142 



NA TURE 



[December 6, 1900 



OSCILLOGRAPHS. 

 'T*HE phenomena connected with the behaviour of alternating 

 currents present a wide field for experimental inquiry 

 which has, up to the present, been but imperfectly explored. The 

 investigation of the wave-forms of alternating potential differences 

 and currents under various conditions of their actual use is a 

 matter, not only of great theoretical and scientific interest, but 

 also of the highest practical importance, since the shape of the 

 wave-form under given conditions, and the alteration of shape 

 produced by any alteration of conditions, are factors which largely 

 affect the efficiency and economy of working. As examples 

 showing the increased efficiency that may be obtained by choos- 

 ing a suitable wave- form, we may quote the results obtained by 

 Messrs. Rossler and Wedding, and by Messrs. Barr, Beeton and 

 Taylor. The former experimenters showed,^ in an investiga- 

 tion on the luminous efficiency of the alternate current arc, 

 that the light per watt when using an alternator giving a flat- 

 topped E.M.F. curve was 44 per cent, higher than when using 

 a machine that gave a peaked curve. Messrs. Barr, Beeton 

 and Taylor, in a research on the efficiency of transformers, found,'-' 

 on the contrary, that a peaked wave-form was the most suitable 

 one to employ. The reactions that take place between alter- 

 nators running in parallel is another case in which the wave- 

 ■ form is of very great practical importance. 



It will be readily understood, therefore, that it is most desir- 

 able that we should have some simple method of observing and 

 studying the wave-form of an alternating current or potential 

 difference. Such a method is supplied by the instruments known as 

 oscillographs. Before the invention of these instruments the only 

 means of studying wave-forms was by the exceedingly laborious 

 " point-to-point" method. Suppose that there is a circuit through 

 which is flowing an electric current which is varying periodically 

 at the rate of, say, n, complete cycles per second, and that it is 

 desired to obtain the wave-form of this current. At any par- 

 ticular instant the current will have a certain definite magnitude 

 at^d direction, and i/«th of a second later the current will again 

 have the same magnitude and direction. If, by means of an 

 automatic contact-maker, a galvanometer is brought into circuit 

 at intervals of i/«th of a second, there will be given to the 

 galvanometer needle a succession of impulses due, in each case, 

 to the same current, and consequently a steady deflection of the 

 needle will be produced from which the particular instantaneous 

 value of the current can be determined. To obtain, however, 

 the complete wave-form, we must determine the value of the 

 instantaneous current at every tnoment during the cycle, or at a 

 sufficient number of moments to enable a smooth curve to be 

 drawn. Having found one point on the wave-form in the vvay 

 described, the contact-maker is shifted so as to bring the gal- 

 vanometer periodically into circuit at some other moment in the 

 cycle, and a second point on the curve is then found. Again 

 the contact-maker is shifted and a third point is obtained, and 

 thus, point by point, the whole curve may be built up. 



This method is open to two objections. In the first place, it is 

 only applicable to cases in which the wave-form is undoubtedly 

 steady, all transient effects being obviously unobservable by such 

 a process, and, secondly, it is so lengthy that elaborate researches 

 are practically precluded. As much as four or five hours may, 

 indeed, be spent in obtaining a single curve, and then, even after 

 all this labour, it is more than possible that the conditions will 

 be found to have altered during the experiment, and the curve, 

 in consequence, to be useless. By means of the oscillograph 

 and kindred instruments, however, experiments can now be car- 

 ried out in a few minutes which occupied days by the old ' ' point- 

 to-point " method, and not only can steady wave-forms be 

 examined, but the most fleeting effects can be studied with 

 equal ease. 



An oscillograph may be defined as a galvanometer the deflec- 

 tion of which, at any instant, is practically proportional to the 

 current flowing through it at that instant, in spite of the current 

 varying very rapidly in strength or in direction. For this to be 

 possible it is necessary for the free periodic time of the moving 

 part of the instrument to be very short, less, generally, than 

 i/30th, of the periodic time of the effect to be observed. The 

 instrument must also be perfectly dead-beat, the moving part 

 taking up instantaneously the deflection proper to the current 

 flowing through the instrument, for if there be any tendency to 

 overshoot the mark it will cause the observed wave-form to be 

 distorted from its true shape. In addition to this, since one is 



1 The Electrician, 1894, vol. xxxiii. 



2 Journal oi the Institution of Electrical Engineers, 1896, vol. xxv. 



dealing with rapidly varying currents, it is necessary for the 

 self-induction of the instrument to be practically nil, and for 

 all effects due to hysteresis or eddy-currents to be eliminated. 

 The original idea of the oscillograph is due to M. Blondel, who 

 pointed out, in 1893, the principles on which such an instrument 

 should be designed, and all the oscillographs since produced owe 

 their inspiration to M. Blondel's work. 



Two other instruments have been invented and developed by 

 which the same end might be attained, namely, the observation 

 or recording of rapidly varying currents or potential differences. 

 These are the Abraham-Carpentier rheograph and 1 he Hess-Braun 

 oscillo radiograph. With these instruments, since they are not, 

 strictly speaking, oscillographs, we do not propose to deal in 

 detail, but must content ourselves by giving a brief account of 

 the principles on which they are constructed. In the rheograph, in- 

 stead of making the free periodic time of the instrument excessively 

 small, M. Abraham uses a galvanometer with a period of about 

 l/iothof a second, and attempts to compensate errors due to the 

 inertia of the moving part by utilising the effects of electromag- 

 netic induction. With this instrument M. Abraham, it is said, 

 has been able to study oscillating discharges having a period of 

 about i/io,oooth of a second ; but the adjustment is not an easy 

 matter, and, moreover, has to be made every time the ini^irument 

 is used. In the Hess-Braun oscillo- radiograph the difficulty of 



NO. 1623, VOL. 63] 



Fig. I. — M. Blondel's oscillograph. 



sufficiently reducing the inertia of the moving part is overcome 

 in a very ingenious manner by using, as the galvanometer 

 " needle," a beam of kathode rays in a vacuum tube. This beam 

 is arranged to produce a bright spot on a fluorescent screen, and 

 the movements of this spot are observed when the beam is 

 deflected by the varying currents passing through two bobbins of 

 wire on either side of the vacuum tube. Unfortunately these 

 bobbins, possessing self-induction, introduce errors. Another 

 difficulty in this apparatus is to obtain good definition, and also 

 sufficient intensity of illumination. On account of the small 

 intensity it is only possible to use this instrument for the study 

 of cyclic phenomena where, as the spot of light can be caused to 

 travel over the same curve again and again, the curve can be 

 observed or photographed. 



Neither of these instruments has been brought to the same 

 degree of perfection as the oscillograph, which has been developed 

 into a very perfect instrument by M. Blondel in France and by 

 Mr. Duddell in England. M. Blondel originally suggested three 

 systems on which oscillographs might be constructed, in which 

 the moving part consisted respectively of a small bar of soft iron, 

 a vibrating plate of iron, and a light coil on a bifilar suspension. 

 The instrument which M. Blondel first perfected, and with 

 which his well known researches on the alternate current arc 

 were carried out, was constructed on the first of these systems. 

 The diagram (Fig. i) shows the chief principles of its construe- 



