BELL SYSTEM TECHNICAL JOURNAL 



CHAPTER I 



INTRODUCTION 



ASTRONOMERS are interested in stars and galaxies, physicists in 

 ' atoms and crystals, and biologists in cells and tissues because these 

 are natural objects which are always with us and which we must under- 

 stand. The traveling- wave tube is a constructed complication, and it can 

 be of interest only when and as long as it successfully competes with older 

 and newer microwave devices. In this relative sense, it is successful and 

 hence important. 



This does not mean that the traveling-wave tube is better than other 

 microwave tubes in all respects. As yet it is somewhat inefficient compared 

 with most magnetrons and even with some klystrons, although efficiencies 

 of over 10 per cent have been attained. It seems reasonable that the effi- 

 ciency of traveling-wave tubes will improve with time, and a related device, 

 the magnetron amplifier, promises high efficiencies. Still, efficiency is not the 

 chief merit of the traveling-wave tube. 



Nor is gain, although the traveling-wave tubes havebeenbuilt with gains 

 of over 30 db, gains which are rivaled only by the newer double-stream 

 amplifier and perhaps by multi- resonator klystrons. 



In noise figure the traveling-wave tube appears to be superior to other 

 microwave devices, and noise figures of around 12 db have been reported. 

 This is certainly a very important point in its favor. 



Structurally, the traveling-wave tube is simple, and this too is impor- 

 tant. Simplicity of structure has made it possible to build successful ampli- 

 fiers for frequencies as high as 48,000 megacycles (6.25 mm). When we con- 

 sider that successful traveling-wave tubes have been built for 200 mc, we 

 realize that the traveling-wave amplifier covers an enormous range of fre- 

 quencies. 



The really vital feature of the traveling-wave tube, however, the new 

 feature which makes it different from and superior to earlier devices, is its 

 tremendous bandwidth. I 



It is comparatively easy to build tubes with a 20 per cent bandwidth at | 

 4,000 mc, that is, with a bandwidth of 800 mc, and L. M. Eield has reported ' 

 a bandwidth of 3 to 1 extending from 350 mc to 1,050 mc. There seems no 

 reason why even broader bandwidths should not be attained. 



As it happens, there is a current need for more bandwidth in the general 

 fiekl of communication. For one thing, the rate of transmission of intclli- 



