380 Pupm — Resonance Analysis of Alternating Currents. 



with little success. There is another reason why a new method 

 of studying alternating current waves seems desirable. It is 

 this : the method of sliding contact is not sufficiently sensitive 

 to detect small deviations from a true sine wave, and conse- 

 quently it is not capable of following up the causes of these 

 deviations, when the effects seem to be absent. For instance, 

 the primary current of a transformer can differ very much 

 from a true sine form when the secondary circuit is open, but 

 when a large current is flowing through an approximately non- 

 self-inductive secondary circuit, then the primary current can 

 be made to differ inappreciably from a true sine wave. The 

 question arises now, what becomes of these causes when the 

 secondary carries a heavy non-self inductive load? 



This question is of deep scientific interest ; it is also of con- 

 siderable technical importance. For, if these causes are present 

 at all loads and only hidden by the principal wave, then, con- 

 sidering that these hidden small causes can produce large 

 effects when conditions favoring resonance arise, it is evident 

 that they must be carefully watched and guarded against in 

 the construction of lines possessing appreciable distributed 

 capacity. I do not think that indicator diagrams obtained by 

 the method of sliding contact are capable of giving a definite 

 answer to this important question. 



The method of analyzing alternating current waves by elec- 

 trical resonance which I employed in the following investiga- 

 tion was suggested by me a year ago.* It is the object of this 

 paper to describe this method at some length and to illustrate, 

 by some of the more definite results so far obtained and relating 

 principally to the causes which produce distortions in simple 

 harmonic waves, the simplicity, sensitiveness, and reliability of 

 the method. I shall also point out that this method of reso- 

 nance analysis works quite satisfactorily even in those cases, 

 alluded to above, where the sliding contact method would in all 

 probability fail to detect any distortion whatever. 



IT. Description of the Method. 

 Consider the following arrangement of circuits : — The non- 

 self-inductive resistance, ah, fig. l a , is inserted in the circuit of 

 an alternator A and the primary B of a transformer. In shunt 

 with ah is a circuit acclh consisting of an inertia coil c of large 

 number of turns of copper wire of low resistance, about 10 

 ohms, but containing no iron, and a mica condenser d divided 

 into subdivisions ranging from "001 M.F. up. In shunt with 

 the condenser d is an electrostatic voltmeter e. The self-induc- 

 tion of the coil c can be varied by throwing a larger or a smal- 



* M. I. Pnpin, '' Electrical Oscillations of low frequency and their Resonance," 

 this Journal, vol. xlv, p. 429, May, 1893. 



