1116 THE BELL SYSTEM TECHNICAL JOURNAL, SEPTEMBER 1956 



guide size and frequency. At 5.4 mm, a J^e" I-I^- waveguide has an 

 appreciably lower attenuation with the circular electric mode than with 

 the dominant mode. A %" I.D. guide has a circular electric attenuation 

 approximately one-fourth that of the dominant mode in the same pipe. 

 It is the purpose of this paper to present some experimental results 

 which have been observed with circular electric wave transmission in 

 the 5-6 mm wavelength region. The attenuation for three different hnes 

 and the transmission variations due to moding effects are reported. Al- 

 lowance for the loss due to oxygen absorption has been included. 



DESCRIPTION OF THE TEST LINES 



The TEoi mode attenuation measurements were made on approxi- 

 mately straight runs of line ranging from about 100 to 200 feet in length. 

 The copper pipe comprising these lines is believed to conform to the 

 best tolerances and internal smoothness which are current manufacturing 

 practice for waveguide tubing. The relative tolerances and their effect 

 upon transmission are considered in a later section. Three kinds of copper 

 line were measured: a waveguide of oxygen-free copper, one line of low 

 phosphorous-deoxidized copper and one line of steel with a 20-mil low 

 phosphorous-deoxidized copper inner lining. The oxygen-free high-con- 

 ductivity-copper with its higher conductivity and somewhat greater 

 ductility was chosen to provide comparative performance data with the 

 low phosphorous-deoxidized copper which is commonly used in wave- 

 guide manufacture. A waveguide whose outer wall is constructed of 

 steel to provide the necessary strength and wall thickness to support a 

 very thin copper inner wall has the advantage that such waveguide re- 

 quires less copper. This composite wall tubing was obtained to ascertain 

 whether the tolerances and the nature of the inner surface would yield 

 transmission data comparable to solid copper waveguide. 



The lines Avere supported on brackets which were accurately aligned 

 and spaced at 6-ft intervals. Although the brackets provided for an 

 accurately straight line, the manufactured pipe was not perfectly straight 

 but, in some samples, varied as much as %" in a 12-ft length. Installing 

 the pipe on the brackets tended to straighten the line and reduce these 

 variations to about half this amount. A general view of the lines is shown 

 in the photograph of Fig. 1. 



The sections of waveguide were joined together with a more or less 

 conventional threaded coupling, but with one very important difference. 

 The threads, which are cut at the ends of each section, are cut relative 

 to center of the inside diameter and not the outside diameter. This is 

 achieved by employing a precision pilot to provide a center for the cut- 



i 



