Spectral Density
Operational limits of angular rate sensors are determined by the random noise present in the output signal when the device is held completely stationary. The rate noise floor represents the continuous, high-frequency white noise that cannot be removed by simple calibration algorithms. This parameter is expressed in degrees per second per square root Hertz, defining the ultimate resolution threshold of the sensor.
A well-defined noise floor allows system designers to predict the performance of the inertial sensor over long navigation runs.
Metrological Limit
Mechanical vibrations, thermal fluctuations, and electrical noise in the sensor pre-amplifier all contribute to this noise level. In high-precision navigation applications, a low noise floor is essential to prevent the accumulation of angular errors when integrating the rate signal to find the orientation.
Characterization Technique
Evaluating this performance parameter requires recording the sensor output over a long period in a temperature-controlled, vibration-isolated test chamber. Analyzing the recorded data using Allan variance curves allows engineers to separate the white noise from other noise sources such as bias instability and rate random walk.
Signal Filter
Reducing the noise level can be achieved by applying low-pass digital filters, but this approach introduces signal delay and reduces the measurement bandwidth. System designers must balance the need for low noise against the response time required by the control loop of the platform.
