TRAVELING WAVE TUBE FOR 6,000-MC RADIO RELAY 1305 



of about ±25 volts. It is not difficult to hold the average pitch variations 

 to less than d=l per cent. The loading, however, is a more difficult prob- 

 lem for not only must the dielectric properties of the support rods and of 

 the glaze material be closely controlled, but attention must also be paid 

 to the size and density of the glaze fillets. The gain of the tube is affected 

 by the amount of loss in the helix attenuator. For the particular loss 

 distribution used in the MI789 a variation of ±5 db out of a total 

 attenuation of 70 db results in a gain variation of about ±1 db. The 

 helix attenuator depends to a large extent on a conducting "bridge" 

 between helix turns and therefore the amount of attenuation is sensitive 

 to the size and the surface condition of the glaze fillets. Thus, the glazing 

 process must be in good control in order to minimize variations in both 

 gain and operating voltage. With our present techniques, we are able to 

 hold the voltage for maximum gain to within ±50 ^'olts of the nominal 

 value. The gain is held to ±2 db — about half of the spread we believe 

 to be caused by variations in loss distribution and about half by differ- 

 ences in beam size. 



Fig. 15 — Enlarged photograph of part of an M1789 helix. Two of the ceramic 

 support rods can be seen. The other is directly opposite the camera behind the 

 lielix and is out of focus. The fillets of glaze which bind the helix to the rods can 

 he seen along the upper rod. This section of helix was free from applied loss. 



