Cells to Pack

TLDR: The cell gets the headlines, but a battery is only as good as how you wire the cells together. Series stacks voltage, parallel stacks capacity, busbars carry the current, and the module is where all of it has to survive heat, vibration, and one bad weld. Here's how a pile of cells becomes a pack.

You can build a world-class cell and still ship a bad battery. There's a lot that happens between the cell and the car…

Let's get into it.

Series and parallel: the only two moves

There are exactly two ways to connect two cells, and they do opposite jobs.

Series (positive terminal to negative): voltages add, capacity stays the same. Two 3.6 V cells become one 7.2 V unit that still holds 50 Ah. This is how you climb from a 3.6 V cell to a 400 V pack. The catch is balance: cells in series must stay matched, because the weakest one hits empty (or full) first and caps the entire string. That is why every series group gets its own voltage tap, feeding a BMS (battery management system, the pack's brain) that keeps the cells even.

A voltage tap (or sense lead) is a thin wire run to each junction in a series string, so the BMS can read the voltage of every individual cell or parallel group, not just the pack total. From the two pack terminals you only see the whole stack; the taps let it watch each cell and bleed charge off any that run high.

Parallel (positive to positive): capacity and current add, voltage stays the same. Two 3.6 V, 50 Ah cells become one 3.6 V, 100 Ah unit. Cells in parallel self-balance, they share the load automatically. The catch is that a parallel group behaves like one big cell, so if one cell inside develops an internal short, its siblings dump their combined current straight into it. That is a safety design problem, and the reason parallel groups get fusible links.

A fusible link is the fix: a deliberately weak, sacrificial piece of conductor, basically a fuse built into the interconnect, sized to carry normal current but to melt open under a fault. Each cell joins the busbar through a thin link (a wire bond or a narrow tab), so when its siblings surge into a shorted cell, that link burns through first and cuts the bad cell loose before it can drag the group into a fire or spread thermal runaway. You sacrifice one cell to save the pack.

Bottom line: series buys voltage, parallel buys capacity. Every real pack uses both.

Reading the label: nSmP

Once you combine them, batteries get described with a shorthand, nSmP, where n is the number in series and m the number in parallel. A "2S2P" pack is two in series, two in parallel.

The math is simple. Voltage = series count times cell voltage. Capacity = parallel count times cell capacity. Energy is the product. So four 3.6 V, 50 Ah cells in 2S2P give 7.2 V, 100 Ah, and 720 Wh. Both numbers doubled, because you used both moves.

Here is the lever nobody outside the industry sees: more series means higher voltage, which means less current for the same power, which means thinner conductors and lower resistive (power a conductor wastes as heat: P = I² × R) losses, but also more cells to monitor and higher-voltage safety. That single trade is the entire 400 V versus 800 V debate happening in EVs right now.

What do real packs use? There is no universal answer, because series sets voltage and parallel sets capacity, so the combo follows the application. But the landmark is EVs: roughly 96S gets you a 400 V pack (Tesla's early packs were 96S74P). The parallel count tracks cell size, small cylindricals stack dozens in parallel, while big prismatic and pouch cells use just 1P or 2P. The move toward large cells (4680, big prismatic) is actually pushing parallel counts down toward 1P.

Busbars: the unglamorous conductors

The connections carrying all that current between cells and modules are busbars, and they are where clean bench results go to fail in the field.

The job: move big current with as little resistance as possible, because every milliohm is heat (I² × R again) and lost range. Material is a trade: copper is the best conductor but heavy and expensive; aluminum is lighter and cheaper but only about 60% as conductive, so it needs more cross-section; nickel and clad bimetals show up where you need weld compatibility.

Then there is the joint. Busbars get attached by laser welding, ultrasonic welding, wire bonding, or bolting, and a bad joint is both a resistive hot spot and a single point of failure. This is the unsexy engineering that decides whether a pack survives ten years of vibration. Busbars are also the fault path, so fusible links sit in the network to keep one bad cell from taking its whole group down.

Module design: where everything has to coexist

A module is the repeatable building block between cell and pack: a serviceable, testable group of cells plus their interconnects and sensing. Designing one means making four kinds of engineering share the same box.

And because it is modular, you can build, test, and replace in units instead of wrestling an entire pack. That is the whole point of the module: it turns a terrifying monolith into serviceable Lego.

The point

The cell is the athlete. The module is the team, the coaching, and the ambulance on standby. You cannot out-chemistry a bad busbar or a hot corner, and a pack that aces the spec sheet can still fail on the tolerance of a single weld. A battery is a systems problem wearing an electrochemistry costume, and the parts that decide whether it ships are the ones nobody puts on the slide.