380 BELL SYSTEM TECHNICAL JOURNAL 



2. Vodas at One End, Echo Suppressor at the Other 



The above possibilities of transmitter overloading and speech 

 reradiation due to echoes might be prevented by adding a voice- 

 operated echo suppressor at the B end as shown in Fig. 5. This device 

 would be operated by received speech so as to disable the transmitting 

 branch of the circuit. Its sensitivity would be limited as would also 

 the received volume at one or both ends. It should be noted that if 

 no cross transmission paths existed, relay 2 of the vodas at the A end 

 could be omitted when this device is used. 



The echo suppressor would not suppress echoes of radio noise or 

 direct transmission of line noise from the B end to the receiving relay 3 

 at the A end. Relay 3 would therefore need to have its sensitivity 

 reduced so as not to be operated by these noises. This gives rise to an 

 additional limitation of received volume at the A end. The amount of 

 the penalty would, of course, depend on the noise conditions, the talker 

 volume and the echoes in two-wire circuits at the B end. Under ex- 

 treme conditions, the necessary reduction in receiving volume at the 

 A end might be as much as 25 db. This is considered to be an impor- 

 tant disadvantage inasmuch as conditions at the two ends are not 

 independent and lack of an anti-singing device at one end penalizes 

 the received volume at the other. 



Another solution would be to limit the transmitting gain at the B end 

 so that the noise transmitted past the echo suppressor would never 

 limit the sensitivity of relay 3 at the A end. This would mean that 

 the received signal-to-noise ratio at the A end would be reduced, 

 particularly if the taker at the B end were weak. 



3. Vodases at Both Ends 



To summarize, it may be said that anything short of anti-singing 

 devices at both ends does not make the two ends of the circuit in- 

 dependent and may penalize the transmission at the vodas end when 

 there is radio noise or line noise at the end without an anti-singing 

 device. The preferred arrangement is shown in Fig. 6. 



Results of Vodas Operation 



In general, the results of operation with the vodas have been good 

 and, when radio conditions are favorable, the circuits are not appreci- 

 ably different from land circuits of comparable length. 



Occasionally, the vodases introduce minor difficulties. False opera- 

 tion by noise and simultaneous talking by the two subscribers both 

 tend to cause speech mutilation. The large transmission advantages 



