8.3 Wideband Scanning Systems

Conventional phased arrays operate over bandwidths that are inversely proportional to the array size. The use of true time delays instead of phase shifts would eliminate the bandwidth restriction due to beam squint, but unfortunately the only viable time-delay technology at the time of this writing consists of switched sections of transmission lines. For example, a large array of 50 wavelengths on a side and scanning to 60 would need a total time delay from zero to 50 sin 60 ,or43l. To obtain precision equivalent to an N-bit phase shifter, about N + 6 bits is necessary. If these units are made with discrete time-delay bits, as is the common practice for phase shifters, the units become too bulky and heavy, and so lossy as to be impractical for most applications, except perhaps for stationary ground-based arrays at relatively low frequencies. The solution to this problem is to devise suboptimum means of providing time delay. Two architectures have been used. The first method consists of matching the time delay at only a fixed number of angles and using phase shifters to scan the array over the small scan ranges between the selected angles of perfect delay. The second method is to divide the array into subarrays and produce true time delay behind each subarray while using phase shift within the subarrays.