332 



BELL SYSTEM TECHNICAL JOURNAL 



Ka is known. If the reciprocal theorem holds, Ka = Kb and Ka can 

 be determined from (23). The optimum efficiency is then 



1 



'Jmax. 



1 + 



(26) 



If the input signal generator has an internal impedance, most efficient 

 energy delivery to the modulator will, of course, result if this im- 

 pedance is made equal to Rp. 



Equivalent T, -k and bridge networks can obviously be drawn from 

 the open and short-circuit measurements as in four-terminal linear 

 networks. 



It appears that even in a plate or grid circuit modulator the formulae 

 of equations (24), (25) and (26) can be applied to the plate or grid circuit, 

 respectively, where the signals are small compared to the carrier, 

 inasmuch as the modulating parts of these circuits are linear and 

 bilateral with no internal energy sources. 



Double-Balanced or Reversing-Switch Modulator ^ 



A number of interesting conclusions can be reached about copper 

 oxide modulators by assuming that the copper oxide acts like a switch 

 having a low-resistance value when the positive half-cycle of the 

 carrier voltage is across the disc and a high-resistance value during the 

 negative half-cycle. The circuits of Fig. 2(c), 2(d) or 2(e) can then be 

 represented by the equivalent circuit of Fig. 8. 



AA^- 



O 



(^ 



Vo 



POSITIVE 

 HALF CYCLE 



NEGATIVE 

 HALF CYCLE 



-wv 



Fig. 8 — Equivalent circuit of a double balanced modulator. 

 2 Referred to in the German literature as the "ring modulator." 



