A novel technique for spectral processing of a fiber optic gyroscope

This paper presents a novel technique for the digital processing of a phase-modulated fiber optic gyroscope (FOG). We have developed a digital processing scheme featuring compact hardware and new digital processing algorithms for the FOG. The computational scheme has been devised using undersampling, simplifying both hardware and software processing. A piezoelectric modulator induces phase modulation and an interference signal is detected using a PIN-diode. It is then amplified by a low noise amplifier. An analog-to-digital converter is used for digital conversion. Detailed analysis is done for selecting the appropriate analog-to-digital converter, taking into account its features which are generally overlooked in oversampling techniques. Full power bandwidth and effective numbers of bits were carefully analyzed in this regard. Data reduction paved the way for developing new algorithms based on the efficient Discrete Fourier Series. The algorithm requires that the frequency of the sampling clock must be an integer multiple of the undersampled signal. For that purpose, a sampling clock was carefully chosen that meets this requirement. The algorithm calculates the amplitudes of spectral components to be used for calculating the rotation rate. We also propose a tracking control loop for eliminating the quadrature bias error to enhance the FOG sensitivity. Test results are presented showing excellent bias stability and scale factor linearity, surpassing those reported with existing open-loop schemes.