Abstract :
It is well known that radiogoniometers of accepted design possess inherent errors due to the coupling between the search-coil and the field-coil windings, and that the magnitude of these errors increases with an increase of this coupling. In practical design it has therefore been usual to effect a compromise between the tightness of the coupling and the permissible errors. A special study of the subject has therefore been made with the object of analysing the causes of these errors and of introducing modifications which, while providing a tighter degree of coupling to the search coil, will at the same time reduce the magnitude of the instrumental errors. A preliminary examination of the various existing types of radiogoniometer shows that the errors are not uniformly ¿octantal¿ in form, and that the value of the maximum error is not necessarily directly proportional to the tightness of the coupling. Several methods of measuring the instrumental errors are described, and further information regarding the causes of apparently non-symmetrical errors is recorded. These are shown to be due to the vector addition of a number of separate errors all harmonically related in frequency but differing in phase relationship. A final summary, in tabulated form, shows the more important results obtained at various stages during the development of a new design of radiogoniometer which is evolved and shown to be at least 6 db better in signal/noise ratio than any existing design, while at the same time possessing a maximum instrumental error of less than ±¿°.
