698 



THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1954 



abstracted from the driven line it is clear that the theory given above 

 does not apply. Dawson has obtained experimental data of interest in 

 this connection. For a GX^ long linear-taper aperture of the form given 

 above (0.33" height at the center and 0.030" height at the ends), the 

 loose coupling theory predicts directivities in excess of 45 db for the 

 wavelength band 3.1 to 3.5 cm. When using sufficient number of wires 

 to obtain 18 db transfer loss, directivities in the range 36 to 48 db were 

 observed. The reason for the 36 db observation being loAver than the 45 

 db theoretical value may be inaccuracy of fabrication (jig per Fig. 36) 

 or inapplicability of the loose coupling theory at 18 db transfer loss. At 

 3 db transfer loss, the observed directivity of a similarly shaped but 5.5 

 Xg long coupling array is shomi at the top of Fig. 38; again loose coupling 

 theory predicts more than 45 db directivity. The reason the observed 

 values are in the 24-33 db range rather than above 45 db is presumed to 



5 40 



O 20 



k-2\g^^ |-«-2\g^ 



3.0 3.1 3.2 3.3 3.4 3.5 3.6 



Aq-cm 

 Fig. 3S — • Directivity and return loss for two coupling distril)utions, each of 

 which produced 3 db transfer loss. 



