computer won't power up - BIOS problem?

[Yup]

I'm running an old 450MHz CPU on an Asus P2B motherboard. Up until
today the computer was working fine. I just got back from my Christmas
holiday, during which I left the computer unplugged while I was away -
about 10 days. I also turned off the heating in my place as well - so
the apartment got cold while I was away (I'm in Norway).

When I plugged the computer back into the wall after the holiday (not
having touched anything), I get nothing from it. No glowing LED saying
it's plugged in, and no response when pressing the power/reset
buttons.

I tried a couple of suggestions from other Google groups posts:
1. Unplugged and replugged the video card.
2. Took out the BIOS battery, cleaned the contacts.

Still no luck.

What I then tried was to just take out the BIOS battery and plug the
computer in. Doing so, all the lights come on at the front (and on the
keyboard), but nothing shows up on the screen and the computer doesn't
start to boot.

Any ideas? Is the problem a result of the battery, or the cold?

Thanks ahead of time for your help.

 

[Vanguard]

I'm running an old 450MHz CPU on an Asus P2B motherboard. Up until
today the computer was working fine. I just got back from my Christmas
holiday, during which I left the computer unplugged while I was away -
about 10 days. I also turned off the heating in my place as well - so
the apartment got cold while I was away (I'm in Norway).

When I plugged the computer back into the wall after the holiday (not
having touched anything), I get nothing from it. No glowing LED saying
it's plugged in, and no response when pressing the power/reset
buttons.

I tried a couple of suggestions from other Google groups posts:
1. Unplugged and replugged the video card.
2. Took out the BIOS battery, cleaned the contacts.

Still no luck.

What I then tried was to just take out the BIOS battery and plug the
computer in. Doing so, all the lights come on at the front (and on the
keyboard), but nothing shows up on the screen and the computer doesn't
start to boot.

Any ideas? Is the problem a result of the battery, or the cold?

Thanks ahead of time for your help.

First off you need to determine if you are getting any power out of the
power supply. When you plug in the computer and power up, do you hear
the drive whine up? Do the fans start spinning? Do you hear any beeps?
Does the Power LED glow on the system case? Do the
CapsLock-NumLock-ScrollLock LEDs flash on the keyboard? If you get none
of those, have you checked that you actually have any power at the
outlet by plugging an incandescent lamp into it? Is the computer
plugged into the outlet or into a power strip (and, if a power strip, is
it powered on and have you tested it circuit breaker and plugged a light
into it)? Is there a circuit breaker on the backside of the power
supply where you push in a button to reset it? Is there a fuse that you
can unscrew from the back of the power supply to check if it blew?

 

[Zotin Khuma]

I'm running an old 450MHz CPU on an Asus P2B motherboard. Up until
today the computer was working fine. I just got back from my Christmas
holiday, during which I left the computer unplugged while I was away -
about 10 days. I also turned off the heating in my place as well - so
the apartment got cold while I was away (I'm in Norway).

When I plugged the computer back into the wall after the holiday (not
having touched anything), I get nothing from it. No glowing LED saying
it's plugged in, and no response when pressing the power/reset
buttons.

I tried a couple of suggestions from other Google groups posts:
1. Unplugged and replugged the video card.
2. Took out the BIOS battery, cleaned the contacts.

Still no luck.

What I then tried was to just take out the BIOS battery and plug the
computer in. Doing so, all the lights come on at the front (and on the
keyboard), but nothing shows up on the screen and the computer doesn't
start to boot.

Any ideas? Is the problem a result of the battery, or the cold?

Thanks ahead of time for your help.

Your BIOS battery is probably dead. As your computer must be
several years old, the battery probably had just about enough energy
left keep going so long as it was kept in a warm environment, aided
by the heat produced inside the computer itself when it was in
regular use. Being left alone in a Norwegian winter for more than
a week must have done it in for good.

Some computers can boot up with a good battery or with none,
but not with a weak one. Yours came closer to booting up with
no battery than with a weak one. Just pop in a new one and set
up your BIOS.

 

[Vanguard]

Your BIOS battery is probably dead. As your computer must be
several years old, the battery probably had just about enough energy
left keep going so long as it was kept in a warm environment, aided
by the heat produced inside the computer itself when it was in
regular use. Being left alone in a Norwegian winter for more than
a week must have done it in for good.

Some computers can boot up with a good battery or with none,
but not with a weak one. Yours came closer to booting up with
no battery than with a weak one. Just pop in a new one and set
up your BIOS.

Guess I've seen a computer that wouldn't boot just because the battery
was dead. The battery is supposed to provide voltage to maintain the
values saved in the CMOS copy of the BIOS' EEPROM values when there is
no power. It isn't needed when there IS power. If you, for example,
have a computer that loses time or loses BIOS settings when you power
off, just don't power off until you get a replacement battery. If the
battery goes dead so the CMOS doesn't have sufficient voltage to retain
a *copy* of the BIOS settings, the original values get read from the
EEPROM and copied back into the CMOS table (since there is now power
after you turned on the power supply) and the machine boots.

The values were originally in the EEPROM. That's what you get when you
first power up your system. That's what got stored in the EEPROM so
there actually would be usable defaults for the mobo until whenever the
jobber happened to slap in a battery to build the machine. If you don't
want to keep the customized settings on a power up, you could leave out
the battery so the defaults get used on every power up.

The only way I could see a battery causing a problem with the CMOS table
(a copy of the table from EEPROM) would be if it wasn't dead but just
didn't have enough oompf to keep the voltage high enough to ensure the
CMOS table didn't get corrupted. Some settings might then lose their
values and you don't get what you expect. However, the BIOS from the
EEPROM gets loaded first to run its bootstrap program (there are no
programs in the CMOS copy) and it performs a CRC against the values in
the CMOS copy and compares that CRC value against the one stored in the
CMOS table. If they don't match, you get some error message telling you
the CMOS copy is corrupted and you'll have to reload the values from the
EEPROM.

If the battery were dead, the EEPROM values get copied into the CMOS and
the boot continues. If the battery is insufficient to guarantee the
values in the CMOS don't get corrupted, the CRC check detects the
corruption and alerts you (so you clear the CMOS or just remove the
battery to let the CMOS drain, and then boot without a battery until you
get around to getting one).

A dead battery won't prevent you from booting up. A failing battery
might force you to wipe a corrupted table in CMOS and reload the default
one from EEPROM, but that still doesn't prevent you from booting up.

 

[kony]

A dead battery won't prevent you from booting up. A failing battery
might force you to wipe a corrupted table in CMOS and reload the default
one from EEPROM, but that still doesn't prevent you from booting up.

Well that's true to a certain extent but not the whole
story. I've a box here right now that had an almost dead
battery earlier today. The board acted completely dead, not
even CPU fan spinning upon attempted power-on, only the
power LED glowing. I cleared CMOS to no avail w/AC power
off, and eventually pulled battery and installed a new one.
Only then did the board respond at all... was an Asus
A7V333.

 

[Spajky]

I'm running an old 450MHz CPU on an Asus P2B motherboard.
When I plugged the computer back into the wall after the holiday (not
having touched anything), I get nothing from it.

Try to re-seat the Cpu in the Slot

 

[Vanguard]

Well that's true to a certain extent but not the whole
story. I've a box here right now that had an almost dead
battery earlier today. The board acted completely dead, not
even CPU fan spinning upon attempted power-on, only the
power LED glowing. I cleared CMOS to no avail w/AC power
off, and eventually pulled battery and installed a new one.
Only then did the board respond at all... was an Asus
A7V333.

I stand corrected. If Kony says it happens then it does. Must be that
I've been lucky and either had a good battery (it boots), a bad battery
(it boots), or a weak battery (CMOS corrupted, checksum error, remove
battery, clear CMOS, it boots, and sometime later replace the battery).

I can see corrupted values in the CMOS table preventing boot up (because
the hard drive and other devices are unknown or incorrectly
parameterized) but a clear of the CMOS should force a reload of the
defaults from the EEPROM even if that reload occurs on every boot until
you finally get around to replacing the battery. Maybe if the defaults
in the EEPROM are not usable with the current hardware configuration
then it won't boot. Defaults that are unusable? Then how could you
ever used the mobo in the first place? If the setup is good and you
replace the battery quickly enough, the values are still in the CMOS and
you boot as before. If the setup is good and you replace the battery
too slowly or leave it out, the CMOS drains and you lose its settings
which can generate a checksum, low battery, or other error, and which
forces a reload of the defaults from the EEPROM (although some let you
save your customized settings in EEPROM and its that table that gets
reloaded). If the settings in the CMOS are corrupt with a good or bad
battery, you have to clear the CMOS to get the good values from EEPROM
(i.e., you have a corrupted table, so you clear it to force a reload of
defaults into it).

Know if the Asus mobo you mentioned will boot WITHOUT a battery (i.e.,
power down, remove battery, clear CMOS, and boot without the battery)?
If it doesn't boot without a battery, did you try clearing the CMOS
(with the battery still out) and try booting again? If that worked,
will the next boot work okay? A mobo that won't boot without a battery
seems like some very wrong circuit design or a defective component.

 

[kony]

Know if the Asus mobo you mentioned will boot WITHOUT a battery (i.e.,
power down, remove battery, clear CMOS, and boot without the battery)?
If it doesn't boot without a battery, did you try clearing the CMOS
(with the battery still out) and try booting again? If that worked,
will the next boot work okay? A mobo that won't boot without a battery
seems like some very wrong circuit design or a defective component.

Well I still have the old battery, still have the board out
too, will try that.

The history of the board was that it was working fine when
shipped to a customer, it sat unpowered (& disconnected from
AC) for about 2 weeeks AFAIK. Customer then proceeded to
try board, couldn't get it working, and since I didn't feel
comfortable telling this customer to do any troubleshooting
on it, I simply sent a replacement board. Customer returned
the A7V333, which I might've tried upon receipt but I can't
recall it now. Months(?) later I was disassembling another
testbed and decided it was convenient to swap in the A7V333
and found it odd that there was no sign of life at all...
except that power LED, but I meant the onboard 5VSB LED, not
the case power LED.

Board consistently posts and boots DOS so I'll swap the
batteries around and report back.

 

[kony]

Board consistently posts and boots DOS so I'll swap the
batteries around and report back.

I've checked that A7V333...

AC power disconnected, battery pulled - Acts dead, no post
(next step, seconds later) AC disconnected, CMOS cleared -
Acts dead, no post

Checked voltage on old battery, only a few dozem mV.
Reinstalled dead battery just for heck of it - no change,
still dead

Pulled batery, cleared CMOS again w/ac-off, still nothing
w/o battery

Reinstall good battery - 1st time system posts since point
when battery was removed

 

[Vanguard]

I've checked that A7V333...

AC power disconnected, battery pulled - Acts dead, no post
(next step, seconds later) AC disconnected, CMOS cleared -
Acts dead, no post

Checked voltage on old battery, only a few dozem mV.
Reinstalled dead battery just for heck of it - no change,
still dead

Pulled batery, cleared CMOS again w/ac-off, still nothing
w/o battery

Reinstall good battery - 1st time system posts since point
when battery was removed

No battery, CMOS cleared, and it won't boot (presumably that means you
can't even get as far as to the BIOS setup screens)? Christ. Not good,
not good at all.

Hmm, does this have those fat boxes with the RTC and memory circuit
inside, like the Dallas box? The picture (even when exploded) at
http://usa.asus.com/prog/spec.asp?m=A7V333&langs=09# is too tiny for me
to see what's on that mobo. That had a memory circuit, crystal, and
tiny battery inside. Maybe the battery inside that NVRAM box is dead (I
never liked that they called it NVRAM because it had a battery inside so
the RAM really wasn't non-volatile just because the battery was hidden).
I recall you could pop off the cover to look inside.

Of the other posts found where A7V333 owners couldn't boot without the
battery, some mentioned that they were replacing their coin cell battery
every 2 to 4 months. I know a circuit to build that can use a cheap
cordless telephone 3.6V Nicad battery, 1n4001 diode, and 10-ohm, 1/4W
resistor that can be used which has the nicad recharging and takes the
place of the battery so you get around the problem of repeatedly having
to replace the battery at short intervals. But that the problem even
exists points to a defective component that is drawing too much current
from the battery (which has no recharge circuit so they just keep
draining). Of course, you would actually have to be the user of that
mobo to know after a long time if that mobo was eating up batteries
every few months instead of lasting 3 to 7 years.

Time to throw some more batteries into the toolkit, I guess.

 

[kony]

No battery, CMOS cleared, and it won't boot (presumably that means you
can't even get as far as to the BIOS setup screens)? Christ. Not good,
not good at all.

Right, no POST, no bios setup, the fans don't spin... it
doesn't turn the power supply on.

I wouldn't necessarily consider it a bad thing though except
for some kind of emergency situation, (or troubleshooting)
as this has never been a problem/issue except for me (and
others) who had to do a board replacement because end-user
didn't check the battery.

Hmm, does this have those fat boxes with the RTC and memory circuit
inside, like the Dallas box? The picture (even when exploded) at
http://usa.asus.com/prog/spec.asp?m=A7V333&langs=09# is too tiny for me
to see what's on that mobo. That had a memory circuit, crystal, and
tiny battery inside. Maybe the battery inside that NVRAM box is dead (I
never liked that they called it NVRAM because it had a battery inside so
the RAM really wasn't non-volatile just because the battery was hidden).
I recall you could pop off the cover to look inside.

No, but here's a pic:

http://arch.pconline.com.cn/pchardware/tpylab/main/10203/pic/rev-a7v333-boardLG.jpg

Of the other posts found where A7V333 owners couldn't boot without the
battery, some mentioned that they were replacing their coin cell battery
every 2 to 4 months.

Well I don't know what their issue was, but this board had
it's original battery still, had been used intermittently
for (maybe?) a couple years here without any problem, being
disconnected from AC quite often as it wasn't in a regularly
used system but for a few of the first months of it's life.
No way to know at this point whether the battery life was
particularly bad but it doesn't seem that way to me.

It was a nice board for it's time, a little on the picky
side PCI-wise due to it being Via chipset, just got eclipsed
by nForce2 and higher levels of integration instead of
separate RAID, USB2, and lower-end (CMI) sound chip.

I know a circuit to build that can use a cheap
cordless telephone 3.6V Nicad battery, 1n4001 diode, and 10-ohm, 1/4W
resistor that can be used which has the nicad recharging and takes the
place of the battery so you get around the problem of repeatedly having
to replace the battery at short intervals. But that the problem even
exists points to a defective component that is drawing too much current
from the battery (which has no recharge circuit so they just keep
draining). Of course, you would actually have to be the user of that
mobo to know after a long time if that mobo was eating up batteries
every few months instead of lasting 3 to 7 years.

Time to throw some more batteries into the toolkit, I guess.

Hard to say, I've come across users who kept buying old
batteries from same source so they were practially dead
already... the "new" battery I tried wasn't particularly
"new" though, just a mystery spare I must've pulled out of
some old board somewhere when throwing things away. I'll
have to remember to replace it with a fresh one if I
redeploy this board.

 

[Zotin Khuma]

Right, no POST, no bios setup, the fans don't spin... it
doesn't turn the power supply on.

I wouldn't necessarily consider it a bad thing though except
for some kind of emergency situation, (or troubleshooting)
as this has never been a problem/issue except for me (and
others) who had to do a board replacement because end-user
didn't check the battery.

The fact is that different mobos behave in different ways with a weak
or no battery. While many (maybe most) will boot up with no battery
installed, some others simply refuse to do so. I've come across such
types though I didn't make a note of which models those were.

 

[Zotin Khuma]

Right, no POST, no bios setup, the fans don't spin... it
doesn't turn the power supply on.

I wouldn't necessarily consider it a bad thing though except
for some kind of emergency situation, (or troubleshooting)
as this has never been a problem/issue except for me (and
others) who had to do a board replacement because end-user
didn't check the battery.

The fact is that different mobos behave in different ways with a
weak or no battery. Some will start acting up when the battery
voltage drops below 2V while others work down to about 1V.
Many (maybe most) boot up with no battery installed but some
others simply refuse to do so. I've come across such types
before though I didn't make a note of which ones.

 

[yaweh32]

Thanks for all the replies, folks. I appreciate it.

I'm now writing this on the previously dead computer, so it's all ok
now. However, fixing it wasn't as straightforward as just putting in
the new battery.

When I slotted the new battery in, the computer was still dead. I
started wondering what the hell was going on - took out the battery,
cleared the CMOS, put the battery back in, still nothing.

Took the battery out, still nothing.

Then, out of simple trial and error reasoning, I unplugged the monitor,
keyboard and mouse off the back, and tried it again (with the battery
out). Surprise, it works. Who the hell knows why. While it's booting, I
plug the mouse, keyboard, monitor, and battery back in (while the
computer was running - I know, bad thing to do), and set up the BIOS
again. All works fine now.

 

[Vanguard]

Thanks for all the replies, folks. I appreciate it.

I'm now writing this on the previously dead computer, so it's all ok
now. However, fixing it wasn't as straightforward as just putting in
the new battery.

When I slotted the new battery in, the computer was still dead. I
started wondering what the hell was going on - took out the battery,
cleared the CMOS, put the battery back in, still nothing.

Took the battery out, still nothing.

Then, out of simple trial and error reasoning, I unplugged the
monitor,
keyboard and mouse off the back, and tried it again (with the battery
out). Surprise, it works. Who the hell knows why. While it's booting,
I
plug the mouse, keyboard, monitor, and battery back in (while the
computer was running - I know, bad thing to do), and set up the BIOS
again. All works fine now.

USB mouse and USB keyboard, maybe?

 

[Trent©]

Right, no POST, no bios setup, the fans don't spin... it
doesn't turn the power supply on.

I wonder if this has anything to do with chassis intrusion...or even
C.O.P.?

It kinda seems like its a safety feature...some feature that needs a
good battery to function properly.

Also...in regard to someone else's post...sorry, lost track...

I had a couple of boards over the years where, if you pulled the
battery to replace it, the CMOS settings were immediately gone. No
5-10 second grace period.


Have a nice one...

Trent©

Budweiser: Helping ugly people have sex since 1876!

 

[kony]

I wonder if this has anything to do with chassis intrusion...or even
C.O.P.?

I wouldn't expect a chassis intrusion problem, the board has
been pulled out of case and though there may be a
header/jumper for that, it hasn't been used nor disturbed.

As for COP technology, I'm sure there's "something" like
that on the board, though just about any Athlon board after
a certain (earlier) data had this, I know not if there is a
specific implementation that would need a battery. It was
one of those fancier Asus boards that had the fan speed
control and voice POST reporter features, but again i don't
know if that matters. I've already shelved the board as it
was just a passing thought to try to get it working after
customer had returned it a while ago, now back to differnt
boards.

It kinda seems like its a safety feature...some feature that needs a
good battery to function properly.

Ther is also a different chip that appears to provide some
functions for an "I-Panel", which displays SMBUS data like
temp, fan RPM, voltage, along with more common features like
audio and sound (unrelated to the chip, common pin-headers
elsewhere on the board). I wonder if that chip inclusion is
the case but I can't very well reverse engineer the board to
get rid of it... at this point I'm apathetic about it beyond
needing to remember this particular board has the
battery-issue.

Also...in regard to someone else's post...sorry, lost track...

I had a couple of boards over the years where, if you pulled the
battery to replace it, the CMOS settings were immediately gone. No
5-10 second grace period.

I've noticed that too, wasn't trying to win a contest by
doing it fast yet had tried to get different battery in,
within 2-3 seconds and settings were still lost.