Measurement distance effects on Bayliss difference patterns

The effects of measurement distance in distorting low-sidelobe difference patterns are examined. Previous calculations have used obsolete suboptimum aperture distributions. The Bayliss linear distribution is a versatile, highly efficient and robust optimum distribution; its use allows a single curve of sidelobe measurement error versus measurement distance (normalized to far-field distance 2D/sup 2// lambda ) for a given sidelobe level. Data are given for patterns from a uniform distribution to a 50-dB Bayliss. Difference patterns require slightly larger measurement distances than sum patterns. For example, the first sidelobe of a 40-dB Bayliss pattern is in error by 1 dB at a distance of 7D/sup 2// lambda . The results should apply approximately for circular apertures as well.<
>