← Back to the chapter · Previous detail

Technical detail · Power Management Tools

How a guessed sleep current distorts lifetime

Lifetime is inversely proportional to average current. A small percentage error in the dominant current therefore produces nearly the same percentage error in lifetime, in the opposite direction. The most sensitive input is the most dangerous one to guess.

Suppose the datasheet promises 10 uA deep sleep and the hourly model predicts about 6.7 years, but the assembled board sleeps at 200 uA because of a regulator and interface chip. Keeping the 50 mA, 0.5 s hourly burst, the real average becomes about 207 uA and ideal device-only life is roughly 0.55 years. One wrong input has overstated lifetime by more than twelvefold. Measuring active current carefully while guessing dominant sleep current puts precision on the wrong number.

Now suppose usable capacity is also 650 mAh rather than the entered 1000 mAh. The measured and entered average currents are about 206.9 and 16.94 uA. The optimism factor is (1000 / 650) x (206.9 / 16.94), about 18.8: roughly 6.74 years predicted versus 0.36 years under these hypothetical conditions.

If the same node reported every few seconds, active current could dominate instead. Rerun the sweep after a duty-cycle change, measure the steepest input on real hardware, and show expected and conservative lifetimes. A single figure hides uncertainty in the assumptions.

← Back to the chapter · Previous detail