that motherboard?

No? As far as I know, memory is memory, I saw a CPU compatibility list but not a memory one. Only spec I saw for memory was speed and type, this is the right speed and type. Let me go poke at GSkill's website and see. But it should work.

Edit: According to GSkill Memory Configuratior the Ripjaws X F3-1600C9D-16GXM should be compatible. Though I did notice in BIOS it's auto detected as 1333, not 1600. Not sure how to change the timings to the recommended 9-9-9-24.

 

[TVeblen]

You will look for the QVL List on GSkills (Qualified Vendors List). Look at the RAM that has been tested and passed in that list and note the voltages and timings. The timings do not all have to be 9-9-9-24. If you see some that are 10-10-10-27 then those would work too. You should not have to change the timings on your sticks.

It's a long shot anyway. Just trying to think of all the possible causes - looking for something to stand out.

 

[BelGarion]

You will look for the QVL List on GSkills (Qualified Vendors List). Look at the RAM that has been tested and passed in that list and note the voltages and timings. The timings do not all have to be 9-9-9-24. If you see some that are 10-10-10-27 then those would work too. You should not have to change the timings on your sticks.

It's a long shot anyway. Just trying to think of all the possible causes - looking for something to stand out.

Well, the bios detected my 1600s as 1333s even though they are on the vendor list (see above link from GSkill's website) for this board. The timings were 11s, not 10s or 9s. Testing now to see if it has gotten worse/better.

 

[TVeblen]

1333 is the default speed for the motherboard.

If the RAM is XMP (Intel motherboard) or AMP (AMD motherboard) you can change the setting in the BIOS (UEFI) to run the RAM in XMP Mode. On Asus boards the setting is under AI Overclock Tuner. Then the RAM will run at the module's rated speed, not the motherboard's default speed.

If the RAM is not XMP rated, then you will need to change the RAM speed in the BIOS.

 

[BelGarion]

1333 is the default speed for the motherboard.

If the RAM is XMP (Intel motherboard) or AMP (AMD motherboard) you can change the setting in the BIOS (UEFI) to run the RAM in XMP Mode. On Asus boards the setting is under AI Overclock Tuner. Then the RAM will run at the module's rated speed, not the motherboard's default speed.

If the RAM is not XMP rated, then you will need to change the RAM speed in the BIOS.

Only thing I see regarding XMP is "XMP Information" which brings up a page of numbers that can't be changed. Well, two BSOD in two days. I run though the BSOD Posting HowTo on this site and am posting that information here instead of duplicating this post. Only thing is the .dmp files are old, like from the 17th and 20th, even though System -> Advanced Settings -> Startup & Recovery shows it is set to create a dump file on system crashes. The rest in the zip should be relevant though.

 

[TVeblen]

No, you should post this over in the BSOD Help & Support Forum. The bluescreen guys hang out there! There's just gear heads over here!

 

[BelGarion]

No, you should post this over in the BSOD Help & Support Forum. The bluescreen guys hang out there! There's just gear heads over here!

Well, if you think so, I don't want to get any flack for duplicate posts. Hopefully someone can figure out what's wrong with my system. I'm at the point where I wish I had a lot of $$$ then I'd ship this thing to a friend who's a guru, makes me look like a tinkerer or dabbler, and let him figure it out for me for reasonable compensation. But, sadly, I don't have much so I gotta figure this weird problem out. Of all the upgrades over the years, motherboard + memory, or CPU, or HDDs, etc never had one be such a pain. But this is my first MSI motherboard, at this point, probably be my last, I've used Asus almost exclusively until now. I've RMAed this motherboard once, the shroud around the front USB 3 pin header fell off during install of the first board. >.<

 

[TVeblen]

It's not a double post when you are asking a different question. I have not learned to read dump files and the folks that have are over in that other part of the forum.

Every motherboard manufacturer has a bad board from time to time. It happens. It always bites when it happens to you.

I'll look for your post over in BSOD to see if it yields any clues as to what is going on.

 

[BelGarion]

It's not a double post when you are asking a different question. I have not learned to read dump files and the folks that have are over in that other part of the forum.

Every motherboard manufacturer has a bad board from time to time. It happens. It always bites when it happens to you.

I'll look for your post over in BSOD to see if it yields any clues as to what is going on.

The crazy thing is this board has such good ratings. And it might turn out to be two RMAs for the same board.

 

[TVeblen]

Two RMA's?
Was not aware that you RMA'd this board previously.

Getting a bad board is one thing, getting a second replacement board that is also bad is highly unusual.
When that happens it strongly suggests it is NOT the motherboard causing the problem, but one of the other components in the system.

Here's the thing: if you are 'lucky' the bad component will show up right away - the system won't start, you get error beeps or codes, the system shuts down, etc. This is a classic hardware failure and you can trace that down logically simply by starting out with the minimal components (motherboard, processor, and power supply) and then start adding components to the system one at a time and testing until you get to the bad component. Actually pretty easy.

But when you are dealing with a system that runs and only has problems in the operating system then things get real complicated. Your problem could be caused by a single bad component, a certain combination of components, a driver for one of the components, or a conflict between drivers for different components. There are also file system errors, incorrect settings, and running program conflicts too.

The clues you have given us so far indicate an error caused by something in the storage area (hard drives) or the memory area (RAM, memory controller, BIOS). But a new clue can change all of that.

Your post in BSOD may result in some clues. Give them time and someone will analyze your files. Sometimes the crash dumps will yield a smoking gun, sometimes just more confusion.
One thing you could try is to generate a recent crash dump. They might be able to give you some instructions on clearing out the old dumps and setting the system to create new dumps over there (since it sounds like you are not getting any new dumps now). Just ask that question directly.

On the hardware side you could try some of your own tests. Try disabling components one at a time - in Device Manager, in BIOS) and running the system without them for a while. See if it reveals an obvious cause.
Since you reinstalled since the new parts (clean install I assume?), I would review which drivers you installed for your components. Be sure you are using the most current drivers offered by the motherboard manufacturer, downloaded from their website for your motherboard - not the (usually) older drivers supplied on the DVD.
Try uninstalling and reinstalling each driver. Given your errors you might start with the chipset driver and SATA drivers. Do not install any of the utilities and add-ons offered by the MB manufacturer, just drivers. Hold off on installing Intel Rapid Storage Technology driver for a while. Many bluescreens have be attributed to installing the wrong version of IRST.

It's got to be something. Stay with it.

 

[BelGarion]

Well, got it nailed down. As for *why* I have _no_ idea. Last Monday, after two nasty freezes with in 30 min I dug out a 640gb hdd out of the closet (it was my recording drive before I got the first 2tb drive) and made *that* the primary and left the new 2tb drive as a second data drive. Installed Win 7 onto the 640gb and a week later, no issues. I'm guessing it's the fact the new 2tb is a NAS drive is all I can do to explain it, as a drive that tests good on a surface scan, the SMART status is green, etc and for all intent purposes good shouldn't be thrashing itself to bits at random moments even after I set the page file to a fixed 16gb and set it for performance / program files and *not* background tasks, Aero on, etc.

 

[TVeblen]

You've narrowed it down to a bad hard drive, despite any positive test results. If it is under warranty, send it back RMA and get a new one.

The difference between a NAS drive and a normal hard drive is Error Recovery Control in the firmware. This has been known to slow down performance in these drives used as a standalone, but otherwise they can work fine as a system drive.

You could try and disable ERC if you can, see if that helps.

Error recovery control - Wikipedia, the free encyclopedia