Computer freezes

[-]

I also have started tyo get these behavior, Radeon AIW 9000Pro
DirectX9.0a (anything above gives me huge problems...thanks STUPID MS
PROGRAMMERS FOR NTSC ONLY!)

For me it is MMC that crashes every time I try to record, not the acctual
Catalyst drivers. I can play games fine and work with "Office" applications
but every time I try to record anything via MMC => System Crash - BSoD or
Dev/Null...

I have tried MMC 8.1, 8.5 and 8.7 same problem with every one regardless of
which Catalyst driver or Omega driver used same problem...Try to to record
via MMC leads to a System Crash every every time!

To me it is obvious that ATI doesn't care if MMC works or not, as long as
the Catalyst driver score good results in magazines etc. Too bad thou

/ J

Hello.

I recently built a machine from spare parts but it is very unstable.

Usually within half an hour to an hour from booting Win2K it will totally freeze.
The only thing that will continue working is mouse pointer positioning.
If I immediately reset it, the situation will repeat after a similar time frame.
Usually happens when using the browser. Did not happen when left under DOS
running the screen saver of Norton Commander (don't ask!) for several hours.

Motherboard: ASUS P5A rev 1.06, flashed latest BIOS (1011.005)
CPU: K6-2/400
Memory: One Infineon(?) 256MB PC133 DIMM (at 3-2-2)
PCI Cards: Diamond Fireport 40 (worked in another box for years without problems)
AGP: ATI AIW 8500 (never tested before), running "Catalist 3.8" drivers
OS: Win2K SP4 with latest patches from "windows update"
Drives: 2 ATA33 hard drives and 1 floppy connected to the board,
1 Plextor CD-ROM connected to the Fireport
Misc: PS/2 style mouse and keyboard, HSF on CPU, case left open

What can cause the problem?

Thank you for your suggestions.

Here is what I did:

Run memtest86 for 24 hours - no errors (memory probably good).
Booted Win2K again - crashes...
Installed the Ali AGP driver - no change (I let is stay).
Upgraded to Catalyst 3.9 plus the recent ATI patch for the AIW - no change.

However, I did not notice any crashing when I removed the Catalyst driver
completely (of course I was stuck at 800x600 16 colors but that's beside the
point) which leads me to believe that it *is* a driver problem.

The Motherboard's voltage monitoring in the BIOS showed me that the 3.3V
rail is at +3.4v and the +12v is at +11.6v (sometimies dipping to +11.5v).
Not ideal but, I believe, within spec. Combined with the fact that the
crashes stopped when I removed the driver that leads me to believe that the
PSU is not the source of the problem. Please correct me if I am wrong.

So, assuming that it is a driver problem, which driver should I use for an
AIW 8500? Which versions are known as stable?

Thank you,
Alex.

 

[Paul]

Hello.



Thank you for your suggestions.

Here is what I did:

Run memtest86 for 24 hours - no errors (memory probably good).
Booted Win2K again - crashes...
Installed the Ali AGP driver - no change (I let is stay).
Upgraded to Catalyst 3.9 plus the recent ATI patch for the AIW - no
change.

However, I did not notice any crashing when I removed the Catalyst
driver completely (of course I was stuck at 800x600 16 colors but
that's beside the point) which leads me to believe that it *is* a
driver problem.

The Motherboard's voltage monitoring in the BIOS showed me that the
3.3V rail is at +3.4v and the +12v is at +11.6v (sometimies dipping
to +11.5v). Not ideal but, I believe, within spec. Combined with
the fact that the crashes stopped when I removed the driver that
leads me to believe that the PSU is not the source of the problem.
Please correct me if I am wrong.

So, assuming that it is a driver problem, which driver should I use for =
an AIW 8500? Which versions are known as stable?

Thank you,
Alex.

Please replace myrealbox with alexoren to reply by email.

Try the following. Install a Catalyst driver and the ALI driver.
Install either Motherboard Monitor (MBM5) or Asus Probe (not both
at the same time, because they can conflict with one another on
the SMBUS), as they both have the ability to log the power supply
voltages in the background while other programs run. Set up logging
to file every 10 seconds, to get frequent readings.

Then, install a 3D game or benchmark (like 3DMark2001SE). Run the
game or benchmark in 3D mode for 30 seconds or so (i.e. you don't
really want it to crash, in case the log file gets corrupted).
Then quit the game or benchmark and examine the voltages recorded
in the log file of MBM5 or Asus Probe. Chances are your +12V is
dipping a lot lower than you think.

If the power supply is the culprit, try another power supply.

Otherwise, use the "SmartGART" tab in the Display control panel,
to change AGP settings. Try a lower AGP transfer rate, for
example. Or disable "Fast Write" if the setting is available.
The ATI driver tends to ignore the AGP BIOS settings, so you must
use SmartGART to control the thing.

HTH,
Paul

 

[Alex]

Same problem with 3.7

Sometimes it takes an hour for the crash, sometimes 10 minutes...
However, now, over 50% of the times, there is a visible display corruption just before the freeze. Often the screen will go blank as well.

install a 3D game or benchmark (like 3DMark2001SE). Run the
game or benchmark in 3D mode for 30 seconds or so (i.e. you don't
really want it to crash, in case the log file gets corrupted).
Then quit the game or benchmark and examine the voltages recorded
in the log file of MBM5 or Asus Probe. Chances are your +12V is
dipping a lot lower than you think.

It happens in @D only - usually when I open another app window (can even be notepad) or some other redraw.
The 3D OpenGL screensaver does not trigger the problem.

If the power supply is the culprit, try another power supply.

Otherwise, use the "SmartGART" tab in the Display control panel,
to change AGP settings. Try a lower AGP transfer rate, for
example. Or disable "Fast Write" if the setting is available.
The ATI driver tends to ignore the AGP BIOS settings, so you must
use SmartGART to control the thing.

SmartGART shows that it's using AGP 2X and "Fast Write" is disabled.

Thanks,
Alex

 

[Paul]

Same problem with 3.7

Sometimes it takes an hour for the crash, sometimes 10 minutes...
However, now, over 50% of the times, there is a visible display =
corruption just before the freeze. Often the screen will go blank as =
well.


It happens in @D only - usually when I open another app window (can even =
be notepad) or some other redraw.
The 3D OpenGL screensaver does not trigger the problem.


SmartGART shows that it's using AGP 2X and "Fast Write" is disabled.

Thanks,
Alex

--=20
Please replace myrealbox with alexoren to reply by email.

Here is a suggestion. I searched in Google for "p5a agp current" and
can find posts like this:

http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&[email protected]

http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&[email protected]

http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&[email protected]

I'm surprised to find posts from people who cannot run mx440's on this
board. The mx440 is a fanless Nvidia card, so power consumption should
be reasonable.

The 3.3V that the AGP I/O pins use, is provided by an onboard circuit
and not by your power supply. Changing power supplies won't fix it.

If you insist on running this config, then try the following:

1) SmartGART - reduce AGP to 1X. This will reduce the AGP I/O current
draw.
2) If that fails to give stability, then reduce the "Hardware Acceleration"
slider to its minimum. Forget using 3D. I think this setting
will still allow you to use the card as a frame buffer, so the
resolution should be changable.

Sorry I didn't think to check this earlier. It is hard to keep track
of the transition from inadequate to adequate AGP current on older
motherboards. The posts in Google suggest the P5A-B was better at
AGP than the P5A.

HTH,
Paul

 

[TMack]

Hello.

I recently built a machine from spare parts but it is very unstable.

Usually within half an hour to an hour from booting Win2K it will totally freeze.
The only thing that will continue working is mouse pointer positioning.
If I immediately reset it, the situation will repeat after a similar time frame.
Usually happens when using the browser. Did not happen when left under DOS
running the screen saver of Norton Commander (don't ask!) for several hours.

Motherboard: ASUS P5A rev 1.06, flashed latest BIOS (1011.005)

This motherboard and all the other ALI chipset socket 7 boards are notorious
for having problems with agp video cards. . It may not be possible to get
it working properly. See if it works ok with a basic pci video card such as
an old S3 Virge or similar if you have one available - you can pick them up
for next to nothing. If it runs OK then try to swap/sell your AIW 8500 for
a good pci video card and get a separate TV card if you want the TV/video
functions that the AIW provides. If you aren't too fussy about 3D gaming
then a PCI voodoo 3500 might do the business and it provides similar
TV/video functions. You might even find that the AGP Voodoo 3500 works OK
as it doesn't actually use the AGP texturing functions (none of the Voodoo3
series utilised the AGP functions), acting rather like a PCI card in the AGP
slot. There are 3 of these currently on ebay - one with a 'buy it now'
price of £25.00. Drivers are available that give full functionality in
Windows 2000 (I have one installed in my wife's pc running under win 2k - 2D
and TV quality is excellent and the 3D is good enough to play UT Quake3
etc )

Alternatively, sell/swap the P5A and get a different socket 7 motherboard
with a different chipset. These are also available for very little but it
would still be a lottery (some of the Via MVP3 chipset boards were also
prone to AGP problems). Basically you are installing a relatively recent
high-performance video card on boards that were not designed for use with
such high-spec hardware.

Tony

 

[Keith]

Alex

I had almost the same symptoms as you when I formatted my drive and
upgraded from Win98 (very stable) to WinXP. I'd get a freezeup or
automatic reboot every 5 or 10 mins. However I have an MSI mainboard
and an ATI Radeon 64 VIVO graphics card. Made whatever changes I could
in my card settings but would get the same instability with the recovery
always indicating it was a problem with my card. I found if I
completely disabled hardware acceleration in my card settings, I could
get stability but some of my 3D stuff would not display. I was near to
the point of either reinstalling Win98 or buying a new video card.

I tried changes in my BIOS settings but there's a lot there that I am
completly unaware of what the settings do. However in the MSI BIOS
setup there are two "global" options - one you pick and it sets
everything in your BIOS for optimal/fast performance (in the manual it
states this setting will almost certainly cause problems) and another
for stable performance. I chose the "stable" one and accepted whatever
changes that made. Since then - 2 days ago - I have not had a computer
crash or glitch.

I'm not familiar with your ASUS board. However you might look there and
see if you can find the fix to your problems as I did.

MSI 645E MaxU MB
P4 2.4 Ghx
512 DDR
WinXP Pro
ATI Radeon 64 VIVO.

Keith

PS to reply directly, please change the country in my address to a more
"com"mon one.