The cell: a tank of electrons with a stubborn voltage
In 1799 Alessandro Volta stacked discs of copper and zinc with brine-soaked cloth between them, and to his amazement the pile produced a steady current — no spinning, no magnets, no friction. He had built the first battery cell, and the principle has not changed since: take two different metals, a chemistry that *wants* to push electrons from one to the other, and let those electrons do their pushing through your circuit instead of directly through the chemical. A cell is, at heart, a controlled chemical reaction with a wire running through the middle of it.
Three numbers describe a cell, and you should hold all three in your head at once. First, its [[voltage|voltage]] — the electrical 'pressure' it pushes with, set almost entirely by the choice of chemistry. A lead-acid cell sits near 2.0 V, an alkaline AA near 1.5 V, a lithium-ion cell near 3.7 V. You don't get to choose this number; the periodic table chose it for you. Second, its capacity, measured in amp-hours (Ah): a 3 Ah cell can deliver 3 amps for one hour, or 1 amp for three hours, or 0.5 A for six. It is the *size* of the tank, not its pressure. Third — the one beginners forget — its internal resistance, a few milliohms of unavoidable friction inside the cell.
Cell terminal model (the number that matters):
I (load current)
+----[ EMF ]----[ R_int ]----o +terminal
| 3.7 V 30 mΩ
|
+-------------------------------o -terminal
V_terminal = EMF − I·R_int
At rest (I=0): V = 3.70 V
At I = 1 A: V = 3.70 − 0.030 = 3.67 V
At I = 10 A: V = 3.70 − 0.300 = 3.40 V (3 W of heat inside!)
At I = 30 A: V = 3.70 − 0.900 = 2.80 V (27 W — getting dangerous)Lithium-ion: why it eats every other chemistry's lunch
The reason your phone is thin and your laptop fits in a bag is one chemistry: the [[ee-lithium-ion-battery|lithium-ion cell]]. Lithium is the third-lightest element and the most eager metal in the periodic table to give up an electron, so it offers a beautiful combination — high voltage *and* low weight. But the genius of Li-ion isn't burning the lithium; it's *not* burning it. Instead the lithium ions shuttle back and forth, sliding into and out of layered host materials (graphite on one side, a metal oxide on the other) like books slotting onto a shelf. Charge pushes them one way; discharge lets them slide back. Nothing is consumed, so the cell can do this hundreds or thousands of times.
To compare chemistries fairly we use [[ee-energy-density|energy density]]: how much energy you store per kilogram (gravimetric, Wh/kg) or per litre (volumetric, Wh/L). A modern Li-ion cell stores roughly 250 Wh/kg; lead-acid manages about 35 Wh/kg, which is why your car's starter battery is a brick and an EV's pack is not seven times heavier than the car. But energy density tells only half the story. Power density — how *fast* you can pour that energy out — is a separate axis, governed mostly by internal resistance and electrode area. A cell can be a deep, narrow well (lots of energy, slow to draw) or a shallow, wide pool (less energy, but you can scoop it out in seconds). No single cell maxes out both.
From cell to pack: stacking voltage and capacity
One 3.7 V cell won't spin an electric motor any more than one AA battery will start a car. To get useful voltage and energy you wire many cells into a pack, and you have exactly two moves. Put cells in series and their voltages add while capacity stays the same — like stacking buckets to build pressure. Put cells in parallel and their capacities add while voltage stays the same — like widening the bucket to hold more water. Pack designers describe the result as 'NsMp': N cells in series, M strings in parallel. A '96s2p' pack means 96 cells stacked for voltage, with two of those stacks side by side for capacity.
Worked example — a small EV pack (96s2p of 3.7 V / 3.4 Ah cells)
Per cell: 3.7 V 3.4 Ah
SERIES adds voltage: 96 × 3.7 V = 355 V (pack voltage)
PARALLEL adds Ah: 2 × 3.4 Ah = 6.8 Ah (pack capacity)
Total cells: 96 × 2 = 192 cells
Stored energy: 355 V × 6.8 Ah ≈ 2.41 kWh
(A real 60 kWh car pack just scales this up:
~96s and dozens of cells in parallel → thousands of cells.)
Energy check via Wh/cell:
3.7 V × 3.4 Ah = 12.6 Wh per cell
192 cells × 12.6 Wh ≈ 2.41 kWh ✓ (same answer, sanity confirmed)The BMS: the computer that keeps the pack honest
You'd think the easiest thing to know about a battery is how full it is. It is, in fact, one of the hardest. [[ee-state-of-charge|State of charge]] (SOC) — the battery's 'fuel gauge', 0% to 100% — cannot be measured directly, because there is no dipstick you can lower into a sealed cell. Worse, for a healthy Li-ion cell the voltage is a treacherous proxy: it barely moves across the whole middle of the discharge curve. From 80% down to 20% a cell might drift only from 3.9 V to 3.7 V — a flat plateau that's lovely for your devices (steady voltage) but maddening for anyone trying to read the tank.
So every serious pack carries a [[ee-battery-management-system|battery management system]] (BMS): a small dedicated computer that watches the pack like a hawk and does four jobs. It *estimates SOC*, usually by coulomb counting — integrating current in and out over time (every amp-second tracked) — and correcting that drifting estimate against the cell voltage whenever the pack rests at the curve's steep ends. It *balances* cells, bleeding charge off the fullest ones (or shuttling it to the emptiest) so the series string stays matched. It *protects*, cutting the pack off the instant any cell strays outside its safe window for voltage, current, or temperature. And it *reports*, handing the rest of the system a trustworthy gauge.
- Sense every cell group's voltage, the pack current, and several temperatures — a 96s pack has 96 voltage taps the BMS must read in turn.
- Estimate SOC by coulomb counting, then re-anchor against open-circuit voltage when the pack rests, where the voltage curve is steep and trustworthy.
- Balance the string by bleeding the fullest cells through small resistors (passive) or moving charge between cells (active), so no single cell hits its limit early.
- Protect by opening a contactor or MOSFET the instant any cell exceeds its safe voltage, current, or temperature — the last line before thermal runaway.
The other tanks: supercapacitors and fuel cells
A battery is not the only way to hold energy, and the alternatives are illuminating precisely because they live at the *extremes* a battery cannot reach. Take the [[ee-supercapacitor|supercapacitor]]. An ordinary capacitor stores energy in an electric field between two plates — no chemistry, just charge sitting on a surface. A supercapacitor pushes that idea to its limit with electrodes so porous their effective surface area is measured in *thousands of square metres per gram*, giving capacitances of thousands of farads in a soda-can shape. Because storing charge physically (no ions burrowing into a lattice) has almost no internal resistance, it can dump or soak up that energy in *seconds* and survive a million cycles. Its weakness is the mirror image: it holds maybe a twentieth of a battery's energy per kilogram.
Now go the other direction. A [[ee-fuel-cell|fuel cell]] breaks the rule that a cell stores its own fuel. A battery is a sealed tank: its energy is whatever was sealed in at the factory, and when it's drained you recharge it. A fuel cell is an engine, not a tank — it combines hydrogen fed in from outside with oxygen from the air, producing electricity (and only water as exhaust) for as long as you keep feeding it. It never 'runs flat'; it runs *out of fuel*, and you refill the tank in minutes rather than recharging over hours. That makes fuel cells attractive where a battery's recharge time or weight becomes the bottleneck — long-haul trucks, forklifts, backup power — though making, storing, and moving hydrogen remains the hard, unfinished part of that story.
Three ways to hold energy — pick by the job
ENERGY (Wh/kg) POWER CYCLES 'refill'
how much how fast how often model
------------ -------------- -------- --------- --------------
Supercapacitor ~5 enormous ~1,000,000 recharge (secs)
Li-ion battery ~250 moderate ~1,000–5k recharge (hrs)
Fuel cell very high* low–mod n/a refuel (mins)
*Fuel-cell 'energy density' depends on the hydrogen tank you
bolt on, not the cell stack — it scales with fuel, like an engine.Putting it together
Step back and the whole field is one design conversation. You start with a chemistry that fixes your cell voltage and your energy-vs-power balance. You stack cells in series for voltage and parallel for capacity until the pack has the kWh and the volts the job needs. You accept that you now cannot directly see how full that pack is, so you hire a BMS to estimate, balance, protect, and report. And when the duty cycle pushes past what a battery does well — needing a second-long burst, or a five-minute refuel — you reach for a supercapacitor or a fuel cell instead. Every electric thing you own, from the AirPod in your ear to the bus on your street, is some answer to this same set of questions.