CH708040B1 - Electronic power converter and method for its control. - Google Patents

Abstract

The invention relates to an electronic power converter and a method for controlling such a converter. An electronic power converter has partial transformers (2) combined to form a sum of output voltages. For each of the phases, a switching device (1a, 1b, 1c) is arranged to apply a positive or negative phase voltage or the voltage zero to the primary-side winding of an associated partial transformer (2). In operation of the converter, a sequence of voltage pulses having the alternating positive and negative respective phase voltages is applied to the primary-side winding of the associated sub-transformer (2) at a modulation frequency, a duration of each of the voltage pulses being proportional to the instantaneous magnitude of the respective phase voltage, and wherein Voltage pulses of the switching devices (1a, 1b, 1c) are generated synchronously to each other and each rising and falling edges of the voltage pulses in each pulse half period are symmetrical with respect to a common center. This results in a power output to a load (16), which is averaged over a period of the modulation frequency, constant.

Description

Description: The invention relates to the field of power electronic circuits and relates to an AC / DC converter for power transmission from a multiphase primary side to a secondary side or vice versa and to a method for its control.

Circuits for rectifying a three-phase mains voltage are known.

It is an object of the invention to provide an electronic power converter for power transmission from a multiphase primary side to a secondary side or vice versa and a method for its control, which compared to conventional transducers have reduced requirements for the components used and thereby have a very constant power consumption ,

The object is achieved by an electronic power converter and a method for controlling it according to the corresponding independent claims.

In the method for controlling an electronic power converter for power transmission from a primary side with a plurality of phases to a DC side or vice versa, wherein • the power converter for each of the phases has a partial transformer, wherein in each case a primary-side winding of a partial transformer is connected to an associated switching device and a sum voltage of the partial transformers is producible on a secondary side of the partial transformers; • for each of the phases, the switching device is adapted to apply an associated phase voltage or the negative phase voltage or the zero voltage to the primary side winding of the associated partial transformer; The power converter has a smoothing element for smoothing a secondary current; Each of the switching devices is controlled in each case by • with a modulation frequency, a sequence of voltage pulses with the alternating positive and negative respective phase voltage is applied to the primary-side winding of the associated partial transformer; Wherein a duration of each of the voltage pulses is proportional to the instantaneous magnitude of the respective phase voltage; and wherein the voltage pulses of the switching devices are generated synchronously with each other and each rising and falling edges of the voltage pulses in each pulse half-period are temporally symmetrical with respect to a common center.

The primary side can be placed on an AC voltage and can then also be referred to as the AC voltage side. The secondary side can be connected to a DC voltage and can then also be referred to as a DC voltage side.

Further preferred embodiments will become apparent from the dependent claims. Characteristics of the method claims are analogously combined with the device claims and vice versa.

In the following, the subject invention based on preferred embodiments, which are illustrated in the accompanying drawings, explained in more detail. Each show schematically:

Fig. 1 an electronic power converter;

Fig. 2 low-frequency processes in the power converter;

FIG. 3 shows high-frequency processes in the power converter; FIG. and

4 shows a variant of the electronic power converter with a load on a series LC resonant circuit.

Basically, the same parts are provided with the same reference numerals in the figures.

Fig. 1 shows an electronic power converter 1 with three partial transformers 2a, 2b, 2c, wherein in each case a primary-side winding of a partial transformer 2a, 2b, 2c to an associated switching device (or phase switching stage) 1a, 1b, 1c is connected. Secondary windings of the partial transformers 2a, 2b, 2c are connected in series, whereby a sum voltage u on the DC side windings can be formed during operation of the circuit.

As shown, the transformers of the phases may be implemented separately as separate partial transformers 2a, 2b, 2c. Instead, the partial transformers can also be combined in a common magnetic component. Such an integrated solution is designed so that each phase is assigned a magnetic core with two legs and on a first leg, the respective primary winding is arranged. The secondary winding is now executed only once and for all phases together, and encloses all second legs of the magnetic cores of the partial transformers. The result is an addition of the magnetic flux through the secondary winding and thus the flux change in the transformer and thus occurs at the secondary winding the same sum of secondary voltages as in the series connection of the separate partial transformers 2a, 2b, 2c. Thus, the series connection of the partial voltages is structurally simple realized.

For each of the phases, the switching device 1a, 1b, 1c (also called phase module) is adapted to the respective associated phase voltage or the negative phase voltage or the voltage zero, so for example a short circuit, to the primary-side winding of the associated partial transformer 2a, 2b, 2c create. In the example shown, a diode bridge 11 is provided as a rectifier and a subsequent (full) bridge circuit 12 with active switches T-n, T12, T21, T22.

An input filter capacitance Cf may be arranged for smoothing the pulse-shaped input currents of the bridge circuit. In the example shown, this input filter capacitance CF is drawn between an input phase connection 13 and a common neutral point 14 of the input filter capacitances CF and also of the switching devices 1a, 1b, 1c. However, it can also be arranged in a functionally equivalent manner between the input phase connection 13 and a second diode bridge branch center 15.

An input filter inductor LF may be present for smoothing a respective phase current associated with the input filter capacitance CF; it may be part of the switching device 1a, 1b, 1c or part of an external circuit.

The switching devices 1a, 1b, 1c here form a star connection with a first neutral point 15a. The input filter capacitances CF likewise form a star connection, with a second star point 14. The neutral points 14, 15a of the two star circuits can be connected to one another, as shown here by dashed lines. In the embodiment as a star connection, at least one of the neutral points 14, 15a is advantageously to be connected to an existing neutral conductor. If no neutral conductor is present, a virtual neutral point 14 is advantageously formed by star connection of three filter capacitors, and the star point 15a of the phase systems or switching devices 1a, 1b, 1c is connected thereto, as in the present example.

Instead of an input-side star connection and a delta connection is used. One advantage of this is: The individual units are then decoupled, since the outer conductor voltages are impressed through the network even when supplied via only three conductors (that is, in a three-core network without a neutral conductor). A disadvantage is that the input voltage is higher than with a star connection (by a factor of 3). Accordingly, a higher voltage loading of the power semiconductors occurs, thus reducing the selection of usable for the practical circuit realization unipolar power transistors with low on-resistance MOSFETS.

The shown switching devices 1a, 1b, 1c for generating in each case the voltage applied to the transformer primary winding of a phase voltage are implemented with a low implementation effort in power and control part by coupling a diode bridge 11 (for rectification) and a subsequent full bridge circuit 12 in two stages. It can be used between these two stages in addition a switching overvias limiting but not significantly influencing the output voltage of the diode rectification capacitor.

As an alternative to a two-stage circuit, the switching devices 1a, 1b, 1c can be realized in each case by a full-bridge circuit with four-quadrant switches. A four-quadrant switch allows a current flow in both directions and a blocking of voltages of both polarities and can be realized in a known manner by combining active and passive unidirectional, unipolar power semiconductors.

On the secondary side, the power converter has a rectifier 3 for rectifying the sum voltage u and an inductance L as a smoothing element 4 for smoothing a secondary-side current i. This secondary-side current i is therefore that which flows through the series connection of the secondary-side windings of the partial transformers 2a, 2b, 2c. An output filter capacitance C may be arranged as a smoothing capacitance for forming a smoothed output voltage u0Ut of the rectifier 3. A load or load 16 is supplied with the output voltage.

The circuit can be operated as follows: Synchronous to each other, with a common modulation frequency, voltage pulses are generated in the switching devices whose width is modulated according to a degree of modulation or duty cycle according to a modulation signal. The modulation signal is formed as the amount of a sine signal. In this case, the modulation signal is synchronous and in phase with the phase voltage with which the respective switching device is fed. The modulation signal for each of the switching devices can therefore be formed by rectification and corresponding amplitude scaling of the respective phase voltage.

Fig. 2 shows low-frequency processes in the power converter, i. Operations in the range of a fundamental frequency of the phase voltages, here over a period, for example 20 ms. As a rule, the fundamental frequency is a mains frequency of a three-phase network, for example 50 Hz or 16 2/3 Hz. The phase voltages are shown in FIG. 2 (a), the associated modulation signals in FIG. 2 (b).

The modulation frequency can be at least as high as 6 times the fundamental frequency, in particular at least one hundred times higher than the fundamental frequency. The modulation frequency can also be chosen so high depending on the rated power that no audible operating noise of the power converter occurs.

Fig. 3 shows high-frequency processes in the power converter. In FIG. 3 (a), the voltage blocks each with the modulation frequency applied to the primary side windings and transformed to the DC side are shown in the course of the secondary side voltages ua, ub, uc. The voltage blocks are alternately generated positive and negative in each phase according to the modulation frequency. The centers of the pulses of the different phases are simultaneous, the modulation of the pulse widths takes place around this common center. In other words, the pulses are each time symmetrical with respect to a common center. Over the time duration of the illustration in FIG. 3 (a), the amplitudes and thus also the modulation (corresponding to the pulse width) of the individual voltage blocks are virtually constant over the visible multiple periods of the modulation frequency. Over a longer period of time, ie over a period of the fundamental frequency or mains frequency, they change in proportion to the respective phase voltage.

From the above-described rule for the modulation of the voltages on the primary-side windings, the following results: • The amplitudes of the individual voltage pulses are proportional to the instantaneous input voltage of the respective phase, thus have a sinusoidal curve at the fundamental frequency. Fig. 2 (c) shows the time course of the voltage pulses in the three switching devices 1a, 1b, 1c as in Fig. 3 (a), but over a period of the low-frequency fundamental frequency. Due to the high-frequency course of the voltage pulses, no individual pulses are visible, but only one envelope at a time. It is therefore visible how the maxima of the voltage peaks of the voltage pulses corresponding to the associated phase voltages replace each other with a phase shift of 120 °. Due to the way in which the modulation signal is formed, the duration of the individual voltage pulses likewise has a sinusoidal profile with the fundamental frequency over several voltage pulses. This can be seen in FIG. 3 (a). • It follows that the voltage-time surface of the voltage pulses of a phase runs according to the square of a sinusoidal function of time. The voltage averaged over several phases or with the modulation frequency thus also runs according to this function. The power flowing through a single phase into the power converter 1, averaged over the pulses occurring at the modulation frequency and at quasi-constant secondary-side current i, is thus also proportional to the square of a sinusoidal function. • As a sum of the power, which flows through the three phases, each with a phase shift of 120 °, from the secondary windings of the transformers 2, results in a constant power. The power converter 1 thus outputs constant power and thus appears, since the power converter 1 has no significant energy stores, the AC voltage as a load with constant power consumption, resulting in a sinusoidal current consumption.

In other words, the control of all phases is synchronous, i. with the same clock frequency according to the modulation frequency and such that the axes of symmetry of the positive and negative voltage pulses of each phase coincide and follow each other in the simplest case pulses of different sign, and so the primary winding of the transformer of a phase stage is fed without equal. For the generation of these voltage forms, a modulator can be used in a known manner in each phase, which can be designed in the sense of a simple pulse width modulation or a phase shift control. For the simple pulse width modulation during the application of positive or negative voltage to the primary winding of one of the partial transformers each diagonally opposite transistors (for example T-n and T22 or T12 and T21) are turned on, and locked in the remaining periods all the transistors of the bridge circuit of the phase.

Alternatively, both bridge branches can be clocked at 50% duty cycle (during the first half of a clock period, the upper power transistor T-n and T12, respectively, during the second half of the clock period, the lower power transistor T21 and T22 of a bridge branch turned on). The desired output voltage waveform is then formed by phase shifting the 50% drive signals of the two bridge branches.

Alternatively, instead of only one (positive or negative) rectangular single pulse per pulse half period and a sequence of multiple unipolar pulses are generated. For example, a positive square pulse is then replaced by a plurality of smaller width positive rectangular pulses whose pulse width is e.g. is sinusoidally changed so that after local averaging a sine wave is approximated whose amplitude is equal to the switching frequency fundamental of the original square wave signal. Here, too, the center axes of the individual pulses of the phases are again advantageously to be selected one above the other or a phase offset is to be provided so that the harmonic content of the ultimately formed sum of the secondary voltage has a minimum value.

In addition to the above-described basic form of the switching devices 1a, 1b, 1c and their control by pulse width modulation, an extension to resonant converters is conceivable, in which case in addition to regulating the change in the pulse width and a control intervention via frequency change can take place in a known form, here However, the frequency of all switching stages is the same to change, so that the collapse of the symmetry axes for the voltage formation described for the pulse width modulation remains analogously further.

On the secondary side, the transformed voltage pulses with the gradients ua, ub, uc are summed up to a total voltage u. The course of the total voltage u is shown in FIG. 3 (b): the total voltage u varies with the modulation frequency, but the amplitude remains constant over several periods with the modulation frequency. The power, averaged over a period of the modulation frequency, remains constant, as already mentioned. Fig. 2 (d) also shows the total voltage u over a period of the low-frequency fundamental frequency. The envelope has a ripple at six times the fundamental frequency. Since the individual pulses of the total voltage, corresponding to FIG. 3 (b), have different widths, the mean value of the amplitude is nevertheless constant.

In order to control the total voltage u and thus also the transmitted power, the degree of modulation can be varied in all three phases in the same ratio. The ratios of the three voltages ua, ub, uc remain the same, and the total voltage is thereby varied in the same ratio. The output voltage is thus adjusted or regulated via the degree of modulation of the switching devices 1a, 1b, 1c. As noted above, the widths of the voltage pulses over the line period vary according to the associated phase voltages and rectifier output voltages, respectively, in the switching devices 1a, 1b, 1c, i.e. the pulse width can be thought of as derived from a sinusoidal modulation signal proportional to the respective phase voltage. When changing the degree of modulation, the amplitudes of these modulation signals are changed the same way in all phases. At full scale in each phase, the voltage pulses in the maximum of the associated phase voltage have a symmetrical rectangular shape, i. E. they remain at the positive value of the phase voltage for the first half of the pulse period and then change to the negative value of the phase voltage for the second half of the pulse period. When lowering the Aussteuergrades occurs in addition to the positive and negative voltage value and the voltage value zero. For zero control no voltage is formed, i. the zero intervals occupy the entire duration of a pulse period.

The total voltage u is rectified by the rectifier 3. The output current of the rectifier is smoothed as already mentioned by the inductance L. The output voltage u0Ut is smoothed by the output filter capacitance C or averaged over the intervals according to the modulation frequency.

Since the smoothing by the output filter capacitance C and the inductance L must take place only in the high-frequency range of the modulation frequency, the requirements for the size or capacity and inductance of these components are substantially smaller. Accordingly, the circuit can be realized with much smaller components.

For the output side rectifier circuit in addition to the full bridge circuit, other known rectifier concepts, in particular a center circuit or a current doubler rectifier ("current doubler rectifier") can be used.

The inductance impresses on the DC side a current i, which flows through the series connection of the secondary-side windings. A transformed current thus flows through the AC-side windings of the partial transformers 2a, 2b, 2c. Depending on the switching state of the respective switching device 1a, 1b, 1c, this current flows • if voltage zero (a short circuit across two of the active switches) is applied to the partial transformer, in a circle through the bridge circuit 12 with active switches; In each case as bus current iA, iB, ic from the diode bridge 11 to the bridge circuit 12. The bus current iA, iB, ic in each phase is thus a pulse width modulated current with a constant amplitude. The pulse width is the same as that of the corresponding voltage pulse, thus also proportional to the corresponding phase voltage. Fig. 3 (c) shows these pulses of the bus currents iA, iB, ic with a resolution corresponding to the modulation frequency. Fig. 2 (e) shows one of the bus currents iA with the individual pulses not visible. • Because of the rectification by the diode bridge, the bus currents iA, iB, ic appear at the phases as phase currents with mains voltage frequency varying sign. This is likewise shown in FIG. 2 (e) for the phase current iitliA belonging to the mentioned bus current iA. The filtering by the input filter capacitance CF and the input filter inductance LF causes a smoothed phase current iin, A to emerge from the pulse width modulated current pulses of the bus current iA. As the pulse widths of these current pulses are, as already mentioned, proportional to the corresponding phase voltage, the smoothed phase current iin, A is also proportional to the corresponding phase voltage. This is shown in Fig. 2 (f).

Thus, the power factor is 1; the electronic power converter 1 appears at each phase as a purely resistive load.

Claims (10)

The circuit may also be designed for a power flow from the DC side to the primary side or for bidirectional power flow. For this purpose, the diode bridges on the primary and secondary side are to be provided with active switches instead of the diodes, respectively, with active switches added in anti-parallel to the diodes. Fig. 4 shows a variant of the electronic power converter with a load on a series LC resonant circuit having a substantially resistive load 16 parallel to the capacitance or a resonant circuit capacitor Cr of the resonant circuit. Thus, the current through the load is substantially independent of the resistance of the load 16. This arrangement is known for feeding lamps or other consumers whose resistance depends on the operating state. The load is resonantly fed with the sum voltage u. This eliminates the rectification of the sum voltage u and simplifies the circuit. In connection with the invention described herein, the combination with a series LC resonant circuit as described above is advantageous because the inductance of the series LC resonant circuit also serves for smoothing respectively impressing the secondary side current i. This simplifies the structure of the combined circuit even further. In principle, a resonant supply of a load is possible even without a series LC resonant circuit or via other resonant circuit circuits. claims 1. A method for controlling an electronic power converter for power transmission from a primary side with a plurality of phases to a DC side where • the power converter has a partial transformer (2a, 2b, 2c) for each of the phases, wherein in each case a primary-side winding of a partial transformer (2a, 2b, 2c) is connected to an associated switching device (1a, 1b, 1c) and a sum voltage (u) of the partial transformers (2a, 2b, 2c) can be formed on a secondary side of the partial transformers (2a, 2b, 2c); • for each of the phases, the switching device (1a, 1b, 1c) is adapted to apply an associated phase voltage, positive or negative, or the voltage zero to the primary side winding of the associated partial transformer (2a, 2b, 2c); The power converter has a smoothing element (4) for smoothing a secondary-side current (i); and in the method, each of the switching devices (1a, 1b, 1c) is respectively driven by • applying a sequence of voltage pulses having the alternating positive and negative respective phase voltages to the primary-side winding of the associated sub-transformer (2a, 2b, 2c) at a modulation frequency ; Wherein a duration of each of the voltage pulses is proportional to the instantaneous magnitude of the respective phase voltage; and wherein the voltage pulses of the switching devices (1a, 1b, 1c) are generated synchronously with one another and in each case rising and falling edges of the voltage pulses in each pulse half-period are temporally symmetrical with respect to a common center. 2. The method according to claim 1, wherein the primary side is fed by a three-phase AC system. 3. An electronic power converter having a polyphase primary side and a DC side, wherein • the power converter for each of the phases has a partial transformer (2a, 2b, 2c), wherein each of a primary side winding of a partial transformer (2a, 2b, 2c) to an associated switching device ( 1a, 1b, 1c) is connected and a sum voltage (u) of the partial transformers (2a, 2b, 2c) can be formed on a secondary side of the partial transformers (2a, 2b, 2c); • for each of the phases, the switching device (1a, 1b, 1c) is adapted to apply an associated phase voltage, positive or negative, or the voltage zero to the primary side winding of the associated partial transformer (2a, 2b, 2c); The power converter has a smoothing element (4) for smoothing a secondary-side current (i); and the power converter has a control device which is set up to control each of the switching devices (1a, 1b, 1c) by transmitting at a modulation frequency a sequence of voltage pulses with the alternating positive and negative respective phase voltage to the secondary side winding of the associated partial transformer ( 2a, 2b, 2c); Wherein a duration of each of the voltage pulses is proportional to the instantaneous magnitude of the respective phase voltage; and wherein the control action is further configured to generate voltage pulses of the switching devices (1a, 1b, 1c) in synchronism with each other and with respective rising and falling edges of the voltage pulses symmetrically with respect to a common center. 4. An electronic power converter according to claim 3, wherein the primary side has three phases. 5. An electronic power converter according to claim 3 or 4, wherein the power converter comprises a rectifier (3) for rectifying the sum voltage (u) before the smoothing of the secondary side current (i) by the smoothing element (4). 6. The electronic power converter according to claim 5, wherein the power converter has a smoothing capacitance (C) for smoothing an output voltage. 7. An electronic power converter according to claim 3 or 4, wherein the power converter feeds a serially resonant circuit whose inductance is formed by the smoothing element (4) and whose capacitance is formed by a resonant circuit capacitor (CR). 8. An electronic power converter according to one of claims 3 to 7, wherein the switching devices (1a, 1b, 1c) each have a diode bridge (11) as a rectifier and a subsequent bridge circuit (12) with active switches (T11, T12, T2i, T22) to apply the respective associated phase voltage or the negative phase voltage or the voltage zero to the primary side winding of the associated partial transformer (2a, 2b, 2c). 9. Electronic power converter according to one of claims 3 to 7, wherein the switching devices (1a, 1b, 1c) are each realized by a full bridge circuit with four-quadrant switches. 10. The electronic power converter according to claim 3, wherein the switching devices (1a, 1b, 1c) form a star connection with a first star point (15a), and input filter capacitances (Cf) form a further star circuit with a second star point (14). and the star points of the two star circuits are interconnected.