P5B-Deluxe won't start unless battery removed/replaced
Dave
forgive any duplication.
****************************
OK. First the admission. I got fooling around in the BIOS and
accidentally set the CPU voltage too high. The comp wouldn't boot
into BIOS after that so I had to reset the defaults by removing the
battery and clearing the CMOS RAM by moving the jumper. I moved the
jumper back to pins 1 and 2, replaced the battery, rebooted and all
was well... until I tried to restart the computer the next day and it
was dead. I checked and the board was getting power so just going
through the changes I had made led me to try taking out the battery
and re-inserting it. The computer fired right up and all was well but
now I have to go through the battery removal/replace every time I shut
the computer down.Tried reseting the jumper to no effect.
Any ideas? I don't know what hardware would be relevant. Q9550 CPU,
4 gigs Kingston RAM, pretty standard set-up.
Any help would be very much appreciated.
Dave
Paul
About the only thing I could suggest, is checking the battery voltage
to make sure the battery was good. A new battery measures somewhere
around 3.0V . Below 2.4V, it may not be guaranteed to hold BIOS settings.
There have been cases, where a completely flat battery prevents POST.
Looking at things like the couple reference schematics I've got, I
don't see why that happens. The battery 3.0V and 3VSB derived from
the +5VSB rail of the power supply, are ORed together, to make VBAT.
Since diodes are used, a flat battery cannot pull down the rail (diodes
are unidirectional). Now, a second connection may run from the battery
to a pin on the SuperI/O, but whether that has something to do with it,
I don't know. Maybe some genius tied the battery voltage into Power_Good
(the thing that tells the motherboard to release the Reset state). You
need Power_Good before a motherboard can start. A Power_Good signal
comes from the power supply, but is supplemented by Power_Good
signals from regulators on the motherboard as well. When they all say
"Go", reset is released and the processor fetches the first instruction.
If the battery is good, and inserted the right way in its socket,
I can't see a logical reason for what you're observing.
Are you certain the Clear_RTC jumper is in the correct position ?
I found this, just by clicking on a link I wasn't particularly looking
for :-) It shows the battery wiring for various Intel chipsets. Too
bad they didn't actually show the Clear_CMOS jumper, which goes across
one of the signals (should not be shorting the battery). The jumper should
be shorting RTCRST# to ground, and the 20K resistor prevents
draining the battery if the jumper is used for short periods of time.
The only reason I'm showing you this, is to demonstrate how
in theory, the state of the battery really shouldn't affect anything.
In the case of the ICH8, no bias is used from the battery for the
oscillator, so the ICH8 has a pretty clean implementation (fig 1-4).
http://www.intel.com/Assets/PDF/appnote/292276.pdf
Paul
Dave
Thanks very much for your help, Paul.
After refusing to start on 10-12 attempts yesterday, it fired right up
first try today and on several subsequent tries. Ghosts?
Thinking back, I might have shorted the battery when I removed it the
first time. Maybe an overnight rest allowed it to repair itself? The
jumper is in the proper position.
I'll replace the battery with a new one as long as I have the comp
apart. Thank you again for your generous help.
Dave
Ian D
"Dave" <sall...@cox.net> wrote in message
news:klnki512n0t1tehoe...@4ax.com...
That's probably what happened. Even an almost fully discharged
battery will have some recovery if allowed to sit for a few hours.
The electrolyte reaction has time to reconstitute. It may still have
some life left, but for the cost of a CR-2032 cell, (the most common
type), you might as well replace it.
That explains the battery, but doesn't explain why a low cell should
inhibit system startup.
Rob
"Ian D" <tau...@nowhereatall.com> wrote in message
news:y-Odna29UIzqSrfW...@giganews.com...
I don't know the P5B and agree with you and Paul that a flat battery
shouldn't prevent booting, but I know some mobos can monitor the
CMOS backup battery voltgage, so maybe custom BIOS code
could prevent it booting?
I've seen the system clock run fast (even in Windows) with a low
CMOS backup battery (on a A7N8X-E Deluxe), so strange things
can (and do) happen!
--
Rob
Dave
wrote:
Thanks for more input.
When I said the comp didn't start,I meant it didn't do anything at
all. No BIOS, no fans, no nothing. As if it were unplugged. I had
to check to see that the light on the MoBo was on (it was).
I did replace the battery and all's well now.
Dave
Ian D
Thanks for the update. I have a 3 year old P5B Dlx WiFi on my other
system, so I'll try to remember this for future reference, just in case,
assuming I still have the board when the battery dies.
One thing I don't understand, is why they bury these batteries in
the guts of laptops. I guess the manufacturers figure that by the
time the cells fail, their cheaply built laptops will be in the scrap heap,
or recycled, to be environmentally correct.
Paul
>> Thanks for more input.
>>
>> When I said the comp didn't start,I meant it didn't do anything at
>> all. No BIOS, no fans, no nothing. As if it were unplugged. I had
>> to check to see that the light on the MoBo was on (it was).
>>
>> I did replace the battery and all's well now.
>>
>> Dave
>
> Thanks for the update. I have a 3 year old P5B Dlx WiFi on my other
> system, so I'll try to remember this for future reference, just in case,
> assuming I still have the board when the battery dies.
>
> One thing I don't understand, is why they bury these batteries in
> the guts of laptops. I guess the manufacturers figure that by the
> time the cells fail, their cheaply built laptops will be in the scrap heap,
> or recycled, to be environmentally correct.
>
The other funny thing about the laptops, is the battery might not be
in a socket. Some use a concept like this, just to annoy.
http://www.smallbattery.company.org.uk/cr2032-3p.jpg
Paul