I would strongly recommend you make sure you understand the relationship between voltage-mode and constant current mode when working with Li+ chargers. A common booboo happens when the current flow into a Li+ cell decreases as the charging cycle is nearing completion, and is not detected within some reasonable amount of time, which causes over-charging. That would be an (ahem) exothermic reaction phase. Think of the news stories about early Apple lapbooks catching fire, which was largely caused by a slow current sense switch in the charger. When you feel you switch has reached the inflection point of current load vs cell potential, then a current limiter needs to be switched-in to keep the cell in low-current trickle charge. If you want to roll your own controller for that would be fine, but it is important to realize that the charging curve would have to be recharacterized each time you change from one Li+ cell manufacturer to another. Or try to have some kind of self-servoing current sense amplifier somewhere on the load. Another thing that is fun to do (I mean that) is figuring out how to switch MOSFET's so that a load can remain connected to the cell even when the charging source (USB) is disconnected (or reconnected to charge up the cell). Make sure, in that case, that you have an output L-C filter on the load-side to squelch the switching transients from the charger. I had to design in all these things in my Blivit charger system, but did not have the real-estate to do it all in discrete elements. I opted for the simple way to do it with a Maxim part as the control element for the charger (MAX1874).
_______________________________________________ dorkbotpdx-blabber mailing list [email protected] http://music.columbia.edu/mailman/listinfo/dorkbotpdx-blabber
