300 BELL SYSTEM TECHNICAL JOURNAL 



No particular precautions were taken to protect the active coating. Even 

 so, lives both in the laboratory and in the field were shown to be well over 

 2000 hours. Much later, the mechanical structure of the cathode and its 

 support was improved and a heavy glass protective housing designed to 

 cover the input leads. The performance chart of Fig. 17 in PART I is that 

 of one of the 718AY-EY series. Other data concerning these magnetrons 

 are given in Table II. 



Quite early in the study of the multicavity magnetron it had been realized 

 that the mode frequency distribution of the resonator system is very im- 

 portant to efficient operation. Evidence had accumulated which indicated 

 that something needed to be done to suppress all but one mode of oscillation. 

 Attempts at influencing mode frequency separation were made in which 

 flu.xbarriers or partitions, either complete or partial, were used in the end 

 spaces of the resonator system. These included what was essentially the 

 distortion of the structure into the form of the so-called "serpentine" men- 

 tioned in PART I. Although striking changes in operation were produced, 

 supporting the basic contention that there are more ways of doing a thing 

 wrong than right, worthwhile improvements were not achieved. 



Following the confirmation, here and elsewhere, of the British results 

 with strapping and the coding of the series of 718AY-EY magnetrons, 

 intensive work was started in the general investigation of the modes of the 

 magnetron resonator system and their relation to the .various possible 

 methods for strapping. This work involved the determination of the mode 

 patterns for mode identification, the technique for which was consequently 

 developed. \'arious strapping schemes were devised, built into magnetron 

 resonator systems, and studied. Among these were the echelon strapping 

 also used by the British, single and double ring strapped systems, and 

 diametral straps which were adaptable to external tuning of the magnetron. 

 The importance of strap shielding and strap assymmetries discussed earlier 

 were brought out in these studies. The effects of end space volume and 

 anode length on mode frequency distribution for given strapping systems 

 were also studied. 



Strapping made possible greater tuning ranges with tunable magnetrons. 

 Efficient magnetrons having broken echelon strapping on one end and 

 diametral straps on the other, connected through a glass seal to an external 

 coaxial tuner were built. Tuning ranges of over ±5 per cent were achieved 

 in this way. There was no demand at the time for tunable magnetrons. 



Following the introduction of the 718AY-EY magnetrons, their properties 

 were reproduced at an extended range near the wavelengths of the 706A-C 

 and at the wavelength of the 714A. The designs for these were arrived 

 at by scaling the 718AY-EY to the appropriate wavelengths. The principles 



