High accuracy frequency extraction based on sigma-delta ADC applied in resonant dew point sensor

Abstract

One of the design challenges for the resonant dew point measurement was optimizing the rapidity and accuracy of frequency extraction to meet the requirements. We have developed a 1-bit second-order sigma-delta (ΣΔ) analog-to-digital (ADC) to be applied in it, which was anticipated to have a highest resolution and design flexibility than traditional ADC. Based on the Butterworth filter, the design of ΣΔ modulator was modeled firstly to evaluate the performance of filters scheme before the hardware implementation. Furthermore, the design and implement of the digital decimation filters was presented to extend the design flexibility and application. To improve the accuracy and instantaneous of frequency estimation, we developed a new frequency estimation algorithm, which combined the Fast Fourier transform (FFT) and the chirp-Z transform. We built a complete hardware implementation system. We tested the performance of the designed ΣΔ ADC and found that has 11-bit resolution and the high flexibility. The experiment results showed the root mean square (RMS) error of frequency estimation was 0.0917 Hz with minimum error 0.01 Hz, and the SNR can achieve 123.04 dB. Results showed the proposed frequency estimation error decreases approximately 35 times under the premise of the same computation with FFT. Moreover, every completion of the frequency estimation only takes less than 1 s. Finally, the experiment of dew point measurement was carried out. The results indicated that the proposed measurement system was feasible for the frequency extraction of the resonant dew point sensor.