EP1766766A2

EP1766766A2 - Anharmonic low-current ac-dc converter

Info

Publication number: EP1766766A2
Application number: EP05738029A
Authority: EP
Inventors: Thierry THALES Intellectual Property ROS; Christophe Thales Intellectual Property TAURAND
Original assignee: Thales SA
Current assignee: Thales SA
Priority date: 2004-05-11
Filing date: 2005-04-21
Publication date: 2007-03-28
Also published as: FR2870403A1; WO2005109615A2; FR2870403B1; CA2566209A1; US7551463B2; US20080031021A1; WO2005109615A3

Abstract

The inventive converter is embodied in the form of an insulated current-hash converter adjustable by means of a controlled switch (5). Said controlled switch (5) is protected by a current-limiting device (9,7) responsive to a peak current instruction proportional to the instantaneous voltage (Vin) of a power supplying source, thereby making it possible to improve the quality of absorbed current and to limit the risk of disturbances of switch-mode power supply and to satisfy the new requirements of avionic networks.

Description

LOW ANHARMONIC AC / DC CONVERTER

The invention relates to AC / DC switching AC to DC voltage converters, with anharmonic current consumption and unit power factor. For a long time, AC / DC type converters have had the major drawback of absorbing a non-sinusoidal current with a large crest factor, a source of harmonic distortions in their power supply networks. The harmonic distortions of the current cause a distortion of the voltage delivered by a supply network, due to the non-zero impedances of the generator and the distribution. These network voltage distortions can then disturb the operation of other equipment supplied by this same network. In addition, a significant crest factor degrades the power factor of the equipment, which makes it necessary to oversize the generator and the distribution network. This results in an additional cost and an increase in mass which is particularly intolerable in the fields of avionics. In order to avoid this, the standards relating to the loads intended to be connected to electrical networks impose increasingly severe limitations on the nuisances induced by the AC / DC converters. These limitations relate in particular to the power factor, the total harmonic distortion also called THD and the maximum level of current harmonics as a function of their rank. In the avionics sector, we also add a tolerance to highly distorted voltage networks as well as operation on a variable frequency network. In addition to these requirements valid for an established operating regime, there are other requirements relating to behavior on transient regimes. During a start-up, short over or under voltages, micro-interruptions and inrush currents must be limited and remain anharmonic.

In the field of AC / DC converters meeting at least partially these requirements, isolated and low power (less than 500 Watts), two types of structures dominate: two-story structures and single-story structures. In two-stage structures, the first stage AC / DC converter is a power factor corrector mainly dedicated to the reduction of the disturbances induced on the network while the second stage is a DC / DC converter. The galvanic isolation between input and output is done at the level of the second DC / DC converter stage. The first stage, which is of the “boost” boost converter type, does not allow control of the inrush current at start-up guaranteeing good performance in the transient phases. In addition, the control circuits of the switches of the two stages are not simple since they require the use of analog multipliers to effect the control of the input current. Single-stage structures include a single AC-DC voltage converter with galvanic isolation input-output. For low powers, the topology most often used is the "Flyback" topology which uses the magnetization inductance of a transformer ensuring galvanic isolation between input and output. A switch in series with the transformer primary controls the periodic injection of a rectified current from the AC power source into the inductance of the isolation transformer. The duration of closure of the switch conditions the extent of a current injection into the inductance of the isolation transformer. It is determined by a switching control circuit which balances the power absorbed from the supply network and the power consumed by the load and the losses of the converter, in order to maintain the output voltage of the converter at a set level while limiting the peak current flowing through the switch. In normal operation, a flyback converter operates in discontinuous current conduction mode. The peak primary current reaches an intensity proportional to the input voltage and the conduction time when the switch is opened. In steady state, the absorption of current from the power source is sinusoidal if the duty cycle is kept constant. However, the transient start-up and end phases of brief outages as well as the output short-circuits give rise to at peaks of the absorbed current which can be clipped by the peak current limitation of the switch control circuit, but this clipping is the source of disturbances and instability of the power source. It is known to overcome this problem during start-up by a slow increase in the duty cycle (so-called “soft-start” technique), but this solution is incompatible with requirements on the shape of the input current during micro- cuts. It is also known, by international patent application WO 89/05057 to limit the peak current passing through the controlled switch of a single-stage converter of the flyback type, as a function of the instantaneous voltage of the power source by means of an analog multiplier modulator circuit performing the product of an image r of the voltage error at the output of the converter by the rectified voltage U _s of the power source. The presence of an analog multiplier makes the control circuit of the controlled switch more complex to produce and less efficient due to the drifts inherent in this kind of circuit.

The object of the present invention is to combat the abovementioned drawbacks and to reduce the disturbances injected by an AC / DC voltage converter on the AC electrical network supplying it, including in the transient operating phases.

It relates to an AC / DC voltage converter with low anharmonic currents and with chopping regulation of a rectified current coming from an AC power supply source, by means of at least one controlled switch subjected by a control device, at switching cycles with variable duty cycle adjusted so as to cancel an output voltage error detected by an error corrector and to comply with a peak current limitation setpoint produced by a peak current setpoint generator depending on the instantaneous voltage of the power source. This voltage converter

AC / Dc is remarkable in that the control device comprises an oscillator equipped with a PWM width modulator controlled by the error corrector in order to cancel an output voltage error, and a logic circuit inserted at the output of the PW width modulator ensuring the blocking of the controlled switch (s) each time the instantaneous current flowing through the controlled switch (s) is exceeded, of the setpoint produced by the peak current maximum setpoint generator. Advantageously, the peak current maximum setpoint generator generates a peak current limit l _{pmm peak} setpoint which is a refined function, in parts, of the instantaneous voltage V _{m IπΛ} of the power source meeting the definition:

I _p ^ _created = V, _n _, _nst + *>, i integer varying from 1 to n

n being the number of parts of the affine function defined by successive ranges of variations of the instantaneous voltage V _m _ _ml of the power source, an ith part corresponding to a range:

V • n_ msl) _Uaχ0 ~ ^U '{ ^v _m _, _nst ) _Max , ^{being the} transition value between i _lé m _Θ and la (i + 1) _lé m _θ blades and k, and bj being constants whose values depend on the part i concerned. Advantageously, the converter has a flyback topology and operates in discontinuous conduction mode. Advantageously, when the converter has an input circuit connected to the power source, containing the controlled switch, an output circuit galvanically isolated from the input circuit and a switch control device having elements distributed between the input and output circuits, its switch control device comprises, at the output circuit, at least one error correcting circuit integrating the difference between the output voltage of the converter and a reference voltage, at the interface between the input and output circuits, a galvanic isolation component carrying the output signal from the error correcting circuit to the input circuit, intended for a circuit generating the control signal of the 'switch controlled. Advantageously, when the converter has an input circuit connected to the power source containing the controlled switch, an output circuit galvanically isolated from the input circuit and a switch control device having elements distributed between the circuits input and output, its switch control device comprises, at the output circuit, at least one error correcting circuit integrating the difference between the output voltage of the converter and a reference voltage, an oscillator with a variable duty cycle controlled by the error correcting circuit providing, for the controlled switch, a two-state control logic signal corresponding one to a conduction command, the other to a blocking, modulated in pulse width and, at the interface between the input and output circuits, a galvanic isolation component routing to the input circuit, at position of the controlled switch, the logic control signal. Advantageously, when the converter uses an inductor, in switching mode, supplied with current from the power source by means of the controlled switch, it comprises a circuit for reducing harmonics of input current formed by an oscillation damping circuit connected to the terminals of the inductor. Advantageously, the input current harmonics reduction circuit is a dissipative circuit comprising a capacitor placed in series with a resistance across the inductance. Advantageously, the input current harmonics reduction circuit is a non-dissipative switched circuit comprising an auxiliary controlled switch placed in series with a non-return diode at the terminals of the inductor and making it possible to short-circuit the inductor during times when it is not supplied with current by the power source. Advantageously, when the converter uses the inductance, in switching mode, of an isolation transformer ensuring galvanic isolation between input and output, it comprises: secondary side of the transformer, - an error corrector delivering a setpoint as a function of the difference presented by the output voltage of the converter with respect to a voltage setpoint, straddling the border separating the primary and secondary sides of the transformer, - a galvanically isolated coupler transmitting the error corrector output setpoint to the primary side of the transformer, and the primary side of the transformer, - a rectifier circuit supplying an unregulated direct voltage, - a pulse width modulation oscillator controlled by the setpoint from the error corrector and transmitted by the coupler, supplying to the controlled switch, a switching signal formed by pulses, - a current sensor measuring the instantaneous current passing through the controlled switch, - a threshold crossing detector circuit of peak current comparing the setpoint delivered by the peak current setpoint generator with the current i snapshot measured by the current sensor, and - a pulse inhibition circuit interposed between the width-modulated oscillator and the command of the controlled switch, triggered by the threshold crossing detector circuit and reset by each pulse of the switching signal delivered by the width-modulated oscillator. Advantageously, when the converter uses the inductance, in switching mode, of an isolation transformer ensuring galvanic isolation between input and output, it comprises: secondary side of the transformer, - an error corrector delivering a setpoint function of the difference presented by the output voltage of the converter with respect to a voltage setpoint, and - an oscillator, called the main oscillator, with pulse width modulation controlled by the setpoint from the error corrector, providing a switching signal, straddling the border separating the primary and secondary sides of the transformer - an isolating coupler galvanic transmitting the signal from the main oscillator to the command of the controlled switch. primary side of the transformer, - a rectifier circuit supplying an unregulated direct electrical voltage, - a so-called auxiliary oscillator supplied by the rectifier circuit placed on the primary side of the transformer and providing periodic conduction pulses for the controlled switch, - a circuit switch with two inputs and an output interposed behind the outputs of the auxiliary oscillator and of the coupler giving priority to the signal, when it exists, of the main oscillator transmitted by the coupler, - a current sensor measuring the current instantaneous through the controlled switch, - a peak current threshold crossing detector circuit comparing the setpoint delivered by the peak current setpoint generator with the instantaneous current measured by the current sensor, and - a circuit inhibiting pulse placed at the output of the routing circuit, triggered by the crossing detector circuit threshold and reset by each pulse crossing the switching circuit.

Other characteristics and advantages of the invention will emerge from the description below of an embodiment given by way of example. This description will be made with reference to the drawing in which: - Figure 1 is a block diagram of an AC / DC converter of the "flyback" type in discontinuous conduction mode according to the invention, - Figures 2 and 3 are diagrams illustrating two examples of distribution of the elements of a controlled switch control device appearing in FIG. 1, - a FIG. 4 is the diagram of an example AC / DC converter of "flyback" type according to the invention, - a FIG. 5 is a diagram of a combined oscillator a switching circuit giving priority to the signal from an external oscillator, - Figure 6 is a diagram detailing an embodiment of a peak current limitation setpoint generator, - Figures 7, 8 and 9 are diagrams illustrating waveforms showing the operating differences between an AC / DC converter according to the invention and an AC / DC converter of the prior art, - Figure 10 is a diagram detailing an embodiment of a circuit of amortisseme nt of oscillations, - Figure 11 is a diagram showing the shape of the parasitic oscillations caused during idle times by the magnetization energy, - a figure 12 is a diagram showing the damping of the parasitic oscillations, produced with the circuit d damping shown in Figure 10, - Figure 13 is a diagram detailing another embodiment of an oscillation damping circuit, and - Figure 14 is a diagram showing the damping of parasitic oscillations achieved with the damping circuit shown in figure 13.

The AC / DC converter illustrated in FIG. 1 supplies a load Z with direct current from an alternating current supply source Vin. Its structure is of the "flyback" type with a transformer 1 whose primary winding Lp is supplied, by forced switching, in rectified voltage delivered by a double altemance rectifier bridge 2 connected to the terminals of the alternating current supply source Vin and whose secondary winding Ls is connected to the terminals of a load Z by means of a rectifying diode 3 and a filtering capacitor 4. The forced switching of the rectified current supplied to the primary winding Lp of the transformer 1 is carried out using a controlled switch 5 connected in series with the primary winding Lp of the transformer 1 to continuous terminals of the rectifier bridge 2. In the discontinuous conduction operating mode where the current in the rectifier diode 3 placed in the secondary circuit of the transformer 1 is canceled before the end of each switching cycle of the controlled switch 5, the magnetic flux in the core of the transformer 1 is canceled at each switching cycle of the controlled switch 5. When the controlled switch 5 closes , the current in the primary winding Lp of transformer 1 starts from the value zero. During this first period, the rectifying diode 3 placed in the secondary circuit of the transformer 1 is blocked and the primary current of the transformer 1 increases linearly. When the controlled switch 5 opens, the flux in the core of the transformer 1 cannot be canceled instantaneously, a current flows in the secondary winding Ls of the transformer 1 making the rectifying diode 3 which charges the capacitor filtering 4 and feeds the load Z. The controlled switch 5 can be produced in various forms and in particular in the form of one or more semiconductor devices placed in parallel, for example one or more cmos transistors placed in parallel. The voltage across the load Z is regulated by varying the duration of conduction of each switching cycle of the controlled switch 5, the total duration of which is kept constant. The control signal of the controlled switch 5 is a logic control signal with two states, one corresponding to a conduction command and the other to a blocking command, of fixed frequency and variable duty cycle. It is provided by a control device 7 through which the galvanic isolation 8 separating the primary and secondary circuits of the transformer 1. It is assumed, in the following explanations, that the high states of the command signal of the controlled switch 5 correspond to a closing command and the low states to a opening command. The control signal is produced, within the control device 7, by a PWM width modulator operating a width modulation on a rectangular binary signal supplied by an oscillator which fixes the frequency of the switching cycles. The PWM modulator makes it possible to vary the duty cycle high state / low state, that is to say closing time / opening time, as a function of an error signal integrating, over a certain period, the difference between the voltage Vs measured at the terminals of the load Z and a reference voltage Vref both applied to inputs of the part of the control device 7 referenced with respect to the mass of the secondary circuit of the transformer 1. The PWM width modulator integrated into the control device 7 extends the conduction time of each cycle if the voltage measured at the terminals of the load Z tends to become lower than the setpoint and shortens it otherwise. The control device 7 comprises, in its part referenced with respect to the mass of the primary circuit of the transformer 1, a device for inhibiting its width modulator PWM which receives, on one side, a peak current maximum instruction lpmax_crête arriving from a primary peak current maximum setpoint generator 9 analyzing the instantaneous voltage delivered by the rectifier bridge 2 and, on the other, a measurement of the instantaneous current lp_mes passing through the controlled switch 5 supplied by a measuring device primary current 6. Figures 2 and 3 give two examples of the distribution of the elements of the control device 7 between its two referenced parts with respect to separate masses galvanically isolated from each other, those of the primary and secondary circuits of the transformer 1. The modulator PWM is shown separately from the other logic functions of the control circuit 7 because it can be placed from the on the input side of the AC / DC converter as on the output side. In the distribution of FIG. 2, most of the elements of the control device 7 are placed in its part 7a, referenced by relative to the mass of the primary circuit of the transformer 1, its part 7b referenced with respect to the mass of the secondary circuit of the transformer 1 containing only a minimum of elements. The part 7a referenced with respect to the ground of the primary circuit of the transformer 1 contains an oscillator 10 providing a rectangular signal at the switching frequency of the controlled switch 5, a PWM width modulator 11 operating on the signal of the oscillator 10 as a function of a modulating signal Sm, a subtractor 12 delivering a primary current limitation setpoint lp_max corresponding to the primary peak current maximum setpoint lpmax_crest delivered by the primary peak current maximum setpoint generator 9, reduced by the measurement instantaneous current lp_mes supplied by the primary current measuring device 6, and a logic circuit 13 combining the primary current limitation setpoint lp_max with the modulated signal coming from the PWM width modulator 1 1. The part 7b referenced with respect to the earth of the secondary circuit of transformer 1 contains only an error corrector 16 providing a signal analog of error Er deduced from the difference existing between the voltage Vs of the secondary circuit of the transformer 1 and a reference value Vref, Between the two parts 7a and 7b of the control device 7 of the controlled switch 5, a coupler 15 transmits, in the form of the analog signal Sm, the error signal Er coming from the error corrector 16 of part 7b, to the modulation input of the PWM width modulator 11 of part 7a. This coupler 15 which provides galvanic isolation between the two parts 7a and 7b of the control device 7, can be produced by means of a linear opto-coupler or a pulse transformer interposed between coding and decoding. In the distribution of FIG. 3, the PWM width modulator 11 and the oscillator 10 which supplies it with the signal to be modulated are transferred to the part 7'b referenced with respect to the mass of the secondary circuit of the transformer 1. This allows to transmit to the coupler 15 ′ the logic control signal of the controlled switch (s) which is a binary signal and not an analog signal, and makes it possible to synchronize on the output side of the AC / DC converter transmitted to the through galvanic isolation by the logic control signal. It is thus possible to synchronize the AC / DC converter with other functions in order to reduce the risks of noise, noise or frequency beats, which is particularly advantageous in avionics techniques. In return it is necessary to provide, in the part 7'a referenced with respect to the primary circuit, of the transformer 1, an auxiliary oscillator 16 taking over from the PWM width modulator 11, by means of the logic circuit 13 ', in the absence of voltage in the secondary circuit of transformer 1. The displacement of the coupler 15 'at the output of the PWM width modulator means that it no longer works in analog but in binary with the advantage of being insensitive to disturbances linked to currents of common mode due to polarization by weak electric currents as well as variations in the optronic transmission gain for signals with a wide range of amplitude variation which are the cause of malfunction. It can be achieved by means of either a fast opto-coupler of binary logic, or a pulse transformer, or by a capacitive link, or even, by a link of the radiofrequency type. FIG. 4 details an exemplary embodiment of an AC / DC converter having a control device 7 for its controlled switch 5, the elements of which respect the distribution illustrated in FIG. 3. The main oscillator 10, the width modulation circuit PWM 11 and the error corrector 16 are supplied by the secondary circuit of the transformer 1. The binary command signal of the controlled switch 5 which they produce is transferred to the primary side of the transformer 1 by means of an opto-coupler 15 'ensuring compliance with the galvanic isolation between the primary and secondary circuits of the transformer 1. Since the transmitted signal is binary, the gain and polarization drifts of the opto-coupler 15' have no effect on it. The error corrector 16 is produced using a comparator mounted as an integrator, in order to deliver an average over a certain period of time, of the difference existing between the DC voltage Vs delivered to the secondary of the transformer 1 and the voltage of setpoint Vref. The binary signal for controlling the controlled switch 5 transmitted by the optocoupler 15 ′ is applied to the controlled switch 5 through a logic circuit 13 'allowing on the one hand, to replace it with an emergency version when it fails due to a lack of DC voltage at the secondary of transformer 1 and on the other hand, to cut to stop the conduction of the controlled switch 5. as soon as a maximum current setpoint is exceeded by the current flowing through the primary winding Lp of the transformer 1. More precisely, the pilot logic circuit 13 ', which is supplied by the rectifier bridge 2 of the primary circuit of the transformer 1 comprises: - an auxiliary oscillator 131 providing a rectangular, symmetrical signal of close frequency and of the same shape as that of the main oscillator 10 but not synchronized with it, - a switching circuit 132 giving priority to the switching control signal coming from the main oscillator 10 via the width modulation circuit 11 and the opto-coupler 15 ′, - one ci rcuit 133 of interruption of the switching control signal constituted here by a logic gate of the “no or” type due to an inversion of the switching control signal by a pulse inhibition circuit placed downstream, and - the pulse inhibition circuit consisting of a logic gate of the “non-and” type 134 left free or forced to the zero state (absence of pulse) by a logic flip-flop of type D 135, the setting of which zero causes that of the logic gate of type "and" 134 and whose setting to one is caused by the rising edge of each pulse of the switching control signal. The control input of the interrupt circuit 133 makes it possible to stop the AC / DC converter at will, for example when the input voltage drops below a minimum operating voltage threshold. The command input of the pulse inhibition circuit, which consists of the reset input of its “D” type flip-flop 135, is controlled by the limiter comparator 12 which receives, on one side, a primary peak current maximum setpoint lpmax_crest reaching it a primary peak current setpoint generator 9 and, on the other, a measurement of the instantaneous current lp_mes passing through the controlled switch 5 taken by a current measuring device 6. The current measuring device 6 can be of any known type. It can proceed for example, by means of a measuring resistor or a current transformer inserted in series with the primary winding of the transformer 1. The embodiment which has just been described relative to the figure 4 easily adapts to a distribution of the elements of the control device in accordance with FIG. 2. During this adaptation, the oscillator 10 and the PWM width modulator 11 are transferred into the primary circuit of the transformer 1 while the oscillator auxiliary 131 and the switching circuit 132 are eliminated. FIG. 5 gives an example of an auxiliary oscillator circuit 131 associated with a switching circuit 132 giving it the status of non-priority. The auxiliary oscillator consists of a logic gate of the “non-and” type 20, the output 200 of which is looped back in response to one of its inputs 201 via an operational amplifier 21 mounted as an integrator. The operational amplifier 21 mounted as an integrator plays the role of a timer making it possible to delay the back propagation of the logic state of the output 200 on the input 201. It makes it possible to adjust the oscillation frequency to that desired . For there to be oscillation, the other input 202 must be at the high logic level (1 logic). This condition is achieved by the bias resistor 22 recalling the other input 202 to the positive terminal of the power source, here the plus terminal of the rectifier bridge 2. The auxiliary oscillator obtained has a duty cycle of 50% and l The integrator is dimensioned so that the oscillator operates at the frequency of the master oscillator. When the signal from the main oscillator 10 is applied to the other input 202, it takes over from the auxiliary oscillator because its low states (logic 0) impose a high state at output 200 of the logic gate of type "no -and »20 and its transitions from a low state (0 logic) to a high state (1 logic) always occur after the operational amplifier 21 has transmitted the high output state (1 logic) at input 201 because it has a duty cycle (high state low state) less than or equal to 50%. The second door “No-and” type logic 23 corrects the signal inversion caused by the first “No-and” type logic gate 20. The primary peak current setpoint generator 9 generates a peak current limitation setpoint ( I _{pmm crêle} ) which is a refined function, by parts, of the instantaneous voltage of the power source (V _{m tnst} ) answering the definition: ^l _P _crê "= ^V " _ "" + ^b , * ^{eTlUQT Variant of} ' ^to "

n being the number of parts of the affine function defined by successive ranges of variations of the instantaneous voltage V _m _ _msl of the power source, an ith part corresponding to a range: (VV m, τιsl>) Max {ι - \) <~ V inst <~ {VV m uât) ⁾ ' _Maχl

V m_ιnsl) _MaJc0 - ^U •

{v _{m ιnst} ) _Uaa being the transition value between the i _lβme and the (i + 1) jth parts and k, and b, being constants whose values depend on the part i concerned. This peak current limitation setpoint is preferably proportional to the instantaneous voltage of the power source, with a progressive limitation of the gain beyond the normal range of amplitude variation of the voltage of the power source ( Wine) (k, ≤ k _{I + i} ). FIG. 6 details a possible diagram for the primary peak current setpoint generator 9 with shaping of the clipped current. This is built around three resistors 30, 31, 32 connected in series, as a voltage divider, at the DC voltage terminals of the full-wave rectifier bridge 2 supplying the primary circuit Lp of the transformer 1 of the AC / DC converter. The maximum peak current setpoint, calibrated in voltage, is the reflection of the instantaneous voltage appearing across the terminals of the rectifier bridge 2, in the reduction ratio corresponding to that of the value of the resistor 32 over the sum of the values of the resistors 30 , 31, 32. To these three resistors 30, 31, 32 is added a device imposing a limit on the range of travel of the upper peak current setpoint value to take account of overvoltages in the supply network. The progressive limitation of the gain of the peak peak current setpoint generator as a function of the input voltage makes it possible to reduce the risks of instability associated with clipping to a fixed maximum value, this progressive limitation is obtained by means of a diode 33 placed in series with a resistor 34 between the junction point of the resistors 30, 31 of the voltage divider and a regulated DC voltage terminal + Vcc lower than the peak voltage at the DC terminals of the rectifier bridge 2, its anode being turned to the regulated DC voltage terminal + Vcc so as to establish a bypass current to the regulated DC voltage terminal + Vcc when the voltage at the junction point of the resistors 30, 31 tends to exceed that of the regulated voltage terminal + Vcc . The diagram of FIG. 7 shows the usual waveform present, at the start of an AC / DC converter provided with a peak current limitation at constant level in accordance with the prior art, at the terminals of its alternating electrical voltage source power. The one in Figure 8 shows the waveform obtained under the same conditions with the proposed peak current limitation. There is a considerable reduction in disturbances. The proposed peak current limitation allows the current absorbed by the converter to be shaped when it starts up or after brief power cuts. The diagrams in FIG. 9 show the comparison of the current absorbed during the starting phases, one a, by a two-stage PFC corrector AC / DC converter with a Boost type corrector and the other b, the same evolution for an AC / DC converter provided with the proposed peak current limitation. The magnetization energy of the transformer 1 of the flyback converter is dissipated by a damping circuit 14, connected to the terminals of its primary winding L1. it reduces the parasitic oscillations coming from the circuit formed by the inductance of the transformer 1 and the parasitic capacities on the primary side, during the period of non-conduction of the controlled switch 5 and of the diode 3 placed at the secondary and makes it possible to bring negligible level the disturbance levels of the input current. FIG. 10 details a possible diagram for the damping device 14. This diagram is that of a dissipative damping circuit consisting of a capacitor 70 connected in series with a resistor 71, in parallel on the primary L _p of the transformer isolation 1. As shown in FIG. 11, the magnetizing energy of the isolation transformer 1 produced at the terminals of the controlled switch 5, at each of its openings, after cancellation of the current in the diode 3 of the secondary circuit , during a dead time phase up to the next closing of the controlled switch 5, a voltage oscillation due to the oscillating circuit formed by the primary inductance L _p of the isolation transformer 1 and the stray capacitances of the primary circuit , mainly that of the output of the controlled switch 5. The presence of this voltage oscillation causes the controlled switch 5 to be turned on, during a voltage alternation network, both at low voltage and at high voltage, which is the cause of switching losses in the controlled switch 5 resulting in significant distortion of the input current. The dissipative distortion reducing circuit with capacitor 70 and series resistor 71 makes it possible, as shown in FIG. 12, to damp this oscillation in a very efficient manner insofar as the dead time is at least greater than two periods of oscillation. FIG. 13 details another possible diagram for the device 14 for damping the magnetization energy of the transformer 1 of the AC / DC converter. This diagram is that of short-circuiting of the transformer 1 during the dead time phase appearing during the opening periods of the controlled switch 5 after the cancellation of the current in the diode 3 of the secondary circuit. This short-circuiting is obtained by means of an auxiliary controlled switch 100 connected in series with a non-return diode 101 in parallel on the primary L _p of the transformer 1. The auxiliary controlled switch 100 is turned on outside. the conduction time of the main controlled switch 5 in order to avoid any simultaneous conduction liable to short-circuit the rectifier bridge 2. The efficiency of the AC / DC converter is optimized by adjusting the non-overlap time between the opening of the auxiliary controlled switch 100 and closing the switch main controlled 5. Indeed, when the auxiliary controlled switch 100 is opened, the oscillation reappears. Thanks to it, the voltage across the main controlled switch 5 passes through a minimum for which the switching losses are lower. This transition to minimum voltage favorable to switching from the blocked state to the on state of the main controlled switch 5 corresponds to the quarter of the oscillation period. This leads to making the end of the non-recovery time coincide with the quarter of the oscillation period. The device 14 for damping the magnetization energy of the transformer 1 of the AC / DC converter operating by short-circuiting the transformer 1 has the advantage of not dissipating the magnetization energy but of storing it temporarily by means of a current flow in the primary winding L _p of the isolation transformer 1. It makes it possible, as shown in FIG. 14, to obtain at the terminals of the main controlled switch 5 a waveform devoid of parasitic oscillations conducive to the reduction of distortions in the supply current and to an increase in efficiency. The AC / DC converter, which has just been described in relation to FIGS. 3 et seq., Has three main operating modes depending on the situation encountered: - starting after a sufficiently prolonged stop so that its output voltage has become insufficient to supply the elements of its secondary circuit, - start-up after a short stop having left an output voltage sufficient to supply the elements of its secondary circuit, and - operation in steady state. During a start-up following a prolonged stop of the converter, the secondary voltage of the isolation transformer 2 is zero. All the elements of the converter placed in the secondary circuit are therefore not supplied and no switching signal is delivered at the output of the galvanic isolation element 15 ′. On the other hand, the auxiliary oscillator 131 placed in the primary circuit of the transformer 1 operates as soon as the primary rectifier bridge 2 supplies energy and delivers a switching signal corresponding to a maximum conduction setpoint for the controlled switch 5 (duty cycle of 50%). In this transient operating phase, the closing of the controlled switch 5 is synchronous with the rising edge of the auxiliary oscillator 131. At each switching period, it is the protection of limitation of the primary peak current which imposes the duty cycle of the controlled switch 5 according to the admissible stresses for the components. The primary peak current limiting protection intervenes in the primary circuit at the level of the reset command for the D 135 type flip-flop of the pulse inhibitor circuit. The tripping of the primary peak current protection resets this D 135 flip-flop to zero, which activates the pulse inhibitor circuit and causes the controlled switch 5 to go out. The device is simple and robust. In a first part of this transient phase, the AC / DC converter operates in direct current conduction mode, at a low output voltage. The decrease in current in the secondary diode 3 is slow and cannot be canceled at these terminals before the end of the switching period. When the secondary voltage is high enough, the galvanic isolation element 15 'transmits the switching signal from the main oscillator to the primary circuit of the converter.

10 via the PWM modulator 11. Synchronization is immediate and this is the switching signal coming from the PWM modulator

11 which is applied to the pulse inhibitor circuit in place of the signal from the auxiliary oscillator 131. The control of the controlled switch 5 becomes synchronous with the rising edge of the main oscillator 10 located in the secondary. In this operating phase which also corresponds to short-term stops and micro-cuts, the duty cycle is maximum since the output voltage is not established and the amplifier

12 secondary voltage error is saturated. The primary peak current limiting protection continues to set the duty cycle, as in the absence of secondary voltage, the only difference being that the command to close the controlled switch 5 has become synchronous to the main oscillator 10 located in the secondary. When the secondary voltage is set, the amplifier ^"Error 12 saturates more and tends towards a value corresponding to an energy balance point in order to provide the necessary power to the load to the desired output voltage. In this mode, the primary peak current limitation protection no longer comes into action. The command received by the controlled switch 5 corresponds to the switching signal coming from the PWM modulator 11 placed at the secondary. The pulse inhibitor device is no longer activated because its D 135 type flip-flop no longer receives a reset command from the primary peak current protection. It is transparent. In summary, with the AC / DC converter which has just been described, the signal digital transmitted by the galvanic isolation element is the image of the command of the controlled switch ensuring the chopping of the primary current when protection of limitation of the primary peak current is not ac When the absorbed current is greater than the maximum peak current setpoint, the AC / DC converter is controlled in current limitation, but the digital signal transmitted by the galvanic isolation element makes it possible to always synchronize the AC / DC converter. DC. For this a minimum duty cycle is imposed on the PWM modulator. An auxiliary oscillator placed on the primary side of its isolation transformer takes over when, as at start-up, the output voltage of the AC / DC converter is insufficient to supply its elements placed on the secondary of its isolation transformer. The advantages of such a device are multiple: - high immunity to noise and other electromagnetic disturbances, - the possibility of simply synchronizing the AC / DC converter from a central clock located in the secondary, - the opportunity to reduce the frequency variations source additional electromagnetic interference, also realization of simple function: variable frequency depending on the output voltage. the generation of synchronous or phase opposition signals for the control of parallel AC / DC converters for example or for the generation of control of complementary MOS-controlled switches

"Idle".

Claims

1. AC / DC voltage converter with low anharmonic currents and chopped regulation of a rectified current from an AC power source (Vin), by means of at least one controlled switch (5) subject by a control device, at switching cycles with variable duty cycle adjusted so as to cancel an output voltage error detected by an error corrector (16) and to comply with a setpoint _limit (I _{pma cru} .) peak current produced by a generator (9) for setting the maximum peak current as a function of the instantaneous voltage of the power source (Vin) characterized in that the control device comprises an oscillator (10) equipped with a modulator PWM width (11) controlled by the error corrector (16) to cancel an output voltage error, and a logic circuit (13, 13 ') interposed at the output of the PWM width modulator (11) ensuring the blocking of the controlled switch (s) (5) each time the instantaneous current flowing through the controlled switch or switches (5) is exceeded, of the set point produced by the peak current maximum set point generator (9).

2. Converter according to claim 1, characterized in that the peak current maximum setpoint generator (9) generates a peak current limitation setpoint (l _{pma peak} ) which is a refined function, in parts, of the instantaneous voltage of the power source (V _{m ^} corresponding to the definition: i _p ^ _created * = ^v , _n _, _ns , + b, i integer varying from 1 to n

n being the number of parts of the affine function defined by successive ranges of variations of the instantaneous voltage V _m _ _{ιns <} of the power source, an ith part corresponding to a range:

{VV m inst ') M, ax (ι- <V insi <- {vV i mn_ mmsstt))) _Max ,

{ ^v , _a , _ns ,) _Uβx , ^{being the} transition value between the i _iό mθ and the (i + 1) _{iόmθ par1iθS} and kj and bj being constants whose values depend on the part i concerned.

3. Converter according to claim 1, characterized in that it is of flyback topology and operates in discontinuous conduction mode.

4. Converter according to claim 1, with an input circuit connected to the power source (Vj _n ) containing the controlled switch

(5), an output circuit galvanically isolated from the input circuit and a switch control device (7) having elements distributed between the input and output circuits, characterized in that the device (7) switch control comprises, at the output circuit, at least one error correcting circuit (16) integrating the difference between the output voltage (V _s ) of the converter and a reference voltage (Vref), at l interface between the input and output circuits, a galvanic isolation component (15) conveying the output signal from the error correcting circuit (16) to the input circuit, intended for the input of a circuit (10, 11) generating the control signal of the controlled switch (5).

5. Converter according to claim 1, with an input circuit connected to the power source (V _in ) containing the controlled switch (5), an output circuit galvanically isolated from the input circuit and a device (7 ) switch control having elements distributed between the input and output circuits, characterized in that the switch control device (7) comprises, at the output circuit, at least one corrector circuit of error (16) integrating the difference between the output voltage (V _s ) of the converter and a reference voltage (Vref), an oscillator (10, 11) with a variable duty cycle controlled by the error correcting circuit (16 ), providing, for the controlled switch (5), a two-state control logic signal, one corresponding to a conduction command, the other to a blocking command, modulated in pulse width and, at the interface between the input and output circuits, a galvanic isolation component (15 ') carrying the logic control signal to the input circuit, destined for the controlled switch (5).

6. Converter according to claim 1, using an inductor (L _p ), in switching mode, supplied with current from the power source (Vin) by means of the controlled switch (5), characterized in what it includes is a circuit for reducing input current harmonics formed by an oscillation damping circuit (14) connected to the terminals of the inductor (L _p ).

7. Converter according to claim 6, characterized in that the input current harmonics reduction circuit (14) is a dissipative circuit comprising a capacitor (70) placed in series with a resistor (71) at the terminals of the 'inductance (L _p ).

8. Converter according to claim 6, characterized in that the input current harmonics reduction circuit (14) is a non-dissipative switched circuit comprising an auxiliary controlled switch (100) placed in series with a non-return diode 101 at the terminals of the inductor (L _p ) and making it possible to short-circuit the inductor (L _p ) during the times when it is not supplied with current by the power source (Vin).

9. Converter according to claim 2, characterized in that it comprises a current limitation setpoint generator (13) generating a peak current limitation setpoint proportional to the instantaneous voltage of the power source, with progressive limitation gain (Jt ,. <k _M ) beyond the normal range of amplitude variation of the voltage of the power source (Vin).

10. Converter according to claim 1, using the inductance (L _p ), in switching mode, of an isolation transformer (1) providing galvanic isolation between input and output, characterized in that it comprises : secondary side of the transformer (1), - an error corrector (16) delivering a setpoint as a function of the difference presented by the output voltage (Vs) of the converter with respect to a voltage setpoint (Vref), straddling the border separating the primary and secondary sides of the transformer (1), - a galvanically isolated coupler (15) transmitting the output setpoint of the error corrector (16) to the primary side of the transformer (1), and the primary side of the transformer (1), - a rectifier circuit (2) supplying an unregulated direct electric voltage, - an oscillator (10, 11), with pulse width modulation controlled by the set point from the error corrector (16) and transmitted by the coupler (15), supplying , at the controlled switch (5), a switching signal formed by pulses, - a current sensor (6) measuring the instantaneous current passing through the controlled switch (5), - a circuit for detecting the crossing of current thresholds crest (12) comparing the co n sign delivered by the peak current setpoint generator (13) with the instantaneous current measured by the current sensor (6), and - a pulse inhibition circuit (134, 135) interposed between the oscillator with modulation of width (10, 1 1 FIG. 2) and the command of the controlled switch (5), triggered by the threshold crossing detector circuit (12) and reset by each pulse of the switching signal delivered by the modulation oscillator width (10, 11).

11. Converter according to claim 1, using the inductor (L _p ), in switching mode, of an isolation transformer (1) ensuring galvanic isolation between input and output, characterized in that it comprises: secondary side of the transformer (1), - an error corrector (16) delivering a setpoint as a function of the difference presented by the output voltage (Vs) of the converter with respect to a setpoint (Vref), and - an oscillator (10, 11), called the main oscillator , with pulse width modulation controlled by the setpoint from the error corrector (16), supplying a switching signal, straddling the border separating the primary and secondary sides of the transformer (1), - an isolating coupler galvanic (15 ') transmitting the signal from the main oscillator (10, 11) to the control of the controlled switch (5). primary side of the transformer (1), - a rectifier circuit (2) supplying an unregulated direct electrical voltage, - an oscillator (131) called auxiliary supplied by the rectifier circuit (2) placed on the primary side of the transformer (1) and supplying periodic pulses to activate the controlled switch (5), - a switching circuit (132) with two inputs and an output interposed behind the outputs of the auxiliary oscillator (131) and of the coupler (15 ') giving priority to the signal, when there is one, from the main oscillator (10, 11) transmitted by the coupler (15 '), - a current sensor (6) measuring the instantaneous current passing through the controlled switch (5), - a peak current threshold crossing detector circuit (12) comparing the setpoint delivered by the peak current setpoint generator (13) with the instantaneous current measured by the current sensor (6), and - an inhibition circuit pulse (134, 13 5) placed at the output of the switch circuit (132), triggered by the threshold crossing detector circuit (12) and reset by each pulse crossing the switch circuit (132).