Phase characteristics of spiral antennas for interferometer applications

Several recent applications of the flat spiral antenna require a well defined phase center as well as a very low ellipticity ratio of the polarization over wide angles of the radiation pattern. One example of this requirement is in the Gemini radar which determines angles by an interferometer measurement with circularly polarized fields. The phase shift for the measurement is obtained by antenna rotation. Thus a precise phase center is required which is independent of antenna rotation. Another example is in the scanning arrays produced by the rotation of the spiral antennas as array elements. The model for the investigation is an attenuated traveling wave upon filamentary conducting arms of a logarithmic spiral. The model is tested by showing that it gives radiation patterns close to measurements by Dyson. It is then used to calculate phase variation across the far field radiation pattern. The phase error of antennas driven from an ideal balun is first found as a function of spiral growth rate. For growth rates, "a," of 0.1 and slower, a phase center may be defined which is behind the spiral approximately "a" radians. The slower the growth rate, the better defined the phase center becomes. At a growth rate of 0.03, which is about as slow as can be fabricated, a phase error of 0.003 radians is found 30 degrees from the polar axis. This is satisfactory for most interferometer applications. The tilt in the amplitude pattern of the spirals with an imperfect balun (and in particular with no balun at all) is well known. In addition, a phase error occurs under these conditions which places a requirement on the unbalanced mode rejection of a balun for spirals having high phase accuracy requirements. It is found that this type of error is also less for spirals of slow growth rate. For the slowest rate practicable, however, it is found that a 30 db unbalanced mode rejection provides for no more than 0.015 radian error 30 degrees from the polar axis. Because the balun perfect- on is difficult to measure directly, it is suggested that the tilt in the amplitude pattern be used as a measure of this type of phase error.