Archived from groups: alt.comp.periphs.mainboard,alt.comp.periphs.mainboard.asus (
More info?)
"Paul" <nospam@needed.com> wrote in message
news:nospam-0110041516510001@192.168.1.177...
> In article <58h7d.211$UP1.165@newsread1.news.pas.earthlink.net>, "knack"
> <zok9NOSPAM@hotmail.com> wrote:
>
> > After replacing the battery of the ASUS SP97-V mainboard with a newly
> > purchased battery from Wal-Mart, I continue to get a checksum error
during
> > POST. The error seems to occur after the power has been shutdown
completely.
> > However, if the PC is restarted immediately, then it will usually not
> > exhibit the problem.
> >
> > Is the power supply supposed to provide a recharging function?
> > Should I install another new battery?
> > Is there something wrong with the mainboard's CMOS?
>
> The power supply should never charge the CMOS coin cell, because
> it could explode. There is a dual diode, a device with three legs,
> usually near the CMOS coin cell, and it prevents the back flow
> of current into the battery. Either +5VSB or the coin cell, power
> the CMOS well on the Southbridge. The "well" is a section of silicon
> which is separately powered from the rest of the chip, and special
> design techniques have to be used for the logic signals inside
> the "well", to be accessable to the rest of the chip. For example,
> when the computer reads out the battery backed real time clock, that
> data crosses the boundary inside the Southbridge.
>
> The spec on a typical Southbridge is the "well" runs on from
> about 3V down to 2V. Allowing for some voltage drop across the
> diode, that means the battery should have about 2.4V minimum,
> to keep the clock running and the CMOS valid. (I'm assuming
> a 0.4V drop across the dual diode, at low current.)
>
> Have you tried refreshing the settings ? Write down all your
> custom settings in the BIOS, and then do a "Load Setup Defaults"
> from the Exit menu. See if that makes the CMOS happy. Clearing
> the CMOS is effectively what you did when you changed the battery,
> so I cannot see repeating that doing anything positive.
>
> About the only other thing that comes to mind, is maybe the dual
> diode is damaged. Asus leaves a ticking time bomb in many motherboards,
> in the form of the dual diode. If you do a "clear CMOS" operation,
> without unplugging the computer, or at least turning off the supply
> so the green LED stops glowing, the computer can be damaged when
> you use the CLRTC jumper. If the green LED is glowing, then the
> CLRTC jumper shorts +5VSB to ground, through the dual diode. The
> dual diode can actually get burnt to the point that the tiny lettering
> on it can no longer be read. Once the dual diode is damaged, then
> the CMOS well generally no longer get the power it needs. In that
> case, you would probably no longer have a working clock between
> power off events.
>
> HTH,
> Paul
Thanks, Paul, for all that very unusual knowledge. I saved it to a Notepad
file.
Whenever I open the cabinet I unplug the power cord. So, yes, the power
supply was deenergized when I changed the battery.
I've had my custom BIOS settings penciled into the mainboard's manual for
several years. I haven't made any changes there since 2000.
I forgot to mention that whenever the checksum error appears, my custom BIOS
settings become lost, so then I must spend 10 minutes reentering the BIOS
settings in order to boot the computer. It never occurred to me to load the
BIOS default settings as you suggested, because if I did my IRQ-jumpered ISA
slot modem will not work. WinNT4 is not plug and play, so I must configure
the modem and set the BIOS to dedicate IRQ-11 to ISA.
This morning the PC started up normally, but if I get the checksum error
again tomorrow, I'll load the BIOS defaults as an experiment.