Memtest86+ V1.55 Crashes With Corrupted Display

[Radagast]

Hi all,

 

I have been having random crashes of my system since I assembled it (last fall!). Windows XP (SP1) just freezes with no error code, no messages, no event log entries, nothing! Most of the time the screen goes black. Occasionally, it shows what was last on display (sometimes with screen corruption), but over 90% of the time, it goes black. If audio was playing, it continues to play the last output "command" (i.e., if it was to play the CD, the CD keeps playing, but if it was from a midi or wave file, the last note, or some distortion, plays until I can hit reset). The only thing to do is hit reset.

 

I have tried many things to resolve this problem. I have uninstalled (completely, using DriverCleaner) ATI drivers and reinstalled them (various versions, currently most recent: Catalyst 5.3), as well as VIA 4in1 drivers (currently ), WDM drivers (), and DirectX (9.0c). I have swapped out the ATI 9600XT video board for an ATI 7200 board (and back, after no change - the 7200 ran slightly longer between crashes, on average, but not enough to identify the 9600 as the problem). This is my first DDR-capable MB, so I cannot swap MB or RAM with any other system I have. I have tried running with nothing but required components installed (CPU, RAM, video, HD & PSU). I have tried changing BIOS settings, based on various forums suggestions (currently all performance BIOS settings are default, except AGP voltage is upped to 1.6V - RAM is default to SPD). However, nothing seems to cure the problem.

 

I finally broke down and downloaded Memtest86+ and burned a CD with it. I have run it several times now, and every time, after about a half-hour of running, it crashes (halts) with an unexpected interrupt (DIVIDE). However, the screen is corrupted with garbage characters all over the place (just enough of final mini-dump is visible to interpolate terms: "unexpected interrupt", "Halting!" & "DIVIDE"), so I do not have any reliable debug information. This occurs whether I run the test with both sticks of memory, or either stick alone. BTW, memory is installed in Slot 2 or Slot2 2 & 3, as the CPU heatsink does not leave enough clearance for Slot 1 memory to be installed (I have been told that this is okay).

 

Prior to these memory tests, Windows would stay up for anywhere from 5 minutes to 5 days before crashing. After the tests, I wound up with the two memory sticks reversed and now Windows crashes either during boot or within 10 minutes after the boot.

 

I have not read about this sort of error in any forums before. I have read many reports of problems with my MB, with ATI VB, with marginal PSU, etc., but most such reports are countered by reports of systems running fine with the same components. I have tried any adjustments that have been reported to fix similar problems, but they don't seem to fix my problem.

 

I have built about a dozen or two different PC's over the last 2+ decades, so I have the skills, and a lot of troubleshooting experience. However, I am stumped by this problem. I can't even determine which component may be the cause and should be returned. Does anybody have any good ideas about what could be the problem with my system?

 

System Profile:

Soyo KT600 Dragon Ultra Platinum (UDMA 133) (Via KT600 Chipset) (400Mhz Front Side Bus) | AMD Athlon XP 3200+ CPU 2.2GHz, 400 Mhz FSB | Arctic Silver | Zahlman CNPS7000A-Cu CPU Cooler | Zahlman ZM-NB32J Passive Chipset Cooler, for Northbridge | Digital Doc 5+ Multifunction Panel, Digital Temperature Sensor, Fan Speed & Voltage Monitor | 1GB DDR SDRAM: Geil Ultra Series Value Dual Channel 184 Pin 1GB (512MBx2) DDR PC-3200 w/Heatspreader | Enermax Noisetaker Plus EG375P-VE SFMA 370W PSU | Thermaltake Tsunami Dream VA3000BNA Aluminum Mid-tower Case | Maxtor 6Y160P0 Ultra Series 160GB UATA/133 HDD | Mitsumi Black FDD 1.44MB 3.5" | Toshiba SD-R5272 8x Dual Format Internal DVD

[martymcfly]

maybe the ram?

[Radagast]

maybe the ram?

458731[/snapback]

 

Yeah. I was hoping for a little more than maybe. It is a major hassle to RMA components and I need the use of my PC, so I don't want to disable the PC for the time it takes to RMA until I have a better idea of which component is the culprit.

[RnDestroyer]

I had that exact problem on my SOYO board OMG.......it was a pain its not nethin its the board maybe a bios flash would do it or just a new board .......because that same exact thing use to happen on my board and i tryed for several weeks to figure it out but mine never crashed as ofter as urs mayb once a day but no warnin no signs no nothin just like u said so try bios flash i would guess if not then chunk that sht thats wut i did....took it back got my money back for it

[martymcfly]

do you have another machine where you can try the ram? or try underclocking it to see if that helps.

[Radagast]

I had that exact problem on my SOYO board OMG.......it was a pain its not nethin its the board maybe a bios flash would do it or just a new board .......because that same exact thing use to happen on my board and i tryed for several weeks to figure it out but mine never crashed as ofter as urs mayb once a day but no warnin no signs no nothin just like u said so try bios flash i would guess if not then chunk that sht thats wut i did....took it back got my money back for it

458735[/snapback]

 

I tried downloading and flashing the latest BIOS for my board (4aa2 - 8/4/2004), but the crash in Memtest86+ still occurred. I haven't run under Windows XP long enough since then to determine if it had any affect on Windows XP stability.

 

do you have another machine where you can try the ram?

Edited April 13, 2005 by Radagast