THE MICROWAVE GYRATOR 5 



as indicated in Fig. 3 and then a gravitational force is suddenly made to 

 act along the y axis so that the net gravitational force acts along A, 

 it is obvious that the gyroscope will begin to preccss about the gravita- 

 tional field direction as indicated by the small dotted circle. However, if 

 after completing a half cycle, the horizontal component of the gravita- 

 tional field is reversed so that now the net gravitational field acts along 

 the vector B, the gyroscope will begin to precess about B as indicated 

 ])y the intermediate size dotted circle. If the horizontal component of 

 the gravitational field is again reversed after the gyroscope completes 

 another half cycle in its precession, the gyroscope will again begin to 

 precess about the direction A and the actual path of the precessional 

 motion will be along the path a-b-c-d. If this process is continued in- 

 definitely, the gyroscope will precess in larger and larger circles around 

 the vertical until the damping becomes large enough to contain the 

 gyroscope in some equilibrium circle (assuming that the damping is 

 large enough to accomplish this). 



The above model affords a classical picture which can be used quite 

 readily to describe the motion of the electrons in a ferrite. If the ferrite 

 is initially saturated along the z axis by a steady magnetic field, the 

 electrons will come to rest with their magnetic moments lying along 

 the axis, as the gyroscope in Fig. 3. If now an alternating magnetic 

 field is applied along the y axis, the electrons will begin to precess in 

 larger and larger circles about the g axis until they finally reach some 

 equilibrium position under the influence of the magnetic fields and the 

 damping. Thus it is apparent in the gyromagnetic resonance experi- 

 ments described above why an alternating field applied perpendicular 

 to a steady magnetic field in a ferrite will give rise to a varying flux 

 perpendicular to both the steady field and the alternating field. It is 

 also apparent why the alternating flux along the x axis is 90° out of 

 phase with the alternating flux along the y axis. Since precession of the 

 top will always be in the same direction regardless of whether the alter- 

 nating field is applied along the x or y axes, consideration of Fig. 3 

 makes it apparent how the two crossed coils with ferrite at their center 

 can constitute a gyrator which violates the reciprocity relation in a 

 manner described by Equations (1). To the present time, however, no 

 practical circuit element making use of this phenomenon has been 

 constructed because the coefficient of coupling between the coils is 

 always small, even in the vicinity of the resonant frequency, and also 

 because the losses in the materials available are so high in the vicinity 



