A single lithium-polymer cell gives you 3.0–4.2 V and anywhere from 50 mAh to 10,000 mAh in standard geometries. Most applications need more voltage, more capacity, or both. How you connect cells to get there determines your balancing requirements, protection complexity, and long-term reliability.

The xSyP notation

Battery pack engineers use the notation xSyP to describe a multi-cell configuration: x cells in series, y cells in parallel per series group.

  • 2S1P: Two cells in series, one parallel group. Pack voltage = 2 × cell voltage (6.0–8.4 V). Pack capacity = 1 × cell capacity.
  • 1S2P: One series group with two cells in parallel. Pack voltage = 1 × cell voltage (3.0–4.2 V). Pack capacity = 2 × cell capacity.
  • 3S2P: Three series groups, each with two cells in parallel. Pack voltage = 3 × cell voltage. Pack capacity = 2 × cell capacity. Six cells total.

Series configuration: adding voltage

Connecting cells in series sums their voltages. This is necessary when the application load requires a voltage above the single-cell range — a 24 V power tool, a 12 V industrial radio, or a 7.4 V drone. Every cell added in series multiplies the voltage by a proportional factor.

The fundamental engineering requirement of a series string is cell voltage balancing. Because no two cells are perfectly identical, the weaker cell in a series string will hit the cutoff voltage (in discharge) or the charge ceiling (in charge) before the others. Without a balancer, the pack terminates early on discharge (the weak cell pulls the whole string down) and the strong cells remain undercharged. Over many cycles, the imbalance grows until one cell is chronically driven outside its safe window.

Passive balancing bleeds current from the stronger cells during the CV phase, slowly equalising voltages. It wastes energy as heat but is inexpensive. Active balancing shuttles charge from stronger to weaker cells using inductors or capacitors, recovering most of the energy. It adds cost (USD 0.50–3.00 per cell) but dramatically improves cycle life in high-cell-count packs. For 2S–4S consumer packs, passive balancing is almost always sufficient. For 8S+ industrial packs, active balancing begins to make economic sense.

Parallel configuration: adding capacity

Connecting cells in parallel sums their capacity and keeps voltage constant. This makes sense when a single cell of the needed geometry cannot provide enough capacity — a smartwatch that needs 450 mAh might use two 225 mAh curved cells in parallel to fit the enclosure without increasing cell thickness.

The requirement for parallel cells is careful capacity and internal resistance matching. Cells in parallel share current proportional to their internal resistance difference. A 10 mΩ difference between two 200 mΩ cells causes only a 5% current imbalance — acceptable. A 50 mΩ difference causes a 25% imbalance — the weaker cell overworks, heats more, and ages faster. Match cells from the same production batch, and match by both capacity (within ±1%) and internal resistance (within ±5 mΩ for small cells).

Parallel cells are particularly dangerous during assembly: connecting two cells with a significant voltage difference creates a large impulse current that can weld tabs, damage cell tabs, or cause thermal events. Pre-screen all parallel cells to within ±50 mV of each other before connecting.

Tab welding and connection topology

How the cells are physically connected determines the pack's internal resistance distribution and its susceptibility to single-cell failure propagation. Two topologies dominate in small-format packs:

PCB-mounted tab welding: Tabs are welded to a rigid PCB that carries the BMS. This is the standard for consumer electronics — compact, manufacturable at volume, and low resistance if the weld quality is controlled. Limitation: the PCB becomes a structural element and cannot flex.

Bus-bar or wire-harness connection: Used in larger industrial packs (≥ 6S, ≥ 20 Ah). Nickel or copper bus bars are welded between cell groups, with separate wires to the BMS balance taps. Higher component count, but allows individual cell replacement and better thermal management.

For any series configuration, the balance tap wires must be routed to a balancing circuit — either on the BMS or on a separate balancer board. Omitting balance taps is a common cut in low-cost pack designs and is the primary cause of premature capacity loss in 2S–4S consumer battery packs.

When to upgrade from a PCM to a smart battery

A simple PCM handles overvoltage, undervoltage, overcurrent and short circuit protection — the four essential safety functions. For applications where the host system does not need to know state of charge, temperature, or remaining runtime, a PCM is sufficient and lowest cost.

The case for upgrading to a smart battery (one that communicates via SMBus or I²C) is:

  • The host OS or firmware needs accurate SoC to display a battery indicator
  • The pack has 3 or more series cells (where manual voltage checking is impractical)
  • The application requires predictive end-of-life warning (service life tracking)
  • The customer's safety or regulatory requirement mandates state reporting (medical, aviation)

The cost premium for a smart BMS is roughly USD 2–8 per pack for consumer-grade SMBus ICs, rising to USD 15–30 for industrial-grade chips with extended temperature range and authentication. The complexity of firmware integration on the host side should not be underestimated — smart battery protocol has subtle quirks that consume 2–4 weeks of embedded engineering time.