WO1992005623A1

WO1992005623A1 - Variable transformer switching power supply

Info

Publication number: WO1992005623A1
Application number: PCT/US1991/006598
Authority: WO
Inventors: Michael A. Knights
Original assignee: Zytec Corporation
Priority date: 1990-09-13
Filing date: 1991-09-12
Publication date: 1992-04-02

Abstract

A power supply having a variable transformer which includes primary, secondary and control windings, wound onto separate power and control cores. The primary winding is excited by a pulsatile current drawn from a bulk supply. The secondary is connected to a load. The control winding has a current flowing therein which reflects the magnitude of the load. The control winding regulates the inductance of the control core and determines the power delivered to the secondary winding.

Description

VARIABLE TRANSFORMER SWITCHING POWER SUPPLY

BACKGROUND OF THE INVENTION

Field of the Invention

The present invention relates to a switching power supply and more particularly to a single-ended, forward converter type, switching power supply which incorporates a variable transformer.

The inductance of the variable transformer primary winding, comprises a "controlled" inductance associated with a load sensitive feedback control circuit, and a "regulated" inductance associated with a secondary winding, used to deliver power to a load.

In operation, power variations in the load are accommodated by feedback variation of the controlled inductance.

Two methods of varying the controlled inductance of the variable transformer are taught.

Description of the Prior Art In general, switching-type power supplies convert bulk DC electrical power into a regulated DC voltage level, for supply to a load. Typically, bulk DC supply is switched by a high frequency oscillator to supply pulsatile current to the primary winding of a relatively small supply transformer. The supply transformer secondary winding may be connected in series with a so called, magnetic amplifier (mag amp) or saturable reactor to regulate power output. Other linear and switching types of regulators are in common use.

With respect to the mag amp type of supply, the mag amp serves as a controlled switch to regulate the current flow in the secondary winding circuit. A feedback circuit monitors the load, and generates a control signal which is used to regulate the "on" time and the "off" time of the mag amp. This operation results in pulse width modulation of the secondary current to accommodate variation in the power demands of the load. This pulsed current is filtered through an LC filter, prior to being supplied to the load.

In this prior art, it has been common to use mag amp switches or magnetic devices which have only one winding coupling to the dipoles in the magnetic material. In general, the core is biased away from saturation by the control circuitry. In operation, the reset level of the mag amp regulates power delivery to the load.

An article entitled "Pulsewidth Modulation Derived Through Control of Flux Distribution in a Dual- Core Transformer" authored by Fred V. Kadri, and appearing in the proceedings of the 1971 Intermag Conference, teaches a dual core transformer of the same general type presented herein. However, the Kadri power supply utilizes a push-pull circuit for primary excitation, and requires the use of timers and attendant SCRs for his pulsewidth modulation of the power supply. Thus the mode of operation of the power supply differs from the present invention. SUMMARY OF THE INVENTION

In contrast to the traditional mag amp based switching power supplies of the prior art, the present invention eliminates the need for a mag amp and provides a variable transformer, regulated by the load, to transfer power to the load.

The variable transformer preferably has two separate and distinct ferro-magnetic cores. These are referred to as the "power core" and the "control core". The transformer primary winding couples to dipoles in both the power core and the control core.

The power core carries a secondary winding directly coupling to the dipoles in the power core, while the control core carries a control winding, directly coupling to the dipoles in the control core.

As a result of this topology, the primary winding inductance comprises the control core inductance and the power core inductance, in series. Also, as a result of this configuration, the voltage appearing across the secondary winding depends on the magnetic state of the control core. The magnetic state of the control core, in turn, depends on the current permitted to flow in the control winding.

In the context of the switching power supply the secondary winding, on the power core, is connected to a load, while the control winding, on the control core, is connected to a feedback control circuit which monitors the load. In operation, the volt-seconds of excitation energy supplied to the primary winding by the bulk supply, are split between the control inductance and the power inductance, and therefore appear across both the secondary winding and the control winding. As a consequence of this configuration, it is possible to control the voltage appearing across the secondary winding by varying the magnitude of the inductance of the control core.

Two illustrative methods for electrically varying the control core inductance are taught in the context of the switching power supply. The first mode is referred to as the "flyback mode" while the second is referred to as the "forward mode".

In another aspect of the invention the control winding circuit uses a MOSFET to provide substantially linear control of the control circuit. The preferred MOSFET is in series with a choke. The choke severs to greatly diminish power dissipation in the secondary by limiting the amount of time that both current and voltage are applied to the MOSFET.

In yet another aspect of the invention, the preferred control methodology is extended to a multiple output power supply.

DESCRIPTION OF THE DRAWING An illustrative embodiment of the invention is shown in several of the figures and identical reference numerals refer to identical structure throughout the several views, in which:

FIG. 1 is an electrical schematic diagram of an illustrative variable transformer switching power supply depicting the "flyback control mode"; FIG. 2 is a mechanical schematic diagram depicting the illustrative and preferred topology for the variable transformer;

FIG. 3 is a waveform diagram depicting circuit operation of a variable transformer switching power supply regulated on the "flyback" principle;

FIG. 4 is a diagram depicting the magnetization level of the control core at various output power levels;

FIG. 5 is a hysteresis diagram depicting operation in the flyback mode;

FIG.6 is a performance diagram depicting the control range for the illustrative variable transformer;

FIG.7 is a waveform diagram depicting circuit operation of a variable transformer switching power supply regulated on the "forward" principle;

FIG.8 is an electrical schematic diagram depicting a preferred power supply module; and

FIG.9 is an electrical schematic diagram depicting a preferred multiple output power supply, based upon the power supply module of FIG.8.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT In the following detailed description of the preferred embodiment, reference is made to an illustrative embodiment of the invention. It is to be understood that other embodiments may be used without departing from the scope of the present invention. Turning to FIG.l, there is shown a switching type power supply 14, for supplying power from a bulk supply connected at terminals 10 and 11, to a load 12 connected across terminals 13 and 15. The switching type power supply 14 converts the unregulated bulk supply power to a regulated constant voltage or constant current source for the load 12. As the load value varies, a conventional regulator circuit 30 provides feedback control information to the variable transformer 16.

As the value of the load rises, more power is transferred from the primary winding 17, to the secondary winding 18. Conversely as the load is reduced, less power is transferred from the primary to the secondary winding. The amount of power transferred is determined by the timing of voltage appearing across the secondary winding 18 which is in turn a function of the inductance of the control core 20.

The value of control core inductance is in turn, a function of the amount of current flowing in the control winding 19.

A further understanding of the operation of the variable transformer and the attendant power supply requires consideration of the topology of the device and several underlying physical principles.

As shown in FIG.2, the preferred variable transformer has two separate and distinct magnetic cores. The power core 21 provides a closed magnetic path for power core flux 23. The control core 20 provides a closed magnetic path for control core flux

22. The cores maybe of identical size and shape as shown or they may differ in size or in other properties.

An important property of the core geometry is that the flux path is closed and lies entirely within the core material. The closure of the core, provides a high permeability path for flux within the core. When the flux path is complete or closed, only a small current is required to align the dipoles and saturate the core.

Gaps in the flux path are to be avoided since the low permeability of an air gap reduces the ability of the windings to influence the effective permeability of the total path and thus limits the ability of a winding to control the inductance exhibited by the winding. Powder technology ferrite toroidal cores are very good for this application because of the absence of air gaps, however pairs of U-cores can be stacked to form acceptable cores as shown in the figure.

The primary winding 17 couples strongly to the magnetic dipoles located in both the power core 21 and the control core 20. This result is preferably achieved by winding the primary winding 17 on a bobbin, which, when assembled, passes the primary winding through the apertures of both of the assembled U-cores.

The control winding 19 is wound only on the control core 20, while the secondary winding 18 is wound only on the power core 21. This configuration causes the control winding to couple strongly to the dipoles in the control core, but only very weakly to the dipoles in the power core, which is desirable. In a similar fashion, the secondary winding 18 couples strongly to the dipoles in the power core 21 and couples very weakly to the dipoles in the control core 20.

The lowest leakage currents occur when the control winding and the secondary winding are placed under the primary winding as depicted in FIG.2. The mobility or lack of mobility of the dipoles in the control core determines the effective permeability of the core, which in turn determines the inductance of the control core.

Therefore, in summary, the core structure and windings are configured to maximize, the coupling between the primary winding and dipoles in both the power flux path and the control flux path; and to minimize the direct interaction between dipoles in the control flux path and the power flux path. Although unitary core structures formed from alternating E and I cores can partially achieve these design goals, the siameseing of the control flux path and power flux path in such a device tends to limit the control range of a switching type power supply operating on these principles.

An illustrative device used to collect the performance data described herein, may be made from pairs of suitable U-cores such as Phillips 376U 250 cores formed from suitable ferromagnetic material such as 3C8 material, for each of the cores. The illustrative transformer used a 42 T primary winding, wound with #23 gage wire; a 4T control winding wound with #23 gage wire; a 4T secondary of #11 gage Litz wire. Pairs of E-cores were assembled into a "square" toruses 20 and 21, and held together by tape.

In general, the inductance of the control core 20 is low when the mobility of the magnetic dipoles in the control flux path 22 are restricted.

The dipoles in this control core 20 may be immobilized by shorting the control winding.

The dipoles may also be immobilized by saturating the control core.

In the absence of an applied magnetic field the core is in a zero remanance state and the dipole alignment is random as determined by the underlying crystal structure of the material. In this resting state, the material is unsaturated and exhibits high permeability.

However, if current is flowing in the control winding the dipoles align with the applied field and are immobilized. This flow of current in the winding results in saturation of the control core when essentially all the dipoles are aligned with the applied field. This corresponds to a low permeability condition and, is reflected as low control core inductance.

The dipoles can be immobilized by shorting the control winding. An applied field which attempts align the dipoles will be resisted by an opposing induced field generated by current flow in the control winding. With near zero resistance in the control winding the opposing field equals the applied field and the dipoles are substantially immobilized. This too results in low, effective permeability of the core and causes the control core to exhibit low inductance.

Consequently the control winding couples to the dipoles in the control core and can influence their magnetic state. The magnetic state of the dipoles in turn determines the inductance exhibited by the control core.

It is important to recognize that the topology causes the primary winding to also couple to the dipoles in the control core. Therefore, a portion of the inductance of the primary winding reflects the magnetization state of the control core which is strongly affected by currents in the control winding.

In general, when the control winding immobilizes the control core dipoles, the control core contribution to the total primary inductance diminishes and more of the applied excitation voltage appears across the "other" inductance, which is the power core winding 18.

The "forward" control mode treats the control winding resistance as an auxiliary secondary load. Voltage is induced on this control winding and power is dissipated in the control resistance 27. The degree of dipole immobilization is inversely proportional to the value of the resistance. This control methodology is not preferred because of the significant power dissipation in the control resistance 27.

The "flyback" control mode relies on regulating the degree of control core saturation. A large induced current from the primary, drives the core towards saturation. The regulator circuit 30 controls the degree to which the control core comes out of saturation. When the control core operates near total saturation the dipoles are substantially immobilized and the control core inductance is low causing more of the excitation voltage to appear across the secondary winding.

These control methods are better understood in connection with the idealized waveform diagrams of FIG.3 and FIG.7.

FIG. 3 depicts voltages and currents developed in a "flyback" controlled variable transformer power supply operating at a two different loads. Panel A depicts the voltage present across the primary winding 17 of the variable transformer 16 taken at point A in FIG 1. This pulsatile waveform results from the periodic activation of the switch elements 24 and 25. As is conventional in this art, the switches are usually MOS transistors operated at a frequency ranging from approximately 100 KHz. to approximately 200 KHz. This switching frequency is usually fixed and is supplied by a conventional oscillator (not shown) . Portion 31 of the waveform of panel A corresponds to the open switch condition and portion 32 corresponds to the closed switch condition.

As shown in FIG. 3 the D3 diode 26 is phased to conduct when the switches 24, 25 are open. Consequently there is a current which will flow in the control winding during the off time of the switches 24 and 25. The panel C waveform depicts this current as measured at point B in FIG 1. When the value of the control resistance 27 is large, a large voltage can develop across the resistance to more fully reset the core 20. The initial magnitude of this reset voltage is depicted as portion 35 of Panel E. The dipoles are more fully reset when the resistance is high and a larger average reset voltage is permitted to develop.

When the control resistance is low, little voltage is induced across the resistance and the control core can not fully reset. The magnitude of the voltage determines the rate of change of current depicted as the slope of waveform portion 34.

Panel B depicts the current flowing in the primary winding 17 and the secondary winding 18 at two di ferent loads. The wave forms in these two windings are similar in shape. Generally, current rises during the on time of the switches 24 and 25. The current in the secondary builds through diode 33 until it exceeds the flywheel current from the choke 36 flowing through diode 37. Beyond this point in the waveform, the current flow in the secondary is limited by the choke 36. The inflection or switch point 38 of panel B depicts the point at which current through diode 33 exceeds the current flow through diode 37. As the switches open eliminating current flow in the primary, the current in the secondary drops to zero as diode 33 is reverse biased. Panel D depicts secondary voltage supplied to the LC filter formed by choke 36 and capacitor 40. The voltage appears across the secondary winding only when the flywheel diode 37 is reversed biased. Consequently, the pulse width of the secondary waveform is determined by the time that is required to reach the switch point 38. Which, for moderate and high level depends upon the time at which the control core reaches the low inductance state. Turning to FIG.4 there is shown magnetization diagrams which relate the control core magnetization level with the power delivered from the secondary winding 18 to the load 12. Panel A of FIG. 4 corresponds to low power delivery from the secondary 18. In this instance the control core inductance must be high to force the majority of the primary winding volt- seconds to appear across the control winding 18. To maintain a high inductance state the control core must exhibit high effective permeability. This is achieved by fully resetting the core to the remanance state (Br) . During the on time of the switches, the induced field drives the control core toward saturation. However since the core was fully reset, saturation (Bs) is not reached. At this power level the current required to reset the control core to the remanance level, is supplied primarily through the Dl diode 28 and the D2 diode 29, while the switches 24 and 25 are open. A minimal contribution to the reset current is supplied through the control resistance 27 because of its high value.

At a medium power level as depicted in Panel B of FIG. 4 the control core inductance is reduced to permit more voltage to appear across the secondary winding 18. More dipoles in the control core are immobilized by retaining the core closer to saturation. This effect is achieved by adjusting the control core resistance to an intermediate level which does not permit as much voltage to develop across it, thus limiting the magnitude of the change, of the resetting current. In this instance the current to reset the core is supplied through The diodes 28 and 29 and through the control resistance 27.

Delivery of maximum power requires very low control core inductance which is achieved by immobilizing most of the dipoles in the control core through saturation. In this instance the value of the control resistance is low and consequently not much voltage can develop across it. The resulting minimal change in reset current is not sufficient to move the dipoles very far out of saturation. The resulting magnetic state lowers the effective permeability of the control core forcing the majority of the volt-seconds of excitation voltage to appear across the secondary winding 18, to supply the load with maximum power. FIG. 5, is a hysteresis diagram which depicts the relationship between the control core and power core saturation at low and high power for the flyback control mode power supply. In general, when the transformer is delivering maximum power the dashed loops 41 and 42 apply. When the transformer is delivering lower power the loops 43 and 44 apply. Point 45 corresponds to the reset point for the control core during high power delivery. Note that the control core operates into the saturation portion of the curve. In this case the dipoles of the control core spend most of the cycle time immobilized. The corresponding low permeability state or low inductance state allows the transfer of primary volt-seconds to the secondary winding to occur more quickly or earlier in the cycle. The secondary core does not operate into saturation, which is desirable for transformer action.

FIG. 6 depicts the broad control range achieved by the illustrative power supply described herein. The minimum current lies in the tens of milliamperes with the maximum current on the order of tens on amperes. The current gain of the device is also quite large with modest control currents on the order of one tenth of the maximum delivered current. FIG. 7 depicts circuit waveforms for the

"forward" control mode. The circuit of FIG. 1 can be altered to operate in the forward control mode by reversing the polarity of diode 26 or by reversing the phase of winding 19. In either case the current in the control winding flows when the switches are closed or in the "on" state. The current in the primary and secondary is depicted on panel B of the figure. In this control configuration the current in the control winding appears as another load on the primary. As a consequence this control mode is not as suitable for switching type supplies.

FIG. 8 depicts a preferred form of the power supply were the controlled resistance 27 is made up form a MOS transistor 46 in series with a choke 47. In this illustrative implementation, the MOS transistor 46 acts as a substantially linear control element controlling control current in the control winding 19 based upon the drive voltage supplied from the regulator circuit 30. A suitable choke 47 for an illustrative supply may be formed from a .6 "toroidal" core with approximately 5 turns of wire. In operation, in the flyback control mode, the choke significantly limits power dissipation in the control secondary 19 circuit. The transistor 46 control the reset point 45 for the secondary core, and although the transistor needs to operate in its linear region it need not conduct continuously for the whole cycle. Therefore, the choke is provided to limit the time period that both a voltage appears across the transistor and current is flowing in the secondary winding. In operation, the voltage appearing across the secondary winding 19 causes a small current t flow in the winding while the choke 47 becomes magnetized. While the reactance of the choke is high the er is essentially no power dissipated in the transistor 46. Only during the brief period of time that the choke becomes saturated, is power dissipated in the transistor 46. This is especially important in multiple supply distribution systems of the type shown in FIG. 9.

FIG. 9 shows a multiple output power supply which has a first output at terminals 48 and 49 from a first power supply module 54. Second power output at terminals are provided at 50 and 51 for a second power supply module 55. Bulk D.C. power is supplied to the multiple power supply modules via terminals 52 and 53, forming the bulk supply buss, while additional power supply modules may be added at terminal 56 and 57, which represent a switched supply buss. It is preferred that the switched supply buss be generated by a pair of synchronously switched switches 62 and 63 which are switched at high frequency by a conventional oscillator.

Each of the power supply modules 54 and 55 follow the general architecture of the supply presented in the drawing of FIG. 8. However, an isolation diode 58 or 59 should be provided to in series with each primary winding, for the supply modules, 60 or 61 to ensure isolation of primary currents.

Claims

WHAT IS CLAIMED IS:

1. A switching type power supply for supplying power to a load comprising: power core means for supporting flux in a power transfer magnetic path; control core means for supporting flux in a control magnetic path; a primary winding coupled directly to both first and said second core means; a control winding coupled directly to said control core means; a secondary winding coupled directly to said power core means; bulk power source means coupled to said primary winding, for providing a source of excitation pulses to said primary winding; load filter means coupled to said secondary winding for supplying regulated power to said load; feedback control means, coupled to said load for monitoring said load and for flyback control of said control core as a function of said load.

2. A switching type power supply for supplying power to a load comprising: power core means for supporting flux in a power transfer magnetic path; control core means for supporting flux in a control magnetic path; a primary winding coupled directly to both first and said second core means; a control winding coupled directly to said control core means; a secondary winding coupled directly to said power core means; bulk power source means coupled to said primary winding, for providing a source of excitation pulses to said primary winding; load filter means coupled to said secondary winding for supplying regulated power to said load; feedback control means, coupled to said load for monitoring said load and for forward control of said control core as a function of said load.

3. A switching power supply for supplying power to a load comprising: power core means for supporting flux in a power transfer magnetic path; control core means for supporting flux in a control magnetic path; a primary winding coupled directly to both first and said second core means; a control winding coupled directly to said control core means; a secondary winding coupled directly to said power core means; bulk power source means coupled to said primary winding, for providing a source of excitation pulses to said primary winding; said load being coupled to said secondary winding for receiving power from said primary winding; flyback mode control means, coupled to said control winding and coupled to said load for monitoring the power delivered to said load and for controlling the reset point for said control core as a function of said load.

4. The power supply of claim 3 wherein said flyback mode control means comprises: regulator circuit means for generating a control signal representative of the value of said load; linear control means in series with said secondary winding, coupled to said regulator circuit means for controlling the reset of said control core; reactance means in series with said secondary winding for reducing control winding power dissipation.

5. A multiple output power supply comprising: bulk supply means for delivering bulk power; switcher means coupled to said bulk supply means for generating pulsatile excitation pulses to a switcher supply buss; a plurality of supply modules coupled to said switcher supply buss, each of said supply modules for providing regulated power output.