EQUIVALENT CIRCUITS OF NETWORKS 601 



as expressions for current equilibrium at any frequency. The four param- 

 eters (3 which appear have simple physical meanings, and it is seen that 



^n is the input admittance with output shorted 

 — i822 is the output admittance with input shorted 

 — jSi2 is the feedback admittance with input shorted 



i82i is the transfer admittance with output shorted. 



Before proceeding to the network representations of (14) it seems well 

 to state briefly some of the reasons which almost force one to adopt the four- 

 pole point of view when dealing with vacuum tubes in the higher frequency 

 range. 



There are the pedagogical reasons that classical methods long employed 

 for passive networks are merely extended into the realm of active networks, 

 thus providing unity in viewpoints. 



The basic analysis involving the four-pole parameters for a particular 

 transducer needs to be performed only once and, once obtained, all problems 

 involving terminal impedances may, in any particular case, be solved in a 

 routine manner. 



There are also further practical reasons. We saw above that with in- 

 crease in frequency the classical equivalent network had to be modified to a 

 considerable extent in order to include the parasitic elements. This poses 

 a serious problem for the tube designer, whose task it is to estimate the tube 

 performance between known terminations. Such a task based upon the 

 modified classical circuit becomes very dijQ&cult and cumbersome. More- 

 over, it is also difficult to segregate and measure the parasitic elements. 

 Hence it appears that one could gain much if design parameters could be 

 developed capable of reflecting parasitic and transit time effects. Finally, 

 it is desirable to develop equivalent circuits with a minimum number of 

 parameters bearing simple relationships to quantities which can be measured 

 directly. 



These general desires arising from the practical needs of the tube designer 

 can be satisfied if tube behavior is specified by means of four-pole parameters. 



All in all the four-pole point of view can be made to satisfy the logical 

 needs of integrated concepts as well as the practical needs of simplicity in 

 the specification of tube performance. 



After this brief presentation of the argument for the four-pole point of 

 view, the network representation of (14) will now be considered. 



Stated in broadest terms: we are seeking a network representation by 

 considering the two equations as a single unit and not by the trivial considera 

 tion of each equation by itself. 



