Blog · 2026-07-05 · Discharge

Series vs Parallel Battery Packs: 2S, 3S, 4S and P-Counts Decoded

Every battery listing is a tiny cipher — 3S, 4S2P, 25C, 2200 mAh. Two letters do all the work: S for series, P for parallel. Decode them and you can read any pack, design your own, and know exactly why voltage and capacity trade the way they do.

A battery pack is never really "a battery." It's a team of individual cells — each one a fixed chemistry with a fixed voltage — wired together in one of two ways, or both. Series stacks voltages; parallel pools capacity and current. Everything on a pack label, every charger setting, every "can I run 12 V motors from this?" question resolves into counting the S and the P. This article builds that fluency from the cell up.

The cell: the atom of every pack

Chemistry fixes a cell's voltage window, and no wiring changes it. A lithium-polymer or standard lithium-ion cell lives between 4.2 V (full) and about 3.0–3.3 V (empty), with 3.7 V as the quoted "nominal" average. NiMH runs 1.2 V nominal; LiFePO₄ 3.2 V; lead-acid 2.0 V per cell. (Which chemistry to pick is its own decision, covered in LiPo vs Li-ion vs NiMH — here we take the cell as given and do the arithmetic.) A cell also has a capacity (mAh), a maximum discharge current, and an internal resistance. Wiring cells together transforms these four numbers in perfectly predictable ways.

Series: stacking voltage

Connect the positive of one cell to the negative of the next and voltages add, like batteries in a flashlight tube:

Pack voltage = cell voltage × S
Capacity (mAh): unchanged · Max current (A): unchanged
Internal resistance: multiplied by S

So "3S" means three lithium cells in series: 3 × 3.7 = 11.1 V nominal, 12.6 V hot off the charger, ~9.9 V empty. The pack's capacity is still whatever one cell holds — the same charge simply falls through more voltage, delivering more energy (Wh = V × Ah). The standard ladder: 1S = 3.7 V, 2S = 7.4 V, 3S = 11.1 V, 4S = 14.8 V, 6S = 22.2 V nominal. Note the full-charge figures (8.4, 12.6, 16.8, 25.2 V) — motors, ESCs and regulators must survive the full voltage, not the nominal one printed largest on the label.

Why raise voltage at all? Because power = voltage × current, and current is the expensive one — current is what heats wires, sags packs and demands thick copper. A 100 W drivetrain at 7.4 V pulls 13.5 A; the same power at 14.8 V pulls 6.8 A through the same wires with a quarter of the resistive losses. Higher S-counts also spin motors faster (RPM scales with voltage) and give regulators comfortable headroom. This is why serious builds trend upward in S as they grow.

Parallel: pooling capacity and current

Connect cells positive-to-positive, negative-to-negative, and they behave as one bigger cell:

Voltage: unchanged
Capacity = cell capacity × P · Max current = cell max × P
Internal resistance: divided by P

Two 2600 mAh cells in parallel make a 5200 mAh "cell" at the same 3.7 V, able to deliver twice the current, with half the internal resistance. That last line is quietly the most valuable: internal resistance is the villain of the companion article Voltage Sag and Brownouts, and parallelling cells attacks it directly. It's also why a physically larger pack of the same chemistry sags less — more material in parallel inside.

Reading the full label: 4S2P and friends

Real packs combine both. "4S2P, 18650, 3000 mAh cells" decodes as: pairs of cells parallelled (2P → 6000 mAh building blocks), four such blocks in series (4S → 14.8 V nominal). Total: 14.8 V, 6000 mAh, eight cells. The order of assembly (parallel first, then series — "P-groups in series") is the standard construction because it lets one balance lead monitor each parallel group as a unit.

ConfigNominal VFull VCapacityTypical use
1S3.7 V4.2 V1× cellMicro robots, tiny whoops
2S7.4 V8.4 V1× cellSmall bots, servo-friendly
3S11.1 V12.6 V1× cellThe hobby default; "12 V-ish"
4S14.8 V16.8 V1× cellFaster drives, bigger arms
3S2P11.1 V12.6 V2× cellLong-runtime rovers
6S22.2 V25.2 V1× cellHeavy platforms, big BLDC

C-rating interacts with P directly: max current = capacity (Ah) × C, and P multiplies capacity, so parallelling is one legitimate route to more amps. A 2200 mAh 25C pack offers ~55 A on paper; the 2P version offers 110 A from the same cell type. Check any candidate configuration against your robot's actual draw with the Battery C-Rating Checker — and remember its lesson that printed C-ratings deserve a haircut.

Balance: the rule that keeps series packs alive

Series is where the safety obligations live. Cells in series carry identical current but are never perfectly identical themselves — tiny capacity differences mean one cell always empties first and fills first. Charge the stack blind and the strongest cell overshoots 4.2 V while the pack total looks fine; overcharged lithium is how fires start. The answer is the balance connector, the thin multi-pin plug on every multi-S LiPo: it gives the charger a voltage tap between every series junction so it can top-balance cells individually. Three rules follow. Always charge through a balance-capable lithium charger. Never charge a pack whose per-cell voltages have drifted more than ~0.1 V apart without investigating — a chronically low cell is announcing its retirement. And never discharge below ~3.3 V per cell; the pack total can hide one cell dying early, which is what per-cell low-voltage alarms are for.

Parallel has one rule of its own: only connect cells or packs in parallel when their voltages match closely (within ~0.1 V). Connect a full pack to an empty one and the difference drives a huge equalizing current through nothing but internal resistance — sparks, heat, damage. Match first, then join. And never parallel packs of different chemistry or different S-count, ever.

Choosing S and P for a robot: the short procedure

Voltage (S) comes from the loads. List what must run: motors (rated voltage), motor driver (input range), regulator (minimum input). Pick the S whose full-charge voltage stays under every maximum and whose empty voltage stays above every minimum. The hobby sweet spots exist for a reason: 2S suits small builds with 6–7.4 V motors; 3S covers the vast "12 V" ecosystem of gearmotors and drivers.

Capacity and current (P, or just a bigger pack) come from the budget. Total your current draw with the Power Budget Calculator, decide your runtime target, and size Ah accordingly — then verify the discharge headroom. If runtime and current demands can't be met by one cell size, that's precisely what P is for.

Then weigh it. Energy is mass; doubling P doubles battery weight, and on a mobile robot that weight raises the very current draw you were budgeting. Iterate once — heavier pack, recompute draw, recheck runtime — and the design converges quickly.

Quick answers

Can I build my own pack from 18650 cells?

Technically yes; practically, only with matched cells (same model, same batch, same tested capacity), proper spot-welded nickel strip (soldering heats cells dangerously), a BMS or balance leads, and real respect for what shorted lithium does. For most robots, a manufactured pack with a warranty is the engineering-mature choice.

Why does my 3S pack read 12.6 V, not 11.1 V?

Both are correct: 11.1 V is nominal (the average over discharge), 12.6 V is full charge (4.2 V × 3). Design calculations use nominal for runtime and full for maximum-voltage safety checks.

Does adding P increase C-rating?

It increases maximum current (which is what you actually care about) because current = Ah × C and the Ah doubled. The printed per-cell C-rating itself is unchanged — the pool of amps just got bigger.

Storage and care by configuration

Configuration also shapes the care routine. Series packs need storage charging: lithium cells kept full or empty age fast, so any pack idle for more than a week should rest at storage voltage — about 3.8 V per cell, a standard button on every balance charger — meaning 11.4 V for a 3S, 15.2 V for a 4S. Check the balance ports of stored packs monthly; a cell self-discharging faster than its siblings is the earliest warning of a pack going bad, visible long before performance suffers. Parallel groups largely self-balance in storage (they're hard-wired to the same voltage), but that same property hides a weak cell: the healthy ones quietly prop it up until the group's capacity and current capability have degraded well below spec. For multi-P packs that matter, an occasional capacity test — full charge, measured discharge through a tester or your own logged run — is the only honest health check. And regardless of S or P: fireproof storage bags, no charging unattended, and retirement without sentiment for anything puffed, dented or drifting.

S buys voltage, P buys amps and runtime, balance keeps it honest. With the cipher decoded, sanity-check any real or hypothetical pack against your robot in the Battery C-Rating Checker, and see how the voltage choice ripples through the whole electrical system in the Power Budget Calculator.