DESIGN FACTORS OF THE 1553 TRIODE 515 



Vith lower values may have excessive input circuit loss or may have 

 narrower bandwidth on the input side than has been assumed. Further 

 measurements, by elucidating this point, might lead to a better design of 

 tube and circuit. 



An entirely similar calculation can be made for the power-band product, 

 rhe additional assumptions required are that the compression function 

 F-{C) has the conservative value of |, and the output coupling coefficient 

 Fo((?2) is taken as 0.9. The power-band product at 4000 megacycles is then 

 computed to be 50 watt megacycles, which is quite close to the figures 

 found by Bowen and Mumford. 



Refinements of the Electronic Theory 



In the electronic computations above, the single-valued theory was used 

 because it is the simplest theory which describes the high frequency case 

 at all accurately. The most important discrepancy between the rigorous 

 theory and the actual situation is the first theoretical assumption listed 

 above, that the electrons are emitted from the cathode with zero velocity. 

 For actual cathodes the velocity of emission is not zero nor uniform but 

 has a Maxwellian distribution such that the average energy away from 

 the cathode is ^ ^ T*, or about equivalent to the velocity imparted by a 

 potential drop of 0.04 volt for an oxide cathode at 1000°K. There result 

 several efifects whose general nature is known but which have not yet 

 been formulated into a rigorous quantitative theory valid at high fre- 

 quencies. 



(1) A potential minimum is formed at a distance on the order of 

 .001" in front of the cathode instead of at the cathode as in the 

 simple theory. This distance is not negligible for close-spaced 

 tubes; so that, for very close spacings, even perfect "physicists' 

 grids" approach a finite trans-conductance limit, [van der Ziel, 

 Philips Research Reports 1, 97-118 (1946); Fig. 2.] 



(2) Because the potential minimum implies a retarding field near the 

 cathode many electrons emerging from the cathode are forced to 

 return to it. These returning electrons absorb energy from the signal 

 and also induce excess noise in it, both effects becoming important 

 at high frequencies. 



The effects of initial velocities on the figures of merit can be measured 

 experimentally. For example, the circuit and electronic impedances of 

 diodes and triodes at 4000 Mc have been measured by Robertson.* Such 

 measurements can determine the electronic loading and noise separately 

 from the circuit degradation effects and are therefore a highly effective 



* loc. cit. 



