514 



BELL SYSTEM TECHNICAL JOURNAL 



120 w.p.m. The arrangements tested were found to have about the same 

 relative susceptibility to level changes at this speed as at 60 w.p.m. 



The distortion shown in Figs. 13 and 15 consisted largely of bias for the 

 arrangements having no limiter. The distortion shown in Figs. 13, 14, 15, 

 and 16 for arrangements which had limiters rose faster than the absolute 

 value of the bias as the level dropped below the cut-off of the limiter. This 

 was partly due to extraneous noise in the laboratory apparatus. 



H 24 



;< 16 



5 



-32 -24 -16 -8 8 16 24 32 40 43 



RECEIVING LEVEL IN DECIBELS (FROM NORMAL) 



Fig. 16 — Distortion vs. receiving level characteristics of wide-band frequency- 

 shift arrangements at 120 w.p.m. (46 d.p.s.). 



Selective Level Changes 



When frequency-shift, two-source, or one-source two-band arrangements 

 are operated over a radio link, selective fading may occur which afifects the 

 marking and spacing frequencies by different amounts. Tests were made 

 only on the two-source arrangements to measure the effect on distortion and 

 bias of differences between received marking and spacing levels. This effect 

 was simulated by setting the level of one of the two carrier oscillators at 

 different values with respect to the other, and measuring the distortion and 

 bias for each level setting. The distortion measurements are shown in Fig. 

 17 and are summarized below in Table L 



It may be seen from Fig. 17 that the distortion rose rather rapidly as the 

 marking level was changed with respect to the spacing level and thus the dis- 

 tortion due to selective fading was not greatly reduced by the use of a limiter. 

 The increase in distortion was mainly bias. 



Tests of Adjacent Channel Crossfire 



In order to investigate the effects on distortion due to interference from 

 adjacent channels, certain arrangements having the same loss characteristic 



