MAGXETROX AS GEXERATOR OF CEXTI METER WAVES 309 



action spaces than direct scaling from 10 cm. would indicate. This had 

 been done to achieve greater cathode size and smaller operating magnetic 

 field. It appeared that 3 cm. magnetrons had been suitable only for opera- 

 tion at high voltage and high magnetic field and were being operated far 

 below their range of maximum efficiency or even of reasonable operation. 

 A decisive step forward was taken when a strapped magnetron was built 

 which was rigorously scaled, according to the scaling principles discussed in 

 PART I, from 10 cm. designs of the same operating voltage and current as 

 that desired at 3 cm. 



The first such magnetrons were scaled from the 10.7 cm. strapped mag- 

 netrons having eight resonators, the 718AY-EY, with only minor changes 

 being made in the resonators for mechanical reasons. The results obtained 

 were \exy encouraging although they were erratic for a variety of reasons. 

 Mechanical construction was difficult and not reproducible before newer 

 tools and techniques for making and assembling small parts were introduced. 

 Output circuit variations in many cases completely masked good electronic 

 operation, and it was not until a carefully considered and executed study 

 of the output circuit design was instituted that consistent results were ob- 

 tained. However, the results were such that it was decided to shift all 

 emphasis in 3 cm. magnetron development at the Bell Laboratories to 

 strapped designs scaled from 10 cm. magnetrons. 



This was by no means an easy decision to make. For example, it meant 

 the use of very small parts and clearances, of order 0.010 in., and ver>' close 

 tolerances. That such a magnetron would be feasible for large scale pro- 

 duction was by no means obvious. A cathode, 0.100 in. or less in diameter, 

 which must deliver a considerably higher current density than any pre\-ious 

 magnetron cathode, would be necessary. Whether such a cathode, at the 

 expected operating conditions, would have any appreciable life was not 

 known. Furthermore, a scaled 3 cm. magnetron would require a much 

 higher magnetic field than previous 3 cm. designs, a demand which might 

 increase the magnet weight. 



On the other hand, it was possible that the improved electronic efficiency 

 of the scaled magnetron would result in less rigorous treatment of the cathode 

 in spite of its small size. In spite of the increase in magnetic field, the 

 decrease in anode length of the scaled magnetron made possible the reduction 

 of the magnet gap to a point where it was conceivable that no greater mag- 

 netomotive force and size of magnet would be required than in previous 

 3 cm. designs. Greater stability and freedom from mode troubles could be 

 expected. 



Early strapped magnetrons having eight resonators had anode diameters 

 of order 0.175 in. and cathode diameters of order 0.065 in. Magnetrons of 



