COITPLKD WAVK THKOHY A\n WAVK(iriDK Al'PI.ICATIONS 



r()7 



comjioneiit. Thus tlic total value of the t I'ansx-crso magnetic intensity at 

 the round \va\ei»;ui(le wall is a measiu'e of the impurity associated with 

 the circular electric wave. (This is very similar to the radial probe t(H'h- 

 niciuc described by 'SI. Aronoff.') Using this method of e\alualioii, the 

 mode impurities ])resent at the output of the transducer were measured 

 as a function of the number of coupling elements, and the results are 

 recorded in Fig. 44. The absolute calibration of the ordinate relates the 

 observed magnetic intensity to that which the same power input used at 

 the rectangular guide would have produced if placed in the round wave- 

 guide in the TEn mode. These measurements show that for all of the 

 modes other than the circular electric mode, the energy components 

 from successive coupling elements suffer destructive interference. Al- 

 though curves are shown only for one and for 66 coupling elements, the 

 patterns for intervening numbers of coupling elements were similar in 

 shape and never exceeded an intensity value greater than about 6 db 

 above that given for the 66 coupling element case; thus the mode dis- 

 criminating property of the coupled wave transducer was verified ex- 

 perimentally. 



Returning to the question of TEio° — TEoi^ transfer loss, it is clear 

 from Fig. 43 that the rectangular Avaveguide has a phase constant which 

 is not equal to that of the circular electric mode in the roimd waveguide. 

 One reason for this inequality lies in the fact that the coupling elements 

 disturb the phase constant in the two waveguides unequally, a conse- 

 quence of the fact that some of the power transferred to the round wave- 



5 25 



PtU 



60 120 160 200 240 



AZIMUTHAL ANGLE IN DEGREES 



Fig. 44 — Distribution of transverse magnetic intensity at the wall for the 

 transducer of Fig. 42. 



