648 BELL SYSTEM TECHNICAL JOURNAL 



to specific values with a high degree of precision. In order to equal the dimen- 

 sional tolerances of the 1553 it was necessary that parallelism and spacing 

 be accurate to 0.1 mil. 



A schematic diagram of the apparatus is shown in Fig. 1. A flat, circular 

 disc having a 250-mil diameter aperture, across which the grid was stretched, 

 was mounted upon the face of the hollow micrometer screw 7^ 1. The latter 

 was mounted so that its face was accurately parallel with the end face of 

 the central conductor of the input coaxial line in the upper part of the figure. 

 By means of the micrometer ^ 1 the input spacing Si, which we shall con- 

 sider as representing the cathode-grid spacing, could be adjusted. The cen- 

 tral conductor of the coaxial line was insulated at d.c. from the outer con- 

 ductor; hence it was possible to use an ohmmeter to indicate when the grid 

 was just touching the coaxial face. The micrometer could then be backed 

 away from the grid by any desired amount. The input coaxial was fitted 

 with a standing wave detector in the form of a probe which could be moved 

 along the line and placed at any arbitrary distance h from the grid. 



On the output side of the circuit, in the lower part of the figure, there 

 was another coaxial line arranged so that its center conductor could be 

 driven by micrometer ^ 2. The latter was insulated from the outer con- 

 ductor of the coaxial by means of a condenser in order that an ohmmeter 

 could be used to determine the position of the micrometer which caused the 

 central conductor to just touch the grid. Spacing ^2 could then be adjusted. 

 The output coaxial line was terminated in its characteristic impedance of 

 62 ohms. At a distance of X/2 from the grid a probe was located for samplmg 

 the power in the output line. 



The diameter of the input coaxial conductor was 180 mils at the end. In 

 the figure it will be noted that at a short distance from the end the diameter 

 increased to a larger diameter (250 mils). Because of the required length 

 of the central conductor, it was necessary to increase its size in this way 

 for mechanical rigidity. The effect of this change in cross-section was 

 computed and allowed for in the final results. The output coaxial conductor 

 was relatively short, so that it was possible to assign a diameter of 180 mils 

 for its entire length. The 180-mil diameter was selected to correspond with 

 the diameters of the cathode and anode in the 1553 triode. 



The procedure for making the measurements was as follows: With a 

 particular set of spacings Si and S-i the standing wave ratio in the input 

 line was measured. This ratio, together with the measurement of the posi- 

 tion of a standing wave minimum, permitted the calculation of an input 

 admittance F to be made. Then with the standing wave detector probe 

 placed at a distance h = X/2 from the grid, the ratio of the voltage at the 

 input terminals of the triode to the voltage appearing at the output probe 

 was measured both as to magnitude and phase as described in a recent 



