Technical detail ยท MCU vs MPU: Measurements and Boundaries
ADC sampling rate and energy
An analog-to-digital converter (ADC) turns a voltage into a number using discrete steps. Quantisation error is the difference between the actual voltage and the nearest represented step. Signal-to-quantisation-noise ratio compares signal strength with that rounding error; a larger ratio means the error is smaller relative to the signal. Aliasing occurs when sampled high-frequency content appears as a different, lower frequency in the recorded data.
The mathematical gist. A signal with a 200 Hz component needs more than 400 samples each second, with practical margin and filtering to limit aliasing. At exactly 400 samples per second, the sample phase can miss that component. The Nyquist bandwidth runs from zero to half the sampling rate; the ideal noise calculation sums quantisation noise over that bandwidth. A 12-bit ADC on 3.3 V divides the input into 4,096 steps of 0.806 mV and gives an ideal 74.0 dB signal-to-quantisation-noise ratio for a full-scale sine wave, with quantisation noise integrated over the Nyquist bandwidth. Real converter noise and distortion can make the result worse. At a catalog-typical 5 mA for 10 microseconds, one conversion costs 165 nJ, so an illustrative budget at 400 samples per second is 5.70 J/day; this rate is not a sufficient acquisition design for the 200 Hz component. At the same assumed conversion cost, 288 door-sensor samples use 47.5 microjoules/day.