High Voltage Side Research Articles

Resistor Divider with Dividing Element on High Votage Side for Impulse Voltage Measurements

The response characteristics of a resistor divider with dividing element on high voltage side were investigated theoretically and experimentally. As the results, it was revealed that it could be realized a divider having an excellent response characteristic by this method.

Crossing High-Voltage Inerfaces with Large Bandwidth Signals

In connection with a 50-kV modulator development, three methods of crossing the high-voltage interfaces at a 1-MHz bandwidth are being developed. The first system uses a pulse transformer driven from ground by solid-state electronics. The transformer requires special design to obtain a very short (≈ 500 ns) risetime combined with the ability to maintain a 1000-μs, 200-V pulse with negligible droop. The second system is an amplitude-modulated rf carrier operating at 5 MHz. The interface is crossed with an untuned toroidal rf transformer. Increased bandwidth is obtained by placing the carrier at the lower edge of the passband and operating in the single sideband full carrier mode. The signal is demodulated with a balanced diode detector having a peak output of +200 V and good linearity. The third method uses a Hewlett-Packard HPA-4309 electro-optical isolator with solid-state amplifier on the high-voltage side. The intrinsic bandwidth of the isolator unit is ~10 MHz. With suitable amplifiers to bring the pulse up to +200-V peak, the bandwidth is ~2 MHz. Because the interface is crossed at a very low level reaching an adequate signal-to-noise ratio has presented special problems. All three systems should be useful on a wide variety of high-voltage laboratory machines; each has special advantages depending on the particular application.

Transient behavior of an Aluminum pot-line

OVER A PERIOD of several years a number of transformers and induction regulators supplying energy to aluminum pot-lines and similar installations have been damaged by insulation failures, apparently resulting from abnormal overvoltages. The transformers used for these applications generally step down from 13.8 kv to a secondary voltage in the range 600 to 900 volts. A basic impulse level of at least 40 kv is normal for the secondary winding. The fact that insulation failures occur on these windings is an indication of the severity of the voltage transients, and clearly shows that they are out of the range of normal switching transients. However, in many cases it is difficult to see how such voltage surges could invade the low-voltage system from the high-voltage side of the power transformers. This points to some other means of surge generation in the low-voltage circuit, and a likely cause is current suppression.

Surge transfer through 3-phase transformers

ALTHOUGH a low-voltage circuit, connected to the system by a transformer, has no direct exposure to lightning, lightning surge voltages can be produced in this circuit as a result of surge voltage impact on the high-voltage side of the transformer. Even though a lightning arrester is connected on the high-voltage side of the transformer, surge voltages transferred to the low-voltage side can be dangerous especially to lower BIL equipments such as rotating machines and dry-type transformers.

Surge Transfer Through 3-Phase Transformers

ALTHOUGH a low-voltage circuit, connected to the system by a transformer, has no direct exposure to lightning, lightning surge voltages can be produced in this circuit as a result of surge voltage impact on the high-voltage side of the transformer. Even though a lightning arrester is connected on the high-voltage side of the transformer, surge voltages transferred to the low-voltage side can be dangerous especially to lower BIL equipments such as rotating machines and dry-type transformers.

A graphite tube resistance furnace and voltage regulator for equilibrium studies in the temperature range 1500-1800 C

A description is given of a simple graphite tube resistance furnace and control gear, which has been found suitable for equilibrium studies at temperatures of 1500 to 1800? C. The equilibrium measurements were made in a 1 in. diameter alumina tube, heated by a surrounding graphite tube. The furnace operated at 3-4 V, the power required at temperature being 2?-3 kW. A control unit, which maintained the high-voltage side of a Variac transformer at 200 ? 0.1 V, made it possible to maintain the temperature within ?3? C when half-hourly adjustments of the Variac were made to compensate for changes in the ambient temperature.

Improvements in transformer-loss compensators

TRANSFORMER-LOSS COMPENSATORS serve for the measurement of energy and demand as of the high-voltage side of power transformer banks with metering equipment located entirely on the low-voltage side. Advantages of their use are: generally lower costs; <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> savings in space; less exposure to lightning and other disturbances; and the conservation of critical materials. The basic compensator <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> comprises copper-loss and iron-loss elements that cause the meter to include these losses in its registration.

Automatic switches to protect transformers

AN AUTOMATIC grounding switch in combination with an automatic air-break switch can be used in place of a circuit breaker in the 110-kv supply to a transformer. Since internal faults in transformers do not occur frequently, these high-voltage circuit breakers can be put to better use at terminals of transmission lines where faults are caused by lightning. The substitute for a circuit breaker on the high-voltage side of a transformer must provide for interrupting load, magnetizing, and fault currents. The low-voltage circuit breakers are necessary in all except very small single-bank stations for all three interrupting duties. An air-break switch on the high-voltage side of a transformer should be able to interrupt the magnetizing current or to break parallel in a station having two or three banks of transformers. The remaining duty is to interrupt current flowing from the 110-kv network into a faulted transformer. This can be done indirectly by putting a fault on the 110-kv bus which will operate circuit breakers at distant ends of the 110-kv transmission line. This is a drastic method of avoiding interruption of 110-kv fault current at the station flowing to transformer with an internal fault; but as this kind of fault occurs very seldom, such a general interruption can be endured occasionally. When two or more transmission lines are terminated at such stations, high-voltage circuit breakers are required at terminals of lines only, because the automatic air-break switch can be used to disconnect a faulted transformer after all sources of fault current have been removed.

Metering with transformer-loss compensators

The transformer-loss compensator is proposed as an economical answer to the problem of metering on the low-voltage side of the transformer, where supplying and billing for energy and demand must be done at distribution or transmission voltages. Use of the compensator, an accessory to the regular watt-hour meter, necessitates no changes in the meter, and its accuracy has equaled that of meters on the high-voltage side.

Metering With Transformer-Loss Compensators

Outstanding economies may be effected by metering high-voltage loads on the low-voltage side of the power transformers, and compensating the metering to include the transformer losses. The transformer-loss compensator is an accessory to the regular watt-hour meter on the low-voltage side and causes the meter to include the transformer losses in its registration. No changes are made in the meter itself, either structurally or in adjustments, it retains its integrity as a standard measuring instrument. The method is intended primarily for use on customers' installations where the contract provides for supplying and billing for energy and demand at distribution or transmission voltages. Based on established principles, the accuracy performance is equal to that of metering on the high-voltage side. Testing techniques have been developed and co-ordinated with meter testing practice, so that its routine application and maintenance can be done by metermen. This paper is the result of experience with compensating meters since 1932, and the more recent application of transformer-loss compensators. It is intended for electrical utility executives and engineers considering the possible advantages of the method for their systems and for guidance in co-ordinating the application of transformer-loss compensators with metering practice.

Compensating metering in theory and practice

Outstanding economies can be effected by metering high voltage loads on the low voltage side of power transformers, a practical method of which is described in this article. A compensating meter is used in combination with a watthour meter, both connected to instrument transformers on the low voltage side; accuracy is equal to, and sometimes better than, metering on the high voltage side. Calculations illustrating the practical application of the method, and performance data for an actual installation are given.

Compensating metering in theory and practice

OUTSTANDING economies can be effected by metering high voltage loads on the low voltage side of the power transformers. A compensating meter in combination with a watthour meter, both connected to instrument transformers on the low voltage side, provides accuracy equal to, and in some cases better than, metering on the high voltage side. The method permits the use of low voltage current transformers of relatively greater thermal and mechanical limits and of improved accuracy, in comparison with high voltage current transformers of relatively smaller current ratings.

115,000-Kw. Turbo-Alternator

Due to the increasing use of large single-shaft steam turbine units, it has been necessary to design very large generators operating at 1,800 r.p.m. The present paper describes a 115,000-kw. unit recently placed in operation. The generator is wound for 18,000 volts and the whole output of the machine, with the exception of a small part used for the operation of auxiliaries, is stepped up to 182 kv. and all switching is done on the high-voltage side of the transformers. It was possible, therefore, to design the generator for a voltage that would permit a simple two-conductor per slot winding. On account of the great axial length of the machine, ventilation required careful study and a new scheme of rotor ventilation was developed to secure uniform cooling throughout the length of the long rotor body. The stator ventilation is of the inward-outward parallel flow type with fourteen parallel paths in each half of the stator. Air circulation is provided by four single inlet motor-driven blowers mounted under the generator. A fin type radiator cooler is mounted under the generator yoke and condensate is circulated through the cooler. Excitation is furnished from a 350-kw., 250-volt coupled, shunt-wound exciter and a 7½-kw. overhung pilot exciter is used to excite the fields of the main exciter.

Diffraction of Low-Speed Electrons by Single Crystals of Copper and Silver

Diffraction beams for normal incidence on the (100) faces of copper and silver crystals.---Previous results for a copper crystal are checked and extended, by using another crystal of exceptionally pure copper. All of the expected diffraction beams which are x-ray analogues in the two principal azimuths and in the range below 325 volts are found. These beams require values of refractive index greater than unity, but the associated values of inner potential are not constant. Another class of weak beams is found characteristic of the copper lattice, but which require a refractive index of approximately unity for the first order beams of the two principal azimuths. These weak beams accompany the main beams on the high voltage side as satellites. Conditions for orders higher than the 1st are more complicated and indicate that the above method of classification is not sufficient. For a silver crystal the number of experimental maxima is considerably greater than the number of theoretical beams in the low-voltage range. These maxima cannot be classified as main beams requiring a refractive index greater than unity, and weak satellites requiring unit refractive index as in case of a copper crystal, but whenever two or more experimental maxima are to be associated with a given theoretical beam, they are, in general, grouped as components of fine structure of a single diffraction beam.Intensity measurements on diffraction beams as a function of angle of incidence.---For copper, as the angle of incidence is changed by only a few degrees from normal, large intensity changes of the various diffraction beams are observed. Some beams increase while others decrease in intensity for a given change in angle of incidence. Primary voltage and collector angle are adjusted for each observation. For silver, as the angle of incidence is changed from normal the relative intensities of the components associated with a particular beam change rapidly. Some components may disappear and others appear at different voltages, as the incident angle is changed a few degrees. The plane grating formula is approximately satisfied over considerable ranges in the angle of incidence, but there are other ranges for which it is not satisfied. The above variations for various beams do not correspond; neither do those for the beams associated with different orders of reflection from the same set of Bragg planes. Intensity measurements are given for most of the beams in the two principal azimuths below 350 volts for ranges in the angle of incidence, including normal incidence, through which the beams are observable.Regular reflection of electrons from the (100) set of planes.---For copper, with angle of incidence equal to angle of reflection, curves obtained by measuring collector current as a function of primary voltage show many irregularities. For silver, the curves contain many irregularities which do not correspond to those for a copper crystal under similar conditions. For the same angle of incidence and the same set of reflecting planes, the results are not the same for the plane of reflection in the (100) and the (111) azimuths, respectively.Diffraction beams due to a surface gas lattice on the (100) faces of copper and silver crystals.---For copper, "additional" beams which were formerly reported have since been found to decrease in intensity after prolonged heat treatment of the copper crystal at temperatures near its melting point. These beams indicate a simple-cubic, single-spaced lattice with the same constant as that for the copper lattice. They require values of refractive index greater than unity. When the surface gas lattice becomes thin it changes in structure from a simple-cubic, single-spaced lattice to a face-centered, double-spaced lattice. The beams characteristic of the latter structure require unit refractive index. Hydrogen also forms in either of the above two lattices on a copper crystal, depending on the pressure of the hydrogen while the lattice is formed. The lattice which is characteristic of the thicker layer is very unstable and soon changes to that of the thinner layer after the pressure is removed. For silver, only a few very weak beams due to a surface gas lattice are found below 50 volts after heating the crystal a short time at red heat. They indicate a double-spaced, face-centered structure.Inner potential. The measured value of the inner potential is neither constant nor a continuous function of the voltage. It is found to change discontinuously, by several volts in some cases, as the angle of incidence is varied. Results for the crystals of copper and silver which have the same lattice structure are found to differ widely. Hence they appear to be a function of the type of atom.Surface action. Experimental methods of distinguishing between effects due to surface action and those due to a space lattice are discussed.

15,000-Kw. turbo-alternator

DUE TO the increasing use of large single-shaft steam turbine units, it has been necessary to design very large generators operating at 1,800 r.p.m. The present paper describes a 115,000-kw. unit recently placed in operation. The generator is wound for 18,000 volts and the whole output of the machine, with the exception of a small part used for the operation of auxiliaries, is stepped up to 132 kv. All switching is done on the high voltage side of the transformers. Therefore it was possible to design the generator for a voltage that would permit a simple two-conductor per slot winding.

Over-voltage on transmission systems due to dropping of load

When a water wheel-driven generator, or a hydroelectric station, loses all or a large part of its load, the voltage rises more rapidly and to a greater extent than has been generally realized. The purpose of this paper is to show the manner in which the voltage will rise under different conditions, and a method of calculating the voltage rise of a water wheel-driven generator when load is lost. The results of a large number of tests are given showing the rise in voltage on an 8500-kv-a., 6600-volt water wheel generator when load was tripped under several different conditions at three different places, namely; at the generator, on the high-voltage side of a transformer bank, and at a substation forty miles away. As it is not considered good practise to subject a transmission system to high over-voltages, consideration is given to methods of checking the voltage rise by use of relays as soon as possible after load has been dropped.

Over-Voltage on Transmission Systems Due to Dropping of Load

When a waterwheel-driven generator, or a hydroelectric station, loses all or a large part of its load, the voltage rises more rapidly and to a greater extent than has been generally realized. The purpose of this paper is to show the manner in which the voltage will rise under different conditions, and a method of calculating the voltage rise of a waterwheel-driven generator when load is lost. The results of a large number of tests are given showing the rise in voltage on an 8500-kv-a., 6600-volt waterwheel generator when load was tripped under several different conditions at three different places, namely, at the generator, on the high-voltage side of a transformer bank, and at a substation forty miles away. As it is not considered good practise to subject a transmission system to high over-voltages, consideration is given to methods of checking the voltage rise by use of relays as soon as possible after load has been dropped.