136 BELL SYSTEM TECHNICAL JOURNAL 



and second, the existence of a relatively high noise level caused mainly 

 by radio-frequency interference from electrical systems within the 

 building. 



The results of a few observations upon signal strength reduction 

 within two buildings are presented in Figs. 10 and 11. Fig. 10, for 

 a non-steel building, shows the field to be roughly halved. In the 

 case of the steel structure building depicted in Fig. 11, the interior 

 field is found to be reduced to as low as a few per cent of that outside 

 the building. For outside rooms, the field strength near the window 

 was found to be about eight times that further in the room. Such 

 severe shielding effects obviously call for picking up the wave energy 

 outside the building and conducting it to the receiving sets by wire 

 circuits, preferably by shielded circuits, in order to protect against 

 local interference. 



Multi-Station Operation 



The discussion given above has been directed chiefly to the relations 

 which might be called internal to a single-channel radio broadcast 

 system. Actually, of course, broadcasting involves the use of the 

 common transmitting medium for a number of channels. This brings 

 with it the problem of frequency selectivity and raises the question of 

 the capabilities of the various types of radio receiving circuits. 



In order to throw some light upon this important factor, measure- 

 ments have been made upon a sample or two of each of a number of 

 different types of radio receiving circuits. The measurements were 

 made in the laboratory,^ simulating as closely as possible the condi- 

 tions under which the receiving sets would be used. The curves of 

 Fig. 12 show the reduction which is to be expected in the detected 

 audio-frequency current, were the receiving set tuned to a transmitting 

 station on 900 kc, and the transmitting station then shifted in fre- 

 quency by the amounts given along the abscissa. In this curve the 

 reduction in current is indicated both as a ratio and in TU, which is 

 a convenient way of indicating power ratios. The relation between 

 TU and current ratio with a given impedance is indicated in the 

 figure. Thus, for a carrier 40 kc. off from the one to which the 

 set is tuned, the single-circuit, non-regenerative type of receiver 



^ The method consists in establishing a small laboratory transmitter and modu- 

 lating it with a single-frequency tone. The receiving set is tuned to the modulated 

 carrier signal as in practice. The gain or sensitivity of the receiver and its coupling 

 with the transmitter are adjusted to produce normal load upon the detector tube. 

 With the receiving set left at this adjustment the frequency of the radio transmitter 

 is shifted each side of the original single frequency in 10 kc. steps throughout a range 

 of 50 to 100 kc. For each of the offside frequencies the reduction caused in the 

 detector output current is measured, this being an indication of the receiving set 

 selectivity. 



