70G 



THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1954 



44 



4 2' 



40 



38 



36 



34 

 en 



UJ 32 



m 

 u 30 



Q 



Z 28 



in 26 

 O 



^ 24 

 ct 



UJ 



U. 22 



CO 



z 



< 20 



cr 



18 

 16 

 14 

 12 

 10 

 8 



3 4 5 6 8 10 20 30 40 



NUMBER OF COUPLING HOLES 



Fig. 43 — Transfer losses versus freciuencj^ for the transducer of Fig. 42. 



circular electric wave but did cut off some of the waves which could 

 propagate in the round guide of the transducer itself. The measured 

 transfer loss under these conditions is recorded in Fig. 43. It is seen that 



the TE 



n _ 



TEoi° coupling was so weak as to be in the region where 



power from successive coupling elements should add inphase all the way 

 up to 40 coupling elements. The observations show the inphase addition 

 for less than 30 coupling elements but show a marked deviation in the 

 vicinity of 40 to 66 coupling elements. This is evidence of inequality of 

 the phase constants for the TEoi° and TEio° waves. More will be said 

 about this matter presently. The transfer loss between the rectangular 

 waveguide and the TEn mode of round waveguide, is also recorded in 

 Fig. 43. As expected, the power from successive coupling elements did 

 not add inphase and no appreciable build-up of power in the TEu mode 

 took place. 



One way of evaluating the total power in all modes other than the 

 circular electric mode, is to measure the value of the transverse magnetic 

 intensity at the wall of the round waveguide. The circular electric wave 

 has no such field component and all other waves do possess such a field 



