MEASUREMENT OF TELEGRAPH TRANSMISSION 



165 



operate the message register which in turn will extinguish the tube by 

 momentarily short-circuiting the anode potential. Other transitions 

 in the signal will not cause the condenser C to be discharged through 

 the coil because the brush of face (c) will not be traversing the proper 

 segment and there will be no path to ground. 



The complete circuit of the device contains a selecting switch which 

 automatically changes the signal combination for each revolution of 

 the brush arm of face (a) and reversing switches to invert the signals 

 and the relay connections. These have been omitted for the sake of 

 briefness and clarity. It is thought, however, that the above descrip- 

 tion will give a good idea of how this device operates to make a record 

 of the frequency of occurrence of distortions. A sample of such a 

 record is plotted in Fig. 15, this being from data obtained over a 



600 



-10 -8 



-6 



-4-2 0+2+4 



•7o DISTORTION 



Fig. 15 — Typical curve obtained with distortion-distribution recorder. 



short-wave radio link. The measurements covered a 15-minute period 

 and one transition of a repeated signal was observed continuously 

 during this period. By inspection it is seen that this circuit has bias 

 of — 1 per cent and the r.m.s. value of distortion is about ± 3 per cent. 

 With miscellaneous signals and with characteristic distortion present, 

 the curve would have several peaks and would be somewhat irregular, 

 but would have the general shape of a normal-law distribution curve. 



