This work presents an parallel analog-to-digital converter based on compressive sensing theory to process compressive/sparse signal below Nyquist Sampling Rate. Linear feedback shift registers (LFSR) are adopted to generate a 0/1 pseudo random sequence and a first-order delta-sigma ADC is presented to computes the inner product of random signals in our design. Besides, random sampling is achieved by using switched-capacitor techniques, therefore the random cut-off of switches regulates the random selection of signals by the pseudo random sequence. The system simulation verification results are given in the end of this paper.