ADMITTAIVCES 01' IW RM.l.EI. PLAN E ELECTRON TUBES 



635 



coefficients (liller in the order of 10-20% over the useful operating range of 

 current densities; so, for j)ractical considerations, the measure coefficients 

 may he regarded as substantially equivalent to the coefficients referred to 

 the fictitious grid plane. Xot that they will be equal to the theoretical values, 

 but they may be regarded as being associated with the same geometry and 

 will serve at least as a qualitative test of the validity of the theoretical values 

 for the physical tube. 



In order to measure the four-pole [)arameters, the 1553 triode was mounted 

 in a coaxial circuit of the type shown in Fig. 15. The grid-anode output 

 circuit of the tube is seen to connect directly with the coaxial output line. 

 The input circuit required a more careful design. Due to the size of the base 

 of the tube it was necessary to taper the input coaxial as shown. In the early 

 stages of this work, difficulty was experienced with higher order modes in 

 the large diameter section of the input coaxial. It was believed that these 

 modes were generated by the action of the parallel wire grid which lacked the 



CATHODE-^ 



'ANODE 



INPUT LINE 



Fig. 15 — Detail of coa.xial mount for measuring four-pole admittances of a triode. 



radial symmetry appropriate to coaxial transmission. The difficulty was 

 overcome by constricting the outer diameter of the coaxial line in the im- 

 mediate vicinity of the grid of the tube, thus inhibiting generation of the 

 higher order mode. 



Before measurements could be made it was necessary lirst to calibrate 

 both the input and the output circuits in a manner similar to that used and 

 described in connection with the diode measurements. The coaxial tester 

 used for calibrating the input circuit was identical with that used for the 

 diode work. For the output circuit a similar tester was use. As one might 

 expect, the value of the cot-cot slope of the output circuit was close to 

 unity. The value actually turned out to be 0.9. In the input circuit the slope 

 was so great that it was difficult to measure, so that it was necessary to 

 introduce a transformer in the coaxial input circuit to permit tuning. 



The complete apparatus necessary to measure Tn and yio is shown in 

 Fig. 16. This equipment, save for the details already discussed, is quite con- 

 ventional in every respect. 



