Output Deviation
Microelectromechanical systems exhibit a slow change in the reported angular velocity even when the sensor remains perfectly stationary. The mems gyroscope drift represents the integration of small errors over time, leading to a growing inaccuracy in the calculated orientation. This phenomenon limits the effectiveness of inertial navigation systems during long periods without external corrections.
Bias Instability
Random walk and bias instability constitute the primary components of the error signal. Allan variance plots identify the different noise sources by analyzing the stability of the output over varying integration times. The flicker noise within the electronics and the mechanical damping of the resonating structure contribute to the baseline fluctuation.
Environmental Sensitivity
Temperature gradients across the sensor package cause physical deformation of the silicon beams, altering the resonant frequency. Because the sensing mechanism relies on the Coriolis effect, any change in the mechanical properties of the vibrating mass shows up as an apparent rotation. Shielding and careful mounting reduce the impact of external vibrations on the drift rate.
Calibration Strategy
Calibration algorithms often use internal temperature sensors to apply a lookup table of correction values. These coefficients must be determined for each individual unit during the factory testing phase. Continuous zero velocity updates provide a way to reset the accumulated error when the device is known to be at rest.
