276 BELL SYSTEM TECHNICAL JOURNAL 



Before the invention of straps the 700A-D magnetrons were scaled to 10 

 cm, to explore the possibilities of a more efficient magnetron design at this 

 wavelength. Straps were introduced before the completion of the experi- 

 ment. The resultant strapped magnetron having six resonators was very 

 efficient— 60 per cent — but required a high magnetic field as can be seen by- 

 referring to equation (16) of PART I. Like other magnetrons, the 700A-D 

 became much more efficient when strapped. At the normal test point the 

 efficiency ranged around 50 per cent, while at higher magnetic field and volt- 

 age, 75 per cent over-all efficiency was achieved. The introduction of straps 

 into the manufactured design was not undertaken. 



One further experiment of interest arose during the development of the 

 700A-D magnetrons from the desire to measure the gas pressure in a sealed- 

 off magnetron. The non-oscillating magnetron itself was used as an ioniza- 

 tion manometer. With the magnetic field set at a high value above cutoff and 

 under conditions of no RF oscillation, electrons which arrive at the anode can 

 do so only after having lost energy by collision with a gas molecule. Under 

 these conditions the anode current is directly proportional to the pressure. 



Although by present standards the 700A-D magnetrons might appear 

 somewhat crude and inadequate, they nevertheless have an important place 

 in the story of wartime magnetron development. They filled an immediate 

 need in the radar system for which they were designed, providing the U. S. 

 Navy with a radar set which saw service in a number of crucial engagements. 

 Furthermore, the development of the 700x\-D magnetrons provided in- 

 valuable experience. 



13.2 The 728 A- J Magnetrons: The 728A-J magnetrons were developed 

 for fire control and search radar systems to supersede those which had 

 used the 700A-D magnetrons In these n v. systems a magnetron gener- 

 ator was to be required which could deliver 200 kw. peak output power 

 in the frequency range 920-970 mc/s (later extended to 900 mc/s). 



In an early design, the resonator system had eight resonators and was 

 strapped with wire straps in the early British configuration [see Fig. 24(a) 

 of PART I]. The anode length was 4 cm., the same as was used in the 700 

 A-D, which on a wavelength basis was about 2/3 that used in the British 

 10 cm. magnetron. The first models were designed for operation at pulse 

 voltages of about 27 kv. When, subsequently, it was decided to reduce this 

 voltage, a redesign involving a reduction of size of the interaction space 

 became necessar}'. Since more had been learned about the technique of 

 strapping in the meantime, it was decided that straps of the double ring 

 type [see Fig. 24(d) of PART I] should be used in the new design. At first, 

 the straps were set on the ends of the anode structure projecting into the end 

 spaces but were later recessed into channels cut into the copper resonator 



