Tom wrote a comparative review of different battery technologies and
commented:
<< The memory effect is a myth. NiCds die from overcharging. >>
If you talk to battery vendors they may or may not agree
completely with this statement.
The situation is a little more complex than revealed in the
statement.
The memory effect is often defined as a (reversible) loss of capacity caused
by only
partial discharge of the cells before recharging them. For this effect to
occur the cells
need to be partially cycled repetitively before the cell "remembers" it's
capacity has
been reduced to the level it was previously discharged to. Some vendors
claim their
cells do not suffer from this effect at all but their competitiors cells do!
In reality it appears
that at least some vendors cells suffer from a "voltage depression" effect
that is
interpretted by the user as memory. This is most common in electronic
equipment
because the electronic cut -off voltage in the equipment is set too near the
normal
NiCd operating voltage. Thus when the voltage is depressed by repetitive
partial
cycling the equipment warns you the voltage is low much too soon and then
shuts off even
though you could have extracted a lot more energy if the equipment was
properly
designed to run at the sligtly lower "depressed voltage". The solution
commonly
recommended to always completely discharge the battery to eliminate memory
is often
as bad as the problem. This is particulary true for consumers who buy single
cells.
In a properly made battery pack the cells are carefully matched in capacity
by the manufacturer, this is generally not true for individual cells
(especially from different lots).
As remarked by others on the list deep discharging a battery usually
reverse charges
the lowest capacity cell in the pack. This is very bad for the cell and
usually causes
low resistance shorts that cannot be cleared by recharging. (You can usually
recover
these cells by applying a 100A pulse from a lead acid D cell for a second or
two but
this is usually only a short term fix). The higher the volage battery pack
the greater the chance of
damaging the pack in deep discharge since the likelihood of mismatch in cell
capacity is
much greater.
As a gross generlization high capacity NiCd cells (much greater than
say 500mAH for
an AA size cell) are significantly less reliable than the more common low
capacity cells.
The high capacity cells have very thin seperators to maximize active plate
material.
In repetitive overcharge the heating mentioned in the above posting causes
the seperators
to be damaged through cyclical mechanical stress from expansion. Some
vendors offer
high temperature cells with thick seperators and hence lower capacity and
these cells
are much more reliable but are generally not available to consumers directly.
A third common failure mode in NiCds is from being allowed to
completely self discharge
("left in cupboard failure"). This results from the fact that the cells are
sealed with a plastic seal.
When charged the cells run at a pressure higher than atmospheric. This helps
compress the seals and
helps prevent electrolyte leakage. When completely discharged the seals are
no longer held closed
by pressure and leak more easily as they age. The cell then grows tell tale
crystals around the
edge of the positive connection (where the seal is). Eventually enough
electrolyte is lost and the
cell fails. When not in use float charging at a very low rate (a few mA )
helps stop this failure mode.
To maximize cell life it is important to keep individual cells
together as a group to keep their
capacity matched when charging and discharging: this reduces the chance of
reverse charging damage
in deep discharge. Since most flashes don't have low voltage cutoff circuits
you should not experience
"memory/volage depression" problems. In fact to maximize battery cycle life
avoid completely
discharging them.
Sorry for this over long post.
Tim Hughes
Hi100@xxxxxxx
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|