_dee wrote:
> I've got a P4C800-E that refused to power up again after being powered
> down. Was working OK except for the clock going way off. Found that
> the lithium battery was dead, but I'm not sure why. Perhaps something
> in that part of the circuit was drawing way too much current.
>
> So, I had powered the system down to check and replace the battery.
> Thought maybe CMOS config was getting munged. Unfortunately, the new
> battery did not help, and that's the last it has powered up.
>
> Any recommendations for debugging?
Have you been playing with the Clear CMOS jumper lately ?
While the 875 chipset has a proper CMOS reset signal on it,
it could be that Asus uses a traditional solution involving
interrupting power flow to the CMOS well. As a safety precaution,
you should unplug the computer, before using the Clear CMOS jumper.
In some cases, an ORing diode gets burned, if +5VSB (standby power)
is still available.
The problem with the P4C800-E Deluxe, is that the CMOS battery is all
the way across the board from the Southbridge. I cannot even
guess where the ORing diode might be physically located. Normally,
the battery, Southbridge, ORing diode, and Clear CMOS jumper
would be close to one another. But a motherboard designer
can throw them where ever they have room, so it isn't essential
for them to be positioned together.
You can download a schematic for an 875 motherboard, here.
http://www.intel.com/design/chipsets...ics/252812.htm
If you open the 25281202.pdf document at PDF page 82, the ORing diode
(a dual diode of type BAT54C), can be seen in the upper left hand corner.
If the diode is burned, it could go open circuit, in which case there
may be no power to the CMOS well. And that could prevent startup, at
least on some motherboards. The BAT54C would resemble a three legged
transistor in a black plastic package. They print an identifying
code on the top, that bears no relationship to "BAT54C", so it isn't
like you'll find "BAT54C" printed on it. I have three older Asus
motherboards here, where that part has "K45" printed on it, and
I need a magnifying glass to read it. One poster, who damaged an
ORing diode on another motherboard, was no longer able to read the
code, due to the device being burned.
So, if you used the Clear CMOS jumper recently, and the power supply
was still switched on at the back of the computer, it is conceivable
that the ORing diode got damaged. Visually, there is no way for
me to determine what kind of circuit design Asus used, so this
remains a wild guess. Intel made a provision to avoid all this
"burning" nonsense, but it is up to the motherboard designer
as to whether they want to use it or not.
If you go to PDF page 37 of that document, in the lower right hand
corner, jumper J9H1 demonstrates a "safe" way to implement a clear
CMOS jumper on an 875 motherboard. But it is just as likely, that
Asus used the traditional "burn-able" design instead.
Which is why I asked my original question.
Can it be repaired ? Sure, if you can find that tiny three-legged
thing, and verify the voltages on it. One poster I helped, soldered
two ordinary diodes in place of the BAT54C, and apparently that
worked. So in at least one case, someone got to keep their motherboard.
Paul