A cell datasheet is a marketing document with engineering units. Nothing on it is false, but almost every number carries a test condition, and the conditions are where the disagreements between your bench results and the supplier's numbers come from. Here is how to read one properly.

Capacity: typical, minimum, and at what rate

A sheet that says "1200 mAh" has told you very little. Ask three questions:

  • Typical or minimum? Typical is the batch mean. Minimum is what every cell must clear. If you are sizing a product's runtime claim, you size on minimum, because half your production will land below typical by definition.
  • At what discharge rate? Rated capacity is almost always measured at 0.2C — a gentle five-hour discharge. Your device may pull 1C in bursts and see noticeably less delivered capacity.
  • To what cut-off voltage? Capacity measured down to 2.75 V is a larger number than capacity to 3.0 V. If your DC-DC drops out at 3.2 V, the tail below that is capacity you paid for and cannot use.

The practical move: specify the capacity you need at your real discharge rate, to your real cut-off, and let the supplier tell you which cell meets it. See sizing a cell from a power profile for how to turn a current trace into that number.

Nominal and charge voltage

Nominal voltage (3.7 V for most Li-Po, 3.8 V or higher for high-voltage cathode platforms) is a nameplate figure roughly representing the mid-discharge plateau. It is used for energy arithmetic and little else.

The number that constrains your design is the charge voltage: 4.2 V for a standard cell, 4.35 V to 4.48 V for the high-voltage platforms discussed in our note on HV LCO. Charging a 4.2 V cell to 4.35 V does not give you free capacity; it gives you accelerated degradation and, eventually, gassing.

Cycle life, and the four conditions behind it

"≥ 500 cycles to 80% capacity" is meaningless without charge rate, discharge rate, temperature and depth of discharge. Change any one and the number moves substantially. A cell rated 500 cycles at 0.5C/0.5C, 23 °C, 100% DoD may deliver far more at 50% DoD and far fewer at 45 °C.

We wrote a whole piece on this — reading a cycle-life curve honestly — because it is the single most misread line on any datasheet.

Internal resistance

Usually quoted as AC impedance at 1 kHz, which is a convenient production measurement and not the DC resistance that determines your voltage sag under load. DC-IR is typically higher. If your design has a tight low-battery cut-off and pulsed loads, ask for DC-IR at your pulse width and temperature, not the 1 kHz number.

Internal resistance also rises as the cell ages and falls as it warms — so an end-of-life cell on a cold morning is the worst case your power budget has to survive.

Dimensions and the thickness you will actually get

Pouch cells are quoted at a nominal thickness with a tolerance, and the tolerance is not symmetric in practice: cells grow slightly over life. A cell specified at 4.0 mm may be built to 4.0 mm +0.2/-0.1 and reach 4.3–4.5 mm late in life under normal cycling.

Design the enclosure cavity for the aged, charged, warm dimension — not the nominal one on the sheet. Our note on what causes Li-Po swelling covers where that growth comes from and which part of it is normal.

Operating temperature: two ranges, not one

Charge and discharge windows differ, and the charge window is the narrower one. A typical cell discharges from −20 °C to 60 °C but charges only from 0 °C to 45 °C. Charging below 0 °C risks lithium plating — permanent capacity loss and a genuine safety concern. If your product can be charged outdoors in winter, your charger needs a thermistor and the firmware to respect it. The CC-CV protocol piece covers how to stage that.

What is usually missing

Four things a good supplier will provide on request and a datasheet rarely prints:

  1. Self-discharge rate at a stated temperature — matters for anything that sits on a shelf.
  2. DC-IR at your pulse profile, as above.
  3. Swelling data over cycle count, not just a static thickness tolerance.
  4. The IEC 62133-2 test report for the exact cell model, not a sister model.

If a supplier cannot produce the last one, that is the answer to several other questions at once.

A short checklist

  1. Capacity: minimum, at your rate, to your cut-off.
  2. Charge voltage, and whether your charger matches it exactly.
  3. Cycle life with all four conditions stated.
  4. DC-IR, not just 1 kHz AC impedance.
  5. Aged thickness, not nominal.
  6. Charge temperature window, and how firmware enforces it.

Send that list to two suppliers and the comparison gets much easier.