596 BELL SYSTEM TECHNICAL JOURNAL 



reduced by an amount corresponding to the back e.m.f. of the lamp. 

 Under these conditions the relation between current — and therefore 

 illumination — and the voltage on the grid of the vacuum tube is as 

 shown by the curve given with the figure. This curve takes into 

 account the change in potential between the plate and filament of 

 the vacuum tube due to the voltage drop in the lamp resistance. If 

 the reactances in the circuit are negligible, this curve may be taken as 

 the dynamic characteristic of this portion of the system. 



Let us assume that to properly build up the desired image at the 

 receiving terminal the light is to be varied between the limits set by 

 the two horizontal lines a and h. It is apparent that two adjustments 

 are necessary in the grid circuit. The amplitude of the impressed 

 alternating current must be such that the difference between its 

 positive and negative maxima is equal to the difference between the 

 grid voltages corresponding to these currents. This is taken care of 

 by suitable adjustments of the amplification. It is further necessary 

 that the bias introduced by the grid polarizing battery be such that 

 the positive and negative peaks coincide with these same values of 

 grid voltage. Under these conditions the grid battery must be looked 

 upon as supplying two absolutely distinct biases, one the bias for the 

 tube and the other the bias for the signal. For example, if the signal 

 wave as delivered to the grid circuit contained the original d-c. com- 

 ponent properly amplified, it would be necessary to adjust the system 

 so that zero current would be obtained with no impressed signal. 

 To accomplish this the tube would require the negative grid bias Eos- 

 Variations in signal voltage would then be considered as taking place 

 about this value of grid potential as the origin. Thus Egz is the 

 operating bias of the tube. To properly locate the signal wave, 

 however, it is necessary to add the positive bias Eg2- It will be seen 

 from the curve that this bias corresponds exactly to the direct current 

 component which is to be restored to the signal. The sum of these 

 two biases, obviously, gives the actual bias, Eai, with which the tube 

 is operated. 



In the circuit as shown the well-known curvature of the vacuum 

 tube prevents us from obtaining a linear relation between the current 

 through the neon lamp and the signal voltage. This condition may 

 be overcome by a number of circuit modifications of which that shown 

 in Fig. 27 is typical. Instead of connecting the neon lamp and the 

 vacuum tube directly in series, a resistance is provided across which is 

 set up a potential. Ex, proportional to the current through it. Across 

 this resistance is shunted the neon lamp and a biasing battery, Eb- 

 The adjustment of this circuit is indicated by the curves shown. 



