A new neural A/D converter architecture

Abstract In this paper, a new A D converter architecture is proposed which includes the benefits of the neural A D converters and advantages, in terms of speed and resolution, of traditional A D converters. This neural A D converter is characterized by a very simple circuit based on a cascade of 1-bit stages including only an analog comparator and an amplifier. The functional description of the new neural A D converter and some design considerations are discussed. Furthermore, the experimental results obtained with a discrete component prototype of the converter are given. Finally, tests used for the dynamic characterisation of conventional A D converters are performed on the prototype and the results are also given.