MULTICHANNEL MODULATION SYSTEM 43 



Radio Interference and Noise. To obtain experimental confirmation of 

 the expected tolerance to high interference levels in the radio path, the 

 output of an oscillator, tunable through the band near 65 megacycles, was 

 superimposed upon the received intermediate-frequency signal at the input 

 to the group selection filters. With this controllable interference tuned near 

 the center of the Group 1 filter, and its amplitude set 6 decibels below the 

 peak amplitude of the (noise-free) pulses, errors were so plentiful that the 

 demodulator did not remain synchronized. Proper framing was restored 

 when the amplitude difference was increased to 7 decibels, but enough 

 decoding errors remained to give intolerable audio noise. At 8 decibels 

 only an occasional crackle of noise was observed, and at 9 decibels reception 

 was perfectly normal. Similar tests of Group 2 gave the same results except, 

 of course, that synchronization was not affected. The fact that noise-free 

 transmission was not maintained quite up to the ideal 6-decibel point is due 

 principally to the width (0.4 microsecond) of the gate which is applied to the 

 rounded PCM pulses entering the receiving equipment. If necessary the 

 ideal could undoubtedly be approached more closely by reducing this 

 width, thus admitting only the extreme peaks and troughs of the signal. 



The effects of actual fluctuation noise were studied by sending the PCM 

 pulses over the radio path at reduced level. The boundary between good 

 and bad transmission was not so sharp as with the continuous-wave inter- 

 ference, as should be expected because of the random nature of the noise. 

 Flawless reception occurred when the root-mean-square signal at the peak 

 of a pulse was greater by 18 decibels than the root-mean-square noise, both 

 measurements being made at the output of a group selection filter. 



VI. Acknowledgment 



The system described is the result of the co-ordinated efforts of many 

 people. In particular, the writers wish to express their appreciation to 

 Mr. R. W. Sears for his development of the coding tube, and to acknowledge 

 the contributions of certain others who have been actively concerned with 

 various phases of the project. These include Messrs. R. L. Carbrey, A. E. 

 Johanson, J. M. Manley, G. W. Pentico, A. J. Rack, P. A. Reiling, L. R. 

 Wrathall, and the mechanics and wiremen who have kept to their usual high 

 standards throughout countless circuit changes. 



The project was under the overall direction of Mr. C. B. Feldman. 



