I'm working on a Microchip IC: MCP73871, a very small linear charger for lithium-based batteries. It has very nifty feature: a load-sharing system that gives priority to the load over the battery in order to avoid to continuously charge-discharge the battery.
The maximum charging current is 1A. How fast the battery is charged could be set on PIN13 "PROG1" by placing a specific resistor following this formula:
Ireg = 1000 / RPROG1
Ireg = mA
RPROG1 = Kohm
I tested a board featuring this IC by writing down current absorbed at the input and battery voltage over time (OCV) and it resulted the following charging profile. The battery is a li-ion 3.7V 6600mAh.
For 05 hours and 40 minutes the IC charged the battery at 1A. Then the battery voltage reached about 4,05V and current started to slow down.
Everything is correct: the charging profile is usually made of two stages:
- Constant current stage
- Constant voltage stage
Charging the battery at 1A when it is almost full could be dangerous and risky for battery life, this is clear. But the second stage is so slow. We are talking about 4 hours and it represents a big percentage of the full charging cycle time. And the average output power is so low.
Smartphones charge don't look to behave like this. They slow down a little bit over time but you won't never see them charging so slow. (In the graph you can see that the current slowed down constantly until a ridiculous 0,01A)
Is it possible that smartphone's batteries are declared 100% when they reach for example 4,05V, and not when they are 4,20V in order to get a shorter charging profile? What is the "state-of-the-art" technique adopted in the industry?