A Novel FEM with MPGA-Assisted Noise Matching Network and DC-offset Cancellation for Ultrasonic Application

A novel analog front-end module (FEM) featuring an auto-generated 4-6 MHz noise matching network (NMN) and DC-offset cancellation (DCOC) circuit is presented to significantly enhance the sensitivity of the ultrasonic measurement. Compared to the conventional ultrasonic FEM architectures, the proposed design achieves a higher output signal-to-noise ratio (SNR) and incorporates DC offset suppression. The NMN is automatically synthesized using a multiple population genetic algorithm (MPGA), thereby circumventing the need for complex Butterworth Van-Dyke (BVD) modeling of the ultrasonic sensor. Furthermore, a dedicated DCOC circuit is integrated into the FEM to suppress the DC noise originating from low noise amplifier (LNA) imbalance, which reduces signal distortion and mitigates baseline drift inaccuracy. The measurement results demonstrate a voltage gain of 28.3 dB and a 10.41 dB improvement in output SNR over the baseline configuration without the FEM. The ultrasonic imaging verification experiments confirm a qualitative sensing improvement, indicating the suitability of the proposed FEM for high-sensitivity ultrasonic measurement and imaging applications. © 1963-2012 IEEE.