A MULTIPLE UNIT STEERABLE ANTENNA 369 



detector. This figure illustrates the manner in which the pattern is 

 built from the six components after the manner of a Fourier synthesis. 

 The vertical and horizontal axes visible on the monitoring oscilloscope 

 in Fig. 16, but which do not appear naturally on the photographs, 

 were drawn in Fig. 22. As mentioned previously, the oscilloscope 

 sweep axis represents one revolution of the "fundamental" phase 

 shifter so that the beginning and end of the sweep represent the same 

 condition. The ends of the sweep are arbitrarily fixed to represent 

 zero (or 360) degrees of phase shift referred to the output of the first 

 antenna, whose phase is not varied. Consequently equiphase inputs 

 result in a principal lobe half of which appears at each end. This 

 would correspond to a wave of zero angle if the velocity of the trans- 

 mission lines was equal to that of light. For a lesser velocity, zero 

 angle may occur at any point on the phase axis, depending upon the 

 wave-length. (See Fig. 28 for a sample angle calibration curve.) 

 The principal lobe as well as the four minor lobes of the monitoring 

 oscilloscope represents the output from one wave as the MUSA is 

 steered through its entire range. The oscilloscope pattern, unlike the 

 directional pattern, does not appear sharper for short wave-lengths 

 than for long wave-lengths; the principal lobe is always 120 degrees 

 wide and the minor ones 60 degrees wide on the phase axis. One 

 degree of phase difference, however, represents a difference in steering 

 angle which depends upon the wave-length and the earth angle. 



The samples of motion picture film shown in Fig. 23 represent fairly 

 typical "two-path " patterns. The camera was focused to include both 

 oscilloscopes and was manipulated by means of a special step-by-step 

 crank. The operator endeavored to expose each frame during one 

 sweep of the monitoring tube. The delay indicator tube shows a con- 

 tinuous pattern produced by the audio frequencies. A correct delay 

 setting is indicated by a straight line. Here, with the two branches 

 steered at the indicated angles of 8.5 and 20.5 degrees, a delay of 950 

 microseconds was required to produce the straight line. The diversity 

 action is apparent in the tilting of this line. When the low angle wave, 

 which corresponds to the left-hand peak on the monitoring tube, is pre- 

 dominant the delay indicator line becomes horizontal and, conversely, 

 when the high angle wave is predominant the line approaches the 

 vertical axis. Automatic gain control is used on the branch receivers 

 supplying the speech outputs but is not used on the monitoring 

 branch. 



Figure 24 shows, in samples 1 and 2, reception of two waves which are 

 just separable by the directivity present in the Holmdel MUSA. The 

 angles are 15 and 22 degrees and the wave-length is 31.6 meters. The 



