894 



THE BELL SYSTEM TECHNICAL JOURNAL, JULY 1953 



assure that all the important effects are sufficiently understood to make 

 intelligent control possible. 



Using the results of this analysis, the off-ground coupling network can 

 be represented by the equivalent circuit of Fig. 8(c), where the values 

 of the pi of admittances are obtained in terms of the fundamental 

 parameters of the transformer and associated elements. This representa- 

 tion is convenient for insertion gain and feedback computations. The 

 most important difference between the equivalent circuits of Figs. 8(b) 

 and 8(c), from a practical standpoint, is the presence in the latter of the 

 admittance Yicn which is a manifestation of the capacity to ground 

 from the high impedance winding and the junction of transformer and 

 peaking coil. 1 2 and the contributions to Yzcn not directly attributable 

 to the obvious shield-to-shield capacity of the transformer can be set 

 equal to zero with little error, but Yicn affects the insertion gain and 

 feedback by about 2 db. 



It will be observed that the transformer is shown with two shields, 

 one connected to the bottom of the high impedance winding, which is 

 the top of the beta circuit, the other connected to ground. The first of 

 these, which is physically adjacent to the high impedance winding, acts 

 to collect, and carry to the beta circuit, the distributed capacity of the 

 high ^vinding, thus avoiding intolerably large capacity from this winding 

 to ground. The second shield prevents capacitative coupling of the large 

 beta circuit signal voltage to the low impedance winding, which would 

 lead to a very poor reflection coefficient performance. Typical curves of 

 reflection coefficient are shown in Fig. 9. Since the amplifier is an active 

 device, the reflection coefficient is to some extent a function of the 

 vacuum tube transconductances, and tends to be degraded as tubes age. 



512 



6 

 6 

 4 

 P 2 



0.4 0.5 0.6 



0.8 1.0 

 FREOUENCy 



1.5 2 3 4 5 



IN MEGACYCLES PER SECOND 



Fig. 9 — Reflection Coefficients. 



