Microprocessor Application in Walsh-Fourier Spectral Conversion

Abstract

The microprocessor has developed over the last few years into a computer built on a few low cost chips. This has facilitated the use of firmware techniques in many areas where, previously, computer applications were too expensive. Thus it is finding widespread use in digital instrumentation and signal processing; this paper describes a microprocessor in such an application. While sinusoids have traditionally been used in spectral analysis, the bipolar Walsh functions, which also constitute an orthonormal set, are increasingly coming into use in this field. In evaluating the coefficients of a Walsh series, a Walsh spectral analyser has only to accumulate sampled values of the signal. The multiplications involved in Fourier analysis are eliminated, thus simplifying the hardware and increasing the speed. A digital Walsh spectral analyzer, to generate the first 64 coefficients of a periodic signal, has been designed and built.