At 09:05 AM 3/22/2004 -0800, William Chops Westfield wrote:

>On Monday, Mar 22, 2004, at 08:51 US/Pacific, David VanHorn wrote:
>
>>>Constant current at 0.1C for 12-14 hours.  Ie a reasonably stable
>>>voltage supply and a resistor...
>>
>>Check the cell spec.  0.1C constant charge takes a special NIMH cell.
>>Sanyo's C series are rated for this.
>>Some NIMH cells are rated for ZERO constant (unterminated) charge
>>current, and they aren't kidding.
>Thus the 12-14 hour comment.  Which assumes dead or nearly dead cells
>to start with.  But it is the "minimum" circuit, and (perhaps sadly) is
>what you'll find in an awful lot of the consumer chargers on the market.

Yes, My phone is like this, resulting in batteries that signal "almost dead" a few minutes after being taken off charge.  Two 1C cycles of discharge and charge restore them to normal function. (Nicad)


>A lot of the remaining ("Fancy") chargers do something special for a
>fast charge, and then "drop down" to .1c or .05c for "float", whether
>it's a good idea or not.  Note that battery manufacturers aren't really
>motivated to provide algorithms that maximize cell lifetime :-(

Hmm.. I've done this in my Nomad printer.

Sanity checks are done on input voltage, battery voltage, and battery temperature. Battery condition is checked, and we launch into either conditioning charge (c/3) or high rate (c) on that result.

High rate checks the battery under load once a second, and monitors absolute and delta on voltage and temperature, as well as an absolute limit on time in high rate.

A really flat battery will go conditioning, high rate, topping, and low rate, with pauses between states to assure that the battery temperature is both below 40C, and descending, before continuing to the next state.

Abnormal conditions are handled as they occur, and the specific error in a given state vectors to the appropriate handler.

>Given their annoyingly fast self-discharge, it's very frustrating to
>not be able to float-charge NiMH without adverse effects...

If you use the Sanyo C series cells, then you can give them as much as C/10 constant without large problems.

If you can accurately measure the current that low, you can set them to zero or slightly positive current budget.  I use a pulsed trickle, letting them discharge a bit, then providing a short 1C charge pulse. This is adaptive, seeking a zero energy budget.

AVR Tiny-26 doing 500kHz PWM controlling the current every 10mS, and managing everything else, acting as a buck constant current source.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads