PART I — THE GROWING NOISE PHENOMENON 835 



III EXPLORATORY MEASUREMENTS 



The electron gun used in these experiments had been designed for use 

 in a helix traveling-wave tube with a longitudinal focusing field of 600 

 gauss. Noise-power and collector-current curves, therefore, were first 

 taken with 600 gauss to study a typical state of affairs in an operational 

 beam. As seen in Fig. 2, the noise power at 3.9 kmc varies periodically 

 with distance for about 4 inches from the gun, then climbs rather 

 smoothly by nearly 23 db to an irregular plateau, where it undulates 

 slowly, and finally levels off. The initial part of the growing noise curve 

 at 10.7 kmc is missing because of inadequate receiver sensitivity, but its 

 later portion is similar to that at 3.9 kmc, with about half the rate of 

 noise climb. With the 0.020-inch aperture, the collector-current varia- 

 tions decrease in amplitude chiefly in the drift region preceding the noise 

 climb; whereas those for the 0.100-inch aperture decrease afterwards. 

 Both curves show a flattening in the growing-noise region itself, as well 

 as a decrease in their average values after that region, signifying an in- 

 crease in the average beam diameter. 



A similar set of curves is shown in Fig. 3, for a focusing field of 279 

 gauss (about twice the nominal Brillouin field). Noise growth at 3.9 kmc 

 starts later, and proceeds less steeply, than at 600 gauss. The noise-power 

 curve for 10.7 kmc is much more articulated, with a semblance of peri- 

 odicity, throughout the drift region. Collector-current curves for both 

 0.020- and 0.050-inch apertures show considerable reduction in current- 

 ripple amplitude with distance, reaching virtually zero in the former case. 



Another type of survey measurement is illustrated in Fig. 4. With 

 the probe stationary at the far end of the drift space (about 18 inches 

 from the gun anode), the main solenoid current is varied smoothly to 

 change the focusing field from to over 600 gauss, and synchronized 

 records are made of collector current and noise power. (In this instance, 

 the current in the auxiliary solenoid was +3.2 amperes.) At low mag- 

 netic fields, both the current and noise-power curves have large ampli- 

 tude variations, which diminish as the field increase. At first glance, the 

 noise peaks and valleys seem to coincide with those of collector current; 

 certainly, some do. Closer inspection, however, reveals significant mis- 

 alignments which cannot be accounted for by experimental error. When 

 the three noise curves, at 3,050, 3,930, and 4,730 mc, respectively, are 

 compared with each other, some characteristic features emerge: 



(1) An average curve drawn through each pattern has one or two 

 broad maxima, which tend to move toward higher field strengths with 

 increasing frequency. 



