charge light cycling

David B. Reiser (starnet!apple!ttown.apci.com!reiserdb)
Fri, 3 Dec 93 09:45:41 -0500

I believe that the cycling of the charge light while the computer remains
plugged in (even while not in use) is normal. If we're lucky, the charge light
being out means that no current is going from the charger to the battery, as
no good can come from overcharging :). As a long time mostly non-user of my
ultralite, I can say with some certainty that the main battery only lasts
about a week, even if the computer is not used at all. So, there is some
power draw even with the computer off (a fairly significant draw IMO).

Different types of batteries have different discharge curves--the voltage of
the battery over time while it supplies either constant current or constant
power to an external circuit. Nicads have an unusually flat discharge curve.
There is a slight drop at the beginning of discharge, but then the voltage
stays relatively constant until the capacity of the battery is reached. Then
you fall off a cliff and the battery is dead. Alkaline cells have more slope
to the curve -- the voltage drops more noticeably while there is significant
life left in the cell. This slope is how HP manages the battery circuit in
the HP95 palmtop. There is even a third party utility that reads the voltage
to something like hundredths of volts and predicts how much longer you can
run. I don't have the utility, but the buzz online suggests it works pretty
well. Even different batteries of the same type will have slightly different
discharge characteristics.

I know nothing about the internal reactions of an NMH battery, but I would
guess that its discharge curve is somewhere between a nicad and an alkaline
cell. If also must (well, almost must) have a slightly different open circuit
voltage (the voltage you would measure with a good VDM without the battery
being connected to anything and with the cell fully charged).

So, since the charger was designed for a combination of Molicel (a lithium
battery) for the main battery and nicad for the memory backup, there will
undoubtedly be some disparity between optimum charging behavior of the original
and of the NMH batteries. What I think is happening is that the NMH batteries
either start out at slightly lower voltage or they have a steeper discharge
curve than the Molicell/Nicad combination so that as the memory draws current
from the main battery, it reaches the voltage threshold to turn on the
charger after a relatively short time. Since not much power has been used up
by the time the charger goes on, the charger also boosts the battery back up
over the "turn-off" threshold fairly quickly.

If this senario holds, we'd better be very hopeful that NMH batteries don't
have a memory effect--as this cycling would really exaggerate the effect.

Dave