On Tue, 30 Sep 2003 09:39:08 -0400, "John Walton"
wrote:

The best thing to do is check the Texas Instruments (www.ti.com) AND Maxim
(www.maxim-ic.com) websites. They have application notes with complete
schematics and their chips incorporate the proper charging algorithm.

Indeed they do, but xpyttl's caution about "long" serires trings
remains just as valid a concern with decent controller chips. They
generally have no way to sense the individual cell SOC differences,
and therefore treat the whole string in terms of per-cell-average
voltage etc.

Additionally, any attempt at charge termination using temperature
sensing (absolute temp limit or rate of change) is VERY difficult to
implement on long strings simply because you need to sense at several
points and theses controllers are invariably designed for a single
temp sensing input. Been there, done that in a string 10, and we
gave up. The only other feasible option is to use MULTIPLE charge
controllers, each treating (and sensing) a small part of the total
string. This is often not a cost-effective approach. Sometimes,
redefining/rethinking the requirement leads to alternative approaches,
such as shorter strings feeding switchmode DC-DC boost converters.

"michael" wrote in message
news
Be careful, NiMH batteries are more difficult to charge than this. There
exists a charging algorithm for NiMH that is rather complex. You
basically charge it as you stated, but when the current draw
changes a certain way, you must detect it then charge at a different rate
for a
predetermined time. There are some IC controllers aware of this
algorithm, so your homebrew charger may be best to employ one of these
IC's...otherwise I suppose you could roll your own with a PIC or BASIC
STAMP and the right monitoring hardware. WARNING: YOU CANNOT
TRICKLE-CHARGE NIMH CELLS!

Good Luck.
--
Michael

To combat spam, I've put my address into the following gif. Sorry
for any inconvenience. http://mswindustries.us/email.gif