A Compressed Sensing(CS) acquisition method is presented that can address both the energy and bandwidth constraints of traditional wireless ENG(electroneurogram) signal sensor. ENG signal which is a kind of bio-physical signal is sparse in time domain making it suitable for CS based acquisition and compression. Simulation of CS samplings demonstrates data compression rate up to 10x while maintaining low perceptual loss in reconstructed signal. Unlike traditional ENG signal acquisition systems which just capture the key part of spike potential, losing information consequently, the presented sampling method enables continuous data acquisition and compression.