634 BELL SYSTEM TECHNICAL JOURNAL 



With the reaUzation that sufficient data are not available to define the 

 phenomena in all detail, it is believed that certain general conclusions can 

 be drawn. From the present work and that of Lavoo^^ and others^'''^*'^', it is 

 ai)parent that the microwave conductance of a close-spaced diode is sub- 

 stantially greater than the low-frequency value. The ratio g/go appears to 

 increase as the spacing decreases. This increase will probably continue until 

 the position of the potential minimum approaches the anode plane. The 

 susceptance decreases with increasing current and appears to level off at 

 high-current densities. The fmal value at a current density of 240 ma cm- 

 varied between 0.5 and 0.9 of the initial value. 



For a given current density, the ratio g/go does not appear to vary ap- 

 preciably as the cathode temperature is changed. 



An attempt was made to study the available diodes at 10,000 megacycles. 

 It was found, however, that the value of Rs was so high at this frequency and 

 that variations in tube conductance were so small in comparison with Rs 

 that accurate results could not be obtained. 



V2Lj,2 V, LJ21 



I I 



Fig. 14 — Equivalent circuit of a triode. 



Four-Pole Admittances of a Triode 



A triode may be considered as an active linear four-pole transducer, and 

 may be defined by the network of Fig. 14. It is apparent that 



yii is the input admittance with the output shorted, 



y22 is the output admittance with the input shorted, 



yn is the feedback admittance with the input shorted, 



yn is the transadmittance with the output shorted. 



The values of the parameters yn , y22 , yn , and y-n to be measured at the 

 grid, cathode, and anode terminals differ from the values of the y admit- 

 tance coefficients given by Llewellyn and Peterson^ who define yn as the 

 admittance of the diode coinciding with the cathode and the fictitious equiva- 

 lent grid plane, and y^o as the admittance between the equivalent grid plane 

 and the anode, and finally y^i as the transadmittance between the two. The 

 relations between the y admittance coefllcients of Llewellyn and Peterson 

 and the coefficients measured by the author are given by Peterson.^ It 

 turns out that, with a high-mu tube, such as the 1553 triode, the two sets of 



