VIII. MICROWAVE TUBE RESEARCH A. H. A. Haus C. R. Russell A. G. Barrett C. Fried L. D. Smullin Prof. L. J. Chu NOISE IN ELECTRON BEAMS The shielding of the apparatus described in the previous Quarterly Progress Report has been completed, and very satisfactory semiautomatic operation has been obtained. The focusing magnet has been rewound and calibrated. With the improved focusing con- ditions the interception current was reduced to below 0. 1 percent for all but very weak Typical interception current curves are shown in Fig. VIII-1. magnetic fields. The Brillouin field corresponding to the curves of Fig. VIII-1 is approximately 370 gauss. A number of noise measurements have been made both for dc and for pulsed Figures VIII-2 to VIII-6 show some of these measurements. operation. that the beam diameter d is 0.20 inch as calculated, VIII-2 to VIII-5 are 262, ponding to Figs. 307, 362, the Brillouin fields corres- and 370 gauss respectively.) curves including the fine structure were well reproducible. a L U.b z w cr 0.4 (If we assume Instability, however, The was r 0.2 z O 0 - w ANODE -02 o -0.4 z -0.6 MAGNET SHIELD a.5 SI 0.68 Fig. VIII-1 Interception current vs cavity position for convergent-flow = 16. 6 4a; pulsed electron beam V = 1500 volts; I pressure 0 7 X 10 7 mm. -35- INTERCEPTION CURRENT< 1/10% , H 344 GAUSS INTERCEPTION CURRENT< 1.5% , H 258 GAUSS SHOT- NOISE LEVEL ' 0 -5 -10 iv -15 -, ,, 9n I\ tf ~_ 4 1l I / 1 ," \/" 10 15 20 DISTANCE (CM) 25 30 32.5 DISTANCE (CM) MAGNET I I L C.) 0.68 0.62 a" INTERCEPTION CURRENT < 1/10%, H- 430 GAUSS o INTERCEPTION CURRENT < / 1/O %, H - 598GAUSS SHOT- NOISE LEVEL 0 -5 20 0 DISTANCE (CM) 5 10 15 20 25 30 32.5 DISTANCE (CM) Fig. VIII-2 Noise power vs cavity position for convergent-flow dc electron beam. Icoll = 1. 7 ma; -7 col V = 700 volts; pressure 4. 5 x 10 mm; beam diameter = 0. 035 inch. IJTERCEPTION CURRENT < IV, *-H 258 GAUSS 5 SHOT- NOISELEVEL 0 5 -10 -15 -20 10 5 15 20 25 30 32.5 DISTANCE (CM) DISTANCE(CM) M MAGNET jODSHEO GAVITY I INTERCEPTION CURRENT< 1/20% . H- 430 GAUSS 0 SHOT-NOISE LEVaEL -5 -10 -15 -20 0 5 10 15 20 25 30 32.5 DISTANCE (CM) DISTANCE (CM) Fig. VIII-3 Noise power vs cavity position for convergent-flow dc electron beam. Icoll = 2. 8 ma; -7 ch mm; beam diameter w 0. 035 inch. 5 x 10 V 0 = 1000 volts; pressure INTERCEPTION CURRENT< 2% , H * 258 GAUSS INTERCEPTION CURRENT< I%, H 344 GAUSS 5 SHOT-NOISE LEVEL S0 ii uJ -5 z.01 / / I P-in V / / -10 -5 v W Ln -15 2- 20 -20 5 10 15 20 DISTANCE (CM) INTERCEPTION CURRENT 25 30 32.5 O < 1/20 % , H430 GAUSS IS 20 DISTANCE (CM) 25 30 325 INTERCEPTION CURRENT < 1/20 % ,H 602 GAUSS SHOT -NOISE LEVEL SSHOT- NOISE LEVEL *0 2 -5 R 10 .0 5 0 5 -5 -10 -20 O 5 10 20 15 DISTANCE (CM) 25 30 32.5 O 5 10 15 20 DISTANCE (CM) 25 Fig. VIII-4 Noise power vs cavity position for convergent-flow dc electron beam. -7 V o = 1500 volts; pressure = 8 x 10 Icoll = 4. 8 ma; mm; beam diameter~ 0.035 inch. 30 32.5 H 301 GAUSS 25 III H 344 GAUSS I 25 r I / 7!:.. I I / / /T / LE/L SHOT-NOISE SHOT- NOISE LEVEL SHOT- NOISE LEVEL W -5 cz 10 z -15 -0 -15 -20 -20 S 5 AwME S 7 ?L 10 15 20 DISTANCE(CM) 25 30 32.5 0 5 10 20 15 DISTANCE (CM) 25 30 32.5 C7VITY II 8 40.62 H- 430 GAUSS H -602 GAUSS ////// / 10 5 0 SHOT- NOISE I LEVEL SHOT-NOISE LEVEL -0 : (n -5 -15 -10 -5 -I5 -15 -20 -20 0 5 10 15 20 25 30 32.5 0 DISTANCE(CM) 5 K) 15 20 DISTANCE(CM) 25 30 32.5 Fig. VIII-5 Noise power vs cavity position for convergent-flow pulsed electron beam. -7 V = 1500 volts; pressure~ 7 X 10 mm; beam diameter Icoll = 16.6 ipa; 0. 035 inch. -~- H - 344 GAUSS H . 258 GAUSS /--_/ 2C 5 I1 o~ IC 5- 10 5 SHOT-NOISE LEVEL SHOT-NOISE LEVEL U cn WC U, - r, 10 15 20 DISTANCE (CM) 5 6ANODE 30 325 5 O 10 15 20 25 30 32.5 DISTANCE (CM) Y C 62 25 5 0.62 CL68 H 581 GAUSS H 430GAUSS 20 15 10 5 SHOT- NOISE LEVEL SHOT-NOISE LEVEL -5 0 5 10 15 20 25 30 32.5 0 DISTANCE(CM) 5 10 15 20 25 DISTANCE (CM) Fig. VIII-6 Noise power vs cavity position for temperature-limited pulsed electron beam. IColl = 1. 1 La; V ° = 1500 volts; pressure - pressure7 volts; =1500 7 x 10 -coil mm. 30 32.5 (VIII. MICROWAVE TUBE RESEARCH) present in the growing waves from approximately 10 db above shot noise. The noise growing wave was considerably more prominent in the unneutralized pulsed beams than in the completely neutralized dc operated ones. Therefore, the growing noise was caused by the ions in the beam. it seems improbable that Assuming that the noise curves consisted of a periodic standing wave and a superimposed growing wave, observe a delay of the growing part as the magnetic field was increased. For high mag- netic fields the growing part was absent in the drift space under observation. well be that in a longer drift space the growing part would still occur. we could It might The level of the successive minima also changed with the magnetic fields, as shown in Figs. VIII-2, VIII-3, and VIII-4. Growing noise waves were present even in temperature-limited beams as Fig. VIII-6 shows. A rise up to approximately 15 db above shot noise level could be observed. Again a delay of the growing portion was noticeable with the increase of the magnetic field. Increasing magnetic field also caused a decrease of the standing-wave ratio in the periodic noise waves. This decrease was due to the raising of the noise-wave minima and the simultaneous lowering of the maxima as well. Further experiments will be made with confined-flow electron beams. C. Fried B. INTERNALLY COATED CATHODES Two electron guns with internally coated cathodes have been designed for low-voltage as well as high-voltage operation. Both guns are under construction at present. C. C. Fried PROPAGATION OF SIGNALS ON ELECTRON BEAMS Data have been taken, both on a direct current electron beam and a pulsed electron beam, following the outline given in the last Quarterly Progress Report (1). A 3000-Mc/sec 1-mw signal, well above the noise level but still small enough for small signal theory to apply, was fed into one (fixed) cavity. picked up by the other (movable) cavity. The alternating beam current was In continuous-wave operation, standing-wave ratios of current as high as 41 db were observed at magnetic fields higher than 1. 5 times the Brillouin field. A typical curve is shown in Fig. VIII-7. The value of the Brillouin field for this case was 300 gauss. A plot of the current standing wave under pulsed operation is figure. The pulse repetition rate was 4000 cps, the length of the pulse 0. 75 4.sec. can be seen that the plasma wavelength is considerably longer. -41- given in the same It This can be attributed (VIII. TUBE RESEARCH) MICROWAVE 32 30 .0 - 0 28 26 -\ BEAM VOLTAGE1.5KV CURRENT 3.2MA (CONTINUOUSWAVE) MAGNETIC FIELD 290 GAUSS SBEAM 22 P ULSED 0- / - 27 920 z 18 I -O S16 o I I! 0 12 I CAVITY GAP0.070 INCHES - (D CAVITY HOLE DIAMETER 0.120 INCHES MEASURED BEAM DIAMETER0.020 INCHES I I - 42 -I 4 2 -2 1 2 4 6 4 6 8 10 12 14 16 IO 16 12 14 IB 20 22 24 26 28 30 32 34 36 20 32 22 24 26 28 30 34 36 POSITION OF MOVABLECAVITY (CM) Fig. VIII-7 Standing-wave pattern of current. to the fact that ions are essentially absent under pulsed operation, whereas in con- tinuous-wave operation the focusing action of the ions confines the beam to a smaller diameter. It may be that the smaller standing-wave ratio observed under pulsed opera- tion can be blamed on the shape of the pulse, which was not ideally square. At high beam current and low magnetic field, growing noise of the nature observed elsewhere (for example, see ref. 2) was known to exist. No anomalies were observed in the standing-wave pattern of the signal current. H. A. Haus References 1. Quarterly Progress Report, Research Laboratory of Electronics, M.I.T. 1953, p. 26 April 15, 2. Quarterly Progress Report, 1953, p. 43 April 15, Research Laboratory of Electronics, -42- M.I. T.