Skip to Main Content
In this paper, we analyze and optimize the PGC demodulation algorithm for optical interferometers in order to extend its usage in low-cost all digital schemes. In our low-cost system, we choose heterodyne interferometer and direct modulation on Distributed FeedBack (DFB) laser and optimize the algorithm, as well as some of the system parameters in order to have the best linearity and noise performance. We first analyze the influence of low-pass filters (LPF) in the PGC-DCM algorithm. A detailed theoretical model of LPF is deduced and reveals that a nonideal LPF is the major limit of linearity in digital systems. Experimental results show that the PGC-atan algorithm will have a 40-dB higher linearity than PGC-DCM one with a 60-order digital FIR filter. Then we analyze the influence of the intensity modulation in PGC-atan algorithm, the main source of the nonlinearity in PGC-atan algorithm, and propose a simple but effective modification. Experiment results verified that it can reduce the influence of intensity modulation by about 20 dB. Further, the impact of Light Intensity Noise (LIN) and circuit noise on the output noise level is analyzed. Equations are derived to calculate the Power Spectrum (PS) of output noise caused by LIN or circuit noise. Simulations show that the theoretical analysis is of high accuracy-less than 0.5 dB. With these equations, a new parameter-Noise Transfer Factor-is defined for better discription of the noise performance of PGC. The results show that the modulation depth C and the DC work point Phi = mean(phi(t)) have great impact on the output noise level. By changing C and Phi , the output noise may decrease as much as about 20 dB. Then three examples show how to choose the system parameter according to the characteristics of LIN or circuit noise. As well, the noise performance of PGC-DCM is analyzed. The result reveals a turning point at the output noise base, which worsens its noise performance at lower frequency. P- - GC-DCM has the same NTF at frequency higher than the turning frequency. A parameter Frequency Conversion Coefficient is introduce to calculate the turning point.