MAGNETRON AS GENERATOR OF CENTIMETER WAVES 325 



was largely determined by the state of development of systems and com- 

 ponents when the work began. It was desirable to produce a magnetron 

 which, while marking a very substantial gain over the 725 A in power output, 

 should at the same time be of reasonable weight, air-cooled, and capable of 

 operation with a pulser of modest dimensions and of working into the 

 system components in existence or under development. 



The design began with the choice of a resonator system having sixteen 

 resonators. The anode diameter was chosen for an operating voltage range 

 of 20 to 25 k\^ The increase in number of resonators from twelve to six- 

 teen and operation at higher voltage necessitated a considerable increase in 

 cathode diameter over that of the 725A. The length of the anode was left 

 as before because it was felt that the desired increase in total cathode emis- 

 sion would be available from the increase in cathode area and that the mag- 

 net weight ought to be kept as low as possible. 



Experimental models incorporating these design changes were built in 

 the 725A type structure with radial cathode mount, "halo" coupling loop 

 and coaxial output terminating in a junction to wave guide. The per- 

 formance of these early models was satisfactory. Operating efficiency 

 better than that of the 725A mgnetron was achieved over the range of 

 currents from 4 to 40 amps. 



However, it was easily to be seen that the 725A type structure was by 

 no means the ideal for a 3 cm. magnetron of increased output power. The 

 output circuit was marginal in its power handling capabilities; it could 

 transmit no more than 300 kw. Furthermore, the opportunity presented 

 itself of eliminating other troublesome features such as the "halo" loop and 

 the coaxial to wave guide junction by designing a wave guide output whose 

 critical dimensions are machined to size. The type of cathode structure 

 of the 725A was limited in its heat dissipation in comparison with the axial 

 type of cathode mount made possible by the use of magnet "packaging". 

 In addition, the axial type cathode mount is superior from the standpoints 

 of DC voltage breakdown and economy of space. Considerations of weight 

 also favored the "packaged" structure, in which the magnet pole pieces 

 form an integral part of the magnetron structure. Many of these first 

 experimental models suffered from a drooping voltage-current characteristic 

 at constant magnetic field in the region of low currents. This effect and 

 its attendant loss of operating efficiency became progressively worse as 

 the magnetron was operated. It was hoped that the parasitic electron 

 emission from the cathode end disks, believed from experiments with non- 

 emissive coating materials to be responsible for the effect, could be elimi- 

 nated or reduced in an axially mounted cathode. 



Accordingly, an entirely new design was undertaken, built around the 



