Design Of The Vector Lock-in Amplifier With Adaptive Frequency Tracking Based On FPGA

In the weak photoelectric signal detection system, the useful signal was buried in strong background noise and other interference, the use of lock-in amplifier to achieve extraction of the useful signal can greatly suppress the noise to improve detection accuracy. Excellent performance's lock-in amplifier at home and abroad have adopted digital structure, such as standford company SR850, SR830 and SR810 type lock-in amplifier. On the digital lock-in amplifier implementation, to implement phase locking algorithm on programmable logic device FPGA has many obvious advantages in parallel processing capabilities and real-time, compared to DSP, Labview and other platforms.This paper based on the actual research background, introduced weak photoelectric signal correlation detection principle, analyzed the present situation of scientific research institutions in the field of lock-in amplifier. Based on Altera's FPGA chip EP4CE115F29C8 digital signal processing platform that combines 16 bit ADDA conversion module, we designed a kind of orthogonal vector digital lock-in amplifier system with adaptive frequency tracking and adaptive single frequency interference filtering out. This system includes the following aspects of development:Firstly, designed the fast frequency adaptive tracking system by use of the FFT-based carrier frequency estimation and digital phase-locked loop technology. Analyzed the DDS technology, FFT carrier frequency estimation module and two kind of excellent performance PLL(square phase-locked loop and Costas loop)which was applied in the system.Secondly, carried out the analysis and design of the filter for the filter system needs a variety of different parameters' filters. To reduce the storage and computation of the phase-locked algorithm and save the FPGA resource effectively, the multi rate signal processing method was used.; realized a narrowband low-pass filter which's bandwidth was 0.1Hz and introduced the internal CIC filters and HBF filters. Designed a adaptive notch filter which was based on error's sign LMS algorithm to filter out the power frequency and 100 Hz lighting interference.Thirdly, tested the system's frequency tracking range, tracking speed and accuracy, as well as the anti-noise performance, and gave the overall result of the measurement system, the adaptive frequency range of system was measured from 0.02 to 8000 Hz, and the simulated frequency tracking accuracy in this range was 10-4Hz with the system bandwidth 0.1Hz. The average relative magnitude measurement error was 0.87%. Analyzed the error sources of the system and supplied corresponding solutions.