CARRIER TELEGRAPH METHODS 491 



One-Source Two-Band Terminal Apparatus 



The one-source two-band sending circuit was similar to that of the fre- 

 quency-shift arrangement with relay modulator, except that the frequency 

 shift was 170 cycles, and the marking and spacing frequencies were adjusted 

 to be at the centers of the pass bands of the marking and spacing sending 

 filters, as shown in Fig. 2. The receiving circuit was the same as that used 

 for the two-source arrangement. A limiter was always used. 



Telegraph Transmission Measuring Apparatus 



In all of the tests, the d-c open-and-close signals in the local sending loop 

 were substantially rectangular and consisted either of reversals (a succession 

 of alternate marks and spaces of equal duration) or of the test sentence: the 



QUICK BROWN FOX JUMPED OVER A LAZY DOG's BACK 1234567890 BTL SEND- 

 ING. The customary 7.42 unit teletypewriter code was used, consisting of a 

 stop pulse, a start pulse, and five code pulses per character^ The distribu- 

 tors supplying these signals were driven by synchronous motors controlled 

 by an adjustable frequency oscillator. The speeds utilized experimentally 

 ranged from 60 to 180 words per minute (about 23 to 68 dots per second). 



In order to measure the telegraph distortion of the signals obtained in the 

 receiving loop, a cathode-ray tube distortion measuring set was used which 

 measured maximum total distortion in per cent of a unit pulse in much the 

 same manner as a start-stop distortion measuring set previously described^ 

 except that electronic circuits were used to replace the distributor and all 

 but one relay, which made possible precise measurements over a wide range 

 of speeds. The bias of received signals was measured on reversals by means 

 of a highly damped zero-center d-c milliammeter inserted in the receiving 

 loop. 



Source of Resistance Noise 



In order to measure the effect of line noise, resistance noise was reproduced 

 from a phonograph record, amplified, and combined with the carrier signals 

 by means of a symmetrical three-way pad (part of the artificial line). A 

 variable attenuator was used to regulate the amount of noise entering the 

 line. The r.m.s. noise power or marking carrier power was measured with a 

 thermocouple. 



Measuring Precision 



The signals generated by the dot and test sentence distributors were dis- 

 torted less than 3 per cent of a dot length. As these distributors were of a 



' E. F. Watson : "Fundamentals of Teletypewriters Used in the Bell System", Bell Sys. 

 Tech. Jour., Vol. XVII, Oct. 1938, pp. 620-639. 



s R. B. Shanck, F. A. Cowan, S. I. Cory: "Recent Developments in the Measurement 

 of Telegraph Transmission", Bell Sys. Tech. Jour., Vol. XVIII, Jan. 1939, p. 149. 



