On Fri, 15 Jan 2010 18:48:02 -0500, "Wild_Bill"
wrote:

I believe that just because a battery cell test shows a full charge, doesn't
mean that the full charge can be gotten out of the cell (imbalance).

Checking ESR at random intervals could indicate if cells are unbalanced, I
think.

I don't understand why you believe ESR and imbalance have any
relationship whatsoever.

Charging at C/10 is said to balance cells, but how often? Having a means to
indicate (actually see) imbalance would give the user an idea of how often.

Yes, BUT most charging regimes for NiXX chemistries are current-based
and overcharge the cells somewhat. This provides SOC balancing.

I prefer to test capacitor ESR and other parameters before I install them,
as there are always inconsistencies in manufacturing.

Do whatever you feel important. I wouldn't bother.

No, I don't routinely retest them at regular intervals after installation,
but when circuit performance is poor, ESR testing is generally the first
step.

Whose first step?

The only reason I would want to zap nicad cells is to actually see if the
internal resistance is improved/lowered. I believe an ESR test would be a
good indication of any change that takes place.

"zapping" NiCd cells will *only* (and then temporarily) remove
conductive dendrites. If the cells have dendrites then they are
already on their last legs, and should be discarded before they cause
you grief. If they don't have dendrites, *zapping" will achieve
nothing, but hey, whatever floats your boat.

Otherwise, an ESR test after cells have been used a while, may be a better
indication of it's condition, and possibly it's reliability.

ESR (or more correctly, internal resistance) tests on a healthy cell
will indicate maybe the extent of irreversible deterioration. To that
extent it may provide some useful information.

I suspect that an unbalanced cell in a battery pack is bad for all the cells
in the pack, as it's likely to generate more heat, and possibly result in an
incomplete or overcharged state.

Yes, but ESR and imbalance are unrelated.

Your test fixture seems like a reasonable test, as a quick accept-or-reject
type of test.

I wouldn't want to automate tests, although it can be accomplished with the
right equipment and quite a bit of effort, by someone brighter than myself.
A dedicated analyzer/tester/reconditioner may be a better solution, but
costly.
I don't have to deal with so many batteries that a dedicated analyzer would
be practical.

I've seen a lot of inconsistent statements regarding battery maintenance and
performance.
One source that appears to be reliable is
http://www.buchmann.ca/default.asp

Pffft!

The medium sized UPS gear we used (3x 300kVA/40minutes) had an
integrated real-time battery condition monitoring system. This simply
applied a known pulsed load to the battery systems (at multiple tap
points) and monitored the corresponding voltage steps, which gave a
direct indication of cell resistance. If any segment of a battery
string reached a threshold value, that string was taken out of service
and cell-by-cell testing was undertaken, follwed by any required
remedial action.

You could set up a similar tester for manual operation (similar to
mike's description) and gain useful insight into internal resistance.