All chapters
Active balancingCapacity testingWeak cell detectionBench procedure

Chapter 1 covered what a pack’s own protection board does on its own, in the field. This chapter is the bench counterpart: a full charge → active balance → discharge cycle run on a balancer / analyser, with the pack disconnected from its normal in-service charge and discharge path.

2.1Purpose

This procedure serves two goals at once:

  • Balance — bring every cell in the series stack to matched voltage using active cell-to-cell energy transfer, rather than the passive bleed-off a BMS performs.
  • Test — use the controlled discharge phase to measure each cell’s actual delivered capacity, and flag any cell that is weak, ageing or mismatched relative to the rest of the pack.

It is not something the pack’s own BMS does automatically — see Chapter 1 for what the board handles on its own, and Chapter 3 for what the capacity numbers this procedure produces actually mean in state-of-health terms.

2.2Equipment class used

Written for the general class of bench active balancer / battery analyser used for this work. Typical capability range:

Table 2.1 — Typical bench balancer / analyser capability

CapabilityTypical range
Series cell count supported2S – 32S
Balancing current (adjustable)1A – 20A
Charge current (adjustable)1A – 20A
Discharge current (adjustable)1.25A – 25A (some units to 50A on the discharge port)
Balancing modesActive, passive, charging balance, discharging balance
Discharge equalisation styleContinuous or pulse mode (selectable)
Voltage measurement resolution0.001V — 1mV class
Supported chemistriesLFP, NMC / Li-ion; some units also support LTO on discharge
MonitoringPer-cell voltage display; some units add Bluetooth or app monitoring
Built-in protectionsReverse-connection, over-series and wrong-connection protection; fuse and protection module; thermal cutback or fan above roughly 45°C internal

2.3Safety precautions

  • Confirm cell count and chemistry (LFP vs NMC) on the unit before connecting — this sets the correct charge voltage target.
  • Check polarity of every cell tap before powering on, even on units with built-in reverse-connection protection.
  • Keep the pack on a fire-safe surface and attended for the full test — a complete charge, balance and discharge cycle can run for hours.
  • Respect the unit’s thermal protection. Do not override automatic cooling or current cutback above roughly 45°C internal.

2.4Phase 1 — Setup and connection

  1. 1Power off the balancer before connecting anything.
  2. 2Connect the full cell-tap harness, lowest cell to highest, one tap per cell junction.
  3. 3Connect the main pack + / − leads to the main charge/discharge port.
  4. 4Power on and confirm on the display that the detected cell count matches the physical pack, and that every channel reads a plausible voltage — no 0V or out-of-range channels.
  5. 5Set chemistry (LFP / NMC) to match the pack.

2.5Phase 2 — Charge

  1. 1Select charging mode.
  2. 2Set charge current conservatively for an unfamiliar or aged pack.
  3. 3Start the charge. The unit brings the pack up toward full while already tracking individual cell voltages.
  4. 4Expect some voltage spread between cells at this stage — that is what Phase 3 corrects.
  5. 5Charge continues to the target charge voltage for the selected chemistry.

2.6Phase 3 — Active balancing

  1. 1Switch to active balancing mode and set the balancing current.
  2. 2Start the cycle. Active balancing moves charge from higher cells to lower cells rather than bleeding it off as heat, so the pack converges without wasting energy.
  3. 3Monitor the per-cell voltage spread narrowing over time.
  4. 4Let the cycle run until the spread stabilises at the unit’s resolution floor, or plateaus.
  5. 5Record the final per-cell table as your balanced baseline.

Table 2.2 — Example convergence log

Time (min)Highest cell (V)Lowest cell (V)Spread (mV)
03.3813.227154
203.3633.28974
403.3543.32232
603.3483.3399
903.3503.3473
Cell voltage convergence during active balancing3.2003.2253.2503.2753.3003.3253.3503.3753.400Highest starting cellLowest starting cell020406080Balancing time (minutes)Cell voltage (V)
Figure 1 — All 16 cell-voltage traces over a 90-minute run. The highest-starting cell falls and the lowest-starting cell rises, converging by roughly 70–80 minutes as the spread drops from about 154mV to about 3mV.

2.7Phase 4 — Discharge capacity test

  1. 1Switch to discharge balancing / discharge equalisation mode — pulse mode for older or weaker packs, continuous mode for healthy ones.
  2. 2Set discharge current within the supported range.
  3. 3Discharge to the pack’s specified low-voltage cut-off, continuing to monitor per-cell behaviour.
Discharge phase capacity test voltage curves2.52.62.72.82.93.03.13.23.32.5V cut-offTypical / healthy cells (15 of 16)Flagged weak cell (1 of 16)100806040200Remaining state of charge (%)Cell voltage (V)
Figure 2 — Fifteen typical cells overlaid against one flagged weak cell. The healthy group holds its plateau to roughly 10% remaining state of charge before dropping toward the 2.5V cut-off; the flagged cell’s plateau breaks early, around 30% — the practical signature of reduced usable capacity.

2.8Phase 5 — Review and data

Example per-cell capacity results for a 16S pack of 100Ah nominal cells:

Table 2.3 — Example per-cell delivered capacity, 16S pack

CellCapacity (Ah)Result
1104.1Pass
2100.6Pass
3100.5Pass
495.9Pass
5101.1Pass
6100.6Pass
7100.2Pass
888.4Flag — below threshold
9101.9Pass
1099.3Pass
1199.5Pass
12100.4Pass
13102.2Pass
14101.9Pass
1598.5Pass
1699.1Pass
Per-cell discharge capacity against the pass threshold020406080100104.11100.62100.5395.94101.15100.66100.2788.48101.9999.31099.511100.412102.213101.91498.51599.11695Ah pass threshold (example)Cell numberMeasured discharge capacity (Ah)
Figure 3 — Measured discharge capacity for each of 16 cells against a 95Ah example pass threshold. Cell 8 is the single outlier at 88.4Ah, visibly short of the line while every other cell clusters tightly between 95.9 and 104.1Ah.

2.8.1Pass / fail interpretation

Table 2.4 — Reading the results

ObservationInterpretation
All cells converge tightly during balancing and stay tight through dischargePack is healthy and well matched
One or two cells consistently hit low-voltage cut-off early on dischargeThose cells likely have reduced capacity — flag for replacement
Balancing takes unusually long or never fully convergesPossible bad connection, or a cell far outside the group’s capacity and impedance range
A channel reads erratic or no voltage throughoutCheck that cell tap’s physical connection and re-test

2.9Troubleshooting quick reference

SymptomLikely cause
Will not detect correct cell countCheck harness seating and the cell-count setting
Thermal warning or fan engagesNormal above roughly 45°C internal — reduce current or let it cool
Balancing current lower than the set pointExpected as cells converge — some topologies self-limit based on the remaining differential
Discharge stops earlyCheck the cut-off voltage matches the chemistry, and check for one cell tripping the whole pack

Important

The numeric datasets and charts in this chapter are illustrative example data, not measured results from a specific test run. Substitute your own logged voltages and capacities from each phase before using this procedure to make pass/fail decisions on a real pack. Consult your specific balancer or analyser’s user manual for exact control names, mode labels and maximum ratings before running it.

Chapter summary

Frequently asked questions

What is the difference between active and passive cell balancing?+

Passive balancing bleeds charge off the highest cells through resistors, dissipating it as heat — it is what most in-pack protection boards do, and it only runs near the top of a charge. Active balancing transfers charge from higher cells into lower ones instead of wasting it, so a pack converges faster and without the energy loss. Active balancing is typically done on a bench unit rather than by the pack’s own BMS.

How do you find a weak cell in a battery pack?+

Balance the pack first so every cell starts from a matched voltage, then run a controlled discharge to the low-voltage cut-off while logging per-cell behaviour. A weak cell reveals itself by delivering noticeably less capacity than the rest of the string and by hitting the cut-off first. On a 16S 100Ah pack, one cell coming in near 88Ah against a group averaging 100Ah is a clear flag for replacement.

How long does balancing a lithium pack take?+

It depends on the starting spread and the balancing current, but a bench active balancing cycle typically runs for one to two hours. A pack starting with a spread of around 150mV can converge to single-digit millivolts within about 90 minutes. If balancing runs unusually long or never converges, suspect a bad tap connection or a cell whose capacity and impedance sit far outside the rest of the group.

What discharge current should be used for a capacity test?+

Use a current within the analyser’s supported range and gentle enough that resistive voltage sag does not trip the cut-off prematurely — a low C-rate gives the most representative capacity figure. Pulse mode suits older or weaker packs, continuous mode healthy ones. Whatever rate is chosen, keep it consistent between tests so results stay comparable over the pack’s life.

Reviewed by

SG

Sahil Goyal

Co-founder, Wingzman

LinkedIn
SG

Sourabh Goyal

Co-founder, Wingzman

LinkedIn

Battery Fundamentals is an original educational series on lithium battery technology. Threshold tables, cycle-life curves and worked examples use representative values drawn from published product specifications and widely-observed industry patterns, not measured data from a specific product. Always consult the current manufacturer datasheet before making design, purchasing, warranty or certification decisions.