Boot fails after shutdown, works after power off.

I am building a new system using a used Q8200, used Gigabyte GA-p31-DS3L, new
2x2GB PC6400 RAM. I have had it overclocked to 2.8 GHz, running fine overnight
with Prime 95 torture test on all 4 cores. The BIOS is set for 400 FSB and
"Auto" power settings.

The problem. The machine has begun to have a problem booting. After shutdown, it
frequently will not boot properly. ( even after being off overnight) If I unplug
the power a few seconds before pushing the power button, it starts fine. In the
failure mode, it usually hangs for awhile during the memory test, actually
"memory tes" for some time before it finally completes that to "memory testing"
and continuing to the point it should be accessing the harddrive, then powering
down until it tries rebooting again. On the next reboot, the overclock has been
turned off, but it still usually won't boot, repeating the process described.


When the problem first showed up, I had a USB harddrive attached. Removing that
solved it. I had seen a problem with that on other computers. I had the bios set
to boot from a USB floppy if it found a bootable disk there (but not there). It
started having the problem again, so I turned off that option, and it worked OK.
Now it seems to do it without that setting.

Any ideas about what might be going on?

Bob F wrote:
I am building a new system using a used Q8200, used Gigabyte
GA-p31-DS3L, new 2x2GB PC6400 RAM. I have had it overclocked to 2.8
GHz, running fine overnight with Prime 95 torture test on all 4
cores. The BIOS is set for 400 FSB and "Auto" power settings.

The problem. The machine has begun to have a problem booting. After
shutdown, it frequently will not boot properly. ( even after being
off overnight) If I unplug the power a few seconds before pushing the
power button, it starts fine. In the failure mode, it usually hangs
for awhile during the memory test, actually "memory tes" for some
time before it finally completes that to "memory testing" and
continuing to the point it should be accessing the harddrive, then
powering down until it tries rebooting again. On the next reboot, the
overclock has been turned off, but it still usually won't boot,
repeating the process described.

When the problem first showed up, I had a USB harddrive attached.
Removing that solved it. I had seen a problem with that on other
computers. I had the bios set to boot from a USB floppy if it found a
bootable disk there (but not there). It started having the problem
again, so I turned off that option, and it worked OK. Now it seems to
do it without that setting.
Any ideas about what might be going on?

I started thinking about USB bus problem, and realized I had an ATI Remote
Wonder II plugged in. Removing it seems to have made the problem go away again.
Could this be a USB bus power problem?

Bob F wrote:
Bob F wrote:
I am building a new system using a used Q8200, used Gigabyte
GA-p31-DS3L, new 2x2GB PC6400 RAM. I have had it overclocked to 2.8
GHz, running fine overnight with Prime 95 torture test on all 4
cores. The BIOS is set for 400 FSB and "Auto" power settings.

The problem. The machine has begun to have a problem booting. After
shutdown, it frequently will not boot properly. ( even after being
off overnight) If I unplug the power a few seconds before pushing the
power button, it starts fine. In the failure mode, it usually hangs
for awhile during the memory test, actually "memory tes" for some
time before it finally completes that to "memory testing" and
continuing to the point it should be accessing the harddrive, then
powering down until it tries rebooting again. On the next reboot, the
overclock has been turned off, but it still usually won't boot,
repeating the process described.

When the problem first showed up, I had a USB harddrive attached.
Removing that solved it. I had seen a problem with that on other
computers. I had the bios set to boot from a USB floppy if it found a
bootable disk there (but not there). It started having the problem
again, so I turned off that option, and it worked OK. Now it seems to
do it without that setting.
Any ideas about what might be going on?

I started thinking about USB bus problem, and realized I had an ATI Remote
Wonder II plugged in. Removing it seems to have made the problem go away again.
Could this be a USB bus power problem?



If you disconnect the USB devices, and then plug them in later
when the system is running, is the computer having any problem
recognizing USB devices ? You would think, if there was a general
USB problem (power or frequency), there'd be other symptoms.

Also, now that it's working with the ATI remote removed, what
happens if you put the USB storage thing back ? (Even leaving
the Legacy USB Storage disabled, would allow the USB storage
device to function as an electrical load.)

I looked in your manual, and I don't see any really useful things
to flip for testing. You've already disabled "Legacy USB Storage",
so that isn't it. And the ATI Remote Wonder II, probably contains
a radio receiver only (low power), and the USB device class is like
HID (human interface device, like keyboard or mouse) on that, so
this likely isn't a USB Mass Storage class issue. (Some other BIOS
in the past, have got tripped up on USB storage, even innocent things
like flash storage inside a USB printer.) You're probably using the
default of "Disabled", for the DOS USB keyboard and mouse support.

ftp://download.gigabyte.ru/manual/mo...ds3l_s3l_e.pdf

That leaves things like a voltage problem with the chipset. You're overclocking,
and on some boards, you might bump the chipset voltage one notch as a
test. But your board would use a separate supply for Northbridge (P31,
1.25V nominal, 1.375V max) and Southbridge (ICH7, 1.05V nominal, 2.1V max).
Adjusting your Northbridge (MCH), wouldn't help the Southbridge where the USB
is located. The USB shouldn't be overclocked in any case, and as far as
I know, the 48MHz it might be deriving a clock from, would run at a constant
rate. Modern clock generator schemes have a lot of stuff locked, so things
like your PCI stay at 33MHz, 48MHz clock stays put, and so on. And
even some clocks, like PCI Express, are locked to 100MHz, unless
you choose to overclock them on purpose via a BIOS setting. So modern
designs are pretty good, not like the older ones, where overclocking
the CPU caused all the system busses to run at the wrong frequency.

Just for the record, what other USB devices are connected to the
machine ?

Paul

Paul wrote:
Bob F wrote:
Bob F wrote:
I am building a new system using a used Q8200, used Gigabyte
GA-p31-DS3L, new 2x2GB PC6400 RAM. I have had it overclocked to 2.8
GHz, running fine overnight with Prime 95 torture test on all 4
cores. The BIOS is set for 400 FSB and "Auto" power settings.

The problem. The machine has begun to have a problem booting. After
shutdown, it frequently will not boot properly. ( even after being
off overnight) If I unplug the power a few seconds before pushing
the power button, it starts fine. In the failure mode, it usually
hangs for awhile during the memory test, actually "memory tes" for
some time before it finally completes that to "memory testing" and
continuing to the point it should be accessing the harddrive, then
powering down until it tries rebooting again. On the next reboot,
the overclock has been turned off, but it still usually won't boot,
repeating the process described.

When the problem first showed up, I had a USB harddrive attached.
Removing that solved it. I had seen a problem with that on other
computers. I had the bios set to boot from a USB floppy if it found
a bootable disk there (but not there). It started having the problem
again, so I turned off that option, and it worked OK. Now it seems
to do it without that setting.
Any ideas about what might be going on?

I started thinking about USB bus problem, and realized I had an ATI
Remote Wonder II plugged in. Removing it seems to have made the
problem go away again. Could this be a USB bus power problem?



If you disconnect the USB devices, and then plug them in later
when the system is running, is the computer having any problem
recognizing USB devices ? You would think, if there was a general
USB problem (power or frequency), there'd be other symptoms.

They work fine if plugged in later.


Also, now that it's working with the ATI remote removed, what
happens if you put the USB storage thing back ? (Even leaving
the Legacy USB Storage disabled, would allow the USB storage
device to function as an electrical load.)

I looked in your manual, and I don't see any really useful things
to flip for testing. You've already disabled "Legacy USB Storage",
so that isn't it. And the ATI Remote Wonder II, probably contains
a radio receiver only (low power), and the USB device class is like
HID (human interface device, like keyboard or mouse) on that, so
this likely isn't a USB Mass Storage class issue. (Some other BIOS
in the past, have got tripped up on USB storage, even innocent things
like flash storage inside a USB printer.) You're probably using the
default of "Disabled", for the DOS USB keyboard and mouse support.

Correct on that last one.


ftp://download.gigabyte.ru/manual/mo...ds3l_s3l_e.pdf

That leaves things like a voltage problem with the chipset. You're
overclocking, and on some boards, you might bump the chipset voltage
one notch as a test. But your board would use a separate supply for
Northbridge (P31,
1.25V nominal, 1.375V max) and Southbridge (ICH7, 1.05V nominal, 2.1V
max). Adjusting your Northbridge (MCH), wouldn't help the Southbridge
where the USB is located. The USB shouldn't be overclocked in any
case, and as far as I know, the 48MHz it might be deriving a clock from, would
run at a
constant rate. Modern clock generator schemes have a lot of stuff
locked, so things like your PCI stay at 33MHz, 48MHz clock stays put, and so
on. And
even some clocks, like PCI Express, are locked to 100MHz, unless
you choose to overclock them on purpose via a BIOS setting. So modern
designs are pretty good, not like the older ones, where overclocking
the CPU caused all the system busses to run at the wrong frequency.

Just for the record, what other USB devices are connected to the
machine ?

USB Trackball. I tried plugging the USB external drive through a powered 7 port
USB 2.0 hub, and it still hangs the bootup process. Haven't tried that with the
romote Wonder II yet.

I am beginning to think I have a problem motherboard, as I have had a couple
hangs, where the disk LED turns on solidly, and the maching fails to reapond for
many minutes. I will post that as a separate question.