US2459069A - Electromagnetic system - Google Patents

Description

Jan. 11, 1949. 5, 172 LD 2,459,069

ELECTROIAGNETIG SYSTEI I Filed Dec. 14, 1946 4 Sheets-Sheet 3 Tia-E INVENTOR r4AN .5: FI'TZGMM M 6.4 A M ATTORNEYS Jan. 11, 1949. A. s. mz GERALD I 2,459,069

I I I I mcmouens'nc SYS'IBI Filed Dec. 14-, 1946 4 Shets-Shaet 4 INVENTOR ags.

ATTORNEYS the said source is capable.

Patented Jan. 11, 1949 ELECTROMAGNETIC SYSTEM Alan S. Fitz Gerald, Wynnewood, Pai, assignor to Warren Webster & Company, Camden, N. J a corporation of New Jersey Application December 14, 1946, Serial No. 716,193

4 Claims. 1

This invention relates to magnetic amplifier systems of the type which employ saturating reactors, and more particularly to systems of the above type wherein a constant relationship is maintained between the input and output magnitudes.

The present invention constitutes an improvement on that forming the subject matter of my co-pending application, Serial No. 666,867, filed May 3, 1946. In this earlier application I have disclosed a magnetic amplifier system primarily intended as a simple sensitive detecting device capable of operating an electro-responsive device such as a relay or the like, whenever the input current reaches a predetermined value.

It is an object of my present invention to provide an improved magnetic amplifier system of a type capable of furnishing a gain of constant magnitude. That is to say, the output is maintained at all times a constant multiple of the.

input.

According to my invention I provide a magnetic amplifier of a type which has .a. floating output-in other words, an output which is of indeterminate value when it is not subjected to a controlling influence-and I provide additional controlling means for maintaining the output at a predetermined value.

In electrical measurements and control systems it is often necessary to energize and operate controlling equipment in accordance with the magnitude of an electrical quantity. A feature of control systems of this type, in respect of which a problem is often presented, resides in the fact that the power levels associated with the said electrical quantities are of a very low order, such that, either these sources are entirely incapable of furnishing the amount of power necessary to operate the control equipment, or, due to their inherent high impedance or regulating characteristics, the withdrawal of any appreciablecurrent therefrom materially affects the magnitude of the voltage furnished thereby.

It is another object of my invention therefore to provide a magnetic amplifying system capable of furnishing an output having a voltage exactly equal to that of a given source, but capable of delivering power output, without change of voltage, substantially greater than that of which This I accomplish by providing connections for comparing the voltage of the output with the voltage of the source.

It will be apparent to those skilled in the art, that when these two voltages are balanced, as by connecting them in opposition, the resultant or 2 difference voltage will be of reversible polarity dependent upon which of these two voltages be the greater. Accordingly a magnetic amplifier arrangement responsive to the D. C. input polarity is required.

For this purpose I may use the polarized magnetic amplifier system which I have disclosed in my co-pending application, Serial No. 666,867, filed May 3, 1946, Fig. 7; or the more sensitive system disclosed in my co-pending application, Serial No. 719,445, filed December 31, 1946.

In other practical applications encountered in electrical problems the magnitude of the control source is of a low voltage or current value insufficient to actuate the desired control equipment and it is necessary to provide a power source of an amplified magnitude.

According to my invention, where this problem is encountered, I provide a voltage divider, as

for example, of the tapped resistor type, which I energize, across its total resistance, from the output of my magnetic amplifier.

I then provide connections for comparing the voltage between the tap on the voltage divider,

which may, for example, be one-tenth, one-hundredth, as may be desired, 0f the value of the total resistance, and the voltage of the input source which is to be amplified.

These and other novel features which I believe to be characteristic of my invention will be set forth with particularity in the appended claims. My invention itself, however, both as to its organization and method of operation, together with further objects and advantages thereof, may best be understood with reference to the following description taken in connection with the accompanying drawings, in which:

Fig. 1 is an electrical circuit diagram showing a single stage of a magnetic amplifier for the purpose of explaining a feature of the operation of my invention;

Fig. 2 is a diagram of a mechanical analogy descriptive of the manner of operation of my invention;

Fig. 3 is an electrical circuit diagram showing an embodiment of my invention;

Fig. 4 is a fragmentary diagram showing degails pertaining to the embodiment shown in Fig.

Fig. 5 an electrical circuit diagram showing another embodiment of my invention.

Referring to the drawings, I show in Fig. 1 a magnetic amplifier circuit comprising an arrangement similar to the circuit shown in Fig. 7 of my above cited co-pending application, in

. 3 which I have used like numerals to identity core i oi the saturating reactor, the A. C. winding 3, the saturating winding 3, together with the core 4, the primary winding 3, and the secondary winding 3, oi the transformer.

I show also resistances 1 and I, conductors, 3, l0, and .II, II, an alternating current source i3, rectifier i4, and conductors II for supplying an fiutplut circuit. I show also the feed-back wind- I have found that the arrangement of the above described circuit, both in carrying it into practice and depicting it in a diagram of connections, is simplified if the order or the-two series-connected elements, to wit, the secondary 3 of the transformer and the A. C. connections to the rectifier ll, be transposed. No eilect or any kind on the electrical action or operation of the circuit results from this transposition.

It will be noted that as this circuit is shown in the figures of my co-pendingappllcation the primary winding 3 oi the transformer is shown above the secondary winding 3. To avoid crossing of wires in the diagram I have, in Fig. 1 of the present application, reversed the position of windings 3 and.

In addition to simplifying the drawing of the diagram there is a practical advantage in this arrangement. In this magnetic amplifier'cir cuit there is a conjugate relationship between the transformer and the saturating reactor. It is therefore logical and convenient to mount these two units together and to seal them into a common housing. If this be done, the above mentioned transposition permits an internal connection and fewer leads need to be brought from the cased assembly.

I have referred above to the use of a magnetic amplifier-oi a type which has a iioating" output. In my co-pending application, Serial No.

tively low order of magnitude and of reversible polarity. For example, I show push buttons at 32 and 33 whereby a current. limited to a low valueby a high resistance 34, may be caused to flow in the input winding 3 from two oppositely connected direct current sources shown at 36 and 33.

The input connections and the polarity of the sources 35 and 38 are such that when push but- 666.867, I described briefly the eil'ect of feedback connections inmagnetic amplifiers of this type indicating therein the manner in which the magnitude of the feed-back eilect applied causes three diiIering qualitative forms of response to be" obtained. a Y

I referred therein to the condition of single stability, which occurs with moderate amounts of feed-back; to the condition of dual stability, which occurs with substantial amounts of feed-' back; andIreferred to a third, and an especially valuable type of action, which may be obtained ture oi the concomitant phenomena and to eX-.

plain the adjustments which are made in order to obtain the desired action.

. In Fig. '1, I show the output conductors ill of the magnetic amplifier connected to energize a load shown as an adiustible resistor II in series with an indicating instrument, as for example, a direct current milliammeter 26. I show also a second adjustable resistor 21 connected as a diverter, or shunt, in parallel with the teed-back winding it, for the purpose of varying the amount of feed-back at will.

I show also in Fig. l, a third resistor 33 which may be of a low, fixed, value, in series with the output of the magnetic amplifier. The P p se 0! the resistance 23 is to provide a slight additional voltage of fixed value, additive with referten 32 is closed, current flows in the winding 3 such that the M. M. F. thereby set up is additivein respect 0! that which is due to any current which may flow in the feed-back winding i3. Conversely, if push button 33 is operated the M. M. F. due to the current flowing in 3 is in opposition to the effect of i6.

In accordance with the operation of this circuit as explained in reference to Fig. 7 01' my co-pending application, Serial No. 666,867, it will readily be understood by those skilled in the art that operation of the push button 32 will tend to increase the output of the magnetic amplifier and that the operation of push button 33 will have the reverse eflect.

However, the manner in which the output of the magnetic amplifier, as indicated by the meter 23, will respond to an input applied to 3 will depend upon the circuit constants which are employed. The principal iactor which determines the nature of this response is the relation between the resistanceof the load circuit i9 and the amount of feed-back. Since it may not be convenient tovary the turns of the feed-back winding II the shunt or diverter 21 permits 01 variation in the amount or effective teed-back. The resistance of the rectifier i4 is also an essential element in the action which takes place. Thus it is not only the resistance of is alone which is of importance but the total circuit resistance comprising the sum of the load resistance is and the resistance of the rectifier ll. While the resistance oi is may be substantially constant, the re sistance of it is of a non-linear character.

Because of the fact that the resistance of N may reach a very high value at current values in the neighborhood of zero it is convenient to operate this circuit over a range of current value that does not actually go down to zero.

In my co-pending application, Serial No. 666,867, I have explained that the value of the resistances I and 8 may ii desired be so adjusted that the amplifier output has an off-set zero.

It is for this reason also that the slight additional source of voltage, in the forward direction, due to the drop across 28, is advantageous.

In practicing my invention I find it desirable to employ a rectifier having a contact area somewhat larger than would be selected on the basis of current carrying capacity alone, so that the resistance of the rectifier will not represent a substantial percentage of the total circuit resistance.'

It the push button 32 be closed, as hereinbefore explained, current will flow in the windin 3 in a direction tending to increase the output of the magnetic amplifier. This direction we may hereinafter refer to as input of positive polarity.

If this current be varied in magnitude, manually, by adjusting 34, and the response of the circuit, as manifested by the movement of the pointer of 26 be observed, the following action may be noted.

If the value. of 21 is quite low in relation to that of II, that is to say, if the amount of effective feed-back action is small in respect of the output circuit resistance, the pointer of the milliammeter will be deflected in an upward direction on. the scale when the current is applied to I; and to an extent related to the magnitude of the current in 3. If the current in 3. be now reduced the deflection will in a corresponding manner be reduced. In other words the action will be, qualitatively, substantially the same as is observed in'the absence of the winding it, except that the magnitude of the deflection will be somewhat increased by the presence of such feedvback action as is effective.

If on the other hand, 21 be adjusted so that a substantial amount of feed-back is present, or if the resistance of I! be reduced, such that in either case, there is a substantial amount of feedback in relation to the load circuit resistance, a condition of dual stability will be noted.

This will manifest itself to the observer of the pointer of 26 as follows.

If a current of positive polarity in 3 be slowly increased at a uniform rate there will at first be no marked response of the instrument pointer which will not appear to follow the increase in input. Under these conditions the circuit has a condition of stability tending to maintain a relatively low value of output current. A substantial amount of input is necessary to detach it from this condition. a

When the positive current in 3 reaches a sufficiently high value a trigger-like action will occur, aphenomenon of discontinuity will manifest itself, and the pointer will swing from a low value to a relatively high value during which movement it does not follow the controlling action of the current in 3 as was the case when low feed-back was used. The circuit will now be found to be translated from the first, low current, condition of stability to another, high current, condition of stability. If the current in 3 be now reduced it will be observed that the pointer of the meter will not return towards zero as the input is diminished; it will tend to remain at the high value of current corresponding to the second stable condition described.

With substantial amounts of feed-back this value of output current will be maintained even though the current in 3 be reduced to zero. and. in order to reduce the current in the output, and to return the pointer to its previous position, negative input must be applied to 3. This may be done by closing the push button 33. If negative input to 3 be slowly increased, similar con-- trolling action in the opposite direction to that already described in reference to positive input application of the external control F.

ment of the pointer 26 and a reliable object situated upon an approximately horizontal curved surface, may be of some assistance.

It is believed that explanation of the manner of operation of this circuit may more briefly be presented if reference be made to the visual aid which I have shown in Fig. 2.

A circular object 31 such as a sphere or cylinder is shown, in Fig. 2, as resting upon the upper surface of a containing member 33 capable of being tilted in either direction so as to induce motion of the reliable object 31. In Fig. 2, nine concepts of the containing member 38 are shown at a through i. It is shown as having a concave, flat, or a convex surface; it is also shown in three different positions, to wit, tilted to the left, horizontal, and tilted to the right.

The appended legends corresponding to the nature of the curvature of the surface of 36 and to its positions, refer, by analogy, to the adjustment of the circuit shown in Fig. 1.

The way in which the pointer of 26 will respond to the flow of input current in 3 will resemble the way in which the object 31 will respond to the application of a force, suggested in Fig. 2 at F, tending to displace it laterally to the left or to the right.

In the first case described, with a relatively small amount of feed-back, and with no input applied to 3, if the resistance of 30 be so adjusted that the pointer of 26 is at about mid-scale, the way in which the pointer of 26 will respond to the operation of the push buttons 32, 33 will be after the manner suggested in Fig, 2, b.

For example, in the mechanical analogy shown in Fig. 2, under the conditions of b, if no force be brought to bear upon the circular object 31 it will take up a position at the lowest portion 6 In both cases similar but opposite action will take place if the actuating effect be such as to cause a displacement in the other direction. That is to say, if the push button 33 be operated the pointer 26 will move to the left, when 33 is released, the pointer will return to its original position.

If the resistance of 36 be increased so that the forward E. M. F. due to 28 becomes less the effect upon the output of the magnetic amplifier will be analogous to that of tilting 38 towards the left as shown in Fig. 2 at a. Similarly if the voltage due to 28 be increased, by reducing the value of the resistance 30, the conditions will be such that the analogy indicated in Fig. 2, at 0, will apply.

On the other hand, the nature of'the action which occurs when there is a substantial amount of feed-back will resemble the conditions suggested in Fig. 2 at 9, h, and i, when the containing object 33 has a convex surface. In this case it will not be possible to retain the roliable object in the center of the curved surface without the It will tend either to remain at one end or the other of the convex surface. If by the application of F it is removed from either extremity, when the controlling effect is withdrawn it will if the containing member 38 is level, as shown at n, roll towards to whichever of the two ends it is nearer.

In the same way, with substantial feed-back applied to the circuit of Fig. 1, the pointer 23 will take up its position either in the neighborhood of zero or alternatively at the upper part of the scale and can only be transferred from one position to the other by an input of critical magnitude of positive or negative polarity as required.

The effect of adjusting the magnitude of the voltage due to 23 will, as before, he analogous to tilting 33 as in g or i.

It will be apparent to those skilled in the art, that if under one condition mid-scale stability is furnished and under another condition dual stability at upper and lower portions of the scale is established, there will exist an intermediate condition. Thus, by adjusting conjointly the value of the resistances 30, 21, and i3 it is possible to secure the floating condition above described.

When the correct adjustment has been arrived at, the following action will be noted. If a substantial value of input of positive value is applied by closing push button 32, the pointer of the milliammeter 26 will move towards the right indieating a current value in the upper portion of the scale. If push button 32 be now released the pointer will remain in the vicinity of the latter position for a substantial period of time. Ultimately, it will tend very slowly to drift away from this position. If, on the other hand, negative excitation be applied to 3 by operating. the push button 33 the pointer will be deflected towards the left and will take up a position corresponding to a low current value.

If 33 be now released again the pointer will for a considerable time remain at or near that position.

If, on the other hand, the pointer be positioned, say, at around mid-scale, as may be done by suitably operating 32 or 33 it will in like manner retain this position for an extended period.

Suppose now 34 be adjusted to a relatively high value such that the input applied to 3, on operating either 32 or 33, is of a low order, if 32 be now operated the pointer of the mllliammeter will be observed slowly to move towards the right, that is, to say, toward the upper portion of the scale. If 32 be operated intermittently the pointer may be inched upwards. So long as 32 is de- Whenever 32 is released this motion will stop. Exactly the opposite action will be noted if 33 be actuated in a similar manner.

Under these conditions the action of my circuit in Fig. 1 will correspond to the condition Suggested in Fig. 2 at e, when the surface of 33 is approximately fiat and is horizontal. If the circular object 31 be displaced in either direction, there is no strong restoring force either tending to return it to the central position, as when the surface of 38 is concave; or tending to cause it to move to either the right or left hand extremity of 38 as in the case in which the surface is convex.

Still considering the mechanical analogy it will be evident to those skilled in the art the force necessary to cause a displacement of given magni-- tude of the circular object 31 when the surface is nearly flat, will be very much less than that necessary when the surface is curved.

In a similar manner when the adjustments of my circuit in Fig. 1 are made so that the "floatment of the pointer of 26 up or down, when input, I

ing" condition obtains, input to the winding 3 of very much less magnitude produces a controlling eflect than is the case when either the condition or mid-scale stability obtains, such as is suggested in Fig. 2 at a, b, and c, or the condition of dual stability as indicated at a, h, and i.

I wish particularly to point out the importance of this feature of my invention which contributes to a major extent, to the extremely effective and sensitive control action which is achieved. a

It will be evident to those skilled in the art, that even though electrical action analogous to an absolutely flat, level surface, as in Fig. 2; e, is not completely achieved, the more nearly this condition is approached the moresensitive response will be obtained. To adjust my circuit of Fig. 1 for optimum results the values of the three resistances ll, 21, and 30 are arrived at in accordance with the following procedure.

With an input applied to 3 of low. magnitude the degree to which levelness" of characteristic is approached is indicated by the rate of, moveof the same magnitude, of positive and negative polarityis alternately applied to 3. If the read ing of 28 goes up, in response to a positive input applied to 3, more readily than it is reduced in response to a negative input of the same rhagni. tude, the resistance of 30 should be decreased. Conversely; if the pointer of 23 move at a greater speed towards the left of the scale in response to input of a negative polarity-but more slowly towards the right of the scale with the same amount of positive input, then 30 should be increased.

The flatness of the characteristic is indicated, in the first place, in making preliminary adjustment, by whether the pointer tends to drift towards mid-scale or whether it tends to drift tc'either the left extremity or the right extremity of the scale in the absence'of any input applied to 3. When action indicative of marked concave or convex characteristic has been eliminated by adjustment, indication of the flatness may be obtained by noting the speed of movement of the pointer of 28, at different portions of the scale, in response to application of inputs of very low magnitude to 3.

Thus, if the pointer of 23 is indicating a low current value towards the left hand end of the scale, and a positive input is applied, the pointer will commence to move towards the right. If,

I with no change in the amplitude of the input, the

rate of movement of the pointer increases, this will be an indication that there is still some"convexity present. If, on the other hand, the rate of movement decreases, and :if perhaps the pointer comes to rest without reaching the upper part of the scale it can be noted that some degree of concave characteristic is still present.

'In other words, the objective which is aimed at in making this adjustment is that, the application to the input winding 3 of a given current of low'value of both positive and negative polarity should, in the first place approach a speed of response which is equal in both directions. In the second place this speed of response should be substantially equal at all points on the scale of and output connections. In addition to this first stage, I show a second and third stage, comprising in each case magnetic amplifiers of the neutral type which may be substantially as shown in Fig. 1 of my co-pending application, Serial No. 969,997, except for the transposed method of showing 9 and I4 as above described.

In order to simplify the drawing in Fig. 3 and to avoid repetition of detail I have indicated, in the diagram, the collective group comprising the saturating reactor and transformer pertaining to the neutral stages, in the manner depicted in Fig. 4 at 49. I have used like reference numerals to identify the structure disclosed in Fig. 1 of my co-pending application, Serial No. 666,867, comprising the core I of the saturating reactor, the A. C. winding 2, and the saturating winding 9, together with the core 4, the primary winding 9, and the secondary winding 9 of the transformer.

I also show connections therefrom at 4! and 42, 49 and 44, 49 and 49, and also at 41.

I show in Fig. 3 the resistance l9 as in Fig. 1. It is to be understood that the major portion of the output circuit resistance comprises the resistance oi the input of the second stage. However, in view of the fact that exact resistance values are desirable in order to get the floating control action referred to, the resistance i9 is provided for trimming purposes. That is to say, the resistance of the input circuit of the second stage is designed to be somewhat less than the expected resistance value which will give the desired characteristic. l9 may then be adjusted so that the total resistance is as desired.

In view of the fact, as hereinbefore described. that it is desirable to operate the floating control stage with an ofl-set zero, the values of the resistances I and 8 and. if necessary, the turns of the transformer windings 9 and 8 in the second stage, are so adjusted that the output of the second stage is substantially zero when the input to the second stage is of the above oiI-set value.

for standardizing potentiometers and the like.

I show also in Fig. 3 an output circuit 49 supplied by the conductors i9 of the last stage. This output circuit 49 may, for example, consist of the slide wire of a potentiometer or any other work or measurement device in which it is desired that a current of constant value be maintained.

As shown in Fig. 3 I connect together the negative terminals of the input circuit 48 and the output circuit 49 and I connect the input winding 9 of the first stagebetween the positive terminals of 49 and 49. 7

It will be evident to those skilled in the art, that, according to these arrangements, when the voltage of the output circuit 49 is exactly equal to that of the source 48 no current will flow in 9.

In the event that the voltage across the output 49 should be reduced, from any cause, to a value less than that of 48, a difference current will flow in 3, which current will be in the positive direction; that is to say, it will tend to increase the output of the first stage and thereby raise the value of the current flowing in the output circuit of the last stage such that the voltage across 49 will be increased and the current in 9 will cease to flow.

In like manner, should the voltage across 49 become greater than that of 49 current will how in 9 in a negative direction which will bring about a reduction in the voltage across 49.

After a lapse of time, due to the floating action of the first stage, some drifting action may take place either upwards or downwards. However, immediately any perceptible deviation occurs, current flows in 9. Therefore. the current in 49 is automatically maintained in very close agreement with that of 49.

While I have shown in Fig. 3 only two additional stages of neutral amplification I wish it to be clearly understood that in accordance with my invention I may add such further stages as may be desired or requisite.

My invention makes it possible to supply to the load circuit 49 quite substantial amounts of power, regulated to the desired voltage. Thus substantial current drain may be withdrawn from the circuit of Fig. 3 with inappreciable how of current in 49. i

Thus, the arrangement of Fig, 3 comprises a magnetic amplifier capable of furnishing any desired magnitude 0! current amplification with power gain to correspond.

I show in Fig. 5 an arrangement somewhat similar to that of Fig. 3 but of a still more sensitive type and intended for voltage amplification.

Ishow in Fig, 5 a "floating control stage and two additional neutral stages identical with the arrangement of Fig. 3.

I show also in Fig. 5 an additional stage of magnetic amplification of the polarized type such as I have disclosed in my co-pending application, Serial No. 719,445, filed December 31, 1946. This polarized amplifier is connected before the "floating ontrol" stage.

In Fig. 5 instead of the single isolating transformer II, I show an isolating transformer 99 having a plurality of secondary windings for supplying separate, isolated, A. C. voltages to the rectifier 29 and the individual stages of the magnetic amplifier.

The first secondary winding 9| supplies the first, polarized type, stage. Secondary winding 92 supplies the rectifier 29. Secondary windings 99, 94, and 99 respectively, supply the floating contro stage, the first neutral stage and the second neutral stage.

The last stage may be designed to operate at a higher voltage than the other stages and the secondary winding 99 may accordingly be designed to furnish a voltage of the desired magnitude.

I show in Fig. 5 an input source 99 which may have an E. M. F. of very low magnitude as, for example, a thermo-couple or pH electrode.

I show also in Fig. 5 two resistors 91 and 98.

The resistance of 99 may conveniently be an exact fraction, as for example, one tenth or one hundredth of the sum 0! the resistance of 91 and 98.

Due to the very great sensitivity of this arrangement, as will readily be understood by those skilled in the art, the balance circuit will some times be found to have a tendency to hunt or oscillate. I have found that the connection of a bilaterally conducting. contact device or limiter may with advantage be connected across the input winding 9 as shown at 99 in Fig. 5.

. This device is substantially non-conducting whenever the circuit is approximately balanced and therefore, involves no reduction in the sensitivity. Whenever hunting takes place so that the circuit swings back and forth through the In Fig. I connect the output leads I80L, I88R of the polarized amplifier 50, to energize the winding 3, I connect I8L to the positive terminal of the source'56 and I connect I8R to the Junction of the resistances 57 and 58.

For a detailed description of the operation of the polarized magnetic amplifier 80, reference may be had to the specification of my above cited cG-pending application, Serial No. 719,445. Briefly, when a direct current input of reversible polarity is applied to terminals ML and I8R, an output, of amplified magnitude, andof corresponding polarity appears at terminals I881... I8IIR. If the polarity of the input applied at "L, |8R is reversed, the output likewise will be of opposite polarity.

The operation of my arrangement as shown in Fig. 5 is very similar to that of Fig. 3 except that it is susceptible of amplifying inputs of very much lower power levels. Thus, the voltage drop across resistance 58 is automatically maintained, at all times within very close limits, equal to the voltage of the source 56.v Nevertheless, the current withdrawn from 58 is negligibly small.

In Fig. 5 I show a pair of output terminals 8| and 82. Terminal 8| is connected to the positive extremity of the resistance 5'! and terminal 62 is connected to the negative extremity of 58. Thus,

the voltage appearing between H and 82 will be the sum of the voltage drops across 51 and 58. If the ratio of the resistances oi 51 and 58 be as indicated, the voltage appearing at 8| and 52 may be ten times, or one hundred times that of the source 56, or such other ratio as may be desired. Power may conveniently be withdrawn at terminals 8| and 62 to supply a load circuit which will receive a voltage regulated at all times to a definite multiple of that of 56. v

It is possible that in some instances a tap may be established at a given fraction of the resistance of the load circuit in which case resistances 57 and 58 may be considered as comprising the output or load circuit. 1

I give below some typical numerical values taken from a practical embodiment of my invention which I have constructed and tested. It should, however, be clearly understood that my invention may be carried into effect on any desired scale of magnitude and may be modified in any manner conformable with the purpose and application for which it is to be employed. I am, therefore, in no wayto be limited by the following data which is included only for the purpose of facilitating the understanding of my invention.

For example, in one embodiment of my invention, the floating control" stage shown by itself in Fig. 1 and also in Figs. 3, and 5, the core I consisted of 0.75" stacking of three-limb laminations having overall dimensions approximately 2.88" x 2.69" wide. The center limb was 0.75". and the two outside limbs were 0.438" wide. The stamping has a window 1.13 square inches. The laminations were made from a nickel iron alloy containing more than '70 per cent of nickel.

The A. C. winding 2 had 150 turns per coil connected in series. turns and a resistance of 500 ohms. The feed- The input winding 3 had 7500 l2 back winding had 350 turns and a resistance of 15 ohms.

The rectifier I 4 consisted of the copper oxide type having one plate of 1%" in diameter in series in each leg. The primary winding 5-0! the transformer had 180 turns and the secondary winding 8 had 100 turns.

The voltage of the source 18 was 25 volts, 60 cycles, and that of the secondary winding of 31 was 2.5 volts.

The total resistance of the load circuit energized by the output of the floating control stage including that of the resistor l9 was ohms.

The resistance of 'l was 180 ohms and the resistance of 8 was 70 ohms.

The resistance of 21 was ohms,that of 28 was 10 ohms. The current in 28 was of the order of 50 ma. so that the positive voltage set up by 28 was about one-half volt.

In one embodiment of my invention according to Fig. 3, in which I usedthree neutral stages in addition to the first stage, a maximum output approaching 5 watts was obtained. When the output was of the order of 0.01 of a watt the current in the input winding 3 was approximately 1.0 microamp. When the output was approximately 3.5 watts the current in winding 8 was 1 microamp. Between these limits the current was less than 1 microamp. These figures indicated the extent to which the desired flatness? cept in so far as is necessitated by the prior art and the spirit of the appended claims.

What I claim as new and desire to secure by Letters Patent oi the United States is:

1. A magnetic amplifier comprising: a saturable core device having an alternating current winding and a plurality of direct current saturating windings; means connecting said alternating current winding in series with a" first impendan-ce across an alternating current source; a transformer having primary and secondary windings; means connecting said primary winding in series with a second impedance across said alternating current source; a circuit containing in series said saturable core alternating current winding, said transformer secondary winding, and a rectifier; a load device; an output circuit connecting said rectifier with said load device in series with one of said direct current saturating windings and an additional source of direct current voltage drop, the relative magnitudes of the excitation of saidsaturating winding, the resistance of said load device, and said direct current voltage, being so apportioned that the system has two difierent conditions of stability giving two difierent values of current in said output circuit; and means for predetermining which of said con-' ditions shall obtain so as selectively to control the current in said. output circuit.

2. A magnetic amplifier comprising: a saturable core device having an alternating current winding and a plurality of'direct current seturating windings; means conecting said alternating current winding in series with a first impedance across an alternating current source; a transformer having primary and secondary windings; means connecting said primary winding in series with a second impedance across said alternating current source; a circuit containing in series said saturable core alternating current winding, said trasformer secondary winding, and a rectifier; a load device; an output circuit connecting said rectifier with said load device in series with one of said direct current saturating windings and an additional source of direct current voltage drop, the relative magnitudes of the excitation of said saturating winding, the resistance of said load device, and said direct current voltage, being so apportioned that the system is in a condition of instability whereby, in the absence of energy applied to another of said direct current saturating windings, the current in said output circuit is of indeterminate value, and whereby a direct current magnetomotive force input applied to said second-mentioned saturating winding of a very low value in the same sense as the excitation of said first-mentioned saturating winding causes a continuing increase in the current in said output circuit and a continuing decrease when applied in the opposite sense, together with further means for at one time applying direct current of one polarity and at another time applying direct current of the opposite polarity to said second saturating winding, thereby to maintain the current in said output circuit at a desired value.

3. A magnetic amplifier comprising: a saturable core device having an'alternating current winding and a plurality of direct current saturating windings; means connecting said alternatin current winding in series with a first impedance across an alternating current source; a transformer having Primary and secondary windings; means connecting said primary winding in series with a second impedance across said alternating current source; a circuit containing in series said saturable core alternating current winding, said transformer secondary winding, and a rectifier; a load device; an output circuit connecting said rectifier with said lead device in series with one of said direct current saturating windings and an additional source of direct current voltage drop, the relative magnitudes of the excitation of said saturating winding, the resistance of said load device, and said direct current voltage, being so apportioned that the system is in a condition of instability whereby, in the absence of energy applied to another of said direct current saturating windings, the current in said output circuit is of indeterminate value, and whereby a direct current magnetomotive force input applied to said second-mentioned saturating winding of a very low value in the same sense as the excitation of said first-mentioned saturating winding causes a continuing increase in the current in said output circuit and a continuing decrease when applied in the opposite sense, together with a second source of direct current voltage of a predetermined value, means for deriving a direct current voltage proportional to the current in said output circuit, and means responsive to the difference between said predetermined voltage and said derived voltage connected to said second-mentioned saturating winding so as to maintain a predetermined relation between the current in said output circuit and said direct current source.

4. A magnetic amplifier comprising: a saturable core device having an alternating current winding and a plurality of direct current saturating windings; means connecting said alternating current winding in series with a first impedance across an alternating current source; a transformer having primary and secondary windings; means connecting said primary winding in series with a second impedance across said alternating current source; a circuit containing in series said saturable core alternating current winding, said transformer secondary winding, and a rectifier; a load device; an output circuit connecting said rectifier with said load device in series with one of said direct current saturating windings and an additional source of direct current voltage drop, the relative magnitudes of the excitation of said saturating winding, the resistance of said load device, and said direct current voltage, being so apportioned that the system has two different conditions of stability giving different values of current in said output circuit, whereby, in the absence of energy applied to another of said saturating windings, the current in said output circuit maintains one of two different magnitudes, and whereby a direct current magnetomotive force applied to said second-mentioned saturating winding of one polarity causes the system to transfer from one of said conditions of stability to another, while application thereof in the opposite polarity causes said system to transfer from said second-mentioned condition to said first-mentioned condition.

ALAN S. FITZ GERALD.

No references cited.

Images