18650 Balancing

Projects General Guidance
1

Hello all,
I'm designing a small battery balancing circuit for 18650 lithium-ion cells. I'm using resistors to dissipate energy from un-balanced cells. I'm need help deciding which resistors to use to do this.
I know the voltage obviously, but deciding on the wattage and resistance required here.

Considering just one cell, with resistance in parallel, I'm using simple OHM's law to calculate the resistance I need. Which also depends on what current flows through, this will also decide the amount of energy that will dissipate. I need to know, if there is an imbalance, what is the percentage of imbalance. Here I'm using Samsung's ICR18650_25J cells out of the same batch. The number of cells in series is 4 (at the moment at least).

I've heard there is usually very little imbalance that occurs, when the cell is of good quality & produced from the same batch. I just need an idea as to how much current I'll have to flow in average through the resistors to achieve balance, example 200mA or 250mA or 500mA whatever is required.I need to know how much imbalance usually occurs if cells like this were used and other factors like temperature etc etc are kept optimal/ideal. :confused:
I know there are ready made ICs availiable for this, but doing this will help us scale it to massive number of cells if I learn from ground up.

Thanks peeps :slight_smile:

2

vikramnayak:
Hello all,
I'm designing a small battery balancing circuit for 18650 lithium-ion cells. I'm using resistors to dissipate energy from un-balanced cells. I'm need help deciding which resistors to use to do this.
I know the voltage obviously, but deciding on the wattage and resistance required here.

Considering just one cell, with resistance in parallel, I'm using simple OHM's law to calculate the resistance I need. Which also depends on what current flows through, this will also decide the amount of energy that will dissipate. I need to know, if there is an imbalance, what is the percentage of imbalance. Here I'm using Samsung's ICR18650_25J cells out of the same batch. The number of cells in series is 4 (at the moment at least).

I've heard there is usually very little imbalance that occurs, when the cell is of good quality & produced from the same batch. I just need an idea as to how much current I'll have to flow in average through the resistors to achieve balance, example 200mA or 250mA or 500mA whatever is required.I need to know how much imbalance usually occurs if cells like this were used and other factors like temperature etc etc are kept optimal/ideal. :confused:
I know there are ready made ICs availiable for this, but doing this will help us scale it to massive number of cells if I learn from ground up.

Thanks peeps :slight_smile:

I have read that some new batteries are shipped from Mfr with, as low as, 50% SOC initially.
But I have never read anything that specified the maximum SOC unbalance.

I don't think there is any specific "Typical % Unbalanced" value.
Maybe, contact the Manufacturer ?
I thought, it was recommended to Charge to 100%, before 1st use.

WORST CASE, in a pack, after much use:
One fully charged battery vs one fully discharged battery
( Battery mAh ) / ( Max Balance Time ) = Balancing milliamps

You could use any balancing milliamps that you want to use ( within limits ) because
the lower the milliamps, then the longer the time balancing
the higher the milliamps, then the shorter the time balancing.

Do you have a maximum time limit to achieve balance?

Imbalance will re-occur as you discharge and recharge the battery pack.

NOTE #1:
Cells with a load will have a lower voltage than Cells with no-load.

NOTE #2:
There is a difference between Top Balancing and Bottom Balancing

3

Balancing relates only to the charge cycle. The maximum current is then the current that the charger can provide.

You don't use resistors to dissipate a fixed current, because you don't know the current. You use shunt voltage regulator circuits which pass whatever proportion of the charging current is necessary to prevent the voltage rising above the set-point.

Someone tell me if I have this wrong?