The best battery site I've seen recently is at:
www.buchmann.ca
As a good starting place, read and learn everything
they have to offer.
1. All the cells in a used Li-Ion battery are the same age.
By the time one cell has failed, those that haven't yet failed
will fail sooner, rather than later. The Li-Ion chemistry has
an aging mechanism and even without use, the cells will all fail
within a few years. Replace them all at the same time.
2. Different manufacturers' cells have different characteristics,
despite the same specifications. The electronics in the battery pack
senses when one cell (in the series string) charges faster than another
and terminates the overall battery charge, so that the aged cell does
not become overcharged and thus become a fire/explosion hazard.
Ideally, one would use matched cells to make a battery. Normally,
this is not practical. But in any case, all cells _WILL_ age differently.
As the cells age differently, they will become unmatched.
3. It is not desirable to over-discharge Li-Ion cells, as then may
not re-charge properly. The discharge cut-off voltage is often given
as 3.0 volts. This is the parameter that any sensing circuit would use.
Therefore larger-capacity cells _SHOULD_ give the longer runtime battery
that you expect, IF all the cells are the same. Larger-capacity battery
packs are already available for many TP's.
My TP 600 uses three series pairs of cells, six total, for 3.6Vx3=10.8Volts.
The connection between pairs of cells is monitored for voltage imbalance,
in order to detect unequal/over charging and discharging.
End-of-discharge is determined when the weakest (any) cell drops below
the cutoff voltage (~3.0V) and charging is stopped when the weakest (any)
cell rises above the maximum peak charging voltage allowed, (~4.1V).
Personal comment:
I have not actually autopsied or investigated TP battery packs beyond the above,
but these principles are almost universally applied to Li-Ion batteries
with cells in series.
In high-reliability applications, elaborate and costly cell-bypass schemes
are sometimes created to prevent overcharging of weak cells and thus fully
charge the remaining cells in a series battery string (of up to 8 Li-Ion
cells in my experience).
There is also much discussion of this subject in the archives, which you
should read, understanding that there may be some inaccuracies and speculation.
Different battery packs should be expected to behave differently. IBM has
had several different model numbers for the battery for my TP 600, and even
one recall, IIRC.
Best of luck!
+++++++++++++++++++++++++++++++++
On Sun, 02Jan05, "Supir Kopaja" <kopaja@gmail.com> wrote, in part:
> ...
>First of all, I am sorry for my poor english. I plan to do a battery
>refill/recell business. I understand that most of the IBM batteries
>consist of 6 or 8 pcs Li-Ion batteries cells [and] that the
>specification are 3.6V or 3.7V and 1800mAH (min).
>
>My question are:
>
>1. Should we replace all the cells since most of the time
> not all cells defected?
>
>2. If we allowed combine brand new cells with current cells,
> can we combine them that have different brand but same spec?
> (for example: current cells are panasonic, we combine with
> brand-new cells brand sony, but both panasonic and sony
> same the same specification)
>
>3. Since batteries pack consist of 6 or 8pcs cells, in what
> way we can verify which cells are functioning and which are not?
> Is there available charger or tester to test these Li-Ion battery cells?
>
>4. Is that true that the battery SMBus (System Management Bus, the
> batteries "mainboard") control how long a battery could operate?
> I mean, the batteries can have a different performance
> even I replace all of the cells?
>
>I hope this is the right mailing-list to ask these questions, if not,
>may anybody suggest me the right mailing-list to do this?
>
>Many thanks.
>Kopaja.
>------------------------------
_______________________________________________
Thinkpad mailing list
Thinkpad@stderr.org
http://stderr.org/cgi-bin/mailman/listinfo/thinkpad
Received on Sun Jan 2 20:13:45 2005
This archive was generated by hypermail 2.1.8 : Fri May 26 2006 - 16:04:11 EDT