What hardware would you replace first?

Status
Not open for further replies.

bockman

Posts: 17   +1
Hi guys,

Taking a tip from Acrobat's stellar post, I am going to try to format my first post much the same. I should point out that some of the information herein is stuff I know directly, some is stuff I have learned through careful study online concerning crashes, and some is reported by a local tech support guy who has spent hours and hours with me trying to diagnose and repair this problem. I am NOT schooled in computer science. I know maybe a little bit more than the average customer or computer user, nothing more.

The Good


  • Motherboard: ASUS A8N-SLI Deluxe
  • 2 VGA EVGA|GF 6600GT 128-P2-N368-TX
  • POWER SUPPLY: ANTEC TP2-550 EPS12V
  • CPU AMD 64 |4200+ ATHLON 64 939P
  • CORSAR D400 512Mx2 TWINX1024-3200XL
  • SANDISK 256M FLDRV SDCZ2-256-A10 R
  • FD 1.44MB|MITSUMI FA404A/404M
  • 2 HD: 160GB|WD 7200 WD1600JD 8MB (SATA)
  • KB&MS LOGITECH|CRDLS INTERNET PRO
  • MNTR SMSNG|20"LCD16msDVI 204T
  • DVD-ROM SONY 16X DDU1615/B2s BK
  • DVD+/-RW 16X SONY DWQ28A SW OEM
  • SOUND BLASTER|AUDIGY 2 ZS PLTM RTL
  • CPU FAN AMD|SILENTBOOST K8 A1838 RT (Replaced under warranty several months ago with a better one)
  • LOGITECH THX Z-5300e 5.1 SPEAKERS
  • CASE: THERMALTAKE|VA3000BWA RT

This computer is used largely for CAD as well as photo manipulation and post-processing. I also have played HalfLife 2 on it frequently with no problems since I bought it in Oct. of 2005. Like Acrobat, everything is at its stock speed. Nothing has ever been overclocked. Windows XP Professional - Service pack 2.

The Bad

Several weeks ago I began enjoying what appeared to be random BSOD's...which never occured at startup, but seemed to occur after an extended period of work time, and/or having multiple applications going at the same time. When they first started I had automatic restart enabled, so I was only briefly able to read the blue screen info, however it was enough to catch various phrases such as 'BAD_POOL_HEADER', 'IRQL_NOT_LESS_OR_EQUAL', and such suggestions as checking and/or updating BIOS. According to some professionals, these errors were occuring either because of driver conflicts or a physical failure of some part of the hardware. I ran every basic diagnostic check one can do: virus scans, windows updates, device manager conflicts, malware/spyware scans, everything.

I ran Memtest 86 for 6 passes, several times, and it found no errors. While I was researching diagnosis and repair options I suffered an overnight BSOD which resulted in an inability to boot up Windows at all; I would get through the DOS startup screens, the black Windows startup would launch with the progress bar, it would get through about two iterations and everything would shut down, even the signal to the monitor would cut off. Yikes.

Called in my tech friend, we were able to secure my critical information via booting up with Linux (Knoppix 5.1 on a CD) and then moving stuff over to an external hard drive. With my drawings and other files safe, we first:

Checked all connections

Everything was tight and fine. At this point it became known that the hard drives were plugged in for RAID array rather than SATA. I didn't specify a RAID'd sytem at purchase but I guess it was an oversight-- not sure if this has anything to do with the crashes ultimately occuring, just stating what was found. The system worked great for over a year in this configuration. They were plugged into the black connections from the red and we moved on.

Ran CHKDSK

Multiple errors were found and repaired on both drives.

PSU was observed to be rock solid and putting out voltages well within acceptable parameters.

He tried various options at repairing Windows XP via the restoration DVD which came with the system when it shipped, none of which worked.

So, a total reinstall was started. CHKDSK ran for a LONG time, over several hours, before re-installation occurred. During installation from the DVD a pop up requested:

'Please insert Floppy Disk 'Nvidia Network Bus Installation Disk #1'

which of course I didn't have. My computer manufacturer's Tech Support said that after installation I should go to the motherboard's website and download the latest drivers. He gave me the motherboard's model (A8N-SLI DELUXE, socket 939). I downloaded and installed with no problems (if you can call successfully navigating Asus's Byzantine download section of their website 'no problem').

System seemed to be stable, I rebooted a dozen times or so with no problems, so we updated every Windows update that was suggested, reinstalled my drafting program, and I pressed on.

The Ugly

While the system is certainly 'more stable', I am still getting the odd BSOD, with messages like, but not limited to:

****STOP: 0X0000008E (0X0000005, 0X8F824861, 0XF09C9B7C, 0X00000000)

***WIN32K.SYS- ADDRESS BF824861 BASE AT BF800000 DATESTAMP 43446A58

I have yet to install any of my other software, such as Photoshop and Sketchup, for fear of causing another horrendous crash. I also set CHKDSK to run a full scan/repair at each morning's reboot.

From what I've read, these wonky symptoms could arguably be the fault of one of my pieces of hardware going bad. Bearing in mind that Memtest did not find any errors at all, I'm wondering just what the heck to do next. I just want my old stable (and fun to use) system back. I'm not a CS major or professional, the amount of time I've spent on this is just staggering really-- I'm sincerely hoping someone can make some suggestions as to what the next steps should be to resolve my problems. Thanks a lot for everyone's time who reads and responds, it's appreciated.

Dave
 
Great post.

Go HERE and follow the instructions. See if that helps at all.

If it doesn`t attach 5 or 6 of your latest minidumps and we`ll see what we can come up with.

It might also be a good idea to physically disconnect anything not required to run your system and see if it becomes stable. if it does, then adding one thing at a time and testing, may reveal the culprit, if it`s hardware related.

Regards Howard :)
 
Thanks Howard,

I've done just about everything suggested in that thread, but I will go back and be scrupulous to do them all. When you write,

It might also be a good idea to physically disconnect anything not required to run your system and see if it becomes stable.

May I ask what items I should be disconnecting? Thanks,

Dave
 
Try running your system with just the absolute basics.

1 stick of ram/cpu/hard drive/graphics card/mouse/keyboard/monitor.

Regards Howard :)
 
Your minidumps crash with 2XWin32k.sys/1Xsr.sys/1Xmemory corruption. 3 have a bugcheck of 8E and 1 has a bug check of 1A.

0x0000001A: MEMORY_MANAGEMENT

This memory management error is usually hardware related. Also, check the Windows system requirements including the amount of RAM and disk space required to load the operating system.

0x0000008E: KERNEL_MODE_EXCEPTION_NOT_HANDLED

A kernel mode program generated an exception which the error handler didn’t catch. These are nearly always hardware compatibility issues (which sometimes means a driver issue or a need for a BIOS upgrade).

Have you tried doing a Bios update?

Regards Howard :)
 
Thanks Howard,

Forgive me, is a BIOS update the same as updating the motherboard drivers?

I reset the pagefile as per that link you mentioned earlier. Also, when you write,

This memory management error is usually hardware related.

do you mean, hardware physically failing or hardware with improper drivers?

You're rapidly earning a case of any beer you desire.

Dave
 
By hardware related, I mean possible hardware incompatibilities or possible hardware failure.

A bios update is a bit like updating a driver, but involves flashing the bios chip on the mobo. Full instructions, along with the bios update itself, if available, will be on the mobo manufacturers website and also possibly in your mobo manual. When flashing the bios, instructions must be followed exactly.

If after updating the bios, you still have problems, we`ll have to look elsewhere for possible causes.

Regards Howard :)
 
Thanks Howard, I will update today sometime.

This morning after bootup I received this BSOD:

'an attempt was made to write to read-only memory'
 
Your minidump crashes at win32k.sys and has a bugcheck of BE.

0x000000BE: ATTEMPTED_WRITE_TO_READONLY_MEMORY

A driver attempted to write to read-only memory. Commonly occurs after installing a faulty device driver, system service, or firmware. If a driver file is named in the error message, try to correct the problem by disabling, removing, or rolling back the driver.

The above error is not caused by the win32k.sys driver itself, but by some other driver/firmware etc.

Let me know how you get on with the bios update.

Regards Howard :)
 
The above error is not caused by the win32k.sys driver itself, but by some other driver/firmware etc.

This is interesting. I was loading a myspace webpage which had 'imbedded music' on it when the crash happened.

Howard, I just googled 'win32k.sys' and it came up with a 2004 thread of yours (you newbie!) which essentially states that the most likely faults are bad ram or vid (card?).

If, after updating BIOS I still get these I'm pullin the ram and heading to Newegg.
 
Successfully updated BIOS today, will keep you guys posted on any BSOD's. Thanks again for the help everyone.
 
Decided to let the system run all night to check for stability, was welcomed with an 'IRQL_NOT_LESS_OR_EQUAL' BSOD when I checked this morning.

IN preparation for such an event, I asked my computer's manufacturer for the RAM specifics. They said I currently run:

Corsair XMS Series 184 Pin DDR, PC 3200. Ah, wait, I noted that (sort of) in my original post:

CORSAR D400 512Mx2 TWINX1024-3200XL

Anyway, I'm going to replace them and see if that solves things. Howard or anyone, could I trouble you to point out the appropriate model over at Newegg?

http://www.newegg.com/Product/ProductList.asp?Submit=BROWSE&Manufactory=1666

Thanks,

Dave
 
Another cause for instability can be a failing power supply. Could you install a monitoring software such as Asus Probe & see how the different rails (+12v, +5v, +3.3v, etc.) react while doing some heavy work ?
 
Didou, thanks for the input. My computer has a voltage monitoring interface which indicates the power is very stable. Granted I am not able to access it while multiple windows applications are open, but I will check to see if it logs voltages during normal operation.

I'll also note that crashes occur at points which would seem to me to be 'low demand' situations, such as what happened overnight.

addendum: Silly me, I installed Speedfan some time ago; it has voltage monitoring.

12 volt rail is always within .20 of volt. Is that acceptable? In other words, it will fluctuate down to 11.80 V.
 
Your voltage fluctuations are within acceptable limits.

It is possible, even though you have a good psu, that it`s failing. As Didou said, a failing psu would cause all manner of bsod`s

Post any further minidumps.

Regards Howard :)
 
Thanks Howard,

Two of the most recent minidumps are attached:

I'm thinking about running it on one stick of RAM for awhile; is it really as straightforward as simply pulling one stick out and powering up?
 
1 minidump crashes at avg7rsxp.sys, AVG antivirus.

1 minidump crashes at NTFS, sys, this is the Windows file system driver.

Check your hard drive with the Powermax ISO programme. Burn it to cd as an image, then boot from the cd.

is it really as straightforward as simply pulling one stick out and powering up?

Yes it is.

Regards Howard :)
 
Well that was interesting. Powermax did not detect either of my hard drives. I think my HD's are Samsung. Will Powermax work on non-Maxtor/Seagate HD's?
 
Hi gang,

Believe it or not, I'm still d*cking around with this desktop. I'd like to replace the hard drives and would appreciate some input as to what are acceptable replacements. It's currently using 2 of the following:

Western Digital WD Caviar SE
Serial ATA Hard Drive
Drive Parameters: LBA 312581808 160 GB
WD PIN: WD1600JD-00HB00

(read right off the drive itself)

Newegg doesn't have the exact model I don't think, I'm wondering how much leeway I have in getting replacements and not creating more problems getting them to work with the existing system (I'm also replacing the MOBO).

Any suggestions greatly appreciated, especially if there's a good deal on them! :)

Dave
 
Shouldn't Be a Problem.....

bockman said:
Hi gang,
Newegg doesn't have the exact model I don't think, I'm wondering how much leeway I have in getting replacements and not creating more problems getting them to work with the existing system (I'm also replacing the MOBO).

Any suggestions greatly appreciated, especially if there's a good deal on them! :)

Dave

More than likely the drives in your computer are SATA 150s (SATA1). Since all the new SATA drives would be SATA 2, you won't find the exact model number unless you by the new drives as replacement parts. Sometimes manufacturers get parts with proprietary numbers for that very reason, to herd you toward buying their direct replacement parts. SATA 2 drives are backwards compatible with SATA 1 installations. Seagates's SATA 2 drives even come with a jumper installed that locks them into SATA 1 operation. WD SATA drives with "SE" prefixes (drives under 400 GB)don't support NCQ (native command queuing) and won't operate in some RAID arrays. Should you think you might consider RAID in the future, you would do well to purchase WD's "RE" prefix drives (Raid Edition), they're about five bucks more, and all sizes support NCQ. Windows drivers for HDDs are generic and you could probably even change the drive brand without running into trouble.
 
Status
Not open for further replies.
Back