326 BELL SYSTEM TECHNICAL JOURNAL 



sixteen resonator anode structure used in the early work but including 

 the features of "packaging," axial cathode mount, and wave guide output. 

 A number of vexing but nevertheless interesting problems was encountered. 

 Principal among these were the tendency to operate in another mode under 

 certain conditions and the problem of obtaining satisfactor}^- magnetic field 

 uniformity in a design which necessitated the removal of a considerable 

 portion of the center of the magnet pole piece to accommodate the axial 

 cathode mount. Each of these difficulties was surmounted as is described 

 in the detailed discussion later. The resonator system was redesigned 

 quite late in the development to take advantage of the possibihty of oper- 

 ating at a lower electronic conductance and hence higher RF voltage for the 

 same output power. The advantages of this step were a reduction in the 

 tendency of the magnetron to "mode" and an increase in unloaded Q. 



The development program briefly sketched above resulted in three 

 coded magnetrons, the 4J50, 4J52, and 4J78. The 4J50 and 4J78 magne- 

 trons, identical except for frequency, were built with magnets large enough 

 to permit operation near the maximum power capabilities of the design and 

 with larger input leads to withstand the higher DC voltage required. The 

 4J52 magnetron, although incorporating the same internal design as the 

 other models, was built with smaller magnets for operation at a set of 

 conditions easily attainable with the higher power design but beyond the 

 reach of the 725A. Developmental work on all three magnetrons was 

 conducted simultaneously. 



In the 4J50 magnetron a hole and slot resonator system was adopted. 

 The remarkable freedom of the 725A magnetron from "moding" difficulties 

 had made it appear desirable to retain as large a mode frequency separation 

 as possible. Equivalent circuit theory indicated that to do this the increase 

 in the number of resonators from twelve to sixteen would necessitate about 

 twice the strap capacitance of the 725A. This would have required ex- 

 tremely wide straps for which the recess channel in the anode structure 

 would be very difficult to trepan. For this reason the resonator system 

 was strapped less heavily; the ii = 8 and )i = 7 mode frequencies differed 

 by 19 per cent rather than 25 per cent as in the 725A. 



In none of the early models built into 725A type structures was any 

 "moding" or "misfiring" observed. When axial cathode and wave guide 

 output were introduced, however, "moding" was experienced both in those 

 with no strap breaks and those with the usual double break at one end of the 

 anode. By enlarging the cathode diameter such that the ratio of cathode 

 to anode radii increased from 0.60 to 0.66, the difficulty was removed in 

 models with broken straps, but it was not possible to ehminate it in those 

 with unbroken straps. Decreasing the rc/ra ratio increased the tendency 

 to "mode." 



