5-6 MM ATTENUATION FOR THE CIRCULAR ELECTRIC WAVE 1125 



elusion of oxygen was not very feasible in the experimental TEoi lines, 

 the effects due to oxygen absorption were included in the measurements. 

 However, in order to simpHfy the presentation of the attenuation data 

 these absorption losses, as indicated in Figs. 6 and 7, have been sub- 

 tracted from the measured data. 



The measured attenuation of the four lines are shown in Fig. 8 as a 

 function of wavelength (5.1-5.8 mm). In each case the dash-dot-dash 

 lines represent the theoretical attenuation for copper. Each plot shows 

 two solid lines which indicate the range of values measured over the mm 

 band. The same range was observed either by varying the length of the 

 line by means of a sliding piston at the far end of the line or by imposing 

 a sweep voltage on the repeller of the signal klystron to produce a small 

 frequency modulation. These variations in attenuation correspond to 

 piston movements which are greater than a half wavelength and are due 

 to the mode interference effects produced by spurious modes generated 

 in the line. The resultant signal fluctuations which are due to mode con- 

 version and reconversion effects have been described in considerable 

 detail by Miller.^o 



Referring again to Fig. 8, the measured data shown by the solid lines, 

 which are for a plain line without mode filters, indicates that the oxygen- 

 free high conductivity copper line gave the lowest measured average 

 attenuation as well as the least variation. The low phosphorous deox- 



Table 1 



'" S. E. Miller, Waveguide as a Communication Medium, B. S.T.J. , 33, pp. 

 1229-1247, Nov. 1954. 



