IMPROVED CIRCUIT FOR TELEPHONE SET 623 



his voice when the ambient noise level is high. A major factor in the 

 impairment of received speech in the noisy location is the masking of 

 received speech by noise picked up by the transmitter and delivered 

 via the sidetone path to the receiver. Therefore, a telephone set which 

 minimizes the masking sidetone noise during the listening interval should 

 operate more satisfactorily in noisy locations. In many previous in- 

 stances where the* noise conditions were extreme, a "push-to-listen" 

 switch was provided to cut off the local transmitter while listening and 

 thus prevent the introduction of noise to the receiver via the sidetone 

 path. With its higher efficiency and better sidetone balance, the 500- 

 type set approaches the good performance of such a * Variable" set under 

 noisy conditions and provides this adequate performance to the user 

 more conveniently and with the expenditure of less effort. 



Laboratory tests have been made which permit appraisal of the 

 merits of various telephone sets or modifications of them under high 

 noise level conditions. These include talking and listening tests between 

 two telephone stations under typical loop and central office conditions 

 with an adjustable amount of typical room noise at the listening end. 

 During the tests the length of trunk between the two stations was in- 

 creased until the received speech was just sufficiently intelligible to per- 

 mit carrying on a conversation. This threshold of intelligibility expressed 

 in db of trunk loss was taken as a criterion of the performance of a given 

 set. 



Fig. 10 shows the results of the tests made with the 500-type set 

 compared with the 302-type set, and with variations of the 500 type set. 

 500C- and D-type sets having silicon carbide varistors were used in this 

 comparison, but in this respect the performance is no different than for 

 the earlier design (500A and B) . The curves shown are for the average of 

 eight observers. 



The abscissae of the curves are the noise levels at the listening end 

 expressed in db above 10~ watts per sq. cm. For reference purposes it 

 should be kept in mind that the average noise in a fairly large business 

 ofl&ce is, on this scale, approximately 65 db. For further orientation as 

 to the significance of the noise levels shown, it should be noted that when 

 a level of about 100 db or higher is reached, it is extremely difficult to 

 carry on a face-to-face conversation where no telephone is involved. 

 The ordinates show the relative trunk loss in db over which it is possible, 

 under the stated conditions, to just carry on a telephone conversation. 

 Therefore, the larger the trunk loss, the better the performance. 



It is seen from Figure 10 that the 500-type set is from 6 to 8 db better 

 than the 302 set over the indicated range of ambient noise conditions. 



