Nanopower sampled data wavelet filter design using Switched Gain Cell technique

In order to realize a nano-power wavelet filter for biomedical applications, this paper applies the singular value decomposition approximation to transform the time domain 1^st-derivative of a Gaussian (gauss1) wavelet base into a 5^th-order z-domain transfer function. Consequently, to realize the approximated transfer function in CMOS technology employing circuitry that operates from a low supply voltage, a sampled data circuit technique, coined dasiaswitched gain cell, (SGC),psila is introduced. Using the SGC technique, standard MOS switches, simple subthreshold (nonlinear) transconductors and their associated parasitic capacitances suffice to constitute the filter, while the scale of the filter can be controlled by the clock frequency. This renders the filter architecture to be simple, modular, and area efficient. Simulation results, using 0.13 mum CMOS model parameters, show that the wavelet filter implements the gauss1 wavelet base well, operates from a 1 V supply and consumes less than 0.47 muW quiescent power.