TRAVELING WAVE TUBE FOR 6,000-MC RADIO RELAY 1333 



becomes very serious and eventually limits the TWT output to about 30 

 watts. If it were necessary to reduce this fading, the envelope shrinking 

 technicjue illustrated in Fig. 16 could be used. The maximum power 

 output after fading is shown as a function of frequency for several beam 

 currents in Fig. 36 and as a function of magnetic flux density in Fig. 37. 

 The theory of the high level behavior of a TWT** predicts that the ratio 

 of electronic efficiency (i.e., E = power output/beam power) to the gain 

 parameter C should be a function of C, QC and 7b (where h is the beam 

 diameter). However, with the range of parameters encountered in the 

 M1789, the variation in E/C should be small. Fig. 38(a) shows E/C as a 

 function of frequency when the TWT is operating at the voltage for 

 maximum gain at low signal levels. Fig. 38(b) shows the maximum value 

 of E/C obtainable at elevated helix voltage. In both figures we show the 

 efficiency as estimated using the results of Tien^ corrected for the effect 

 of intrinsic loss following the procedure of Cutler and Brangaccio.^ 

 All etticiencies in these two figures are the electronic efficiency before 

 fading. It would be quite difficult to compare the efficiency after fading 

 with theory because the intrinsic attenuation in this case varies along 

 the helix in an unknown manner so that we cannot properly take it into 

 account. From the figures we see that the calculated value of E/C at 

 6,000 mc and 40 ma is not far from the experimental value but the ex- 

 perimental points show more variation with frequency than is predicted 

 by theory. The low efficiency at 20 ma results from the fact that there 

 is insufficient gain between the helix attenuator and the output. As a 

 result, the TWT "overloads in the attenuation." 



4.4 Noise Perjormance 



A new and important noise phenomenon was observed in the course of 

 the Ml 789 development. It was found that the noise figure is strongly 

 dependent on the magnetic flux linking the cathode and on the rf output 

 level of the TWT. For example, with the TWT operating near maximum 

 output and with a cathode completely shielded from the magnetic field, 

 noise figures of about 50 db were observed. By allowing 20 gauss at the 

 cathode, the noise figure was reduced to 30 db. Fig. 39 shows the noise 

 figure as a function of magnetic flux density at the cathode for several 

 values of rf power output. We see that there is a peak of noise figure 

 roughly symmetrical about zero flux at the cathode, and that the magni- 

 tude of this peak is considerably increased by operating the TWT at 

 high output levels. 



Some additional observed properties of the noise peak are: 

 (1) The magnitude depends on the synchronous voltage of the helix. 

 For a 1,600-volt helix it is about 10 db higher than shown in Fig. 39 and 



