cmos clock malfunction repairable?

 > Flashing the BIOS won't fix the clock.
I was afreaid of that.

 > Do you "shut the box down" by a APM/ACPI powerdown or front-panel button
 > that leaves the 5 volt power on or by killing power externally? I believe
 > the clock uses the battery only when 5 vdc is not available on ATX boards.
I let the board shut itself down. I don't pull the power externally.

 > You might try an "external" battery ... one that has a three-pin connector
 > that plugs into m/b. It might end-run what might be a failure on the
 > motherboard.
Since the BIOS is holding all the other settings correctly, wouldn't that
suggest that the problem is not power related and thus something that an
external battery wouldn't solve?

 > It's also possible (and I think the most likely situation) that the
hardware
 > clock itself has failed. Windows runs with a software-based clock, which
 > is synchronized to the hardware clock at boot-up.
I was afraid of that, too. Do you happen to know if the clock is physically
part of the cmos chip itself (it's a Award BIOS)?


Thanks - Gil Theissen<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: cmos clock malfunction repairable? 

Gil Theissen wrote:

  > > Flashing the BIOS won't fix the clock.
 > I was afreaid of that.
 >
 >
  >>Do you "shut the box down" by a APM/ACPI powerdown or front-panel button
  >>that leaves the 5 volt power on or by killing power externally? I believe
  >>the clock uses the battery only when 5 vdc is not available on ATX boards.
 >
 > I let the board shut itself down. I don't pull the power externally.
 >
 >
  >>You might try an "external" battery ... one that has a three-pin connector
  >>that plugs into m/b. It might end-run what might be a failure on the
  >>motherboard.
 >
 > Since the BIOS is holding all the other settings correctly, wouldn't that
 > suggest that the problem is not power related and thus something that an
 > external battery wouldn't solve?

Not necessarily. The clock function needs a higher voltage to run than what
the CMOS needs to simply keep from losing it's data.

 >
 >
  >>It's also possible (and I think the most likely situation) that the
 >
 > hardware
 >
  >>clock itself has failed. Windows runs with a software-based clock, which
  >>is synchronized to the hardware clock at boot-up.
 >
 > I was afraid of that, too. Do you happen to know if the clock is physically
 > part of the cmos chip itself (it's a Award BIOS)?
 >
 >
 > Thanks - Gil Theissen
 >
 ><!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: cmos clock malfunction repairable? 

Gil Theissen wrote:

  > > Flashing the BIOS won't fix the clock.
 > I was afreaid of that.
 >
  >> Do you "shut the box down" by a APM/ACPI powerdown or front-panel button
  >> that leaves the 5 volt power on or by killing power externally? I
  >> believe the clock uses the battery only when 5 vdc is not available on
  >> ATX boards.
 > I let the board shut itself down. I don't pull the power externally.
 >
  >> You might try an "external" battery ... one that has a three-pin
  >> connector
  >> that plugs into m/b. It might end-run what might be a failure on the
  >> motherboard.
 > Since the BIOS is holding all the other settings correctly, wouldn't that
 > suggest that the problem is not power related and thus something that an
 > external battery wouldn't solve?
 >
  >> It's also possible (and I think the most likely situation) that the
 > hardware
  >> clock itself has failed. Windows runs with a software-based clock, which
  >> is synchronized to the hardware clock at boot-up.
 > I was afraid of that, too. Do you happen to know if the clock is
 > physically
 > part of the cmos chip itself (it's a Award BIOS)?
 >
 >
 > Thanks - Gil Theissen

An external battery often has a different circuit path through the
motherboard. Sometime you get lucky and do an end-run around a failed
component this way.

As another poster pointed out, voltage low enough to stop the hardware clock
is usually still enough to maintain the BIOS memory. If you haven't
already, you oughta check the voltage on the battery while it's connected
(and with the line power disconnected), that is, under load: one voltmeter
probe on the '+' side of the coin cell, the other to ground. It's worth a
try, even though you substituted another battery before.

I once saved a m/b that was leaky and drained a coin battery in less than a
month by adding a homebrewed external battery: two AA-cells last a couple
of years (!) It's been that way since 1997 and is in daily use at a small
business.

I don't know where the hardware clock lives on your m/b. I have two plastic
digital wristwatches that came with my Cheerios that I will send you if you
wish. When I buy cereal, I really-really don't want jewelry thrown in.

Roby<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: cmos clock malfunction repairable? 

OK, I'm game, but this is uncharted waters for me, so I have a couple of
questions. First, I assume I could use the coin sized extra 3 volt lithium
battery I have instead of 2 double AA's - at least for testing purposes -
right?
Second, do I simply affix one wire to the positive side and another to
the negative side of the battery?
Third, where on the board do I affix the other ends of the two wires?
Into the positive and groud prongs of plug for an unused fan? Somewhere
else?

 > When I buy cereal, I really-really don't want jewelry thrown in.
How come? And what kinda cereal you been buying anyway?

Thanks -
Gil Theissen<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: cmos clock malfunction repairable? 

Gil Theissen wrote:

 > OK, I'm game, but this is uncharted waters for me, so I have a couple of
 > questions. First, I assume I could use the coin sized extra 3 volt lithium
 > battery I have instead of 2 double AA's - at least for testing purposes -
 > right?

It's often easier to find a battery holder for AA's than a coin cell. And
in either case, a battery holder is better than cobbling together something
that might have poor contacts. AA's also have much higher current
capabilities. The BIOS normally draws only a few microamps, but a failing
component can drive that current way up.

You would be in a better position to fix the problem if you could measure
battery voltage under load; i.e., in the circuit it must supply. That's
the only way to know that your "new" battery is OK. You don't need a
laboratory instrument to do this. Just a simple multimeter will do ... and
will come in handy when other electrical mysteries arise.

 > Second, do I simply affix one wire to the positive side and another
 > to the negative side of the battery?

If you solder them, be sure to put on your flak jacket, helmet, goggles,
etc. A battery holder is a great idea. Or be sure your health insurance
is paid up.

 > Third, where on the board do I affix the other ends of the two wires?
 > Into the positive and ground prongs of plug for an unused fan?

AAIIIIIIEEEEE!!!!

 > Somewhere else?

Connect the battery *ONLY* to the external battery pins on the motherboard.
They're located near the on-board battery and identified as such on the
motherboard. Typically, a four-pin strip with one pin missing, connections
to the two end pins. +/- polarity marked on the board must be obeyed.

Roby<!-- ~MESSAGE_AFTER~ -->
 >> Stay informed about: cmos clock malfunction repairable?