Using the MATLAB support package for Arduino, I am trying to characterise the analog input channels on an Arduino Uno board. The underlying theory is that all sensors are noisy and there are no perfect sensors in the world. After all, the arduino has got only a 10-bit ADC, and there should be at-least some quantization noise, if not pick-up electrical noise.
But when I ground any of the 5 analog pins on my Uno (i.e. a hook-up wire to the on-board GND pin), all that MATLAB can read is a perfectly flat '0' (ideal 0) even after running for long time-periods. This is a bit strange, as the sensor on-board the Arduino is certainly not super hi-fidelity.
i would really like to characterize the analog channels (for coming up with the "R" (i.e. sensor covariance) matrix when I implement a Kalman filter using data collected from this Arduino.
Are you grounding one analog pin and reading zero on all other? When pin is grounded, it would give 0v i.e 0 ADC count. Please help understand why a perfect flat '0' is not correct when pin is grounded.