MAGNETRON AS GENERATOR OF CENTIMETER WAVES 285 



magnetron. To circumvent this ditTiculty, the input leads and their glass 

 housing were redesigned. The glass of the leads was lengthened and folded 

 to provide a greater surface distance over which the arc must strike. 

 Although the redesign was limited somewhat by the space available in the 

 transmitter unit, it was possible to make leads which could stand 60 cycle 

 peak voltages of almost twice the normal pulse voltage applied to the mag- 

 netron. This step all but eliminated flash-over in operation. To take 

 care of the occasional flash-over, a spark gap across which the breakdown 

 could occur was added in the equipment. 



The 4J21-30 magnetrons are illustrative of carefully designed non- 

 tunable magnetrons of wavelengths 20 to 30 cm. developed at the Bell 

 Laboratories. They were the first magnetrons designed here completely 

 on the basis of CW impedance measurements. 



14. Tunable Magnetrons for Wavelengths of 20 to 45 Centimeters 



14.1 General: As fixed frequency magnetrons became available in the 

 various frequency bands designated for radar use, the interest quite naturally 

 turned to the development of tunable magnetrons. The operational rea- 

 sons for this were to enable one to set the frequency of a radar system at will, 

 thus avoiding interference between sets in a large group of aircraft on naval 

 vessels, to enable one to vary the frequency at will from time to time or as 

 the need arises to avoid jamming, and to permit the stocking of fewer mag- 

 netrons to cover a given frequency band. 



Early work on tunable magnetrons at the Bell Laboratories was done 

 with 10 cm. models. Although no such developments were carried to the 

 stage of production, the ideas and techniques evolved were used at other 

 frequencies. Somewhat later the Bell Laboratories committed itself to a 

 program of development of tunable magnetrons for pulsed radar use in the 

 20 to 45 cm. wavelength range. 



The program initially was not directed toward the goal of some particular 

 magnetron of fixed specifications. Rather, it was the intention to explore 

 the field of possible tuning methods and to find that one which appeared both 

 electrically and mechanically to be best suited for large magnetrons. Work 

 in this initial stage was done on anodes of the 4J21-30 series and may 

 be divided into two main channels characterized by the degree of symmetry 

 involved in the tuning scheme. In one, the tuning of the entire anode block 

 is accomplished by modification of a single one of the resonators and may 

 thus be characterized as unsymmetrical tuning. In the other, each resonator 

 of the multiresonator block is tuned in a symmetrical fashion. 



Among the unsymmetrical types studied were those employing an auxiliary 

 loop in one resonator connected to a reactive element which could be either a 



