US3015059A - Stepless compensation of reactive current - Google Patents

Description

Dec. 26, 1961 M. SANGL ETAL 3,015,059

STEPLESS COMPENSATION OF REACTIVE CURRENT Filed June 15, 1959 5 Sheets-Sheet 1 Dec. 26, 1961 M SANGL ETAL 3,015,059

STEPLESS COMPENSATION OF REACTIVE CURRENT Filed June 15, 1959 3 Sheets-Sheet 2 Dec. 26, 1961 M. SANGL ETAL STEPLESS COMPENSATION OF REACTIVE CURRENT Filed June 15, 1959 3 Sheets-Sheet 3 Patented Dec. 26, 1961 3,015,059 STEPLESS COMPENSATION OF REACTIVE CURRENT Michael Sang] and Werner Elischer, Erlangen, Germany, assignors to P. Gossen & Co. G.m.b.I-I., Erlangen, Bavaria, Germany Filed June 15, 1959, Ser. No. 820,527

Claims priority, application Germany Sept. 6, 1958 2 Claims. (Cl. 323-101) The invention pertains to a device for the stepless and, particularly, automatic compensation of a continuously changing inductive or capacitive reactive current by a controllable reactive load.

Inductive or capacitive reactive currents, varying comparatively rapidly and continuously in magnitude, occur in various electric installations owing to the specific characteristics of the operation involved. This is the case, for example, in an induction melting furnace, where the material to be melted is inside a winding carrying alterhating current. The inductive reactance of this winding is subject to considerable fluctuation in the melting of ferromagnetic, paramagnetic, and diamagnctic substances because of the changes of the ohmic resistance and the magnetic conditions within the material being melted. To secure as high a possible power consumption-by the melting furnace or to relieve the load on the input leads and the generator, the continuously varying inductive current of the furnace must be continuously compensated by capacitances in parallel. Up to the present time the effective parallel capacitance has had to be adapted to the condition of the melt by connecting or disconnecting capacitances by means of contactors, by hand, or by automatic switchgear.

It is obvious that such switchgear is of complicated design and subject to trouble as the result of wear of the moving parts. This requires continuous or at least regular maintenance of such installations. Furthermore, the contactors and capacitances must meet particularly high requirements, as switching is done with full voltage across the switch contacts. This entails a considerable expenditure for the various capacitances and contactors. Moreover, the induction furnace reactive current can be compensated only in steps, that is not exactly enough.

The reactive current compensation of the invention, on the other hand, consists entirely of stationary and sturdy parts that are not subject to any wear. The expensive contactors that are othervn'se required are eliminated completely, and the entire capacitance, which heretofore has been subdivided into several units, may be combined into a fixed unit. Furthermore the arrangement of the invention makes it possible to compensate the reactive current continuously and in such a way that there is a small capacitive component in the line.

This is achieved by connecting a fixed capacitance in parallel with the induction melting furnace, which is able to compensate the maximum inductive reactive current of the furnace, and, secondly, by usinga special regulating reactor as a sort of current-regulating magnetic amplifier. This regulating reactor, controllable by direct current, automatically and continuously compensates the resultant capacitive reactive current in combination with a circuit for measuring the reactive current I -sin Other advantages and objects of the invention will be made apparentby the following description and the accompanying drawing; wherein, as an example, the invention is applied to an induction melting furnace, as mentioned above. 7

FIG. 1 is a basic circuit diagram of the invention.

FIG. 2 illustrates the construction of the transductor.

FIGS. 3a, 3b and 4 are diagrams explanatory of the operation of the invention.

FIG. 5 is a diagram of an automatic compensating ar rangement according to the invention.

According to the basic circuit diagram of FIG. 1 the controllable inductance is embodied in the load winding 1, connected in parallel with the furnace winding 2. In parallel with the latter there is the capacitance 3. These circuit elements are connected to the A.C. source 5-, a medium-frequency generator for, say, 10 kilocycles, via the lines 4 and 4 The capacitance 3 is dimensioned so as to compensate the maximum inductive current of the furnace.

The mechanical construction of the transductor consisting of the two parts T and T is shown in FIG. 2. It consists of the two adjacent iron cores 6 and 7, which are designed as continuous-strip toroidal cores, the load winding 8 and 9, and the control winding 10, which is wound around the two other legs of cores 6 and 7 for both. The load winding consists of two parts, wound in opposite directions, and connected in parallel. As the same A.C. voltage is impressed on both parts of the load winding, the actions of the induction fluxes flowing in the center legs of the core upon the control winding 10 cancel out.

To explain the mode of operation two idealized magnetization curves for the two transductors T and T respectively, are shown in FIG. 3a. The D.-C. biasing should be chosen so that when there is no A.C. voltage applied, core 6 is saturated negatively at point P and core 7 saturated positively at point P When A.C. voltage is impressed across the load winding, an additional alternating flux is superposed on the constant biasing D.-C. fiux in each transductor.

The cores of the transductors and the load windings 8 and 9 are so dimensioned with respect to the A.C. voltage, that when no bias magnetization is present, they are not saturated by the A.C. voltage.

When polarization (biasing) is present, on the other hand, the magnetization curves are no longer traversed symmetrically with respect to the H-axis, but the A.C. fluxes are shifted in the direction P or P depending upon the nature of the biasing, so that we get the sections of the oscillation curves of the two transductors (magnetic amplifiers), started at the time t shown in FIG. 3b, 11 denoting the A.C. voltage applied, and i and i denoting the currents in windings 8 and 9, respectively.

During the positive half-wave of voltage the current i can only rise to the value of the control current in the load winding of T As the voltage continues to rise, the core of T emerges from the region of saturation, because the saturating action of the control voltage is exact- 1y balanced out in this case. Because of the very high inductance of the load winding along the steep portion of the demagnetization characteristic, the current i remains constant to begin with, until the field has again been displaced so far into the negative region toward the end of the subsequent negative half-wave that the lower break in the curve is reached. When this is so the coreof T is again saturated, and i jumps to a high value, which depends upon the ohmic resistance of the load winding 8, on the one hand, and upon the degree of polarization, on the other, for a given voltage a. If this polarization (biasing) is greater than that shown in FIG. 3a, the lower break in the magnetization characteristic isreached sooner as the field is reversed, and the time during which the core of T is saturated becomes longer, so that we get a larger current-time area. In principle, the same conditions prevail in the right-hand curve shown in FIG. 3b. Because of the direction of the turns in winding 9, which is opposite t'o-that of winding 8, the region is reached during the positive voltage half-wave, so that at the end of the latter current pulses flow in the positive direction. The constant components of the currents i and i which depend upon the degree of biasing and have already been mentioned, flow in opposite directions, so that their effects cancel out in the external circuit.

As the current pulses always flow at the end of the corresponding voltage half-wave, this is equivalent to an inductive phase displacement of nearly 90..

The currents and voltages present in the furnace connected together with the capacitance and the regulating reactor are shown in FIG. 4 for the case where the amplitude of the capacitive reactive current i is some 50% greater than that of the inductive reactive current i Accordingly, a resultant capacitive reactive current will flow in the lines to the generator. As is readily seen in the figure, the current pulse icomp produced by the compensation circuit described coincides in time with the opposite half-wave of the capacitive current. Thus, compensation takes place. To secure complete compensation,

the polarization (biasing) is so chosen that the currenta time area of the compensation pulse, together with that of the half-wave of the inductive reactive current, equals the current-time area of the capacitive half-wave. Hence, the regulating reactor described makes it possible to achieve stepless compensation of the capacitive reactive current by means of a simple and inexpensive D.-C. control.

There is no external resistance connected in series with the load winding of the regulating reactor. Hence the amplitude of the current pulses through it depends only upon its own ohmic resistance for constant polarization. Moreover, since the ends of these impulses always coincide more or less'with the passage of the voltage curve through zero for different values of polarization, the shorter and steeper the pulses are for equal current-time areas, the better the 90phase displacement. Hence the ohmic resistance of the load winding is kept as low as possible.

In FIG. 2 the two legs of cores 6 and 7, which have no windings 8 and 9 on them are wound with a common control winding 10. As a result, as already mentioned briefly in connection with this figure, the effects of the A.-C. fluxesinthe adjacent cores are cancelled out with respect to the control circuit. To be sure, this would also be the case in the usual division into two spatially separated coreswith their own control windings. But the disadvantage of such an arrangement is that a high induced. is produced across each control winding, which is cancelled out externally only as the result of the connection of the two control windings in opposite directions. The resultant high potential acrossthe insulation of the two windings is avoided in the setup of FIG. 2, because here the A.-C. fluxes flowing in the cores are in opposite directions. This already provides a sort of magnetic compensation.

The connections of'the reactive current load of the invention with an l-sin meter and with a magnetic amplifier connected inuseries with the latter to constitute an automatic reactive current compensator is shown in FIG. 5.

The circuit for measuring 1 sin qt consists of the current transformer 11, which is connected on its secondary side with another current transformer 12, the rectifiers 13, 14, 15, and 16, the condenser 17, and the two condensers 1S and 19 connected in series. A rectifier 20 as well as the control windingsof 'a magnetic amplifier 21 are connected to the output end of this'I-sin g measuring circuit, formed by the terminalsofthe'series'circuits 18 and 19. Another rectifier 22 is connected in the line to these control windings; The rectifiers 13-16 constitute a remote-controlled switch, being opened and closed by the voltage across the lines 4 and 4;. Either rectifiers 13 and 14 are opened and rectifiers and 16 closed, or vice versa, .iepending'upon theinstantaueous polarities. of the control voltage. Accordingly, the control voltage opens up two different current paths for the secondary current in transformer 12. But as any change in these current paths also changes the polarity of the secondary current in the current transformer 12, a pulsating DC. current flows through the elements connected to the output ends of the I -sin measuring circuit.

The phase of the control current produced by the voltage across the lines 4 and 4 is displaced through by the capacitor 17 and by the capacitors 18 and 19 that act in parallel in this case. Hence this measuring circuit delivers maximum current to the subsequent elements when the reactive current in lines 4 and 4 is at a maximum, i.e. I -sin =I. The polarity of the pulsating D.-C. current at the output ends of the I-sin measuring circuit just described depends upon whether the reactive current flowing through the lines 4 and 4 is capacitive or inductive. Rectifiers 20 and 22 act so as to allow a control current to flow through the control windings of the transductor (magnetic amplifier) 21 only when this reactive current is capacitive.

The control winding 10 of the regulating reactor of the invention, whose output windings 8 and 9 are connected in parallel with the furnace winding 2, serves as a load for the'magnetic amplifier 21, being connected to the latter through the rectifier bridge 23, which contains a smoothing capacitor, not shown.

The mode of operation of this circuit is readily understood frorn whathas been said above. If the inductive reactive current in the furnace'windings drops during the melting process, a resultant capacitive reactive current flows in lines 4 and 4 This generates a D.-C. voltage at the output of the I -sin measuring circuit, which causes current to flow through the control winding of the magnetic amplifier 21. The control parameter is amplified in the latter, then controlling the current flowing through the load windings 8 and 9 of the regulating reactor in such a way as to nearly completely compensate the capacitive reactive current, establishing a state of equi librium with a small and desired capacitive residual current. The large capacitance in parallel with the furnace winding effectively short-circuits the upper harmonics generated by the pulse-shaped currents through the regulating reactor.

If the input amplifier 21 is provided with an additional control winding, and a D.-C. current that is fixed or is proportional to. the apparent current flows through the line, the reactive current compensator regulates the line current to a fixed reactive current I sin 4) that corresponds to the given direct current in one case, while in the other it regulates the line current to any desired power factor cos 4:. This supplementary winding is required whenever a power factor is wanted on the line that differs from unity.

Small reactive currents can also be controlled by employing a so-called inductance-controlled regulating reactor, in which the working point is shifted back and forth by the biasing control parameter within the steeper and not yet fully saturated regions of the magnetization characteristic curve. This arrangement has the advantage of being particularly free of upper harmonics.

For high frequencies a core material formed of ferrite is advisable instead of the cores 6 and 7, which are preferably made of thin iron strip in the present case.

What is claimed is:

1. Apparatus for continuously compensating for reactive current in an alternating current line connected to an inductive load, comprising circuit means for producing a direct currentwhich is a measure of. the reactive component of the line current, said means including a rectifier switching circuit, an input transformer connected to the input of' said rectifier circuit, a current transformer connecting said alternating current line to the primary winding of the input transformer, acapacitor connecting one side of said line to a midpoint of the secondary of the input transformer, and means for connecting the midpoint of the output of said rectifier circuit to the other side of the alternating current line, capacitor means connected in parallel with said load, said capacitor means having a greater magnitude than that required to compensate for the inductive load, a magnetic amplifier, said magnetic amplifier including two contiguous ring cores, a control winding Wound around both cores, and a load winding connected across said line and said control winding, said load winding comprising two coils connected in parallel and wound on each of said cores, respectively, to produce op- 10 posing magnetic fields in said control Winding, and means for impressing said direct current on said control Winding with a polarity such that the magnetic amplifier is controlled only in response to capacitive reactive current on said line.

2. Apparatus according to claim 1 including a second magnetic amplifier having load windings and a control Winding, a rectifier connecting said last-mentioned control Winding to the output of said rectifier circuit, and a second rectifier connected in series with the load windings of the second magnetic amplifier across the line, the output of said second rectifier being connected to the control winding of the first-mentioned magnetic amplifier.

References Cited in the file of this patent UNITED STATES PATENTS 1,227,302 Osnos May 22, 1917 1,836,886 Thompson Dec. '15, 1931 2,421,786 Haug June 10, 1947 Rhyne May 21, 1957

Images