292 



BELL SYSTEM TECHNICAL JOURNAL 



resulted in the 5J26 magnetron. It dictated the nature of the interaction 

 space and resonator system and thus the cathode size and output coupUng 

 loop. Initial experiments were performed with a vane type resonator 

 system [see Fig. 21(b)] having channel straps and an interaction space of 

 approximately the size used in the 4J21-30 series. 



From studies of dynamic V-I plots on the experimental tunable models 

 and other magnetrons under a variety of operating conditions, it became 

 clear that the phenomenon had to do with the rate of buildup of RF oscilla- 



Fig. 57— The 5J26 tunable magnetron (600 k\v., 1220 to 1350 mc/s)— the tunable re- 

 placement for the 4J21-30 magnetrons. 



tion in the tt mode and with the rate of rise of the applied DC voltage. The 

 "mode skip" shown in the sequence of V-I plots of Fig. 38 and discussed in 

 PART I is an example of the behavior of one of the experimental models 

 used in this work when failure to oscillate in the tt mode occurs. 



Thus at constant magnetic held there appears to be a maximum DC volt- 

 age above which w mode oscillation is impossible. This voltage is presum- 

 ably that beyond which the mean angular velocity of the electrons around 

 the interaction space becomes enough greater than that of the traveling field 

 component for the phase focusing to fail to maintain synchronism. The 

 term "maximum current hmitation" is thus in a sense a misnomer as Fig. 

 38(b) indicates. Here, ir mode oscillation fails even to start and does not 



