Enter the operating envelope and get a ranked chemistry with real cells to look at — filtered first on the constraints that actually eliminate candidates: pulse droop through the cell’s internal resistance, and whether it can be charged at your minimum temperature, not just discharged.
Runs entirely in your browser. Nothing is uploaded, stored, or sent anywhere.
Battery selection is usually presented as a search for the highest watt-hours per kilogram. In practice that is seldom the binding constraint. Two other things eliminate candidates far earlier, and both are absent from the front page of every datasheet.
A CR2032 holds about 225 mAh, which sounds ample for a sensor that wakes once a minute. Its internal resistance is somewhere between 10 and 40 Ω, and it rises as the cell ages and as it gets cold. Ask it for a 100 mA radio transmit burst and Ohm's law takes one to four volts away instantly. The 3 V cell delivers well under 2 V at the load, the microcontroller browns out mid-packet, and no amount of remaining capacity helps.
This is the single most common battery mistake in low-power design, and it does not show up in a spreadsheet that only tracks milliamp-hours. That is why peak current is an input here and why the ranking rejects on droop rather than deducting points for it. The fix, when you are stuck with a high-impedance cell, is to put a supercapacitor or a hybrid layer capacitor across it: the cell supplies the average, the capacitor supplies the burst.
Li-ion will discharge cheerfully at −20 °C. It must not be charged below 0 °C. Below freezing, lithium plates as metal on the anode surface instead of intercalating into it; capacity falls immediately and the plated metal can eventually grow through the separator. A solar-powered outdoor device that charges whenever the sun is up will do exactly this every winter morning, and the failures arrive a year after deployment.
The honest options are a heater before charging, refusing to charge below freezing and sizing the battery to ride through, or a chemistry that tolerates it — LTO charges to −30 °C, and NiCd is still unmatched at cold-temperature charging even though the cadmium restriction makes it hard to specify. This is why the tool asks whether you must charge at the minimum temperature as a separate question from the operating range: answering yes eliminates most of the lithium catalogue outright.
For a twenty-year meter, the load is often not what flattens the battery — the battery is. At 3 %/year an alkaline cell has lost half its energy to self-discharge before the design life is up, regardless of what the circuit did. Lithium thionyl chloride sits below 1 %/year, which is precisely why it dominates utility metering and long-life industrial sensing despite being expensive and awkward to pulse.
The awkwardness is real and worth planning for. Li-SOCl2 passivates: a protective film grows on the anode while the cell sits idle, and the first current draw after a long sleep meets a much higher resistance than the datasheet suggests until that film is broken down. Designs that sleep for months and then transmit need either a depassivation routine or a capacitor to carry the first burst.
It applies hard feasibility filters first — rechargeable or not, operating temperature, charge temperature, pulse droop, cycle life — and only ranks what survives them. Anything eliminated is listed with the reason, because "why not this one" is usually more useful than the recommendation itself.
What it cannot know is your certification path, your supply chain and your assembly process. Shipping lithium cells brings UN 38.3 testing and transport classification. A cell in a sealed product needs a different safety argument from a user-replaceable one. Coin cells in consumer products carry ingestion regulations that have become considerably stricter. Sodium-ion looks excellent on paper and may or may not be purchasable in your format in your production window. None of that is in a watt-hours-per-kilogram number, and all of it can overrule this ranking.
The figures here are representative of mainstream cells of each chemistry rather than of any specific part, and internal resistance in particular varies enormously within a chemistry — a spiral-wound Li-SOCl2 cell and a bobbin one differ by more than an order of magnitude. Use this to build a shortlist, then take the real numbers from the datasheets of the two or three cells you are actually considering.
Once a chemistry is chosen, the battery life estimator takes a duty-cycled load profile through to runtime with self-discharge included, and the capacitor life calculator covers the other half of most power designs. If the pulse problem above is what you are solving, the PDN target impedance tool is the same argument one level up: what impedance the source has to present so the load's transient does not move the rail.
Each of these runs entirely in your browser. Nothing is uploaded, stored or sent anywhere, and none of them needs an email address.
See all 49 engineering tools →
We add tools here fairly often and write up the things worth writing up — a stackup that behaved oddly, a standard that turned out to be obsolete, a calculator that was quietly wrong. Join and you get told when something new lands.
One email a month at the very most, and usually less. No drip sequence, no sales cadence, no sharing your address with anyone. Unsubscribe whenever you like.