---
title: "Why does a LiPo pack need balancing?"
topic: "Battery balance"
answer: "Because a plain charger watches only the total. On a 3S pack with one cell faded to 90%, a plain charge stops at 12.60 V with that cell at 4.25 V, past its 4.20 V limit. A balance charge watches every cell and stops each one at 4.20 V."
url: "https://nearfieldnotes.com/notes/battery-balance/"
site: "Near Field Notes"
author: "Michael Walker"
note: 15
family: "Power"
published: "2026-11-05"
revised: "2026-11-05"
revision: 1
video: "https://nearfieldnotes.com/videos/battery-balance/"
---

Note 15 · [Power](https://nearfieldnotes.com/families/power/) · Rev 1 · [Watch the video · 2:26](https://nearfieldnotes.com/videos/battery-balance/)

Battery balance

# Why does a LiPo pack need balancing?

**Short answer:** Because a plain charger watches only the total. On a 3S pack with one cell faded to 90%, a plain charge stops at 12.60 V with that cell at 4.25 V, past its 4.20 V limit. A balance charge watches every cell and stops each one at 4.20 V.

The figure charges the same pack twice at 2.2 A. First plain, through its main plug alone. Then with its balance lead plugged in, so the charger can bleed 300 mA off the cell that leads.

**Note 15’s opening figure: A 3S pack, charging.** Plain charge, then balance charge

![Note 15’s opening figure: A 3S pack, charging](https://nearfieldnotes.com/notes/battery-balance/fig-opening.png)

*What the figure shows:* A 3S LiPo pack opened into its three cells, with each cell's voltage beside it and the pack's total, over a chart of each cell's voltage while it charges. The plain charge takes 29 min and ends with the pack at 12.60 V: the two healthy cells at 4.17 V and the weak cell at 4.25 V, over the 4.20 V limit. The balance charge bleeds the weak cell through a resistor and ends with every cell at 4.20 V.

Cell 3, at the top, holds 1980 mAh. The other two hold 2200 mAh. All three start at 3.80 V. On the balance pass, the dashed line is cell 3 on the plain pass.

**The numbers**

| Charge | Took | Cells 1 and 2 | Cell 3, weak | Pack | Lost as heat |
|---|---|---|---|---|---|
| Plain | 29 min | 4.17 V | 4.25 V | 12.60 V | none |
| Balance, 300 mA bleed | 29 min | 4.20 V | 4.20 V | 12.60 V | 0.41 Wh |

## A 3S pack and its balance lead

*Two plugs, two jobs*

A 3S pack is three cells in series. Two thick leads carry the charge and the load through the main plug, an XT60 here. Across them, a charger sees only the total.

Four thin wires run to the balance plug, a 4-pin JST-XH. They reach both ends of the pack and both joints between cells, so a charger can read every cell.

**Fig. 15: The pack.**

![Fig. 15: The pack](https://nearfieldnotes.com/notes/battery-balance/fig-15.png)

*What the figure shows:* Figure 15: a 3S 2200 mAh LiPo pack in blue shrink wrap, 106 by 34 by 26 mm. Two thick leads, red and black, run to a yellow XT60 main plug. Four thin wires run to a small white balance plug.

A 2200 mAh 3S pack, 106 × 34 × 26 mm. Under the wrap are three flat pouch cells, stacked as in the opening figure.

**The numbers**

| Plug | Wires | Carries |
|---|---|---|
| Main, XT60 | 2 thick | The charge and the load current, across the whole pack |
| Balance, 4-pin JST-XH | 4 thin | Each junction between cells, for reading and balancing each cell |

Each balance pin sits one cell above the last. With every cell full, they read 0, 4.2, 8.4 and 12.6 V.

A charger works out each cell from the difference between two neighboring pins.

**Fig. 15b: The pack as a circuit.**

![Fig. 15b: The pack as a circuit](https://nearfieldnotes.com/notes/battery-balance/fig-15b.png)

*What the figure shows:* Figure 15b: the pack as a circuit. Three cells in series. Thick main leads run from the top and bottom of the string to the XT60's + and − pins. Thin wires run from all four junctions to the balance plug's pins 1 to 4, which sit at 0, 4.2, 8.4 and 12.6 V when every cell is full.

The thick leads carry the pack’s current to the XT60. The thin wires to the balance plug only read and balance.

**The numbers**

| Balance pin | Joins | At full |
|---|---|---|
| 1 | The pack’s − | 0.0 V |
| 2 | Cell 1 + and cell 2 − | 4.2 V |
| 3 | Cell 2 + and cell 3 − | 8.4 V |
| 4 | The pack’s + | 12.6 V |

## S and P

*Cells in series, cells side by side*

S counts cells in series. Each adds its voltage, and the same current flows through all of them.

P counts cells side by side in each group. They share the current and add their capacity. Joined, they always sit at the same voltage, so they balance as one.

The balance plug has a pin for each end and each joint, so a 3S pack takes 4 pins, whatever its P. Hobby makers’ guides draw 3S2P as two 3S strings side by side. Large packs join each group first, as drawn here.

**Fig. 15c: Three wirings.**

![Fig. 15c: Three wirings](https://nearfieldnotes.com/notes/battery-balance/fig-15c.png)

*What the figure shows:* Figure 15c: three ways to wire 2200 mAh cells. 2S1P: two cells in series, 8.4 V full, 2200 mAh, a 3-pin balance plug. 3S1P: three in series, 12.6 V, 2200 mAh, 4 pins. 3S2P: three groups in series, each two cells side by side, 12.6 V, 4400 mAh, 4 pins. Each open circle is a balance pin.

2200 mAh cells. Each open circle is a balance pin.

**The numbers**

| Wiring | Cells | Full | Capacity | Balance pins |
|---|---|---|---|---|
| 2S1P | 2 | 8.4 V | 2200 mAh | 3 |
| 3S1P | 3 | 12.6 V | 2200 mAh | 4 |
| 3S2P | 6 | 12.6 V | 4400 mAh | 4 |

## A plain charge

*What the charger sees*

A plain charge leaves the balance lead out and works on the total. It pushes 2.2 A until the pack reads 12.60 V, then holds 12.60 V while the current [tapers off](https://nearfieldnotes.com/glossary/#g-cc-cv). It stops at 220 mA, after 29 min.

`V_pack = V_1 + V_2 + V_3`

where V_pack: pack voltage (V); V_1: cell 1’s voltage (V); V_2: cell 2’s voltage (V); V_3: cell 3’s voltage (V).

`4.173 V + 4.173 V + 4.254 V = 12.600 V`

The total comes out right. The cells don’t. Cell 3 holds 10% less, so the same current fills it faster. It passes 4.20 V after 26 min and ends at 4.25 V, 82 mV above the others.

One LiPo maker gives 4.25 V as the absolute maximum, never to be passed. A cell only 5% short still ends at 4.23 V.

Cells from one batch still differ in capacity, resistance and self-discharge, and TI notes that the gaps grow with use. The weak cell is the one that ends high, and each overcharge weakens it further.

**Fig. 15d: A plain charge.**

![Fig. 15d: A plain charge](https://nearfieldnotes.com/notes/battery-balance/fig-15d.png)

*What the figure shows:* Figure 15d: a plain charge of the 3S pack at 2.2 A. The pack's voltage climbs to the charger's 12.60 V limit and holds there. The charge ends at 29 min. Underneath, the two healthy cells end at 4.17 V and the weak cell at 4.25 V, crossing the 4.20 V line after about 26 minutes.

Blue is cells 1 and 2, which charge as one. Orange is cell 3.

**The numbers**

| Minute | Pack, V | Cells 1 and 2, V | Cell 3, V |
|---|---|---|---|
| 0 | 11.426 | 3.809 | 3.809 |
| 5 | 11.560 | 3.852 | 3.857 |
| 10 | 11.748 | 3.910 | 3.927 |
| 15 | 11.969 | 3.982 | 4.006 |
| 20 | 12.206 | 4.057 | 4.092 |
| 25 | 12.462 | 4.138 | 4.187 |
| 28.7, end | 12.600 | 4.173 | 4.254 |

## What 4.25 V costs

*Cycle life against charge voltage*

Nobody has published cycle life at 4.25 V. The nearest data is a 2020 study by Juarez-Robles and others. They cycled 5 Ah LiCoO₂ pouch cells at 1C to seven charge voltages, until each held only 4 Ah.

**Fig. 15e: Charged higher, worn out sooner.**

![Fig. 15e: Charged higher, worn out sooner](https://nearfieldnotes.com/notes/battery-balance/fig-15e.png)

*What the figure shows:* Figure 15e: commercial 5 Ah LiCoO2 pouch cells cycled at 1C to seven charge voltages, from Juarez-Robles and others, 2020. Charged to 4.2 V: 5.05 Ah at first, 290 cycles until it held 4 Ah. Charged to 4.3 V: 5.46 Ah at first, 209 cycles until it held 4 Ah. Charged to 4.4 V: 5.83 Ah at first, 198 cycles until it held 4 Ah. Charged to 4.5 V: 6.48 Ah at first, 134 cycles until it held 4 Ah. Charged to 4.6 V: 6.97 Ah at first, 28 cycles until it held 4 Ah. Charged to 4.7 V: 8.77 Ah at first, 29 cycles until it held 4 Ah. Charged to 4.8 V: 9.05 Ah at first, 12 cycles until it held 4 Ah.

Their Table I. Charged higher, a cell holds more at first and lasts fewer cycles. Orange is 4.3 V, the first row above this pack’s 4.25 V.

**The numbers**

| Charged to | Capacity at first | Cycles to 4 Ah |
|---|---|---|
| 4.2 V | 5.05 Ah | 290 |
| 4.3 V | 5.46 Ah | 209 |
| 4.4 V | 5.83 Ah | 198 |
| 4.5 V | 6.48 Ah | 134 |
| 4.6 V | 6.97 Ah | 28 |
| 4.7 V | 8.77 Ah | 29 |
| 4.8 V | 9.05 Ah | 12 |

Overcharge

### A little more capacity, a lot fewer cycles

**Field marks.** A cell that reads high at the end of every charge and low at the end of every flight. Past 4.5 V, a pouch that has puffed up.

**Look-alike.** The rule that each 0.1 V halves cycle life. Battery University gives it with no source. Measured, 4.3 V instead of 4.2 V cost 28%: 209 cycles against 290.

**Range.** 4.3 V gave 8% more capacity at first. Past 4.4 V, lithium metal built up on the graphite. Past 4.5 V the cells swelled, and the one charged to 4.8 V almost doubled in volume. A fresh cell pushed slowly failed at 5.1 V.

**Habitat.** Any series pack charged on its total alone: hobby packs on a plain charge, and packs with no cell monitor.

## In flight, it empties first

*10 A from the plain-charged pack*

The weak cell runs out first, too. The same 10 A comes out of every cell, and it has the least to give.

A flight controller running Betaflight reads only the pack. Its docs’ example settings warn at 3.4 V a cell and alarm at 3.3 V. On a 3S pack, that’s 10.2 and 9.9 V.

Here cell 3 is empty after 12 min, while the pack still reads 10.0 V. The warning came 5 s before. The alarm never sounded.

**Fig. 15f: The same pack in flight.**

![Fig. 15f: The same pack in flight](https://nearfieldnotes.com/notes/battery-balance/fig-15f.png)

*What the figure shows:* Figure 15f: the plain-charged pack discharged at 10 A. The weak cell falls away near the end and is empty after 12.2 minutes, reading 3.02 V under load, while the pack still reads 10.0 V, above a 9.9 V alarm. The healthy cells are at 3.50 V.

Voltages under the 10 A load. Cell 3 is empty where its line ends, at 3.02 V.

**The numbers**

| Minute | Pack, V | Cells 1 and 2, V | Cell 3, V |
|---|---|---|---|
| 0 | 12.478 | 4.132 | 4.214 |
| 2 | 11.917 | 3.962 | 3.994 |
| 4 | 11.504 | 3.830 | 3.845 |
| 6 | 11.228 | 3.743 | 3.743 |
| 8 | 11.081 | 3.695 | 3.691 |
| 10 | 10.925 | 3.647 | 3.630 |
| 12.2, empty | 10.018 | 3.501 | 3.016 |

## Balance charging

*Passive: burn the extra*

A balance charge plugs in both leads. It charges through the main plug and reads every cell through the balance plug.

When one cell leads, the charger switches a resistor across it. Part of the charging current flows around that cell as heat, and the others catch up.

`I_bleed = V_cell/R`

where I_bleed: bleed current (A); V_cell: voltage of the cell being bled (V); R: bleed resistor (Ω).

`= (4.20 V)/(14 Ω) = 300 mA and 1.26 W of heat`

**Fig. 15g: A bleed across each cell.**

![Fig. 15g: A bleed across each cell](https://nearfieldnotes.com/notes/battery-balance/fig-15g.png)

*What the figure shows:* Figure 15g: passive balancing. Each of the three cells has a resistor and a switch, a MOSFET, across it. The switch across cell 3, the fullest, is on, so part of the charging current flows through its resistor and turns into heat.

Cell 3 leads, so its switch is on. The lit loop is its bleed current.

**The numbers**

| Part | Value |
|---|---|
| Bleed resistor, each cell | 14 Ω |
| Bleed current at 4.20 V | 300 mA |
| Heat while on | 1.26 W |
| Switch | A MOSFET, on while its cell leads |

Passive balancing

### Simple and cheap, and the extra becomes heat

**Field marks.** A resistor and a small MOSFET across each cell, in a charger or on a BMS board. Hobby chargers bleed 0.3 to 1 A a cell.

**Look-alike.** Active balancing, on the same balance lead. It moves charge between cells instead of burning it.

**Range.** A 300 mA bleed balances this pack in 29 min, about as long as the plain charge, and turns 0.41 Wh into heat. TI’s BQ76920 bleeds up to 50 mA, and never two neighbors at once. On this pack it takes 3 hours.

**Habitat.** Hobby chargers, and most commercial packs. Even the Tesla Model S balances with resistors: 39.5 Ω across each group of 74 cells, about 94 mA.

All three balancers end with every cell at 4.20 V. They differ in time and heat.

A bleed fixes the overcharge. It leaves the run time alone: cell 3 still holds 10% less, and the pack still stops when cell 3 does.

**Fig. 15h: Three balancers, one pack.**

![Fig. 15h: Three balancers, one pack](https://nearfieldnotes.com/notes/battery-balance/fig-15h.png)

*What the figure shows:* Figure 15h: the same 3S pack, one cell at 90%, balance-charged three ways. Charger, 300 mA: full and balanced in 29 minutes, 0.412 Wh lost as heat. BQ76920, 50 mA: full and balanced in 183 minutes, 0.435 Wh lost as heat. Active, inductor: full and balanced in 29 minutes, 0.035 Wh lost as heat.

Each charge at 2.2 A from 3.80 V, as in Fig. 15d, until every cell is within 5 mV of 4.20 V.

**The numbers**

| Balancer | Current | Full and balanced in | Lost as heat |
|---|---|---|---|
| Charger’s bleed | 300 mA | 29 min | 0.412 Wh |
| BQ76920 bleed | 50 mA, 70% of the time | 183 min | 0.435 Wh |
| Inductor, active | 825 mA out of the high cell | 29 min | 0.035 Wh |

## Balance near the top

*Where a gap shows*

A balancer goes by voltage, and a cell’s voltage barely moves through the middle of its charge. A gap between cells hides there. A [LiFe](https://nearfieldnotes.com/glossary/#g-lifepo4) hides it almost to the end, as [Note 18](https://nearfieldnotes.com/notes/lipo-vs-life/) shows.

The same 5% gap in charge reads as 20 mV at 3.70 V and 70 mV at 4.15 V. Near full, the gap is easy to see.

**Fig. 15i: A cell’s voltage at rest.**

![Fig. 15i: A cell’s voltage at rest](https://nearfieldnotes.com/notes/battery-balance/fig-15i.png)

*What the figure shows:* Figure 15i: the voltage of a LiPo cell at rest against its charge, from 3.0 V empty to 4.2 V full, flat through the middle and steep near the top. The same 5.1% gap in charge makes a 20 mV gap at 3.70 V and a 70 mV gap at 4.15 V.

Each pair of dots is 5.1% of a full cell apart, the gap the plain charge leaves between cell 3 and the others. Past 100%, a cell is over 4.20 V.

**The numbers**

| Cell at | Charge | 5.1% more charge | Voltage gap |
|---|---|---|---|
| 3.70 V | 20.9% | 3.720 V | 20 mV |
| 3.80 V | 52.0% | 3.822 V | 22 mV |
| 3.90 V | 69.0% | 3.945 V | 45 mV |
| 4.00 V | 80.6% | 4.048 V | 48 mV |
| 4.10 V | 91.2% | 4.152 V | 52 mV |
| 4.15 V | 96.1% | 4.220 V | 70 mV |

## Active balancing

*Move the charge instead*

An active balancer moves charge out of a full cell and into an empty one. With an inductor, it takes three steps, 50,000 times a second.

**Fig. 15j: An inductor between two cells.**

![Fig. 15j: An inductor between two cells](https://nearfieldnotes.com/notes/battery-balance/fig-15j.png)

*What the figure shows:* Figure 15j: an inductor balancer between two cells, at each step of its 20 µs cycle, and the inductor's current over three cycles. 1: the switch closes across the high cell and the current through L climbs to 3.6 A in 9 µs. 2: the switch opens and L's current runs on through the diode into the low cell, falling to zero at 17.3 µs. 3: nothing flows until the next cycle. On average 0.83 A leaves the high cell and 0.74 A reaches the low one.

10 µH switched at 50 kHz, between cells at 4.10 and 4.00 V. The high cell is on top. The lit loop is where the current flows. One settled cycle of it, repeated.

**The numbers**

| Step | From | To | Conducting | Inductor current |
|---|---|---|---|---|
| 1, fill L | 0 µs | 9 µs | The switch, from the high cell | 0 to 3.63 A |
| 2, empty L | 9 µs | 17.3 µs | The diode, into the low cell | 3.63 A to 0 |
| 3, wait | 17.3 µs | 20 µs | Nothing | 0 |

Less arrives than leaves. The rest heats the switch, the diode and the inductor’s copper:

`η = V_low I_in/V_high I_out`

where η: share of the energy that arrives (%); V_low: low cell’s voltage (V); I_in: average current into the low cell (A); V_high: high cell’s voltage (V); I_out: average current out of the high cell (A).

`= (4.00 V × 0.741 A)/(4.10 V × 0.825 A) = 88%`

A capacitor can carry charge too. It connects across the higher cell, then across the lower one, and takes a little charge down each time:

`I = C · ΔV · f`

where I: average balancing current (A); C: capacitance (F); ΔV: voltage gap between the two cells (V); f: switching frequency (Hz).

`= 100 µF × 100 mV × 10 kHz = 100 mA against 97 mA simulated`

The current follows the gap: 19 mA at 20 mV, 4.8 mA at 5 mV. It slows as the cells close in.

**Fig. 15k: A capacitor between two cells.**

![Fig. 15k: A capacitor between two cells](https://nearfieldnotes.com/notes/battery-balance/fig-15k.png)

*What the figure shows:* Figure 15k: a switched-capacitor balancer's current against the gap between its two cells, simulated at three gaps: 4.8 mA at 5 mV, 19 mA at 20 mV, 97 mA at 100 mV. A dashed line shows capacitance times gap times switching frequency.

100 µF switched at 10 kHz. The dots are the simulated currents, the dashed line *C*·Δ*V*·*f*.

**The numbers**

| Gap | Simulated | C × gap × f |
|---|---|---|
| 5 mV | 4.8 mA | 5.0 mA |
| 20 mV | 19 mA | 20 mA |
| 100 mV | 97 mA | 100 mA |

Active balancing

### Charge moved, not burned, for more parts

**Field marks.** An inductor or a transformer beside the cells, or capacitors switched between them, with a switch for each.

**Look-alike.** Passive balancing, on the same balance lead, with only a resistor and a switch a cell.

**Range.** The inductor balancer moves 0.825 A at 87.7% and loses 35 mWh balancing this pack. ADI’s LTC3300 uses a transformer, at up to 92% and up to 10 A. A switched capacitor reaches only its neighbors, and slows as the gap closes.

**Habitat.** Electric vehicles and grid storage, and packs where run time is worth the parts.

Fast enough, an active balancer can work during a flight. It feeds cell 3 from the others while the pack gives 10 A.

The balanced pack then gives 2058 mAh instead of 1936 mAh before any cell reads 3.3 V, 122 mAh more. A bleed can’t do that, since it only takes charge away.

**Fig. 15l: Balancing in flight.**

![Fig. 15l: Balancing in flight](https://nearfieldnotes.com/notes/battery-balance/fig-15l.png)

*What the figure shows:* Figure 15l: the balanced pack discharged at 10 A until any cell reads 3.3 V. With no balancing it delivers 1936 mAh. With the inductor balancer moving charge to the weak cell during the flight, 2058 mAh.

The pack balance-charged first, then run at 10 A. The active balancer is the inductor one above.

**The numbers**

| In flight | Delivered | Ends with |
|---|---|---|
| No balancing | 1936 mAh | cell 3 at 3.29 V |
| Active, during the flight | 2058 mAh | cell 3 at 3.30 V |

## Which balancer

*A key*

Key to balancing a pack

### Which balancer fits?

1. 1a The pack charges on a hobby charger Balance mode, lead plugged in
2. 1b It has its own BMS board Go to 2
3. 2a Only overcharge matters Passive, like the BQ76920
4. 2b Run time matters too Go to 3
5. 3a Charge has to cross the whole stack Inductor or transformer, like the LTC3300
6. 3b Neighbors only, and slow is fine Switched capacitor

This pack’s answer is 1a. The Tesla Model S balances with resistors, as in 2a.

## Fade a cell yourself

*24 charges, every one drawn*

Pick how far cell 3 has faded and how the pack charges. The further it has faded, the further a plain charge takes it past 4.20 V.

A bleed and the active balancer both end with every cell at 4.20 V. The difference is the heat.

**Note 15’s explorer: Each cell, charging.**

![Note 15’s explorer: Each cell, charging](https://nearfieldnotes.com/notes/battery-balance/fig-explore.png)

*What the figure shows:* Each cell's voltage while the pack charges, for the chosen weak cell and charge.

2.2 A from 3.80 V, as in Fig. 15d. Bleed is 300 mA. Active is the inductor balancer. Under 86%, a plain charge takes cell 3 past the end of the cell model.

**Every case, as a table**

| Weak cell | Charge | Weak cell ends at | Others end at | Took | Lost as heat |
|---|---|---|---|---|---|
| 100%, 2200 mAh | Plain | 4.20 V | 4.20 V | 30 min | none |
| 100%, 2200 mAh | Bleed 300 mA | 4.20 V | 4.20 V | 30 min | none |
| 100%, 2200 mAh | Active | 4.20 V | 4.20 V | 30 min | none |
| 98%, 2156 mAh | Plain | 4.21 V | 4.19 V | 29 min | none |
| 98%, 2156 mAh | Bleed 300 mA | 4.20 V | 4.20 V | 29 min | 78 mWh |
| 98%, 2156 mAh | Active | 4.20 V | 4.20 V | 29 min | 7 mWh |
| 96%, 2112 mAh | Plain | 4.22 V | 4.19 V | 29 min | none |
| 96%, 2112 mAh | Bleed 300 mA | 4.20 V | 4.20 V | 29 min | 160 mWh |
| 96%, 2112 mAh | Active | 4.20 V | 4.20 V | 29 min | 13 mWh |
| 94%, 2068 mAh | Plain | 4.23 V | 4.18 V | 29 min | none |
| 94%, 2068 mAh | Bleed 300 mA | 4.20 V | 4.20 V | 29 min | 244 mWh |
| 94%, 2068 mAh | Active | 4.20 V | 4.20 V | 29 min | 20 mWh |
| 92%, 2024 mAh | Plain | 4.24 V | 4.18 V | 29 min | none |
| 92%, 2024 mAh | Bleed 300 mA | 4.20 V | 4.20 V | 29 min | 327 mWh |
| 92%, 2024 mAh | Active | 4.20 V | 4.20 V | 29 min | 28 mWh |
| 90%, 1980 mAh | Plain | 4.25 V | 4.17 V | 29 min | none |
| 90%, 1980 mAh | Bleed 300 mA | 4.20 V | 4.20 V | 29 min | 412 mWh |
| 90%, 1980 mAh | Active | 4.20 V | 4.20 V | 29 min | 35 mWh |
| 88%, 1936 mAh | Plain | 4.26 V | 4.17 V | 28 min | none |
| 88%, 1936 mAh | Bleed 300 mA | 4.20 V | 4.20 V | 29 min | 497 mWh |
| 88%, 1936 mAh | Active | 4.20 V | 4.20 V | 28 min | 42 mWh |
| 86%, 1892 mAh | Plain | 4.28 V | 4.16 V | 28 min | none |
| 86%, 1892 mAh | Bleed 300 mA | 4.20 V | 4.20 V | 33 min | 587 mWh |
| 86%, 1892 mAh | Active | 4.20 V | 4.20 V | 28 min | 50 mWh |

#### Conditions

- **Pack:** Three LiPo cells in series, two of 2200 mAh and cell 3 faded to 1980 mAh, all from 3.80 V. Sized like the Gens Ace 2200 mAh 3S 25C, GEA22003S25X6, with a plain label.
- **Cells:** An equivalent circuit for each: the open-circuit voltage of a LiCoO₂/graphite pouch cell from PyBaMM’s Ai2020 parameters, then 4 mΩ in series and 2 mΩ in parallel with 10 kF. The resistances and the capacitance are assumed. No LiPo maker publishes them.
- **Charger:** 2.2 A until the pack reads 12.60 V, then 12.60 V until the current falls to 220 mA. A balance charge also waits until every cell is within 5 mV.
- **Bleed:** 14 Ω, 300 mA at 4.20 V, the SkyRC B6 Evo’s most. The BQ76920 case uses 84 Ω, 50 mA, on 70% of the time. Each bleeds any cell more than 1 mV above the lowest, once the spread passes 3 mV.
- **Active:** 0.825 A out of the highest cell, 87.7% of the energy into the lowest, from the inductor balancer’s settled cycle
- **Balancers:** ngspice, cells as sources behind 4 mΩ. Inductor: 10 µH with 20 mΩ, 50 kHz, on 45% of each cycle, an ideal 10 mΩ switch and a diode. Capacitor: 100 µF with 5 mΩ, 10 kHz, 10 mΩ switches.
- **Flight:** 10 A, until a cell is empty or, for Fig. 15l, until a cell reads 3.3 V under load
- **Renders:** Blender. Cells, tabs and plugs from the pack’s published size and photos.

Simulated only. Nothing on this page has been measured on a board yet.

#### Limits

- Past 4.20 V the cell model extends its fitted curves, so voltages there are indicative.
- The cells’ resistance is a guess. A higher one would make every cell read higher while charging.
- One temperature throughout, with no heating and no aging during a run. Self-discharge is left out.
- Fig. 15e is another cell, 5 Ah, tested in 0.1 V steps. No one has tested 4.25 V.
- The active balancer’s current is held at its value between cells at 4.10 and 4.00 V.

#### Files

- sim/pack.py The cell model, charger and balancers
- sim/scenarios.py The charges and flights
- sim/balancers.py The inductor and capacitor balancers in ngspice
- sim/site.py The explorer’s sweep
- render/scene.py The renders
- refs/research.md Every datasheet and paper figure and its source

#### Sources

- [Juarez-Robles and others, Overcharge and aging analytics of Li-ion cells, J. Electrochem. Soc., 2020](https://iopscience.iop.org/article/10.1149/1945-7111/ab9569)
- [TI, SLYT322, cell balancing buys extra run time and battery life](https://www.ti.com/lit/pdf/slyt322)
- [TI, BQ76920 datasheet](https://www.ti.com/lit/ds/symlink/bq76920.pdf)
- [ADI, LTC3300-1 datasheet](https://www.analog.com/media/en/technical-documentation/data-sheets/LTC3300-1.pdf)
- [Izadi and others, switched-capacitor equalizers, IEEE Access, 2025](https://cronfa.swansea.ac.uk/Record/cronfa71288/Download/71288__36075__7d157b7a8e714bd89097334a9cf33eee.pdf)
- [Blömeke and others, balancing resistors in battery systems, Commun. Eng., 2024](https://www.nature.com/articles/s44172-024-00203-6)
- [Battery University, BU-808](https://batteryuniversity.com/article/bu-808-how-to-prolong-lithium-based-batteries)
- [J&A, JA-301730P LiPo cell specification](https://www.olimex.com/Products/Power/BATTERY-LiPo110mAh/resources/JA301730P-Data-Sheet-3.7V-110mAh--170112.pdf)
- [SkyRC, B6 Evo manual](https://www.rcscomponents.kiev.ua/datasheets/B6_Evo_Instruction_Manual_EN_V1.0.pdf)
- [Betaflight, battery docs](https://betaflight.com/docs/wiki/guides/current/Battery)
- [JST, XH connector datasheet](https://www.jst-mfg.com/product/pdf/eng/eXH.pdf)
- [Gens Ace 2200 mAh 3S 25C, AMain listing](https://www.amainhobbies.com/gens-ace-3s-lipo-battery-25c-11.1v-2200mah-ga-b1113/p792030)
- [Gens Ace, LiPo basics, on S and P](https://genstattu.com/blog/the-basic-knowledge-of-lipo-battery)
- [ngspice, the circuit simulator](https://ngspice.sourceforge.io/)
- [PyBaMM, the battery models](https://pybamm.org/)

What should the next note be? [Ask for one](https://nearfieldnotes.com/#ask-for-a-note)

---

Source: https://nearfieldnotes.com/notes/battery-balance/

Cite as: Michael Walker, “Battery balance”, Near Field Notes, Note 15, Rev 1, 5 Nov 2026. https://nearfieldnotes.com/notes/battery-balance/

Figures: free to reuse under CC BY 4.0 (https://nearfieldnotes.com/license/). Each one's image is linked under its name. Credit Michael Walker, Near Field Notes, and link the figure: this address plus #fig- and its number, like https://nearfieldnotes.com/notes/battery-balance/#fig-15.