STANDARD VOLUME INDICATOR AND REFERENCE LEVEL 113 



The explanation probably lies in the physiological and psychological 

 factors involved in the ear's appreciation of overload distortion, which 

 permit to pass unnoticed considerable amounts of distortion on rarely 

 occurring instantaneous peaks of very short duration. 



Peak Checking 



A very important use of volume indicators is that of checking the 

 transmission losses or gains along a program network by measurements 

 made on the transmitted program material [item (6) in" the list given 

 earlier]. The program circuits which make up the large program 

 networks are in continuous use for many hours each day, and during 

 that period are switched together in many combinations as called for 

 by the operating schedules. It is not convenient to interrupt service 

 for sine-wave transmission measurements; hence to check the trans- 

 mission conditions during service hours, it is the custom to take simul- 

 taneous readings at two or more points in the program networks on 

 particular impulses of whatever program wave is being transmitted, 

 coordinating these readings by the use of an order wire. On such 

 readings, the r-m-s type of instrument is far superior to the peak- 

 reading type, because of phase distortion and slight non-linearity in 

 the program circuits. These effects are undetectable to the ear, but 

 change the wave shape of the program peaks sufificiently to cause 

 serious errors in the readings of the peak-type instrument. On the 

 other hand they have no noticeable efifect on the r-m-s instruments. 



Tests were made on this effect by taking readings on several kinds 

 of program at the beginning and end of a program circuit extending 

 from New York to Chicago and return (about 1900 miles). The circuit 

 was lined up so that either volume indicator read the same at both 

 ends of the circuit on a 1000-cycle sine wave. In all the tests, the 

 readings obtained on program material with the r-m-s instrument at 

 the two ends of the circuit agreed within a very few tenths of a decibel. 

 The readings of the peak instrument, however, disagreed by the values 

 shown in Table III, when the program material was applied to the 

 circuit at the normal maximum operating level. 



It is of interest that the errors shown by the table are affected by 

 the frequency range of the program material transmitted, being greater 

 for the broader band. The frequency range was controlled by the 

 use of low-pass filters inserted between the source of program and the 

 line before the point at which the sending end levels were read. Tests 

 were also made of the effect of a 180-degree phase reversal at the center 

 of the loop. This was found to increase the errors in some cases and 

 to decrease them in others. 



