WO2004006416A1

WO2004006416A1 - A power supply

Info

Publication number: WO2004006416A1
Application number: PCT/DK2003/000458
Authority: WO
Inventors: Nils-Ole Harvest
Original assignee: Danfoss Drives AS
Current assignee: Danfoss Power Electronics AS
Priority date: 2002-07-04
Filing date: 2003-07-02
Publication date: 2004-01-15
Anticipated expiration: 2005-01-04
Also published as: AU2003243927A1; DK174716B1

Abstract

The invention concerns a power supply comprising an electric converter, primarily for converting AC to DC. The converter comprises a self-induction, a rectifier element and a switch member controlled by a control circuit, and preferably a capacitor. The control circuit controls the switch member on the basis of a reference and a measurement of a voltage at a given location in the power supply and comprises a filter with a first cut-off frequency for increasing voltage and a second cut-off frequency, different from the first cut-off frequency, for decreasing voltage.

Description

A power supply

The present invention concerns a power supply comprising an electric converter, primarily for converting AC to DC, which converter comprises a self- induction, a rectifier element and a switch member controlled by a control cir- cuit, which controls the switch member on the basis of a reference and a measurement of a voltage at a given location in the power supply, a method for controlling a power supply, and the use of such a power supply for motor control.

For the current supply of many electrical devices, mainly devices connected by one phase to the mains, circuits are used, which correct the power factor, the so-called PFC circuits. When using such circuits, a substantially sine-shaped current can be drawn from the mains, that is, a current with a low content of harmonics, which today is required by standards. Further, with such a PFC- circuit, a power supply can be achieved, which is relatively independent of the —supply voltage, so that with one and the same power supply a unit can be realized, which can be connected directly to different supply voltages, for example 110 V or 240 V, without requiring a switch action. An example of such a PFC- circuit can be found in US-B1 -6,215,287, in which the PFC-circuit is based on a forward converter circuit in the form of a boost converter circuit.

In such boost converter circuits, as described in US-B1-6,215,287, a transistor is ON -OFF controlled with a switch frequency that is substantially higher than the mains frequency, for example 50 to 100 times higher than the mains fre- quency, which is typically 50 Hz or 60 Hz. When the transistor is ON, the current flows from the mains through a rectifier bridge, a self-induction and the transistor, so that energy is stored in the self-induction. When the transistor is OFF, the current flows from the mains through the rectifier bridge, the self- induction, a rectifier element in the form of a diode and a load. During this course, the current in the self-induction decreases, while supplying its energy to the load. To ensure that the load can always be supplied with the required instant power, independently of whether or not the transistor is ON, an inter- mediary circuit capacitor is normally used, which is connected in parallel with the load.

The control of the transistor can, for example, be a pulse width modulation, as described in US-B1 -6,215,287. The pulse width modulation is typically controlled by a cascade control, that is, a control comprising two loops.

One loop is a fast-controlling inner loop, which controls the current in the mains, for example, by controlling the current in the self-induction. This control can be made on the basis of a measuring resistor inserted at a suitable location in the circuit, for example between the rectifier bridge and the transistor emitter, or, as shown in DE-C1-33 13 124, in series with the self-induction. In this connection, it must be noted that it has no influence on the measurement that DE-C1-33 13 124 concerns a different type of forward converter. Alterna- tively, like in US-B1-6,215,287, a not clearly defined detection device can be used, which measures on the phase or the 0-conductor on the mains side of the rectifier bridge.

The second loop is a slow-regulating outer loop, which controls the output volt- age of the circuit, that is, in practice, the voltage across the load including the conductors to the load. To obtain a sine-shaped current in the mains, the output signal from the outer control loop is multiplied by the instant value of the mains voltage, for example based on a measurement across the rectifier bridge output, and the result of this multiplication is used as a reference for the inner control loop.

Further to this, a PFC-circuit can typically comprise compensation for variations on mains voltage and load, and monitoring functions for overload and overvoltage.

As mentioned, the inner control loop is fast-regulating, which, in this connection, means that it controls substantially faster than the mains frequency, but still slower than the switch frequency of the transistor. Typically, this control is a fast l-control. In comparison, the outer control loop is slow, that is, substantially slower than the mains frequency. If this control is chosen too fast, it will equalise the ripple voltage, which will otherwise, due to the limited size of the intermediary circuit capacitor, inevitably occur on the output voltage. This would cause that the input current is not sine-shaped, but distorted, and thus contains exactly the harmonics, which are to be avoided by using the PFC- circuit.

However, the PFC-circuits of the state of the art have certain disadvantages.

For example, a sudden drop in the load, for instance its disconnection, will give rise to a hefty overshoot of the output voltage. Such an excess voltage from an overshoot can cause an unwanted damage to the circuit components if these are not over-rated with regard to voltage, which would of course not be desirable for economic reasons, as this makes the components more expensive.

This problem is dealt with in US-BI-6,215,285, which proposes a solution, in which a low-pass filter with variable cut-off frequency is inserted in the control loop. However, this solution is unnecessarily complicated, as it requires adjustable filter functionality. Another problem is that instabilities may occur in the slow control loop, if the load contains disturbing frequencies, for example from an inverter, which is loaded by a motor. As it must be possible to vary the motor speed at random by means of the inverter, such a load may contain ran- dom frequencies. This is particularly a problem, if these motor related frequencies and the mains frequency, or multiples of these, are close to each other, so that low-frequent mixed frequencies occur. Such problems can be further aggravated by the fact that motors normally have inherent resonant frequencies, some of which may be low-frequent. These frequencies and the loading of the converter circuit influence on the ripple on the output voltage, which again influences on the voltage control circuit and gives rise to the instability mentioned. The traditional way of solving this problem is to use a relatively large, and thus expensive, intermediary circuit capacitor to isolate the mains and the load from each other with regard to control, and at the same time to keep the ripple of 5 the voltage at an acceptably low level, which of course makes the circuit more expensive.

On the basis of the above, the task of the present invention is to provide a power supply as mentioned in the introduction, which is cost-effective, and o which does not suffer from the problems mentioned above with regard to stability and voltage overshoot.

According to a first aspect of the present invention, this task is solved by a power supply comprising an electrical converter for converting AC to DC, said 5 converter comprising a self-induction, a rectifier element and a switch member controlled by a control circuit, which controls the switch member on the basis of a reference and a measurement of a voltage at a given location in the power supply, the power supply being characterised by comprising a filter with a first cut-off frequency for an increasing voltage and a second cut-off frequency, o which is different from the first cut-off frequency, for a decreasing voltage.

According to a second aspect of the present invention, the task is solved by a method for controlling a power supply comprising an electrical converter, primarily for converting AC to DC, the converter comprising a self-induction, a 5 rectifier element and a switch member controlled by a control circuit, and preferably a capacitor, said method comprising control of the switch member by means of the control circuit on the basis of a reference and a measurement of a voltage at a given location in the power supply, which method is characterised by a filtering of the signal in the control circuit by means of a filter with a o first cut-off frequency at an increasing voltage and a filtering of the signal in the control circuit by means of a filter with a second cut-off frequency, which is different from the first cut-off frequency, at a decreasing voltage. When selecting suitable cut-off frequencies, it is thus achieved that the control of voltage increases is faster than that of voltage drops, so that the undesired voltage overshoots are reduced without causing the circuit to attempt removing the ripple voltage.

According to a third aspect of the invention, the power supply is used as a motor control, meaning that the above-mentioned problems with regard to disconnection and low-frequent mixed frequencies are overcome.

According to a preferred embodiment of the invention, the power supply is made so that both the first cut-off frequency and the second cut-off frequency are fixed. This gives the advantage that the filter can be realized by means of simple components, for example, merely a diode, two resistors and a capaci- tor.

These simple components are also comprised in a further preferred embodiment, according to which the measurement of the voltage at the given location in the power supply is made by means of a voltage measuring circuit and that the filter is realized with different time constants in this voltage measuring circuit.

In an advantageous variant of this embodiment, the different time constants in the voltage measuring circuit are realized in the form of an RC-link comprising at least one capacitor and two parallel-connected resistors, a diode being connected in series with at least one of the parallel-connected resistors.

Particularly advantageous is an embodiment, in which the diode is part of a circuit for realising an ideal diode. Hereby it is achieved that the cut-off fre- quency is not influenced by the size and the speed of the voltage increase. In another preferred embodiment, the converter comprises a boost converter circuit, which is an inexpensive and simple way of realising the converter circuit.

In the following, the invention will be explained in detail by means of examples of embodiments of the invention with reference to the drawings, showing:

Fig. 1 a schematic view of a first embodiment of a power supply according to the invention,

Fig. 2 a schematic view of a partly digital, second embodiment of a power supply according to the invention,

Fig. 3 a flowchart schematically showing examples of the digital control in the embodiment according to Fig. 2

Fig. 1 is a schematic view of a substantially analogue embodiment of the invention. The power supply circuit is supplied with single-phase AC from the inlets F and 0. A bridge rectifier D1-D4 rectifies the AC. The circuit comprises a pulse-width modulator 1 , which controls a transistor T1 to be conducting and blocking with a switch frequency that is substantially higher than the mains frequency, for example 50 to 100 times higher than the mains frequency, which is typically 50 Hz or 60 Hz. When the transistor T1 is conductive, the current flows from the mains through the rectifier bridge D1-D4, a self-induction L1 and the transistor T1 , so that energy is stored in the self-induction L1. When the transistor T1 blocks, the current from the mains flows through a rectifier bridge D1-D4, the self-induction L1, a rectifying element D5 in the form of a diode and a load 2. During this, the current through the self-induction L1 drops, while it supplies its energy to the load 2. In order to ensure that the load will always receive the required instant power, no matter whether or not transistor

T1 is conductive, an intermediary circuit capacitor C1 , connected in parallel with the load 2, is inserted in the circuit. Control loops are used for controlling the pulse-width modulator 1. One loop is a fast-controlling inner loop, which controls the current in the mains, for example by controlling the current in the self-induction. This control takes place on the basis of current measuring derived from a voltage measuring over a measuring resistor R1 , which is inserted between the rectifier bridge D1-D4 and the emitter of the transistor T1. This current measuring in itself is known state of the art, and can of course be made at any suitable location in the circuit and in any manner known. For example, as mentioned above in connection with the state of the art.

The voltage signal 3 corresponding to the voltage measurement is supplied to a subtraction node 4, where it is subtracted from the output signal 5 from a multiplication device 6. The resulting signal 7 is filtered in a low-pass filter 8, so that a filtered signal 9 is obtained, which controls the pulse-width modulator 1 , which again controls the transistor T1.

The second loop is a slow regulating outer loop, which controls the output voltage of the circuit. In the present embodiment of the invention, this voltage is measured across the capacitor C1 , in the form of a voltage signal on the conductor 10. This signal is filtered through an RC-link comprising a capacitor C3 and two parallel-connected resistors R2 and R3, a diode D6 being arranged in series with the resistor R2. The output signal 11 from this RC-link is subtracted from a reference signal 13 in a subtraction node 12. The reference signal 13 indicates the nominal output voltage to be supplied by the circuit, for example the 110 V or 240 V mentioned in the introduction. According to need, the output signal 14 from the subtraction node 12 can be scaled in a scaling device 15 to obtain an output signal 16 from the outer control loop. In order to achieve a sine-shaped current in the mains, output signal 16 from the outer control loop is multiplied by a signal representing the curve shape of the mains voltage, for example as shown on the basis of a measuring across the output of the rectifier bridge D1-D4, and the result of this multiplication, which is sup- plied as the output signal 5 from the multiplication device 6, is used as reference for the inner control loop.

The RC-link, which comprises a capacitor C3 and two parallel-connected re- sistors R2 and R3, a diode 6 being arranged in series with the resistor R2, serves the purpose of providing two different time constants, one for increasing voltage on conductor 10 and one for decreasing voltage on conductor 10, that is, it provides a low-pass filter with two different cut-off frequencies, depending on whether the control circuit must control an increasing or a decreas- ing output voltage for the circuit.

This is achieved in that a diode D6 is connected in series with one of the two parallel-connected resistors R2 and R3, in the present case R2. If the voltage on the conductor 10 increases, the diode 6 will conduct, and a charging current will run through both parallel-connected resistors R2 and R3 to the capacitor C3. If, on the other hand, the voltage on the conductor 10 drops, the diode D6 will block, so that the capacitor is only discharged through the resistor R3. As the total resistance of the parallel connection of the resistors R2 and R3 is lower than the resistance of R3 alone, the capacitor of the RC-link will thus see a lower time constant for increasing voltage than for decreasing voltage on the conductor 10.

Preferably, the values of the resistors R2 and R3 have been chosen so that R3 is in the range of 100 times larger than R2, which ensures a sufficient differ- ence between the two time constants.

Thus, the compensation of the circuit for increasing voltages is faster than that for decreasing voltages, which mainly causes that the damaging and thus unwanted overvoltages can be regulated away faster. In a simple manner, this provides the desired filtering with two substantially fixed cut-off frequencies. In order to compensate the conduction voltage drop over the diode D6, which will influence the time constant of the RC-link, particularly during slow voltage increases on the conductor 10, it is advantageous to use an ideal diode coupling with an operation amplifier 17, as shown in Fig. 1.

In this connection, it must also be mentioned that, according to an embodiment not shown, the two different time constants can also be achieved by using two oppositely polarised diodes arranged in series with the respective resistors R2 and R3. This means that in series with the resistor R3 is inserted a diode, whose polarisation is opposite to that of the diode D6, which is arranged in series with R2 in Fig. 1.

Fig. 2 shows a second embodiment of the invention, in which the control circuit is digitally implemented in, for example, a microprocessor or a digital signal processor, as suggested with a dotted line. For better understanding, corresponding elements in Fig. 1 and Fig. 2 have the same reference symbols. The control circuit comprises a multiplexer 21 , which combines the output voltage signal 10, the voltage-measuring signal 3 and the signal representing the curve shape of the mains voltage to a common signal 22. This signal 22 is supplied to an analogue/digital converter 23, in which it is converted to a digital signal 24 for use with a calculating unit 25. Based on this digital signal 24, the calculating unit 25 calculates a signal 29, which controls the pulse-width modulator 1 , which again controls the transistor T1.

Fig. 3 shows an example of a control algorithm, which can be used by the calculating unit 25. This algorithm can be stored in a memory 26 in connection with or integrated in the microprocessor or the digital signal processor. This memory can be RAM, ROM or, in principle, any suitable memory medium.

The mode of operation of the algorithm shown in Fig. 3 will be explained in the following. The algorithm, which is iterative, starts in a step 100. Then, in a step 101 , the calculating unit 23 awaits the next pulse-width modulated pulse and then, in step 102, reads a UDC, which is a variable representing the actual output voltage on the conductor 10. In the step 103, the calculating unit then compares with the voltage value represented by the variable UDCI, which the pulse-width modulator is presently set to supply. If the measured value, repre- sented by the variable UDC is larger than UDCI, a new value for UDCI is calculated in the step 104. If, however, UDC is smaller than or equal to UDCI, the new value for UDCI is calculated in the step 105. In this connection it must be noted that in the case, where UDC is equal to U_Dcι, the calculation of the new UDCI can take place in any of the steps 104 or 105, as both calculations will con- tinue to give the same value of UDCI, namely equal to UDC- After calculation in step 104 or 105, the new value of UDCI is sent to a step 106 as control signal for the pulse-width modulator 1 , which controls the transistor T1.

The calculation of the new value of UD_CI takes place in the two steps 104 and 105 with two different weightings x and y, so that UDCI approaches UDC with two different speeds, depending on whether UDCI is larger than UDC or not.

In the above, the present invention has been described on the basis of an analogue embodiment and a digital embodiment, respectively. However, it is un- derstood that the embodiments described are merely illustrating examples, and that, within the frames of the claims, many other analogue and digital embodiments and combinations thereof can be imagined.

Further, the expert will understand that other types of converter circuits than the described forward-converter of the boost type can be used, for example SEPIC-converters (Single Ended Primary inductance Converter), and also that other modulation forms than the described pulse-width modulation can be used.

Claims

Patent Claims

1. Power supply comprising an electric converter, primarily for converting AC to DC, which converter comprises a self-induction, a rectifier element and a switch member controlled by a control circuit, which controls the switch member on the basis of a reference and a measurement of a voltage at a given location in the power supply, characterised in that the control circuit comprises a filter with a first cut-off frequency for an increasing voltage and a second cut-off frequency, which is different from the first cut-off frequency, at a decreasing voltage.

2. Power supply according to claim 1 , characterised in that both the first cutoff frequency and the second cut-off frequency are invariable.

Power supply according to claim 1 or 2, characterised in that the measurement of the voltage at the given location in the power supply is made by means of a voltage measuring circuit and that the filter is realized with different time constants in this voltage measuring circuit.

4. Power supply according to claim 3, characterised in that the different time constants in the voltage measuring circuit are provided in the form of an RC-link comprising at least one capacitor and two parallel-connected resistors, a diode being arranged in series with at least one of the parallel- connected resistors.

5. Power supply according to claim 4, characterised in that the diode is part of a circuit for realising an ideal diode.

6. Power supply according to any of the claims 1 to 5, characterised in that the converter comprises a boost-converter.

7. Application of a power supply according to any of the preceeding claims for motor control.

8. Method for controlling a power supply comprising an electric converter, primarily for converting AC to DC, the converter comprising a self- induction, a rectifier element and a switch member controlled by a control circuit, and preferably a capacitor, said method comprising the control of the switch member by means of the control circuit on the basis of a reference and a measurement of the voltage at a given location in the power supply, characterised in that the method comprises a filtering of the signal in the control circuit by means of a filter with a first cut-off frequency at increasing voltage and a filtering of the signal in the control circuit by means of a filter with a second cut-off frequency at decreasing voltage, whereby the second cut-off frequency is different from the first cut-off frequency,