LIBRARYMeasuring power and battery runtime.
Runtime is capacity divided by average draw, discounted for the usable fraction and the regulator loss. How to measure real current and why sleep current dominates battery life.
Runtime is capacity divided by average draw, with an honest discount for the voltage window you can actually use and the power the regulator loses. Guess the draw and you get a fantasy; measure it and you get a number you can trust. This guide, and the calculator below, do it the honest way.
Average draw is what counts
Runtime is set by the average current, not the peaks. A board that sleeps at 20 uA and wakes for a 100 mA burst once a minute averages far closer to the sleep figure than the burst. Add up the current over a full cycle of the board's behavior and divide by the time to get the true average.
The runtime estimate
Start with the capacity, multiply by the fraction of it you can actually reach before the cell hits its empty voltage, then divide by the average draw. The usable fraction is never one: a regulator stops working below its dropout, and a lithium cell is empty at 3.0 V, not 0 V, so a real board reaches perhaps 80 percent of the printed capacity.
Inputs
Result
13 h 20 m
estimated runtime
13.3 hours at 120 mA average
Efficiency and regulator loss
Every regulator between the cell and the load takes a cut. An LDO wastes the voltage it drops; a switcher loses a fixed percentage; a boost draws more current from the cell than the load ever sees. Fold that loss into the average draw at the battery, not at the load, or the estimate runs long.
Measuring the real current
You cannot estimate what you have not measured. Put a meter in series with the supply and read the current in each state, asleep and awake. For the tiny sleep current a plain multimeter often lacks resolution, so a low-value sense resistor or a dedicated power analyzer reads the microamps that decide a battery board's life.
On most battery boards the board is asleep almost all the time and awake for a few milliseconds now and then. That means the sleep current, not the flashy active burst, sets how long the battery lasts. Halving a 100 uA sleep current can nearly double the runtime, while shaving the active burst barely moves it. When a design must last months on a cell, the whole fight is the sleep number.
▸Deep dive· Self-discharge and cold, the slow leaks
Two effects erode runtime that no current meter on the load will ever show. A cell self-discharges, slowly losing charge on its own even with nothing connected, so a board that sleeps for months loses capacity to the cell as much as to the load. And cold cuts capacity: a lithium cell delivers noticeably less charge in the cold than at room temperature, because the chemistry slows down. For a board that must last a season outdoors, budget both, or the bench estimate will beat the field every time.
Checkpoint
Quick check
One Thousand Drones engineering team · verified 2026-07
Want the whole Power & Batteries cluster as one book?