DE4344709C2 - Process for converting DC or AC voltages of different sizes into an arbitrarily specified voltage - Google Patents

Description

The invention relates to a method for converting equals or AC input voltages according to the preamble of claim 1, and an arrangement to implement the method.

For the energy transfer, in which different supply voltages with different frequencies in a uniform, from the feeding System potential separated supply voltage are to be converted have been using medium frequency transmission systems lately become known resonant switching converters. In doing so, or alternating voltages initially in a medium-frequency alternating voltage of, for example, 25 kHz and converted into on the consumer side different form provided. In converters, which are assigned to the consumers, the voltage required in each case and frequency changed one last time. Has this energy transfer The following advantages: Once everyone has potential isolation and Voltage adjustment required transformers as well as those in Power supply devices used inductors and capacitors the high frequency only a fraction of the volume, weight and thus the losses of 50 Hz components with the same power. also can at high frequency without affecting the dynamic Properties of controls switching points of the power semiconductors in the Zero crossings of currents or voltages of the medium frequency vibrations are placed.  

This resonant switching reduces the switching losses of the Power semiconductors and those based on switching operations electromagnetic interference significantly. In addition, the at non-resonant switching operations required wiring of the Power semiconductors saved.

By P. K. Sood and T. A. Lipo: "Power Conversion Distribution System Using a Resonant High Frequency AC Link ", IEEE-IAS, Annual Conference Record 1986, pp. 533 is a medium frequency transmission system according to the The preamble of claim 1 is known. Then DC voltages or alternating voltages of a first designated as an input circuit Circuit in which an inductor stores energy and the Rate of change of the input current limited, in thereof galvanically isolated direct or alternating voltages of a second as Output circuit designated circuit converted, the galvanic Separation takes place via a transformer element that with a parallel-connected parallel resonant circuit one between the input circuit and Output circuit lying AC voltage intermediate circuit forms. The Frequency of the intermediate circuit is determined by the resonance frequency of the Given parallel resonant circuit, which is much higher than the highest occurring frequency of the AC input voltage. Electronic Switches which convert to converters in the manner of an alternating or Direct converters are interconnected, control the power flow between input circuit and parallel resonant circuit.

To the parallel resonant circuit of such Medium frequency transmission system can have multiple consumers additional power converters can be connected. The one for every consumer individually required voltage and frequency can be determined according to the Pulse density modulation method by stringing and omitting are formed by half vibrations of the center frequency.

The underlying idea here is that half a period the high-frequency voltage as a unit for the synthesis of the mean a low frequency voltage is used to in an inductor to generate any desired current profile.  

Are the voltages of the input and output circuits alternating voltages, are the electronic switches bidirectional switches that for example from combinations of GTOS or transistors and diodes, or IGBTs and diodes.

The publication by O. Wasynczuk and P. C. Krause: "Simulation and Dynamic Performance of a 20 kHz Space Force Power System ", IEEE 1990, CH2881-1 / 90, pp. 343 shows a medium frequency transmission system to which as loads, a rectifier via a transformer and a medium fre quenz three-phase direct converter for feeding an induction motor are connected. The power feed into the medium frequency Power system with the parallel resonant circuit is carried out in pairs Output transformers connected in series as "mapham converters" designated thyristor inverters. Each pair of inverters feeds a series-connected filter power in the parallel resonant circuit. The Resonant circuit voltage and the level of the supply currents of the inverters Pairs are made by changing the phase angle of the partial inverters tensions between a mapham converter and the phase angle between the output currents of the inverter pairs.

A major disadvantage of this known medium frequency transmission systems is the limitation of the supply voltage of the input circuit the reverse voltage of the as an electronic switch or converter used power semiconductors with transmission powers in the 10 kW Area and above.

The limited input voltage leads to high levels of power Amperages and thus to complex parallel connections from Power semiconductors and large cable cross sections of connections cables and windings, which because of the high frequency from one elaborately manufactured and therefore expensive medium frequency braids exist. The realizable input voltage must be chosen the lower, depending higher the desired transmission frequency and thus the volume and Weight advantage for the electromagnetic components and condensates is fool. This is because, due to physical laws, the The higher the blocking voltages of the power semiconductors, the smaller the Switching speed and frequency is for which the power semiconductors be interpreted.  

The disadvantage of the limited input voltage becomes particularly serious, if, for example, the standardization of the supply voltage Mains voltage has a high value. For example, the Input voltages for the power supply of passenger coaches in the European rail network between 1000 and 5000 V. Basically Power converters for high input voltages with series connected Power semiconductors can be implemented, but the uniform must Voltage distribution on the power semiconductors during the Switching processes that take less than 1 µs due to complex Circuits are ensured.

The Lipo medium frequency transmission system has another Disadvantage in its control properties and those based thereon Impact on dimensioning is justified. With this principle sinusoidal currents in the supply lines to the supply network the method of pulse density modulation described by Lipo. The current in the inductances of the input lines increases, if the converter disconnects the parallel resonant circuit from the input and shorts the entrance. If, on the other hand, half vibrations of Resonant circuit voltage switched against the driving supply voltage, the input current decreases. This is the same method Relationship between the supply voltage and the medium frequency voltage on Parallel resonant circuit adjustable within limited limits. Rough changes the input voltages at, for example, approximately constant Medium frequency voltage can only be corrected using this method if the energy content of the parallel resonant circuit and the Input inductors is dimensioned so that it is a multiple of that in a Half period of the medium frequency transmissible energy. In the In practice, you have to expect ten times this amount of energy. At low input voltages are the times in the gaps between those fed into the input circuit by the converter Voltage half-oscillations a multiple of the half-oscillation period. In In these gaps, the parallel resonant circuit will have no energy over the Input circuit fed. The parallel resonant circuit therefore needs so much energy have saved that he without major voltage drop Consumers can dine in times of gaps.  

The demand for the coarse storage capacity of the input inductor activities follows from the limitation of the ripple of the input current.

In the Wasynczuk and Krause transmission system, the Setting the voltage on the parallel resonant circuit and the supply currents about the angular position of the vibrations in the partial inverters and quasi steadily within each half period. The energy content of the Parallel resonant circuit can therefore be dimensioned smaller than that Transmission system according to Lipo, but this is the converter effort significantly higher. In addition, the Mapham converter must be connected to a Voltage source with a low input impedance must be connected. The desired low harmonic feed currents are with a Inductance in the input circuit easier to implement.

The invention is therefore based on the object, a method and a Arrangement for carrying out the method by the preamble of To create claim 1 marked type, the Energy content of the parallel resonant circuit and the input inductance and thus the volume, weight and losses of these components limited fluctuations in the resonant circuit voltage as small as possible should stay. The harmonics of the input current should also stay limited.

This object is achieved according to the invention by the in the characteristics of Features listed claim 1 solved.

Further training and special designs, as well as circuits for Implementation of the method are specified in the subclaims.

The following advantages are achieved with the invention:

• - The high input voltage is shared by the electronic switches in exactly the same partial voltages without wiring measures.
• - The power semiconductors occur after switching off the current no sudden voltage changes on what losses and cause electromagnetic interference. So there are none Wiring measures to limit the voltage rise rate required.  
• - The means available for the potential-free transmission of power frequency transformers are also used in a particularly advantageous manner for transferring the power semiconductors to different Potential control power used.
• - The supply voltages can differ from DC or AC voltages their frequencies and high different voltage values have a multiple of the reverse voltage of the fast switching Power semiconductors can be.

The invention is explained below with reference to the drawing.

It shows:

Figure 1 shows a generalized switching principle of an arrangement for carrying out the method.

Figure 2 shows the assignment of the symbols of the generalized switching principle to the components of an embodiment of a converter arrangement.

Fig. 3 is the used to explain the operation of the switching principle of the transducer assembly;

FIG. 4 shows a control process by the method of pulse density modulation, caused by an increase in the input voltage;

Figure 5 shows the pulse density modulation in the regulation of the input current which is to take the sinusoidal waveform of the line voltage.

Fig. 6 shows another example of a switching principle of the transducer arrangement according to the invention, in which a single-phase input AC voltage is converted into a DC output voltage via a series-connected direct converter with transformer center circuit;

Fig. 7 is a equipped with a particularly advantageous auxiliary power converter arrangement which allows the supply of the medium-frequency intermediate circuit from the output circuit and an auxiliary voltage source, and

Fig. 8 is a equipped with a switching converter array which allows a selectable series and parallel connection of sub-transducers with system-related changes in the mains voltage.

In the generalized switching principle according to FIG. 1, the input voltage U _E and the output voltage U _{A can be} DC voltages of any polarity or AC voltages. The input current i _E and the output current i _A can be direct currents in any direction or alternating currents. It must be noted that the generalized switches, S₁₁-S _1n and S _A1 -S _{A4 are} electronic switches such as power semiconductors, which block voltages and conduct currents in the required directions. In principle, therefore, the power flow between the input voltage U _E and the output voltage U _{A is} possible in both directions.

When using bidirectional semiconductor switches, the z. B. according to Fig. 2e consist of two diodes and transistors or IGBTs, according to the inventive method with the same arrangement, all conversions between the different voltage types are possible with any direction of power flow.

As shown in FIG. 1, the converter arrangement can be divided into four parts: the input circuit EK, the intermediate circuit ZK, the output circuit AK and the control and regulating unit SR. First, the input circuit EK is described. It consists of the input voltage source U _E, an input choke L _E and a number n of switch groups B _i connected in series, electronic switches S _i1 -S _{i4 being connected together} to form bridge-type converters in the manner of an inverter or direct converter. Primary windings w _i of transformers TR _{i are} connected to the outputs A _i1 , A _{i2 of} the converter. The switches of the converters can pass the input current i _E through the primary windings of the transformers or past them.

For example, if the switches on the left or right side of a converter bridge B _{i are} closed, the current is conducted past the corresponding winding W _i . This is the case for example with B₂ in Fig. 1. At B _n in Fig. 1, however, the switches are alternately opened or closed. As a result, current i _E flows through winding W _n .

The intermediate circuit ZK consists essentially of the parallel resonant circuit C _P and L _P and the secondary windings W _{S1 of} the transformers connected in parallel. The vibrations of the parallel resonant circuit are detected by the control and regulating unit SR as voltage U _s and current i _SM and used for the synchronous control of the switches on the input side of the transformer element. In a first approximation, the temporal mean value of the input current i _E is constant or changes only slowly over the duration of a half-oscillation of the parallel resonant circuit. Preferably, the transformer element, as shown in Fig. 1, consists of a number of identical transformers, but in a further advantageous embodiment of the invention, all n transformers could also be combined to form a transformer with n primary and one secondary winding.

The output circuit A _K consists of the electronic switches S _A1 -S _A4 connected to a converter B _A in a bridge circuit, an output choke L _A connected in series with the converter and the output voltage source. The output voltage U _A is z. B. in the power flow direction from the input voltage source to the output voltage source, the counter voltage of the load. The control and regulating unit SR detects, in addition to the input voltage U _E and the output voltage U _A, further measured variables which represent the internal state of the converter, in accordance with the control task. It is also supplied with the setpoint variable U _{AS of} the output voltage U _A. Depending on these variables, the switches S₁₁-S _{n4 of} the converters are actuated via the control signals K₁₁-K _{n4 in such a} way that the number α of the converters feeding into the parallel resonant circuit is changed on average over time so that the desired resonant circuit voltage U _S and output voltage U _A and a desired time course, for. B. set a sinusoidal shape of the input current i _E.

In cases where a variable ratio between the resonant circuit voltage and output voltage, an AC output voltage or a change in the direction of energy flow occur, the electronic switches S _A1 -S _{A4 of} the output converter B _A from the control and regulating unit S _R via the control signals K _A1 -K _A2 controlled.

FIGS. 2a to 2d arrange the symbols of the general principle diagram of Fig. 1 for the most common application, the conversion differed Licher DC and AC voltages to an output DC voltage with directional power flow only to the latter, the functions of real circuit elements.

According to FIG. 2a, the input voltage U _E is a DC voltage of a certain polarity which forms a bridge rectifier from AC voltage or DC voltages of either polarity.

In part b of the figure, the switches S₁₁. . . S _{n4 exemplified} as IGBTs with upstream blocking diodes. The blocking diodes are necessary because the IGBTs cannot absorb any voltage against the switching direction.

Fig. 2c shows a diode as an example of one of the switches S _A1 to S _A4, which are connected in the output circuit AK for the formation of a DC output voltage to a rectifier bridge circuit B _A.

According to FIG. 2d, the output DC voltage U _{A is formed} across the load resistor R _A by smoothing the output DC current i _A with the aid of the buffer capacitor C _A.

With these assignments, the simplified switching principle shown in FIG. 3 is created from FIG. 1 and is used to further explain the mode of operation. According to the switching principle shown here, a regulated DC output voltage U _{A is} formed from a high, strongly fluctuating input voltage U _E. With sinusoidal input voltage | U _E | the input current i _E can also be modulated sinusoidally, so that the converter loads the supply network like an ohmic consumer. For the high resonant circuit frequency of 25 kHz, for example, an input voltage U _E that changes sinusoidally with the mains frequency is to be treated like a slowly changing DC voltage.

The losses of the converter are taken into account in FIG. 3 by the ohmic resistances R _VE in the input circuit and R _VA in the output circuit.

In the example shown, n = 3 converters are connected in series in the input circuit and connected via n = 3 identical transformers with the transmission ratio ü = W _i / W _si to the parallel resonant circuit L _P , C _P. In the example shown, the output circuit consists of a rectifier bridge BA with the smoothing inductor L _A and a buffer capacitor C _A , the capacitance of which is dimensioned so roughly that ripples in the rectifier output current i _A and the load current i _LA do not cause any significant change in the DC output voltage U _A. With reference to FIG. 3 and the diagrams in FIG. 4, it is now shown that the control of the power flow over several converters B 1 connected in series in the input circuit EK. . . B _n is carried out in such a way that the converters selectively and independently feed the common input current i _E either independently via a separate winding of the transformer element TR in parallel into the parallel resonant circuit or bypass the primary windings and the parallel resonant circuit.

The number α of the converters feeding into the parallel resonant circuit is changed as a function of the input voltage U _{E over} time so that the desired output voltage U _{A is} set.

The diagram a in FIG. 4 shows the input voltage U _E with an arbitrary change ΔU _E and an assumed input current i _E which changes by the amount Δi _E as a result of the change in the input voltage with constant output variables U _A and i _LA . Diagram 4b shows the time course of the resonant circuit voltage U _s ′ = Us · ü transmitted to the primary side of the transformers TR. The amplitude of the resonant circuit voltage deviates only temporarily by the amount ΔÛ _s 'from the otherwise constant amplitude Û _s '.

The diagrams c, d and e of FIG. 4 show the bridge B1, B2, B3 of FIG. 3 inserted into the input circuit EK voltages U 1, U 2, U 3, which are composed of sequences of half-oscillations of the resonant circuit voltage. By activating the switches S _i1 -S _{i4 of} a converter bridge, it is determined whether the voltage U _i inserted by the converter into the input circuit is positive for the duration of a half-oscillation, as shown in FIG. 4, and is therefore opposed to the input current i _E , so that energy is supplied to the parallel resonant circuit, or whether this voltage U _{i is} equal to zero.

In the latter case, the input current i _E is conducted past the associated transformer TR _i by two conductive switches located one below the other. In principle, negative half-oscillations of the voltages U _i can also be inserted into the input circuit, energy being removed from the oscillating circuit and being supplied to the input inductor L _E. The switching of the electronic switch S₁₁-S _n4 is preferably as close as possible to the zero crossing of the resonant circuit _voltage . This avoids switching losses in the power semiconductors and electromagnetic interference which occurs when switching higher voltages. After a semiconductor switch has been switched off, the voltage to be picked up by it in the forward or conducting direction increases only slowly with the rate of rise du / dt predetermined by the resonant circuit voltage.

This avoids the risk of faulty switch-on due to excessively high du / dt values in certain semiconductor switches such as GTOs and thyristors and the losses caused by "aftercurrents" (or tail currents) caused by current flow are kept low. The parallel connection of all transformer secondary sides to the parallel resonant circuit also has the effect that the same voltage u · U _{s is} transmitted in all partial converters and the same reverse voltages occur at the electronic switches of the partial converters.

The sum of the voltages inserted into the input circuit by the partial converters forms the counter voltage U _g = U₁ + U₂ +. . . U _i to the input voltage U _E. If the number of the partial converters feeding into the parallel resonant circuit and causing a counter voltage is designated by α and the exact switching of the power semiconductors at zero crossing of the resonant circuit voltage U _{s is} required, the arithmetic mean of the counter voltage is Ug during a half oscillation

U _g = Ûs · α · ü · 2 / π.

The input inductance L _E is dimensioned such that the input current i _E can only change significantly in the course of several half oscillations. For the change of the input current, only the value _{g of} the counter voltage U _g averaged over several semi-oscillations is decisive.

Assuming that the amplitude change ΔÛ _s of the oscillating circuit voltage remains limited to small values in dynamic processes, the arithmetic mean of the number of partial converters feeding into the oscillating circuit determines the mean counter voltage

_g = Û _s ·· ü · 2 / π.

Diagram f in FIG. 4 shows the course of the counter-voltage U _g and the mean value _g thereof in broken lines.

In the steady state, the arithmetic mean value of the input current _{E is} constant and the mean voltage value _L = 0 at the input inductance LE, so that the following applies to the voltage of the input circuit:

_E = ( _E · R _E + Û _s ·· ü · 2) / π.

The corresponding equation for the output circuit is:

U _A = Û _s · 2 / π- _A · R _A.

With a constant amplitude in the parallel resonant circuit, the power supplied to it must be equal to the power dissipated over time. The relationship between the mean value of input current i _E and output current i _A follows:

_A = _E · ü ·.

By eliminating the resonant circuit voltage U _s and the output current _A , the relationship from the last three equations for the input current _{E is obtained} :

_E (ü² · ² · R _A + R _E ) = _E -ü U _A.

The left side of this equation expresses the voltage drop across the converter in the steady state and is usually a small quantity compared to the voltages on the right side. With constant input voltage U _E and constant output voltage U _A, small changes are therefore sufficient to adjust the input current i _E and thus also the power to be transmitted from the input to the output in a larger range.

If, on the other hand, the input voltage U _{E is} variable, neglecting the current-dependent influence from the right-hand side follows the above equation for the average number of partial converters feeding into the parallel resonant circuit

The diagrams in FIG. 4 each show a steady state for two different values of the input voltage U _E at the beginning and end of the time scale. At the smaller value at the beginning of the time scale, the values α = 2 and α = 1 alternate, so that ₁ = 1.5. Long after the input voltage increases at the end of the time scale, α = 3 and α = 2 partial inverters alternately feed into the parallel resonant circuit, the mean value is now ₂ = 2.5.

Since the power to be transmitted remains unchanged when the input voltage is increased by ΔU _E , the input current decreases by Δi _E. According to the law of induction, for every current change in an inductance, a voltage time area of the appropriate size must be present there. This voltage time area is in FIG. 4f

∫ U _L dt = Δi _E · L _E

hatched. With the partial voltages in the input circuit, this is Equation:

∫ (U _E -i _E · R _VE -α Û _s · ü · 2 / π) dt = Δi _E · L _E.

This relationship and the diagram f in FIG. 4 show that the number α of the partial converter feeding into the parallel resonant circuit has to be changed temporarily in order to generate the voltage time area at the inductor L _E required for the current change Δi _E. Since the current i _E · ü · α fed into the oscillating circuit also changes with α until a new steady state has occurred, a deviation Δs in the oscillating circuit amplitude occurs temporarily.

The number α of the partial converters, which insert the resonant circuit voltage U _s with the factor α · ü into the input circuit, is the manipulated variable of the converter method.

The arithmetic mean determines the ratio of the input voltage U _E to the output voltage U _A and the instantaneous value α the change in the input current during the half-oscillation period. With a sufficiently large capacitance C _A at the output of the converter or a voltage source at the output, the voltage of which changes only slowly compared to the medium-frequency voltage at the parallel resonant circuit, it is therefore possible to determine the temporal course of the input current by means of dynamic changes in the quantity α.

If, for example, the input voltage U _{E has} the time profile of a rectified sinusoidal mains voltage with the angular frequency Ω and the output DC voltage U _{A is to} remain constant, then the control and regulating device SR outputs when the converter is not under load, ie the output current i _A and the input current i _E are approximately Zero, the mean values depending on the output voltage U _A and the input voltage U _E according to the following relationship:

If the voltage U _A deviates from the output voltage setpoint U _AS given to the control and regulating device, for example as a result of a current i _{LA loading} the output, then the control and regulating device SR makes a small change Δ in the value compared to the value occurring during idling, so that a current is transmitted from the input circuit to the output circuit until the deviation of the output voltage U _A from its target value U _AS becomes a minimum. In the case of sinusoidal input voltage U _E , a sinusoidal course of the input current i _E is generally also desirable. The control and regulating device SR therefore influences the change Δ in its time profile, for example with the aid of a current control subordinate to the output voltage control, in such a way that the input current i _{E has the} following time profile:

_E = | _E · sin Ωt |.

The temporal course of the current _{A is} calculated from the last two relationships and the equation for the mean value _{A of} the current in the output circuit or from the equality of the input and output power

Fig. 5 shows an example of the modulation of a matching with the time course of the input voltage U _E rectified sinusför-shaped input current i _E.

In diagram 5a, the resonant circuit voltage U _s and in diagrams 5b, 5c and 5d the voltages U 1, U 2 and U 3 formed as a result of the modulation on the partial inverters are shown. The sum of the voltages U₁ + U₂ + U₃ gives the course of the voltage U _G shown in diagram 5e. If the ohmic voltage drop U _R in FIG. 3 is neglected, its arithmetic mean corresponds to the input voltage U _E shown in broken lines, which, like the input current i _{E, has} the shape of a rectified sine function. Diagram 5f shows the output current i _{A of} the rectifier B _A in FIG. 3. It has twice the frequency with a constant output voltage compared to the mains voltage and the mains current. The mean value of i _A corresponds to the constant load current i _LA .

The converter method according to the invention was explained on arrangements according to FIGS . 1 and 3, in which in the input circuit EK n converters are connected in series in a bridge circuit. A modified method is described with reference to FIG. 6, which is based on the series connection of converters, combined with transformers with center tap on the converter side.

The transformers TR _i connected in parallel on the secondary side and feeding the resonant circuit have on the primary side two identical windings W _i1 , W _i2 connected in series, at the connection point of which the input current i _E common to all converters B _i , i = 1-n is supplied. Via the electronic switches connected to the outer ends of the primary windings, the current path to the transformer center tap on the primary side of the subsequent converter connected in series is switched.

In the direction assumed in Fig. 6 of the input current i _E , this flows either over the upper primary winding W _i1 when the switch S _i1 , or over the lower primary winding W₁ when the switch S _{i2 is} closed. The instantaneous resonant circuit voltage U _s ′, transformed from the secondary side to the primary side, can counteract the input current i _E , for example at the upper partial windings W _i1 , or, as with the lower partial windings W _i2, have a driving effect in the direction of the input current. According to these assignments of the current and voltage directions, energy is either transferred to the parallel resonant circuit or withdrawn from the parallel resonant circuit. For the resulting number α determining the voltage ratio, the converters feeding into the parallel resonant circuit now have α = n-2e where n is the total number of existing converters with transformers in the center circuit and e is the number of converters that draw energy from the resonant circuit.

The arrangement shown in Fig. 6 has in the input circuit EK an AC voltage source with the voltage U _E , the frequency of which is significantly lower than the converter frequency determined by the resonance frequency of the parallel resonant circuit and is, for example, 50 Hz or 16 2/2 Hz Fig. 6 no rectifier is present, also of the converters and transformers supplied input current i _e changes its direction at the low frequency of the voltage U _e. The electronic switches connected to the external connections of the transformer primary windings must therefore be suitable for switching currents in both directions of current flow. While, for example, for the direction of the current i _E shown in FIG. 6, the switches S _i1 and S _i2 in combination with diodes connected in series according to FIG. 2b alone enable the method according to the invention to be carried out, for the opposite direction of the converter i _E additional switches S _i3 , S _i4 and diodes can be arranged in accordance with FIG. 2e. This creates bidirectional semiconductor switches.

As is well known, electronic switches both Switching information, which determine the times of the switch as also a certain power or energy can be supplied to one Perform switching operation or a switching state, such as. B. the to maintain the conductive state of a transistor or IGBT.  

As is generally known, the switching information and the control power can either be transmitted to the electronic switches as uniform control signals, for example as pulses, with a sufficiently high electrical power via the same lines and electrical isolation transformers, or in physically separate ways. The almost instantaneous transmission of the switching information via optical fibers and the transmission of the control power via transformers are an example of physically separate transmission paths to the electronic switches S _i1 -S _{i4 of} the converters B connected in series. For the implementation of the converter process, the distinction between uniform and separate Transmission of switching information and control power without meaning. The signals K₁₁-K _n4 of Fig. 1, 3 and 5 therefore only show the switching information.

However, as the exemplary embodiment in FIG. 7 shows, in connection with the converter method according to the invention there is a particularly advantageous method for transmitting the control power, which makes the generally complex isolation of control power transformers superfluous. Here, the power required within a converter B _i for controlling the electronic switches S _i1 -S _i4 at the connections A _i1 , A _{i2 of} the transformers TR _i feeding into the parallel resonant circuit is attacked and in amplifier stages V _i1 -V _i4 , which control the z. B. amplified by optical fiber switching information K _i1 -K _{i4 of} the electronic switch, fed via isolation transformers T _i1 -T _i4 . The isolation of these electrical isolation transformers only has to be dimensioned for the potential differences occurring at the electronic switches within a converter.

In the control power supply shown in FIG. 7, the control power is withdrawn from the parallel resonant circuit L _P , C _P via the transformers TR _i , which also ensure the insulation for the maximum potential differences occurring in the series connection of all converters. It should be noted that the parallel resonant circuit, in assemblies, for example according to Fig. 3 or 6, as the diagrams in Fig. 5e and Fig. 5f show, no power is supplied to the zero crossings of the input voltage U _E and the input current i _E and therefore no power can be withdrawn without the vibration subsiding.

The oscillation in the parallel resonant circuit has to remain at least for a limited period of time with an approximately constant amplitude of the voltage U _s when the power supply from the input circuit is interrupted if the parallel resonant circuit through the amplifier stages V _i1 -V _i4 , (i = 1-n) and further consumer VB energy connected to the collective power SL (see FIG. 7) is withdrawn. Therefore, an auxiliary inverter BH consisting of controllable power semiconductors SH₁-SH₄ feeds the required differential power from the output capacitor C _A into the parallel resonant circuit.

The auxiliary inverter BH can also be controlled via the control signals KH₁-KH₄, in particular by changing the phase angle between these signals and the resonant circuit voltage U so that it always takes a current i _H from the output capacitor C _A via the inductors L _H / 2 connected in series and feeds into the parallel resonant circuit. It then flows back of the output current i _A via the output converter B _A and the inductors L _A / 2 to the output capacitor C _A as a subcomponent. The circuit currents i _K1 , i _{K2 formed in this way} ensure a continuous current i _A via the output rectifier B _A, and thus a constant, load-independent ratio U _s / U _A = π / 2 between the resonant circuit voltage U, even when the load current i _LA approaches zero _s and the output voltage U _A.

The auxiliary inverter is also used to start up the converter when the control power for the electronic switches of the converters B _i , as shown above, is taken from the parallel resonant circuit via the transformers TR _i . For this purpose, the output capacitor C _{A is} charged via a switching device S _u and a resistor R _H by an auxiliary voltage U _H until the control and regulating device sets the auxiliary inverter B _H via the signals KH 1 -KH-into operation. This initiates the oscillation of the parallel resonant circuit and the voltage supply to the amplifier stages V _i1 -V _i4 for controlling the converter B _i . In order to get by with the smallest possible charging current i _R , the load current i _LA flowing via contact 1 of the switching device S _{u is} interrupted during the charging of the capacitor via contact 2 .

In principle, it is also possible to charge the capacitor C _A instead of an auxiliary voltage from the input voltage U _E for starting up the converter.

Since it is important for applications in different countries that power supply units automatically switch to the other mains voltage, a switch is provided here which switches the individual switch bridges in parallel and in succession in different ways. The circuit type and the corresponding switch positions are shown schematically in FIG. 8. Depending on the position of the switches UK₁, UK₂, UK₃, six of these modules B1. . . Connect B6 in series or connect two in parallel or three in parallel. If you use three diodes RD1, RD2, RD3, you can save yourself a pair of contacts with the switches.

Through the additionally provided switch UK₁, UK₂, UK₃ an adaptation to a system-related change in the mains voltage U _{o is} carried out. The total input voltage present at the switch bridges (U _E ) is divided into a series parallel connection with a corresponding number of switch bridges so that after the adjustment of the voltage drop across each switch bridge (B _i ) has the same order of magnitude as before the switchover, the switch bridges with respect to the series or parallel connection can be combined into groups of the same number of switch bridges.

The changeover of the number of bridges connected in parallel and in series during the transition to a different mains voltage is carried out when the voltage U _{E has changed,} for example, by more than 100 V over a time of at least 10 ms. The changeover is triggered by a voltage detector and the adaptation to the new mains voltage is carried out according to the result of the evaluation in a discriminator by triggering a switching mechanism.

Claims (19)

1. Method for converting different sized direct or alternating voltages (U _E ) of a first circuit designated as an input circuit (EK), in which an inductance (L _E ) stores energy and limits the rate of change of the input current (i _E ). in it galvanically isolated direct or alternating voltages (U _A ) of a second circuit designated as an output circuit (AK), the galvanic isolation via a transformer element (TR) which follows with a parallel resonant circuit (L _P , C _P ) connected in parallel between the input circuit and output circuit lying AC voltage intermediate circuit (ZK), whose frequency is given by the resonance frequency of the parallel resonant circuit and which is substantially higher than the highest frequency of the input AC voltage (U _E ), and wherein electronic switches (S₁₁.. S _n4 ), which to n converters (B₁.. B _i ... B _n ) in the manner of an alternating or direct mrichters are interconnected, control the power flow between the input circuit and the parallel resonant circuit, characterized in that
that the power flow is controlled via several converters (B 1... B _n ) connected in series in the input circuit (EK) in such a way that
that the converters selectively and independently feed the common input current (i _E ) either in parallel to the parallel resonant circuit via separate windings of the transformer element (TR) or bypass the primary transformer windings and the parallel resonant circuit, and that the number (α) of those in the parallel resonant circuit feeding converter is changed as a function of the input voltage (U _E ) over time,
that the desired resonant circuit voltage (U _s ) or output voltage (U _A ) is set. 2. The method according to claim 1, characterized in that the temporal course of the input current (i _E ) by deviations (Δ) formed in a control and regulating device (SR) of the temporal mean () of the number (α) of those feeding into the resonant circuit Converter is determined by the ratio of the input voltage (U _E ) to the output voltage (U _A ). 3. The method according to claim 1 and 2, characterized in that the control and regulating device (SR) influences the number (α) of the converter feeding into the resonant circuit in such a way that the input current (i _E ) has the same time profile as that Input voltage (U _E ) assumes. 4. The method according to any one of claims 1 to 3, characterized in that the amplitudes of transients are limited by the change in the number (α) of the converter (ZK) feeding in the converter (B) in the control and regulating unit (SR) _i ) is limited in the unit of time. 5. The method according to any one of claims 1 to 4, characterized in that the control and regulating device (SR), the electronic switches (S _i1 -S _i4 ; i = 1 _... n) at the times of the zero crossings of the resonant circuit voltage (U _s ) switches on and off. 6. The method according to any one of claims 1 to 4, characterized, that the control and regulating device (SR) for stepless regulation the phase voltage of the switching times compared to the output voltage the zero crossings of the resonant circuit voltage by one changeable phase angle shifts.   7. The method according to any one of claims 1 to 6, characterized in that the control of the power flow over several in the input circuit (EK) series-connected converters (B₁... B _n ) and transformers (TR₁... TR _n ) with center tap at the input winding in such a way that the electronic switches (S₁₁.. S _n4 ) of the converter selectively and independently of each other the common input current (i _E ) supplied to each center tap of the transformers via one of the two input _{partial windings} (W _i1 , W _i2 ) conduct and thereby supply or withdraw energy from the parallel resonant circuit via the transformers, depending on whether the voltage U _s ′ of the parallel resonant circuit transformed on the transformer input side is opposite or rectified to the input current, with the resultant number (α) of the converters α feeding into the parallel resonant circuit = n-2 · e, and where n is the total number of currents mrichter and e is the number of converters that draw energy from the parallel resonant circuit. 8. The method according to any one of claims 1 to 7, characterized in that the power required to control the electronic switches (S _{i1 ...} S _i4 ) of the converter (B _i ) at the converter-side connections (A _i1 , A _i2 ) tapped into the parallel resonant circuit transformers (TR _i ) and in amplifier stages (V _i1... V _in ), which amplify the control signals (K _i1... K _i4 ) of the electronic switches, via potential _{isolating transformers} (T _i1... T _i4 ) is fed. 9. The method according to any one of claims 1 to 8, characterized in that when suspending the energy supply from the input circuit to the parallel resonant circuit, for example in the zero crossing of the input current (i _E ) with sinusoidal input voltage (U _E ), one of controllable power semiconductors (SH₁... SH₄) existing auxiliary inverter (B _H) power draws the charged to an output voltage (U _a) output capacitor (L _a) and in the parallel resonant circuit, so that even in the absence of power supply from the input circuit, the voltage (U _s) feeds the parallel resonant circuit as a continuous and almost constant supply voltage for the amplifier stages (V _i1... V _i4 , i = 1-n), the control and regulating device (SR) or other consumers (VB) connected to the parallel resonant circuit via the bus (SL) remains. 10. The method according to claim 9, characterized in that when the load current (i _LA ) of the output circuit of the auxiliary inverter (B _H ) approaches zero, the output capacitor (LC) always takes as much power and thus a current (i _H ) and feeds it back into the parallel resonant circuit that circulating currents (i _K1 , i _K2 ) flow through the auxiliary inverter and the output converter, as well as through the inductances (L _{H / 2} , L _{A / 2} ) connected in series to smooth the currents (i _H , i _A ), which ensure a continuous current (i _A ) via the output converter (B _A ) and thus a constant, load-independent ratio between the resonant circuit voltage (U _s ) and the output voltage (U _A ). 11. The method according to claims 7 and 8, characterized in that for starting the output capacitor (L _A ) via a switch (S _u ) and a resistor (R _H ) from an auxiliary voltage source (U _H ) is charged, the switching device at the same time prevents the flow of a load current (i _LA ) and discharging of the output capacitor (C _A ), and that after the output capacitor is charged to a sufficiently high output voltage (U _A ), the control and regulating device (SR) via signals (K _H1 . ... K _H4 ) first starts the auxiliary inverter (B _H ) and thus initiates the oscillation of the parallel resonant circuit and the voltage supply to the amplifier stages (V _{i1 ...} V _in ) for controlling the power semiconductors of the converters (B _i ) of the input circuit. 12. The method according to any one of claims 1 to 11, characterized in that the primary winding in the diagonal of the associated switch bridge (B _i ), which has contactless switches (S _i1 , S _i2 , S _i3 , S _i4 ) by crosswise closing the Switch is turned on, the current direction is changed by changing the bridge branches. 13. The method according to any one of claims 1 to 12, characterized in that the switches (S _i1 , S _i2 , S _i3 , S _i4 ) are actuated to switch off the primary winding i in such a way that the current is conducted past the winding in question. 14. The method according to claim 1, characterized in that by additionally provided switches (UK₁, UK₂, UK₃) to adapt to a system-related change in the mains voltage U _o the total applied to the switch bridges input voltage (U _E ) on a series parallel connection with accordingly numerous switch bridges is divided so that after the adjustment of the voltage drop across each switch bridge (B _i ) has the same order of magnitude as before the switchover, and that the switch bridges are combined into groups of the same number of switch bridges with respect to series or parallel connection. 15. The method according to claim 14, characterized in that the switching of the number of bridges connected in parallel and in series is carried out at the transition to a different mains voltage if the voltage U _{E has changed} by more than 100 V over a time of at least 10 ms . 16. The method according to claim 14 or 15, characterized, that the switch is triggered by a voltage detector and the Adaptation to the new mains voltage according to the result of the evaluation device in a discriminator by controlling a switching mechanism is carried out.   17. Arrangement for converting different coarse direct or alternating voltages (U _E ) of a first circuit designated as an input circuit (EK), in which an inductance (L _E ) stores energy and limits the rate of change of the input current (i _E ), in it Galvanically isolated direct or alternating voltages (U _A ) of a second circuit designated as an output circuit (AK), the galvanic isolation being carried out via a transformer element (TR) which has a parallel resonant circuit (L _P , C _P ) connected between the input circuit and Output circuit lying AC voltage intermediate circuit (ZK), the frequency of which is given by the resonance frequency of the parallel resonant circuit and is at least one order of magnitude higher than the highest frequency before the incoming AC voltage (U _E ), and wherein contactless switches (S₁₁.. S₁₄), which to n Power converters (B₁... B _i ... B _n ) in the manner of a Wechselric hters or direct converter are interconnected, control the power flow between the input circuit and the parallel resonant circuit via the transformer element (TR), for carrying out the method according to one of claims 1 to 13, characterized in that switch bridges (B _i ) are provided which are provided by additional switches (U _K ) in groups of at least two series-connected switch bridges (B _i ) can be connected in parallel and / or in series, the primary windings (W 1. . . W _n ) of a transformer element (TR), the secondary side of which feeds energy into a parallel resonant circuit, and that a circuit is provided which enables the decoupling of energy to the load L. 18. Arrangement for performing the method according to one of the claims 1 to 17, characterized, that the transformer element (TR) from a transformer with n There are primary windings and a secondary winding. 19. Arrangement for performing the method according to one of the claims 1 to 18, characterized, that the transformer element (TR) from n transformers with parallel switched secondary windings.