US3019382A - Control apparatus - Google Patents

Description

Jan. 30, 1962 D. J. ROTIER 3,019,382

CONTROL APPARATUS Filed Dec. 18, 1958 2 SheetsSheet 1 EI6 zgl5 O 40 FIG I INVENTOR. DONALD J. ROTIER ATTORNEY Jan. 30, 1962 D. J. ROTIER 3,019,332

CONTROL APPARATUS Filed Dec. 18, 1958 2 Sheets-Sheet 2 kSO FIG 2 INVENTOR.

DONALD J. ROTIER ATTORNEY United States Patent 3,019,382 CONTROL APPARATUS Donald J. Rotier, Crystal, Minn., assignor to Minneapolis- Honeywell Regulator Company, Minneapolis, Minn, a corporation of Delaware Filed Dec. 18, 1958, Ser. No. 781,345 5 Claims. (Cl. 323-89) This invention relates to magnetic control apparatus and relates more particularly to self-biasing in magnetic amplifiers.

Many arrangements have been devised for biasing, or resetting, the saturable cores in magnetic amplifiers in order to establish the desired operation. In some magnetic amplifiers the bias current and the control current are superimposed in a single Winding; in others, separate bias windings are provided, In any case, it is desirable for the effect of the bias magnetomotive force to be independent of control signal input, magnitude of output signal, and any other variable associated with operation of a magnetic amplifier. One way to establish such independence is to furnish the bias current from a separate power supply. Thus, it is not uncommon in the art to encounter magnnetic amplifier circuits having a separate source of bias current in the form of rectifiers and all the attendant paraphernalia. On the other hand, such cumbersome arrangements may sometimes be avoided by selfbiasing circuitry. It is to the latter type of arrangement that my invention particularly pertains.

My invention contemplates powering bias windings in a magnetic amplifier with signals derived from currents flowing through the devices power windings. While these latter currents vary in magnitude in accordance with control signals applied to the magnetic amplifier, it has been found that, since in some circuits an increase in current through one power winding is accompanied by a corresponding decrease in current through another power winding, it is possible, by providing means capable of combining such oppositely varied power Winding currents, to obtain bias power having the same magnetomotive effect over each half-cycle of supply power regardless of changes in current through the power windings due to control signal changes.

It is therefore an object of my invention to provide a new arrangement whereby a magnetic amplifier may be self-biased.

Another object ofmy invention is to provide a selfbiasing arrangement wherein bias current is derived from current flowing through the power windings of the magnetic amplifier.

These and other objects of my invention will be more thoroughly understood through reference to the following description, claims, and drawings, in which:

FIGURE 1 is a schematic drawing of a four core, fullwave, push-pull magnetic amplifier embodying the invention, and

FIGURE 2 is a pictorial representation of some of the signals appearing during operation of the magnetic amplifier of FIGURE 1.

Structure of FIGURE 1 The magnetic amplifier of FIGURE 1 has four saturable cores 10, 11, 12, and 13. Associated with saturable core are a power winding 14, a bias winding 15, and a control winding 16. Likewise, a power winding 17, a bias winding 20, and a control winding 21 are associated with saturable core 11; a power winding 22, a bias winding 23, and a control winding 24 are associated with saturable core12; and a power winding 25, a bias winding 26, and a control winding 27 are associated with saturable core 13. The saturable cores 10, 11, 12, and 13 are preferably, although not necessarily, made of magnetic material 3,019,382 Patented Jan. 30, 1962 having a substantially rectangular hysteresis characteristic.

Such magnetic materials are well known in the art.

The controlwindings 16, 21, 24, and 27 are connected in series to operate as an entire control winding, indicated generally by the numeral 30, for the magnetic amplifier. Control winding 30 is connected between the control signal input terminals 31 and 32. A conductor 33 connects a first end of power winding 14 to a first end of power winding 25. A bias resistor 34 connects conductor 33 with a power input terminal 35, Another conductor 36 connects a first end of power Winding 17 With a first end of power winding 22, and conductor 36 is also connected through another bias resistor 37 to power input terminal 35. Two rectifiers, or asymmetrical conductive devices, 40 and 41, are connected in series between the second end of power winding 14 and the second end of power winding 17; rectifiers 4t) and 41 are poled so as to present an easy current flow path from winding 14 to winding 17. An output resistor 42 is connected from the junction 47 between rectifiers 40 and 41 to another power input terminal 43. Two more rectifiers, or asym metrical conductive devices, 44 and 45 are connected in series, the series combination being connected from the second end of power winding 25 to the second endof power winding 22. Rectifiers 44 and 45 are so poled as to provide an easy current flow path from winding 25 to winding 22. Another output resistor 46 is connected from the junction 48 between rectifiers 44 and 45 to power terminal 43.

A conductor 50 is connected from conductor 36 to a indicate the winding sense of one winding relative to the others. Current flowing into the dotted ends of the coils has a given magnetic effect on the associated saturable core, and current flowing in the opposite direction has the op osite ma netic effect. I

With winding polarities as shown in FIGURE 1, the circuit operates so that unidirectional current flowing through the power windings 14, 17, 22, and 25 tends to saturate the corresponding core in a given magnetic sense, current flowing through the bias windings. 15, 20, 23, and 26 from junction 51 to junction 53 tends to have the opposite magnetic effect on the magnetic cores, that is, tends to reset the cores, and current flowing through control windings 16, 21, 24, and 27 tends to have one magnetic effect on cores 10 and 11 and the opposite magnetic effect on cores 12 and 13.

Operation of Figure I The amplifier of FIGURE 1 is a full-wave push-pull device. It is arranged to give an alternating current out put signal which depends in magnitude and phase upon the magnitude and polarity of an'input signal. In operation, a source of input signal (not shown) is connected to terminals 31 and 32, an alternating current power source (not shown) is connected to terminals 35 and 43, and an output signal may be taken from between junction points 47 and 48. If desired, a differential load may be connected in place of load resistors 42 and 46, the net difference in currents flowing therein then being the output signal of the magnetic amplifier. For example, onehalf of a motor winding may be connected 'as output resistor 42 and the other half of a motor winding may be connected as resistor 46, the winding halves being connected in opposing relation so that the net magnetomotive force produced by the entire winding depends upon the net difference in currents flowing therethrough.

It should be understood that the saturable cores 1t), 11, 12, and 13 when unsaturated allow practically no current (only a small magnetizing current) to flow through their respective power windings 14, 17, 22, and 25; but that when a core saturates it allows current to flow freely through its associated power winding. Furthermore, with suitable power applied to power terrninals35 and 43 and a zero control signal applied to control winding 30, the cores are biased, or reset, so that they saturate at a point half way through their respective conducting half-cycles. By conducting half cycle is meant a half-cycle of applied supply power having the correct polarity to allow current conduction through the diode in series with the power winding (or associated magnetic core) being referred to. Specifically, those half-cycles of supply power during which power input terminal 35 is more positive than power input terminal 43 are referred to as the conducting half-cycles of power windings 14 and 25. Likewise, those half-cycles or supply power during which terminal 43 is more positive than terminal 35 are referred to as the conducting half-cycles of power windings 17 and 22. In normal operation, then, power windings 14 and 25 are allowed to conduct during one half-cycle and power windings 17 and 22 are allowed to conduct on the following halfcycle. The actual current flow through these windings, of course, is substantial only when the associated cores are saturated. With zero control signal, each winding will allow a substantial current to flow only during the last half of its conducting half-cycle. During a conducting half-cycle for power windings 14 and 25, the current flow may be traced from power terminal 35 through resistor 34, winding 25, diode 45 and output resistor 46 back to terminal 43 and also from resistor 34 through the power winding 14, rectifier 40, and output resistor 42 back to power terminal 43. The current flowing through output resistor 42, in this case, equals the current flowing through the other output resistor 46, and the sum of the currents flowing through output resistors 42 and 46 flows through bias resistor 34. During the next half cycle of supply power, current paths may be traced from power terminal 43 through output resistor 46, rectifier 44, power winding 22, and resistor 37 back to power terminal 35 and from power terminal 43 through output resistor 42 through rectifier 41, power winding 17, and resistor 37 to power terminal 35. Thus, it is seen that during this half-cycle current is flowing through output resistors 42 and 46 in the opposite direction of that during the previous half-cycle, and that bias resistor 37 has flowing through it the sum of the currents through output resistors 42 and 46. With a zero control signal, of course, current flowing through one of output resistors 42 and 46 is, exactly matched by the current flowing through the other; and the difierential output, that is, the voltage difference between junction points 47 and 48 or the difference in currents flowing through output resistors 42 and 46, is Zero.

To help visualize the operation of a magnetic amplifier as so far explained, attention is directed to the curves of FIGURE 2. Curve 56 depicts the voltage from terminal 43 to terminal 35, to which terminals a source of supply power (not shown) is connected. Curve 57,- drawn to the same time base ascurve 56, represents the current that flows through output resistor 42 and also represents the current that flows through output resistor 46, these currents being equal when the control signal is-zero, It willbe noted that current flows through each outputresistor during the latter half of each power supplyhalf-cycle. As to output resistor 42, positive pulses of current flow through power winding 14 and negative pulses of current flow through power windings 17, as these current pulses are depicted in wave form 57. As

to output resistor 46, positive pulses of current flow through power winding and negative pulses of current flow through power winding 22, as these current pulses are depicted by wave form 57.

It should be noted that, while output resistors 42 and 46 each have currents flowing through them in both direc tions, bias resistor 34 conducts current only in the direction away from power terminal 35 and bias resistor 37 conducts current only in the direction towards power terminal 35. With a zero control signal applied to the magnetic amplifier, then, current flowing through resistor 34 is similar in waveshape to, although twice the magnitude of, the positive half-cycles of curve 57 and is shown as curve 60 of FIGURE 2; the current flowing through resistor 37 has a waveshape similar to, although twice the magnitude of, the negative half-cycles of curve 57 and is shown in FIGURE 2 as curve 61. When current flows through resistor 34 no current flows through resistor 37, and vice versa. Further, the direction of current flow through resistors 34 and 37 is such that the net difference in voltage resulting therefrom, which is the voltage appearing between conductors 33 and 36, is the same polarity during every half-cycle of supply power. Thus, the voltage developed between conductors 33 and 36 is represented in FIGURE 2 by curve 62, the excursions of curve 62 referring to the voltage of conductor 36 taken with respect to conductor 33. It is 7 this voltage that powers the bias windings 15, 20, 23,

and 26. The voltage difference between conductors 33 and 36 is applied to junctions 53 and 51, respectively, through conductors 52 and 50. Bias current is thereby forced from junction 51 through resistor 54, bias winding 15, and bias winding 20 back to junction 53 and is also forced through resistor 55, bias winding 26, and bias winding 23 back to junction 53. Resistors 54 and 55 acts as means for controlling the magnitude of current flowing through the bias windings and for balancing the current flowing through windings 15 and 20 with the current flowing through windings 26 and 23. Resistors 54 and 55 are shown for the sake of completeness only, since they may not be necessary in many devices embodying my invention; it is suflicient that the bias winding circuit have the correct impedance to allow the desired amount of bias current to flow therethrough.

It is clear from the above explanation that, with zerocontrol signal, the bias signals applied to the bias Winding circuit at junctions 51 and 53 resets each core between its conducting half-cycles by the amount required for proper operation of the circuit, that is, so that the cores saturate half way through their respective conducting half-cycles. It will now be shown that the same effect is produced when control current is flowing.

As will be recalled, a current flowing through control winding 30 has one magnetic eflect upon cores 10 and 11 and the opposite magnetic effect upon cores 12 and 13. As a result, a control current of given direction will cause, for example, cores 10 and 11 to be reset less and cores 12 and 13 to be reset more than was the case in the zero control signal example above. Thus, wind' ings 14 and 17 conduct over a larger portion of their respective conducting half-cycles and windings 22 and 25 conduct over a smaller portion of their conducting half-cycles, the increase in conducting time of windings 14 and 17 being equal to the decrease in conducting time of windings 22 and 25. In this condition, the current flowing through output resistor 42 has a wave shape that is represented by curve 63 of FIGURE 2. Curve 64 represents the current wave; form of output resistor 46. The positive portions of curves 63 and 64 repre-' sent that portion of output current flowing through bias resistor 34; these positive portions, when added together, result in a curve 66, which represents the wave shape of the current flowing through resistor 34. Likewise, the negative portions of curves 63 and 64, when added together, result in a curve 67, which represents the cur rent flowing through bias resistor 37. Due to the direction of flow of the currents represented by curves 66 and 67, the resultant voltage difference between conductors 33 and 36 has a waveshape as represented by curve 70. Since this is the signal that powers the bias windings, it must now be shown that the signal represented by curve 70 has the same magnetic effect upon the saturable cores as does the signal represented by curve 62. This requirement is necessary in order to avoid signal feedback through the bias circuit of this magnetic amplifier.

As is well known in the art, a change in magnetic state of saturable cores is dependent upon both the magnitude of signal applied to windings associated therewith and the time over which such a signal is supplied. More precisely, the change in magnetic state of a core is substantially equal to the time integral of the voltage applied to its winding. This applies, of course, only to changes over the unsaturated region of operation. Since this is the case for resetting operation in this magnetic amplifier, the relationship is valid as here applied.

Since, as mentioned above, the amount of resetting taking place depends upon the time integral of the ap plied signal, it is seen that the area under curve 62 must equal the area under curve 70 in order that the bias circuit have the same resetting effect in both cases. A comparison of these areas for a single half-cycle is sufficient, since the curves are repetitive from one halfcycle to the next. The area under the curve 62 during the first half-cycle is proportional, as will be recalled, to twice the area under curve 57 over the first halfcycle. Taking the peak value of curve 57 equal to one arbitrary unit,

where 0 equals the angle in radians of the applied power signal.

The area over the first half-cycle of curve '70, on the other hand, equals the sum of areas under curves 63 and 64 over the first half-cycle. Thus where z is the angle in radians at which curve 63 first attains a non-zero value and which is the point at which the respective core saturates.

The areas in question are therefore equal, as desired. Inasmuch as the above showing represents the general case, it is clear that the signals applied to the bias windings in this magnetic amplifier are independent of the control signal, and therefore also independent of the output signal, over the entire range of operation.

The mode of operation described above may be referred to as class A, for the current flowing through each power winding with a zero control signal is one-half the maximum current that can flow through any of the power windings with a suitable control signal applied. The current that flows with a Zero control signal, usually called the quiescent current, may cause difficulties where the load connected to the magnetic amplifier is a motor winding. For example, a two-phase motor having a center tapped control phase winding may be connected to the magnetic amplifiers so the one half of the control phase winding replaces load resistor 42 in FIGURE 1 and the other half of the winding replaces load resister 46. The MMF produced by the entire motor control phase winding depends upon the net difference in the currents flowing through its two halves. The phase of the MMF depends upon the phase of the input signed to the magnetic amplifier. All of this is just as it should be; however, it should be noted that the total current flowing through the motor control phase winding is just as great when the magnetic amplifier is in the quiescent state as when it is producing maximum output. This, of course, may cause overheating of the motor and, in addition, results in lowefficiency operation.

To overcome the undesirable effects of high quiescent current in a motor load as explained before, the self bias circuit of the magnetic amplifier may be proportioned to produce operation approaching class B. The quiescent current is thereby reduced to a relatively small value. While operating a magnetic amplifier of the sort described herein in other than a class A mode introduces signal feedback, a useful output may still be obtained with a motor load. Hence, a designer may employ my invention in a variety of applications by taking into account the permissible quiescent current, permissible signal feedback, and other controlling factors of a particular use.

Since many variations and modifications of this invention will undoubtedly occur to those skilled in the art, I wish it to be understood that I intend to be limited by the scope of the appended claims and not by the specific embodiment of my invention which is disclosed here for the purpose of illustration.

I claim:

1. A self-biased magnetic amplifier comprising: first and second pairs of power windings, each of said windings being discretely associated with a saturable core; first conductive means connecting one end of a power winding of said first pair with one end of a power winding of said second pair; second conductive means connecting one end of the other power winding of said first pair with one end of the other power winding of said second pair; first and second series pairs of rectifiers; means connecting said first series pair of rectifiers between the other ends of said first pair of power windings; means connectingsaid second series pair of rectifiers between the other ends of said second pair of power windings, said first and second series pairs of rectifiers being so poled as to present easy current flow paths from said first conductive means through said windings and rectifiers to said second conductive means; a series pair of bias impedances; means connecting said series pair of bias impedances between said first and second conductive means; load impedance means; means connecting one end of said load impedance means to a point intermediate the rectifiers of said first series pair of rectifiers; means connecting the other end of said load impedance means to a point intermediate the rectifiers of said second series pair of rectifiers; means connected intermediate the ends of said load impedance means and connected intermediate the said bias impedances for applying alternating current power; bias winding means magnetically associated with said saturable cores; means connecting said bias winding means in energizable relation between said first and second conductive means; and control winding means magnetically associated with said saturable cores and connected so that any current flowing therethrough has one magnetic effect on the cores associated with said first pair of power windings and has the opposite magnetic effect on the cores associated with said second pair of power windings.

2. A self-biased magnetic amplifier comprising: first and second pairs of power windings, each of said power windings being discretely associated with a saturable core; first current conductive means connecting one end of the power winding of said first pair with one end of a power winding of said second pair; second current conductive means connecting one end of the other power winding of. said first pair with one end of the other power winding of said second pair; first and second asymmetrical conductive means; means connecting said first asymmetrical conductive means between the other ends of said first pair of power windings; means connecting said second asymmetrical conductive means between the other ends of said second pair of power windings, said first and second asymmetrical conductive means being so poled as to present relatively easy current flow paths from said first current conductive means through said windings and asymmetrical conductive means to said second current conductive means; bias impedance means; means connecting said bias impedance means between said first and second current conductivemeans; load impedance means; means connecting one end of said load impedance means to an intermediate point on said first asymmetrical conductive means; means connecting the other end of said load impedance means to an intermediate point on said second asymmetrical conductive means; means connected to a point intermediate the ends of said load impedance means and to intermediate point on said bias impedance means for applying alternating current power; bias winding means magnetically associated with said saturable cores; means connecting said bias winding means in energizable relation between said first and second current conductive means; and control winding means magnetically associated with said saturable cores and connected so that any current flowing therethrough has one mag netic efiect on the cores associated with said first pair of power windings and has the opposite magnetic effect on the cores associated with said second pair of power windings.

3. In a magnetic amplifier having means allowing a first set of power windings to conduct only during a first and subsequent like half-cycles of supply power and allowing a second set of power windings to conduct only during a second and subsequent like half-cycles of supply power, the improvement comprising: bias impedance means; means connecting a first portion of said bias impedance means in circuit with said first set of power windings and a second portion of said bias impedance means in circuit with said second set of power windings whereby any current flowing through said first set of power windings also flows through the first portion of said bias impedance means and any current flowing through said second set of power windings also flows through the second portion of said bias impedance means; bias winding means; and means connecting said bias winding means across said bias impedance means so that the current flowing through the first and second portions of said bias impedance means, during the first and second half-cycles of supply power respectively, provides a constant polarity energization to said bias winding means.

4. In a magnetic amplifier having means allowing a first set of power windings to conduct only during a first and subsequent like half-cycles of supply power and allowing a second set of power windings to conduct only during a second and subsequent like half-cycles of supply power, the improvement comprising: first and second bias impedance means each having first and second terminals; means connecting the first terminals of said first and second bias impedance means in common to said supply power; means connecting the second terminal of said first bias impedance means in circuit with said first set of power windings so that any current flowing through said power windings also flows through said first bias impedance means; means connecting the second terminal of said second bias impedance means in circuit with said second set of power windings so that any current flowing through said second power windings also flows through said second bias impedance means; bias winding means; and means connecting said bias winding means from the second terminal of said first bias impedance means to the second terminal of said second bias impedance means so that the current flowing through said first and second bias impedance means develops a constant polarity energization for said bias winding means.

5. In a magnetic amplifier having means allowing a first set of power windings to conduct only during a first and subsequent like half-cycles of supply power and allowing a second set of power windings to conduct only during a second and subsequent like half-cycles of supply power, the improvement comprising: first and second bias impedances connected in series circuit with said first and second sets of power windings respectively; bias winding means; and means connecting said bias winding means across said bias impedances so that" the current flowing through said bias impedances develops a constant polarity energization for said bias winding means.

References Cited in the file of this patent UNITED STATES PATENTS

Images