644 



BELL SYSTEM TECHNICAL JOURNAL 



The remaining parameters for the triode of Fig. 21 are very similar 

 to those of Fig. 20. 



Figure 22 shows the variation of the phase of the transadmittances yn 

 for the two triodes. The figure also shows the theoretical curve of the 

 Llewellyn analysis for purposes of comparison. As in the case of Fig. 13 the 

 abscissa do not represent transit time for the experimental values. The 

 quantity x is equal to the cathode-grid spacing. 



It is of interest to compare the triode measurements with those of the 

 diode. It was expected that gn for the triode should correspond with g for 

 the diode. Within the limits of reasonable experimental accuracy this 

 appears to be the case. For the triode at low frequencies go ^ gm- The triode 



5 

 (26 X 



1/3 



Fig. 22 — Phase of triode transadmittance. 



results indicate that the ratio gn/gm is quite comparable in magnitude with 

 the corresponding ratio g/go for the diode. This was expected. The behavior 

 of ill for the triode was unexpected. It was thought that, as the grid voltage 

 was varied so that the input space changed from a condition of zero space 

 charge to one of maximum space charge, hn would vary from its initial 

 "cold" value to a value approaching 60% of the latter. This was not so. 

 In the figures one observes that it drops to a much lower value. This effect 

 has not been explained from a theoretical standpoint. There are several 

 qualitative interpretations, but as yet no way of determinmg which of them 

 is correct m a quantitative sense has been found. The observed phenomenon 

 could, for example, be explained by an increase in the effective series resis- 

 tance of the tube caused perhaps by an increase in the resistance of the 



