Alternating Currents of High Frequency - A Reply to Mr. Tesla

I read, with great interest, Mr. Tesla’s and Prof. E. Thomson’s account of their experiments with alternating currents of very high frequency; the greater, as my experience with high speed alternating machinery enabled me to fully appreciate the enormous difficulties connected with the mechanical production of such enormous frequencies.

Indeed, in my experiments, I never went by far so high as they did, and I am not musician enough to judge whether such a machine would be fit to accompany a thoroughly Wagnerian opera, as Mr. Tesla thinks; but, at least, I doubt it, because the one machine - of a design somewhat similar to his - which I once built, gave out a howling resembling much more an Indian war whoop, only a little more frightening, than anything else. And it was a very remarkable machine, too; the boys in the shop called it the “humming bird,” and it goes now by this name. Perhaps I shall refer at a later occasion to the interesting experiments made with this machine, which, among other features, gave continuous magnetisation from alternating currents without commutation.

It seems, indeed, that the advantages of reaching such high frequencies, not by static discharges, but with machines powerful enough to yield perceptible currents of electricity, amply compensate for all the difficulties encountered therewith. Because these experiments made with frequencies of 20,000 to 30,000 per second form the connecting link between the phenomena of the alternate currents used in practical engineering, and those electric disturbances of millions or hundreds of millions of vibrations per second, which give rise to the Hertzian ether waves of radiating electricity.

And this fact explains all the experiments shown by Mr. Tesla and Prof. Thomson. Most of them, indeed, show truly static effects of oscillating electric disturbances, like the oscillating discharge of the induction coil in Hertz’s experiments, only that in this case where the electric disturbances are backed by currents, which count by horse-powers, much more brilliant effects appear.

It is certainly not conduction through the air which makes an incandescent lamp, entirely disconnected from the machine, light up, but it is the action of the radiating energy of the electric ether waves. Not the secondary current of the induction coil is felt by the hand which holds the incandescent lamp, but the electric charge influenced in the hand as the outer condenser coating, just as we draw sparks from the outer coating of a Leyden jar when the inner coating is charged by a static-electric machine, and feel every spark which jumps over from the machine conductor to the inner coating.

Various forms of the screening effect of metals, with regard to those waves, are shown by Mr. Tesla. Reflection and refraction, indeed, he could hardly expect to observe, because, even with 30,000 alternations per second, the wave length of the electric disturbance in space amounts still to something like seven miles.

But the whole space surrounding such a machine has to be considered as an electro-magnetic field of force, and a very powerful one, too, because of the enormous strain exerted upon the dielectric by the high pressure of the induction coil.

It is only to be regretted that in all the reports of Mr. Tesla’s experiments nowhere anything is stated about the E.M.F. of those electric secondary currents. This omission rather impairs the value of these experiments somewhat. The E.M.F. of the primary current Mr. Tesla gives as 500 volts. But how did he measure the alternating pressure of 30,000 vibrations per second; by static electrometer? Because any dynamometer would be unreliable under these circumstances.

I think, although any closed circuit voltmeter would be entirely out of question for such high frequencies, a modified electro-meter, built for very low capacity, would, perhaps, answer the purpose of measuring the secondary E.M.F., and could be built so that it could be used for determining the E.M.F. and also the electric energy, that is, the power consumed by the circuit. This seems to me of very great importance, because it would give the means to determine by experiment the amount of electric energy which radiates into space as electric ether waves, that is, the time hysteresis of air.

I should suggest to determine the whole energy of the electric current by an electrostatic wattmeter, and the energy converted into heat in the conductor by a calorimeter. The difference would give the energy of radiation. Indeed, enormous practical difficulties would have to be overcome to carry out this scheme.

__________________

The condenser, such an inefficient and cumbersome apparatus for circuits with the usual number of periods, becomes very powerful and compendious for such high frequencies as 30,000 per second, and systems of electric distribution by means of condensers would be very well conceivable, if much higher frequencies than now could be used in practical engineering. Already some years ago such a system of distribution by condensers, with a frequency of, I believe, 800 per second, was proposed by a European electrician, but never tried, I suppose. (Account of it is to be found somewhere in the “Centralblatt für Electrotechnik,” Munich, Germany.)

However, I believe that for electric light and power distribution such frequencies as Mr. Tesla used are quite out of question, and I am much more inclined to think that even the frequencies now in vogue in this country, of in maximo 138 complete periods per second (Westinghouse), or 16,000 alternations per minute, will not be kept much longer, but will be reduced as soon as possible. They are now used only because of the small unit of converters found in this country, which for efficient work need a high frequency.

But I entirely agree with Mr. Stanley that it is most desirable for efficient work to proceed to larger converter units, and that then a reduction of the number of alternations will be made. In the old country, where the practice is to use the converter as a kind of secondary station and to feed a whole district from one converter or a bunch of converters, the converter unit is much larger generally, and therefore a much lower frequency, only 42 complete periods per second, permissible. This reduces the loss by hysteresis very greatly, and this loss is known as the most important factor, because of being about the same for no load as for full load, and especially determines the all-day efficiency.

Indeed, even in this country, for arc lighting by means of alternating currents and series converters, the number of alternations has been reduced already to one-half, 8,000 reversals per minute, in the Stanley-Westinghouse constant current alternator, I think, because it was very difficult to work any electro-magnetic regulating mechanism with the frequencies of incandescent light currents.

It is not the question whether magnetism can be produced at all by very rapidly varying currents, but the question of self-induction, that is, whether a current strong enough to work the regulating mechanism can be forced through its energising coil against the counter E.M.F. of self-induction, without consuming too much pressure, which limits the number of vibrations permissible in electro-magnets and motors.

I will not try to decide which is the highest number of periods permissible for effective working of alternating motors, and whether this limit is higher than the number of periods practically used; but it must be borne in mind that until now no alternate current motor is on the market which runs with halfway fair efficiency with the number of alternations used for incandescent light circuits, 16,000 per minute, while in Europe, where a much lower number of alternations, only 5,000 per minute (Ganz & Co.) is used, alternate current motors are already on the market and in practical use and reach a very fair efficiency. And, nevertheless, in the use of electricity as motive power, the United States are undoubtedly a long way ahead of other countries.

As before said, it is not the impossibility of producing magnetism by rapidly varying currents which makes them unfit to run motors, but the enormous self-induction, which necessitates the waste of an enormous pressure to send even a small current through the inductive resistance.

For magnetism is produced with still much more rapidly oscillating currents than those which Mr. Tesla had at his disposition, as the fact proves that magnetisation, and even permanent magnetisation of steel rods, has been produced by oscillating condenser discharges and lighting strokes. Steel rods were found permanently magnetised at such occasions, and the curious phenomena was, that by grinding off the surface of the steel rod, the magnetism decreased - reached zero, reversed its polarity - increased by further grinding off to a maximum in the opposite direction as that of the magnetism of the integer rod, and then decreased again and even reversed its polarity once more.

__________________

The explanation for this phenomenon is that the discharge was oscillating, with increasing wave length but decreasing intensity. The first wave magnetised only the very surface of the steel rod strongly, but because of its very short duration and the screening effect of the metal, did not enter deeper into it. The second wave, of opposite direction, was not strong enough to reverse the magnetism produced by the first wave, and only weakened it somewhat; but because, of its longer duration it entered deeper into the steel and magnetised the layer underneath in the opposite direction to the surface, so that by grinding off the surface layer the magnetic moment of the rod reversed; and if the resistance of the electric circuit was low enough to allow a third wave to be produced with perceptible intensity, underneath the second layer a third one was found, magnetised in the same direction as the surface. This phenomenon shows that even such very rapid oscillations produce magnetism.

I was very much interested in reading the arguments which Prof. Thomson proffered for the existence of carbon vapour in the electric arc, and I must consider the arguments as very strong. The only fact which seems to me not quite in agreement with the existence of carbon vapour, is the spectrum of the electric arc. The copper arc, the iron arc, show the characteristic spectra of those metals, with their many lines. The carbon arc, if containing carbon vapour, ought to show the - until now unknown - carbon spectrum. But nothing of that kind; it gives an entirely continuous spectrum, quite similar to the spectrum of glowing solids.

This fact, I believe, needs still a satisfactory explanation, and I should like to hear how Prof. Thomson explains the absence of a characteristic carbon spectrum in the electric arc.

I should like to hear, also, whether any experiments have been made to determine the exact temperature of the arc and the glowing carbons. The only temperature determinations of which I know are already of rather old date and made by an Italian, Rosetti, which measured the heat radiation by means of a thermopile and found the temperature of the tip of the positive carbon to 3,900° C., the tip of the negative carbon to 3,150° C., but the temperature of the flame higher - 4,800 to 4,844° C. These figures do not agree with the data given by Prof. Thomson, who considers the positive carbon as the hottest part of the circuit. Rosetti used currents from 3 to 6 amperes.

The explanation of the instability of the electric alternating arc, because of the alternating cooling off of the carbons during the zero period of the current is certainly the nearest one, and, therefore, generally accepted. But whether it agrees with the real proceedings must still be doubted. Theoretically, at least, these fluctuations of temperature should be expected to be so small that they could have no influence on the conductivity of the arc.

An experiment to determine the fluctuations of temperature of an electric conductor, caused by the alternations of the current, was made on a piece of copper wire, one millimetre thick and 657 centimetres long. A current of 6•4 ampères was sent through with only 5,840 alternations per minute, and the lengthening of the wire measured. The increase of temperature was 37° C.; but the maximum variation of temperature calculated to only •019 degree C., that is, one-twentieth per cent, of the whole rise of temperature. (“Das Gesetz zwischen Ausdehnung und Stromstärke für einen von galvanischen Wechselströmen durchflassenen Leiter,” von Dr. Cranz, Zeitschrift für Mathematik und Physik, Dresden, 1889, p. 92.) Although this calculation is not quite conclusive for the somewhat different conditions in the electric arc, it is sufficient for making the assumption of so great and rapid cooling down of the arc during the time zero current doubtful.

But if we consider the energy consumption of the arc as caused by a counter E.M.F., a kind of polarisation, and at the hand of the facts must assume that this counter E.M.F. of the arc in very wide limits is nearly independent of the current strength, then the apparent resistance of the arc must vary inversely as the current intensity, and in consequence of this, in an ordinary alternating arc light circuit, the height of the current wave must be increased at its crest, decreased near its minimum position, so that the time of minimum current is greatly increased, and the shade of the wave becomes more triangular.

But if the alternator is so well self-regulating that the current wave, in spite of the varying resistance, is kept sinusoidal, that is, regulating in time so short that it is small even in comparison with the wave length of the alternating current, as is the case in the Stanley-Westinghouse constant current alternator, then the alternating arc should become just as steady as the continuous current arc. Now, in Mr. Tesla’s high frequency machine this condition was certainly fulfilled, its self-induction being very large, as the tendency of yielding constant current shows, and self-induction regulated the impressed E.M.F., according to the sine wave of the current producing the rectangular shape of the wave of impressed E.M.F., as shown in the curves given by Messrs. Tobey and Walbridge, for the Stanley-Westinghouse alternator, in their paper read before the American Institute of Electrical Engineers.

This particular shape of E.M.F. and current might have a great deal to do with the steadiness and unsteadiness of the arc.

Therefore, it would be of very great interest, in the same way as it has been done in Cornell University with the Stanley-Westinghouse alternator, to determine the shape of the E.M.F. and the current wave for an alternator of very low self-induction, for instance, with a disc armature, running arc light. Then I should expect to find the opposite phenomenon; the impressed E.M.F. retaining its sinusoidal form, and the current wave assuming a more triangular, shape, at the same time causing unsteadiness of the arc.

Yonkers, N.Y., March 26th, 1891.