BACKGROUND NOISE IN BROADCASTING 



343 



give the proper change in detector current. It was then modulated 

 30 per cent with a pure tone, and the gain of the audio-frequency out- 

 put ampHfier was adjusted until a fairly loud, but entirely comfortable, 

 level was delivered to an observer placed about six feet in front of the 

 loud speaker. The output level of the audio-frequency amplifier was 

 read on a meter so that its gain might be checked later on. 



The Linear Rectifier 

 The detector of a radio receiver was adjusted to have a linear recti- 

 fier characteristic in the manner just described and manual gain con- 

 trol was employed. In the first set of runs the carrier ratio was de- 

 termined at which the flutter effect at low frequencies, or the carrier 

 beat-note at higher frequencies, became just noticeable, the frequency 

 being the variable. In Fig. 5 is shown a curve representing a number of 



. —^- 



< 10 



10 20 30 40 50 60 70 80 90 100 110 



BEAT FREQUENCY IN CYCLES PER SECOND 



Fig. 5 — Carrier ratio for perceptible flutter with a linear detector. Noise equivalent 

 to 9.5 per cent modulation. 



observations of this type. The noise level was constant at 10 db 

 down from a 30 per cent modulated signal. By this it is meant that 

 when the noise was impressed upon the receiver, together with a car- 

 rier the level of which had been fixed as described above, the audio- 

 frequency output, as measured on a copper oxide level indicator, was 

 10 db below the audio output resulting from a 30 per cent modulation 

 of the same carrier in the absence of noise. 



A very interesting fact to be noted from this curve is that, for beat 

 frequencies of less than about 20 cycles, the carriers must be very nearly 

 equal before any flutter effect whatever may be detected. The 

 average curve has been drawn through a value of 1.5 db. The ob- 

 served values vary from this figure by not more than ±0.5 db. 



