SINE WAVE OSCILLATORS 



amplifier gain of 3 and oscillates at co = 1//?C . The phase shift oscillator 

 requires a gain of about 8 and oscillates at ll(3y^^RC if a is about 10. If 

 a = 1, the gain required is 29 and co = \l{()fl^RC. 



HT+ 



HT-t 



(a) 



(b) 



Figure 14.29 



Arrangements for maintaining the level of oscillation have to be more 

 elaborate than the simple self-biasing device which is adequate for LC 

 oscillators. This is because the RC frequency-determining networks have 

 of themselves only a very small Q value ; one could not possibly tolerate the 

 pulsatory currents which are permissible in LC oscillators because there is no 

 device with a g of several hundred to re-create a pure sine waveform. Thus 

 the valves must amphfy in class A, without distortion, and some other 

 means must be found of regulating the gain. 



D' Argidmbeau' s method — There is a variety of pentode known as 'variable 

 fi' {Figure 14.30). As a result of grading the mesh size of the control grid 

 along the axis of the valve it can be shown that the ampHfication obtainable 

 from the valve is a function of the mean grid bias {Figure 14.31). Clearly if the 



input signal is large enough to sweep over an appreciable proportion of this 

 curve, the output of a variable /n stage is distorted. However, if the signal is 

 small enough only a short piece of the curve is used and this may be regarded 

 as being straight. Use of this fact may be made in the Wien bridge oscillator. 

 In Figure 14.32, part of the output is tapped off and shunt rectified by the small 

 diode in the second valve to produce a negative bias which controls the gain 

 of the first valve. The capacitance of C^ has to be sufficiently large for its 

 reactance to be negligible compared with the frequency determining R's and 



222 



