The Theory of the Operation of the Howling 

 Telephone with Experimental Confirmation 



By HARVEY FLETCHER 



Synopsis: A general theory of the sustained oscillations of electro- 

 mechanical systems is presented in the paper. The electrodynamical 

 properties of the telephone transmitter and receiver are described and 

 sufficient numerical data are given to enable one to calculate the intensity 

 and frequency of howling for various tj'pes of systems. Detailed con- 

 sideration is given to the following three systems, namely, one where the 

 transmitter and receiver disphragms are coupled together mechanically 

 by a lever system, one where they are coupled by a small box of air, and 

 one where they are coupled by a long tube of air. The type of electrical 

 circuit to use with each of these systems depends upon the type of perform- 

 ance desired. 



WHEN the telephone receiver of a subscriber's set is held in 

 front of the mouthpiece of the transmitter, a shrill note is 

 emitted. A sustained oscillation is set up in the electro-mechanical 

 system which is frequently called "howling" or "singing" or "hum- 

 ming." 



This phenomenon was first observed by A. S. Hibbard of 

 the United States in 1890. Frank Gill was the first to publish an ac- 

 count of the phenomenon. He first noted that the pitch of the howling 

 note was changed by reversing the telephone recei\"er connection. In 

 summarizing further his experimental results, he states "that the pitch 

 of the note appears to be determined by the length of the column of air 

 between the two diaphragms and the conditions of the circuit. As the 

 periodic time of the circuit is increased, the time of the note rises. To 

 some extent, the pitch is governed by the rate of the diaphragm, but I 

 do not think this is so important a factor as the others. The main 

 factors appear to be the angle of lag and the length of the column of 

 air between the diaphragms. Although the vibration is a forced 

 one, we could almost see that its rate is largely dependent on the 

 free period of the circuit." ^ 



In 1908 Kennelly and Upson extended Gill's work and made ex- 

 tensive experimental investigations of the case in which the trans- 

 mitter and receiver are coupled together acoustically by means of a 



1 Taken from a paper on "Notes on the Humming Telephone" by F. Gill, read 

 at a meeting of the Dublin Local Section of the Society of Telephone Engineers 

 and published in the Journal of the Institution of Electrical Engineers, \'ol. XXXI, 

 1901. 



27 



