Re: Battery problems

Actually, it's not a myth.  NiCd batteries do indeed have a memory. 
Some rechargers have a deep-cycle function to combat it.  They work by
placing a load on the battery until its voltage drops to a certain
point, then removing the load, waiting for a short period, and then
adding the load back.  The battery will then discharge a little more,
and the cycle is repeated until the battery can no longer provide useful
current.  This process retrains the battery.  I have had NiCd batteries
for my video camera that appeared to be dead until deep-cycled.  

You can easily see how this works if you have an old electric razor or
something with a NiCd battery that has been used regularly at least for
several months.  Simply use it until it won't run anymore, turn it off,
wait a minute, and then turn it back on.  You will find that it will run
for a very short period and then die.  You can repeat this quite a few
times before it will not run anymore.

Another thing you can do if you're really interested is to use the
device only while plugged in for about a year, and then see how long it
lasts without being plugged in.  Of course, if you want the test to be
more valid, you should probably have another razor that you always
discharge fully before recharging.

On Thu, 2003-06-05 at 14:06, John S. Andersen wrote:
> On 4 Jun 2003 at 23:47, J/ wrote:
> 
> > Either way, the
> > basics of it all runs down to -- if you make sure you fully discharge
> > your battery at least every other day, you should be fine. 
> 
> Tha'ts old advice, based on the long debated theory of NiCad Memory.
> Several battery companies stated that it never occurs in real life
> but has been seen in such things a sattelites etc.
> 
> But the advice is totally wrong with respect to Lith-Ion
> batterys.
> 
> The Li-ion batteries prefer a shallow discharge. Partial discharges 
> produce less wear than a full discharge and the capacity loss per 
> cycle is reduced. A periodic full discharge is not required because 
> the lithium-based battery has no memory. 
> 
> The links posted by SJB are excellent (although even they
> quote un-named "studies" supporting the NiCad myth.)
-- 
Lawrence MacIntyre <[email protected]>
High Performance Information Infrastructure Technology Group

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)

iD8DBQA+342yCNjP8rawCW4RApkkAJ945GbPjcaeAhIVscxJ9deApE4/4gCdHAlr
gK3leRQcL/hD60YJXQANiAY=
=6/30
-----END PGP SIGNATURE-----