450 



BELL SYSTEM TECHNICAL JOURNAL 



point of the tube is changed during manipulation for balance. 

 Everitt's modification,^ which consists in separating the direct and 

 alternating current paths by a retard coil and condenser in the usual 

 manner, is therefore preferable in this respect, but a complicating 

 factor enters in the introduction of a reactive component due to the 

 retard coil which cannot be balanced out by the variable resistances 

 originally provided. This may be taken care of in a more or less 

 obvious way by shunting a reactance around the grid resistance as 

 shown in Fig. 3, the effect of which is to correct for the introduced 



Af 



— HWWW^- 



R+j^ 



«^— A/WVW- 



Ru 



TO AMPLIFIER 



— e- 



^ 



Fig. 3 — Modification of Miller's method for determining ix and Ro 



phase unbalance due to the retard coil and so lead to a precisely 

 determinable null point instead of to a broad minimum, as is other- 

 wise found. 



The effect of the inserted reactance may be calculated by direct 

 methods. Referring to the figure, there is no potential of funda- 

 mental frequency across the amplifier used to indicate balance at the 

 null point, and if the grid-filament impedance of the tube is much 

 greater than R,j (50 ohms), the total oscillator current passes through 

 Rg and Xg in parallel, with R^, in series. The potential impressed on 

 the grid is then 



^g = JRgXgioJ(Rg -i-jXg), 



which appears in the plate multiplied by the amplification factor of 

 the tube and reversed in sign. The nomenclature is clearly indicated 

 in the Figure. An alternating current flows in the plate circuit which 

 is just balanced by the drop across R^ so that we may write 



jflRgXg 



R+jX 



= R. 



Rg+jXg Ro + R-\-jX 



^ See p. 201 of van der Bijl's "Thermionic Vacuum Tube. 



