A new approach to increasing sensitivity and resolution of A/Ds

Signal processing in software defined receivers often requires increasing sensitivity and resolution of analog-to-digital converters (A/Ds). Two approaches to this problem are currently used. The first approach is based on oversampling, and the second one is based on a generalized successive-approximations technique. Although these approaches have not completely exhausted their potential, they encounter formidable technological difficulties when it is necessary to maximize the product of the A/D resolution and sampling rate. Therefore, the development and investigation of new approaches and methods that increase sensitivity and resolution of A/Ds are important. A new approach to increasing the A/D sensitivity and resolution is presented and analyzed in this paper. The approach is based on joint processing of M signal samples. It uses the effect of reduction of each sample contribution to the quantization error of algebraic sums of the samples when the individual samples and the sums are quantized by A/Ds with the same quantization steps. To utilize this effect, only the least significant bits of the quantized samples and the corresponding portions of the samples' analog values are summed, and the sums are formed to be components of the Walsh spectrum of these portions of the processed samples. The proposed approach allows increasing sensitivity and resolution of even the fastest A/Ds with high initial resolution. Several new methods and devices illustrating this approach are described in the paper.