CHARACTERISTICS OF TOLL TELEPHONE CABLES 295 



together to make a quad. The nominal capacitance of a pair is .062 mf. 

 per mile. There are various twist lengths of both pairs and quads in a given 

 cable as well as cables ranging in size from 12-quad cable to the oversize 

 19-gauge cable containing 225 quads. Some type K is operating over 

 "paired" cable, i.e., cable in which only the wires of each pair are twisted 

 together. 



Operation in two directions is accomplished by using either a separate 

 cable for each direction or a single cable with two groups of conductors 

 separated by a layer shield. This avoids serious near-end crosstalk effects 

 which would result from the large level difference existing between opposite 

 directions at a repeater point. 



Methods op Measurement 



As mentioned above, 250-foot reels of standard toll cable were placed 

 in a special room which could be accurately maintained at any desired 

 temperature from about zero to 120 degrees, Fahrenheit, and measurements 

 made for various frequencies and temperatures. For the most part, these 

 consisted of open-circuit admittance and short-circuit impedance measure- 

 ments on part of the pairs in the cable at temperatures of about zero, 30, 

 50, 90 and 120 degrees F., over the frequency range from 4 to 100 kc. From 

 these measurements computations could then be made to determine the 

 resistance, inductance, capacitance and conductance as well as the attenu- 

 ation, phase constant and characteristic impedance of this type of cable at 

 the different temperatures and frequencies. Detailed data on frequency 

 and temperature variations of these quantities are given below. Most of 

 these data were obtained from measurements made on 16- and 19-gauge 

 pairs in a typical reel-length of standard toll cable. The temperature is 

 difficult to maintain at a constant level and d-c. resistances of certain pairs 

 were measured at frequent intervals during the process in order to get ac- 

 curate temperature readings by comparing with resistance-temperature 

 curves of these pairs. Thermocouples were also attached to the cable 

 at various points along its length and sheath temperatures determined 

 from them. After stabilizing the room temperature as closely as possible, 

 the variations in cable temperature took place slowly enough to be allowed 

 for in the computations. 



After the selection of the Toledo-South Bend route for a trial installation, 

 further measurements were made on certain of the pairs in this cable. The 

 test sections, extending out of Lagrange, Indiana, were made about 10 

 miles long in order to obtain the averaging effect of length. For this dis- 

 tance it was not possible to use open and short-circuit measurements as 

 was done in the laboratory, and a substitution method^ was devised (Fig. 1). 



' This was devised by H. B. Noyes and will be described by him in a paper in the 



Bell Laboratories Record. 



