Voltage at 25 °C, moderate C-rate
| Chemistry | Full V | 50% V | Empty (nominal) | Curve shape |
|---|---|---|---|---|
| Li-ion (NMC) | 4.20 V | 3.80 V | 3.00 V | Gradual drop, knee at ~3.3 V |
| LiFePO₄ | 3.65 V | 3.30 V | 2.50 V | Flat plateau — poor SoC from voltage alone |
| Alkaline (AA) | 1.55 V | 1.25 V | 0.90 V | Sloping |
| NiMH (AA) | 1.40 V | 1.25 V | 1.00 V | Very flat, knee at end |
| Lead-acid 12 V | 12.7 V | 12.2 V | 11.8 V (50% DoD) | Sloping, depends on rest |
| Lithium primary | 3.60 V | 3.30 V | 2.00 V | Long plateau |
Observations
- Flat curves (LiFePO₄, NiMH): voltage is a poor SoC indicator — use coulomb counting.
- Sloped curves (Li-ion NMC, lead-acid): voltage reading with load removed gives decent SoC.
- Under load: terminal voltage drops due to internal resistance — wait for it to rest for an "open-circuit" reading.
- Temperature effect: cold batteries show lower voltage and less usable capacity.
- High C-rate: steeper voltage sag; rated capacity is at specified (often low) discharge rate.
Peukert effect
| Peukert law | C = I^k · t, where k ≈ 1.1–1.3 for lead-acid, ≈1.05 for Li-ion |
|---|---|
| Implication | Lead-acid delivers much less than rated capacity at high current |
Worked example: estimating state of charge
A rested lead-acid battery reads 12.2 V. The table above gives 12.7 V at full charge and about 12.2 V at 50%, so the battery is about half full. Read the same battery while a 20 A load is running and it may show 11.8 V; that is mostly the internal resistance drop (for example 0.03 ohm × 20 A = 0.6 V), not an empty battery. Always let the pack rest, typically 30 minutes to a few hours, before judging charge from voltage.
For an NMC lithium cell the same method works well: 4.20 V is full, 3.80 V is about half and 3.30 V is nearly empty. A LiFePO₄ cell is the opposite. Between roughly 20% and 90% it sits within about 100 mV of 3.3 V, so a 20 mV measurement error moves the estimate by tens of percent. Use a coulomb counter (a shunt and integrator) and recalibrate it at full charge.
Choosing a cut-off voltage
- Cut off under load, not at rest. A device that stops at 3.0 V per lithium cell while drawing heavy current leaves more capacity behind than one that stops at the same voltage with a light load, because the loaded voltage includes the resistance drop.
- Stay above the damage threshold. Discharging a lithium cell far below about 2.5 V can damage it permanently. Lead-acid should not go much below 10.5 V under load on a 12 V battery, and regular deep discharge shortens its life.
- Capacity is a function of rate and temperature. The same cell can deliver 100% at 0.2C and 25 °C, but noticeably less at 2C or at 0 °C. Compare curves only at the same rate and temperature.
- Datasheet curves are measured. Use the manufacturer curve for the exact cell when sizing; generic values in the table are typical, not guaranteed.