Electronic – Lithium Ion charging problems – power outages

I think what they're trying to do here is 'trick' the phone's battery charging intelligences. Li-On batteries are very touchy and have somewhat complex charging strategies. It all boils down to determining something called State-Of-Charge (SOC). SOC is just a percentage in the end, but arriving at the SOC number relies on a large number of factors that are not always easy to read and sometimes must be indirectly inferred. For instance, let's assume that you have a cell phone with a Li-On battery that is 3.7V and 1000mAH. We'll start with it being fully charged, so we know SOC is 100%. As you use your device you're drawing current out of the battery and the battery's voltage will drop - eventually. By measuring the current and monitoring the voltage you can guess what the SOC is. One problem is that the voltage isn't terribly useful in determining SOC because it doesn't change very much until the battery is nearly empty - that is NOT something you want to do to a Li-On battery. So you're mainly relying on the current.

So your SOC is being estimated throughout the usage. It gets low - 50% maybe - so you plug it in to charge. While it's charging, it monitors the charge current and battery voltage to determine when SOC is 100% again. Only, due to errors in measurement it says that the charge is complete when SOC is actually only 95%. Now your phone thinks 95% is fully charged - and it remembers this for future reference because it doesn't want to over-charge the batteries (this is also very bad). So essentially it's trying to read when the battery is full by measuring what goes in and guessing where that puts the SOC based on past results.

The errors aren't large so during normal charge/discharge usage you won't notice a problem. But sometimes the errors can stack up and your phone thinks its fully charged when it has little or no charge - it goes straight from full to empty and due to the incorrect SOC calculation the phone won't try to charge the battery more because it doesn't want to damage it.

In these cases you have to reset the SOC state. I have a Droid Incredible 2 and I've done it by removing the battery and holding the power button for 30 seconds, then putting the battery in and charging the phone while its off. This always fixes the issue where the battery thinks its full but drops down to something like 10% very quickly and the issue where it thinks its at 10% but has much more charge left.

The strategy outlined in your post is obviously trying to recalibrate the SOC or trick the algorithm somehow. Having never developed a charger that relied on SOC I can't say whether it will work but it seems like a lot of effort for a questionable amount of benefit. If your battery is acting really funny try what I suggested first.

LiIon is usually charged at constant current until a max allowed voltage is reached and then is held at that voltage while current tails off under "control" of chemistry of battery until Ichg = k% of Imax where k% is chosen according to longevity or max energy concerns.50% or 25% of Imax gives longer life. 10% or 5% tail gives max capacity but lower life.

Lowering Vpedesatl by 0.1V greatly assists battery life.

Discharging to higher cutoff voltage aids cycle life.

LiIon also has calendar life and starts self destructing from day one so a lightly used battery still dies.

Best cycle life is achieved by stopping charge when Vpedestal is reached and systen changes from CC to Cv. By monitoring voltage this point can be observed. You could even do a "dumb" system that simply watched delta Vbat and declared constant V when delta fell to zero. Only slightly more than a comparator and an RC delay in one input would achieve that.(While Vin is ramping a delayed vin is lower. When Vin pedestals the delayed Vin almost catches up. An offset voltage is needed to allow comparator towork).

LiIon cells mechanically flex the cell as metallic Lithium is "plated" in and out of the cell*. Cycle life is in large part due to battery beating itself to death mechanically.(This is why LiFePO4 lasts much longer and has lower capacity - the material is held in an Olivine matrix that maintains constant shape as active material is moved in/out BUT it takes up some space. )

*Note: Bill Dubuque has suggested that this sentence would be better replaced with " 'LiIon cells mechanically flex the cell as Lithium ions are intercalated".
The distinction is a finer one than may be apparent. However, it is true to say that if you cut open a LiIon cell you would not usually find metallic Lithium in it. Bill notes that this makes primary Lithium cells, which do contain metallic Lithium, a greater fire hazard than LiIon cells.
If you charge a LiIon cell with excessive voltage metallic Lithium will be 'plated out' and "vent with flames" mode usually occurs at about the same time.

Charge to CV level as often and as soon as possible.

If charging all the way their "disconnect message" is a sign of bad ethos. They are probably trying to minimise the risk of fire without telling you.

For longest storage life (as opposed to long life in regular use) storing at a lower voltage than Vmax is in order. Probably at about 3.6 V and only about 30% state of charge. The various Mars Rovers use LiIon batteries and have a design life of about 8000 cycles - but charge to about 3.6 - 3.7 V maximum.
8000 / 365(~=) ~= 22 Terran years.