694 



THE BELL SY8TEM TECHNICAL JOUKNAL, MAY 1954 



Z 7 



2 4 6 8 10 12 14 16 18 20 22 24 



CC. LOSS PER COUPLING UNIT IN DECIBELS 



Fig. 34 — Overall loss to the undriven line versus loss per coupling unit, with 

 the number of coupling units as a parameter. 



the driven and luidriven lines respectively are 



E — cos [ui sin^ a\ + rii sin~ 0:2 + • • • Uk sin aj , (39) 

 and 



V = sin [jii sin~ ai + n-i sin~ a-i -\- • • • Uk sin~ a/] . (40) 



These are amplitude factors due to coupling, and the normal attenuation 

 effects in the uncoupled lines must be added separately. For complete 

 power transfer we set the bracketed quantity of (39) and (40) equal to 

 7r/2, which gives the desired information about number and strength of 

 point couplings. Other transfer losses may similarly be prescribed or 

 determined. 



For multiple coupling units of the same coupling strength, Fig. 34 

 shows the overall transfer loss to the undriven line versus loss per cou- 

 pling units as a parameter. The shape of these curves from the complete 

 transfer point toward higher losses is very nearly the same. Fig. 35 shows 

 the loss per coupling unit versus number of coupling units, with overall 

 transfer loss to the undriven line as a parameter. 



SOME RESULTS OF EXPERIMENTS IN DOMINANT-MODE WAVEGUIDE 



In a pre^'ious paper on dominant-mode waveguide directional cou- 

 plers, complete power transfer between dominant-mode rectangular 



