FR2496362A1 - Isolated input power switching transistor control circuit - has opto-isolator or pulse transformer input pulse to initiate transistor turn on with self sustaining action and blocking - Google Patents

Abstract

Control impulses or variable delays are generated in a part tied to a cold earth. Only the input, driver and switching circuit are powered directly by a rectifier connected to the supply. Four foward biased diodes in series connected to the power transistor emitter provide passive security by blocking the transistor in the absence of a control signal. Control signal input isolation is provided by a pulse transformer or opto-isolator. The phototransistor or pulse transformer output transistor has an anti-saturation diode connected to its collector which is connected to the base of the first of the driver transistor pair. A secondary winding of the power transformer is connected by diodes and a limiting resistor to the power transistor to provide self-drive in the saturated state. Turing the control transistor on blocks the driver transistors switching on the power output transistor. The emitter diodes are bridged by a long time constant capacitor reservoir. The two diodes in series providing regenerative feedback are connected by resistor to the supply to lower the level necessary to start saturation using the feedback winding and by an anti-saturation diode to the collector of the output transistor. This limits the base voltage to within five times the forward biasses diode voltage drop.

Description

CONTROL CIRCUIT OF A SWITCHING TRANSISTOR
POWER AND DEVICE COMPRISING SAME
The invention relates to a control circuit for a power transistor comprising a galvanic isolation element at its input, such as a photocoupler or a pulse transformer controlling a switching transistor. The use of a photocoupler makes it possible to achieve an insulating dielectric strength of 4 to 10 kV approximately.

In the control circuits generally used, in particular, in switching power supply devices or inverters, that is to say DC-DC or DC-AC converters, where the collector-emitter path of a switching transistor power is generally inserted with a chopping or switching inductor which stores an energy (LI2MAX / 2) increasing with the duration of the saturated state of the transistor and restores it in the form of a voltage pulse during the blocked state of the latter, the control of its base generally requires a non-negligible power supplied by a drive circuit which may include an amplification stage equipped with a switching transistor and a pulse transformer (see, for example, the article by PEGLAU published on pages 235 to 238 of No. 8, of the German review "FUNKSCHAU" of the year 1971, or the articles by HETTERSCHEID and HETTERSCHEID and VAN SCHAIK published respectively on pages 131 to 140 and 203 to 215 from the American quarterly magazine "IEEE TRANSACTIONS ON BROADCAST
AND TELEVISION RECEIVERS ", volume BTR-16, numbers 2 and 3, from the months of May and August 1970 or in the articles of VAN
SCHAIK respectively entitled "AN INTRODUCTION TO SWITCHED MODE POWER SUPPLIES IN TV RECEIVERS" and "CONTROL CIRCUITS FOR SMPS IN TV RECEIVERS", respec
published in the British journal "MULLARD TECHNICAL
COMMUNICATIONS "number 135, from July 1977, on pages 181 to 195, as well as in the Dutch English-language journal" PHILIPS 'ELECTRONIC APPLICATION BULLETIN "volume 34, number 3, on pages 93 to 108, from the year 1976 , and on pages 210 to 226 of number 136, of the month of October 1977 of the first of these journals and on pages 162 to 180 of volume 34, number 4, of the year 1976 of the second of these). In the applicant's previous publication EP-A 0 005 391, it has also been proposed to use as the "attack" stage (called "driver" in English) the cascade combination of a phase-shifting stage (called "phase- splitter "in English) and a power amplifier assembly of the" push-pull series "type composed of bipolar junction transistors and frequently used in logic integrated circuits of TTL type.

Another method of controlling the base of a switching transistor uses, in its basic circuit, a reaction winding magnetically coupled either to the switching inductance, as is, for example, the case in the publications DE -B -21 60 659, 23 36 11Q and 24 17 628 (or FR-A- 2 267 654) from SIEMENS
AKTIENGESELSCHAFT, in the articles of DANGSCHAT respectively published on pages 245 to 255 of the American review "IEEE
TRANSACTIONS ON BROADCAST AND TELEVISION RECEIVERS ", volume BTR-17, number 4 of the month of November 1971, on pages 782 to 784 of the German review" FUNK-TECHNIK "number 20, of the year 1971 and of DANGSCHAT et al. pages 40 to 43 of the German review "FUKSCHAU" number 5, of the year 1975, in the publications FR-A- 2 345 762, 2 447 639 and 2 448 820, as well as in the French patent application n ° 79 16905 of June 29, 1979 of the present applicant, either to the primary winding of a current transformer, which is connected in the collector circuit of the switching transistor, in series with the switching inductance, as is the case , for example, in publication FR-A
1,403,260 or in French patent application no. 80 14912 of July 4, 1980 by the applicant. In self-oscillating circuits (of the blocked oscillator type, known as "blocking oscillator" in English or "Sperrschwinger" in German) or auto -excited, the conduction and saturation of the switching transistor results from a cumulative process due to the reaction and its blocking results from the ignition of a thyristor short-circuiting the reaction winding and possibly allowing application simultaneous negative bias voltage between the base and emitter of this transistor in order to accelerate the evacuation of minority carriers stored in the base during its saturation, i.e. to reduce the storage time t5 (called "storage time" in English) which causes an additional delay during the blocking of the transistor. It is known to have for this purpose a "battery" whose positive blade is coupled to the emitter of the switching transistor and whose negative pole is coupled by via the primed thyristor, at the base of this transistor. In publication DE-A- 21 60 659 and the first two articles of DANGSCHAT mentioned above, this "battery" consists of a capacitor (8, C1) of strong value (100 microfarads) positively charged through a diode (32, D8-B2540) by a secondary winding (9-33, 7-8) of the transformer (4, AZ 3600) including the primary winding (3, 1-4 ) constitutes the switching inductor located in the collector circuit of the switching transistor (1, T2). In French patent application No. 80 14912 this battery (80) is constituted by several diodes connected in series to drive in the same sense that the base-emitter junction of the switching transistor, connected in parallel with a capacitor retaining its charge during the blocking of the diodes, the assembly being inserted between the base of this transistor and one of the terminals of the reaction winding the other terminal of which is coupled, via a measuring resistor e of its base current, to the emitter thereof, the thyristor being coupled in parallel with this winding.

 In publication DE-A- 23 04 545 (PHILIPS), the "battery" is inserted between the emitter of the NPN switching transistor and the negative power pole and comprises a capacitor which is charged by the emitter current of it. In parallel with the capacitor, a resistance or a Zener diode is provided at the terminals of which the emitting current causes a voltage drop which is applied, when the transistor is blocked by means of an electronic switch uniting the base with the aforementioned negative pole, which also short-circuits the reaction winding, at the base-emitter junction of it, so as to polarize it upside down.

When a chopping supply device is of the so-called "accumulation" type (called "flyback" in Anglo-American literature), the energy stored in the chopping inductor (L I2MAX / 2) is restored during the interval for blocking the transistor, using a rectifying diode the cathode of which is connected to the junction of the chopping inductor with the collector of the transistor (non-isolated load) or one of the electrodes of which is connected to one of the terminals of a secondary winding magnetically coupled to the chopping inductor and the other electrode of which is connected by means of another filtering capacitor to the other terminal - of this secondary winding. then leads, during the transistor blocking intervals, a linearly decreasing current which charges this other filtering capacitor to an output voltage Vs which is, in particular, a function of the stored energy, of the transformation ratio between the primary and the and the resistive component of the impedance connected between the terminals of this other capacitor. The value of the supply voltage VA between the output terminals of the rectifier assembly being determined by the alternating network being able to fluctuate, the switching inducance LD or switching and the transformation ratio being given constants of the circuit, the main element allowing a regulation of the output voltage
VS is constituted by the ratio between the duration of the saturation interval of the transistor TS and either the duration of its interval of
TB blocking, i.e. the sum of the respective durations (T5 + TB) of the TS saturation and TB blocking intervals, which is equal to the
feed device repeat or recurrence period TR
tation. This TS / TR = TS / Ts + TB ratio then constitutes the ratio
cyclical (called "duty cycle" or "time ratio" in English literature
American).

 The duty cycle can then be varied by varying the duration, either of the saturation interval T5 while keeping the recurrence period TR constant, or of this latter TR (or its inverse which is the repetition frequency FR) with (! In constant saturation interval TS, in order to maintain an output voltage Vs constant regardless of the supply voltage VA or the average value of the rectified current ïRM consumed by the load (within a range of In the arrangements of publications DE-B- 21 60 659, 23 36 111 and 24 17 628 - (= FR-A- 2 267 654) mentioned above, it is in particular the frequency of repetition FR which varies with the load and the duration of the saturation interval TS which varies with the supply voltage VA derived from the network. In the device described in the aforementioned publication FR-A- 2 345 762 of the applicant, the repetition frequency FR determined by a pulse generator is substantially constant and l he respective durations of the saturated TS and blocked TB states of the switching transistor, the sum of which (TS 4 TB = TR) is therefore constant, vary jointly in opposite directions, simultaneously according to respective functions of the power consumed by the load, of the supply voltage A continues and the output voltage Vs, in order to maintain the latter constant.

In another type of device described in the publication DE
B- 1 234 836, where the electronic switching switch is made bidirectional by connecting a diode in anti-parallel with the collector-emitter path of the switching transistor which is controlled on its basis by rectangular signals with frequency and duty cycle constant and or the chopping inductance constituted by the primary winding of a transformer (THT) forms a parallel resonant circuit with a tuning capacitor, during the opening of the switch and which is connected in parallel with the latter , it is this period of resonance which determines the duration of the opening of the switch. Indeed, when the transistor saturated beforehand and conducts a collector current linearly increasing which crosses the chopping inductance, is controlled on its base to be blocked, the bidirectional switch it forms with the anti-parallel diode opens after the evacuation of the minority carriers stored in its base (i.e. after a n storage time, the duration of which is a function of the collector and base currents of the transistor at the time of application of a reverse or zero polarization at its base-emitter junction, increased by the delay time for decay specific to each type of transistor even without saturation), and the chopping inductor resonates with the tuning capacitor for a little more than half a period of sinusoidal oscillation of the voltage across its terminals and cosine of the current, in the way classic of the scanning-line transistor output stages. In this circuit, the regulation of the output voltage Vs rectified and filtered after sampling at the terminals of a secondary winding of the transformer, is carried out by the variation in function thereof, of the phase delay between the opening time of this switch and the instant of connection, using another bidirectional electronic switch, of another tuning capacitor in parallel with the first in order to modify the duration of the half-resonance period.

 When the delay is greater than the half-period of the primitive resonant circuit i there is no regulation, but when it is less than this, it has the effect of prolonging it so as to decrease the duration of the switch closing interval and, therefore, the energy stored in the chopping inductor. When the oscillation voltage, modified or not, after its first zero crossing, exceeds the conduction threshold of the anti-parallel diode, it becomes forward biased so as to close the switch and conduct a negative current linearly increasing to a zero value, through the chopping inductor, in the DC power source, hence the name of the diode called parallel or shunt recovery in the scan- line output stages. . note that the rectangular signal controlling the base of the switching transistor must keep it blocked beyond the end of the resonance half-period of maximum duration which corresponds to a minimum or zero delay, where the two capacitors tuning are connected in parallel from the moment of blocking of the switching transistor, the device then supplying a minimum voltage Vs at its output. The upper limit of the duration of the transistor blocking signals is not critical, thanks to the presence of the anti-parallel diode, since the base-emitter junction of the switching transistor should only be biased live shortly before the cancellation of the current through this anti-parallel diode. In summary, this device comprises a variable delay generator as a function of a voltage.

 In a switching power supply device combined with a scan-line output stage, described for example, in the publications FR-A- 2 275 092 and US-A- 3 999 102 (embodiments of FIGS. 6 and 14 ) or FR-A- 2 425 186 and EP-A- O 005 391, the cutting circuits as well as the scanning-line output stages respectively comprise bidirectional switches of the type described and capacitors tuning their switching inductances (which in the case of the output stages are called respectively return capacitor and deflection-line coils in parallel with a first winding of the transformer-line).

The output stage is no longer connected here to a DC voltage source, but the deflection coils are connected in series with a capacitor called "S effect" or "go" (because it is he who supplies them with voltage during the sweep-line intervals) and the first winding is connected in series with another so-called supply capacitor (because it is this which supplies this stage and which is charged using the cutting circuit).

 The cutting circuit is, on the other hand, supplied with the aid of a rectifier assembly providing a high direct voltage, between the terminals of which the cutting inductance and the bidirectional switch are connected in series, with the tuning capacitor in parallel either with one or with the other. A second winding of the line transformer, well isolated from the first, constitutes or forms part of the chopping inductor, so that the transfer of energy between the two circuits takes place via this transformer. resonance period of the parallel resonant circuit formed by the deflection coils, the first winding and the return capacitor determines the duration of the line return interval which is normalized for each standard (12 microseconds for European standards of 625 lines per image and 25 frames per second each with two interlaced frames).

 When that of the resonant circuit formed by the chopping inductor and the tuning capacitor is appreciably higher, such as, for example, around half a line period (32 microseconds), the regulation of the amplitude of the current sawtooth traversing the line-deflection coils during the forward periods by that of the voltage across the terminals of the supply capacitor of the output stage, is carried out by varying the phase delay between the respective cut-off times of the sweep and cut-off switches, obtained using a so-called variable delay generator as a function of this voltage or one of the peak amplitudes of the line-return pulses taken from another secondary winding , called auxiliary, of the line transformer.

 In any case, it is necessary to isolate the ground of the television receiver of which the stage of - sweep-line output, called cold, forms a part, from that connected to the network of the cut-out power supply device, called hot. protect the user. This isolation can be obtained, for example, by inserting between the output of the variable delay generator constituted by a pulse width modulator (called "pulse-width modulator" or "PWM" in English) providing pulses whose front rise coincides with those of synchronization-line signals or return-line pulses, the duration of which depends on the voltage across the supply capacitor and the differentiated rear edge of which serves to control the blocking of the circuit switching transistor cutting, a pulse and isolation transformer (see US-A-3,999,102 cited above). The variable delay generator is then connected to the cold ground.

 Another means of isolation is to use the auxiliary winding of the line transformer, on the one hand, for the triggering of a sawtooth waveform generator at the line frequency and, on the other hand part, to obtain a DC voltage proportional to the voltage across the supply capacitor by correcting that of the signal alternations across this winding which occur during the sweep intervals, either to change the slope of the teeth of saw as a function of a current derived from this rectified voltage, or to apply this as a reference voltage to one of the inputs of an analog voltage comparator, the other input of which receives the saw teeth with constant slope. variable delay generator is then connected to the hot ground and requires an autonomous supply of low DC voltages derived, by rectification and filtering, from the AC network.

 The control circuit of a switching transistor according to the invention is applicable in particular to most of the line-scanning circuits and known chopper power supply devices mentioned above, and in particular it allows good insulation between the power circuit (chopping) connected to the network and the circuits of use, without requiring a separate auxiliary power supply for the driving circuit of the chopping transistor and allowing to place the generator of command pulses or variable delay in the part connected to the cold ground, which however must be supplied in isolation from the network. The rectifier assembly connected to the network therefore directly supplies only the input and drive stages and the chopping circuit.

 The control circuit, object of the present invention, allows in particular a new use of the "battery" inserted here between the transmitter and the negative power pole as well as a galvanic isolation element located at a relatively power level. low, making it possible to improve the performance of the aforementioned circuits forming part of the state of the art while ensuring passive safety (blocking of the power transistor in the absence of control).

 According to the invention, a control circuit of a power switching transistor whose collector-emitter path is connected in series with its inductive load between the poles of a first high-voltage direct source and whose base is supplied by means of a reaction winding magnetically coupled to this load or to the primary winding of a current transformer, traversed by the collector current of this transistor, is mainly characterized in that it comprises in cascade: a stage galvanically isolated, the output element of which consists of another transistor and a transistor drive stage, consisting of a logic inverter, the output transistor of which is controlled so as to be in a complementary state relative to this other transistor , galvanically joined by its collector at the base of the switching transistor in order to control the blocking of the latter, in the absence of a command signal of predetermined polarity and amplitude inée at the input of the isolation stage causing the saturation of this other transistor which, as well as the drive stage, is supplied by the first source by means of a network of resistors.

The invention will be better understood and other of its characteristics and advantages will appear on reading the following description and the attached drawing, given by way of example, in which:
- Figure 1 is the block diagram of an embodiment of the control circuit according to the invention applied to a first type of known switching power supply device; and
- Figure 2 is the block diagram of the application of this control circuit to another type of known switching power supply.

 In these two figures, the same elements are designated by the same references.

 As the switching power supply device of FIG. 1 has an operation analogous to the conventional output stage of a line-scanning circuit of a television receiver, it is obvious that the control circuit according to the invention can equal ment be used in such an output stage equipped, in particular, with a switching transistor self-powered by reaction. It will also be noted that the control circuit can also be used in switching power supply devices of the type described, in particular, in the publications DE-B- 2 160 659, DE-A- 2 304 545 or US-A- 3 925 717, if one substitutes there for the "battery" existing between the emitter of the power switching transistor and the blade of the high-voltage power supply which is joined to it, or if one inserts therein, a battery of the type described in the following, known per se.

In Figure 1, there is shown an advantageous embodiment of the control circuit of the invention, used in a switching power supply device of the type described in the French patent application No. EN 80 14 912 filed on July 4, 1980 by the applicant company. This device comprises a high-voltage source of continuous supply 1 which includes a two-wave rectifier bridge 2 whose diagonally opposite input terminals 3 and 4 are intended to be respectively connected to the blades 5 and 6 of the AC supply network (220 \ 'effy 50 Hz), and whose output terminals 7 and 8, also diagonally opposite, are respectively connected to the positive and negative armatures of an electrochemical filtering capacitor 9 of capacity high and constituting pale 7 positive and negative 8 from the source
I, the latter of which is connected to an earth 10, called hot or primary, because it is not galvanically isolated from the AC network. When it is desired, as is generally required by most new safety standards, isolate the part accessible to the user for current adjustments or controls (volume or gain of the sound channel, contrast or gain of the luminance channel, brightness or brightness, saturation or gain of the chrominance channel, change of program or band frequencies, etc.) of the AC network, the part supplied by the chopping supply device is generally isolated by means of a transformer whose primary winding constitutes or forms part of the chopping inductance This inductor or this primary winding are connected in the collector circuit of the power switching transistor to store energy during the conduction (saturation) intervals thereof, in the form of a current. collector increasing substantially ~ linearly over time.

The energy stored in the chopping inductance LD is sensi i2 'L equal to LD IL MAX12, where 1L MAX is the peak value of the current flowing through the inductor, equal to VA.TS / LD, where VA is the voltage d power supply across the inductance LD with the transistor saturated and T5 is the duration of the time interval during which the latter is saturated.

 The positive pole 7 of the source 1 is joined, by means of a protective fuse 1 1, to a terminal 12 which constitutes the common positive supply terminal of the cutting circuit 20, of the control circuit 30 and d part of the isolation circuit 50 or input.

With regard to the cutting circuit 20 which, arranged in a similar manner to that described in the aforementioned French patent application no. 80 14 912 of the plaintiff taking up in part the teachings of the publication FR-A- 1 403 260 (or US-A3 317 816) mentioned above, includes a chopping inductor 21 which can be constituted by or include a primary winding of a supply or line transformer, connected by one of its terminals to the common positive supply terminal 12.The other terminal of the inductor 21 is connected in parallel to one of the terminals of a primary winding 31 of a current transformer 32, the other terminal of which is connected to the collector of a bipolar transistor power switching 22, of NPN type, at the cathode of a first diode 23 connected in anti-parallel with the collector-emitter path of transistor 22 and forming with the latter a bidirectional switch with unidirectional control 24 and at one of the terminals d 'a first con tuner densifier 25. The anode of the first anti-parallel diode 23 is connected to the primary or hot ground 10 constituting the common negative supply terminal connected to the negative pale 8 of the source. The other terminal of the first capacitor 25 can be connected either to, as shown here, the primary ground 10 or to the positive terminal 12 (see FR-A
2 382 812) so as to be connected in parallel with the inductor 21 to form with it a parallel resonant circuit during the opening of the switch 24.

 The emitter of the power switching transistor 22 is joined here to the primary ground terminal 10 through a battery 26 constituted by the parallel mounting of a second storage or storage capacitor 27, of a first resistor 28 and d 'a voltage generator assembly 29 comprising four diodes 290, 291, 292 and 293 in series leading in the same direction as the transitor 22 and whose conduction threshold voltages in the direct direction (called diode forward voltage threshold "or VFD in added together, provide the voltage of the battery 26 (equal to 4 VFD or approximately 2.8 V for silicon diodes) at which the second capacitor 27 charges to polarize the emitter of transistor 22 positively with respect to ground primary or hot 10.Note here that a similar battery with diodes in series and with capacitor in parallel, has already been described in particular in the aforementioned French patent application No. 80 14 912, where it was placed in the circuit of transistor base switching, in series with the secondary "reaction" winding of the current transformer.

The secondary winding 33 of the current transformer 32 which is part of the control circuit 30, is connected here, by one of its terminals, to the primary ground 10, its other terminal being joined, by means of two others diodes 34 and 35 in series, at the base of the switching transistor 22, so that it is supplied with base current, during the intervals of saturation thereof, by the current induced in the secondary winding 33 by the current linearly increasing collector which then traverses the primary winding 31. The second diode 35 of the series arrangement 34-35, the cathode of which is connected to the base of the transistor 22 and the anode of which is connected to the cathode of the first 34, is connected in parallel with a second resistor 36. The junction 37 of two other diodes 34, 35 and of the resistor 36 is connected, on the one hand, to one of the terminals of a third resistor 38 and, on the other hand, to the collector of a second bipolar transistor 39, of the same type (NPN) as the t switching transistor 22, whose ltem - ;; eur is connected, on the one hand, to the base of a third transistor 40, of the same type (NPN) as the other two 39, 22, and on the other hand, to the by means of a fourth resistor 41 to the primary ground 10. The emitter of the third transistor 40 is directly connected to the primary ground 10 and its collector is assembled, by means of an inductive dipole q2 composed of a coil of shock 43 used to protect against excessive rates of rise of the current and of a sixth resistor "t4 connected in parallel to the base of the switching transistor 22
The base of the second transistor 39 is joined, on the one hand, by means of a mounting of a seventh resistor 45 and an eighth resistor 46 in series, to the common positive supply terminal 12 and, d on the other hand, at the junction of the output terminal 51 of the isolation stage 50 constituted here by a photocoupler of the type
FCD 810 or 820 or MCT-2 respet -iverrient from the American companies "FAIRCHILD" and "MONSANTO CHEMICAL", for example, which is connected to the collector of the photo-transistor 52 forming the output element of the latter, with the cathode of 'an eighth diode 47 called anti-saturation, also called "catch-collector" (or "collector-catcher diode" in English - see 9 for example, on pages 203204 of DELHOM's work entitled "DESIGN AND APPLICATION OF TRANSISTOR SWITCHING CIRCUITS ", published by Mc GRAW-
HILL BOOK COMPANY in 1968) .This isolation stage 50 is therefore provided with an open collector transistor 52, as an output element.

This diode 47, by preventing the complete saturation of the switching transistor 52 to the collector to which it is connected, allows the acceleration of the cut-off or decrease response by significantly reducing the storage time (of minority carriers in the base - see, for example, on pages 429-432 of the book edited by WALSTON and MILLER, entitled "TRANSISTOR
CIRCUIT DESIGN ", published by the above-mentioned publisher, in 1963), because it does not allow the collector-emitter path of transistor 52 to reach its minimum saturation voltage VCE sat (which is a few tenths of volts).

 The use of this anti-saturation diode 47 to calibrate the minimum collector voltage V, MIN to the voltage VFD corresponding to the direct conduction threshold of a diode and to prevent saturation of the photo-transistor 52, is made possible by the particular arrangement of the drive stage 30, the input of which is brought together, by means of the two respective base-emitter junctions of the second 39 and of the third transistor 40, to the primary ground 10, so as to render the threshold voltage required to saturate the two transistors 39, 40 equal to 2 tFDX, that is to say greater than Vc MIN = VFD. This placing in series of two base-emitter junctions then makes it possible to control the simultaneous blocking of the two transistors to which they belong, by means of a voltage equal to VFD (0.7 V).

 It should be noted here that, if the gain of the common emitter stage comprising the third transistor 40 was sufficient, the base-emitter junction of the second transistor 39 could be replaced by a diode leading in the same direction as this junction to obtain blocking of the third transistor 40 by the saturation of the photo-transistor 52.

 The junction 48 of these two resistors 45 and 46 in series joining the common positive supply terminal 12 at the junction of the base of the second transistor 39 with the collector of the phototransistor 52, is connected to the other terminal of the third resistance 38 , so as to supply the collector of this second transistor 39 through the latter.

According to the invention, the anode of the anti-saturation diode 47 is connected to a branch 294 of the battery 26, located in the emitter circuit of the switching transistor 22. This branch 294 is formed here by the junction between the cathode of the third 291 and the anode of the fourth diode 292, which provides continuous potential
of 2 VFD equal to approximately 1.4 V, at least during the intervals of
conduction of the switching transistor 22 during which it is
necessary to maintain the collector-emitter voltage of the photo
transistor 52 above a predetermined value.
direct voltage drop across the anti-saturation diode 47 is
of VFD = 0.7 V approximately, the minimum collector-emitter voltage of the
phototransistor 52 will be calibrated (called "clamped" in English) at + 0.7 V
approximately with respect to the primary mass 10, to which is connected, to
through the second output 53 of the photocoupler 50, its transmitter.

It is the absence of saturation of the phototransistor 52 which allows the
operation of the photocoupler 50 at the line frequency.

The base of the phototransistor 52, connected to the third output 54
of the photocoupler 50, is brought together, via a ninth
resistance 55, to primary mass 10 in order to reduce the time of its
switching to the cutoff and ensuring that it remains blocked
in the absence of illumination (of kradiati.on) of its base by a diode
light emitting (LED) or electroluminescent 56 which constitutes
kills the input element of the photocoupler 50. The anode and the cathode of
the diode 56 are respectively connected to the first 57 and to the
second photocoupler 50 input terminal 58, to allow sound
power supply using a current in the form of recurrent rectangular pulses through a tenth input resistor 59.

The other terminal 60 of the resistor 59 can be connected to a generator
recurring pulses (not shown) which can be formed,
for example, by a variable delay generator depending on a
tension, such as those of the type described in the aforementioned French patent application no. 79 16 905 (not yet
published) or in - publications EP-A- O 005 391 or US-A
3,999,102 or DE-B- 1,234,836, or also, when the
power circuit (chopping) 20 is constituted by a stage of
output from a scanning-line circuit of a television receiver, by
a time base lines supplying command pulses to the
line frequency, such as, for example, a mono integrated circuit
lithic type TBA 920 or TDA 2590 (from the company THOMSON
CSF, SESCOSEM semiconductor division, for example) or
TDA 2571 (from NV PHILIPS 'GLOEILAMPEN
FABRIEKEN).

It should be noted here that when the power circuit 30 of the switching power supply device does not include a tuning capacitor 25 and an anti-parallel diode 23 and when, optionally, the reaction winding is magnetically coupled to the chopping inductor 21 which is constituted by the primary winding of a supply transformer (that is to say in the absence of a current transformer), the control circuit supplying the emitting diode of light 56 can be arranged in the manner described in the above-mentioned VAN SCHAIK articles (monolithic integrated circuits of the TDA 2640 or TDA 2581 type of
NV PHILIPS 'GLOEILAMPENFABRIEKEN), for example, allowing operation at fixed frequency and regulation, by duty cycle, of the output voltage.

 When energizing the circuit of Figure 1, by connecting these terminals 5 and 6 to those of the AC network (220 Vefff 50 Hz), the bridge rectifier 2 charges the filtering capacitor 9 so that after a short time interval it provides a high continuous voltage of around 300 V between its output terminals 12 and 10. The base of the phototransistor 52 being connected to ground and no current passing through the light-emitting diode 56, the latter remains blocked and the second NPN 39 transitor mounted in a hybrid assembly close to that called a common collector, becomes conductive because its base is then supplied by the positive voltage of terminal 12, through resistors 46 and 45 in series and that its collector is also supplied through resistors 46 and 38 in series. The emitter of the second transistor 39 is connected to the primary ground 10 via, on the one hand, the fourth resistor 41 and, on the other hand , of the diode constituted by the junction ion base-emitter of the third transistor 40, the latter becomes saturated substantially at the same time as the second 39, the emitter current of which supplies the base current of the third 40, as soon as the circuit is energized.

The collector currents of the two saturated transistors 39 and 40 add up in the third resistor 38 causing between its terminals a voltage drop proportional to their sum which is added to the base current of the second transistor 39 in the eighth resistor 46 by causing between the terminals of the latter another corresponding voltage drop which is to be subtracted from the 300 V on the terminal 12 to obtain the voltage V48 between the junction 48 and the primary ground 10 which constitutes the supply voltage of the stage d attack 30 of the switching transistor 22, when the transistors 39 and 40 are saturated. The above considerations are involved in part in the calculation of the values of these resistors which can be chosen, in certain cases, the following:
R38 = 1.5 kiloohms, R45 = 15 kiloohms and R46 = 75 kiloohms, for example.

 The third transistor 40 then forms a common emitter stage, connected to the primary ground 10, with its collector assembled, through the inductive dipole 42 which intervenes only during the blocking of the switching transistor 22, at the base thereof and also, through the assembly composed of the seventh diode 35 which is conductive at least during the saturated state of the two driving transistors 39 and 40, and of the second resistor 36 at the junction 37 of the diodes 34, 35, of the third resistor 38 and the collector of the second transistor 39.

 Even when the battery 26 ensuring the positive polarization of the emitter of the switching transistor 22 is not charged by a first conduction period thereof, the collector of the third saturated transistor 40 polarizes the base of the switching transistor 22 at VCE sat of a few hundred millivolts at most, relative to its transmitter (joined by the first resistor 28 to the primary ground 10), so as to maintain it in the blocked state.

 When the control circuit (not shown) supplies a positive current through the tenth resistor 59, to the light-emitting diode 56, the latter emits electromagnetic radiation (due to the phenomenon of electroluminescence of the materials constituting its junction) which irradiates the base of phototransistor 52 so as to make it saturated.

 The saturation of the phototransistor 52 which, when the battery 26 is not charged, brings the base of the second transistor 39 from VB 3g = 1.4 V approximately to VB 39 = VCE S2sat of a few tens or hundreds of millivolts, controls the blocking almost simultaneous of the transistors 39 and 40 of the driving stage 30, thanks to a rapid evacuation of the carriers stored in the bases of these respectively through the phototransistor 52 and the fourth resistor 41 joining the base of the transistor 40 to its emitter .

 The approximate grounding of the output 51 of the photocoupler 50 causes, in addition to blocking the transistprs 39 and 40, the positive polarization (relative to its emitter) of the base of the NPN transistor 22, via the voltage divider. formed by resistors 45 and 46 in series, supplied by high voltage (300 VDC) and from the network formed by resistors 38 and 36 in series, the last 36 of which is "shunted" by diode 35 so as to limit the voltage V37 at junction 37 to 5 VFD + VBE 22max = 6 VFD (approximately 4.2 volts). The sixth diode 34 joining the junction 37 to the so-called "hot" terminal, not connected to ground 10, of the secondary winding 33 (reaction) of the current transformer 32, is then biased backwards so as to ensure the isolation between junction 37 and primary ground 10. This positive polarization of the base causes the switching transistor 22 to conduct first through its emitting resistor 28 until the voltage drop caused at these terminals exceeds the conduction threshold voltage of the four diodes 29 in series (4 VFD = 2.8 V), the base current from junction 37 must also first pass through resistor 36 before reaching the conduction threshold of the diode 35. We can then say that thanks to the presence of resistors 36 and 28 shunting the diode 35 and the four diodes 29 of the battery 26 respectively, the conduction threshold of the transistor 22 is reduced from 6 VFD to approximately VFD + VBE = 2 VFD.

qui doit être rendu supérieur à 1B 22min. However, for the positive reaction via the current transformer 32 to be established, it is necessary to exceed a minimum value of the base current of the transistor 22 which is around 1B 22min = 1 milliampere. This makes it possible to establish a relationship between the basic current ib, the supply voltage V12, the conduction threshold voltage of the diodes 35, 290 to 293 in series with the base-emitter junction of the transistor 22 (6 VFD) and the values of resistors 46, 45 and 38, which constitutes another criterion for the calculation of the latter. We can then write (in the absence of resistances 28 and 36):

which must be made greater than 1B 22min.

The values of resistors 46, 45 and 38 may however be increased, when the values of resistors 36 and 28 are chosen to be relatively low (by a few hundred ohms, for example). For the values of R38, R45 and R46 previously mentioned, it is advantageous to choose R28 = 220 ohms and
R36 - about 75 ohms.

 As soon as the basic current has exceeded its minimum value IBmin mentioned above, the collector current begins to increase and the collector - increasing current flowing through the primary winding 31 of the current transformer 32, induces in its secondary winding 33 an equally linearly increasing current which generates a voltage at its terminals greater than the conduction threshold of the diode 34 whose cathode is polarized at 6 VFD. I1 therefore requires a voltage of 7 VFD at the terminals of the secondary winding 33 so that the self-supply of the base and, consequently, the regenerative effect bringing the transistor 22 to its saturation is established. The increasing collector current then travels through the four diodes in series 29 so as to charge the capacitor 27 to 4 VFD and to polarize the collector of the phototransistor 52 - to + VFD through the antisaturation diode 47. During all the time when the diode light-emitting 56 is supplied with current of sufficient intensity to keep the phototransistor 52 saturated, the switching transistor 22 will remain saturated, the two transistors 39, 40 of the driving circuit 30 will remain blocked, and the chopping inductor 21 is traversed d 'a current increasing linearly with time.

 As soon as the light-emitting diode 56 of the photocoupler 50 ceases to be supplied with current, it goes out and, consequently, the phototransistor 52 stops driving after the expiration of a time interval called total time at decay of the collector current successively composed of the storage time t5 (known as "storage time" in English) explained previously and of the delay time with decreasing tf (known as "fall time" in English), to then represent substantially an open circuit between the terminals output 51 and 53 of the coupler 50. This storage time ts is reduced to its minimum value, on the one hand, thanks to the resistor 55 joining the base of the phototransistor 52 to its emitter and the value of which is however chosen large enough to that it can saturate (R55 = 47 kiloohms) and on the other hand, according to the invention, due to the setting of the maximum collecsrur-emitter voltage VCE 52min to + VFD, which reduces the amount of carriers stored in the base and prevents the total saturation of transistor 52.

The phototransistor 52 being blocked, the current coming from the common positive supply terminal 12 through the resistors 46 and 45 in series causes the saturation of the two transistors 39 and 40 due to the fact that the emitter current of the second 39 directly feeds the base. of the third 40 which stabilizes at a positive potential of
VBE 40 equal to VFD (0.7 V) relative to the primary ground 10 to which the latter's transmitter is connected. Consequently, the potential of the base VB 39 of the second transistor 39 relative to the ground 10 will be 2 VFD, the voltage applied between the terminals of the voltage divider 45-46 then being equal to V12 - 2 VFD (3001.4 V).

The saturation of the third transistor 40 brings together the junction of the base of the switching transistor 22 and I the cathode of the seventh diode 35, through the dipole 42, so as to derive towards ground 10 the current induced in the secondary 33 of the transformer 32 by the collector current which continues to flow through its primary 31 during the storage time of the switching transistor 22 and to allow a negative base-emitter voltage to be applied to the latter
VBE 22 = - 2.8 V due to the battery 26 charged at 4 VFD polarizing its transmitter in order to accelerate the evacuation of the carriers from the base. The inductive dipole 42 already described in the publications FR-A-2 345 762 , 2 447 639 or 2 448.820 above, is a voltage differentiator and current integrator circuit preventing the sudden rise of the current passing through it, due to the high initial reactance presented by the shock coil 43 (of several microhenry) during transitions fast, compared to - the resistor 44 (of one or a few tens of ohms) which is connected to it in parallel. The dipole 42 therefore prevents too rapid growth, on the one hand, of the collector current iC 40 of the third transistor 40 and, on the other hand, of the reverse base current - iB 22 of the switching transistor 22 at the start of the storage time. In any case, the complete saturation of the third transistor 40 is well ensured, by the positive polarization of its base through the base-emitter junction of the second transistor 39 which will have to provide it with a sufficient base current for its collector to accept the high intensity current which is then supplied by the secondary reaction winding 33 and derived in part, through the collector-emitter path of the second transistor 39, at the base of the third transistor 40 and in part towards the collector-emitter path The presence of the charged battery 26 in the emitter circuit of transistor 22 and this complete saturation of the third transistor 40, resulting in a negative bias of the base of transistor 22 compared to rt to its transmitter, allow it to work during its blocking with a higher collector-transmitter voltage VCEX than if its base-transmitter voltage was only zero (VcE0). As soon as the switching transistor 22 is blocked, the diodes 29, the VFD conduction thresholds of which determine the battery voltage, become blocked and the second capacitor 27 which is connected in parallel with them, begins to discharge exponentially through the first resistance 28 which is parallel to it, with a time constant RC equal to the product of their respective values. The capacitance C27 of the capacitor 27 must therefore be chosen sufficiently large so that the negative bias of the base with respect to the emitter of the transistor 22 is maintained at least during the time interval when its direct collector-emitter voltage (VCE 22) is higher than the value VCE0 specified for a zero VBE polarization.

The peak value of the direct collector-emitter voltage is determined, on the one hand, by the supply voltage V12 (300 V) and, on the other hand, by the duration of the half-resonance period TR / 2 of the parallel resonant circuit formed by the chopping inductor 21 and the tuning capacitor 25 (see FR-A-2 272 092 and 2 425 186 above), as well as the duration of the interval TON during which the transistor switching 22 was previously kept saturated, which with the inductance L21 and the supply voltage V12 determine the energy stored in the inductance 21 by the maximum collector current IC 22 MAX = Vl2.tON / L2l, is equal to
C25 'C 25
CIC 22 MAX12 as well as C25 V2 / 2. For a half-period of oscillation TR / 2 substantially equal to a line half-period (TH = 64 microseconds) and a saturation time t ON substantially equal to a quarter of the horizontal scanning line period of a television and for a resistance R28 of 220 ohms, a capacity
C27 of the order of microfarad seems sufficient to maintain a substantially constant charge of the battery 26 (because R28C27 = 220 microseconds is then significantly larger than tOFF which is less than 3 TR / 4 = 48 microseconds).

environ jusqu'au retour à VCE 22 = l2 - 4 VFD, suivi d'une chute quasi4inéaire jusqu'à VCE 22 = - VFD, lorsque la première diode anti-parallèle 23 devient conduc::rice en courtcircuitant le condensateur d'accord 25 ainsi que le transistor 22 en série avec l'enroulement primaire 31 du transformateur 32 et avec la pile 26.Pendant que la tension à na jonction de l'inductance 21 et du condensateur d'accord 25 eL-ectue une excursion demi-sinusoYdale positive, le courant résonnant parcourant ces deux éléments subit une variation demi-cosinusoïdale entre une saleur maximale positive
IL 21 MAX et une valeur maximale négative ou minimale 1L 21 MIN avec un passage par zéro coïncidant dans le temps avec la valeur crête de l'excursion de la tension collecteur

(qui doit être inférieure à la valeur VCEX du type du transistor 22 choisi, spécifiée dans les catalogues des fournisseurs).As in the case of the switching power supply circuits of the publications DE-B- 1 234 836, FR-A- 2 272 092 and 2 425 186 and of the above-mentioned French patent applications No. 79 16 905 and No. 80 14 912 , the blocking of the switching transistor 22 causes, due to the recovery of the current IL 21 MAX in the inductor 21 by the tuning capacitor 25, a quasi-linear rise in the collector-emitter voltage VCE 22 thereof. ci, with a value of VCE 22 sat close to zero up to VCE 22 equal to V12 - VE 22 or V12 4 VFD and thereafter half a sinusoidal oscillation period of equal duration

approximately until the return to VCE 22 = l2 - 4 VFD, followed by a quasi-linear fall to VCE 22 = - VFD, when the first anti-parallel diode 23 becomes conductive by short-circuiting the tuning capacitor 25 as well as the transistor 22 in series with the primary winding 31 of the transformer 32 and with the battery 26. While the voltage at na junction of the inductor 21 and the tuning capacitor 25 takes a half-sine positive excursion , the resonant current flowing through these two elements undergoes a semi-cosine variation between a positive maximum salt
IL 21 MAX and a maximum negative or minimum value 1L 21 MIN with a zero crossing coinciding in time with the peak value of the excursion of the collector voltage

(which must be less than the VCEX value of the type of transistor 22 chosen, specified in the suppliers' catalogs).

 The anti-parallel diode 23 begins to conduct a current of opposite direction with respect to the collector current of the transistor 22, which is linearly increasing from a negative value IL 21 MIN (slightly higher than -IL 21 MAX due to the ohmic losses in the power circuit 20, in particular in inductance 21) towards a zero value. In order for transistor 22 to be able to pick up the linearly increasing current at the time of its zero crossing, its base must be previously polarized so as to allow it to become conductive. It is therefore necessary to apply the rising edge of the current pulse to the light-emitting diode 56 of the photocoupler 50 at an instant preceding the cancellation of the current in the anti-parallel diode 23, by at least the time necessary to block the two transistors 39, 40 of the driving circuit 30, which comprises successively the storage times and the delay in the respective decrease thereof. The instant of saturation of the phototransistor 52 is not critical provided that it is after the conduction of the anti-parallel diode 23 and earlier, at least of the cumulative delay of the transistors 39, 40 at decay, at the cancellation of the current in the latter. The base of the switching transistor 22 will therefore be positively biased with respect to its emitter as soon as the driving transistors 39, 40 whose quasi-simultaneous saturation causes its blocking, become blocked in turn , to enable it to become conductive as soon as the current in the diode 23 passes through zero and to subsequently become saturated by the well-known regenerative action of the current transformer 32.

 The extinction of the light-emitting diode 56 and the subsequent blocking of the phototransistor 52 make it possible to restart the process of blocking the switching transistor 22 as described above.

 When the switching transistor 22 suddenly turns off, that is to say when its collector current quickly becomes zero after reaching its maximum value IC 22 MAX "the fact of having connected one of the armatures of the tuning capacitor 25 (as well as the cathode of the anti-parallel diode 23) at the junction between the chopping inductor 21 and the primary winding 31 of the current transformer 32, causes the cancellation of the current flowing through it and the inversion of the voltage drop between this junction and the collector of the transistor and, consequently, a corresponding inversion of the electromotive force induced in its secondary winding 33. As this electromotive force is not negligible, it can go as far as to cause a discharge of the "avalanche" type in the sixth diode 34 through which an additional negative bias could be applied to the junction 37 respectively directly connected to the collector of the second transistor 39 and in directly connected to the collector of the third 40 and to the base of the switching transistor 22. Such a phenomenon is not unacceptable if the reverse breakdown voltage known as the avalanche voltage of the sixth diode 34, possibly added to that of the seventh 35, is not notably lower than the voltage at the terminals of the reaction secondary 33 due to this induced electromotive force, so as not to cause temporary desaturation of the transistors 39 and 40. Such an "avalanche" type breakdown would then carry out a partial evacuation carriers of the base, via the resistor 36, the diode 34 and the secondary 33 in series. If this phenomenon becomes annoying, it is easily eliminated by the connection in parallel with the secondary reaction winding 33 of a damping circuit (not shown) of reverse overvoltages comprising a diode whose anode is connected to the cold terminal. (connected to ground 10) thereof and the cathode of which is joined through a damping resistor to its hot terminal (connected to the anode of diode 34).

 In summary, the control circuit 30 of a switching transistor 22 whose load circuit 21 is inductive and whose base is supplied by a reaction winding 33 coupled to this inductive load 21 or by a current transformer 32 having a winding in series with the latter is remarkable in particular by the fact that it comprises in cascade: an isolation stage 50 whose output element is constituted by a transistor 52 optionally switching, and a drive stage 30 constituted by a logic inverter with switching transistors whose output is galvanically connected to this base to control the blocking of the switching transistor 22 by the saturation of a transistor whose collector-emitter path joins this base to ground, when the stage d isolation 50 does not receive a control signal on its input. In other words, the main advantage of the control circuit of the present invention is that in the absence of a control signal at the input of the isolation stage 50, the output transistor 52 of it ci, as well as the power switching transistor 22 controlled in phase with the previous one, remain blocked, thus conferring self-protection against nuisance tripping by a possible failure, such as a short circuit of a transistor or a capacitor, of the driving circuit 30, for example.

 FIG. 2 represents a block diagram of another possible application of the control circuit according to the invention to a supply device by chopping substantially of the type described in the publication FR-A-2 345 762, for example.

 In FIG. 2,1, the chopping (or switching) inductance of the power circuit 200 is constituted by the primary winding 211 of a supply transformer 210 of which a first secondary winding 212 constituting the reaction winding, is connected by one of its terminals to primary ground 10 and joined by its other terminal, through a current limiting resistor 301 (of a few or ten ohms) in series with the two diodes 34 and 35 , at the base of the switching transistor 22 to ensure self-supply thereof during the saturated state of the latter.

 The junction 37 of the cathode of the first 34 with the anode of the second 35 of the two diodes 34, 35 in series, is connected, in addition to the common point of the collector of the second transistor 39 and of the resistor 38, to the anode of a third diode 302 whose cathode is connected to the collector of switching transistor 22.This third diode 302 whose anode is connected to that of diode 35 whose cathode is connected (without resistance in parallel in this case) at the base of the switching transistor 22, plays the role of a second anti-saturation diode or "catch-collector" by fixing the minimum potential of the collector VC 22 MIN> when the latter is saturated, at the maximum value of base potential VB 22 MAX (equal to 5 VFD with respect to ground 10) so as to prevent it completely saturating when reaching VCE 22 sat (when this value is less than VBE 22 sat) When the base of transistor 22 and its collector are at the same potential, part of the current supplied by the winding ent ent 212 is diverted to the collector, so as to further reduce the number of minority carriers stored in the base.

The collector of the switching transistor 22 is joined here, in addition, on the one hand by a damping circuit 303 constituted by a parallel mounting of a diode leading in the same direction as the collector-emitter path thereof, and of a resistance of the order of a kiloohm or more, described in the aforementioned FR-A- 2 345 762, at one of the terminals of the primary winding 211 of the transformer and, on the other hand, by a circuit of classic protection (see, for example, in the article by Mc MURRAY entitled "OPTIMUM
SNUBBERS FOR POWER SEMICONDUCTORS "published in the American journal IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, volume IA - 8, no 5, from September October 1972, pages 593 to 600) against excessive rates of rise in collector-emitter voltage 304, to primary mass 10.

 The supply transformer 2 comprises at least one other secondary winding 213, one of its terminals of which. is connected to the secondary or cold ground 100, isolated from the primary or hot ground 10 which is, in galvanic connection with the alternating network.

The other terminal of this secondary winding 213 is connected to one of the electrodes (anode or cathode) of a rectifying diode 71 whose other electrode (cathode or anode) is connected to one of the armatures (positive or negative) of a second filtering capacitor 72 of the electrochemical type. The other armature (negative or positive) of this capacitor 72 forming with the diode 71 a rectifier assembly 70, is connected to the junction of the secondary ground 100 with the other secondary winding 213. Because the electronic circuit switch of power 200 is constituted by the single switching transistor 22 (without anti-parallel diode), the energy accumulated in the inductance of the primary winding 211 cannot be restored after its blocking and the cancellation of its collector current, than by one of the secondary windings 212 or 213. abrupt cancellation of the collector current flowing through the primary 211 causing an inversion of the voltage across all the windings 211, 212 and 213 of the transformer 210 and due to the presence of the isolation diode 34 which is then polarized at the upside down, the reaction secondary 212 is an open circuit, the other secondary supply winding 213 must be connected in such a way to the terminals of the rectifier assembly 70 so that the rectification diode 71 is conductive, when a voltage of reverse polarity to that which is present at its terminals during the saturation interval t ON of transistor 22, is applied to it.

correspondant à l'énergie ac cumulée dans le primaire à la fin de l'intervalle de saturation toN de ce dernier.The diode 71 then conducts, at least during the initial portion of the blocking interval tOFF of the transistor 22, a linearly decreasing current from a maximum value ID 71 MAX =

corresponding to the energy ac accumulated in the primary at the end of the latter's saturation interval toN.

The current in the diode 71 can decrease either until it is canceled, the switching power supply device of the accumulation type ("fly-back") is then at full discharge from the transformer 210, or up to a value ID0 different from zero and in the same direction as ID MAX 'function of the power consumed by the equivalent resistive component of the load (not shown) connected between the armatures of the second filtering capacitor 72. In the latter case, it is a device with incomplete accumulation and discharge of the transformer 210 in which the saturation of the transistor 22 must take place prior to the cancellation of the current in the rectifying diode 71 which then stops driving.

 The control or driver circuit 300 of FIG. 2 is, apart from the absence of the resistor 36 in parallel with the diode 35, (which can then be omitted or replaced by another resistor of appreciably higher value), is similar to that 30 of FIG. 1, and its functioning with regard to blocking and restoring conduction of the switching transistor 22, substantially in phase with the output transistor 502 (or 52) with open collector of the stage of isolation 500 (or 50), the collector of which is also joined, on the one hand, to the shunt 294 of the battery 26 by means of the anti-saturation diode 47 and on the other hand, to the input of the circuit control 300 (or 30) which is in turn joined on the one hand by the respective base-emitter junctions of the two transistors 39, 40 forming the inverter, connected in series to the primary ground 10 and, on the other hand by resistors 45 and 46 in series, to the pale positive supply 12.

 The isolation stage 500 here essentially consists of a pulse and isolation transformer 515 (known from articles in the aforementioned Dutch and British magazines) whose secondary winding 504 is connected in series with a bias resistor of the base and limiting the base current to saturation 505, between the base and the emitter of a bipolar switching transistor 502, which is of the NPN type.

 This emitter constituting one of the outputs 503 of the stage 500 is connected to the primary ground 10. The other output 501 of the isolation stage 500 is constituted by the open collector of this transistor 502 forming its output element . The primary winding 506 of the transformer 515 can equip an amplification or switching stage comprising a bipolar input transistor 507 of which it can constitute the load placed in the collector circuit (assembly with common emitter or common base) or emitter ( In the circuit of Figure 2, there is shown a common emitter stage which is connected to a secondary ground 100 isolated from the network and whose collector is joined by the primary winding 506 and a terminal power supply 512 to pale positive + VCC from a low-voltage DC source, isolated from the network, the negative pole of which is connected to this secondary ground 100 and which also supplies the pulse-cutting control generator (not shown) whose outputs are respectively joined to the inputs 510 and 511 of the circuit 500 to provide it with pulses posed recurrent (when the input trisistor 507 is of the NPN type) with a duty cycle in operation. n of the output voltage between the two armatures of the second filtering capacitor 72. The base of the input transistor 507 is connected to its emitter through a resistor 508 which polarizes it so as to keep it blocked in the absence of positive pulses on its input, as well as possibly, via another resistor 509, at the input 510 of the isolation stage.

It should be noted here that the primary winding 506 of the pulse transformer 505 can be supplied with any other
classical, provided that one applies between its hills pulses of constant (rectangular) or slightly increasing (trapezoidal) voltage different from zero and that it is, therefore, traversed by a linearly increasing current, having the effect inducing in the secondary winding a voltage which applied between the base and the emitter of the output transistor 502 brings it to saturation (VBE greater than VFD).

 However, the higher cost price of a transformer isolation stage 500 (called "transformer-coupled pulse gate" in English) is, due to the additional presence of an input transistor 507 and a component wound (transformer 515) requiring significant labor, as well as its lower insulating power for a low cost price, confer significant advantages to the use of the photocoupler 50 (of FIG. 1).

 Because the output transistor 52 or 502 is controlled in phase with the switching transistor 22, the latter is protected against nuisance tripping but the circuits of FIGS. 1 and 2 do not include any device for limiting the collector current of this transistor 22, which increases linearly throughout its saturation interval tON, unlike the devices described, in particular, in the publications DE-B 21 60 659, 23 36 111 and 24 17 628, FR-A- 2 345 762, 2 447 639 and 2,448,820, as well as in the aforementioned French patent applications Nos. 79 16 905 and 80 14 912, or therefore of automatic starting devices which would make them self-controlled (at a repetition frequency lower than that of the controlled operation by pulses applied to the inputs of isolation stage 50 or 500). It is therefore necessary to provide safety and start-up circuits for the chopping circuit 20 or 200 and its inverter driving circuit 30 or 300, both supplied by the rectified and filtered mains voltage, which would then be part of 'A control circuit (with independent power supply and isolated from the rectifier assembly 1 of the figure but with simultaneous energization, for example), which are not part of the invention.

 It is obvious that one can use to equip all the stages of the circuit, bipolar transistors of PNP type on condition of reversing, on the one hand, the blades of the high voltage supply and, on the other hand, the diodes, without departing from the scope of the invention.

Claims (10)

 1. Control circuit of a power switching transistor (22) whose collector-emitter path is connected in series with its inductive load (21, 211) between the blades (12, 10) of a first high source -continuous voltage (1) and the base of which is supplied by means of a reaction winding (212, 33) magnetically coupled to this load (211) or to the primary winding (31) of a current transformer (32 ) traversed by the collector current of this transistor (22), characterized in that it comprises in cascade: a galvanic isolation stage (50, 500) whose output element is constituted by another transistor (52, 502 ), and a transistor driver stage (30, 300), consisting of a logic inverter including the output transistor (40), controlled so as to be in a complementary state relative to this other transistor (52, 502), is galvanically joined by its collector at the base of the switching transistor (22) in order to control the blocking of the latter in the absence of a command signal of polarity and predetermined amplitude at the input (58-59, 510-511) of the isolation stage (50, 500) which should cause the saturation of this other transistor ( 52, 502) which, as well as the drive stage (30, 300), is supplied by the first source (1) by means of a network of resis-ances (45, 46, 38).  2. Control circuit according to claim 1, characterized in that the galvanic isolation stage (50) consists of a photocoupler whose input element is a light emitting diode (56) which, when is traversed by a direct current, illuminates the base of a phototransistor (52) whose collector and transmitter are respectively connected to the two outputs (51, 53) of this stage (50).  3. Control circuit according to claim 2, characterized in that the base of the phototransistor (52) is connected to a third output (54) of the isolation stage (50), which is joined by means of a resistor of polarization at the emitter (53) of the phototransistor (52).  4. Control circuit according to claim 1, characterized in that the isolation stage (500) comprises a pulse transformer (505) whose secondary winding (504) is connected in series with another resistor (505 ) between the base and the emitter of the other transistor (502).  5. Control circuit according to one of the preceding claims, characterized in that the drive stage (30, 300) comprises a second transistor (39) whose emitter is connected to the base of a fourth transistor ( 40); the base of which is joined, on the one hand, by means of two resistances in series (45, 46), to the positive pale (12) of the first source (1) and on the other hand, to the collector (51, 501) of the photo- (52) or of the other transistor (502), the emitter (53, 503) of the latter being connected together with that of the fourth transistor (40) to the negative pole (10) of this first source (1 ), the collector of the second transistor (39) being joined, on the one hand, by means of another resistor (38) at the junction (48) of the two resistors (45, 46) connected between the positive blade (12) and its base and, on the other hand, by means of a diode (35) leading in the same direction as the collector-emitter path of the transistor (40) output from the driver stage (30, 300) to the collector thereof, which is galvanically joined to the base of the switching transistor (22), and in that the junction (37) of the other resistor (38) with the diode (35) is joined by means of a separation diode (34) leading in the same direction that the latter (35), at one of the terminals of the reaction winding (33, 212), the other terminal of which is connected to the negative pole (10) of the first source (1).  6. Control circuit according to claim 5, in a power circuit (20, 200) of the type in which the emitter of the switching transistor (22) is joined to the negative pole of the first source (1) via a second source of low continuous voltage (26) comprising an assembly (29) of at least three diodes (290-293) leading in the same direction as the collector-emitter path of this transistor (22), in series, to be periodically traversed by the emitting current of this, and, respectively connected in parallel with the assembly (29) a capacitor (27) and a resistor (26) whose product is chosen so as to be greater than the interval between two successive conduction periods of this transistor (22), characterized in that the anode of that (292) of the diodes (290-293) whose cathode is joined by a single diode (293) at the negative pole (10 ) constitutes a derivation (294) of the second source (26) which is connected to the anode of an anti-saturation diode ration (47) whose cathode is connected to the collector of the photo- (52) or of the other transistor (502, in order to prevent its complete saturation while allowing to polarize the respective base-emitter junctions of the second (39) and of the third transistor (40), connected in series, so as to block them, when it is saturated.  7. Control circuit according to one of claims 4 to 6, characterized in that the junction (37) between the cathode of 3a separation diode (34), the resistor (38) and the anode of the diode (3f ') must the cathode is connected to the base of the switching transistor (22), is also joined to the latter> by means of a resistor (38) making it possible to reduce the threshold voltage necessary for switching on the transistor ( 22) and to its regenerative saturation by means of the reaction winding.  8. Control circuit according to one of claims 4 to 6, characterized in that the junction (37) between the cathode of the separation diode (34), the resistor (38) and the anode of the diode (35 ) whose cathode is connected to the base of the switching transistor (22), is connected to the anode of another anti-saturation diode (302) whose cathode is connected to the collector thereof.  9. Device for generating recurrent sawtooth-like currents in an inductive load (21) and or voltage pulses of substantially semi-sinusoidal shape across the latter, of the type in which this load (21) is connected in series with a bidirectional current and unidirectional voltage switch (24), constituted by an anti-parallel circuit, a switching transistor (22) and a diode (23), between the two poles (12, 10 ) of a first DC voltage source (1) and in which the junction of the load (21) and the switch (24) is connected to one  terminals of a capacitor (25) of which your terminal is connected to one  (10) or the other (12) pale from the source (1) to constitute with the inductance of the load (21) a parallel resonant circuit, during the periods of opening of the switch (24), characterized in that the switching transistor (22) is controlled on its base using a control circuit according to one of the preceding claims  dentes, the inductive load (21) possibly comprising the primary winding of a supply or line transformer which is connected in parallel with or coupled to the line-deflection coils.  10. Cutting supply device of the type in which the cutting inductor is constituted by the primary winding  (211) of a power supply transformer (210)> connected in series with a switching transistor (22) between the poles (12, 10) of a high-voltage- continuous source (1), in which a first secondary winding (212) of the transformer (210) which constitutes the reaction winding is connected between the negative blade (12, 10) of the source and the anode of a separation diode (34) whose cathode is galvanically coupled at least in the direction of conduction - of this diode (34) and of the base-emitter junction of the switching transistor (22), at the base of the latter and in which at least a second secondary winding (213) of this transformer (210) supplies at least one rectifier assembly (70) of which at least one rectifier diode (71) conducts during the blocking intervals of the switching transistor (22), characterized in that the latter is controlled on its base by means of a common circuit according to one of claims 1 to 8.