US3579088A

US3579088A - Ferroresonant transformer with controllable flux

Info

Publication number: US3579088A
Application number: US814393A
Authority: US
Inventors: Taylor C Fletcher; Lawrence M Silva; Bruce L Wilkinson
Original assignee: Individual
Current assignee: Individual
Priority date: 1969-04-08
Filing date: 1969-04-08
Publication date: 1971-05-18
Anticipated expiration: 1988-05-18

Abstract

A ferroresonant transformer with means for varying saturation flux capacity for controlling the transformer output. A ferroresonant transformer with at least a portion of the core structure carrying the secondary winding having two separate sections providing parallel magnetic paths for the secondary and resonant flux, with a control winding on one of the sections and a switching circuit for opening and closing the control winding. A low frequency version utilizing E and I laminations with the control winding on one of the outer legs of the E. Another low frequency version utilizing E and I laminations with the control winding encircling a portion of the secondary magnetic circuit. A high frequency version utilizing a plurality of toroid cores with the control winding on one of the cores.

Description

United States Patent [72] Inventors Taylor C. Fletcher 2,598,617 5/1952 Stimler 323/48(X) 1534 Sunnycrest, Fullerton, Calif. 92632; 3,079,546 2/ 1963 Kuba 323/50 Lawrence M. Silva, 4110 Mlritine, 3,1 17,274 1/1964 Essinger 323/56 Portuguese Bend, Calif. 90274; Bruce L. 3,123,764 3/ 1964 Patton 323/56 gllllikinsongosasm Sharynne Lane, Torrance, 3,361,956 1/1968 Sola 323/56(X) [21] No 814,593 Primary Examiner-J. D. Miller [22] Film Apt 8 1969 Assistant Examiner-Gerald Goldberg I [45] Patented May 18, 1971 Attorney-Harris, Klech, Russell & Kern [54] FERROREmNm TRANSFORMER WITH ferroresonant transformer with means for CONTROLLABLE FLUX varying saturation flux capacity for controlling the trans- 17 Claims, 16 Drawing Figs fonner output. A ferroresonant transformer with at least a portion of the core structure carrying the secondary winding [52] US. CL 323/6, having two separate sections providing parallel magnetic paths 32 323/56, 323/60, 132 /8 for the secondary and resonant flux, with a control winding on [51] Int. Cl G051 1/38, one f th ti ns and a switching circuit for opening and 6051' closing the control winding. A low frequency version utilizing [50] FR ofSearch 323/6, 48, E and laminations with the co ug] winding on on: of

lhe

50, 56, 57-61, 321/57, 68 outerlegs of the E. Another low frequency version utilizing E. and llaminations with the control winding encircling a portion Mm Cited of the secondary magnetic circuit. A high frequency version UNITED STATES PATENTS utilizing a plurality of toroid cores with the control winding on 2,519,425 8/ 1950 Barlow 323/56(X) one of the cores.

:9 in, 1111/ 1l? 1: 22 '1' PR/. c D g '1 I; /2 llln' 111] CONTROL (Z/ c/Rcu/T Patented May 18, 1971 3,579,088

3 Sheets-Sheet 5 Era. 12. F1614} .307 RESONANT 305 SECONDARY 302/; 303 30 1 /522 1m.- 3 W/ //Z 300 L 3/0 A 3% 3/4 3/3 304 30/? /6 7 PRIMARY 7 FIG RESMW 3w $ECONDAEY CONTROL 30 I; 303 W 2 F 307 Z'SONW se'cowoAlav CONTROL 302 sec. 306

16 /N VENRJRS 72010;? C. FLETCHER, RES. RES. 307 LA wee-wee M. 5/L VA 6: 55c. CON7'ROL309 BRUCE L. VV/LK/NSON 3/2 3/5 32/ 322 BY THE/2 ATTOENEYB HAZE/5, K/ECH, RUSSELL & KEEN FERRORESONANT TRANSFORMER WITH CONTROLLABLE FLUX This invention relates to ferroresonant transformers and in particular, to a new and improved ferroresonant transformer incorporating means for varying the saturation flux capacity in the core structure for controlling the transformer output. Ferroresonant transformers are widely used today for a variety of regulating and control purposes and the basic design is shown in the US. Pat. to Sola, No. 2,143,745. A conventional ferroresonant transformer utilizing toroid cores is shown in the US. Pat. to Sola, No. 2,753,5l3.-Various modifications and improvements are shown in other patents to the same patentee. It is an object of the present invention to provide a new and improved ferroresonant transformer construction which may be utilized with any of the presently known ferroresonant transformers and which provides for saturation flux capacity variation and transformer output control.

The conventional ferroresonant transformer has a magnetic core structure with a primary winding, a secondary winding and a resonant winding thereon. A capacitance is connected across the resonant winding. The resonant winding may be a separate winding or may actually be the secondary winding, with the capacitance connected across the secondary winding. A leakage flux path is provided in the core structure. In a transformer utilizing E and l laminations for the core structure,.a.quantity of core material is installed between the primaryand secondary windings to provide a magnetic shunt for the leakage flux path. In transformers utilizing a'toroid for the corestructure, at least two toroids are utilized with the primary winding linking all of the toroids and with the secondary winding not linking all of the toroids. The ferroresonant type of operation may also be achieved with a core structure having two or more separate core units, and several such devices utilizing a saturating transformer, without a shunt, and a series choke are shown in the U5. Pats. to Schmutz et al., No. 2,179,353, John et al., No. 2,505,620, Buie No. 2,764,725, and Kohn No. 2,967,271. The theory of operation and the details of construction of these conventional regulating devices may be obtained from various prior art publications, including the aforementioned patents. The term ferroresonant transformer as used herein is intended to include all such regulating devices.

In all ferroresonant transformers, the saturation flux capacity of the secondary magnetic circuit is predominant in determining the output voltage for a given number of secondary turns and a given frequency.

Because of external drops such as rectifiers and wiring re sistance and because of other parameter changes such as frequency and saturation drifts with temperature, means for controlling the saturation flux capacity and hence the output voltage is desirable. Attempts have been made to achieve such control in the past by using windings with DC bias which affect the efiective saturation flux density of the core material. These techniques require a sizeable mmf. in order to achieve control and the cost and complexity of the required driving 1 circuits offset the advantages of using a ferroresonant approach.

The'present invention provides for changing the saturation flux capacity of the secondary magnetic circuit in a ferroresonant transformer by isolating a portion of the secondary core material andcontrolling the flux in this section. In the present invention .the secondary magnetic cross-sectional area is divided into two sections, which may be referred to as an uncontrolled section and a controlled section. Means are provided in the controlled section to limit the maximum value of the instantaneous flux passing through this section. Said means comprises a control winding encircling the controlled section and a switch element for shorting or opening said winding. By varying the time in the cycle at which the switch is operated, the maximum instantaneous flux in the controlled section can be varied from zero to a maximum value equal to the saturation flux capacity of the controlled section. As used herein, the term saturation flux capacity means the sum of (I) the product of saturation flux density of the magnetic material multiplied by the cross-sectional area of the core at the uncontrolled section and (2) one-half the total change of flux in the core at the controlled section.

Accordingly, it is an object of the invention to provide a new and improved ferroresonant transformer with a core construction incorporating two separate sections in the secondary path, with a control winding on one of the sections, and switch means for opening and closing a circuit across the control winding. A further object is to provide transformers incorporating the invention for operation at low frequencies and for operation at high frequencies. An additional object is to provide such a transformer which may incorporate the various features of conventional ferroresonant transformers and which may be constructed utilizing present day manufacturing techniques, including E and I laminations and toroidal cores.

Other objects, advantages, features and results will more fully appear in the course of the following description. The drawings merely show andthe description merely describes the preferred embodiments of the present invention which are given by way of illustration or example. Various configurations for the magnetic core structure may be utilized including simple rectangular cores, E and I lamination cores, toroids,C- cores, and separate choke and transformer elements, and several specific forms are illustrated. The switch means for shorting and opening the control winding typically may include a switch element and a control circuit for actuating the switch element. Various devices may be used as the switch element, including a simple mechanical switch or relay, a transistor, an SCR, a Triac, athyratron, an ignitron, a gas triode, a magnetic amplifier, and a saturable reactor. Various 7 circuitry arrangements may be used as the control circuit, in-

cluding a synchronized oscillator, a phase shifter, a magnetic amplifier, and a saturable reactor. Several specific examples of the switch means are set out herein. The switching operation may be performed every cycle or every half-cycle, and circuits for both modes are illustrated. 1

In the drawings:

FIG. I is a diagram of -a ferroresonant transformer incorporating an embodiment of the present invention;

FIG. 2 is a view of an alternative form of construction of a ferroresonant transformer of the present invention suitable for use at power frequencies;

FIG. 3 is a diagram indicating typical flux waveforms in the .controlled and uncontrolled sections and the total secondary FIG. 4 is a schematic diagram of a circuit providing a synchronized oscillator with half-wave control for use with the transformers of FIG. 2 and FIG. 11;

FIG. 5 is a view of an alternative form of construction of a ferroresonant transformer incorporating another embodiment of the present invention;

FIG. 6 is a view of another alternative form of construction of a ferroresonant transformer incorporating a preferred embodiment of the present invention for use at higher frequencies;

shifter control for use with the transformer of FIG. 6;

FIG. 8 is a block diagram illustrating the circuit of FIG. 7; FIG. 9 is a schematic diagram of an alternate circuit providing a synchronized oscillator with full-wave control for use with transformers described herein;

FIG. 10 is a diagram of a ferroresonant transformer of this FIG. 7 is a schematic diagram of a circuit providing a phase FIGS. l4, l and 16 are sectional views taken along the lines 14-14, 15-15, and 16-16, respectively, of FIG. 13.

The transformer of FIG. 1 includes a magnetic core which may be constructed in the usual manner of interleaved or butt stacked laminations. The various conventional transformer manufacturing processes may be utilized. Typical core constructions for use with E and l laminations are illustrated in FIGS. 2 and 11-16. Core constructions for use with toroids are illustrated in FIGS. 5 and 6 and will be described later. Referring to .FIG. 1, an input or primary winding 11 is provided on a leg portion 12 of the core 10. An output or secondary winding 13 is provided on another leg portion 14. In the embodiment illustrated, the secondary winding also serves as the resonant winding, and a capacitor 16 is connected across the secondary winding 13. In an alternative arrangement, a separate resonant winding could be provided on the leg portion 14, with the capacitor 16 connected thereacross.

A leakage flux path is provided in the core 10 between the primary winding ll and the secondary winding 13, and comprises a magnetic shunt of

core sections

20, 21 with a nonmagnetic gap 22, usually an air gap, therebetween. The elements described thus far are found in the conventional ferroresonant transformers such as described in the aforementioned Sola patents and reference may be made to the prior art for a study of the theory of ferroresonance. A power source is connected to the primary winding and a load is connected to the secondary winding, and the transformer functions to maintain the output voltage of the secondary winding constant within predetermined limits for variations in load and variations in voltage of the power source.

In the core of the invention, as illustrated in FIG. 1, the secondary leg portion 14 is divided into two

sections

23, 24, which may be identified as the uncontrolled section 23 and the controlled section 24. The uncontrolled section is designed to saturate. A control winding 25 is provided on the section 24. A control circuit 26 operates a switch 27 connected across the control winding 25. The control circuit serves to open and close the switch at a particular phase angle of the output waveform. This phase angle is a prescribed function of a controlling signal. If a closed loop regulating device is desired, the controlling signal is derived from the. output voltage. While a variety of devices may be utilized as the switch, contemporary solid state devices are presently preferred, such as a siliconcontrolled rectifier, and two examples of control circuits and switches are described herein below.

When the switch 27 is open, the control winding 25 is open circuited and the secondary and resonant flux varies in both the uncontrolled section 23 and the controlled section 24. With the switch 27 closed, the control winding 25 is short circuited and further variation of flux in the controlled section is prevented. Since the flux in the controlled section is maintained constant by the action of switch 27, the total saturation flux capacity is reduced as compared to the total saturation flux capacity when the switch is closed for a minimum duration; the time required for the capacitor ring-over cycle.

In order for the device to operate in the ferroresonant mode the uncontrolled section must saturate. During periods when the mmf. across the controlled section exceeds the mmf. required to maintain the desired flux in this section the switch must be shorted or closed. The maximum switch closure time is a half-cycle and this condition corresponds to the the minimum total secondary saturation flux capacity. Since increasing the switch closure time causes a reduction in total secondary saturation flux capacity, a corresponding reduction in the output voltage is obtained. Therefore by opening or closing the circuit across the control winding, the saturation flux capacity and hence the output voltage may be changed.

Prior to closing the switch and shorting the control winding the flux variation in the controlled section follows the instantaneous secondary voltage. The magnitude of the flux rate of change in the controlled section is determined by the instantaneous secondary voltage and the ratio of the reluctances of the two sections. After the switch in the control winding closes, and thereby shorts the control winding, the flux in the controlled section remains constant because any further changes in flux are prevented by the induced mmf. in the short-circuited control winding.

As a consequence of shorting the control winding and forcing the flux in the controlled section to remain constant, the rate of change of flux in the uncontrolled section must then increase to a level sufficient to maintain the voltage that existed in the secondary winding at the time of switch closure.

The secondary voltage prior to and after switch closure at time t (FIG. 3), must be identical because the resonant capacitor across the secondary magnetic circuit prevents any instantaneous change of voltage. The instantaneous variation of fluxes in the two sections and the control action of the switch and control winding is indicated in FIG. 3. In the drawing, t, and t are the times when the fluxwaveforms pass through zero, is the time when the switch is closed, and I is the time when the switch is opened.

After the switch closes, the flux in the uncontrolled section continues to increase until the magnetic material in the uncontrolled section saturates. When this occurs, at time 1 the secondary resonant winding reflects a low impedance to the resonant capacitor and initiates a ring-over cycle that inverts the voltage on the secondary and resonant circuits. As a result of the inversion of the secondary voltage by the capacitor ringover cycle the rate of change of fluxes reverses and the flux in the uncontrolled section begins to decrease. At time 2 when the flux in the uncontrolled section decreases to a value approximately equal to the magnitude of flux that existed at the time of switch closure, the switch is opened. With the switch open the flux in both sections can now change. The total rate of change of flux in both sections is again determined by the instantaneous value of the secondary voltage, and the flux division between the two sections is again given by the ratio of the reluctances. After switch opening the magnitude of the rate of change of flux in the uncontrolled section decreases since the rate of change of total flux before and after switch opening must be identical.

The process then repeats itself in the following half-cycle with the polarity of fluxes and voltages being reversed.

The peak value of the total flux is equal to the peak value of the flux in the controlled section plus the peak value of flux in the uncontrolled section.

By varying the time of switch closure the, peak flux in the controlled section is varied, since the flux in this section increases up the time the switch is closed. If the switch is closed at the time of flux zero crossing the peak flux in the controlled section will be approximately zero. If the switch is closed at .the time the secondary flux waveform has a maximum, the

peak flux in the controlled section will be a maximum and will be equal to the saturation flux capacity of the controlled section.

The peak value of the flux in the uncontrolled section is not affected by varying the switch closing point since the uncontrolled section is driven into saturation each half-cycle.

As a consequence, the peak value of the total secondary flux will vary as the peak value of the controlled flux.

Since the average secondary output voltage is directly proportional to the peak value of the total flux, the magnitude of the average secondary output voltage can be varied by varying the length of the interval when the switch is closed.

Since the device of this invention is operating in the ferroresonant mode and hence the secondary voltage does not ring over until the uncontrolled section saturates, the shorting or opening of the control winding does not introduce a discontinuity in the output voltage waveform. The only effect of shorting the control winding is to cause a change in the amplitude of the waveform over the entire half-cycle.

For operation of the transformer as a voltage regulator, a smooth control between the two conditions is desirable. This may be achieved by shorting the control winding for a certain portion of each cycle or each half-cycle of the primary supply frequency. By varying the duty cycle between open circuit and closed circuit conditions, the output may be smoothly adjusted from minimum to maximum voltage. Suitable circuits for effecting this control'are illustrated in FIGS. 4 and 7 and 9.

FIGS. 4 and 7 are half-wave control circuits and operate once per cycle. FIG. 9 is a full wave control that closes the switch each half-cycle. i

The ferroresonant transformer of the invention circulates sizeable mmf. in the control winding. One advantage of the transformer of the present invention is that this mmf. is generated by the transformer in the form of induced current and the control circuit must merely contain a switch capable of carrying this current. No separate power supply is required to supply the mmf. in the control winding.

When the controllable ferroresonant transformer of the invention is used as a closed loop regulator, a method for determining when in the cycle to close the control winding is desired. Most existing phase control methods are troubled with the fact'that the instantaneous phase angle of the ferroresonant transformer varies with load changes and varies during dynamic voltage variations in the output. Such phase variations can cause severe instabilities in a phase controlled loop.

A more satisfactory mode of control is the type embodied in the circuit of FIG. 7. In this circuit the control winding voltage is integrated with respect to time and the switch across the control winding is closed when this integrated value reaches the required level. This level is determined by comparing a signal proportional to the output voltage with a reference voltage and amplifying the error signal. By varying the proportionality relationship between the output voltage and the reference voltage, the output voltage amplitude can be adjusted, as by resistors 109 and 110. The integrator may be reset at the time the short circuit is applied to the control winding so that the integrator will be ready for the next halfcycle.

In some applications, it is only necessary to short the control winding once every cycle instead of every half-cycle because the magnetic core will automatically saturate at the desired time during the other half-cycle. This occurs when a core with square loop material is used in the controlled section of the transformer. Under these conditions, the integrator of the control circuit may be reset at any time between the application of the short circuit and the beginning of the next cycle. A variety of integrators may be utilized, such as a resistor and capacitor circuit, a Miller integrator, an operational amplifier connected as an integrator, a magnetic amplifier, and the like.

Turning now to the embodiment of FIGS. 2 and 4, a core 30 is fomted of butt stacked E and I laminations 31, 32, respectively. A primary winding 33 is positioned about the center leg 34 of the stack of E laminations. A shunt 35 is positioned between the center leg 34 and the outer leg 36 and a similar shunt 37 is positioned between the center leg 34 and the outer leg 38. These shunts may be conventional in construction and typically each comprises a stack of laminations.

A'secondary winding 41 is positioned about the center leg 34 and a control winding 42 is positioned about portion 44 of the outer leg 38 so that the portion 44 functions as the controlled section of the core and the portion 43 of the outer leg 36 functions as the uncontrolled or saturating section of the core. In the embodiment illustrated, the leg 36 is reduced in cross-sectional area at 43 for the purpose of assuring-saturation in this section and the leg 38 is reduced in cross-sectional area at 44 for the purpose of minimizing the induced voltage in the control winding.

In the diagram of FIG. 4, the resonant capacitor 48 is connected across the secondary winding 41. The AC output volt- 7 plying holding current to SCR 53. A trigger diode 57 is connected between the junction point 58 of the integrator and the control element of the silicon-control rectifier 53. A resistor 59 is connector from the silicon-control rectifier control element to the point 50 to provide a load to ground for the trigger diode. The trigger diode 57 passes substantially zero current until voltage thereac'ross builds up to a given level, after which the diode conducts with a relatively low impedance. A typical diode would be an MPT 28.

In operation, the control rectifier 53 and the diode 57 are initially not conducting. As the output voltage of the transformer builds up, the capacitor 55 is charged through the resistor 54. When the voltage level at point 58 reaches a predetermined value, the diode 57 conducts, discharging the capacitor 55 into the control rectifier 53 and turning the con-.-

trol rectifier on. The control rectifier remains conducting until the end of the half-cycle. when it automatically turns off. The operation is repeated for every alternate half-cycle.

The resistor 54, the capacitor 55 and the diode 57 function as a relaxation oscillator which normally runs in synchronism with the line frequency under equilibrium conditions. If the output voltage of the transformer increases, the DC voltage at

points

49, 50 increases and the frequency of the relaxation oscillator increases, resulting in an earlier shorting of the control winding 42 and a reduction in the output of the transformer which in turn causes the oscillator to return to synchronism. Similarly, a reduction in the transfonner output causes a reduction in frequency of the oscillator and a later closing of the switch 53 with a subsequent increase in transformer output voltage.

A preferred embodiment of this invention is illustrated in FIGS. 11-16, and may be used with the control circuit of FIG. 9. FIG. 11 is an overall view of the complete transformer assembly which includes a magnetic core 300, a primary coil 301, a secondary coil assembly 302, shunts 305. and 306 and gaps 303 and 304 (FIG. 15). The secondary coil assembly includes a resonant winding 307, a secondary winding 308 and control winding 309 (FIGS. 13 and 16). The secondary magnetic circuit includes a controlled section 310, and uncontrolled sections made up of 311 and 312 (FIG. 16). To assure saturation in the

uncontrolled section

311 and 312, a window 315 is cut in the tongue of the uncontrolled section (FIGS. 12, 13 and 16). This window 315 reduces the cross-sectional area of the uncontrolled secondary magnetic circuit relative to the cross-sectional area of the

outer legs

314 and 313 and'the total primary core cross-sectional area of

core tongues

316 and 316a and 310.

The control winding 309 encircles the controlled section 310. The resonant winding 307 and the secondary winding 308 encircle both the-controlled core section 310 and the

uncontrolled core section

311 and 312. The control winding 309 is encircled by secondary coil 308 and the resonant coil 307 encircles both the secondary winding 308 and the control winding 309 (FIG. 16).

FIG. 15 is a view through the magnetic shunt structure.

Shunts

305 and 306 appear in plan view and have an L-shape to provide magnetic flux shunting of the primary flux existing in the primary

magnetic circuit

316 and 316a and the controlled section core tongue 310. The primary coil 301 encircles both

core tongues

310 and 316, 316a. If the cross-sectional area of tongue 310 is small relative to the

crossssectional area

316, 316a, then the

shunts

306 and 305 may vbe straight sections that only extend the length of the

primary core tongue

316, 316a.

FIG. M is a view showing the primary coil structure. The primary coil 301 encircles both the controlled section core tongue 310 and the primary core tongue 316, 3160.

The core 300 of the transformer structure is assembled from modified standard E andl laminations. The laminations in stack 320 are only encircled by the resonant winding 307 and secondary 308 winding and consist of standard B's and Is with a window 3115 cut at the back of the E. In the stack 321, space for the control winding 309 is obtained by cutting off a portion of the tongue of a standard E lamination. The stack 322 is made from the same standard E and l laminations with the center leg 310 of the E reduced in width to receive the control winding 309.

A particularly simple form of full-wave control for the transformer of FIGS. 11-16 is shown in FIG. 9. The circuit of FIG. 9 utilizes a Triac for the switch element. Since the Triac operates each half-cycle, the nominal frequency of the relaxation oscillator must be twice the transformer operating frequency.

Diodes

200, 201, 202, and 203 form a rectifier bridge which rectifies the output of the transformer. This rectified AC is applied to resistor 204 and the anode of rectifier 205. The cathode of rectifier 205 is connected to capacitor 206. This capacitor filters the rectified signal to DC. This DC in turn supplies current through resistor 208 to the Zener diode 211 which results in a constant potential appearing on the cathode of the Zener 211. The DC appearing on capacitor 206 also causes resistor 207 to supply current to capacitor 209 which charges until the firing point of the unijunction 212 is reached. At this point, the unijunction switches to a conducting state which causes the charge on capacitor 209 to be delivered to the control electrode of the Triac 213 which in turn causes the Triac to conduct, shorting the control winding.

The function of resistor 204 and Zener 210 is to supply a synchronizing pulse to the B2 electrode of unijunction 212. This is accomplished as follows. The rectified voltage out of the bridge is not filtered because rectifier 205 isolates this point from the filter capacitor 206. As a result, when the AC voltage into the bridge crosses through zero, the voltage out of the bridge drops to zero which stops the current flow in resistor 204. At this time, the voltage on the B2 electrode of the unijunction 212 is given by the Zener voltage of Zener 21! minus the Zener voltage of Zener 210. During the remainder of the half-cycle, when the voltage is high, current flows in resistor 204 which causes the Zener 204 to become forward biased. The voltage on the B2 electrode of unijunction 212 is then given by the Zener voltage of Zener 211 plus the forward voltage of Zener 210. This latter voltage appears during most of the half-cycle but during the zero crossing period of the AC, the voltage momentarily drops to the former level resulting in a negative pulse on the B2 electrode of unijunction 212.

Since the firing point of the E electrode of unijunction 2ll2 is a fixed percentage of the voltage on the B2 electrode, reducing the B2 electrode potential reduces the firing point of the E electrode. This will cause the unijunction to fire at this time if capacitor 209 is charged to a potential near the firing level when the voltage on the B2 electrode is at its high level.

In operation, resistor 207 is adjusted such that capacitor 209 will charge to the firing level of the E electrode of unijunction 212 in one-half cycle of the AC period when the DC voltage on capacitor 206 is at the desired potential. If the voltage is too high, the capacitor 209 charges faster, causing the firing angle of unijunction 212 and Triac 213 to advance which in turn reduces the voltage on capacitor 206 until equilibrium is established with the voltage on capacitor 206 at the proper potential for capacitor 209 to charge to the firing level in one-half cycle. Likewise, if the voltage on capacitor 206 is too low, the charging time of capacitor 209 lengthens which retards the firing angle until equilibrium is again established as before.

In the event the transformer is unable to supply the desired voltage, the firing angle is delayed until the following zero crossing of the AC waveform is reached. At this point the previously mentioned synchronizing pulse causes the unijunction to fire which prevents any further delay in the firing angle and the resultant loss of synchronization. A ferroresonant transformer constructed as illustrated in FIGS. l1--16 and operated with the circuit of FIG. 9 at 60 Hz. as a voltage regulator provided substantially constant output voltage (i.e., 10.3-volts variation at 166.6-volts output) over the range of no load to full load for input voltage variations in the range of 100 to 135 volts.

FIGS. 5 and 6 illustrate alternative constructions for the transformer of the invention particularly suitable for operation at higher frequencies, such as 20 kHz., and utilizing a plurality of toroids for the magnetic core structure. In FIG. 5, a primary winding 65 is linked through toroids 66, 67 and 68. A secondary winding 69 is linked through the toroids 67 and 68. A control winding 70 is linked through the toroid 68. The resonant capacitor 71 is connected across the secondary winding, and the control circuit 72 provides for closing the switch 73 across the control winding. The cores 67, 68, comprise the secondary portion of the core structure, with the core 67 corresponding to the uncontrolled section 23 of FIG. 1 and with the core 68 corresponding 'to the controlled section 24 of FIG. 1. The operation of the transformer will be the same as described in conjunction with the transformer of FIG. 1 and the transformer of FIGS. 2-4.

FIG. 6 illustrates an alternative form for the transformer of FIG. 5. The toroid 67 is replaced by two toroids 67a, 67b to provide the desired amount of magnetic material. A compensation winding 74 may be provided on one of the toroids, here the toroid 66. The compensation winding is used in the same manner as in the conventional ferroresonant transformer as described in some of the aforementioned patents. The toroids of FIGS. 5 and 6 may be formed as unitary molded elements, or pairs of C-cores, or wound ribbons, or in other forms as desired.

FIG. 7 illustrates the use of a ferroresonant transformer as shown in FIGS. 5 and 6, as the output transformer in a transistor inverter circuit. A typical inverter utilizes a pair of transistors connected in series opposition across a primary winding of an output transformer, with the AC output appearing at a secondary or load winding of the transformer. The DC power source is connected to the junction point of the transistors and to a center tap of the primary winding, and a drive circuit provides a drive current to the transistors for turning the transistors off and on. The drive circuit may be energized from a winding on the transformer providing a self oscillating inverter or the drive circuit may be energized from an external source. The two transistors operate as switches with first one and then the other being closed to connect the DC source current through the transformer primary winding alternating in the positive and negative directions to provide the AC voltage on the transformer. This basic inverter circuit is well known and several variations thereof are shown in US. Letters Pat. Nos. 2,990, 5 l9; 3,256,495; 3,317,856; and 3,405,342.

The output of an inverter is connected to the primary winding 65. The secondary winding 69 is connected to a fullwave

rectifier comprising diodes

81, 82 and a filter comprising a capacitor 83 and an inductance 84, to provide a DC output at the

terminals

85, 86. The resonant capacitor 71 is connected across a portion of the secondary winding 69.

The DC output voltage is connected to a control circuit via

lines

91, 92. The control circuit 90 includes an integrated circuit component 94 which provides a reference voltage and a DC amplifier, a threshold or trigger device in the form of an unijunction transistor 95, a switch in the form of a silicon-controlled rectifier 96, and an integrator comprising a resistor 97 and a capacitor 98.

The function of the control circuit is to vary the area under the voltage vs. time waveform of the control winding 70. The reset period occurs in the following half-cycle and will exhibit the same area as the. controlled area because a magnetic device will not support a DC unbalance. As a. result, it is only necessary to control during one-half cycle out of every cycle. Since the prime objective is to control the area, a means of controlling the firing angle as a direct function of area is used in this control circuit. The basic operation of the circuit is shown in the block diagram of FIG. 8.

The voltage on the control winding 70 is integrated with time by the

integrator

97, 98. Since the value of the integral is proportional to the area to be controlled, it is only necessary to close the switch 96 when the integral reaches a predetermined level. This triggering function is accomplished by the the switch 96 whenthe integral exceeds the preset threshold value. In order to obtain control however, it is necessary to vary the area and hence the threshold value. The threshold device has another input which controls the threshold level in proportion to the voltage applied to that input. To provide a control loop, the voltage applied to the other input of the threshold device is derived by comparing the output voltage of the transformer at

terminals

85, 86 with a reference voltage and amplifying the resulting error signal with a DC amplifier 94.

One circuit arrangement suitable for use as the control circuit 90 is illustrated in FIG. 7. A diode 100 is connected in series with the rectifier switch 96 to enhance the reverse blocking of the switch. Diode 101 is connected across the capacitor 94 of theintegrator to limit the swing during the reset period of the control winding to essentially zero so that the integrator will start at zero for the next control interval.

When the voltage on the electrode E of theunijunction ing an impedance that is dependent on load, line, frequency or other desired parameter, in conjunction with the control winding to obtain close loop control or open loop compensation.

FIG. illustrates the utilization of the transformer of this invention in combination with a saturable reactor to obtain DC load compensation. I

The circuit of FIG. 10 utilizes the transformer of FIG. 1 and corresponding elements are identified by the same reference numerals. A full-wave rectifier 225 is connected across the transistor 95 reaches a certain percentage of the voltage on the electrode B2, usually 60 to 80 percent. the device switches from essentially an open circuit between the electrodes E and B1 to a verylow impedance. This action causes the charge on capacitor 98 to be delivered to the gate or control electrode of the rectifier96, firing the rectifier into conduction. The impedance between electrode B2 and electrode B1 is substantially reduced during this period and a resistor 102 is included as a current limiting resistor.

. The integrated circuit 94 contains a voltage regulator which supplies a fixed reference voltage and contains a DC amplifier. A typical integrated circuit'may be aFairchild p. A723C. Terminals 7 and 8 are the positive input terminals, terminal 5 is the negative terminal, terminal 4 is the reference voltage output and is connected to terminal 3 which is the noninverting or reference input for the DC amplifier. Terminal 4 is used as the return point for diode 101, capacitor 98, diode 103, and rectifier 96, enabling a reverse bias to be developed on the gate of rectifier 96 and the leakage current of transistor 95 to be bled of? by resistor 104. Diode 103 serves to limit the reverse bias on the rectifier 96.

Terminal 6 of the integrated circuit 94 is the DC amplifier output and provides the varying control signal voltage for the electrode B2 of the threshold device 95. Resistor I08 func tions to prevent the voltage on electrode B2 from becoming so low that the charge from the capacitor 98 is insufficient to fire the rectifier 96. Terminal 2 of the integrated circuit 94 is the inverting input to the DC amplifier and resistors I09, 110 form a voltage divider which divides the output voltage down to an appropriate level for the amplifier. Terminal 9 is a terminal provided for frequency compensation of the DC amplifier to prevent high frequency instability. Capacitor 11] provides this compensation. Additional frequency compensation is provided by resistor 112, capacitor 113, resistor 114, and capacitor 115. Capacitor 116 serves to prevent noise from disturbing the reference voltage on terminal 4.

A ferroresonant transformer connected as illustrated in FIGS. 6-8. operated at 20 kHz. as a voltage regulator provided substantially constant output voltage (i.e., 0.06-volts variation at 30.03-volts output) at no load and at full load for input voltage variations in the range of 20 to 36 volts for a 28- volt rated input.

Thus, it is seen that the objects of the invention are achieved in providing very close control of a ferroresonant transformer utilinng flux switching and without requiring separate power supplies.

The control winding can be used in other ways. By way of example, a mechanical switch may be used to change the output voltage or the operating frequency for a fixed output voltage. By connecting a suitable frequency sensitive impedance, such as a series resonant circuit, across the control winding, frequency compensation can be obtained. The ferroresonant transformer of the invention can be used with any device havsecondary and resonant winding 13. The rectified output from the rectifier 224 is connected through one winding of a saturable reactor 226 to the load 227. The other winding of the saturable reactor 226 is connected across the winding 25.

The saturable reactor 226 offers a low impedance to the control winding 25 when it is saturated and a high impedance during unsaturated periods. Essentially, the reactor functions as a switch. To obtain load compensation, the reactor windings are interconnected in a manner that will cause the conduction period of the saturable reactor to be a maximum at no load and a minimum at full load. With this configuration of elements it is possibleto obtain a variety of load compensation characteristics.

' We claim:

1. A ferroresonant transformer including a magnetic core structure, a primary winding and a secondary winding,

wherein the core structure with the secondary winding includes two separate sections providing parallel magnetic paths for the secondary and resonant flux;

a control winding on one of said sections; and

switch means for opening and closing a circuit across said control winding, with no current source connected to said control winding whereby the only current in said control winding is self induced current resulting from flux in said one section.

2. 2. A ferroresonant transformer as defined in claim 1' wherein the core structure includes an opening therethrough, with said two sections forming opposite walls thereof.

3. A ferroresonant transformer as defined in claim 1 wherein said'core structure includes stacked E and l laminations, with said two sections comprising the outer legs of the E laminations.

4. A ferroresonant transformer as defined in claim 1 wherein said core structure includes at least two groups of stacked E and l laminations, with said two sections comprising the center legs of two different groups of said stacked laminations.

5. A ferroresonant transformer as defined in claim 1 wherein said core structure is formed of a plurality of toroids, with said two sections comprising separate toroids.

6. A ferroresonant transformer as defined in claim 1 wherein said core structure includes three sets of stacked E and l laminations assembled side by side, with the center set inverted with respect to the outer sets, and with said two sections comprising the center legs of the outer sets, respectively.

7. A ferroresonant transformer including a magnetic core structure, a primary winding and a secondary winding,

a control winding on one of said sections; and

switch means for opening and closing a circuit across said control winding as a function of the secondary voltage, with no current source connected to said control winding whereby the only current in said control winding is self induced current resulting from flux in said one section.

8. A ferroresonant transformer as defined in claim 7 wherein said switch means includes a synchronized oscillator.

9. A ferroresonant transformer as defined in claim 7 wherein said switch means includes a phase shifter.

10. A ferroresonant transformer as defined in claim 7 wherein said switch means includes a saturable reactor.

II. A ferroresonant transformer as defined in claim 7 wherein said switch means provides for switching operation every half-cycle of the supply voltage.

12. A ferroresonant transformer as defined in claim 7 wherein said switch means provides for switching operation every cycle of the supply voltage 13. A ferroresonant transformer as defined in claim 7 including rectifier means connected across the secondary winding for producing a DC control signal voltage; and

said switch means includes a switch element and a control circuit having said control signal voltage as an input for closing said switch element when said control signal voltage exceeds a predetermined magnitude.

14. A ferroresonant transformer as defined in claim 13 in which said control circuit includes an integrator connected for charging from said control winding, and a trigger unit having the integrator output and said control signal voltage as inputs with said trigger unit actuating said switch element at least once each cycle of the supply voltage connected to the primary winding.

15. A ferroresonant transfonner as defined in claim 14 in which said integrator comprises a resistor and capacitor circuit.

16. A ferroresonant transfonner as defined in claim 14 in which said integrator comprises an operational amplifier connected as an integrator.

17. A ferroresonant transformer including a magnetic core structure, a primary winding and a secondary winding,

wherein the core structure with the secondary winding includes two separate sections providing parallel magnetic paths for the secondary and resonant flux; and means for varying the saturation flux capacity in the core structure by switching the secondary and resonant flux between a first condition with flux change occurring in both of said sections and a second condition with substantially all of the fiux change occurring in one of said sections, said means for varying including a control winding on the other of said sections, and means for closing and opening a circuit across said control winding, with no current source connected to said control winding whereby the only current in said control winding is self induced current resulting from flux in said other section.

zgggg UNITED STATES PATENT OFFICE CERTIFICATE OF CORRECTION Patent No. 3,579,088 Dated May 18, 197].

Inventor(s) TAYLOR C. FLETCHER It is certified that error appears in the above-identified patent and that said Letters Patent are hereby corrected as shown below:

Column 10, line 11, "224" should read --225--.

Claim 2, line 1, "2.2." should read --2.--.

Signed and sealed this L th day of January 1 972.

(SEAL) Attest:

EDWARD M.FLETGHER,JR. ROBERT GOTTSCHQLLK Attesting Officer Acting Commissloner of Patents

Claims

1. A ferroresonant transformer including a magnetic core structure, a primary winding and a secondary winding, wherein the core structure with the secondary winding includes two separate sections providing parallel magnetic paths for the secondary and resonant flux; a control winding on one of said sections; and switch means for opening and closing a circuit across said control winding, with no current source connected to said control winding whereby the only current in said control winding is self induced current resulting from flux in said one section.

2. 2. A ferroresonant transformer as defined in claim 1 wherein the core structure includes an opening therethrough, with said two sections forming opposite walls thereof.

3. A ferroresonant transformer as defined in claim 1 wherein said core structure includes stacked E and I laminations, with said two sections comprising the outer legs of the E laminations.

4. A ferroresonant transformer as defined in claim 1 wherein said core structure includes at least two groups of stacked E and I laminations, with said two sections comprising the center legs of two different groups of said stacked laminations.

6. A ferroresonant transformer as defined in claim 1 wherein said core structure includes three sets of stacked E and I laminations assembled side by side, with the center set inverted with respect to the outer sets, and with said two sections comprising the center legs of the outer sets, respectively.

7. A ferroresonant transformer including a magnetic core structure, a primary winding and a secondary winding, wherein the core structure with the secondary winding includes two separate sections providing parallel magnetic paths for the secondary and resonant flux; a control winding on one of said sections; and switch means for opening and closing a circuit across said control winding as a function of the secondary voltage, with no currEnt source connected to said control winding whereby the only current in said control winding is self induced current resulting from flux in said one section.

11. A ferroresonant transformer as defined in claim 7 wherein said switch means provides for switching operation every half-cycle of the supply voltage.

12. A ferroresonant transformer as defined in claim 7 wherein said switch means provides for switching operation every cycle of the supply voltage.

13. A ferroresonant transformer as defined in claim 7 including rectifier means connected across the secondary winding for producing a DC control signal voltage; and said switch means includes a switch element and a control circuit having said control signal voltage as an input for closing said switch element when said control signal voltage exceeds a predetermined magnitude.

15. A ferroresonant transformer as defined in claim 14 in which said integrator comprises a resistor and capacitor circuit.

16. A ferroresonant transformer as defined in claim 14 in which said integrator comprises an operational amplifier connected as an integrator.

17. A ferroresonant transformer including a magnetic core structure, a primary winding and a secondary winding, wherein the core structure with the secondary winding includes two separate sections providing parallel magnetic paths for the secondary and resonant flux; and means for varying the saturation flux capacity in the core structure by switching the secondary and resonant flux between a first condition with flux change occurring in both of said sections and a second condition with substantially all of the flux change occurring in one of said sections, said means for varying including a control winding on the other of said sections, and means for closing and opening a circuit across said control winding, with no current source connected to said control winding whereby the only current in said control winding is self induced current resulting from flux in said other section.