ABIT KD7A BIOS and CMOS Battery Anomaly

  • Thread starter Thread starter W. Watson
  • Start date Start date
W

W. Watson

I'm using the Subject line board with an Athlon XP 2000+ processor.
Recently, it all acted dead. Nothing would bring it alive. I checked the
cmos battery and the voltage was 2.8. I found if I left it out, the system
(W2k) would boot properly. Each time I rebooted though, I'd have to set the
date/time and processor speed. When I put in a new battery, it wouldn't
boot. After several boot attempts with the new battery, it finally began
booting into W2K properly. I hypothesize that keeping the battery in longer
allowed the capacitors in cmos to charge up enough to work fully.

In any case, while doing all the above, I had lots of problems with the
on/off and reset switch. Once things began to clear up, the problems seemed
to go away. I suspect cmos had something to do with that. Is that reasonable?

After several hours of just leaving W2k up, I shut it down. It seemed like
it was doing just fine when the screen went blank but power was still on. I
waited but it stayed that way. Finally I turned off the power. Is this
another bios related problem? Maybe more charging is required. Later today,
I'll try the same shutdown and observe carefully what does happen.
 
W. Watson said:
I'm using the Subject line board with an Athlon XP 2000+ processor.
Recently, it all acted dead. Nothing would bring it alive. I checked the
cmos battery and the voltage was 2.8. I found if I left it out, the
system (W2k) would boot properly. Each time I rebooted though, I'd have
to set the date/time and processor speed. When I put in a new battery,
it wouldn't boot. After several boot attempts with the new battery, it
finally began booting into W2K properly. I hypothesize that keeping the
battery in longer allowed the capacitors in cmos to charge up enough to
work fully.

In any case, while doing all the above, I had lots of problems with the
on/off and reset switch. Once things began to clear up, the problems
seemed to go away. I suspect cmos had something to do with that. Is that
reasonable?

After several hours of just leaving W2k up, I shut it down. It seemed
like it was doing just fine when the screen went blank but power was
still on. I waited but it stayed that way. Finally I turned off the
power. Is this another bios related problem? Maybe more charging is
required. Later today, I'll try the same shutdown and observe carefully
what does happen.
Click to expand...

Don't forget, that when +5VSB is available (i.e. switch is on, on the back
of the computer), that "charges the capacitors" *instantly* to their normal
voltage. The +5VSB is a "hard" path. So hard in fact, that it can burn
the steering diodes. The capacitors won't argue, when +5VSB is present.

The battery does have a 1K ohm series resistor, which might slow the charging
rate, but we're talking about short time intervals here. Not long periods
of time. Otherwise, people would be flummoxed every time they change
batteries, and we'd have many more reports of strange symptoms right
after a battery change. The last time I changed a battery, my system was
good to go on the next boot.

I suspect you are dealing with two problems at the same time, and the
battery is more or less a red herring. Something else much more critical
is going on, such as a power supply failure.

For example, if I look at some specs, a chipset might specify a Vbatt
minimum of 2.0V. There are Schottky steering diodes in the battery
path, to prevent charging of the CMOS battery, which adds about 0.4V
to the required minimum voltage. Which means the CMOS battery should
still function at 2.4V. A fresh battery is over 3V, and your battery
at 2.8V should still have worked. So I'm not convinced that the battery
is the whole story. You are probably right to replace it, because the
battery sounds like it is beginning to fail, but since your system
switched itself off, you have to look elsewhere for an explanation,
like the power supply. Or even an overheating processor, that has
shut off power to protect itself (not all S462 boards have thermal
protection, and the ones that do, use custom solutions to do it -
only some of those have programmable thresholds for example).

Paul
 
Back
Top