NATURE OF POWER SATURATION IN TRAVELING WAVE TUBES 867 



as operations on the helix pitch alone, or the use of an auxiliary output 

 circuit. 



The results being in normalized form, are believed to be generally 

 applicable to conventional traveling wave tube design. With determina- 

 tion of an equivalence yi beams, they should even be a useful guide in 

 the design of tubes using hollow beams or other configurations. 



The work described could not have been done without the able assist- 

 ance of G. J. Stiles and L. J. Heilos and the helpful council of many of 

 my colleagues at Bell Telephone Laboratories. 



Appendix 

 scale model tube design 



There were a larger number of factors to be accounted for in the de- 

 sign of this tube. Its proportions should be such as to make it repre- 

 sentative of the usual design of traveling wave tube. Its size should be 

 such as to make it easy to define the electron beam boundary, and to 

 dissect the beam. The size should also be such that the electron beam 

 velocity analysis could be done before the beam character would be 

 changed either by space charge, or its velocity spread. The voltage should 

 he low so that further acceleration in the velocity analyzer would not 

 lead to an inconveniently high voltage. Finally, the availability of suit- 

 able measuring gear over a 3-1 frequency range, and the size of the 

 laboratory must be considered. All of these factors led to low frequency 

 operation, limited principally by the laboratory size and the mechanics 

 of construction. 



A moderate perveance of around 0.2 X 10~ was taken, with a 7a of 

 1.2 and 7ro of 0.8 in a representative helix with small impedance reduc- 

 tion due to dielectric and space harmonic loading. This is representative 

 of practical tube design in the microwave range and is centered on the 

 parameter values of most general interest. At a frequency of 100 mc and 

 a beam potential of 400 volts this resulted in a helix 10 feet long and l}^ 

 inches in diameter, with an electron beam 1 inch in diameter. The 

 choice of frequency was finally determined by the availability of meas- 

 uring equipment, and the voltage was selected to give a convenient size 

 for dissection of the electron beam. 



By changing frequency, beam current, and beam diameter it was 

 ! possible to cover a reasonable range of yro , and QC, and to make some 

 observations into the region of large C operation. 



In all of the measurements described, a very strong uniform magnetic 

 field was used to confine the beam, and therefore scaling of the magnetic 



