DIRECTIVITY STEERING FOR FADING REDUCTION 209 



ANGLE IN DEGREES 



17 20 17 20 



HALF-WAVE VERTICAL 

 (^=69°-7l° (/) = 7l°-73° 4,= 73°-76' 



RHOMBUS 



RHOMBUS A 



Fig. 16 — Pulse pattern changes with steering, April 8, 1933, 

 2:40 P.M., E. S. T. Station GBW on 20.78 meters. 



the selective fading worse only four per cent of the total time of ob- 

 servations. 



Occasionally, and in particular on twenty meters, only slight selec- 

 tive fading was observed. When pulse transmissions were available 

 during these times, only one major pulse could be seen. Really bad 

 fading invariably occurs when multiple pulses, which are widely spaced 

 in time, are observed. 



It may be evident, from the previous discussions in this paper, that 

 the change in antenna output, with steering, is closely related to the 

 number and spread of the waves arriving and to the selective fading 

 improvements obtainable. Figure 17 shows three cases of results se- 

 cured by reading relative gain changes, as shown by automatic 

 recorders. 



Case 1 is typical of a closely spaced wave cluster arriving at an 

 average angle of about ten degrees above the horizontal . Case 2 can be 

 explained as due to a narrow wave cluster at eleven degrees plus another 

 of less amplitude at eight degrees. We would ordinarily expect annoy- 

 ing selective fading in such an event. Should we deal with many closely 

 spaced waves having a large angular spread, very little gain change 

 would be evident while steering the rhombic antenna, but selective 

 fading improvements over the comparison antenna might still be pos- 

 sible. 



Curve 3 is of considerable interest in that it served as one of the 

 experimental checks of the theoretical directive pattern calculations. 

 The change in gain with steering is so well defined that probably only 

 one wave-direction existed. This belief was supported by an absence of 

 noticeable fading. Independent measurements, made by an average 

 angle measuring installation^ consisting of two horizontal dipoles at 

 different heights which determines the average angle by the ratio of 

 the respective outputs, gave the arrival angle at from eighteen to nine- 

 teen^degrees above the horizontal. Figure 4 indicates that a <^-angle 



