🚀 go-pugleaf

486 motherboard, with a 3v replacable Lithium battery rather than Ni-Cad
pack, newly installed, and the real-time-clock works erratically, if at
all.

If the computer is turned off for more than a few minutes, the RTC
stops, or slows down dramatically.  Sometimes it is not more than 4 or 5
hours slow in 8 or 9 hours.  Sometimes it seems to have stopped almost
when the main power is turned off.

So far I have seen no sign of CMOS settings being lost.

Anyone had this problem when it wasn't a dodgy battery?  I'm replacing
the battery when I get the chance -- probably mail-order from Maplin,
since I don't have the free time to get into town.

Is the BIOS date of july 1994 any guide to when the motherboard was
assembled?

According to the manual, and the jumper settings, the battery shouldn't
have been powering the CMOS until I switched to the internal battery
setting -- the jumper is in the 'clear CMOS' setting when the board
leaves the factory (and it was).

--
David G. Bell -- Farmer, SF Fan, Filker, Furry, and Punslinger..

Never criticise a farmer with your mouth full.