Funny battery values (nx6325)

[Ian Smith]

On Sun, 21 Mar 2010, Joerg Wunsch wrote:
 > As Dan Lukes wrote:
 > 
 > > Three or four cycles of charging / depth depleting may recalibrate
 > > the internals and reported values may become more appropriate.
 > 
 > I wonder whether I'll better do that outside the computer though.
 > After all, the Coulomb counter in the battery does not care much
 > whether it is attached to a computer querying it or not.

Hard to second guess what's going on in the battery chip, apart from 
measuring current in and out, over a shunt of some sort I suppose.
It's most likely got constants (brand/model/serial/design cap etc) 
in EPROM, no? and presumably its estimate of present state in RAM.

 > > You should not stop the discharging prematurely - it affect the
 > > results.

That's my understanding too.  I don't think that process is linear, ie 
you really do need to drain it to some set point before it resets, such 
that the next full charge can calculate a new 'last full' capacity, so 
several such cycles should bring that up nearer your 3Ah estimate.

 > I did want to reboot it into normal operation in order to allow the
 > nightly backup to run.  I simply ran out of time waiting for the
 > battery to drain completely.  Anyway, at a voltage level of < 9.0 V,
 > it wouldn't have lasted for another quarter of an hour.  (Worst case
 > cutoff level for a LiIon cell is around 2.7 V.)

The chip's estimation of its voltage is likely much more accurate than 
current, given the cost of accurate small-current shunts, so I expect 
you can trust it to know its cutoff / reset level.  I find running it 
down from the BIOS setup screen more comfortable, perhaps after taking 
it down near empty while you can load it enough to get on with the job.

BTW, I see my presumed-dead battery which was down to 8.6V is now back 
to 11.4V (still saying Critical Charging) after 2 days suspended on AC, 
after hearing its inverter's fan revving during periods overnight, ie 
taking some charge now and again.  Truly mysterious, _seeming_ to work 
differently suspended under 8.0 than 7.2-S .. but that's ridiculous!

 > Never trust an operating system you don't have sources for. ;-)

As in the BIOS, the EC, and the battery chip? :)

cheers, Ian