MAGNETRON AS GENERATOR OF CENTIMETER WAVES 297 



The variations in the resonator system of the British magnetron were 

 made primarily for the purpose of studying the parameters affecting fre- 

 quency of operation. Magnetrons were built incorporating such variations 

 as dimensional changes in the hole and slot resonators; the use of other 

 forms of resonators such as slot resonators; and variations of factors affecting 

 the coupling between adjacent resonators, as, for example, the volume of the 

 end space region between the ends of the resonator system and the end covers 

 of the magnetron. A series of magnetrons was built having from four to ten 

 resonators, including odd numbers, but the majority of the experimentation 

 was done with eight resonators as in the original system. 



Besides affecting frequency, most of these resonator changes markedly 

 affected the electronic operation as well. Undoubtedly this was the result 

 of changes occurring in the mode frequency distribution. At the time, the 

 signilicance of this distribution to magnetron operation was not fully appre- 

 ciated nor, indeed, was it known for these magnetrons. As discussed in 

 PART I in connection with curve (a) of Fig. 25, it was later found that the 

 frequencies of several modes including the tt mode were very nearly equal, 

 making the electronic operation quite sensitive to the exact nature of the 

 frequency distribution. This made the experiments difficult to interpret. 

 None of the resonator systems, involving only changes in anode length and 

 diameter with corresponding changes in the cathode, appeared to be appre- 

 ciably better than that in the British magnetron. 



In one rather important experiment the cathode diameter, and thus rjr^ , 

 was varied to determine the value for optimum efficiency using the British 

 anode dimensions. The result, confirmed later by comparison with British 

 work, showed the original dimensions to be very nearly optimum. 



Some variations of the output coupling to the resonator system were also 

 tried. In one such variation the coupling loop was placed in the end space 

 of the magnetron between two resonators. In this position the loop was 

 coupled magnetically by the flux linking the two adjacent resonators, as well 

 as directly by virtue of the fact that the end of the loop is fastened to the 

 anode segment (see the discussion of output coupling in PART I). It will 

 be recognized that this type of coupling is that used later in the 700A-D 

 magnetrons. The output circuits in the first American 10 cm. magnetrons 

 were like those in the British design. 



The constructional techniques used in the British magnetron were fol- 

 lowed with some variations. The most bothersome technique was that of 

 making the vacuum seals between the end covers and the body of the 

 magnetron. This was done by means of tin plated gold ring seals between 

 the members. Little difficulty was encountered with this seal in produc- 

 tion, however, and it was used throughout the production at the Western 



