Strange behavior from an ASUS A7V600 mobo

[Erik Ch. Ohrnberger]

After assembling the system and making sure that the CPU temp did not peak
over normal levels (stays steady at 119F) I booted up memtest-86 V3.00 in
order to test all the memory. The memory tests ran fine and did not detect
any errors.

Now, to install Windows XP. I have the CD-DVD ROM drive set as secondary IDE
controller slave, and I have a WD SATA drive as Master on Channel 0.
However, when I boot from the Windows XP CD, I get the prompt booting from
CD, and no further response. I see no IDE controller activity via the front
panel light, nothing.

OK, so I unplug the SATA drive and I put a regular EIDE drive connected on
primary IDE controller as master. Now I can boot from the Windows XP CD and
install the OS. Except that when it starts setting up the devices, I get an
IRQL_NOT_LESS_OR_EQUAL Stop 0x0000000A BSOD. I've tried to install Windows
XP multiple times, and I get this error most often, while I've only received
one STOP 0x0000004E PFN_LIST_CORRUPT BSOD.

Now, from what I've researched on the Internet, when you get a
IRQL_NOT_LESS_OR_EQUAL BSOD, a device driver tried to grab an IRQ higher
than what it should have. Could this be the onboard Gigabit Ethernet? A
STOP 0x0000004E PFN_LIST_CORRUPT BSOD, it's usually bad memory. But if
that's the case, why did the memory test pass? I've made sure to get good
memory for the system:

Crucial 184 Pin 256MB DDR PC-3200 4T - OEM
Speed: DDR400(PC3200)
Type: 184 Pin DDR SDRAM
Error Checking: Non-ECC
Registered/Unbuffered: Unbuffered
Cas Latency: 3
Support Voltage: 2.6V
Bandwidth: 3.2GB/s
Organization: 32M x 64 -Bit
Warranty: Lifetime
Model#: CT3264Z40B.4T

The other thing that bothers me is when I try to boot from the CD with just
the SATA drive in, that it freezes when the CD is just starting to boot.

The goal that I want to get to is to have the SATA drive installed and be
the main boot disk. There will be at least 2 OSs installed there, and a
third separate Linux installation an EIDE disk, probably as a slave on the
primary IDE. I'll have a DVD-ROM as master on primary, a CD-RW as master on
secondary, and a DVD-RW as slave on secondary.

Any advice that you can offer about booting with just the SATA drive
installed as well as the Windows XP installation issues would be greatly
appreciated.

Thanks,

Erik.

 

[kony]

After assembling the system and making sure that the CPU temp did not peak
over normal levels (stays steady at 119F) I booted up memtest-86 V3.00 in
order to test all the memory. The memory tests ran fine and did not detect
any errors.

Now, to install Windows XP. I have the CD-DVD ROM drive set as secondary IDE
controller slave, and I have a WD SATA drive as Master on Channel 0.
However, when I boot from the Windows XP CD, I get the prompt booting from
CD, and no further response. I see no IDE controller activity via the front
panel light, nothing.

OK, so I unplug the SATA drive and I put a regular EIDE drive connected on
primary IDE controller as master. Now I can boot from the Windows XP CD and
install the OS. Except that when it starts setting up the devices, I get an
IRQL_NOT_LESS_OR_EQUAL Stop 0x0000000A BSOD. I've tried to install Windows
XP multiple times, and I get this error most often, while I've only received
one STOP 0x0000004E PFN_LIST_CORRUPT BSOD.

Now, from what I've researched on the Internet, when you get a
IRQL_NOT_LESS_OR_EQUAL BSOD, a device driver tried to grab an IRQ higher
than what it should have. Could this be the onboard Gigabit Ethernet? A
STOP 0x0000004E PFN_LIST_CORRUPT BSOD, it's usually bad memory. But if
that's the case, why did the memory test pass? I've made sure to get good
memory for the system:

Crucial 184 Pin 256MB DDR PC-3200 4T - OEM
Speed: DDR400(PC3200)
Type: 184 Pin DDR SDRAM
Error Checking: Non-ECC
Registered/Unbuffered: Unbuffered
Cas Latency: 3
Support Voltage: 2.6V
Bandwidth: 3.2GB/s
Organization: 32M x 64 -Bit
Warranty: Lifetime
Model#: CT3264Z40B.4T

The other thing that bothers me is when I try to boot from the CD with just
the SATA drive in, that it freezes when the CD is just starting to boot.

The goal that I want to get to is to have the SATA drive installed and be
the main boot disk. There will be at least 2 OSs installed there, and a
third separate Linux installation an EIDE disk, probably as a slave on the
primary IDE. I'll have a DVD-ROM as master on primary, a CD-RW as master on
secondary, and a DVD-RW as slave on secondary.

Any advice that you can offer about booting with just the SATA drive
installed as well as the Windows XP installation issues would be greatly
appreciated.

Thanks,

Erik.

Underclock the system to lowest supported FSB (probably 100MHz) and
synchronous (same) memory bus speed, with conservative (high numbers)
timings or "auto", "SPD", memory timings in the BIOS.

With the system underclocked, try to replicate the error condition(s).

 

[Erik Ch. Ohrnberger]

Kony,
Thanks for the hint.

OK, so I changed the Memory Frequency BIOS from 400 to 266 MHz (lowest),
and memory timing was already to SPD, so done. Try again. I plugged in the
SATA drive again, and booted with the XP CD.
Boot from ATAPI-CD-ROM, No Emulation (waited at least 30 seconds no 'Press
any key to boot from CD' prompt)

Disconnected the power to the SATA drive. Rebooted, and now XP installation
proceeds normally with a successful installation. Installed all the drivers
for the SATA Raid (which I don't plan to use) as well as the Gig-Ethernet
card, display (ATI), ASUS probe, etc.

Installed XP SP1 and then web based updates (oh God! 76 MB of patches and
updates! Well, I'm glad that I'm on a cable modem at least! I'll do that
later). Installed SETIAtHome to produce a compute load on the system.
Installation went fine.

Flipped the memory speed back up to 'Auto' to see if the system is now
stable. It's not. Seems like SETIAtHome causes the system to crash with a
BSOD (which I can't see due to an instantaneous reboot in spite of having
that turned off in XP). Also seems like the ATI Radeon drivers have fallen
back to software rendering. Updating with the latest from the ATI web site.

Switched memory speed from 'Auto' to 333 MHz. See if this stabilizes things.
If this Motherboard cannot sustain 400 MHz FSB speed, it'll have to go back
(or the memory will have to). I already have a system that runs at 266 speed
and 333 is not that much of a gain (basically what I had 2 years ago).

Will keep everyone informed as to how things proceed.

Erik.

 

[kony]

On Sat, 28 Feb 2004 19:19:53 -0500, "Erik Ch. Ohrnberger"

Flipped the memory speed back up to 'Auto' to see if the system is now
stable. It's not. Seems like SETIAtHome causes the system to crash with a
BSOD (which I can't see due to an instantaneous reboot in spite of having
that turned off in XP). Also seems like the ATI Radeon drivers have fallen
back to software rendering. Updating with the latest from the ATI web site.

MIght be a power supply problem. What make/model power supply? Have you
taken voltage readings during heavy load?

Switched memory speed from 'Auto' to 333 MHz. See if this stabilizes things.

Set memory speed to synchronous setting, same speed as FSB... Not just for
testing, but use.

If this Motherboard cannot sustain 400 MHz FSB speed, it'll have to go back
(or the memory will have to). I already have a system that runs at 266 speed
and 333 is not that much of a gain (basically what I had 2 years ago).

It's probably not the motherboard. "Could" be, but failing any reports of
similar issues with the board I'd suspect the memory or power supply.
Memtest may not catch all errors, sometimes a module barely stable will
pass memtest86 but err in windows, especially if you didn't run memtest86
for at least a dozen hours. Another test with memtest86 might be to
manually raise FSB speed to 5% over desired value, test at that speed.
For my systems I insist that it test stable at higher than actual speed at
which it'll be used, regardless of whether it'll be running overclocked or
not.

 

[Erik Ch. Ohrnberger]

MIght be a power supply problem. What make/model power supply? Have you
taken voltage readings during heavy load?

The Power supply is a Cogen 350 W, Model 250 X1.

things.

Set memory speed to synchronous setting, same speed as FSB... Not just for
testing, but use.

I'm not sure where in the BIOS to set this. Nothing obvious pops out at me.

It's probably not the motherboard. "Could" be, but failing any reports of
similar issues with the board I'd suspect the memory or power supply.
Memtest may not catch all errors, sometimes a module barely stable will
pass memtest86 but err in windows, especially if you didn't run memtest86
for at least a dozen hours. Another test with memtest86 might be to
manually raise FSB speed to 5% over desired value, test at that speed.
For my systems I insist that it test stable at higher than actual speed at
which it'll be used, regardless of whether it'll be running overclocked or
not.

I did run memtest-86 V3 for at least 28 hours without fail at the 400 MHz
FSB setting hence my assumption that the memory was up to 400 MHz FSB. I'll
try the 5% increase in FSB for a stress test. It sounds like the right thing
to do at this point in time. I think that you are wise in your testing
standards.
I set the CPU External Freq to 205/43, which the BIOS says is 410 MHz.

Back to memtest! Let you know how it works out.

 

[kony]

On Sat, 28 Feb 2004 20:21:57 -0500, "Erik Ch. Ohrnberger"

The Power supply is a Cogen 350 W, Model 250 X1.

Replace the Codegen, even if it isn't the current problem (which it likely
is), it may be soon enough, even damaging the board.

You don't need more than 350W but that PSU isn't even worth 250W.

 

[Erik Ch. Ohrnberger]

Well, OK, if you say so (a new power supply is pretty cheap insurance I
figure)

But I watched the voltage for around 10 minutes: (Seems stable to be, ASUS
probe never alarmed)
With SETI: WO SETI:
12.544 4.73 3.216 1.728 12.608 4.757 3.216 1.728
12.608 4.73 3.216 1.728 12.544 4.73 3.232 1.728
12.672 4.757 3.216 1.728 12.672 4.757 3.232 1.728

Again, thanks very much for your insight and advice.

Erik.

 

[kony]

Well, OK, if you say so (a new power supply is pretty cheap insurance I
figure)

But I watched the voltage for around 10 minutes: (Seems stable to be, ASUS
probe never alarmed)
With SETI: WO SETI:
12.544 4.73 3.216 1.728 12.608 4.757 3.216 1.728
12.608 4.73 3.216 1.728 12.544 4.73 3.232 1.728
12.672 4.757 3.216 1.728 12.672 4.757 3.232 1.728

Again, thanks very much for your insight and advice.

Yep, that pretty much sums it up...

Your power supply is inadequate on it's 5V rail, so to increase that 5V
voltage it's also increasing the 12V rail. Even though ATX specs allow a
pretty large % deviation in voltage, in reality the faster newer system
have larger, faster changes in current, are more susceptible to low
voltage, slow response.

All (that I've even seen) power supplies regulate the 5V rail. It should
be very close to 5V. A poorly adjusted power supply, or one with a part
or two with bad tolernace, might have a 5V rail slightly too high or low,
that's not necessarily a sign of a problem, but the 12V rail should also
reflect this, not be drifting in the opposite direction.

 

[Erik Ch. Ohrnberger]

korny,
Thanks for all your help. I've shut it down for now, pulled the PSU,
and plan on getting a better one tomorrow. Hopefully, it'll fix all that
ails it. ;-)

Thanks very much, I owe you, just a question as to how to make good on
it.

Erik.

 

[Erik Ch. Ohrnberger]

<snip>
Revised Voltage Study:
Replaced the PSU with a new 450W unit, and have the following numbers to
report:
13.056 4.919 3.184 1.728 (SETIAtHome)
13.056 4.919 3.168 1.728 (SETIAtHome)
12.736 4.999 3.216 1.728 (No SETIAtHome)
12.8 4.999 3.216 1.728 (No SETIAtHome)

However, even with this new PSU, when I change the RAM speed to 400 MHz, I
get spontaneous reboots when running SETIAtHome. Very strange. I'm
beginning to think that it might very well be the memory.

Erik.

 

[kony]

<snip>
Revised Voltage Study:
Replaced the PSU with a new 450W unit, and have the following numbers to
report:
13.056 4.919 3.184 1.728 (SETIAtHome)
13.056 4.919 3.168 1.728 (SETIAtHome)
12.736 4.999 3.216 1.728 (No SETIAtHome)
12.8 4.999 3.216 1.728 (No SETIAtHome)

However, even with this new PSU, when I change the RAM speed to 400 MHz, I
get spontaneous reboots when running SETIAtHome. Very strange. I'm
beginning to think that it might very well be the memory.

Erik.

The voltage readings still look a bit off. Is that a name-brand power
supply? You didn't need a higher wattage power supply, rather a decent
name-brand that could be trused to be built well and have an accurate
rating on it's label. A generic 450W might only be worth 250W, and
believe it or not could be exactly the same power supply under that metal
casing... for years now I've seen almost identical supplies sold, with
increasing wattage ratings on the label but still the same interior as 5
year old supplies labeled as 235-250W. Either way, the other power
supply was clearly struggling, better you replaced it now rather than
after it stressed the motherboard. Hopefully the replacement is a decent
unit.

You're adding a lot of unnecessary random info to your posts instead of a
clear, concise description of the system and then the exact problem. That
will discourage many people from wading through it all to respond. If you
had some other memory available you might try that or do some searching
for issues of that particular motherboard, at web forums like
http://forums.amdmb.com

 

[Erik Ch. Ohrnberger]

<snip>

I figured out what needed to be done.

I had version 1005 of the BIOS. After upgrading to version 1006, the
problem appears to bo resolved, fixed solved!

Am now running at 400 MHz memory access speed!

Thanks for all the help and support everyone!

Erik.