MAGNETRON AS GENERATOR OF CENTIMETER WAVES 257 



not yet entirely clear. It is certain, however, that the RF oscillation in the 

 magnetron is responsible for the fact that such currents may be attained. 

 This is made evident experimentally by measuring the current that may be 

 drawn when the magnetic field is reduced so that oscillation is impossible 

 and comparing it with the current drawn during oscillation. At tempera- 

 tures even in excess of the operating temperature such pulsed emission cur- 

 rents of excellent magnetrons may run as low as 1% of the currents flowing 

 during oscillation. Furthermore, the fields at the cathode during these 

 measurements may actually exceed those present during oscillation because 

 of the absence of the dense space charge clouds in the interaction space. 



In the process of phase selection of electrons, as previously discussed, 

 those electrons starting from the cathode in a phase such as to gain energy 

 from the RF field are removed from the interaction space and driven into 

 the cathode. They impart to the cathode the energy they have gained, 

 causing secondary electrons to be emitted and the cathode temperature to 

 increase. In most magnetrons the amount of this returned energy is about 

 5% of the input, though in some cases it may become as high as 10%. This 

 means that the average energy of back bombardment may run as high as 

 75 watts per square centimeter of cathode surface. This amount of energy 

 must be dissipated under equihbrium conditions by radiation and conduc- 

 tion. In general, the short wavelength magnetrons are run with no heater 

 power supplied to the cathode, the cathode temperature being maintained 

 by back bombardment. In many cases, cathode overheating by this 

 process is actually the limitation on the operating capabiUties of the mag- 

 netron. 



There is some experimental evidence that the large current drawn from 

 the magnetron cathode is not primarily made up of secondary electrons but 

 may result from an "enhanced" primary emission. This is supported by 

 the fact that the secondary electrons emitted at the return to the cathode 

 of the "out of phase" electrons are themselves, assuming negligible emission 

 time, largely "out of phase" electrons later to return to the cathode. How- 

 ever, the emission of large numbers of secondary electrons may lead to an 

 "enhanced" primary emission by a process not now understood. As pos- 

 sible processes may be mentioned field emission or an actual lowering of the 

 cathode work function, each brought about by the fields in the cathode 

 coating which result from the charge loss attendant upon the secondary 

 emission. Ionic conduction or electrolytic action in the coating may also 

 contribute in some manner to a lowered work function and to the "enhanced" 

 emission, although such ionic processes would generally involve time in- 

 tervals longer than a microsecond. The actual mechanism involved, how- 

 ever, is still speculative. 



A word about the relation of magnetron scaling to magnetron cathode 



