20 



THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1957 



Table II — RMS Xoise Objectives in Busy Hour 



busy hour was agreed as 38 dba (i.e. —46 dbm or 3.9 mv). This was 

 allocated between the various links as in Table II. 



For the program channels, the agreed noise objective was —50 dbm 

 as measured on a C.C.I.F. psophometer with a 1951 program weighting 

 network. 



Statistical data on probable speech le^'els and distributions at London 

 and New York terminals were provided as a basis for repeater loading 

 studies. 



Early planning studies indicated that these objectives would probably 

 be met on all, or nearly all channels without resort to compandors. If, 

 as the system aged, the noise increased owing to increasing misalign- 

 ment, the use of compandors would offer a means for reducing message 

 circuit noise below the objectives. 



The minumum equal-level crosstalk loss between any two telephone 

 channels was set at 56 db for any source of potentially intelligible cross- 

 talk. For channels used for VF telegraph, the eciual-level crosstalk loss 

 between go and return directions was set as a minimum of 40 db; for 

 all program channels the minimum crosstalk attenuation would be 55 db. 



Restrictions of Telegraph and Other Services 



Since the system was being designed primarily for message telephone 

 service, it was agreed that a channel used for any other service should 

 not contribute more to the system rms or peak load than if this channel 

 were used for message telephone, except by prior agreement between 

 Post Office, Bell System and Canadian Overseas Telecommunication 

 Corporation engineering representatives. 



Signalling Objectives 



In order to conserve frequenc.y space, it Avas decided to transmit all 

 calling and supervisory signals within the telephone channel bands and, 



