x25-m BSOD and lockups

[imported_BradCube]

Hi All,

After searching the internet for ages, I am having trouble finding anyone that is having the same troubles as me with their x25-m 80gb.

I'm currently having lockups around every 15 minutes or so that last for 5-10 secs as well as BSOD's every few hours or so.

I should probably mention that the x-25m wasn't the only thing to go into the system when I upgraded it recently. I also threw in an extra 4gb's of ram, and a PCI SATA card (as I had no ports on the mobo left). The x-25m is running on the motherboards integrated SATA controller on channel 0.

I tried to use HDDErase tonight to wipe everything back to start afresh, but cannot get it to detect the x25-m at all.

My current system configuration looks like so:

MSI P6N-SLI Platinum
E6600 Core 2 Duo 2.4ghz (OCed to 2.8ghz)
8gb (4x2gb) Corsair XMS2 DDR2
NVIDIA 8800GTX
Corsair HX620w
Blu-Ray Burner
DVD Burner
x25-m 80gb
Samsung 640gb
Hitachi Deskstar 1TB
+ two external 1Tb drives

Vista - Home Premium 64-bit SP1

Any ideas?

 

[Forumpanda]

10 panda points says memory problems, possibly bad memory or timings set too tight.
http://www.memtest.org/ .. burn to a cd or boot from a usb drive.. if that test gives you any errors at all then that is where your problem is.

 

[imported_BradCube]

I haven't adjusted timings on them so far (as in, I'm not pushing them beyond their specs)

However, Memtest is a good idea anyway. I'll let it run for the next 20hrs or so and report back. Thanks

Cheers.

 

[lopri]

Or you could just take the two sticks out, and see if the symptom persists. It'd be a pain to open the case but it could be a quicker way to diagnose.

 

[taltamir]

Memory problems produce the issue you are describing, failing HDD produces different issues. if the HDD was the issue it will also trigger a checkdisk scan after every lockup. If it doesn't than that is another sign its not the drive.

 

[imported_BradCube]

Aha ha.

You people are brilliant troubleshooters! Memtest has already found some errors while I left it overnight.

I have a slight problem in that I realized that all 8gb are new to my system. As I said I used to have 4gb but it was made up of 4x1gb sticks. In order to actually fit 8gb of RAM in the system I swaped these 4 sticks with my brothers system which was made up of 2x2gb. Then I added the new 4gb for a total of 8gb.

In short, looks like I'm going to have to test all of the sticks and see if I can find the problematic one/s. I'll post back when I know more

Thanks so much for the help guys. I'm so delighted that the problem is ram and not the x25-m

 

[taltamir]

it could also be because of overheating of the northbridge or the ram sticks caused by running all 4 sticks at once... but if a single stick is erroring by itself then you have a definite culprit.

 

[imported_BradCube]

So, I managed to find one faulty stick of RAM. I removed this from the system and then ran MemTest with the 6gb and it tested fine with no errors.

However, I'm still getting lock-ups and BSODs. Any more ideas?

Should I try running the system with just 4gb of ram for a while and see if the problem persists?

 

[taltamir]

you should reinstall windows, installing windows or ANY windows updates with a faulty stick of ram -!WILL!- corrupt windows files, causing permanent irreversible system instability.
the only (practical) solution is to reinstall windows to fix the corrupt files...

just to verify, with the 6GB of ram, all in together, you get ZERO errors with memtest even after running it ALL NIGHT LONG?

Oh, and you lose quite a lot of speed by running 3 modules in a dual chanel system, it forces it to single chanel mode and halves your ram speed. Keep your ram in pairs!

 

[Forumpanda]

As the guy said, the damage is done, files are corrupt due to the faulty ram.
A reinstall should make it work just fine (I should probably have mentioned this in my first post).

 

[imported_BradCube]

Thanks for this guys. I've been looking for an excuse to try out the Windows 7 RC. Guess, this will give me that opportunity!

I'll re-install today sometime and let you know how I go

Just to confirm, the 6gb tested fine in Memtest after running it for a whole night. As per your suggestion, I may pull out that extra 2gb stick so I can run the memory in dual channel until I get the faulty stick replaced

 

[taltamir]

good idea, also if it is a dual chanel pack (aka, you bought a 2x2GB package) then you MUST send them back together to get it warranty replaced, they will not just replace one unit (they must be MATCHED correctly... make sure you take out the right stick and don't break up the other pair, if you use the same model but not a matched pair you might end up getting errors).

 

[imported_BradCube]

Will do. It will be easy enough for me to identify which stick belongs with the faulty one, as they have different timings. The new pair runs at 5-5-5-18 while the older ones run at 4-4-4-12.

Just getting drivers prepped for the re-install now. Is there anything I should know before going ahead? For drivers that don't have a Windows 7 release, should I just download and install the 64-bit Vista versions?

 

[imported_BradCube]

So, I've just installed Windows 7 RC with 4gb of Memtested ram. I then installed 64-bit NVIDIA Forceware drivers for a 650i chipset (which the P6N-SLI Plat is). It wanted a reboot, so I went ahead. The system then locked up on boot. I restarted the system, and Win 7 kindly offered to system restore with it's start up recovery utility. System restore work and then booted fine. Since then I've had another two crashes with the exact same symptoms as when I was on Vista. The only drivers that I have installed (seeing as the Forceware ones were removed) are the GeForce display drivers for my 8800GTX and anything that windows update automatically installed (from memory drivers for my two flat panel displays). Everything else is running on native Win 7 drivers.

This rules out memory problems I think (although we did at least find that one stick which was probably reeking its own havoc upon the system). I've pulled out the add-in PCI SATA card, on the off chance that it is causing the problems. No crashes so far, but it's only been 10 mins or so.

When the crash happens the HDD LED slows down to about one flash per second (regardless of whether it's trying to launch something or not), then after a few minutes it will remain completely on. By the time it is completely on, the computer has become completely unresponsive and the mouse will not even move.

Any more ideas guys? Could it be that my x-25m is faulty?

 

[IntelUser2000]

To find out if the X25-M is at fault, test it on another system. That's really the only way. I have a feeling that it really is your X25-M drive. Surprisingly I have been seeing lots of errors with these drives on forums. Go RMA it or something. It does sound like similar to when my friend's hard drive started dying.

 

[imported_BradCube]

So after weeks of troubleshooting, I have finally found a solution to my problem (or at least my system has been running for a record 36hrs straight now).

The problem has to do with an incompatibility issue between the x25-m line and some NVIDIA 650i and 680i chipsets. I found the solution on the Intel forums. All I had to do was to disable "command queuing" in Windows for the x25-m (although it has been said that enabling/disabling ACHI could also help).

So how does one disable command queuing? Simply go to device manager > IDE ATA/ATAPI controllers > NVIDIA nForce Serial ATA Controller > Choose the relevant "Port" tab for your disk drive > Uncheck "Enable Command Queuing" > Hit OK and reboot.

It's probably wise to make sure your chipset drivers are up to date before this. Just wanted to thank all that have helped me solve this issue. Hopefully this info will be of use to someone in the future.

On a separate note, my motherboard also seems to be limiting the speed of the drive. With "Command Queuing" enabled HDTach reports average read speeds of around 130MB/sec. With it disabled (and stable), I'm on about 160MB/sec. An improvement, but still a far cry the 220-250MB/sec I was expecting. I threw the drive into my brothers system (motherboard, MSI P35 Neo2-FR) and instantly the same test shot up to 230MB/sec. Would most of you guys expect that what I am seeing on my motherboard is a limitation of the drive controllers processing speed? Thanks once again

 

[Idontcare]

Sata performance on my 680i board was abysmal as well.

 

[imported_BradCube]

Ahh ok. My motherboard is actually on the 650i chipset, but it's good to know that it's not another problem I am able to sort out (short of replacing the motherboard anyway ).

 

[taltamir]

never use system restore AFAIK, it only makes more problems...