Using the TP4056 to charge 3S Li-ion pack

I was trying to make a 3 cell lipo pack charger using 3 of the TP4056, one for each cell. I encounter the same problem as (this gentleman). This is a comment from that forum question: "If you can charge batteries at different intervals then use relays to sequence one charge at a time." - @spicetraders Summary, commune ground, short.

The battery pack I'm using is a 3 cell pack, each cell being a 3.7v @ 2200mAh. Instead of buying a battery balance module, I'd like to try to make my own.

Because it is very inconvenient to use 3 different power sources or use a galvanic isolated power source which I read somewhere else, I'd rather use one power source.

3 TP4056 IC's (1 per cell) and a "supervisor" attiny85 IC + 3 MOSFET's.

The idea is to connect/charge 1 cell at a time for 1 second, then move on to the second cell for one second, and then, move to the third cell for another second. Repeat.

Now, in theory, would this work? if so, what would it be better, connect each cell to the charger for 1 second at a time, use milliseconds instead or minutes?

TP4056 Datasheet PDF

You need a balancing circuit really - anything else is awkward and clumsy. The are IC sets to help with this, but honestly its much much easier and cheaper to buy one.

And do you mean Li-ion pack? I would have thought you are talking about LiPo.

MarkT: You need a balancing circuit really - anything else is awkward and clumsy. The are IC sets to help with this, but honestly its much much easier and cheaper to buy one.

And do you mean Li-ion pack? I would have thought you are talking about LiPo.

I do need a balancing circuit, but I'd like to make my own for experience purposes instead of buying one. The problem is that I only have a power source of 5v @ 1A and the battery pack is a 3 cell lipo, which when fully charged is 12.6v. This is why I asked if I could charge each cell individually to 4.2v and move on to the next one. (By charging one cell at a time I mean charge cell one for 5 second and move to cell #2, charge 5 seconds and move to cell #3, charge 5 seconds and go back to cell #1.)

I was wondering what effect would on the batteries if they are constantly being charged in a sequence of about 5 seconds each vs 100 times a seconds pulses. Since a MCU and MOSFET's will be doing the switching.

Thanks!

Sure, you can do it your way, but the battery will only last for ~3-4 charges. If one cell is charged more than the next one, the higher cell will try to charge the lower cell. This kills a battery. You need to be able to measure and keep track of each cell.

Pringles: I do need a balancing circuit, but I'd like to make my own for experience purposes instead of buying one. The problem is that I only have a power source of 5v @ 1A and the battery pack is a 3 cell lipo, which when fully charged is 12.6v. This is why I asked if I could charge each cell individually to 4.2v and move on to the next one. (By charging one cell at a time I mean charge cell one for 5 second and move to cell #2, charge 5 seconds and move to cell #3, charge 5 seconds and go back to cell #1.)

I was wondering what effect would on the batteries if they are constantly being charged in a sequence of about 5 seconds each vs 100 times a seconds pulses. Since a MCU and MOSFET's will be doing the switching.

Thanks!

You can quite happily charge the cells individually.

If connected in series however you have to be sure that no individual cell gets below its minimum voltage. For series connected cells it is normal to have them of equal capacity of equal age and balanced charged so they have equal voltage . You can do this yourself with an arduino but it is generally cheaper to buy a chinese balance charger. 5V 1A you could charge individual cells but for a multi cell series pack you would need switches to get a higher voltage output..

Boardburner2: You can quite happily charge the cells individually.

Would the batteries have any negative effect if they are charged with a square wave? Because of my limited voltage, I can only charge one at a time. Of course I will not wait until 1 out of 3 cells is charges, because I'm planning in charging cell 1 for 500milliseconds and move to the next, and so on. Basically, batteries will be charging by pulses I think. Is going faster or slower best better efficiency?

This is how balanced chargers work.

Mine is on for about 20 secs then stops to measure the voltage off load then repeats.

I see no advantage for faster switching.

One other thing some batteries come with under voltage protection built in some do not.

You can add this yourself but its more expensive that way.

Downside is the circuit has s small drain on the cells so for long time storage they need monitoring periodicaly as the circuit will continue to draw current even after the shut off voltage is reached.

Lipos and LIion batteries should both be charged the same way, in domestic drills they generally are not and i suspect this to be the main cause of failure.

I have a recent aquired makita that balance charges and the batteries easily last twice as long as the cheap ones did.

codlink: Sure, you can do it your way, but the battery will only last for ~3-4 charges. If one cell is charged more than the next one, the higher cell will try to charge the lower cell. This kills a battery. You need to be able to measure and keep track of each cell.

How will the higher cell charge the lower cell of the battery when the battery isn't connected to any load? The higher cell is in series with the lower cell. Is this about leaking current within the battery?

Why not use three charging circuits wired like these shown attached? It might work charging each cell to 4.2V

The disadvantage of your concept charger is that each unit will need floating supplies.

That is they cannot share a common ground.

Parallel charging is done by some modellers and there are boards available specifically for that.

Personally i do not do it.