EP0413806A1

EP0413806A1 - Electronic starting and power supply device for preheated electrode fluorescent tubes

Info

Publication number: EP0413806A1
Application number: EP90904853A
Authority: EP
Inventors: Jean-Claude Harel
Original assignee: Individual
Current assignee: Individual
Priority date: 1989-03-10
Filing date: 1990-03-08
Publication date: 1991-02-27
Also published as: KR920700524A; FI905547A0; CA2028122A1; FR2644314A1; HU902766D0; HUT55578A; WO1990011005A1; JPH03504908A

Abstract

Dispositif électronique de démarrage et d'alimentation pour tubes fluorescents comprenant un ou plusieurs circuits de charge résonnant (51/56, 52/57, 55/59, 53/58); un étage d'amplification comprenant deux transistors (34, 43) montés en série; un circuit oscillant (130) pilotant les transistors (34, 43) et un circuit (120) variant la fréquence du circuit oscillant (130). Plus particulièrement, la valeur du condensateur de filtrage (16), lorsque présent, à la sortie du circuit redresseur (6) est inférieure ou égale à 0,04 muF. Le dispositif comprend de plus pour l'alimentation en courant continu du circuit oscillant (130) et du circuit variateur de fréquence (120) le long de lignes (e) et (f) un circuit d'alimentation temporaire (110) actif lors de la phase de démarrage et un circuit d'alimentation (150) dépendant de l'un des circuits de charge, lequel circuit (150) est actif en phase de fonctionnement permanent. Le dispositif comprend également, pour l'alimentation de l'étage d'amplification, un circuit (100, 105, 110, 115) assurant un minimum de tension.Electronic starting and feeding device for fluorescent tubes comprising one or more resonant load circuits (51/56, 52/57, 55/59, 53/58); an amplification stage comprising two transistors (34, 43) connected in series; an oscillating circuit (130) driving the transistors (34, 43) and a circuit (120) varying the frequency of the oscillating circuit (130). More particularly, the value of the filtering capacitor (16), when present, at the output of the rectifier circuit (6) is less than or equal to 0.04 muF. The device further comprises for the direct current supply of the oscillating circuit (130) and of the frequency converter circuit (120) along lines (e) and (f) a temporary power supply circuit (110) active during the starting phase and a power supply circuit (150) depending on one of the load circuits, which circuit (150) is active in the permanent operating phase. The device also comprises, for the power supply of the amplification stage, a circuit (100, 105, 110, 115) ensuring a minimum voltage.

Description

ELECTRONIC STARTING AND SUPPLY DEVICE FOR FLUORESCENT TUBES WITH PREHEATABLE ELECTRODES

The present invention relates to an electronic starting and supply device for gas or vapor discharge lamps provided with two preheatable electrodes such as fluorescent tubes.

Initially, a fluorescent tube was connected in parallel to a choke and this assembly was connected with an inductor in series to the network either to an alternating current of 280 V, 50 or 60 Hz. The start being random and the energy consumption of the inductor during the operating regime being very high, we then turned to electronic circuits generating a high supply frequency more reliably and more economically. Multiple devices have thus made it possible to significantly improve the overall performance of these fluorescent tubes.

The majority of these devices operate according to a principle of conversion of the low frequency alternating current of the sector into direct current, itself reconverted into high frequency alternating current. A new parasitic phenomenon then appears owing to the fact that the conversion alternating current / direct current is carried out normally thanks to the use of four diodes assembled in bridge of Graetz followed by a filtering of the pulsed current unidirectional resulting using a necessarily high value capacitor. Indeed, it can be observed that the absorption of current on the AC electrical network occurs over time in the form of a series of peaks corresponding to the maximum of the alternations, that is to say at the moment when the mains voltage exceeds the voltage present at the terminals filter capacitor. In other words, this filtering capacitor supplying the electronic circuits permanently retains the majority of its charge, which results in long periods of non-consumption of energy from the sector, followed by brief moments of recharging. Since these current draws are usually weak, this does not result in any perceptible change in the shape of the mains voltage as a function of time.

However, this is no longer quite the case when considering the installation of a large number of devices in a building. So the total intensity at the level of the entire building to be supplied on time is no longer negligible. This results in a notable deformation of the form of the voltage at the maximum of the half-waves, that is to say the appearance of low frequency harmonics of odd predominance. With regard to the electricity network, this effect is unacceptable. In addition, recent decrees of electrical standards define the maximum value of the harmonics authorized with respect to the current at 25% of the effective value of this current for a harmonic of rank 3, ie 150 Hz.

To overcome this problem, several solutions are currently used. A first solution consists in adding an inductance L forming on the mains line with the capacitor C of filtering a low-pass filter with a threshold equal to 150 Hz. The usual calculation of the resonance frequency for an LC circuit quickly shows that for a high capacitor value C the value of the inductance L must also be. There then appears a significant drop in potential at the terminals of this inductance, ie a loss of energy released into heat by the Joule effect, which amounts to the same drawback as the primitive branch circuits.

A second alternative consists in using a 1/1 ratio transformer providing galvanic isolation. But there too, energy losses by Joule effect are noted.

A last solution consists in carrying out a first high-frequency cutting of the low-frequency alternations of the sector before rectification, which effectively makes it possible to obtain galvanic isolation preventing the return of harmonics on the supply network but complicates the general device and limits the efficiency. global.

The ideal solution consists in fact of getting rid of this parasitic filtering generator, or at least of reducing it to a very low value. This has the immediate consequence that the supply voltage of the electronic circuits remains alternating with each period equivalent to a pure positive sinusoidal function (corresponding to values of input angles ranging from 0 to 180 °). Thus, this supply voltage passes cyclically through an almost zero value. We then see that such a form of voltage becomes unsuitable for supplying electronic devices for known fluorescent tubes.

In fact, let us consider a first type of device called "self-oscillating" such, for example, that described in the presentation FR 2 599 208. This device essentially consists of a charging circuit comprising an inductor connected in series with a fluorescent tube and a capacitor connected between the terminals of this same tube; two transistors mounted in "push-pull" between the supply terminals and connected at their common point to the load circuit by means of a small transformer whose two secondary allow to take an alternating high frequency feedback signal driving these two transistors. This device further comprises a circuit generating a starting pulse. Then, the regular passage of the supply voltage to an almost zero value no longer makes it possible to ensure by the two secondary transformers a minimum of voltage on the gates of these transistors which stop working. The complete start-up cycle described must then be repeated at each alternation of the supply voltage, ie 100 times per second. Even if we let this start-up cycle repeat itself with real risks of failures, it would also result in the appearance of consequent tones of polluting frequency 100 Hz.

A second type of electronic device comprises an independent oscillator supplying, through an amplification stage, a resonant charging circuit which supplies one or more fluorescent tubes with electrical energy suitable for the starting and keeping in the on condition. In devices of this type, described for example in the presentations FR 2520575 and EP 0 065 794, there is, as before, an amplification stage comprising two power transistors connected in series between the supply terminals and a connected load circuit at their common point, which circuit includes a capacitor connected in parallel with a fluorescent tube and an inductor mounted with this same tube. More particularly in these circuits, the control element of these transistors, bases or gates, is controlled by an independent oscillating circuit alternately saturating one then the other transistor. As explained in these presentations, it appeared moreover necessary to provide a second independent circuit modifying the frequency of the oscillator during the starting of the device in order to carry out a first phase of preheating, a second phase of priming and a last phase of permanent operation. Here again, as disclosed, there is always provided a converter of the AC mains voltage into stabilized direct current for the overall supply of the device by the presence of semiconductor elements, such as operational amplifiers or flip-flops, present in these frequency converter or oscillator circuits, this conversion circuit having the disadvantages mentioned above.

Furthermore, the fact that the unidirectional pulsed supply voltage passes cyclically through an almost zero value also causes cyclic disarming of the fluorescent tubes in the absence of supply. To overcome this drawback responsible for faster wear of the tubes and higher overall consumption linked to the energy required for repetitive restrikes, it should be prevented that this supply voltage falls below a certain value corresponding to the threshold of defusing the tubes.

The object of the present invention is an electronic starting and supply device for fluorescent tubes with preheatable electrodes no longer requiring a capacitor for filtering the rectified mains voltage constituting the only polluting low frequency harmonic factor.

More particularly, in a device comprising:

- a rectifier circuit (6) of the AC alternating current delivering a unidirectional current drawn from a first and second line,

- one or more resonant load circuits connected in parallel, each circuit being in the form of a series connection of an inductor, the first electrode of the fluorescent tube, a capacitor and the second electrode of the fluorescent tube ,

an amplification stage comprising two transistors mounted in series between the first and second supply lines of the device and connected at their common point to the load circuit, an oscillating circuit connected to a circuit for driving the transistors so that each either alternately in the conduction state, and

- a circuit varying the frequency of the oscillating circuit when the device is started, 8

energy along the third line to the frequency converter circuit and the oscillating circuit; and switching means, the control element of which is controlled by a capacitive element, which means disconnects the return by the fourth line to the second line of the direct current from these circuits during a first phase of acquisition of energy by the means accumulation then, once the capacitive element is charged after a predetermined time, restores this return.

Usefully, the permanent supply circuit of the oscillator and frequency converter circuits includes:

- either a winding around the inductance of one of the load circuits or a secondary of a transformer whose primary is connected in parallel to the load circuits, winding one of whose ends is connected to the second line d power supply to the device and the other end is connected to the anode of a rectifying diode,

- a filter capacitor, a charge capacitor and a voltage regulation zener diode, these being connected in parallel on one side to the second line and on the other side on the one hand to the cathode of the diode rectification and on the other hand to the third supply line of the variator and oscillator circuits.

Preferably then, the circuit ensuring a minimum of voltage at the amplification stage comprises: - a transformer, the primary of which is connected in parallel to the load circuits and the secondary of which is on the one hand connected to the the present invention aims to solve the problem of starting from an unidirectional alternating supply voltage of 100 Hz unfiltered.

Obviously, the starting sequence must be able to comply with a first phase of preheating the electrodes and a second instantaneous and reliable tripping phase of the fluorescent tubes. The energy efficiency of the device should preferably be the best possible of the order of 90%. Finally, it would be desirable for such a device also to easily modulate the light intensity delivered by these tubes.

These goals are achieved by a device whose value of the filtering capacitor at the output of the rectifier circuit, if present, is less than or equal to 0.04 μF and further comprising for direct current supply only the oscillating circuits and frequency converter along a third and fourth line a temporary supply circuit active during the start-up phase and a supply circuit dependent on one of the load circuits, which circuit is active in the permanent operation phase , as well as for supplying the amplification stage, therefore fluorescent tubes, a circuit ensuring a minimum of voltage corresponding to the defusing voltage of the tubes.

Advantageously, the temporary supply circuit comprises means for accumulating electrical energy in continuous form charged by the rectifier circuit and applying this second supply line of the device and on the other hand to the anode of a first rectifying diode,

- a capacitor connected between the second supply line and the cathode of the first diode, - a second isolation diode the anode of which is connected to the common point of the first diode and of the capacitor and the cathode of which is connected to the first supply line.

According to a first preferred embodiment, the oscillator circuit then comprises two dividing bridges mounted in parallel between the third and fourth supply lines, a first essentially consisting of three identical resistors, a second consisting of two resistors and a capacitor, the voltage at the corresponding intermediate points being compared by two comparators respectively, the first acting on the input R, the second on the input S of a rocker whose output is firstly amplified by an amplifier before application to the primary d 'a transformer constituting the control circuit, on the other hand connected to an inverter which controls a switching element short-circuiting or not the capacitor. The frequency converter circuit can then comprise a switching element connecting or not connecting a capacitor in parallel to the capacitor of the divider bridge and the control element of which is connected through a diode to one of the terminals of another capacitor. which terminal is also connected to the third supply line through a resistor, the other terminal of the capacitor being connected to the fourth line. According to a second preferred embodiment, the oscillating circuit can comprise a first inverter whose output is connected to the input of a second inverter, the output of which is connected by a capacitor and a resistor connected in series to the input of the first inverter, these two inverters being supplied by the third and _. the fourth line; an adjustable resistor connected between the output of the first inverter and the common point of the capacitor and the resistor; as well as a parallel connection of at least two inverters amplifying the signal from the output of the second inverter before application to the control circuit. The frequency converter circuit can then comprise a first voltage divider bridge essentially consisting of a capacitor and two resistors connected in series between the third and the fourth line and a series connection of a resistor and a diode between d on the one hand the common point between the two resistors of the divider bridge and on the other hand the input of the second inverter. The oscillating circuit may also include a diode interposed between the first and the second inverter and the anode of which is connected to the adjustable resistor while the cathode is connected to the input of the second inverter.

It may be useful to include a filtering circuit of frequencies between 50 and 150 KHz upstream of the rectifier circuit comprising two windings respectively connected in series on the mains supply lines and in phase opposition one with respect to the other, a first capacitor mounted upstream of the windings between the sector terminals and a bridge of capacitors mounted downstream of the windings between the sector terminals and having their common point connected to earth.

According to preferential arrangements when the control circuit essentially consists of a transformer with two secondaries mounted in phase opposition, the transistors of the amplification stage are of the MOS type and they are protected against overvoltage between their gate and their source. by a zener diode bridge, in overvoltage between the drain and the source thanks to a diode, and in overcurrent by a circuit comprising a transistor short-circuiting through a diode the gate and the source according to a difference of potential appearing across a resistor connected in series with the source, which voltage is transmitted to the base of the transistor by diodes.

The invention will be better understood from the study of an embodiment taken by way of nonlimiting example and described by the following figures:

FIG. 1 is the block diagram of the device,

FIG. 2 is the detailed plan of the electronic components forming a first embodiment of the device,

FIG. 3 is a detailed plan of the oscillator 29 associated with the first embodiment of FIG. 2,

FIG. 4 is the detailed plan of the electronic components forming a second embodiment of the device, and - Figure 5 is the detailed plan of the electronic components forming a variant of the second embodiment of the device.

With reference to FIG. 1, the device according to the invention is connected to the low frequency alternating voltage 50 or 60 Hz of the network supplied by the National Office for Electric Power Production. This voltage first passes through an anti-interference circuit 100 before being applied to the input terminals of a rectifier circuit 6. This anti-interference circuit 100 prevents the rise of any residual interference of frequency around 100 kHz from the device to the sector. These parasitic voltages being low, that is to say less than 1%, this circuit 100 is only considered as an option for very specific applications, in particular premises intended to house computer equipment or radars. The rectifier circuit 6 is preferably a bridge of four diodes mounted according to the known method known as Graetz. Since the capacitor 16 connected to the output of this rectifier circuit 6 is very low or even non-existent, the voltage present on the output lines c and d is initially a positive unidirectional alternative, the form of the alternations of which corresponds to that present on the mains network but of frequency double, that is to say that these alternations have a duration of l / 100th (1 / I20th) of a second and a sinusoidal form whose input values go from 0 to 180 ° exactly. Line d is hereinafter considered as the return ground line which is however different from the grounding line E explained below. Between these two supply lines c and d are connected two transistors in series 34 and 43, that is to say that, in the case of MOS transistors, the drain of transistor 34 is connected to line c, the source of this transistor 34 is connected through a protection circuit 142 to the drain of transistor 43 and the source of transistor 43 is connected through a protection circuit 144 to line d. Two capacitors 54 and 67 are connected in series between lines c and d. Between the common point M of the transistors 34 and 43 and the common point Q of the capacitors 54 and 67 is connected a charging circuit making it possible to supply one or more tubes with electrical energy in a suitable form during their start-up then during their operation. In the case of two fluorescent tubes, this charging circuit is in the form of two parallel branches, each branch being in the form of a series connection of an inductor 51 or 56, of the first electrode of the fluorescent tube 52 or 57, a capacitor 55 or 59 and the second electrode of the fluorescent tube 53 or 58. In other words, on each branch, the inductor 51 or 56 is connected in series with the fluorescent tube, both of which electrodes are respectively connected to each other by a capacitor.

The transistors 34 and 43 are provided to be controlled by a circuit 135 so that cyclically one is in the conduction state and the other in the blocking state and then vice versa, this variation of situation taking place at very high frequency of the order of one hundred KHz. This makes it possible to show at the intermediate point M the voltage present between the lines c and d, but chopped at the switching frequency of transistors 34 and 43. The capacitors 54 and 67 then authorize a return of electrical energy passing through the charge circuit to line c or d depending on the conduction state of the transistor 34 or 43 while cutting all residual DC components at this level.

As mentioned previously, the voltage present on the output lines c and d is initially a positive unidirectional alternative and therefore passes cyclically through an almost zero value, which defuses the tubes in a detrimental manner. To overcome this problem, a minimum voltage maintenance circuit includes a transformer 100 whose primary is connected between points M and Q. The secondary is firstly connected to the second line d and secondly to the cathode d 'a fast rectifying diode 105. The anode of this diode is connected to the common point of a series connection of a capacitor 110 and a diode 115 located between the first and the second line c and d, the cathode of the second diode 115 being connected to the first line c.

During operation, a high voltage is induced in the secondary, the ratio of the transformer 100 being substantially equal to unity. This secondary voltage is rectified by the diode 36 then filtered by the capacitor 11 before being reinjected through the diode 41 in the first supply line c. This diode 115 has two functions. On the one hand, it prevents the voltage from the AC network from coming to charge the capacitor 11 and consequently producing harmonics low frequency whose purpose of the invention is precisely to eliminate. On the other hand, it does not allow the capacitor 110 to discharge until the voltage on the first line c becomes insufficient. Thus, this supply voltage of the amplification stage cannot fall below the value of the charge voltage of the capacitor 110, thus preventing the fluorescent tubes from deionizing.

The output of the rectifier circuit 6 also supplies via lines a and b a temporary supply circuit 110, more precisely during start-up, of stabilized DC voltage on lines e and f. A second circuit 150 is provided which, by taking electrical energy by means of either a winding 66 around the inductance 56, or a secondary of the transformer 100, makes it possible to supply the line e permanently via line k in DC voltage stabilized once the device has started.

Lipes e and f supply two circuits, namely an oscillator circuit 130 and a circuit 120 which by action on line g modifies the oscillation frequency of circuit 130. The oscillating signal generated by circuit 130 is transmitted to a control circuit 135 which controls the alternating conduction of transistors 34 and 43 through individual protection circuits 142,144.

This control circuit can be constituted, as in the embodiments illustrated in FIGS. 2 and 4, of a transformer 135 comprising a primary and two secondary in phase opposition. We then obtain an alternative conduction of transistors as soon as their gate / source potential exceeds approximately 4 V.

As an alternative and as illustrated in FIG. 5, this control circuit 135 may comprise an integrated circuit operating according to a technique known as "Botstrap" and receiving on two inputs 10, 12 the signals phase shifted at 180 ° from commands coming from the oscillator . At each conduction of the transistor 43, the capacitor 32 finds itself charged through the resistor 38 and the diode 39. This diode 39 prevents any discharge of the capacitor 32 by the resistor 38 during the following cycles. Once the transistor 43 is blocked, a voltage close to that present on the third line is recorded across the capacitor 32. The "Botstrap" effect then makes it possible to switch this voltage on the control circuit 142 of the transistors 34 in floating potential. The resistance torque 30/37 and diode 31/40, constituting in this embodiment the individual circuits 142/144, allows a progressive charge of the trigger of the power transistors, while their discharge is instantaneous.

The frequencies during start-up and in the permanent operating phase of the signal generated by the oscillator 130 and its variator 120 are determined as a function of the values of the inductors 51.56 and of the capacitors 56.59 so that in the start-up phase the load circuits do not enter into resonance and allow the passage of a preheating current through the electrodes then, in the operating phase, the load circuits enter into resonance therefore apply across the terminals of 21

capacitor 24 for breaking continuous components. The second end of the winding 30 is connected to the line f. Furthermore, the output of flip-flop 330 is also applied to an inverter 350 controlling, through a resistor 360, the base of a transistor 370. This transistor 370 may or may not short-circuit the capacitor 20.

When the circuit 130 is energized, the potential at points y1 and y2 is equal to two thirds and one third of the voltage respectively. The capacitor 20 being discharged, the voltage x2 is almost zero and the voltage xl is intermediate between y2 and yl. Then, taking into account the connections of the comparators, the input R of the flip-flop 330 is in the low state while the input S is in the high state. The output of the flip-flop 330 is then in the high state and the output of the inverter 350 controlling the transistor 370 is in the low state, which transistor is then in the non-conduction state. The capacitor 20 being charged little by little, the voltage x2 becomes greater than y2 then the voltage xl also becomes greater than yl. Then the input R switches to the high state while the input S has already gone to the low state. This results in the change in the low state of the output of the flip-flop 330 therefore in the high state of the output of the inverter 350, ie the saturation of the transistor 370 which now short-circuits the capacitor 20. This capacitor 20 discharging, the voltages in xl and x2 drop to values lower than the voltages present in yl and y2 causing a new change of state of the flip-flop 330 therefore the opening again of the transistor 370. It is thus understood that this circuit automatically starts to oscillate with a frequency 22

essentially dependent on the charge and discharge speed of the capacitor 20 therefore on its eigenvalue.

The input C for controlling the electronic component 29 provides access to the point yl of the divider bridge, thus making it possible to impose a different voltage of threshold vx at point yl and vx / 2 at point y2, which in another way modifies the oscillator frequency. We can take advantage of this terminal C to apply a desired voltage and change at will the frequency of this oscillator, therefore the resulting light intensity of the fluorescent tube.

The frequency converter circuit 120 associated with the first embodiment of circuit 130 is now described in relation to FIG. 2. This circuit essentially comprises a capacitor 27 connected on the one hand between the point common to the resistor 25 and to the capacitor 20 and on the other hand to the collector of a transistor 14 whose emitter is connected to line f. The control circuit of the base of this transistor 14 comprises a first series connection of a resistor 23 and a zener diode 26 between the line e and the base of this transistor 14 as well as a parallel connection of a charge capacitor 22 and a discharge resistor 21 between the line f and the common point between the resistor 23 and the zener diode 26.

During the application of the filtered DC voltage on line e, the capacitor 22 being still discharged, the voltage at the input of the zener diode 26 is low and the transistor 14 remains 17

capacitors and electrodes a Q factor overvoltage of the order of 4 which allows the fluorescent tubes to be triggered and kept on.

The details of the circuits mentioned above will now be described in relation to FIGS. 2, 3, 4 and 5.

In the same way in FIGS. 24 and 5, the block 110 first comprises a transistor 15 controlling the return of the line f to the line d via the line b. The drain of this transistor 15 is connected to line f and the source is connected to line b back to line d. A resistor 19 of high ohmic value connected between the lines f and d fixes a drain / source potential necessary for the switching of this MOS transistor 15. The gate of this transistor 15 is controlled by a first circuit generating a time constant Tl and composed two resistors 7.8 connected in series between lines a and d, a resistor 12 and a zener diode 13 connected in series between the common point of resistors 7 and 8 and the gate of transistor 15 and a capacitor 17 and a discharge resistor 18 connected between the common point of the resistor 12 and the zener diode 13 and the line d. As can easily be understood, the voltage portion present between resistors 7 and 8 allows the capacitor 17 to be charged slowly. Once the voltage present at the terminals of this capacitor exceeds the opening voltage of the zener diode 13 plus 2 V, ie at the end of a delay T1, this voltage is applied to the gate of transistor 15 which, saturating, only then connects line f to line b then d. 18

Furthermore, the temporary supply circuit 110 comprises a second circuit composed of a resistor 9 connected between the rectifier circuit 6 and the line e as well as a capacitor 11 and a discharge resistance 10 connected between the line e and the line d (in FIG. 5: capacitor 61 after the direct crossing 3-3 of the integrated circuit). During the period T1 during which the transistor 15 disconnects the return from line f to line d, the variator 120 and oscillator 130 circuits cannot make any current draw on line e. The capacitor 11 can then be charged supplied by the rectifier circuit 6 through the resistor 9. This period T1 is determined by the value of the resistors 7,8 and 12 and of the capacitor 17 so that it is long enough to allow a sufficient charge of the capacitor 11. In fact, once this period T1 has ended and the transistor 15 switched thus connecting the line f to line b, this capacitor 11 must supply the line e therefore the circuits 120 and 130 during a sufficiently long period T2 to allow preheating and tripping of fluorescent tubes.

Identically in FIGS. 2 and 4, the permanent supply circuit 150 comprises a winding 66 around the inductor 56 one of the branches of which is connected to line d and the other of which is connected to the cathode of a rectifier diode 64 via a resistor 65. A filtering capacitor 63 is connected between the anode of this diode 64 and the line d. A charge capacitor 61 and a zener diode 60 are connected on one side to the line d and on the other side on the one hand to the anode of the diode 64 through a 19

resistance 62 and on the other hand to the line k joining the line e mentioned above. The permanent supply circuit in FIG. 5 is substantially identical except for the fact that the starting point is the secondary 66 of a small transformer 100 whose primary is connected between the points M and Q, therefore in parallel with the circuits of charge. In addition, the charge capacitor 61 is coincident with that 11 of the temporary supply circuit. Thus, the alternating voltage sampled by the winding or the secondary 66 is rectified by the diode 64 then prefiltered by the capacitor 63. The current issuing, passing through the resistor 62 of low ohmic value, will charge the capacitor of chemical type 61 until to the value corresponding to the zener voltage of the diode 60 which has the function of regulating this voltage. This voltage thus filtered and regulated can then be applied via the lines k and e to supply the circuits 120 variator and 130 oscillator. The resistors 62 and 65 are intended to limit the current passing through the diodes 64 and 60. It is easily understood that as long as the device has not started, this supply circuit 150 remains inoperative. But taking into account the presence of the temporary supply circuit 110 ensuring the supply of direct current during the entire period T2 corresponding to the preheating and then to the triggering of the fluorescent tubes, it is understood that at the end of this period T2 the tubes being tripped and the voltage across the inductor 56 having dropped to a lower value, then the circuit 150 can immediately take over and keep the variator 120 and oscillator 130 circuits in operation. 20

Given the limited capacity in time, ie during the period T2, of direct current supply to the capacitor 11, the variator 120 and oscillator 130 circuits should be as economical as possible.

A first advantageous embodiment of the circuit

130 will now be described in relation to FIG. 3. This circuit 130 comprises an integrated electronic component 29 and a first voltage divider bridge consisting of the resistors 28, 25 and of the capacitor 20 connected in series between the lines e and f. This integrated component 29 also includes a voltage divider bridge produced by three identical resistors 300 connected between the terminals v and m themselves connected to the lines e and f respectively. It further comprises two comparators 310 and 320. The positive input of comparator 310 is connected via the terminal Th at the intermediate point xl between the resistors 28 and 25; and the negative input is connected to the intermediate point yl between the first and the second resistor 300. The negative input of the second comparator 320 is connected via the terminal T to the intermediate point x2 between the resistor 25 and the capacitor 20. The input positive of this second comparator 320 is connected to the point y2 intermediate between the second and the third resistance 300. The output of the first comparator 310 acts on the input R and the second on the input S of a rocker 330. The output of this flip-flop 330 is amplified by an operational amplifier 340 to provide a current of about 200 mmA for both high and low states. This output signal is subsequently applied to the primary winding 30 via a in the state of non-conduction. The oscillator circuit 130 then operates exactly as explained above at a high frequency but different from the resonance frequency of the load circuit. The electrodes of the fluorescent tubes can then be preheated. When the positive terminal of the capacitor 22 charged by the resistor 23 has reached the threshold voltage of the zener diode 26 plus the base-emitter voltage of the transistor 14, then the latter will enter into conduction and will allow the connection via its collector / emitter junction in parallel with the capacitor 27 with the capacitor 20. Therefore, the frequency delivered by the circuit 130 will drop suddenly so as to then obtain the frequency putting the charging circuits in resonance which immediately ignite the fluorescent tubes. Conversely, when the device is cut off from the network, the resistor 21 allows the capacitor 22 to be discharged more quickly than by its internal leaks, which systemically starts the system with a preheating phase.

A second advantageous embodiment of the circuit 130 will now be described in relation to FIG. 4. This oscillator circuit 130 comprises a first inverter 80 whose output is connected through a diode 82 to the input of a second inverter 83. The output of the inverter 83 is itself connected to the input of the first inverter 80 by a capacitor 84 and a resistor 86 connected in series. A variable resistor 85 is connected between the output of the inverter 80 and the common point of the capacitor 84 and the resistor 86. The signal present at the output of the inverter 83 is amplified by a parallel connection of four inverters 86 before being applied to the capacitor 24. The two inverters 80 and 83 as well as the four inverters 86 are of the MOS type grouped in a single housing supplied between lines e and f .

Assuming the diode 82 short-circuited and the capacitor 84 initially discharged, the operation of the oscillator is as follows. When the six inverters are simultaneously energized, the potential at the output of the inverter 83 corresponds to a low logic level due to the discharge of the capacitor 84. This implies a high level at its input and therefore at the output of the 'inverter 80 is a low level on the input of this inverter 80. The adjustable resistor 85 being on the one hand connected to the output of the inverter 80 at the high level and on the other hand to the capacitor 84 discharged therefore at the low level , a current corresponding to the charge of this capacitor 84 will flow through this adjustable resistor 85. The voltage at the common point of the capacitor 84 and the adjustable resistor 85 which is applied to the input of the inverter 80 thanks to the resistor 86 will therefore increase as the capacitor 84 charges. When this common point voltage has reached 50% of the supply voltage present on the line e, the inverter 80 will switch imposing then an n high level at the input of the inverter 83 which itself will switch implying a low level at its output. The potential at the terminals of the resistor 85 and of the capacitor 84 is therefore reversed, causing the passage of a current discharging this capacitor 84. Once the latter is discharged, the voltage at the point common between the adjustable resistor 85 and the capacitor 84 is again low, this voltage being found at the input of the inverter 80 by the resistor 86. This inverter 80 switches over and the initial conditions of the circuit are then found. The cycle described above begins again. This circuit 130 thus oscillates on a frequency determined by the value of the adjustable resistor 85 and of the capacitor 84. This oscillation present at the output of the inverter 83 is amplified by the parallel mounting of the other four inverters 86 before being applied to the primaries 30 of the transformer through the capacitor 24.

If a minimum voltage, for example 10% of the voltage present on line e, is imposed by a line g between the output of the inverter 80 and the input of the inverter 83, the capacitor 84 can no longer be discharge at a value lower than this minimum imposed voltage. In other words, this capacitor 84 will reach a voltage corresponding to 50% of the supply voltage more quickly, causing the inverter 80 to switch earlier than previously. There is therefore a greater rapidity of the oscillatory phenomenon described above, ie an increase in its frequency. The diode 82 precisely allows the application of such a control voltage on the line g by preventing this voltage from disturbing the output of the inverter 80.

The frequency converter circuit 120 associated with this second embodiment of the circuit 130 is now described always in relation to FIG. 4. This circuit essentially comprises a voltage divider bridge consisting of the capacitor 73 and resistors 71 and 72 connected in series between lines e and f. A resistor of high ohmic value 70 connected in parallel to the capacitor 73 makes it possible to discharge the latter when the device is stopped. This frequency converter circuit 120 also includes a series connection of a resistor 75 and a diode 76 connecting on the one hand the common point between the resistors 71 and 72 and on the other hand, by the line g, the point common between the cathode of the diode 82 and the input of the inverter 83. This series connection makes it possible to apply to the input of the inverter 83 the voltage present at the common point of the resistors 71, 72.

During the application of the filtered DC voltage on line e, the capacitor 73 having been previously discharged by the resistor 70, a strong charging current appears in the resistors 71 and 72. A potential difference therefore appears at the terminals of the resistor 72 which is applied via resistor 75 and diode 76 to the input of the inverter 83. Diode 76 allows this voltage to be injected into the oscillating circuit 130 without a backtracking of the oscillatory signal being made possible. The resistor 75 limits the intensity supplied to the input of the inverter 83 because only a voltage parameter matters. The oscillator circuit 130 then began to operate as explained above at a high frequency which is higher than the resonance frequency of the charging circuit allowing the electrodes of the fluorescent tubes to be preheated. As the capacitor 73 charges, the current passing through the resistor 72 decreases this which also lowers the oscillation frequency of circuit 130. When this capacitor 73 is fully charged, no more voltage appears on line g and circuit 130 oscillates at a frequency now imposed by the values of capacitor 84 and the adjustable resistor 85, which frequency corresponds to that of resonances of the charging circuit which immediately initiates the fluorescent tubes.

This possibility of controlling the frequency of oscillations of the circuit 130 by applying a voltage on the line g can be taken advantage of during the operation of the fluorescent tubes to modulate their luminous flux by slight variations in this frequency of oscillation around the resonance value. This is achieved by an auxiliary circuit receiving a DC voltage at point 22 on line h. This additional circuit includes a series connection of a resistor 91 and a diode 95 connecting the line h to the resistor 75. A capacitor 92, a resistor 93 and a zener diode 94 are connected in parallel between on the one hand the line f and on the other hand the common point between the resistor 91 and the anode of the diode 95. The resistor 91 limits the intensity of the current. The capacitor 92 provides anti-parasite filtering. The zener diode 94 prevents the reported voltage from reaching a value greater than half of the supply voltage on line e so as not to block the oscillator circuit 130. Finally, the diode 95 allows this voltage to be injected. control without disturbing the preheating phase. In the embodiment illustrated in Figure 5, we recognize the integrated control circuit of the transistors 135 with the resistor 38, the diode 39 and the associated capacitor 32, the circuits 30/37 and 31/40 as well as the holding circuit minimum voltage 100, 105, 110 and 115 described above. The oscillator and frequency converter circuits are identical to those described with reference to FIG. 4. However, the tubes must be primed under the best possible conditions, an additional device is provided for locking the annex modulation circuit of the luminous flux during the preheating of the tubes. This device comprises a transitor 130, the collector and the transmitter of which are respectively connected to the terminals of the capacitor 92, therefore between the anode of the diode 95 and the fourth line f, and the base of which is connected by a series connection of a zener diode 134 and a resistor 133 at the common point between the capacitor 73 and the resistor 74.

When charging the capacitor 73, the zener diode 134, of value substantially equal to a quarter of the voltage present on the first line, allows the circulation of a current in the base of the transistor 130 via the resistor 133 which has the effect the cancellation by earthing any voltage present on the control input, for the entire duration of preheating. When the latter is finished, the voltage present at the common point of the capacitor 73 and of the zener diode 134 becomes lower than the value of the diode 134 which is blocked as well as the transistor 130 by absence of basic current. The voltage present at the input of the annex modulation circuit becomes active again.

The transistors 34 and 43 present in FIGS. 2 and 4 being preferably of the MOS type, it is desirable to provide identical protection circuits 142, 144 for each of the transistors. More particularly in the circuit 142, any overvoltage between the drain and the source of the transistor 34 is protected by the diode 50 and any overvoltage between the gate and the source is protected by a bridge of zener diodes connected in series but opposite between the gate and source. The protection against an overcurrent which may appear between the drain and the source comprises a resistance 39 of low ohmic value connected between the source of the transistor 34 and the intermediate point M. The potential drop caused by the passage of current through this resistance is applied by a diode 38 between the base and the emitter of a transistor 37 whose emitter-collector junction short-circuits through a diode 35 the voltage between the gate and the source of the transistor 34. Thus, if the current passing through the resistance 39 exceeds 4 A for example, the potential difference appearing between the base and the emitter of the transistor 37 brings it to the saturation state thus immediately lowering the gate voltage therefore putting the transistor 34 in the non-conduction state. The protection circuit 144 for the transistor 43 is strictly identical to the circuit 142 previously described.

As can be seen, the device according to the invention succeeded in supplying the load circuit with voltage rectified and hatched sector at a high frequency of the order of a hundred KHz, and this with a very low or nonexistent filter capacitor 16. Thus, by eliminating this filtering capacitor, it was possible to eliminate the very cause of the parasites of the order of one hundred or so truly polluting Hz by inducing distortion on the mains voltage. In addition to meeting the new electrotechnical standards in force, this device allows its installation in large numbers in buildings without calling into question the electrotechnical wiring of the latter. In very specific cases requiring an almost perfect network voltage, it has however been found that this device still induces extremely low interference of the order of a hundred KHz. For this, an optional interference suppression circuit 100 has been provided which comprises two windings 2 and 3 mounted in phase opposition respectively on lines a and b, a capacitor 1 mounted upstream of the windings and a capacitor bridge 4 and 5 mounted downstream windings, that is to say close to the rectifying circuit 6. The common point of the capacitors 4 and 5 is then connected to the physical earth. Thus, the high frequency parasites appearing in the windings 2 and 3 but in the opposite direction cancel themselves out while the low frequency voltage of the network crosses this circuit 100 towards the rectifying circuit 6 without any difficulty.

By way of non-limiting example, the preferred numerical values for the important components of the device are presented below. R9 = 50 KΩ R23 = 100 KΩ

Eyelash = 470μF C22 = 10μF

R12 = 100 KΩ R25 = 1 KΩ

C17 = 10 μF C27 = 47 pF C20 = 10 pF

C110 = 22 _μ F 129 = IC 555

C73 = 10 μF T34.43 = IRF 830

R74 = 10 KΩ R32.40 = 22Ω

R72 = 47 KΩ L56.51 = 1 mH

R75 = 100 KΩ L66 = 3 turns on L56

R86 = 10 KΩ C61 = 10 μF

R85 = 20 KΩ aj Z60 = 16 V / l 3

C84 = 1 nF C54 = 100 nF / 500 V C67 = 100 nF / 500 V.

It should be noted that if the device described above has a certain complexity, the components mentioned above the constituent are generally of small size which makes it possible to assemble this device easily on a small plate of the order of 80 mm. long by 10 mm wide and 15 mm high. The material production of such devices can easily be carried out by fully automatic machines ensuring a high level of reliability. In addition, by the particular choice of transistors 34,43 associated with resistors 32,40 this device has increased reliability due to a lower overall operating temperature.

Many improvements can be made to this device in the context of this invention.

Claims

1. Electronic starting and supply device for fluorescent tubes with preheatable electrodes comprising: - a rectifier circuit (6) of the AC mains supply delivering rectified current on a first and second line (c, d),

- one or more resonant load circuits connected in parallel, each circuit being in the form of a series connection of an inductor (51,56), of the first electrode (52,57) of the fluorescent tube, a capacitor (55,59) and the second electrode (53,58) of the fluorescent tube,

an amplification stage comprising two transistors (34,43) connected in series between the first and second supply lines (c, d) of the device and connected at their common point (m) to the charging circuit,

an oscillating circuit (130) connected to a control circuit (135) of the transistors (34,43) so that each is alternately in the conduction state, and

- a circuit (120) varying the frequency of the oscillating circuit (130) when the device is started, characterized in that, if present, the value of the filtering capacitor (16) at the output of the rectifier circuit (6) is less than or equal at 0.04 μF and in that the device further comprises for the direct current supply of the oscillating circuit (130) and the frequency converter circuit (120) along a third and fourth line (e) and (f) a temporary supply circuit (110) active during the start-up phase and a supply circuit (150) dependent on one of the load circuits, which circuit (150) is active in permanent operating phase. as well as, for the supply of the amplification stage therefore of the fluorescent tubes, a circuit ensuring a minimum of voltage corresponding to the defusing voltage of the tubes.

2. Device according to claim 1, characterized in that the temporary supply circuit (110) comprises means for accumulating electrical energy in continuous form (11) charged by the rectifier circuit (6) and applying this energy along the third line (e), and switching means (15), the control element of which is controlled by a capacitive element (17), which means disconnect the return by the fourth line (f) to the second line (d) of the direct current of the frequency converter (120) and oscillator (130) circuits during a first phase of energy accumulation by the element (11) then, once the capacitive element (17) is charged after a predetermined time, restores this return.

3. Device according to claim 1, characterized in that the permanent supply circuit of the oscillator and frequency converter circuits (150) comprises

- either a winding (66) around the inductance (56) of one of the charging circuits or the secondary of a transformer whose primary is connected in parallel to the charging circuit, winding one of whose ends is connected to the second line supply (d) and the other end is connected to the anode of a rectifying diode (64),

- a filter capacitor (63), a charge capacitor (61) and a voltage regulation zener diode, these being connected in parallel on one side to the second line (d) and on the other side d on the one hand to the cathode of the rectifying diode (64) and on the other hand to the line (k) then the third line (e) of supply of the variator (120) and oscillator (130) circuits.

4. Device according to claim 1, characterized in that the circuit ensuring a minimum of voltage in the amplification stage comprises:

- a transformer (100) the primary of which is connected in parallel to the charging circuits and the secondary (66) of which is on the one hand connected to the second supply line of the device and on the other hand to the anode of a first rectification diode (105), a capacitor (110) connected between the second supply line and the cathode of the first diode (105), and

- A second isolation diode (115), the anode of which is connected to the common point of the first diode (105) and of the capacitor (110) and the cathode of which is connected to the first supply line (c).

5. Device according to claim 1, characterized in that the oscillator circuit (130) comprises two dividing bridges mounted in parallel between the third and fourth supply lines (e) and (f), a first essentially consisting of three identical resistors ( 300), a second consisting of two resistors (28,25) and a capacitor (20), the voltage at the points corresponding intermediates (xl / yl, x2 / y2) being compared by two comparators (310,320) respectively, the first (310) acting on the input (R), the second on the input (S) of a rocker (330 ) the output of which is on the one hand amplified by an amplifier (340) before application to the primary of the transformer 135 constituting the control circuit, on the other hand connected to an inverter (350) which controls a short switching element (370) -circuiting or not the capacitor (20).

6. Device according to claim 4, characterized in that the frequency converter circuit (120) comprises a switching element (14) connecting or not connecting a capacitor (27) in parallel to the capacitor (20) and whose control element is connected through a diode (26) to one of the terminals of a capacitor (22) which terminal is also connected to the third supply line (e) through a resistor (23), l 'other terminal of the capacitor (22) being connected to the fourth line (f).

7. Device according to claim 1, characterized in that the oscillating circuit (130) comprises a first inverter (80) whose output is connected to the input of a second inverter (83), the output of which is connected by a capacitor (84) and a resistor (86) connected in series to the input of the first inverter (80), these two inverters being supplied by the third (e) and the fourth line (j); an adjustable resistor (85) connected between the output of the first inverter (80) and the common point of the capacitor (84) and the resistor; as well as a parallel connection of at least two inverters amplifying the signal from the output of the second inverter (83) before application to the control circuit (135).

8. Device according to claim 6, characterized in that the oscillating circuit comprises a diode (82) interposed between the first (80) and the second inverter (83) and the anode of which is connected to the adjustable resistor (85) then that the cathode is connected to the input of the second inverter (83) and that the frequency converter circuit (120) comprises a first voltage divider bridge essentially consisting of a capacitor (73) and two resistors (71, 72) connected in series between the third (e) and the fourth line (f) and a series connection of a resistor (75) and a diode (76) between on the one hand the common point between the two resistors of the divider bridge and on the other hand the input of the second inverter (83).

9. Device according to claim 1, characterized in that a circuit for filtering frequencies between 50 and 150 kHz (100) located upstream of the rectifier circuit (6) comprises two windings (2,3) respectively mounted in series on the mains supply lines (A, B) and in phase opposition to each other, a first capacitor (1) mounted upstream of the windings between the terminals (A, B) and a bridge of capacitors (5,4) mounted downstream of the windings between the terminals (A, B) and having their common point connected to earth.

10. Device according to claim l, characterized in that the control circuit consists essentially of a transformer with two secondaries mounted in phase opposition, in that the transistors (34,43) are of the MOS type and in that they are protected against overvoltage between their gate and their source by a zener diode bridge (33/36 , 41/45), in overvoltage between the drain and the source thanks to a diode (50,48), and in overcurrent by a circuit comprising a transistor (37,46) short-circuiting through a diode (35, 42) the grid and the source as a function of a potential difference appearing across a resistor (39,49) connected in series with the source, which voltage is transmitted to the base of the transistors (37,46) by diodes (38,47).