754 THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1953 



lished its value as, 



= 1.12 mc/volt. 



df 



dV, 

 Substitution into equation (4.2) then gave the value of primary Q as, 



Q = .(3.8)^10^^ ^|Sy£- " "^^-^^^ = ^^^' 



The entire Q-determination, incidentally, as outlined above, was re- 

 peated for the 2 + % mode and resulted in Q = 128.8 thus affording an 

 excellent and independent double check. 



4.3.2 Calibration of Secondary Resonator. To facilitate the establish- 

 ment of controlled and reproducible conditions of secondary Q and 

 coupling coefficient, two cahbration curves had to be obtained. One, 

 relating the values of secondary Q with the readings of the micrometer 

 controlling the depth of insertion of the resistance vane and the other, 

 relating the coefficient of coupling, /c, with the coupling iris width. The 

 latter could be varied by means of gold plated spring shutters as shown 

 in Figs. 14 and 15. 



This calibration of the secondary cavity was carried out in three 

 distinct phases. The first phase involved the determination of the varia- 

 tion of Q, with micrometer setting for values of Qa ranging from 500 to 

 2000. The second phase, which was based on the results obtained in 

 phase one, yielded the complete calibration of the coupling iris and the 

 third and last phase, in turn dependent on results of phase two, yielded 

 values of Q, down to 65. The reasons for this particular sequence of 

 measurements will become apparent in the following more detailed 

 description. 



By means of a Q-measurement technique^ based on an oscillographic 

 display of reflected power, points on the calibration curve were obtained 

 as indicated by the circles in Fig. 18. It is seen that the lowest value of 

 Q, which could be determined by this method was 550. For lower values 

 of 0, the sweep range of the signal generator became insufficient to 

 display the required fraction of the resonance curve; in addition, the 

 cavity proved excessively undercoupled to permit reliable measurements 

 of bandwidths. 



Using the values of Q, thus determined and a particular property of 

 coupled resonators covered earlier in this text and further elaborated 

 below, the relation between coupling coefficient and iris width was 

 established. It was shown earlier that an overcoupled secondary cavity 

 gives rise to load hysteresis manifesting itself in mode shapes as sketched 



