WO1994014230A1

WO1994014230A1 - Series resonant converter having a three part resonant inductor

Info

Publication number: WO1994014230A1
Application number: PCT/US1993/011816
Authority: WO
Inventors: George W. Oughton; Steven R. Widener
Original assignee: Exide Electronics Corporation
Priority date: 1992-12-07
Filing date: 1993-12-07
Publication date: 1994-06-23

Abstract

Methods and apparatus for converting DC voltage (Vin) provided from a source into AC voltage are disclosed in a novel resonant converter (10). Such converter is shown to include an output circuit branch (12), wherein voltage and current in the branch are provided to a load. The output branch includes a primary inductor (22). A first circuit branch (14) is connected to the output branch for periodically causing current to pass through the output branch in a first direction. The first branch having a first inductor (24) and a first circuit portion having a first reverse recovery charge capacity. A second branch (16) is also connected to the output branch, for periodically causing current to pass in a second direction. The second branch includes a second inductor (32) and a second circuit portion having a second reverse recovery charge capacity. The first and second inductors are connected between the direct current source and the respective circuit portions. The first and second inductors operate to minimize reverse recovery currents occurring in relation to the first and second reverse recovery charge capacities.

Description

SERIES RESONANT CONVERTER HAVING A THREE PART RESONANT INDUCTOR

Field of the Invention

This invention relates to the field of power supplies and, more particularly, to resonant converters used in such power supplies wherein over-voltage conditions and time based variations in current are controlled.

The invention will be described in association with an uninterruptable power supply (UPS) system as the invention was developed for such use, however, it is to be understood that the invention is not limited to such use. The invention may be used in other power supply systems utilizing resonant converters. Background of the Invention Resonant converters are utilized in UPS systems for the conversion of direct current (DC) provided from a power source into alternating current (AC) . Such power sources can include either rectified AC or a battery. Resonant converters include either a parallel arrangement of resonant components (Parallel Resonant Converters) or a series arrangement of such components (Series Resonant Converters) . For purposes of illustration, the present invention is described in terms of series resonant converters, however, the invention is not so limited. Series resonant converters or SRCs generally can be divided into three electrical sections or portions, namely two busses and a resonant "path." The busses typically each include a bi-directional switching element, such as a

SUBSTITUTESHEET transistor connected in parallel with a diode which together operate complementarily. The resonant path will normally contain a capacitor, an inductor and the output element, such as a transformer, connected in series. Examples of such converters can be found in U.S. Patent Nos . 4,679,129 Sa akibara et al . and 4,992,919 - Lee et al . The converters depicted in those patents are half bridge converters. It is noted that application of the invention is not limited to half bridge converters, but, will also have application to full bridge converters.

In operation, out of phase driving signals applied to the base of the transistors cause the transistors to turn ON and OFF in an alternating fashion. However, it will be noted that periods exist within each "cycle" of the transistors where both transistors are OFF and one of the diodes is conducting. This selective switching of the transistors in turn causes current provided by one or more DC sources to flow in one direction through the resonant components and then in the reverse direction. As recognized in U.S. Patent No. 4,992,919 it is desirable to design a DC/AC converter which is capable of operating at high frequencies. However, switching an inductive load at high frequencies imposes high switching losses and switching stresses on semiconductor devices . Such losses and stresses are typically embodied in the generation of heat and may have an effect on component life. The frequency at which the transistors can be switched is limited by such switching losses also known as "turn-on" and "turn-off" losses. Transistor switching losses will increase as the switching frequency increases.

One form of switching loss is caused by the reverse recovery phenomena that results from the charge stored in the junction of the diode portion of the bi-directional switch. Such phenomena is generated by the forward current level on the bi-directional switch, i.e. the diodes, being suddenly reversed. In particular, the sudden attempt at reversing the

SUBSTITUTE SHEET voltage on the diodes produces a reverse recovery current (i_rr) .

In these situations, reverse recovery current (i_rr) results from a reverse recovery charge (Q_rr) existing in the junction region which charge results from the forward current level on the diode. The ability of a charge to exist in the junction region is referred to herein as a reverse recovery charge capacity. When the forward current level is reversed, the charge must be swept away before the diode will block current flow. During the time period when charge is being swept away a reverse current will flow. It is noted that diodes exhibit a charge absorbtion rate wherein a portion of the stored charge is absorbed. Any charge which is absorbed will reduce the amount of stored charge available to be swept out into the external circuit thereby causing a reverse recovery current .

In converter applications where di/dt is low, sufficient time exists in the switching process for a significant portion of the stored charge to be absorbed. Thus, little if any reverse recovery current will occur. However in converter applications where di/dt is high, little time exists in the switching process . Consequently, significant (i_rr) would occur, thereby limiting the frequency at which switching could occur. In other words, in conditions where di/dt is high, the amount of absorbed charge is small and the amount of i_rr that will occur is greatest.

Moreover, in applications involving relatively low voltage DC sources, such reverse recovery phenomena may not seem significant enough to attempt a solution, even though the existence of such current effects the operating temperature and therefore life and efficiency of a converter's electronic components. However, in high power applications, i.e., those applications where the DC source is providing several hundred Volts, the reverse recovery current is significant in relation to both component life and converter efficiency. U.S. Patent No. 4,679,129 - Sakakibara et al . discloses a number of embodiments of a series resonant

SUBSTITUTESHEET converter. The problem of reverse recovery current is not addressed. In one embodiment (Figure 7) , the resonant portion of the circuit incorporates a parallel arrangement of inductor and capacitor in series with a further inductor and capacitor. In another embodiment (Figure 8) , the parallel arrangement is incorporated into each of the busses. In a still further embodiment (Figure 10) , a series arrangement of inductor and capacitor are incorporated into each of the busses together with the diodes being directly connected to the DC source. Unfortunately, none of these arrangements will control or limit the reverse recovery current (i_rr) .

The suggestion of a parallel arrangement of inductor and capacitor in the resonant path does not limit i_rr because as switching frequency increases, increasing the frequency of the signal passing through the capacitor, the capacitor becomes more of a short circuit. Consequently, reverse recovery current remains un-controlled, i.e., the inductor has minimal effect limiting i_rr. Since switching losses are not minimized, the frequency at which such a converter can operate is limited. The inclusion of a series inductor and capacitor in a resonant path in which the diodes are directly connected to the DC source also provides no control for reverse recovery current . Since the diodes are directly connected to the DC source, voltage will remain high even though the transistors have been switched. Such an arrangement will still result in high changes in current with respect to time i.e., di/dt, and consequently reverse recovery current will remain high.

One possible solution to controlling di/dt and thereby reverse recovery current would be to add a buss inductor or make the buss inductor, if present, larger in an attempt to slow changes in current. Unfortunately, such a solution has the undesirable result of causing an over voltage to occur in the converter. An over voltage is a condition where the voltage across the switching element exceeds the voltage provided by the DC source. An over voltage condition can damage the switching elements, particularly if the over voltage

SUBSTITUTE SHEET exceeds transistor parameters, and is a condition to be avoided.

U.S. Patent No. 4,992,919 - Lee et al . discloses a series resonant converter in which a variable inductor is incorporated in the resonant path (Figure IB) . The inductance value is varied in order to maintain the so-called quality factor constant. In this way, switching when the voltage across the buss capacitor is zero can be maintained even during variations in load current. U.S. Patent No. 3,678,368 - Popp discloses an apparatus for preventing over voltage conditions in inverters and converters. An arrangement of inductors and a fuse are said to provide surge protection and over voltage protection, respectively. When an over voltage condition occurs, the fuse acts to halt current flow. Unfortunately, the converter cannot be used until the fuse has been replaced.

From the above it will be clear that a need still exists for a resonant converter which minimizes the reverse recovery current phenomena and at the same time does not result in excessive over voltages.

Summary of the Invention

The previously described problems are resolved and other advantages are achieved in a method and apparatus for converting DC voltage provided from a source into AC voltage in a novel resonant converter which includes a three part resonant inductor. Such converter includes an output circuit branch, wherein voltage and current in the branch are provided to a load. The output branch includes a primary inductor. A first circuit branch is connected to the output branch for periodically causing current to pass through the output branch in a first direction. The first branch having a first inductor and a first circuit portion having a first reverse recovery charge capacity. A second branch is also connected to the output branch, for periodically causing current to pass in a second direction. The second branch includes a second inductor and a second circuit portion having a second reverse recovery charge capacity. The first and second inductors are connected

SUBSTITUTE SHEET between the direct current source and the respective circuit portions. The first and second inductors operate to minimize reverse recovery currents occurring in relation to the first and second reverse recovery charge capacities. It is preferred for the first and second circuit portions to each include a bi¬ directional switch which in one embodiment can include a transistor and a diode connected in parallel. In such an embodiment, the reverse recovery charge capacity is a characteristic of the diodes. It is preferred for the inductance value of the first and second inductors to be approximately equal. In such an embodiment, it is especially preferred for the primary resonant inductor to have a larger inductance value. The resonant inductor is composed of three inductors. At any instant the effective resonant inductance is is the sum of the primary resonant inductor plus one of the buss inductors. Description of the Drawings

The foregoing and other objects and advantages of the invention will become more readily apparent from the following description of the preferred embodiment of the invention when taken in conjunction with the accompanying drawings which are a part hereof and wherein:

Fig. 1 is a schematic diagram showing a half bridge series resonant converter constructed in accordance with the invention;

Fig. 2 is a graph depicting various voltages and currents occurring in the converter of Fig. 1, wherein the buss inductors have values much less than the primary resonant inductor; Fig. 3 is an enlarged graph of Fig. 2 depicting additional currents occurring in select components;

Fig. 4 is a graph depicting various voltages and currents occurring in the converter of Fig. 1, wherein the one of the buss inductors has a value much less than the resonant inductor and the other buss inductor has a value equal to one half of the primary resonant inductor;

SUBSTITUTE SHEET Fig. 5 is an enlarged graph of Fig. 4 depicting additional currents occurring in select components;

Fig. 6 is a graph depicting various voltages and currents occurring in the converter of Fig. 1, wherein both of the buss inductors have inductance values approximately equal and the inductance value of the primary resonant inductor is twice as large; and

Fig. 7 is an enlarged graph of Fig. 6 depicting additional currents occurring in select components. Description of the Preferred Embodiment

A new and novel resonant converter constructed in accordance with the present invention is shown in Fig. 1 and generally designated 10. Converter 10 converts the direct current to an alternating current using a novel three part resonant inductor and provides such alternating current to a load. It will be recalled that although the invention is being described in relation to a series resonant converter, the scope of the invention is not so limited.

DC voltage (V_iπ) supplied by a source (not shown) is provided at the input to converter 10. Converter 10 includes a resonant path or output branch 12, a first circuit portion or buss 14 and a second circuit portion or buss 16. Output branch 12 includes a capacitor 18, a transformer 20 and an inductor 22, all connected in series. First circuit portion 14 includes inductor 24 and bi-directional switch 26, which is shown to include transistor 28 and diode 30 connected in parallel. Second circuit portion 16 includes inductor 32 and bi¬ directional switch 34, which is shown to include transistor 36 and diode 38 connected in parallel. It will be seen that inductors 24 and 32 are connected between bi-directional switches 26 and 34 and the input to converter 10 to which the DC power is provided.

Bi-directional switches 26 and 34 each exhibit reverse recovery charge capacity. Such capacity primarily existing in diodes 30 and 38, an explanation of such capacity having been previously given. As will be described below, inductors 24 and 32 collectively operate to minimize reverse recovery currents

SUBSTITUTESHEET occurring in relation to the reverse recovery charge capacities by limiting changes in current with respect to time, i.e., di/dt. Put another way, inductors 24 and 32 limit the speed by which the current can be reversed. A pair of input capacitors 40 and 42 are provided.

The capacitance value of each capacitor 40 and 42 is relatively larger than the capacitance value of capacitor 18. Accordingly, capacitors 40 and 42 will act as a voltage divider, dividing the input voltage between branch 14 and branch 16.

A clamp circuit 44 is shown to include capacitor 46 and diode 48. Clamp circuit 44 clamps the maximum voltage applied to switches 28 and 36. A bleed resistor 50 is provided to limit the so-called "flyback" voltage which will occur in inductors 24 and 32.

Having described the components of converter 10, consider now the operation of converter 10 under various conditions. Figs. 2-7 depict various graphs showing operating voltages and currents for various of the above described components. However, the inductance value of inductors 24 and 32 will change between certain of the graphs. In Figs. 2 and

3 the inductance value of inductors 24 and 32 is relatively small compared to the value of inductor 22. For example, inductor 22 is 2.0 μH while inductors 24 and 32 are each 0.1 μH. It will be recognized that the inductance value of inductors 22 and 32 is so small as to have little or no effect in limiting changes in current with respect to time. In Figs.

4 and 5, inductor 22 is 2.0 μH, inductor 24 is 1.0 μH, while inductor 32 remains 0.1 μH. In Figs. 6 and 7, inductors 24 and 32 are each 1.0 μH, while inductor 22 is 2.0 μH.

In operation, series resonant converter 10 receives unregulated DC voltage and provides voltage and current to the designated LOAD. As indicated previously, bi-directional switches 26 and 34 are arranged in a half bridge configuration with the switches being turned fully ON and OFF alternately via suitable control circuitry (not shown) . The circuitry necessary to alternately control switches 26 and 34 is known.

SUB Although such circuitry does not constitute the invention, a general description of the control of such switches will be given.

Pulses of substantially constant width are applied by control circuitry to base of transistors 28 and 36 to turn them ON. The frequency at which these pulses are applied determines the level of the resonant current IR which is coupled by the transformer 20 to the LOAD. As the load requirements increase, the operating frequency of the resonant converter 10 will also be increased by the control circuitry such that a constant output voltage is maintained. As transistors 28 and 36 switch, current is caused to flow through output branch 12 in a first direction and then in a second direction resulting in the provision of an AC voltage to the LOAD. Consider first the voltage and current waveforms associated with Fig. 2. V_in is shown as remaining constant. The voltage across capacitor 18 is shown as a sinusoidal waveform. The voltage across transistor 36 is depicted in a square waveform-type pattern. When transistor 36 turns ON, the voltage drops from about 170 volts to zero. At point 52, transistor 36 turns OFF. The voltage across transistor 36 however, remains zero because current, having already been reversed, is now flowing through diode 38. It is noted that at this point both transistors 28 and 36 are OFF. At point 54 transistor 28 turns ON and the voltage across transistor 36 begins to rise. At the point where the voltage across capacitor 18 and coincidently transistor 36 reaches its peak, the resonant current through inductor 22 reverses. At point 56 transistor 28 turns OFF. The voltage across transistor 36 however, remains approximately constant because current, having already been reversed, is now flowing through diode 30. It is noted again that at this point both transistors 28 and 36 are OFF. It is noted that at point 56 a slight over voltage exists, i.e., the voltage across transistor 36 is greater than V_in. The cycle of turn ON and OFF thereafter repeats . In order to understand the reverse recovery phenomena, reference is now made to Fig. 3.

SUBSTITUTE SHEET Fig. 3 is an enlarged view, i.e. , a time expanded view of Fig. 2. Fig. 3 again shows the voltage across transistor 36, however the current flowing through diode 30 and the current flowing through transistor 36 are also shown. The cross-hatched region designated 58 represents the reverse recovery current charge (Q_rr) occurring in this embodiment. The size of region 58 is a measure of the reverse recovery losses that will occur. The existence of Q_rr is shown to cause a corresponding spike like current region 60 for transistor 36. The area of region 60 should match the area of region 58. Such current spikes effect transistor efficiency and thus are to be avoided. As indicated previously, one solution to reduce such currents would be to make inductor 22 larger. The problem with adding inductance to converter 10 is the causation of an over voltage condition. In Fig. 4,' buss inductor 24 has instead been increased to 1.0 μH.

As shown in Fig. 4, increasing inductor 24 has resulted in an over voltage at 56, which in turn means that the safety margin associated with transistor 36 has been reduced. However, in order to appreciate whether increasing the value of inductor 24 has reduced the reverse recovery losses or the region indicative of Q_rτ consider Fig. 5. As shown in Fig. 5, although the reverse recovery losses have been reduced, they still exist. It is also noted that the effect of increasing inductor 24 is to slow down the rate at which current in diode 30 reverses. In other words, di/dt has been lowered, thereby allowing more of the charge in diode 30 to be absorbed, in turn reducing the amount of charge to be swept away when the current level is reversed. It has been discovered that if inductor 32 is increased to the value of inductor 24, reverse recovery losses can be significantly reduced. In other words, inductors 24 and 32 are made symmetrical. In such an embodiment the extra inductance used to limit the change in current with respect to time is split between buss 14 and buss 16. Such a condition exists for the waveforms of Fig. 6.

SUBSTITUTE SHEET As shown in Fig. 6, even though the inductance of inductor 32 has been increased to match the inductance of inductor 24, the over voltage at 56 has remained relatively the same. However, Fig. 7, clearly shows that region 58, the measure of Q_rr, has been significantly reduced and the corresponding current region 60 in transistor 36 has also been reduced. Consequently, for the same amount of over voltage di/dt is cut in half, resulting in lower reverse recovery current and reduced reverse recovery losses. Conversley, the amounts of over voltage could be cut in half without increasing di/dt and reverse recovery losses buy splitting the single dominant inductor 24 as depicted in Fig. 4 into two equal parts inductor 24 and 32. The amount of overvoltage at point 56 would be cut in half without increasing reverse recovery losses in region 58 of Fig. 5.

It will be appreciated from the above that the resonant inductor is effectively composed of three inductors. At any instant the effective resonant inductance is is the sum of the primary resonant inductor 22 plus one of the buss inductors 24 or 32.

While the invention has been described in detail herein in accord with certain embodiments thereof, many modifications and changes therein may be effected by those skilled in the art. As for example, operation without an isolation transformer, with other power switching devices and in systems other than UPS systems, are contemplated by the applicants. Accordingly, it is intended by the appended claims to cover all such modifications and changes as fall within the true spirit and scope of the invention.

SUBSTITUTE SHEET

Claims

ClaimsWhat is claimed is:

1. A resonant converter, said converter being connected to a direct current source, for converting said direct current to an alternating current and for providing said alternating current to a load, said converter comprising: an output circuit branch, wherein voltage present in said branch and current passing through said branch are provided to said load, said output branch having a primary inductor; a first circuit branch, connected to said output branch, for periodically causing current to pass through said output branch in a first direction, said first branch having a first inductor and a first circuit portion having a first reverse recovery charge capacity, said inductor being connected between said direct current source and said first circuit portion; and a second branch, connected to said output branch, for periodically causing current to pass through said output branch in a second direction, said second branch having a second inductor and a second circuit portion having a second reverse recovery charge capacity, said inductor being connected between said direct current source and said second circuit portion, wherein said first and second inductors operate to minimize reverse recovery currents occurring in relation to said first and second reverse recovery charge capacities .

2. The resonant converter of claim 1, wherein said first circuit portion comprises a transistor and diode connected in parallel.

3. The resonant converter of claim 1, wherein said second circuit portion comprises a transistor and diode connected in parallel .

4. The resonant converter of claim 1, wherein the inductance value of said inductors are substantially equal .

SUBSTITUTE SHEET

5. The resonant converter of claim 1, further comprising first and second capacitors, wherein said first capacitor is connected between said first inductor and said output circuit branch and said second capacitor is connected between said second inductor and said output circuit branch.

6. The resonant converter of claim 5, said output circuit branch further comprises a resonant capacitor and wherein the capacitance value of each of said first and second capacitors is significantly larger than the capacitance value of said resonant capacitor.

7. The resonant converter of claim 1, further comprising a clamp for clamping the voltage applied to said bi¬ directional switches.

8. A method for converting direct current provided by a source to alternating current and for providing said alternating current to a load, said method comprising the steps of: providing an output branch, wherein voltage present in said branch and current passing through said branch are provided to said load and providing said output branch with an inductor; providing first and second bi-directional switches connected to said output branch, said first and second bi¬ directional switches having first and second reverse recovery charge capacity, respectively; periodically causing current to pass through said output branch in a first direction by controlling said first bi-directional switch; controlling the change in current with respect to time flowing through said first bi-directional switch; periodically causing current to pass through said output branch in a second direction by controlling said second bi-directional switch; and controlling the change in current with respect to time flowing through said second bi-directional switch, wherein reverse recovery currents occurring in relation to said first and second reverse recovery charge capacities are reduced.

SUBSTITUTE SHEET

9. The method of claim 8, wherein said step of controlling the current flowing through said first bi¬ directional switch comprises the step of connecting a first inductor to said first bi-directional switch.

10. The method of claim 9, wherein said step of controlling the current flowing through said second bi¬ directional switch comprises the step of connecting a second inductor to said second bi-directional switch.

11. The method of claim 10, wherein said steps of connecting first and second inductors comprises connecting inductors having substantially equal inductance values.

12. The method of claim 8, wherein the step of providing first and second bi-directional switches comprises the step of providing first and second transistors and first and second diodes, wherein each of said first and second transistors are connected in parallel with said first and second diodes.

SUBSTITUTE SHEET