The end user charger has no control of how a battery is drained, or how long it is left drained - so that is beyond its scope.
Charging then, breaks down into five phases.
- Battery detection and charge level determination - health tests (resistance, et al)
- Begin of battery charging - for a fully discharged battery (assuming testing shows it is at that level)
- Middle of charging
- Approaching level where battery-full might occur
- Final charging arrangements, when fullness is detected.
I suspect the charger can detect when "something" is placed between the positive and negative connectors of the charger. Even if this "thing" might just be a piece of metal. The circuit is completed in some way.
So check for things that might break the charger - reversed battery, a piece of metal, whatever comes to mind.
After it has determined a battery is what is closing the circuit...
Check for resistance.
Check approximate level of charge. In order to do this it must, as far as I know, place the battery under some load and determine voltage.
A fully emptied battery, with a passive voltage check on a VMM will show somewhere around 1.2 volts. So to detect empty - put it under load and the voltage will drop fairly quickly down to 1.1 and below - this would be categorized as phase one or empty battery.
A partially full battery can be determined by testing under load also. These generally stay around 1.20 to 1.27 volts, without much change.
A mostly full battery would show as being 1.27 to 1.39 or so, under load.
2. Begin of charging
I've seen several examples of other chargers amps patterns at beginning of charge. I'm not an electrician so I'm not certain of the effect of Pulse Width Modification (PWM) vs a fixed current, but all the charting I've seen seem to practically equate the two charging methods (purists prefer fixed it seems).
In any case to be as gentle as possible on an empty battery, and its chemistry, it would seem to make sense to start slowly with a lower current. Perhaps C/40 or C/50 for 5-10 minutes.
2a. Transition between charge voltages.
I propose that all changes of voltage occur in a gradual fashion. Which from a graphing perspective would appear to be a slope or curve. Maybe 2 minutes for transition from C/40 to middle charging current.
Example of Gradual increase of voltage (SkyRC MC3000 charger designer chose to take ~8 minutes for slope)
3. Middle of charging
Keep up the current, watch the voltage, check for loss of circuit (ie: someone removing battery during charge).
Test voltage while charging and also while no current is applied, between pulses (assuming PWM). Perhaps, even occasionally checking voltage under load (brief).
4. Approaching end of charge
Whichever charge ending criteria was chosen, which might be: specific voltage, inflection (tm Paul), a voltage plateau, "Peak Sensing tech.", a selected amount of mAh pushed into battery, -dv/dt detection...
At some point it is decided that the battery is practically full. So what some chargers do is lower the voltage to avoid as much heating of battery as possible. How much to lower is something that should be discussed. 75%, or multiple levels, 80% for a while, 60% for a bit, 40% etc. As in 2a above always transitioning. If the lowered voltage somehow interferes with "fullness" detection (as in missing a crossover for -dv/dt) some solution could be developed.
This charger halves the amps about 25% before the end of the charging. (Opus BT-C700) and example of tiered decrease of voltage (Opus BT-C3100)
5. Final charging arrangements
Transition the voltage down to much lower value. Depending on charge setting a low C/10 or C/20 top off charge could be applied...although usually not. If battery is the old non LSD, then trickle charge might be used (although how to tell the charger that? and also would a user want to?) - probably best not to trickle.
Example of a downward slope for voltage - Note: this was for a discharge but the principal is the same.