How fast does a 65 watt charger actually charge the battery cells?

Question

My laptop (A thinkpad t440s) has a 6-cell battery: three pairs of 18650 cells in series, for a total of about 11 V. My charger is 65 watts at 20 volts, which works out to 3.25 amps. My question is, do most laptops step this down linearly, so that it charges each cell with about (3.25 amps)/(2 cells in parallel) ~= 1.6 amps (which seems fairly reasonable for a high capacity 18650: "0.5C"), or is it more likely using a switching regulator to put the full power into the batteries, which would work out closer to 3 amps per cell?

I ask because I'm planning to replace the cells and want to ensure that the cells I buy are able to safely handle the charging current.

Answer 1

Let's make some assumptions:

0. the 6-cell battery has around 72 Wh which breaks down to single cells that have roughly 3.4 Ah each)
1. the charger is a switch-mode charger, so we must regard that the charger can push more amps to the battery as it get's on the input
2. he cells are all empty, they could be down to let's say 3.0 V that is 9 V for the series of 3 groups
3. they are just not yet considered deep-discharged so the charger does start in fast-charge mode (if they are lower than 3 V, according the the datasheet of samsung cells, IT equipment chargers should employ pre-charge with I translate to reduced current charge up to 3 V)
4. the computer is off - maximum power available to the battery
5. power supply is 65 W @ 20 V, with 90% efficiency that could give 58.5 W for charging, that means at 9 V no more current than 6.5 A is available in the first few minutes of charging
6. the efficiency may decrease a bit at lower load, but we ignore that
7. the cells are charged in a string 3S2P with balancing

Discussion

If we assume that the two parallel cells are in good shape they consume each half of the 6.5 A so 3.25 A in the beginning of the charging process.

If 3.6 V per cell are reached (pack at 10.8 V) the 58.5 W translate to 5.4 A max. so the current per cell is already below 2.7 A.

Justme's comment about larger power supplies is useful. The charging can and will be faster. But it is not said that it will charge with 120W, because the charging system will limit the current to what the original cells can handle (and the charging system itself also has a current limit!).

If it would use 120W for charging that would mean 12 A of current under the same assumptions as above, so 6 A per single cell.

There are also 170W power bricks available with the same jack so you can connect them to any mating Thinkpad - without harming the battery. (My P52 supply has 170 W (20 V and 8.5 A) and it can safely charge a Carbon X whose original supply has only 65 W (20 V and 3.25 A).

Conclusion

• The new cells should have the same rating than the original cells
• we see with the 65 W brick the cells are more or less charged with 1C
• we just can and do not know the max. charging rate limit of the original cells, but we can assume that with a larger power supply they may get charged with up to 2C
• If the original cell type is unknown, take new cells from a trustworthy supplier and manufacturer that allow continous charging at 2C rating. For 3.4 Ah cells this means 6.8 A max per cell, which means >12A for the string which is not possible from a 120W supply. Thus you are on the safe side.

Answer 2

Assume charging at 90% efficiency using a switching regulator. That is, the cells receive up to about 60W of power in total, with the laptop switched off and the batteries empty. That should be conservative enough. But leave yourself a factor of safety of 1.1, that is whatever currents you figure out must be multiplied by 1.1 before you use them to specify the cells.

Answer 3

That's easy enough to answer without looking closely: a linear regulator would turn almost half the energy into heat, and with the laptop switched off, no fan is running. To charge the batteries with so little heat production would require drawing only a rather small fraction of the available current, not matching the achievable charge times.

So you better plan for cells that are comfortable with the maximum available power, and it would be saner to not rely on the power being limited by the particular brick in use: better plan with the charging times achievable with a brick that is not maxed out.