MAGNETRON AS GENERATOR OF CENTIMETER WAVES 335 



In the development of the 3J21, the ''rising sun" resonator system used by 

 the Columbia Laboratory was adopted practically without change. The 

 ratio of the natural frequencies of the large and small resonators is approxi- 

 mately 1.8. The anode diameter is 0.160 in. and the anode length 0.190 in. 

 The structure is fabricated by the so-called "hubbing" technique,'-^ which 

 had been brought to a high state of development at the Columbia Laboratory 

 for the purpose. In this technique, a hardened steel die or hub, machined 

 to be the "negative" of the desired contour, is forced by high hydraulic 

 ])ressure (of order 250,000 Ib./sq. in.) into a copper blank. The hubbed 

 blank, after trimming and turning to size, becomes the resonator system 

 and body of the magnetron. The proper contour to receive the wave guide 

 output circuit is then bored. 



The cathode of the 3J21 magnetron has a diameter of 0.096 in. and a 

 length of active coating of 0.165 in. At 15 amps, peak current, the current 

 density at the cathode surface is about 50 amps./sq. cm. — a considerably 

 higher value than that in magnetrons of longer wavelength. Furthermore, 

 the back bombardment of the cathode in this magnetron is about 10 per 

 cent of the input power as compared to the 3 to 6 per cent in longer wave- 

 length magnetrons. For these reasons, the cathode was one of the severest 

 problems of the whole design. 



The first axially mounted cathodes were supported on Kovar tubes sealed 

 to the input glass structure. The heater lead was carried down the center 

 of this tube through a glass bead in the input end. The structure resembled 

 that of the 2J51 seen in Fig. 71. Since the hole through the magnetic pole 

 piece was initially 0.100 in. and the cathode end disks 0.130 in. in diameter, 

 the cathode was assembled onto its support after the Kovar tubing was in 

 place. However, in quantity production the cathode could better be 

 assembled as a unit before attachment to the outer glass and pole piece 

 structures. At the expense of magnetic field strength and uniformity, the 

 hole in the pole piece was enlarged to accommodate the entire cathode 

 structure. No ill effects were observed and no special features such as the 

 permendur cathode end structures of the 4J50 magnetron were found 

 necessary. 



The problem of the dissipation of the considerable heat of back bombard- 

 ment was complicated by the necessity of activating the cathode at a 

 temperature of 1050°C. The requirements of heat dissipation and activation 

 oppose one another, one calling for low thermal impedance of the cathode 



-* The technique is here called "hubbing" rather than bobbing for two reasons: The 

 term bobbing has a meaning in machine practice quite apart from the technique in ques- 

 tion. The term hultbing is used for processes in which the interior of a piece is removed by 

 forcing a member into it. It presumably' arose from the early practice of making axle 

 holes by forcing a hardened steel member through the wrought iron wheel hub. Further- 

 more, it has an analogous usage in coining. 



