THERMIONIC VACUUM TUBES 



55 



Then, imaging for the moment that the condenser C\ has been re- 

 moved, the output voltage e 2 of the transformer is 



emp 



e<i 



rp+jhp 



(8) 



in which e is the input voltage, h is the inductance of the primary 

 winding, m is the mutual between the windings, and p is 2tt times the 

 frequency. This neglects resistance of the winding and also capacity 

 effects. Inspection of Equation 8 shows that e% varies with the fre- 



T'equc^y f Cyt lei per 5earo 



Fig. 26b — Curve A corresponds to 1 = 1 henry. Curve C to 1=2 henries, C 1 = .141 /. 



quency or with p in the manner shown in curve A, Fig. 26b, from 

 which it is seen that the transformer tends to suppress the lower 

 frequencies. 



Curve A shows the performance of a transformer as calculated from 

 Equation 8 assuming r p = 5,000 ohms and h = l henry. Such a trans- 

 former would be quite unsuited for a speech frequency amplifier as it 

 introduces very serious distortion below 1,000 cycles. Curve B is 

 calculated on the assumption that / 2 = 2 henries and shows marked 

 improvement over A for the lower frequencies. 



In input transformer design it is ordinarily necessary to limit the 

 inductance of the two windings not only because of the limited winding 

 space but also because of the need of keeping down the capacity be- 

 tween windings and the capacity within each winding. Curve C 

 shows the performance of the same transformer as Curve B when 

 the capacity C\ (Fig. 26a) is put in the primary circuit, C\ having a 

 value of .141 m.f. and being so chosen as to tune l\ to 300 cycles. With 

 the capacity present Equation 8 becomes 



ei 



emp 



r > +j ('' p -ctp) 



(8a) 



Use of the capacity improves the transformer characteristic for all 

 frequencies above about 200 cycles and the combination therefore 

 gives better results in a speech amplifier than the transformer alone. 



