1911 Encyclopædia Britannica/Meter, Electric
METER, ELECTRIC. In the public supply of electric energy for lighting and power it is necessary to provide for the measurement of the electric energy or quantity by devices which are called electric meters. Those in use may be classified in several ways: (i) according to the kind of electric supply they are fitted to measure, e.g. whether continuous current or alternating current, and if the latter, whether monophase or polyphase; (ii) according to whether they record intermittently or continuously; (iii) according to the principle of their action, whether mechanical or electrolytic; (iv) according to the nature of the measurement, whether quantity or energy meters. The last subdivision is fundamental. Meters intended to measure electric energy (which is really the subject of the sale and purchase) are called joule meters, or generally watt-hour meters. Meters intended to measure electric quantity are called coulomb meters and also ampere-hour meters; they are employed for the measurement of public electric supply on the assumption that the electromotive force or pressure is constant. Most of the practical meters in use at the present time may be classified under the following five heads: electrolytic meters, motor meters, clock meters, intermittent registering meters and induction meters.
Electrolytic Meters are exclusively ampere-hour meters, measuring electric quantity directly and electric energy only indirectly, on the assumption that the pressure of the supply is constant. The first electrolytic house meter in connexion with public electric supply was described by St. George Lane-Fox. He was followed by F. J. Sprague and T. A. Edison, the last-named inventor elaborating a type of meter which he employed in connexion with his system of electric lighting in its early days. The Edison electric meter, like those of Sprague and Lane-Fox, was based upon the principle that when an electric current flows through an electrolyte, such as sulphate of copper or sulphate of zinc, the electrodes being plates of copper or zinc, metal is dissolved off one plate (the anode) and deposited on the other plate (the cathode). It consisted of a glass vessel, containing a solution of sulphate of zinc, in which were placed two plates of pure amalgamated zinc. These plates were connected by means of a german-silver shunt, their size and the distance between them being so adjusted that about 11000 part of the current passing through the meter travelled through the electrolytic cell and 9991000 of the current passed through the shunt. Before being placed in the cells the zinc plates were weighed. The shunted voltameter was then inserted in series with the electric supply mains leading to the house or building taking electric energy, and the current which passed dissolved the zinc from one plate and deposited it upon the other, so that after a certain interval of time had elapsed the altered weight of the plates enabled the quantity of electricity to be determined from the known fact that an electric current of one ampere, flowing for one hour, removes 1·2133 grammes of zinc from a solution of sulphate of zinc. Hence the quantity in ampere-hours passing through the electrolytic cell being known and the fraction of the whole quantity taken by the cell being known, the quantity supplied to the house was determined. To prevent temperature from affecting the shunt ratio, Edison joined in series with the electrolytic cell a copper coil the resistance of which increased with a rise of temperature by the same amount that the electrolyte decreased. Owing to the cost and trouble of weighing a large number of zinc plates, this type of meter fell into disuse.
A more modern type of electrolytic meter is that due to C. O. Bastian.[1] The whole current supplied to the house flows through an electrolytic cell consisting of a glass tube containing two platinum electrodes; the electrolyte is dilute sulphuric acid covered with a thin layer of oil to prevent evaporation. As the current flows it decomposes the liquid and liberates oxygen and hydrogen gases, which escape. The quantity of electricity which is passed is estimated by the diminution in the volume of the liquid. A third electrolytic meter of the shunted voltameter type is that of A. Wright. In this meter the electrolyte is a solution of mercurous nitrate which is completely enclosed in a glass tube of a particular form, having a mercury anode and a platinum or, carbon cathode. The current is determined by measuring the volume of the mercury delivered at the cathode. In the Long-Schattner electrolytic meter a solution of sulphate of copper is electrolyzed.
Motor Meters.—Amongst motor meters one well-known, type belonging to the ampere-hour species is that of S. Z. Ferranti, who introduced it in 1883. It consists of an electromagnet within the iron core of which is a flat disk-like cavity containing mercury, the sides of the cavity being stamped with grooves. The thin disk of mercury is therefore traversed perpendicularly by lines of magnetic force when the magnet is excited. The current to be measured is passed through the coils of the electromagnet, then enters the mercury disk at the centre, flows through it radially in all directions, and emerges at the periphery. The mass of mercury is thus set in motion owing to the tendency of a conductor conveying an electric current to move transversely across lines of magnetic force; it becomes in fact the armature of a simple form of dynamo, and rotates with a speed which increases with the strength of the current. The roughness of the surface of the cavity serves to retard it. The rotation of the mercury is detected and measured by means of a small vane of platinum wire immersed in it, the shaft of this vane being connected by an endless screw with a counting mechanism. The core of the electromagnet is worked at a point far below magnetic saturation (see Magnetism); hence the field is nearly proportional to the square of the current, and the resistance offered to the rotating mercury by the friction against the sides of the cavity is nearly proportional to the square of the speed. It follows that the number of the revolutions the mercury makes in a given time is proportional to the quantity of electricity which is passed through the meter. In order to overcome the friction of the counting train, Ferranti ingeniously gave to the core of the electromagnet a certain amount of permanent magnetism. Another well-known motor meter, working on a somewhat similar principle, is that of Chamberlain and Hookham. In its improved form this meter consists of a single horseshoe permanent magnet formed of tungsten-steel having a strong and constant field. Two air-gaps are made in this field parallel to each other. In one of these a copper disk, called the brake disk, revolves, and in the other a copper armature disk. The latter is slit radially, and the magnetic field is so arranged that it perforates each half of the disk in opposite directions. The armature is immersed in a shallow vessel filled with mercury, which is insulated from the vessel and the armature, except at the ends of the copper strips. The current to be measured passes transversely across the disk and causes it to revolve in the magnetic field; at the same time the copper brake, geared on the same shaft, revolves in the field and has local or eddy currents produced in it which retard its action. The principle, of the meter is to make the breaking and driving action so strong that the friction of the train becomes immaterial in comparison. This meter is an ampere-hour meter and applicable only to continuous current circuits. Another form of motor meter which is much used is that of Elihu Thomson. It takes the form of a small dynamo having an armature and field magnets without any iron core. The armature carries on its shaft a commutator made of silver slips, and the current is fed into the armature by means of brushes of silver wire. The current to be measured passes through the fixed field-coils, whilst through the armature passes a shunt current obtained by connecting the brushes across the supply mains through a constant resistance. The driving force is balanced against a retarding force produced by the rotation of a copper disk fixed on the armature shaft, which rotates between the poles of a permanent magnet. Induced or eddy currents are thus created in the copper disk, and the reaction of these against the magnetic field offers a resistance to the rotation of the disk. Hence when a current is passed through the meter, the armature rotates and increases its speed until the driving force is balanced against the retarding force due to the eddy currents in the copper brake disk. In these circumstances the number of rotations made by the armature in a given time is proportional to the product of the strength of the current flowing through the armature and that flowing through the field-coils, the former being the current to be measured. Hence the meter is a watt-hour meter and measures electric energy. In order to overcome the friction of the train the field-coils are wound with an auxiliary shunt coil which supplies a driving force sufficient to overcome the friction of the counting train. This last is geared to the shaft of the armature by an endless screw, and the number of revolutions of the armature is reckoned by the counting-dials, which are so arranged as to indicate the consumption in Board-of-Trade units (1 Board-of-Trade unit＝1000 watt-hours). A modification of the above meter with some mechanical improvements has been devised by S. Evershed.[2]
Clock Meters.—Among clock meters the best known is that of H. Aron, which is based upon a principle described by W. E. Ayrton and J. Perry in 1882. It can be constructed to be either an ampere-hour meter or a watt-hour meter, but is usually the latter. Its principle is as follows: Suppose there are two pendulum clocks, one having an ordinary pendulum and the other having a pendulum consisting of a fine coil of wire through which a current is passed, proportional to the potential difference of the supply mains—in other, words, a shunt current. Below this pendulum let there be placed another coil through which passes the current to be measured; then when currents pass through these coils the pendulum of the second clock will be either accelerated or retarded relatively to the other clock, since the action of gravity is supplemented by that of an electric attraction or repulsion between the coils. Hence the second clock will gain or lose on the other. The two clock motions may be geared to a single counting mechanism which records the difference in the rates of going of the two clocks. If the difference of the number, of oscillations made by the two pendulums in a given time is small compared to the number made by either of them separately, then it is easy to show that the power given to the circuit is measured by the gain or loss of one clock over the other in a given time, and can therefore be indicated on a counting mechanism or registering dials. By the use of a permanent magnet instead of a shunt coil as the; bob of one pendulum, the meter can be made up as an ampere-hour meter. In this form it has the advantage that it can be used for either continuous or alternating currents.
In Intermittent Registering Meters some form of ampere-meter or watt-meter registers the current or power, passing into the house; and a clock motion electrically driven is made to take readings of the ampere-meter or watt-meter at definite intervals—say, every five minutes—and to add up these readings upon a set of registered dials. The arrangement therefore integrates the ampere-hours or watt-hours. These meters, of which one well-known form is that of Johnson and Phillips, have the disadvantage of being unsuited for the measurement of electric supply in those cases in which it is irregular or intermittent—as in a theatre or hotel.
Induction Meters are applicable only in the case of alternating current supply. One of the most widely used forms is the Westinghouse-Shallenberger. It consists of a disk of aluminium, the axis of which is geared to a counting mechanism and which runs between the poles of permanent magnets that create eddy currents in it and therefore exert a retarding force. In proximity to the upper side of the disk is placed a coil of wire having an iron core, which is a shunt coil, the ends of the coil being connected to the terminals of the supply mains. Under the disk are two other coils which are placed in series with the supply. When these last coils are traversed by an alternating current they induce local or eddy currents in the disk. The current in the shunt coil lags 90 degrees behind the impressed electromotive force of the circuit to be measured; hence if the main current is in step with the potential difference of the terminals of the supply mains, which is the case when the supply is given wholly to electric lamps, then the field due to the main coil differs from that due to the shunt coil by 90 degrees. Since the eddy currents induced in the disk are 90 degrees in phase behind the inducing field, the eddy currents produced by the main coil are in step with the magnetic field due to the shunt coil, and hence the disk is driven round by the revolution due to the action of the shunt coil upon the induced Currents in the disk. Hence the disk will be accelerated until the driving force is balanced by the retarding force due to the induced currents created in the disk by the permanent magnets. When this is the case, the number of revolutions of the meter in a given time is a measure of the watt-hours or energy which is passed through the meter. The counting mechanism and dials may be so, arranged as to indicate this energy directly in watt-hours. The meter is made up also in a form suitable for use with two or three fixed electric currents. (See Electrokinetics.)
Requirements of a good House Meter.—A gas meter which has an error of more than 2% in favour of the seller or 3% in favour of the customer is not passed for use. An electricity meter should therefore have approximately the same accuracy. As a matter of fact, it is difficult to rely upon most electric meters to register correctly to less than 4% even between quarter-load and full load. Out of nearly 700 current motor meters of various makes tested at Munich in 1902, only 319 had an error of less than 4%, whilst 259 had errors varying from 412% to 10%. If possible, however, the departures from absolute accuracy should not be more than 2% at quarter-load, nor more than 3% at a full load. The accuracy of a meter is tested by drawing calibration curves showing the percentage departure from absolute accuracy in its reading for various decimal fractions of full load. Such a test, is made by determining with an accurate ammeter or watt-meter the current or power supplied to a circuit for a period measured by a good clock and comparing with this the actual reading of the meter during the same time. A common source of trouble is the short circuiting of the shunt coils owing to the shellaced cotton covering of the wire becoming moist.
A good meter should start with a current which is not more than 2% of its full load current. With a supply pressure of 200 volts a 5 c.p. carbon filament lamp takes only 0·1 ampere; hence unless a meter will begin to register with 110 ampere it will fail to record the current consumed by a single small incandescent lamp. In a large supply system such failure would mean a serious loss of revenue. The resistance of the meter coils causes a fall in voltage down the series coil which reduces the supply pressure to the consumer. On the other hand the resistance of the shunt coil absorbs energy which generally varies from 1 to 3 watts and is a loss either to the consumer or to the supply company, according to the manner in which the shunt coil is connected. In those meters which are compounded—that is, have a shunt coil wound on the field magnets to compensate for the friction of the train—it is important to notice whether the meter will operate or continue operating when there is no current in the series coil, since a meter which “runs on the shunt” runs up a debt against the consumer for which it gives no corresponding advantage.
Generally speaking, the price of the meter is a subordinate consideration. Since the revenue-earning power of a supply station depends entirely upon its meters, inaccuracy in meter record is a serious matter. The cost of measuring current by the aid of a meter is made up of three parts: (1) the prime cost of the meter, which varies from £2 to £6 for an ordinary 25-light house electric meter; (2) the capital value of the energy absorbed in it, which if the cost of the energy is taken at 2d. per Board-of-Trade unit, with interest and depreciation at 6%, may amount to £10 per customer; and (3) the annual working costs for repairs and also the wages of the staff of meter men, who take the required monthly or quarterly readings. In the case of small and irregular consumers, such as the inhabitants of model dwellings and flats inhabited chiefly by working-class tenants, coin-in-the-slot meters are much employed. The customer cannot obtain current for electric lighting until he has placed in a slit a certain coin–say, a shilling—entitling him to a certain number of Board-of-Trade units—say, to 2 or 4, as the case may be. In the Long-Schattner electrolytic meter, the insertion of the coin depresses a copper plate or plates into an electrolytic cell containing a solution of sulphate of copper; the passage of the current dissolves the copper off one of the plates, the loss in weight being determined by the quantity of the electricity passed. As soon as the plate has lost a certain amount of weight corresponding to the value of the electric energy represented by the coin, the plate rises out of the liquid and cuts off the current.
Authorities. H. G. Solomon, Electricity Meters (London, 1906); C. H. W. Gerhardi, Electricity Meters: their Construction and Management (London, 1906); L. C. Reed, American Meter Practice (New York, 1904); J. A. Fleming, A Handbook for the Electrical Laboratory and Testing Room (London, 1904); T. P. Wilmshurst, “Electricity Meters,” Electrician (1897), 39, 409; G. W. D. Ricks, “On the Variation of the Constants of electricity Supply Meters, with Temperature and Current,” Electrician (1897), 39, 573. (J. A. F.)