DISCUSSION ON

METALLIC TUNGSTEN ANd Some of itS APPLICATIONS (COOLIDGE), AND "THE CONVECTION AND CONDUCTION OF HEAT IN GASES" (LANGMUIR), BOSTON, MASS., JUNE 25, 1912.

C. M. Green: Dr. Weintraub of the research laboratory at Lynn has suggested a new use of tungsten, that is as a substitute for platinum leading-in wire for rectifier tubes. I am thoroughly satisfied from experiments I have made that it will be better than platinum. The specific resistance is about one-half that of platinum, or about 5 to 6 microhms per cu. cm., while the platinum which we use at the present time is about 13 to 14 and the platinum used a few years ago had a specific resistance of 30 to 40 microhms.

A Member: I was very much interested in Mr. Coolidge's paper and I tried in the first instance to see if the tungsten could be plated with copper. I was interested in his statement that it could be wet with copper, and I would like to inquire if he has ever tried to find out whether the metal could be plated with copper, because, if it could, it would be very easy to solder it to other metals.

William J. Hammer: I would like to ask Mr. Coolidge whether he has observed any special changes in the characteristics of tungsten while it is being gradually heated. I recall that in Mr. Edison's very early experiments at Menlo Park, whereas his platinum would only give a light of say from three to four candle power before it melted, by gradually putting little increments of current through the platinum the occluded gases were driven out and the metal was made more dense and became so hard that a sharp file would not mark it, and in this condition the platinum could be brought up to an incandescence of from 30 to 34 candles without melting. Mr. Edison also alloyed the platinum with iridium in many of his lamps.

Carl Hering: I would like to ask Mr. Coolidge how the tungsten acts when used as electrodes, and whether it will stand the oxidizing influence in being used as an anode in ordinary electrolysis.

W. D. Coolidge: The question has been raised as to whether tungsten could be plated with copper. That is possible and fairly easy, and it is possible to go ahead as suggested and solder such a copper-plated piece of tungsten. That method we have not found so satisfactory as the method we are now using with molten copper, the adhesion not being anywhere near so firm as we get from the present method. The plating goes best in a copper sulphate solution. Mr. Hammer raised a question about occluded gas. We have done a great deal of work along this line and have found that the greater part of the gas which is in the drawn wire comes off very quickly when the wire is heated in a vacuum. Dr. Hering raises the question of whether tungsten can be used as an anode for electrolysis. As a matter of fact tungsten is oxidized and goes into solution very readily indeed

under such conditions. We had hoped that it could be used in place of platinum. H. M. Hobart: In the latest edition of the A. I. E. E. Standardization Rules, paragraph 269 relates to the application of temperature corrections to take into account the room temperature on the occasion of tests and to permit of arriving at values corresponding to the reference room temperature of 25 deg. cent. The rule is based on the assumption that the observed temperature rise will be greater the higher the temperature of the surrounding air. My experience does not conform to this rule, but shows rather that the temperature rise will generally be less, the higher the temperature of the surrounding air. It is very important to engineers that any uncertainty in this matter should be cleared up and I should be pleased if Dr. Langmuir would give us the advantage of his experience and opinions bearing upon this point.

If an electrical machine is to be operated in a place where the surrounding temperature is likely to be high, say 40 deg. cent., then it is very important to be able to know in advance whether a certain machine, tested at, for example, 20 deg. cent., and then sustaining a temperature rise of 40 deg. cent., will, under the conditions of its practical operation in the hot location, have a rise of 44 deg. cent., or only, say, 38 deg. cent. At present the data are very conflicting, and some engineers would be of the opinion that the rise when the surrounding temperature is 40 deg. cent. would amount to 44 deg. cent., thus bringing the temperature of the machine to 84 deg. cent., while the tests to which I allude would indicate that the rise would be somewhere between 36 deg. and 40 deg. cent., say 38 deg. cent., thus making the actual temperature of the machine, when the surrounding atmosphere is 40 deg. cent., only some 78 deg. cent.

Any information which Dr. Langmuir could furnish in this matter would be of much value.

Irving Langmuir: In reply to Mr. Hobart's question, I would like to point out that there are apparently only two factors involved in convection: first, the heat conductivity of the air, and second, the thickness of the film which determines the shape factor. In quiet air this shape factor can be calculated in the cases of wires and plane surfaces. However, in other cases (as where several wires are close together or where we have a confined space, such as might exist in the armature of a dynamo) this shape factor would be difficult to calculate, but in any case it is probably nearly independent of the temperature, and we may say that the heat convection would vary with the temperature practically only because of the temperature coefficient of the heat conductivity of the air itself. Since the latter nearly doubles with the 300 deg. rise in temperature, it is apparent that the convection should increase with increase in temperature of the air in which convection takes place.

of the American Institute of Electrical Engineers, Boston, Mass., June 25, 1912.

Copyright, 1912. By A. I. E. E.

CHARACTERISTICS AND APPLICATIONS OF VIBRA

TION GALVANOMETERS

BY FRANK WENNER

In the vibration galvanometer we have a type of synchronous motor which is distinctly different from all the ordinary types of dynamo-electric machines. Further, it does not have any of the characteristics of any of the ordinary galvanometers, and except for the fact that it is used in the detection or measurement of small currents and voltages, it should not be called a galvanometer. In galvanometers, except when used in the measurement of transient currents or quantity of electricity, the moving system is displaced until we have an equality of static couples acting on it. In the vibration galvanometer the equilibrium condition is an equality between integral values of the product of the current and generated voltage and the mechanical power dissipated in various ways as in ordinary electric motors when operated without a load. It therefore behaves more like an electric motor than like a galvanometer. Further, since it is used only with alternating currents and operates in synchronism with the current it must necessarily have some of the characteristics of a synchronous motor. As a motor the efficiency of conversion was found in a particular case to be as high as 97 per cent, while the power required to maintain an easily discernible amplitude of vibration was of the order of 10-11 watts. One of the large turbo-generators would therefore furnish the power necessary to operate a thousand-million-billion such machines.

While the vibration galvanometer will probably never be used in the ordinary way for driving other machines, it is of considerable interest from a theoretical standpoint and possesses

certain characteristics which make it a most valuable instrument for use in a large number of alternating-current measurements.

A number of different forms of galvanometers have been and are being used. One form is very similar to an oscillograph except that in general the period is longer and the oil for damping is omitted. Another form is very similar to the D'Arsonval galvanometers of the marine type except that the coil is narrower and the suspensions tighter. In all cases provision is made for changing the period of the moving system. This is usually done by changing the length or tension of the suspensions. Where a large range in frequency is desired the suspensions are often made bifilar.

Passing an alternating current through the coil causes it to vibrate back and forth past its equilibrium position, hence the name vibration galvanometer. The amplitude of the vibration, which is a measure of the current, is ordinarily determined by observing the broadening of a line image as seen in a small mirror attached to the moving system. As regularly used in the detection or measurement of current or voltage the natural or free frequency of the moving system is made to correspond with the frequency of the alternating current to be detected. As a result the amplitude of the vibration is much larger than it would be under almost any other condition.

The sensitivity is large only to currents of the frequency to which the moving system of the galvanometer is tuned. It is therefore possible to use currents of almost any wave form, even in those null methods in which an exact balance can be obtained only with a current free from all higher harmonics. In all such measurements we may make our observations and calculations just as if we were using current having a sine wave form, since the galvanometer responds only very feebly to the 3rd, 5th and higher harmonic components. It is this characteristic combined with its extremely high sensitivity at low frequencies which makes the vibration galvanometer a most valuable instrument in various kinds of alternating-current measurements.

Before we can make much progress in the design of an instrument or machine it is necessary that we know definitely the relation between its various constants. In some cases this knowledge is necessary before we can even use a well designed and constructed instrument or machine to its best advantage. We shall therefore show the relation which exists between the amplitude of the vibration and the impressed voltage in terms of the

intrinsic constants of the instrument and of the electric circuit in which it may be used. Since here we have a mechanical and an electrical oscillating system so connected that they must necessarily operate at the same frequency, we shall make free use of the analogy existing between such systems. Passing a current through the winding produces a mechanical torque tending to displace the moving system of the galvanometer from its equilibrium position. The torque is proportional to the current and we shall let G be the proportionality factor or the displacement constant. In addition we shall use the following notation, in which vectors are designated by bold-faced type, and make use of the following well-known relations:

or, if we let X represent the electrical reactance and M represent the mechanical reactance,

Here E' is the total voltage available for producing current or the sum of the impressed voltage E and generated voltage