Can anyone identify the main issue here?

[dataDude]

I'm faced right now with a humdinger of an issue. First I'll get my system specs out in the open:

Athlon 64 3000+ 754
Chaintech VFN3-250 mobo (BIOS updated)
1.5 GB RAM
111 GB HDD
Radeon 9700 video card
Windows XP streamlined w/ SP2. Current updates are installed.

I really have no idea where the exact origins of this problem are, but I have clues to it. It seemed to start at the time when I started playing Unreal Tournament GYOT edition. My system would either only let me play on 32-bit with the generic system drivers, or with Direct3D software on 16-bit. Trying to use Direct3D at 32-bit, the system would freeze after a few seconds but the game would still be going (but not respond to mouse/keyboard). I hadn't played UT for a long time and the system never gave me issues before that, though I've had the game installed on here for quite a while, along with Doom 3. I also tried running Doom 3 and after about 5 minutes of play the system would freeze. I ran memtest86 and found no problems. I ran chkdsk and it found no problems. I figured it was the video drivers, so I reinstalled those. Nope, problem continued. I've also been doing some DVD->DIVX conversions with FlaskMPEG while I go to bed. I'll wake up, check the progress and find that the system has been rebooted, the conversions never completed (nobody but me touches this system). I went into system settings and unchecked the automatically reboot at system error. The next time it tried to reboot it blue-screened. Sometimes it showed IRQL_NOT_LESS_THAN_OR_EQUAL and sometimes it only showed that a file was to blame (something like "ati2d-something" created an error). I uninstalled and reinstalled the video driver, still had problems.

After the system froze so many times I decided to reinstall the OS. Not long after that the system kept freezing again. This time after having done so a few times I ran chkdsk and it showed that there was a problem with the file system. Not good. I also then finally remembered to look at the event viewer. It showed a system error log with event ID 1003 with a similar description of: Error code 00000024, parameter1 001902fe, parameter2 aec5b640, parameter3 aec5b33c, parameter4 804e7926. Also, the error report was seem as corrupted.

That's the latest error I received. In an effort to correct the file system error, I reinstalled Windows last night. After installing the video driver (without the Catalyst Control Panel or WDM), it stopped any visible activity when the OS was showing a blue background (the HDD LED was blinking a little though). The mouse still functioned, but nothing else seemed to be so. The video driver installation did say that the system would be performing a couple of tests for a minute upon reboot and would cause the system to slow down or nonadvance during that time. I waited for about 10 minutes and nothing advanced. I manually rebooted and the system seemed just fine. That is, until I went to install ZoneAlarm, where it wouldn't advance during the process at 11% and became nonresponsive. It took me several tries to finally have it install properly. Since that time last night the computer rebooted when I was trying to do DVD conversion and then relogging into the system caused the computer to get stuck at "Loaded personal settings..." I manually rebooted again and this time everything worked flawlessly.

After that long story, can anyone clue me in to a possible solution or even to the main cause? Thanks.

[J1978]

HI it looks like every time you do a cpu intensive task it freeze's that could hint to a heat problem check your temps.
you should check you Hdd as well.
go to the manufacturer's website and download and run there diagnosis tool.

Hope that helps

good luck

[jayb1234]

2 things to check.

Dust buildup in the case, blow it out with canned air. Loose CPU heatsink, figure out why it's lose and repair or reseat as appropriate.

[dataDude]

I went into the BIOS to check voltages. The readings:

+3.3V power level showed 3.26V
+5.0V power level showed 5.13V
+12V power level showed 12.22V
-12V power level showed switched back and forth between -11.95 and -12.03V

Do any of those seem a bit far from where it should be? The 12.22V seemed kinda iffy to me, but I'm not sure. The CPU temp remains stable at 47 and 48 degrees. The fan switched between 3443 and 3515 RPM.

Inside looks pretty clean. It wasn't very long ago (a few months) when I had upgraded from Athlon XP to 64. I made sure to clean any dust out of there when I performed the board replacement. Other than the voltage levels, nothing seems suspicious.

[jayb1234]

Might be a RAM problem. Try running it with each stick 1 at a time and see if you get the same problem with one of them.

Also i have seen a few posts where the new version of Zonealarm is causing problems, Try an older version and that might solve the problem.

[GaryRouth]

I like the suggestion that was made to run the manufacturer's disk diagnostics [Western Digital = DataLifeGuard, Maxtor = PowerMax, Seagate = SeaTools, etc.]. Chkdsk is a good tool, but the manufacturer's diagnostics are usually a bit more thorough.

Have you run full-system malware scans from Safe Mode yet (to rule out the possibility of malware tripping things up)? If not, try that - just to be sure. And to run a double-check on that, you can visit an online scanner like TrendMicro's HouseCall or Panda's. I particularly like HouseCall's because it scans for vulnerablilities and spyware along with it's antivirus scan (if you choose it's "Complete Scan" feature). HouseCall is over at http://housecall.trendmicro.com You will likely have to allow it's ActiveX control to run (it's specific to the scan, and OK).

The video driver area might still be something to continue investigating. If you're overclocking, try running at stock timings and voltages (no matter what you're overclocking - buses, cpus, gpus, etc.). And check in on your Bios settings to make sure you're not overclocking by accident. You can check your temps again while in there, to make sure that your Windows-based hardware monitor isn't reading them improperly.

Some DVD's have fairly aggressive copy-protection schemes - I imagine you may have heard of some of the uproar over the scheme that Sony put on some disks (though I think the worst infractions were on it's CDs so far). Some copy-protection antics were installing stealth programs that hid files on the PC, and were allegedly hiding files not only from Sony, but also hiding the PC users files as well (when the program considered them to endanger it's copyright). Worse yet, the removal of such copy-protection schemes has been leaving behind gaping security holes. Microsoft plans to include it's own removal tool in it's next "Malicious Software Removal Tool" (that it releases every month) - targeted to remove Sony's latest bad idea.

Best of luck
. . . Gary

[P.S. -- if your hardware and Bios settings continue to test OK, and you want to be completely sure nothing uninvited is on your hard drive (and if you have good recent backups, and your product key handy) - you can use the manufacturer's disk diagnostics to zero-write the drive & just start over from scratch. Not even a rootkit can survive a zero-write.

Another option for a quick-fix would be to try a different power supply]

[dataDude]

Thanks Jay and Gary, I'll check into those and post results.

Some additional recent events: I got the BSOD a couple of times. Last night I got one that said "PAGE_FAULT_IN_NONPAGED_AREA" and then pointed to netbt.sys to blame. Then a few minutes ago I got the other BSOD that pointed blame to tcpip.sys. Is it possible there's something wrong with the NIC?

[dataDude]

I ran a chkdsk and it found a problem with the file system. I then went to a web page that reminded me about the Windows utility SFC. I'm not sure if it's supposed to alert me about system issues it finds or just automatically fix them. I guess only time will tell. I also went to the Samsung web site (this hard drive's manufacturer) and downloaded the drive diagnostic that's run through a bootable floppy. It found no problems. I hope some of this stuff will ameliorate the issue. The system's getting annoying. Any more ideas?

In response to jay's reply, testing the memory one stick at a time was one of the first things I did to troubleshoot. Sorry I hadn't mentioned it before.

[dataDude]

My PC just BSOD'ed again and said there was a problem with win2k.sys. In the past 2 days alone I've had 3 BSOD's all of them blaming a different file. My antivirus shows nothing's wrong, so does my spyware software. There's no indication of anything wrong with anything, except the BSOD's. I'm also not overclocking anything on here. It's as if my friend was dying and all medical tests show perfectly healthy.

[GaryRouth]

Since the files involved keep changing, but several have been system files, the usual two suspects are malware infection or hardware failure. Your local malware scans have come up clean - did you try a double-check at HouseCall or Panda yet? Can't hurt.

If it's possible that a past infection is at fault (and that the infection & it's removal have left damaged files behind) - the idea to try the System File Checker (or Windows File Protection, as it is often referred to in Windows XP) is worth a try. With your slip-streamed Windows XP & SP2 disk in the cd-drive, go to the Start/Run box & type:

sfc /scannow

That will run for a bit - if your CD is in the drive, it should find the files it needs there (you may have to point to it at first).

Another way to refresh many of your system files is to run a repair installation (which I mention again in the next paragraph = along with borrowing a different power supply).
___________________

It's likely that the more probable cause is faulty hardware: something is leading to data corruption. Since you've tested the memory and the hard drive already - see if you can round up a high-quality power supply to test with. With the borrowed supply installed, try a repair reinstallation of Windows XP with the slip-streamed CD. Be sure to be unplugged from networks while doing this, and don't reconnect to the Internet until your firewall and malware protection are running again. Then head to Windows Update for all the Security Patches you'll need.

___________________

Have you considered the "nuke & pave" idea? [zero write]. Should the problem recur after such a drasticly clean reinstall - it would pretty much point to an as-yet-uncovered hardware issue.

Let us know if you catch the culprit
. . . Gary

[dataDude]

I tried the HouseCall checker and selected all tests - nothing bad found. The computer in question is one of a few that I have and it's the one with a history of having the least number of virus/spyware infections. I'm keeping in mind though that this PC was only just reformatted a couple weeks or so ago and it almost seems like it was from day one this thing was giving me issues.

Yeah, I know how to use SFC, I just rarely ever need to use it, so naturally I tend to forget about it in the line of normal troubleshooting measures.

I'm thinking that maybe I should run a repair install if this thing gives me crap again. Zero filling the drive is an option, but a last resort. I finally got all the stuff that belongs on this PC onto it. I may have to redo the process.

In the beginning (before the reformat), the files that were giving me trouble were video driver files. I've taken measures to avoid having to install the problem files while still keeping the video accelerator running as I want it to. Since the reformat, it's been the netbt.sys, tcpip.sys, and win2k.sys. As previously mentioned, I wonder if it's something to do with the NIC?

[dataDude]

Microsoft is helping at least with one thing on debugging systems. I found they have a program that will actually view memory dumps saved to the hard drive (Dump Check Utility found at http://www.microsoft.com/whdc/devtoo...g/default.mspx - requires symbols file). The following shows what the problem was (a bit cryptic):

*******************************************************************************
* *
* Bugcheck Analysis *
* *
*******************************************************************************

PAGE_FAULT_IN_NONPAGED_AREA (50)
Invalid system memory was referenced. This cannot be protected by try-except,
it must be protected by a Probe. Typically the address is just plain bad or it
is pointing at freed memory.
Arguments:
Arg1: 8b3ea62c, memory referenced.
Arg2: 00000001, value 0 = read operation, 1 = write operation.
Arg3: b1fc89c6, If non-zero, the instruction address which referenced the bad memory
address.
Arg4: 00000000, (reserved)

Debugging Details:
------------------


Could not read faulting driver name

WRITE_ADDRESS: 8b3ea62c

FAULTING_IP:
netbt!SendCompletion+29e
b1fc89c6 ff09 dec dword ptr [ecx]

MM_INTERNAL_CODE: 0

CUSTOMER_CRASH_COUNT: 2

DEFAULT_BUCKET_ID: DRIVER_FAULT

BUGCHECK_STR: 0x50

LAST_CONTROL_TRANSFER: from b1fe3dd2 to b1fc89c6

STACK_TEXT:
af1a4bf4 b1fe3dd2 89fd02f8 80506836 80506836 netbt!SendCompletion+0x29e
af1a4c34 804eddf9 89fd02f8 8883d4f8 806d02d0 netbt!NTSendDatagram+0x232
af1a4c58 805749c9 89fd02f8 8883d4f8 89feedc8 nt!FsRtlAcquireFileForModWriteEx+0xe9
af1a4d00 8056d326 00000990 00000000 00000000 nt!MmAccessFault+0x1d39
af1a4d34 8053c808 00000990 00000000 00000000 nt!MmWriteTriageInformation+0x58
af1a4d64 7c90eb94 badb0d00 0190f684 00000000 nt!MiDeleteSystemPagableVm+0xa44
WARNING: Frame IP not in any known module. Following frames may be wrong.
af1a4d68 badb0d00 0190f684 00000000 00000000 0x7c90eb94
af1a4d6c 0190f684 00000000 00000000 00000000 0xbadb0d00
af1a4d70 00000000 00000000 00000000 00000000 0x190f684


FOLLOWUP_IP:
netbt!SendCompletion+29e
b1fc89c6 ff09 dec dword ptr [ecx]

SYMBOL_STACK_INDEX: 0

FOLLOWUP_NAME: MachineOwner

SYMBOL_NAME: netbt!SendCompletion+29e

MODULE_NAME: netbt

IMAGE_NAME: netbt.sys

DEBUG_FLR_IMAGE_TIMESTAMP: 41107ecc

STACK_COMMAND: kb

FAILURE_BUCKET_ID: 0x50_W_netbt!SendCompletion+29e

BUCKET_ID: 0x50_W_netbt!SendCompletion+29e

Followup: MachineOwner



One thing I noticed was the "DEFAULT_BUCKET_ID: DRIVER_FAULT". Well, the good thing is that since a couple nights ago there has been no prob. The next minidump file shows:

KERNEL_MODE_EXCEPTION_NOT_HANDLED_M (1000008e)
This is a very common bugcheck. Usually the exception address pinpoints
the driver/function that caused the problem. Always note this address
as well as the link date of the driver/image that contains this address.
Some common problems are exception code 0x80000003. This means a hard
coded breakpoint or assertion was hit, but this system was booted
/NODEBUG. This is not supposed to happen as developers should never have
hardcoded breakpoints in retail code, but ...
If this happens, make sure a debugger gets connected, and the
system is booted /DEBUG. This will let us see why this breakpoint is
happening.
Arguments:
Arg1: c0000005, The exception code that was not handled
Arg2: b1ff655c, The address that the exception occurred at
Arg3: af014370, Trap Frame
Arg4: 00000000

Debugging Details:
------------------


EXCEPTION_CODE: (NTSTATUS) 0xc0000005 - The instruction at "0x%08lx" referenced memory at "0x%08lx". The memory could not be "%s".

FAULTING_IP:
tcpip!IsBCastOnNTE+0
b1ff655c 3b9330010000 cmp edx,[ebx+0x130]

TRAP_FRAME: af014370 -- (.trap ffffffffaf014370)
ErrCode = 00000000
eax=8889963c ebx=00faf008 ecx=00000280 edx=00000001 esi=00000200 edi=af0148e4
eip=b1ff655c esp=af0143e4 ebp=af0144a4 iopl=0 nv up ei pl zr na po nc
cs=0008 ss=0010 ds=0023 es=0023 fs=0030 gs=0000 efl=00010246
tcpip!IsBCastOnNTE:
b1ff655c 3b9330010000 cmp edx,[ebx+0x130] ds:0023:00faf138=????????
Resetting default scope

CUSTOMER_CRASH_COUNT: 1

DEFAULT_BUCKET_ID: DRIVER_FAULT

BUGCHECK_STR: 0x8E

LAST_CONTROL_TRANSFER: from 8889963c to b1ff655c

STACK_TEXT:
af0143e0 8889963c 89faf938 af014628 80505aad tcpip!IsBCastOnNTE
WARNING: Frame IP not in any known module. Following frames may be wrong.
af0144a4 b1ff6e13 00faf008 8889963c 88c56d90 0x8889963c
af014600 b1ff6b95 00000000 00000038 baef7d48 tcpip!InvalidateRCEOnIF+0x21
af014880 b1ff6a15 af014818 000006d0 00000000 tcpip!InvalidateRCE+0xaf
af014918 b1ff603c 89b15aa8 89b15b18 89b15b18 tcpip!GetLocalNTE+0x123
af014934 b1ff5ff4 89b15aa8 89b15b18 af0149b0 tcpip!GetFWBuffer+0x1a
af01496c b1f54841 89bbec70 89b15aa8 b1f5454f tcpip!AllocateCopyBuffers+0xd6
af014c34 804eddf9 89bbec70 89b15aa8 806d02d0 vsdatant+0x3d841
af014c58 805749c9 89bbec70 89b15aa8 88e06500 nt!FsRtlAcquireFileForModWriteEx+0xe9
af014d00 8056d326 000004b8 00000000 00000000 nt!MmAccessFault+0x1d39
af014d34 8053c808 000004b8 00000000 00000000 nt!MmWriteTriageInformation+0x58
af014d64 7c90eb94 badb0d00 0161f510 00000000 nt!MiDeleteSystemPagableVm+0xa44
af014d68 badb0d00 0161f510 00000000 00000000 0x7c90eb94
af014d6c 0161f510 00000000 00000000 00000000 0xbadb0d00
af014d70 00000000 00000000 00000000 00000000 0x161f510


FOLLOWUP_IP:
vsdatant+3d841
b1f54841 ?? ???

SYMBOL_STACK_INDEX: 7

FOLLOWUP_NAME: MachineOwner

SYMBOL_NAME: vsdatant+3d841

MODULE_NAME: vsdatant

IMAGE_NAME: vsdatant.sys

DEBUG_FLR_IMAGE_TIMESTAMP: 4313a6c6

STACK_COMMAND: .trap ffffffffaf014370 ; kb

FAILURE_BUCKET_ID: 0x8E_vsdatant+3d841

BUCKET_ID: 0x8E_vsdatant+3d841

Followup: MachineOwner



That shows a driver fault as well. Maybe I won't have to reformat, but merely repair next time this happens. By the way, does anybody know if there are any known problems with the Windows Backup utility? I have a saved image of that that I use in case I have to reformat so that I can bring all basic settings and files that I will always want on here. I have a gut feeling that maybe Microsoft hasn't perfected such a utility. Dunno. As for now, I can only wait to see if this system messes up again.

[dataDude]

Here's something interesting. I got the blue screen again (this time it didn't point to anything to blame, almost like it's teasing me lol) and performed a repair. When everything was finished, I rebooted into Safe Mode and then ran chkdsk and no problem showed up whatsoever. Then I rebooted normally and ran chkdsk but I didn't get the file system error this time, but something much worse. The checker hit so many errors that it couldn't even finish. So I ran the preboot checker and it didn't tell me anything was wrong. That tells me that it's a definite software issue, but I have no idea how to repair it, considering it was a brand new install. Perhaps a bad install? I might have to risk doing the whole darn thing over again. But this time I'll take your advice, Gary, and do a zero fill first and then reformat, repartition, yadda yadda.

[glc]

Run manufacturer's diags on the hard drive first - and run memtest86+ on your ram while you are at it.

[dataDude]

I already have and they both showed up negative for errors.

I had the exact same problem & found a solution. When I built the system about 4 years ago with an XP 2000+, the system would lock up and get random errors of all kinds. RMA'd the mainboard & the same thing happened...really strange for a new system. I never overclocked this system, but having done oc before, realized that more voltage = stability, at the cost of heat.

So, the cpu was turned up from 1.65v too 1.85v, and ddr up from 2.55 to 2.85. That completely solved the problem for almost 4 years, until a month ago when the buildings power transformer blew & it took out my cpu, psu, & mainboard. So, with a new cpu, psu, and mainboard, guess what happened....locked up just like the first system did; this with voltage at cpu default. So, I turned up the voltages as above and the system would either lockup, blue screen at login, or just not boot at all; and I noticed cpu heat was over 50c.

Solution was to set ddr voltage from 2.55 up to 2.85 as before, and get this...turn DOWN cpu voltage from 1.65 to 1.575. Totally stable at this point.

My advise to you is to do the same...turn up ddr voltage about .2 volts, and play with your cpu voltage; first go up .25v at a time to a max of 1 v up. Then if it's getting worse as you go up, and it might due to the extra heat, go DOWN in .25v steps below cpu default setting until it stablizes. My guess is ddr up, cpu down will work. These AMD chip run hot already, and turning down the voltage will help it run cooler. Remember, high heat always causes lockups & failures.

One thing you should do first, is remove the cpu fan, clean the cpu & heatsink, apply some silver thermal paste; put a good thick even layer to ensure full contact and crank down the heatsink properly. You can get a syringe of silver thermal paste at any Mac repair shop for about $5; all G5's use it per factory spec.

I'm almost totally convinced this will be the solution for you...good luck & let us know how it works.


Athlon 2400+
Abit KR7A-R
512 MB RAM
160 GB HDD
MSI GeForce 5700
Windows XP streamlined w/ SP2. Current updates are installed.

[glc]

I would really hesitate recommending voltage changes from default unless you know exactly what you are doing and why. That's a good way to damage components. Recommending a .25 volt increase is way too much in one step even if you are trying to overclock, and a whole volt is just about suicide. Baby steps, please.

The Athlon XP's may run hot, but the Athlon 64 is a different animal. Some brands of ram respond well to voltage increases, others do not.

If you have a retail processor, removing the factory applied thermal solution and using silver compound will void the processor warranty.

Originally Posted by glc I would really hesitate recommending voltage changes from default unless you know exactly what you are doing and why. That's a good way to damage components. Recommending a .25 volt increase is way too much in one step even if you are trying to overclock, and a whole volt is just about suicide. Baby steps, please. The Athlon XP's may run hot, but the Athlon 64 is a different animal. Some brands of ram respond well to voltage increases, others do not. If you have a retail processor, removing the factory applied thermal solution and using silver compound will void the processor warranty.

All that may be true; however what works in the real world often differs from manufacturers specs. My personal experience with AMD processors is that you initially have to change the voltages to stabilze the system. Factory thermal solutions are quite often inadeqaute, and that would be the first thing I'd do; not hard to improve in that area. You have to consider the current strength in the building he's at. Just measuring the socket to check for 120v a/c is not enough; neither is a surge suppressor. Ideally, you would have a high quality power filter which would the reduce or eliminate the need for voltage changes. What I can see from the original post is that the cause of the problems is with little doubt a power quality/current stength issue.

[glc]

Well, I've been building and repairing computers for a living for over 10 years, and I have *never* had to play with any voltages to "stabilize" ANY processor or ram that's running at stock speeds. I have also never had to improve on stock cooling under the same parameters - everything running stock. If you feel that the source of the problem is a "power quality/current strength" issue, I'd be recommending a different power supply first, then a motherboard replacement. Those are the only 2 things that affect the DC voltages in that computer unles there's another FAULTY component that's pulling the whole thing down.

[dataDude]

During my time here there has been evidence of power outages and brown outs. While I have everything on surge protectors, is it true that they're not functional anymore after the first surge it encounters? However, Since the last time I fixed or reinstalled the OS, I don't believe there have been issues with the voltage in the building.

I'm not sure what to do at this point, glc. Does 12.22 for the +12V setting seem a little off to you? The only things I could think of to check the stability of the system were the HDD diag (which came up negative for errors, so no reason to zero fill), and running the chkdsk after logging into Windows and then before the login screen (chkdsk /f). After logging in to Windows I ran it without the /f param and it showed the following right after finishing the security descriptor:

CHKDSK discovered free space marked as allocated in the master file table (MFT) bitmap.
Windows found problems with the file system.
Run CHKDSK with the /F (fix) option to correct these.

So I ran it with the /f param and the test showed absolutely nothing wrong, no errors of any sort. Is there something I'm missing in the error message? Some sort of "read between the lines" thing? Let me know, cuz one other interesting thing is that for every computer onto which I'm installing this XP slipstreamed with SP2, they're all getting this same error. I'm pretty sure, however, that I was receiving that error for the first before first time usage of that particular CD (I was using an XP install without slipstreaming before then). Maybe it's an SP2 problem? I'm lost.

[GaryRouth]

Originally Posted by dataDude . . . one other interesting thing is that for every computer onto which I'm installing this XP slipstreamed with SP2, they're all getting this same error. . . Maybe it's an SP2 problem?

Hmm . . . it's certainly possible that a driver or other software has a conflict with SP2: several older versions of firewalls and antivirus products, as well as some video drivers, had troubles initially. Most of these conflicts have been resolved since either by updates from Microsoft, or from the software/driver vendors.

I'd be interested to see if there's any change if you tried the repair using a different CD (if any of your friends have the XP disk with SP2 already on it, it's OK for you to use it to repair your system: you simply continue using your current product key, & not the one from the borrowed CD).

The other two remaining hardware choices do seem limited to power supply and motherboard. Since the 9700 draws considerable power, I'd start with the power supply. --> The issues that were resolved on sclark1000's machine were obviously power issues: and the symptoms seem fairly similiar. Going to overclocking for a solution might work, but it's a chancy business (glc noted the dangers - you're risking a fairly pricey cpu and video card). My view of the overclocking to solve the possible power issue is that it's a workaround, rather than a repair/replacement of the misperforming unit. Your choice, though, since it's your box.

Best of luck
. . . Gary

[glc]

Boot with your slipstreamed CD and log into the repair console. Run CHKDSK /P and see what happens. Repeat till it comes up clean. Sometimes this finds and fixes errors that chkdsk /f won't find. If you suspect bad sectors, run CHKDSK /R instead in the console.

12.22 is well within specs, officially the tolerance is 10% but I prefer to see 5%. This doesn't tell you what it's doing under load when you are looking in the bios though, nor does it show the stability or lack thereof. What brand and wattage is the power supply? A Radeon 9700 is pretty hungry.

Most consumer grade surge strips are one shot devices. If your power supply has been taking surges, it's very possible there are some weak or faulty components in there now that are impacting on the quality of your regulated DC power, and there's a possibility the motherboard voltage regulators are in the same boat. I just replaced a board that had leaky caps - it was showing about 11.4 volts on the 12 volt rail, a new power supply showed the same. Replacing the motherboard brought it back to 12.1 volts.

Originally Posted by GaryRouth The other two remaining hardware choices do seem limited to power supply and motherboard. Since the 9700 draws considerable power, I'd start with the power supply. --> The issues that were resolved on sclark1000's machine were obviously power issues: and the symptoms seem fairly similiar. Going to overclocking for a solution might work, but it's a chancy business

Oveclocking will only make whatever the problem is worse, so skip that; at least until everything else is running perfect. In my case, we also had power outages/brownouts; everything was on good surge suppressors. After the problems began, the first thing I did was swap out the surge protection, and rma'd the power supply - Nexus 4090 special edition. Even with a new cpu, mainboard, power supply, and swapped out vid card & ram...with no other periperhals connected, the problems continued. That is until the cpu voltage was dropped in .25v increments; get this - down 1.25v below the factory spec. It sounds crazy, but it TOTALLY solved the issue, and the cpu runs 10 deg c lower.

I've been working on computers for 25 years and have seen a lot of things, never had to undervolt one, always over, if needed. But it DOES work, so don't be afraid of trying it. And I completley disagree about stock cooling. It's cheap, noisy and there are many ways to improve on it. Do you really think they'd add $50 to the cost of a retail processor if its not completly neccessary, not a chance.

Good luck datadude!

[GaryRouth]

[ . . . sclark1000 -- the "under-clocking" and "under-volting" is sometimes mentioned over in the "Quiet PC" sites and forums = not only do you get a cooler running machine, it's usually a bit quieter, too (especially if you switch to the Zalman sort of cooling). It's true that there's much less risk in that. When it works, "silence is golden" . . . - Gary ]

[ one more P.S. ... 25 years? you must be my age! (I've been in IT for that same length of time = you must remember ATs, XTs, Apple IIs. . . and perhaps the megalith mainframes of the era - IBM, Control Data, Sperry, DEC, Tandem, Sun, etc. . . . ]

Originally Posted by GaryRouth [ . . . sclark1000 -- the "under-clocking" and "under-volting" is sometimes mentioned over in the "Quiet PC" sites and forums = not only do you get a cooler running machine, it's usually a bit quieter, too (especially if you switch to the Zalman sort of cooling). It's true that there's much less risk in that. When it works, "silence is golden" . . . - Gary ] [ one more P.S. ... 25 years? you must be my age! (I've been in IT for that same length of time = you must remember ATs, XTs, Apple IIs. . . and perhaps the megalith mainframes of the era - IBM, Control Data, Sperry, DEC, Tandem, Sun, etc. . . . ]

Nothing like carrying around punch cards & 8" floppies.

[dataDude]

Well, I've just come in with a major breakthrough. I decided to see what would happen if I were to install Windows XP (without SP2) and then before installing anything further to run chkdsk. The result, no issues found. Then I went to install SP2 and in the middle of installing it the system spontaneously rebooted (I had neglected to set the computer to not reboot when an error occurred). I checked the Event Viewer and found that once again I had encountered an error 1003 (the same one I've been receiving all along). I decided to uninstall SP2 (or what remnants of it were on the system) and reinstall. When the process was finally completely, I tried chkdsk and found the scanned problem had reappeared. Looks like I found the culprit. This isn't the first time I've encountered problems with SP2, but it's the first time that I've encountered this kind of problem (the other time caused my system to become unstable and have all the resources consumed almost all the time the computer was run and had to uninstall it) What the heck can I do about that though? I want to put all the Windows XP updates on there, but it looks like the only way to do so is to screw up the system in the first place. Has anyone else had to deal with this situation? I know there have been several people with the issue, but I also know that SP2 has worked very well for others. What to do now?

[GaryRouth]

Curiouser & curiouser.

Thought I'd take a glance at other posts and forums where folks had the same motherboard as yours = and came up with some possibilities.

Some of the posts mention problems with the "Cool & Quiet" technology - usually featured in both the Bios Setup & in Windows. And some problems with system lockups were solved by disabling this feature in both the Bios and in Windows.

And - the under-volting procedure was indeed mentioned as a workaround!

Interesting note about "Cool & Quiet" motherboards & memory modules - here's a quote from the silentpcreview.com forums (it is referring to socket 754 Athlon 64s):
-->"Athlon 64's memory controller can without help handle only two DIMMs. The third one requires external memory buffer present on motherboards (memory buffer is on my KT600 mobo as well). Cool'n'Quiet is not supported if three DIMMs are used because while FID_Change command is issued, A64 can not control the mobo memory buffer. That results in crash or data corruption. This rule does not apply to s939 as there are two memory controllers present, therefore it can handle 4 DIMMs without any problem with Cool'n'Quiet."

A fair sensitivity to brand, size, and placement of memory modules popped up on several of the threads for your Chaintech model. Not likely to apply to your situation, though = since you've already tested pretty thoroughly with memtest86.

Other posts pointed to ATi drivers & SP2... I recall that some early 9xxx ATi cards were sensitive to whether Service Pack 1 was installed or not - to use more recent versions of the Catalysts. I'd be curious to see what happens if you keep the video at a Standard SVGA built-in driver until after SP2 is in = and test the chkdsk error then. If it then crashes after installing the latest Catalysts, see if an intermediate driver-version helps.

Pretty good puzzle.
. . . Gary

The problem is definetly a hardware issue, so reinstalling anything is most likely a waste of time until you fix the source of the problem.

I would go with undervolting the cpu in small increments .25v at a time; keep a log of time change was made and the results. Run the system and note when/if it crashes. As you continue to lower volts, is the system staying stable for a longer period of time? My guess is yes it will. As mentioned before you may also have to raise memory voltage to stabilize. And with the video card you have, it's required to have external power (other than the agp slot) to run a stable system; not doing so will easily crash you system in a hurry, if it even boots at all.

Focus on the keywords: hardware issue, voltages, stablize. And forget about trying software solutions; it's not sp2 or anything else thats causing you difficulites.

[dataDude]

Y'know what the funny part about the Cool n Quiet system is? After I read what it was and that a more updated BIOS version download was available that would add its capability to my system, I decided right away that I wanted it. Now I find out it could be part of if not "the" cause for my errors. Bummer. There's a setting in Windows for the Cool n Quiet capability? I've never seen it.

I'm not using 3 sticks of RAM in my system. I have a 1 GB stick in the first slot and the 512 MB in the second.

I've used more than one type of ATI driver for my video card. I've used the one from the official site, and downloaded one from guru3d.com. I had told an acquaintance about my problems with gaming and he responded with guru3d and how the people behind that web site look through the official drivers, remove what they believe to be unnecessary software code, and then publish it one the site. I personally didn't get any different activity from it (freeze freeze freeze).

Currently, only the standard Windows drivers are in my system. But as was said in my last post, the error showed up after the installation of SP2, so the video driver seems to be irrelevant (unless it actually irritates the matter).

I'll check to see if I can do something with the voltages. I'm always a bit uneasy about doing so; I get annoyed with having to reset the CMOS repeatedly if the settings don't work right with the video.

[dataDude]

I decided to be daring and change some voltage settings. The BIOS showed CPU VCore: 1.53V and Mem VCore: 2.60V. I changed CPU to 1.5V (not even sure if it makes a difference), and the mem to 2.7V. When I booted back to Windows and checked the chkdsk, it reported no erros. Excellent. So I went and tried to install AVG7. The install went ok, but then when I was updating it the system blue screened, blaming avg7core.sys about the matter. Ok, no prob. I then uninstalled the program and then ran chkdsk again. The problem came back, event ID 1003. I tried running it with /f and again no error. When I rebooted, it said a serious system error had occurred and the microsoft result that came up said a device driver was to blame and there was no further info. The only drivers I have on here right now are the mobo drivers and any additional ones added through SP2. This keeps getting weirder and weirder. I think I've been showing some good patience through this, but this unexplained 1003 is really starting to get to me. Help?