CONSTANT FREQUENCY OSCILLATORS 



97 



equation may be written 

 1 



iZ ( a;Z/2 — 



C0C4 



— cjLi ( C0L9 -pr ) + ork^LiL-. 



(38) 



The condenser, d, was varied until the reading of the vacuum tube 

 voltmeter became zero. This means that Z was zero, and hence (38) 

 gives the value of r4 required by (37) which is in turn the value needed 

 to stabilize the oscillator. 



How well this checked the actual values needed for the case shown 

 in Figs. 20 and 21 may be seen by the following: For Fig. 21 the value 

 of d measured as above, was 4000 /i/^f. The experimental value was 

 3000 jUjuf. For Fig. 20 the measurement gave a rather broad zero on 

 the vacuum tube voltmeter, which was, however, estimated at 8400 

 ixni. For a check, a measurement was made at 7 megacycles which 

 gave a sharper zero, and a value of 120 /x/xf. This must be reduced to 

 its equivalent value of 1 megacycle by multiplying by 7^ which gives 

 5880 jjifxi. The experimental curves of Fig. 20 show a noncritical value 

 of 6000 p.p.i. which is nevertheless in good accord with the above meas- 

 urements, while in Fig. 21 the agreement is somewhat more striking. 



As an example of stabilization of oscillators in the altogether differ- 

 ent frequency region from 7 to 40 kc, the following table was taken 

 from data kindly supplied by F. J. Rassmussen: 



TABLE I 



This table again emphasizes the less critical adjustment required 

 when the coupling is tight, as in the last seven rows. 



It is hoped that the foregoing data and comments will serve as a 

 guide to design methods for constant frequency oscillators, in so far 

 as dependence of the frequency on operating voltages is concerned. 

 Combinations and permutations of the various circuits dealt with will 



