The design of Neural A/D converter using neuron-MOS transistor

Abstract

A novel circuit realization of two parallel A/D converters based on the Hopfield's neural network is presented. The two neural A/D converters are constructed with comparators and Schmitt triggers, respectively, and they are all designed by using neuron-MOS transistors. The benefit of the proposed circuit realization with neuron-MOS transistors is that the structure of the circuit has been simplified significantly. The proposed circuits can be fabricated by standard CMOS process with a 2-ploy layer. From the HSPICE simulation results using TSMC 0.35μm double-polysilicon CMOS technology, the effectiveness of the proposed approach is validated.