ZA200304926B - Synchronous rectifier circuit. - Google Patents

Description

Minebea Co., Ltd.

M30183PCT(L)

A sychronous rectifier circuit

The invention relates to a sychronous rectifier circuit for use in a push-pull d.c. voltage con- verter.

It is the task of d.c. voltage converters to transform a direct voltage applied at their input into a direct voltage which is output at a different value, and to do so most efficiently. The output value may be greater or smaller than the input value and may be adjustable. In a d.c. voltage converter the direct voltage input, first, is transformed with the help of switching stages into alternating voltage having a rectangular waveshape. During the switch-on period "chopped" direct voltage is stored in the form of magnetic energy in a choke. During the switch-off pe- riod, it acts as a self-induction voltage at the output of the d.c. voltage converter. D.C. voltage converters operating according to this principle are referred to as choke converters. They have their inputs and outputs galvanically separated. It is known in the art to use transformers to achieve separation of potential. Here, the induced voltage occurs at the secondary winding, and the voltage transformation can be determined by the number of windings. Fig. 1 is a block diagram illustrating a transformer-type converter used as d.c. voltage converter.

Fig. 1 depicts the basic elements of a transformer-type converter, including an input switching stage 10, a power transformer 12, a rectifier circuit 14, and an output filter 16. A distinction is made in the art between single phase d.c. converters and push-pull voltage transformers. A single phase d.c. converter may be regarded as being a simple electronically controlled switch, whereas switch-over operations occur with push-pull voltage transformers, and a transformer 12 having two primary windings may be required. Push-pull transformers can be derived from two single phase d.c. converters connected in parallel. The electronic switch- over is accomplished by two switching stages, and current always flows through one of the two primary windings. The invention relates to the field of push-pull voltage transformers.

In practice, such push-pull voltage transformers are employed in switch mode power supply, such as server architecture for telecommunications applications, in PCs, industrial applica- tions, and many other situations. The invention is especially advantageous in distributed en- ergy supply systems where several stages are connected in succession. In new server archi-

0. tectures, for instance, power supply units are used in which the mains voltage, to begin with, 1s converted into a bus voltage of some 48 to 50 V. A second conversion to +12 V, for exam- ple, then takes place within the server sub-system. The specific voltages required for the vari- ous components, such as the microprocessor, RAMs, etc. are produced locally by so-called voltage regulator modules which are connected to the 12 V rail.

Each of the converting stages must operate as efficiently as possible in view of the series con- nection of different power stages. For maximum efficiency to be obtained with switch mode power supply, optimization in terms of energy loss is required regarding each and every source thereof. Energy losses depend not only on the type of converter or transformer chosen, be it a single phase d.c. converter or a push-pull voltage transformer, but are determined deci- sively by the mode of operation of the rectifier circuit.

Making good use of the inductivities as well as driving the transducer in positive and negative directions present great advantages of push-pull voltage transformers. Another advantage is the great efficiency of the transformer and the high output power attained.

Fig. 2 is a schematic circuit diagram of a push-pull voltage transformer provided with Schot- tky diodes for rectification at the secondary side. The push-pull forward converter illustrated in fig. 2 is known; its function is described, for example, in Billings, Keith "Switch Mode

Power Supply Handbook", 2nd edition, McGraw-Hill, New York, 1999. This converter com- prises a power transformer 18 having a primary side 18a and a secondary side 18b. The pri- mary side 18a and the secondary side 18b each comprise two winding sections. For driving purposes, two power transistors 20, 22 are associated with the two winding sections at the primary side 18a. Two secondary diodes 24 and 26 are associated, respectively, with the two winding sections at the secondary sidc 18b. These diodes are connected to an output filter stage made up of a storage choke 28 and a storage capacitor 30, as may be seen in fig. 2. The power transistors 20, 22 are driven, for instance, by a control IC (not shown).

As the transistor 22 is driven, current will flow through the associated winding section of the power transformer 18 and also through the transistor 22. The polarity of the associated wind- ing section at the secondary side 18b of the power transistor 18 causes the diode 26 to cut off,

At the same time, voltage 1s induced also in the other winding section at the secondary side 18b, thus causing current to flow through the diode 24 via the storage choke 28. When a suf-

ficient amount of energy has been transmitted from the primary side to the secondary side, the transistor 22 blocks. During the next cycle the transistor 20 is driven. The current now flow- ing through the second winding section of the primary side 18a causes reversal of the polarity of the associated winding at the secondary side 18b. Diode 24 blocks, while diode 26 is con- ducting, thus permitting current to flow through the choke 28, as during the first cycle. To make sure that the two power transistors 20, 22 will not be conducting at the same time, a compulsory break, the so-called freewheeling phase, is provided between the first and second cycles described above. During this freewheeling phase, the electric circuit at the secondary side 18b of the transformer 18 is formed of the storage choke 28, the storage capacitor 30, the two conducting diodes 24 and 26, and the connected load (not shown in the drawing).

Fig. 3 presents idealized waveshapes of the output voltages ug, ug; at the two winding sec- tions of the secondary side 18b of the power transformer 18, forward currents ig;, ig; through the diodes 24 and 26, and the output current iy through the storage choke 28.

A positive voltage ug; is generated during the first time interval from t; to t;. Diode 24 is con- ducting. The overall output current ig is passed through the same and through the upper sec- ondary winding section of the power transformer 18 toward the output. The rise in output cur- rent is determined by the voltage difference ug;-ug; (output voltage) and the sum of the induc- tivities of the secondary circuit.

The second time interval from t; to t3 corresponds to the so-called freewheeling phase. The output voltages ug; and ug; of the transformer 18 are zero. The current ig is determined by the inductivities of the secondary circuit. If the upper and lower winding sections at the secondary side 18b are identical the output current ig is divided in two. Each of the diodes 24, 26 will carry one half of the output current iy. During this time interval the output current drops.

During the third time interval from t3 to t4 a positive voltage ug is produced and diode 26 is conducting. The resulting behavior corresponds to that of the first time interval.

During the last time interval from t4 to ts of period T both power transistors 20, 22 are turned off. The voltages ug, and uy, once again are zero, which corresponds to the freewheeling phase.

In the embodiment of the d.c. voltage converter shown in fig. 2 the secondary rectifier is em- bodied by diodes. The rectifier diodes produce losses which depend on the forward voltage of the diodes 24, 26 and are composed of forward losses and switching losses of the diodes.

The forward loss Ppc of a diode is given by the product of its forward voltage drop uf and its forward current ip (see also fig. 4)

Ppc = Uf 1p.

The forward voltage rises as the load increases; it lies between 0.5 V and 1.5 V, depending on the type of diode provided. If the transducer output voltage is 3.3 V, for example, which would correspond to a processor voltage, as much as 30 % of the voltage will drop at the rec- tifier diodes. With higher transducer output voltages, e.g. 48 V in telecommunications appli- cations, the voltage drop at the diodes is comparatively less, but still not negligible.

The switching loss of a diode can be estimated by the following equation:

Pps =Qf-0-f where QF is the recovered load during the fall time of the reverse current of the diode, f is the reciprocal value of period T, and - is the peak value of the diode turnover voltage.

A reduction of the forward loss discussed above can be achieved only by reducing the voltage drop.

One solution resides in the use of a MOSFET connected in parallel with the diode. That is shown in fig. 4. The MOSFET is turned on when current in forward direction is applied to the diode, and it is turned off when the current is reversed. This is called synchronous rectifica- tion. The diode, such as diodes 24, 26, in a circuit may be replaced by a MOSFET. If the

MOSFET used is of vertical structure its antiparallel diode or inverse diode (body diode) is utilized. This is illustrated in fig. 5 which shows like members identified by like reference numerals as in fig. 1. Fig. 5 diagrammatically illustrates the replacement of the diode-type rectifier 14 by a synchronous rectifier circuit 32 on the basis of MOSFETs.

With reference to figs. 4 and 5 it becomes clear that the voltage drop ups at the MOSFET is determined by the switch-on resistance Rpsiony of the MOSFET and the actual drain current which must equal the diode current ip. The following must apply if the energy loss is to be reduced: lups| = [Rps(on) * ip[>uF.

The forward loss, therefore, can be reduced by selecting a MOSFET of which the forward resistance Rps(ony is small.

A control signal is required for the MOSFET to be switched on and off. Generating the con- trol signal has a decisive influence on the switching behavior. Moreover, the energy losses in this circuit must be taken into consideration. There are various known methods of driving synchronous rectifiers comprising a MOSFET, and they may be roughly classified as self- controlled, IC controlled, and current controlled.

Fig. 6 is a simplified diagram of the secondary side of a self-controlled synchronous rectifier circuit. Again the same reference numerals are used to designate members corresponding to those of fig. 2. The power diodes 24, 26 shown in fig. 2 are replaced by MOSFETs 34 and 36, respectively. The first winding section at the secondary side 18b is designated Lg; and the second winding section is designated Ls.

In the case of the self-controlled synchronous rectifier according to fig. 5 the output voltage of the power transformer 18 is used to control the MOSFETs 34, 36. This circuit has the advan- tage of necessitating only little expenditure in circuitry since no additional driver circuits are needed to drive the MOSFETs 34, 36.

With reference to fig. 3, the output voltage uy, at the first winding section Lg, is positive dur- ing the time interval from t; to t;, while ug, is negative at the second winding section Lg,,

With these conditions, the p-channel MOSFET 34 is switched on because its gate voltage is negative. Corresponding switching behavior is true of the p-channel MOSFET 36 during the time interval from ts to ts; MOSFET 34 is turned off, while MOSFET 36 is turned on. Opera- tion of the MOSFET switches 34, 36 during these two time intervals is satisfactory. During the freewheeling phase, however, no control voltage is generated. The current flows through the inverse diodes of the MOSFETs, whereby higher forward losses are produced than neces- sary. ~ Furthermore, it is disadvantageous with the self-controlled synchronous rectifier design that p-channel MOSFETs are needed which are much more expensive and have a higher forward resistance than comparable n-channel MOSFETs. Moreover, the transformer output voltage range is restricted due to the gate voltage of the MOSFETs 34, 36. It must be higher than the threshold voltage and lower than the maximum permissible gate voltage of the MOSFETs of approximately 30 V.

Fig. 7 is a circuit diagram of the secondary side of a synchronous rectifier with 1C drive, cor- responding members being designated by the same reference numerals as in fig. 6. With this synchronous rectifier design, driver ICs 38, 40 are provided to drive the MOSFETs 34, 36.

There are only a few manufacturers who offer such specific driver ICs for synchronous recti- fiers. The IC components 38, 40 scan the secondary voltages of the transformer 18, and the

MOSFETs 34, 36 are turned on or off, depending on the potential profile. The electronic con- trol assures synchronous switching on and off of the synchronous rectifier. However, the scarce commercial availability and the relatively high costs and greater expenditure involved in connecting and feeding the driver ICs 38, 40 are points against making use of driver ICs 38,40.

Fig. 8, finally, is a circuit diagram of the secondary side of a current controlled synchronous rectifier, showing only the upper part of the secondary side 18b which includes the first sec- ondary winding section Lg). The structure of the lower part, including the second secondary winding section Ls; is mirror inverted.

In the current controlled synchronous rectifier, the power MOSFET 34 (and 36, too, not shown in fig. 8) is controlled through a current transformer 42. The current transformer 42 is connected in series between the upper winding section Lg; at the secondary side 18b of the power transformer 18 and the MOSFET 34 and comprises a primary winding 42a and a sec- ondary winding 42b. The secondary winding 42b is connected to the gate of the MOSFET 34 by way of a voltage divider composed of two resistors 44, 46.

AMENDED SHEET

® When the transformer 18 is controlled such that current flows through the winding section Ls: of the secondary side 18b current also will flow through the inverse side of the MOSFET 34, the current transformer 42 thus generating current in its secondary winding 42b. This current : brings about a voltage drop across the resistor 46 of a magnitude equal to the gate voltage of the MOSFET 34. The value of the voltage drop is adjustable by the ratio between the two resistors 44, 46.

MOSFET 34 is tumed on during the time interval from t, to t, i.e. it is on also during the freewheeling phase. When current flows in the opposite direction the output voltage at the secondary side 42 of the current transformer 42 becomes negative and MOSFET 34 turns off.

The second section Lg; of the power transformer 18 and the second MOSFET 36 (not shown in fig. 8) behave accordingly, yet with opposite sign. .

Controlling synchronous rectifiers by means of current transformers, such as illustrated in fig. 8, has certain disadvantages. On the one hand, a MOSFET requires a great current pulse to be turned on, which means that the windings ratio n2/nl of the current transformer must be low.

In the switched-on state, on the other hand, the gate current of the MOSFET is negligible, which means that a high windings ratio n2/n1 of the current transformer is required.

A description of prior art similar to what has been described above, but relating to a single phase forward transformer with synchronous rectification, including a current transformer, is to be found in "Synchronous Rectification Circuit Using A Current Transformer" by Y. Ku- bota et al., NTELC Conference Proceedings, September 2000, pages 267 to 273.

A need exists, starting from the state of the art as described above, to indicate a synchronous recti- fier circuit for a push-pull voltage transformer that attains the fastest possible switching of the metal oxide semiconductor field effect transistors (MOSFETs) while, at the same time, causing the least possible power dissipation. This aim is to be reached, above all, by generating a higher switch-on current for the MOSFETs so as to keep the time of flow through the inverse diodes as short as possible, and of keeping the drive current as small as possible when the MOSFETs are in the on-state so as to minimize power dissipation.

The need is fulfilled by a synchronous rectifier circuit comprising the features recited in claim 1 which is to be considered as an integral part of this specific description.

In a synchronous rectifier circuit comprising a current converter of the kind described above, the invention more specifically provides for designing the transformer of the current converter such as to include first and second secondary windings to drive the MOSFET in two stages.

The first secondary winding yields a relatively high current gain, while the second secondary winding yields a relatively low current gain. Therefore, the first secondary winding can be utilized for quickly turning on the MOSFET, the gate capacitance of the MOSFET being charged rapidly by the relatively great switch-on current. In the second stage, the MOSFET which has been turned on is kept in the on state with relatively low current gain, and that re- quires a smaller current. Thus the MOSFET is susceptible of being driven quickly and, when turned on, being kept with low loss in its switched state.

In the preferred embodiment of the invention, therefore, the electronic switches of the syn- chronous rectifier circuit are embodied by MOSFETs and the invention will be discussed be- low in the context of that type of transistors. However, the electronic switches likewise may be implemented as bipolar transistors or any other suitable type of switch.

In addition, the invention provides for a transconductance choke to reset the secondary side 42b of the current transformer 42, the choke supporting the resetting or demagnetizing of the current transformer 42 after the switching operation.

While transconductance chokes are known in principle, they were not employed so far either in synchronous rectifiers or for resetting current transformers. The specific provision of the transconductance choke in the circuitry of the secondary side of the current transformer per- mits the resetting operation to be accomplished in clearly less time so that the freewheeling phase can be chosen to be shorter. That offers the advantage of permitting the push-pull volt- age transformer, on the whole, to be operated at higher frequency and/or in the sense of greater variability of the clock ratio. As the freewheeling phase becomes shorter, the clock ratio of the transformer can be adjusted more flexibly, in response to the energy to be trans- mitted. A shorter freewheeling phase also contributes to increasing the efficiency.

Preferred embodiments of the invention are indicated in the dependent claims.

The invention will be described further by way of a preferred embodiment with reference to the drawings, in which:

Fig. I is a block diagram of a push-pull voltage transformer circuit in which the invention may be used;

Fig. 2 is a schematic circuit diagram of a push-pull voltage transformer, including a diode- type rectifier;

Fig. 3 shows idealized waveshapes of the output voltages and output currents of the push- pull voltage transformer shown in fig. 2;

Fig. 4 is an equivalent circuit diagram of a MOSFET which may replace the diodes shown in fig. 2;

Fig. 5 illustrates the replacement of the diode-type rectifier of fig. 1 by a synchronous recti- fier including a MOSFET, in the form of a block diagram;

Fig. 6 is a schematic circuit diagram of the secondary side of a self-controlled synchronous rectifier;

Fig. 7 is a schematic circuit diagram of the secondary side of an IC controlled synchronous rectifier;

Fig. 8 is a schematic circuit diagram of the upper half of the secondary side of a synchronous rectifier including a current transformer, the lower half being mirror inverted;

Fig. 9 is a schematic circuit diagram of the upper half of the secondary side of a synchronous rectifier circuit according to the invention, the lower half being mirror inverted;

Fig. 10is a schematic circuit diagram of the upper half of the secondary side of the synchro- nous rectifier circuit according to the invention, similar to the illustration in fig. 9 but with further details, in a first operating phase;

Fig. 11 shows a circuit diagram similar to fig. 10, in a second operating phase;

Fig. 12shows a circuit diagram similar to figs. 10 and 11, in a third operating phase;

Fig. 13 shows idealized waveshapes of the output voltages and output currents of the synchro- nous rectifier circuit according to figs. 10 to 12.

Fig. 9 is a circuit diagram of the upper part of the secondary side of the synchronous rectifier according to the invention. The lower part of the secondary side is mirror inverted. The structure of the primary side may be of known design.

Components corresponding to those in the circuitry shown in figs. 6 to 8 are designated by the same reference numerals.

At its secondary side, the synchronous rectifier circuit according to the invention comprises a first secondary winding section Lg; of the power transformer 18 connected in series with a current converter transformer 48 and a MOSFET. The current converter transformer 48 com- prises a primary winding 48a and a secondary winding 48b, the latter being divided into two winding sections 48bl and 48b2. The first secondary winding section 48b1 is connected to a first branch which, in the embodiment illustrated in fig. 8, includes a diode 50 and a transcon- ductance choke 52. The second secondary winding section 48b2 is connected to a second branch including a diode 54 and a resistor 56. The transconductance choke 52 is also known as "magnetic amplifier" or "mag amp".

The first and second branches are connected via another resistor 58 to a common reference potential of the synchronous rectifier circuit.

As is known and also shown in fig. 8, a storage choke 28 and a storage capacitor 30 are pro- vided to form an output filtering stage at the output of the synchronous rectifier circuit.

The function of the synchronous rectifier circuit, in principle, is based on the following switching behavior: To turn on the MOSFET 34, a relatively high current is generated across the second secondary winding section 48b2. At the same time, relatively low current flows through the first secondary winding section 48b1 and the first branch in which the transcon- ductance choke 52 is located. To that end, the winding ratio of the first and second winding sections 48bl, 48b2 to the primary winding 48a is suitably chosen. In particular, a low wind- ing ratio n21/nl in the order of magnitude of 10:1 of the second secondary winding section 48b2 to the primary winding 48a is chosen so that a high switch-on current will be generated and the MOSFET 34 be turned on very rapidly. A high winding ratio n22/n] in the order of magnitude of 100:1 of the first secondary winding section 48b1 to the primary winding 48a together with the high inductivity of the transconductance choke then lead to slow current rise in the first branch. As the current increase, the inductivity of the transconductance choke 52 decreases - the transconductance choke 52 is "switched on" - and consequently the current across the resistor 56 becomes very small again.

This switching behavior is due to the property of the transconductance choke 52, namely having a high inductivity at low current and acting like an open switch. As the current in-

creases, the inductivity decreases until the transconductance choke acts like a closed switch.

This switching behavior is exploited advantageously by the invention.

In practice, the switch-off behavior of the circuit is somewhat more complicated than in the other circumstances described above. In a practical implementation, therefore, some addi- tional components are necessary and advantageous in combination with a non-ideal transcon- ductance choke, as will be described below.

Figs. 10 to 12 illustrate detailed circuit diagrams of the synchronous rectifier circuit according to the invention. The main constituent parts of the secondary rectifier continue to be the cur- rent transformer 48 and the MOSFET 34, the current transformer 48 comprising a first output branch which includes the transconductance choke 52 and the diode 50, and a second output branch which includes the diode 54 and the resistor 56. In addition to the circuit of fig. 9, figs. 10 to 12 show a diode 60 which is connected parallel to the resistor 56 and another diode 62 connected in series to the resistor 58. The first and second branches connect the secondary winding 48b of the current transformer 48 through the diode 62 and resistor 58 and, further- more, through a transistor 64 connected in parallel with them, especially a bipolar pnp- transistor, to the gate of the MOSFET 34.

Figs. 10 to 12 further show the transconductance choke 52 having a first winding 52a in the first branch and a second winding 52b connected in series to a diode 66 and a resistor 68, this series connection being connected in parallel with the first and second branches, as may be seen in figs. 10to 12.

Operation of the synchronous rectifier circuit according to figs. 10 to 12 will be described in greater detail below, with reference to the waveshapes illustrated in fig. 13. The idealized waveshapes of fig. 13 demonstrate the output voltage ug, of the upper winding section Lg, of the secondary side 18b of the power transformer 18, furthermore the output current ig; of this winding section Ls), and the drain-source voltage ups of the MOSFET 34, as well as the reset current irss (see fig. 12) to reset the current converter transformer 48 and the transconduc- tance choke 52.

The output voltage ug) is zero during the time interval from to to t; (see fig. 13). During this phase, which corresponds to the freewheeling phase, both MOSFET switches 34, 36 are con- ducting (not shown in the figs. but mirror inverted to the presentation in figs. 10 to 12).

Attime ty current ip; flows through the inverse diode of MOSFET 34 (fig. 10) and the current transformer 48. Accordingly, current also flows through the diode 54 and begins to charge the input capacitance of the MOSFET 34. This current depends in the first place on the winding ratio n21/n1 of the second secondary winding section 48b2 to the primary winding 48a of the current converter transformer 48 and on the load.

A negligibly small current also flows through the transconductance choke 52 during the first time interval TD (see fig. 13). The winding ratio n22/n1 of the second secondary winding section 48b2 to the primary winding 48a of the current converter transformer 48, the transconductance choke 52 itself and the load all influence the switching behavior of the

MOSFET 34 (see fig. 11). The winding ratio n22/nl must be chosen depending on the transconductance choke 52 such that drive losses will be minimized. At the end of the time interval TD the transconductance choke 52 is saturated and the MOSFET 34 fully turned on (see fig. 13).

The diode 69 and the resistor 56 determine the level of the gate-source voltage of the

MOSFET 34. The resistor 58 serves to attenuate oscillations in the gate circuit.

At time t; the MOSFET 34 continues to be conducting, while the other MOSFET 36 (not il- lustrated in the figs.) turns off. Consequently, all the output current flows through MOSFET 34.

During the freewheeling phase, time interval t; to t3, the behavior of the circuit corresponds to what has been described above. Beginning at time t3, the voltage ug, across the secondary winding section Lg; becomes negative. Thereupon, the current through the MOSFET 34 de- © 30 creases. As a consequence, the voltages across the secondary winding sections 48bl and 48b2 of the current converter transformer 48 are inverted, the transistor 64 becomes conducting and, thereby, accelerates turning off of the MOSFET 34.

The negative voltages at the secondary winding sections 48bl, 48b2 of the current trans- former 48 start the resetting process of the transconductance choke 52. A reset current ies flows through the resistor 68 and the diode 66, resetting the current transformer 48 and the ~ transconductance choke 52 (see fig. 12).

The reset current is limited by the resistor 68. The diode 66 makes sure that current will flow during this interval alone. Reset current irgs stops to flow no later than at time ty when the freewheeling phase begins.

If both MOSFETs 34, 36 should be non-conducting when the synchronous rectifier is switched on the respective inverse diodes take care of the flow of current when it begins to flow through the secondary winding 18b of the power transformer 18. At the same time, cur- rent also flows through the primary winding 48a of the current converter transformer 48 and thus through the first and second secondary winding sections 48b1, 48b2. The transconduc- tance choke 52 prevents current from flowing through the first branch, but current serving to charge the gate capacitances of the MOSFETs 34, 36 does flow through the diode 54 of the second branch. The small transmission ratio in the order of 10:1 of the second secondary winding section 48b2 to the primary winding 48a of the current converter transformer 48 brings about a relatively great switching current and results in extreme acceleration of the driving of MOSFETs 34, 36. This quick charging in tum cuts the time of current flow across the inverse diodes. The MOSFET 34, 36 becomes conducting as the threshold voltage is reached. In this context, the resistors 56, 58 serve as current limiters and the diode 60 as a voltage limiter. The transistor 24 is blocked at this time. As already described, also the transconductance choke 52 is blocked during this first switch-on phase.

The transconductance choke 52 acts as a magnetic switch in this circuit. It is only after a cer- tain charging time that the transconductance choke 52 becomes saturated, which corresponds to a closed switch condition, so that the gate capacitance of the MOSFET 34, 36 is fully charged through the diode 50 and thus the synchronous rectifier is fully controlled. The selec- tion of a greater transmission ratio in the order of 1:100 between the primary winding 48a and the first secondary winding section 48b1 of the current converter transformer 48 brings with it a smaller charging current and, therefore, less driving losses of the synchronous rectifier.

As already mentioned, it is an advantageous feature of the invention that the transconductance choke 52 and the current converter transformer 48 can be reset or demagnetized quickly and easily after a switching operation. To this end, the transconductance choke 52 is provided with a second winding 52b, as may be seen in figs. 10 to 12, which may have four turns, for example. The winding direction, here, is opposed to that of the first winding 52a of the transconductance choke 52. The second winding 52b is fed through both the resistor 68, serving as current limiter, and the diode 66 preventing current to flow in the opposite direc- tion. The reset operations of the transconductance choke 52 and the current converter trans- former 48 each take place during the blocking phase of the MOSFETs 34, 36. During this time period the voltage at the corresponding winding section of the secondary side 18b of the power transformer 18 is negative. Due to this negative potential, current will flow through the resistor 68, diode 66, transconductance choke 52 and secondary winding 48b of the current converter transformer 48. The voltage at the secondary side of the current transformer jumps to negative potential, and the subsequent falling current flow magnetizes the current trans- former and the transconductance choke in opposite direction. The return current iggs breaks off (see fig. 13) as the voltage at the secondary side, e.g. Ls), of the power transformer 18 returns to 0 volt (freewheeling phase). Thereafter, both inductivities, of the transconductance choke 51 as well as the current converter transformer 48, are demagnetized and ready for the next switching operation.

The guidelines below may be followed for selecting and dimensioning the various compo- nents of the synchronous rectifier circuit according to the invention.

The power transformers, MOSFETs, in the first place are selected with a view to a low for- ward resistance, preferably a forward resistance Rps < 50 m(2, a drain-source voltage and a drain-source current depending on the desired output voltage and output current.

The optimum transmission ratio of the current converter transformer should be determined experimentally, depending on the particular case of application. Winding ratios of the primary winding 48a to the first secondary winding section 48bl to the second secondary winding section 48b2 of 1:50:5 and 1:100:10 have provided good test results in this configuration.

In the rectifier circuit, the transconductance choke 52 assumes the function of a switch. It is responsible for a second stage of driving the MOSFET to take place after a certain time TD, as described above.

A transconductance choke essentially consists of a saturable annular soft magnetic core in- cluding one or more windings. The hysteresis curve of the core is almost rectangular. The magnetic switch is either on or off, depending on the degree of magnetization of the choke.

The behavior of the transconductance choke will be described briefly. When voltage is ap- plied to the choke the inductivity of the choke, initially, is very high, and no current flows through the choke. After the lapse of a period of time TD, the choke enters into a state of satu- ration, and then its inductivity is very small. At this state, current flows through the winding of the transconductance choke, the magnetic switch is closed. The magnetic flux density re- mains almost constant as long as this current flows. When the flow of current through the choke is interrupted the magnetic field strength decreases while the magnetic flux density continues to remain constant. If a current now flows in the opposite direction (reset current) or if a negative voltage is applied the inductivity of the choke goes down towards zero so that the transconductance choke will become demagnetized. If the reset current flows for a long enough time the transconductance choke will end up being fully demagnetized and thus be reset. In case the reset current continues to flow the transconductance choke will become magnetized in the opposite direction, and the magnetic switch is opened again.

The function of further individual components is as follows:

The diodes 50, 54, and 58 prevent reset current from flowing during the blocking phase of the

MOSFETs 34, 36. Schottky diodes, for example, may be selected since their loss is rather small due to their small forward voltage and as they allow relative great currents to be ap- plied.

The resistor 56 and the Zener diode 60 serve for adjusting the gate-source voltage. The resis- tor 58 serves as a gate resistor to suppress oscillations. Furthermore, the provision of the re- sistor 58 upstream of the transistor 64 abbreviates switching-off of the MOSFET 34, whereby the efficiency of the overall circuit is improved once more. The transistor 64 serves for dis- charging the gate capacitance of the MOSFETs 34. 36. The resistor 68 and the diode 66 serve for demagnetizing the transconductance choke 52 and the current transformer 48.

AMENDED SHEET

® The features disclosed in the specification above, in the claims and drawings may be signifi- cant for implementing the invention in its various embodiments, both individually and in any combination.

The invention provides a synchronous rectifier for use in a push-pull voltage transformer, being exceptionally efficient and quick in switching. This makes it possible, on the one hand, to apply the circuit according to the invention particularly also in multi-stage switch mode power supply and, at the same time, to obtain switch mode power supply operating at high frequency and/or being flexible as regards the clock ratio. The invention is particularly ad- vantageous for switch mode power supply with low output voltage, such as <24 V, where the losses of a diode-type rectifier are especially significant. Examples of applications are switch mode power supply for telecommunications systems, computers and industrial applications, as well as voltage supply for processor cores, and particularly all those applications which require low voltages and high currents. "Comprises/comprising” when used in this specification is taken to specify the presence of stated features, integers, steps or components but does not preclude the presence or addition of one or more other features, integers, steps or components or groups thereof.

The claims which follow are to be considered an integral part of the present disclosure.

Reference numbers (directed to the drawings) shown in the claims serve to facilitate the correlation of integers of the claims with illustrated features of the preferred embodiment(s), but are not intended to restrict in any way the language of the claims to what is shown in the drawings, unless the contrary is clearly apparent from the context.

Minebea Co., Ltd.

M30183PCT(L)

List of reference numerals : input switching stage 12 power transformer 14 rectifier circuit 16 output filter 10 18 power transformer 18a primary side of the power transformer 18b secondary side of the power transformer 20 power transistor 22 power transistor 24 secondary diode 26 secondary diode 28 storage choke 30 storage capacitor 32 synchronous rectifier circuit 34 MOSFET, electronic switch 36 MOSFET, electronic switch 38 driver IC 40 driver IC 42 current transformer 42a primary winding of the current transformer 42b secondary winding of the current transformer 44 resistor 46 resistor 48 current converter transformer 48a primary winding of current converter transformer 48b secondary winding of current converter transformer 48b1 first secondary winding section 48b2 second secondary winding section 50 diode 52 transconductance choke

52a first winding of the transconductance choke 52b second winding of the transconductance choke 54 diode 56 resistor

58 resistor 60 diode 62 diode 64 transistor 66 diode 68 resistor

Claims (21)

i ® Minebea Co., Ltd. M30183PCT(L) CLAIMS:

1. A synchronous rectifier circuit, comprising a power transformer (12; 18) which has a primary side (18a) including first and second primary winding sections and a secondary side (18b) including first and second secondary winding sections, a rectifier circuit (32) at the secondary side of the power transformer (12; 18), which rectifier circuit comprises first and second electronic switches (34, 36) associated with the first and second secondary winding sections (18b), respectively, first and second current transforming means (48) associated with the first and second secondary winding sections, respectively, and first and second drive circuits (50 - 58) for the first and second electronic switches (34, 36), respectively, each current transforming means (48) generating first and second currents which are dependent on the current of the associated secondary winding section (18b) of the power transformer (18), and each drive circuit (50 - 58) comprising a first branch (50, 52) and a sec- ond branch (54, 56) to receive the first and second currents, respectively, generated by the current transforming device (48), and the first branch comprising a transconductance choke (52).

2. The synchronous rectifier circuit as claimed in claim 1, wherein each current trans- forming means comprises a converter-type current transformer (48) including a primary wind- ing (48a) and first and second secondary windings (48b), the first and second branches of the drive circuit being associated with the first and second secondary windings, respectively, of the converter-type current transformer (48).

3. The synchronous rectifier circuit as claimed in claim 2, wherein the electronic switches (34, 36) each comprise a MOSFET.

4. The synchronous rectifier circuit as claimed in claim 2 or 3, wherein the primary winding (48a) of the first current transforming means (48) 1s connected in series to the first secondary winding section (18b) of the power transformer (18) and the primary winding of the second current transforming means is connected in series to the second secondary winding section of the power transformer (18).

AMENDED SHEET ® Cs, The synchronous rectifier circuit as claimed in claim 3 or 4, wherein the first switch (34) is connected in series to the first secondary winding section (18b) of the power trans- former (18) and the second switch (36) is connected in series to the second secondary winding section (18b) of the power transformer (18).

6. The synchronous rectifier circuit as claimed in any one of claims 3 to 5, wherein the first branch connects the first secondary winding (48bl) of the converter-type current trans- former (48) via a diode (50) and the transconductance choke (52) to the gate of the associated MOSFET (34), and the second branch connects the second secondary winding (48b2) of the converter-type current transformer (48) via a diode (56) to the gate of the associated MOSFET (34).

7. The synchronous rectifier circuit as claimed in claim 6, wherein the first and second branches are connected to the gate of the associated MOSFET (34) through a switching ele- ment (64).

8. The synchronous rectifier circuit as claimed in claim 7, wherein the switching element (64) is a bipolar transistor.

9. The synchronous rectifier circuit as claimed in any one of claims 2 to 8, wherein the _ first secondary winding (48bl) of the converter-type current transformer (48) has a greater number of windings than the second secondary winding (48b2) of the converter-type current transformer (48).

10. The synchronous rectifier circuit as claimed in claim 9, wherein the winding ratio (n1:n22:n21) of the primary winding (48a) to the first secondary winding (48b1) to the second secondary winding (48b2) of the converter-type current transformer (48) is in the order of 1:100:10.

11. The synchronous rectifier circuit as claimed in any one of the preceding claims, wherein the transconductance choke (52) comprises a switch winding (52a) and a reset wind- ing (52b), the switch winding (52a) being disposed in the first branch and the reset winding (52b) being disposed in a reset branch parallel to said first branch.

AMENDED SHEET -H- ® 12. The synchronous rectifier circuit as claimed in claim 11, wherein the winding ratio of the switch winding (52a) to the reset winding (52b) lies in the order of 1:4.

13. The synchronous rectifier circuit as claimed in claim 12, wherein the reset branch in- cludes a second diode (66).

14. The synchronous rectifier circuit as claimed in any one of the preceding claims, wherein an input switching stage (10) is connected upstream of the power transformer (18).

15. The synchronous rectifier circuit as claimed in any one of the preceding claims, wherein an output filter stage (16) is connected downstream of the rectifier circuit (32).

16. A push-pull transformer comprising a synchronous rectifier circuit as claimed in any one of the preceding claims.

17. A synchronous rectifier circuit including any new and inventive integer or combina- tion of integers, substantially as herein described.

18. A synchronous rectifier circuit according to the invention, as hereinbefore generally described.

19. A synchronous rectifier circuit as specifically described with reference to or as illus- trated in the accompanying drawings.

20. A push-pull transformer as claimed in claim 16, substantially as hereinbefore de- scribed and exemplified.

21. A push-pull transformer including any new and inventive integer or combination of integers, substantially as herein described.