WO1982003744A1

WO1982003744A1 - Device for the supply of at least one fluorescent tube

Info

Publication number: WO1982003744A1
Application number: PCT/CH1982/000052
Authority: WO
Inventors: Elektronik Apparatebau Ag Koch
Original assignee: Jucker Peter
Priority date: 1981-04-15
Filing date: 1982-04-01
Publication date: 1982-10-28
Also published as: NO824199L; EP0063284A1; JPS58500633A

Abstract

The fluorescent tube (20), supplied by this device, comprises electrodes (21, 22) having each two connection ends (21a, 21b; 22a, 22b). A direct current source (31) is connected to one of the ends (21a, 22a) of each electrode (21, 22) through an induction coil (23) and an electronic switching circuit (32, 33) to automatically adjust the current (IR?) to a given rated value. Switches (24, 25) are used for periodically inverting the current circulation direction (IL?) through the fluorescent tube (20). The other ends (21b, 22b) of the electrodes (21, 22) are connected to an electronic switch (26) which may be controlled by a control signal (U26) generated by a control device (27). This control signal (U26) controls the switch (26) with a pulse frequency of at least 50 Hz, for opening and closing it periodically. When the switch (26) is closed, the electrodes are heated and the fluorescent tube (20) is shunted, whereas during the opening, the induction coil (23) induces a voltage sufficient to light the tube. The control device (27) comprises means to modify the ratio between the closing duration and the opening duration of the switch (26), thereby allowing the luminosity of the fluorescent tube (20) to be controlled.

Description

Device for feeding at least one fluorescent lamp

The present invention relates to a device for feeding at least one fluorescent lamp with glow electrodes, each having two connection ends, one of which can be connected to a power source and the other can be connected to one another by means of a switch arrangement, and to devices having an induction coil for limiting the strength of the current flowing from the power source to the lamp.

In known devices of this type, the induction coil has a relatively high impedance in order to be able to serve as a ballast for limiting the current strength. The switch arrangement for connecting the two glow electrodes only serves to start the fluorescent lamp, in that the two glow electrodes are first connected to one another by means of the switch arrangement in order to heat up the glow electrodes, and afterwards the connection is interrupted in order to close the fluorescent lamp by the induction voltage of the induction coil ignite. Recently, fluorescent lamps have become commercially available, which are filled with krypton gas, among other things, and which enable an increased luminous efficacy compared to previously customary fluorescent lamps and thus a corresponding saving in operating energy. However, such new fluorescent lamps cannot be ignited with the conventional devices for feeding fluorescent lamps which have been customary up to now.

It is therefore the object of the present invention to design a device of the type mentioned at the outset in such a way that it is particularly suitable for supplying and igniting the new fluorescent lamps mentioned. The device found to solve this problem is characterized by the features defined in claim 1.

Appropriate and advantageous developments of the device according to the invention are defined in the dependent claims. It should be particularly emphasized that the device according to the invention can be designed in a relatively simple manner in such a way that brightness control of the fluorescent lamp is possible, specifically without phase gating control. For this purpose, the control device for the electronic control of the switch arrangement connecting the two glow electrodes of the fluorescent lamp is provided with devices for changing the ratio of the time intervals with the conductive and non-conductive state of said switch arrangement.

Details of the device according to the invention and various advantageous embodiments of the same result from the following description and the associated drawings. Show it:

1 shows a block diagram of a first exemplary embodiment of the device according to the invention;

2 shows two diagrams belonging together, which illustrate the time course of a control signal which is dependent on the current strength and the resulting strength of the current;

3 shows two diagrams belonging together, which represent the time course of a control voltage for controlling the switch arrangement connecting the two glow electrodes of the fluorescent lamp and the resulting current through the fluorescent lamp; FIG. 4 shows a more detailed circuit diagram of the exemplary embodiment according to FIG. 1;

FIG. 5 shows several diagrams belonging together, which represent the time course and the relationship between control voltages in the circuit diagram according to FIG. 4 and the current flowing through the fluorescent lamp; FIG.

6 shows the circuit diagram of a second exemplary embodiment of the device according to the invention;

7 shows several diagrams belonging together, which illustrate the time course and the relationship between control voltages in the circuit diagram according to FIG. 6 and the current flowing through the fluorescent lamp;

8 shows the circuit diagram of a third exemplary embodiment of the device according to the invention;

9 shows three diagrams belonging together, which show the time course and the relationship between control voltages in the circuit diagram according to FIG. 8 and the strength of the current flowing through the fluorescent lamp;

10 shows several diagrams belonging together, which illustrate the temporal course and the relationship of voltages in the circuit diagram according to FIG. 8 and the current flowing through the fluorescent lamp;

11 shows the circuit diagram of a fourth exemplary embodiment of the device according to the invention;

FIG. 12 shows several diagrams belonging together, which show the time course and the relationship of voltages in the circuit diagram according to FIG. 11 and the current flowing through the fluorescent lamp. Reference is first made to FIG. 1, in the middle part of which a fluorescent lamp 20 with two glow electrodes 21 and 22 is shown. Each of the two glow electrodes 21 and 22 has two connection ends 21a and 21b or 22a and 22b. One connection end 21a of the glow electrode 21 is connected via an induction coil 23 to a first switchover device 24, while the corresponding connection end 22a of the second glow electrode 22 is connected directly to a second switchover device 25. The other connection ends 21b and 22b of the two glow electrodes 21 and 22 are connected to a switch arrangement 26, by means of which the connection ends 21b and 22b can be connected to one another. The switch arrangement 26 can be controlled electronically by a control voltage U26, which is generated and supplied by an associated control device 27. The configuration of the switch arrangement 26 and the control device 27 is explained in more detail below.

The two switching devices 24 and 25 can also be controlled electronically by means of control voltages U24 and U25, which are generated by an associated control device 28. The design of the switching devices 24, 25 and the associated control device 28 is also explained in more detail below. For the time being, it is sufficient to point out that the two switching devices 24 and 25 can be switched synchronously alternately into opposite switching positions, so that the switching device 24 connects the induction coil 23 alternately to the positive pole 29 and the negative pole 30 of a direct current source 31, respectively, during each at the same time the other switching device 25 connects the connection end 22a of the glow electrode 22 alternately to the negative pole 30 or the positive pole 29 of the direct current source 31. Devices 32 and 33 for automatically regulating the strength of the current I _R supplied to the fluorescent lamp 20 are connected between the negative pole 30 of the direct current source 31 and the two switching devices 24 and 25. These devices for regulating the current intensity have a second electronically controllable switch arrangement 32 and a current sensor device 33, the latter providing a control signal U33 for controlling the switch arrangement 32. The configuration of the switch arrangement 32 and the current sensor device 33 is explained in detail below. Finally, a free-wheeling diode 34 is connected on the one hand to the positive pole 29 of the direct current source 31 and on the other hand to a connecting conductor 35 between the switch arrangement 32 and the current sensor device 33.

1 is briefly as follows:

It is initially assumed that the two switching devices 24 and 25 assume the switch positions illustrated in FIG. 1 and remain therein. It is further assumed that the switch arrangement 26 is in the conductive state and thus connects the connection ends 21b and 22b of the two glow electrodes 21 and 22 to one another. A current then flows from the positive pole 29 of the direct current source 31 via the switching device 25, through the glow electrode 22, via the switch arrangement 26, through the glow electrode 21, through the induction coil 23, via the switching device 24, through the current sensor device 33 and through the switch arrangement 32 back to the negative pole 30 of the direct current source 31. Here, the glow electrodes 21 and 22 are heated in order to enable thermal electron emission. If the switch arrangement 26 is then controlled in the non-conductive state, calls the electrical energy stored in the induction coil 23 produces a relatively high induction voltage which causes the fluorescent lamp 20 to ignite. Instead of the switch arrangement 26, the current I _L now flows through the fluorescent lamp 20.

The switch arrangement 32 and the current sensor device 33 ensure that practically the same current intensity I _R flows both when the switch arrangement 26 is in a conductive state and when it is not in a conductive state. In order to keep the power loss in the current control as small as possible, the principle of the two-point control is used, the switch arrangement 32 being controlled in the conductive state as soon as the current I _R falls below the setpoint by a predetermined amount or controlled in the conductive state as soon as the current strength I _R exceeds the setpoint by a predetermined amount. The control signal U33 generated by the current sensor device 33 for controlling the switch arrangement 32 has the form of pulses with pulse pauses in between, as shown in the upper diagram in FIG. 2. Each time the current I _R falls below the critical value, the control signal U33 goes high and the switch arrangement 32 is conductive. The ratio of the pulse duration to the pulse pause of the control signal U33 depends on the respective current current I _R. Mainly because of the inductance of the induction coil 23, the current rise when switching the switch arrangement 32 into the conductive state does not occur suddenly, but continuously as a function of time. Each time the switch arrangement 32 is controlled in the non-conductive state, the current flow to the negative pole 30 of the current source 31 is interrupted, but due to the electrical energy stored in the induction coil 23 via the Free-wheeling diode 34, through the fluorescent lamp 20 and through the current sensor device 33, the current continues to flow with decreasing intensity until the switch arrangement 32 is controlled again in the conductive state. The lower diagram in FIG. 2 shows the time profile of the current I _R flowing through the current sensor device 33 in relation to the control signal U33 shown in the upper diagram.

It can be seen from the above description that the switch arrangement 32 and the current sensor device 33 are a ballast for current intensity limitation, which is otherwise customary for the operation of fluorescent lamps, e.g. a choke coil or a capacitor. An induction coil 23 is also necessary in this device, but this is only used to generate the induction voltage sufficient to ignite the fluprescent lamp 20 and can therefore have a smaller inductance than a choke coil which is also suitable for current limitation. Furthermore, the automatic current control by means of the devices 32 and 33 ensures that when the switch arrangement 26 is in a conductive state, a current with a nominal current flows correctly through the two glow electrodes 21 and 22, and that influences of voltage fluctuations of the supplying current source 31 on the brightness and the service life the fluorescent lamp are largely switched off.

The device shown in FIG. 1 and described with reference thereto also allows the brightness of the fluorescent lamp 20 to be regulated between zero and the nominal value in a relatively simple manner. For this purpose, only the control device 27 for controlling the switch arrangement 26 has to be designed as a clock pulse generator. With the clock pulses U26, the switch arrangement 26 with a repetition frequency of at least 50 Hz is periodically alternately controlled during a first time interval in the conductive state and during a second time interval in the non-conductive state, the ratio of the time intervals with conductive and non-conductive state being adjustable to regulate the resulting average brightness of the fluorescent lamp. In this connection, reference is made to FIG. 3, in the upper diagram of which the clock pulse signal U26 generated by the control device 27 is shown. It is assumed that the switch arrangement 26 is controlled in the conductive state for the duration of a clock pulse of the signal U26. Thus, the current flow through the fluorescent lamp 20 is interrupted for the duration of a clock pulse, since then the switch arrangement 26 and the two glow electrodes 21 and 26 together form a low-impedance shunt path, whereas the current I _{L flows} through the fluorescent lamp in the pulse pauses of the signal U26, as can be seen in the lower diagram of FIG. 3. In the left part of FIG. 3, the clock pulses of the signal U26 have a duration t1 which is shorter than the duration t2 of the pulse pauses. Accordingly, the resulting interruptions in the current I _L flowing through the fluorescent lamp 20 are each comparatively short, so that the fluorescent lamp has a fairly high brightness. 3, the clock pulses of the signal U26 have a duration t1 'which exceeds the duration t2' of the pulse pauses. This results in comparatively longer interruptions of the Sjromes I _L by the fluorescent lamp, so that the resulting brightness is lower. It is understood that the brightness of the fluorescent lamp 20 is greatest when the clock pulses of the signal U23 are only very narrow or when the clock signal U 23 is completely switched off becomes. Conversely, the brightness of the fluorescent lamp is at its lowest or even zero if the pauses between the pulses of the clock signal U23 are very narrow or if a continuous voltage is supplied as a control signal to the switch arrangement 26 instead of the clock pulses.

It should be noted that the frequency of the pulses of the control signal U33 for the automatic current regulation adjusts itself to a value which considerably exceeds that of the clock frequency of the control signal U26 for the brightness control of the fluorescent lamp 20.

The two switching devices 24 and 25 are only intended to reverse the direction of flow of the current I _L through the fluorescent lamp 20 from time to time. Thus, the repetition frequency of the control signals U24 and U25 for the reversal of the switching devices 24 and 25 is in any case chosen to be considerably lower than the clock frequency of the control signal U26 for the brightness control of the fluorescent lamp 20.

FIG. 4 shows the device shown in FIG. 1 again in more detail, the same reference numerals being used insofar as the components are the same.

4, the direct current source 31 has a rectifier bridge 36, the one diagonal of which is connected to two terminals 37 and 38 for connection to an alternating current network with, for example, 220 V effective voltage. The other diagonal of the rectifier bridge 36 is connected to a charging capacitor 39 and connected to the positive pole 29 or the negative pole 30 of the direct current source. 4, the switch arrangement 26 for controlling the brightness of the fluorescent lamp 20 has a power transistor 40, the emitter-collector path of which is connected to a diagonal of a rectifier bridge 41. The other diagonal of the rectifier bridge 41 is connected to the connection ends 21b and 22b of the two glow electrodes 21 and 22 of the fluorescent lamp 20. The control device 27 belonging to the switch arrangement 26 contains an astable multivibrator 42 for generating clock pulses which are fed to a monostable multivibrator 43 which in turn generates the pulses of the control signal U26 and delivers them to the base-collector path of the transistor 40. The monostable multivibrator 43 is assigned, for example, a manually adjustable setting resistor 44, with the aid of which the width t1 (FIG. 3) of the pulses of the control signal U26 can be changed for the purpose of controlling the brightness of the fluorescent lamp 20.

It can also be seen in FIG. 4 that the switch arrangement 32 provided for current regulation has a power transistor 45 and two transistors 46 and 47. The emitter-collector path of the power transistor 45 is switched into the supply current path between the negative pole 30 of the direct current source 31 and the current sensor device 33. The base of the power transistor 45 is connected via a resistor 48 to the emitter of the transistor 46, the collector of which is connected to an auxiliary voltage source (not shown). The base of this transistor 46 and the emitter of the power transistor 45 are connected to a first control signal input 321, to which the control signal U33 already mentioned with reference to FIG. 1 is fed. The other transistor 47 is connected with its emitter-collector path between the base and the emitter of the power transistor 45. The base and the emitter of transistor 47 are connected to a second control signal input 322, to which a further control signal U60, the purpose of which will be explained, can be supplied.

The current sensor device 33 has a measuring resistor 50 located in the feed current path to the fluorescent lamp 20 and a DC voltage amplifier 51, the input of which is connected to the ends of the measuring resistor 50. The output of the amplifier 51 is connected to an optocoupler 52, which outputs the control signal U33. The optocoupler 52 is used in a known manner only for the electrical isolation of the amplifier 51 and the control signal output of the current sensor device 33.

4 that each of the two switching devices 24 and 25 has a pair of complementary power transistors 55 and 56, the emitters of which are connected to one another and to the induction coil 23 and to the connection end 22a of the glow electrode 22, respectively. The collector of the one power transistor 55 is connected to the positive pole 29 of the direct current source 31, while the collector of the other power transistor 56 is connected to the current sensor device 33. The control voltage U24 and U25 already mentioned with reference to FIG. 1 is applied to the base-emitter paths of the power transistors 55 and 56 via resistors 57 and 58.

4, the control device 28 for generating the control voltages U24 and U25 is designed as follows: an astable multivibrator 60 is connected to the input of a flip-flop 61. The latter has two signal outputs, each of which is provided for electrical isolation see optocouplers 62 and 63 are connected to the input of an amplifier 64 and 65 respectively. The outputs of the two amplifiers 64 and 65 supply the mentioned control voltages U24 and U25 for reversing the switching devices 24 and 25.

4 is basically the same as has been described with reference to FIGS. 1, 2 and 3. In addition, only the functioning of the switching devices 24 and 25 and the associated control device 28 will now be explained with reference to FIG. 5. The top diagram of FIG. 5 illustrates the chronological sequence of the voltage pulses U60 generated by the astable multivibrator 60. The repetition frequency of these pulses is considerably lower than the clock frequency of the pulses of the control signal U26 for the brightness control of the fluorescent lamp 20. The flip-flop 61 is controlled by the falling edge of each pulse U60, the two outputs of which are inverted signals U62 and U63 according to the second and deliver the third diagram of FIG. 5. By means of the optocouplers 62 and 63, the signals U62 and U63 are transmitted to the amplifiers 64 and 65, the outputs of which deliver the control voltages U24 and U25 also shown in FIG. 5 to the power transistors 55 and 56 of the switching devices 24 and 25. Each time the control voltage U24 is negative and the control voltage U25 is positive, the power transistor 56 of the switching device 24 and the power transistor 55 of the other switching device 25 are controlled to be in the conductive state, so that the current _IL in the direction of the arrow in FIG. 4 flows through the fluorescent lamp 20. This is shown in the bottom diagram in FIG. 5 as a positive current flow. Conversely, if the control voltage U24 is positive and the control voltage U25 is negative, the are The power transistor 55 of the switching device 24 and the power transistor 56 of the other switching device 25 in the conductive state, while at the same time the two other power transistors are non-conductive, so that the current I _L flows through the fluorescent lamp 20 against the arrow in FIG. 4. This is illustrated in the bottom diagram in FIG. 5 as a negative current flow. The regular, brief interruptions in the current I _L shown in FIG. 5 are caused by the switch arrangement 26 for the brightness control of the fluorescent lamp, as has been explained in detail with reference to FIG. 3.

The device shown in FIG. 4 also has a special feature: the voltage pulses U60 generated by the astable multivibrator 60 are also supplied to the second control signal input 322 of the switch arrangement 32 provided for current regulation via two conductors 68 and 69. Each pulse U60 controls the transistor 47 in the conductive state, whereby the power transistor 45 is simultaneously controlled in its non-conductive state. This is done completely independently of the control signal U33 for the automatic current control. In each case, when the power transistor 45 is controlled into the non-conductive state by a pulse U60, the current flow to the negative pole 30 of the direct current source 31 is interrupted. Because of the electrical energy stored in the induction coil 23, however, the current I _L continues to flow via the freewheeling diode 34 until the stored energy is used up, the current strength decreasing exponentially. This can be seen in the bottom diagram of FIG. 5 in the time intervals t3. The duration of the pulses U60 and thus of the time intervals t3 is now chosen such that the current intensity practically drops to zero in each case before the two switching devices 24 and 25 are reversed. The periodic reversal of the transistors 55 and 56 thus takes place in a practically currentless state, as a result of which the power loss and heating of these transistors can be kept low.

The embodiment of the device according to the invention shown in FIG. 6 corresponds in its basic structure to that according to FIGS. 1 and 4. In detail, only the two switching devices 24 and 25 and the associated control device 28 are designed differently. 6, the two switching devices 24 and 25 each have two thyristors 71 and 72 (instead of the power transistors 55 and 56 in FIG. 4). The control device 28 again has an astable multivibrator 60, which is connected both to the two conductors 68 and 69 and to a monostable multivibrator 74. The latter in turn is connected to the control input of a flip-flop 75, which is followed by an amplifier 76. An input winding 77 of a pulse transformer 78 is connected to the output of the amplifier 76 and also has four electrically isolated output windings 79, 80, 81 and 82. Each of these output windings 79 to 82 is connected, on the one hand, to the cathode and, on the other hand, to the ignition electrode of one of the thyristors 71 and 72, a diode 83, 84, 85 and 86 being interposed in each case in order to transmit negative voltage pulses from the ignition electrodes of the thyristors 71 and keep 72 away.

The direct current source 31, the switch arrangement 32 provided for current regulation and the associated current sensor device 33 are only indicated in the form of blocks in FIG. 6, but have the same off in detail formation as in Fig. 4, so that no further words are needed.

The mode of operation of the device shown in FIG. 6 is basically the same as has been explained with reference to FIGS. 1, 2 and 3. In addition, only the mode of operation of the switching devices 24 and 25 and the associated control device 28 according to FIG. 6 will now be explained. The astable multivibrator 60 generates a voltage pulse train U60, as shown in the top diagram in FIG. 7. The repetition frequency of these pulses is considerably lower than the clock frequency of the pulses of the control signal U26 for the brightness control of the fluorescent lamp 20. Via the conductors 68 and 69, the pulses of the signal U60 are fed to the second control input 322 of the switch arrangement 32. As has already been explained with reference to FIGS. 4 and 5, each pulse of the signal U60 uses the control transistor 47 to control the power transistor 45 of the switch arrangement 32 into the non-conductive state, in each case for the entire duration t3 of the pulse U60 in question. The pulse duration t3 is chosen so large that the current flow through the freewheeling diode 34 has decayed completely to zero before the end of the pulse U60 in question.

The voltage pulse sequence U60 generated by the astable multivibrator 60 is, however, also fed to the trigger input of the monostable multivibrator 74, which is triggered by the falling edge of each pulse of the signal U60 and also supplies a voltage pulse sequence U74 at its output, which is shown in the second diagram in FIG. 7 is shown. The duration t4 of each pulse of the signal U74 is relatively short in comparison to the duration t3-the abstei The leading edge of each pulse of signal U74 controls flip-flop 75, which now supplies signal U75, also shown in FIG. 7, to the input of amplifier 76. The output of the amplifier 76 supplies a voltage U76 to the input winding 77 of the pulse transformer 78, the time course of which is also shown in FIG. 7. Only short, needle-shaped voltage pulses are induced in the four output windings 79 to 82 of the pulse transformer, which coincide in time with the rising and falling edges of the voltage U76 supplied to the input winding 77. The control voltages U241 and U242 or U251 and U252 shown in the third lowest and second lowest diagram of FIG. 7 are supplied to the thyristors 71 and 72 in the form of needle pulses. A suitable polarity of the output windings 79 to 82 ensures that a needle pulse U251 or U242 reaches the thyristor 71 of the switch arrangement 25 and the thyristor 72 of the switch arrangement 24 in each case on the rising edges of the voltage U76 in order to simultaneously turn these thyristors into the conductive state control, and that in each case on the rising edges of the voltage U76, a needle pulse U241 or U252 reaches the thyristor 71 of the switch arrangement 24 and the thyristor 72 of the switch arrangement 25 in order to control these two thyristors in turn in the conductive state.

As is known, a thyristor has the property of maintaining the conductive state of the array-cathode path even after the control voltage has ceased to exist, as long as a current flows through the latter. In the present case, the pulses of the control voltage U60 supplied to the second control signal input 322 each cause an interruption in the current flow, so that before the end of the pulse U60 in question, all thyristors become deenergized and therefore non-conductive. By alternately firing thyristors 71 and 72 in the two switching devices 24 and 25, the direction of flow of the current I _L through the fluorescent lamp 20 is periodically reversed, as the bottom diagram in FIG. 7 shows.

The purpose of the monostable multivibrator 74 is to delay the rise and fall of the voltages U75 and U76 and consequently the occurrence of the needle pulses of the control voltages U241, U242, U251 and U252 with respect to the end of the pulses of the signal U60 by the time period t4. This ensures that the needle pulses provided for firing the thyristors 71 and 72 only occur after the power transistor 45 (FIG. 4) in the switch arrangement 32 has resumed its conductive state at the end of the previous pulse of the signal U 60 and thus the anode-cathode sections of the thyristors are already live again. Without the aforementioned delay by the time period t4, reliable firing of the thyristors 71 and 72 could be called into question.

The further exemplary embodiment of the device according to the invention shown in FIG. 8 differs from the examples described so far mainly in that only a single switching device 25, but two freewheeling diodes 341 and 342 and a direct current source 90 with a positive pole 91, a negative pole 92 and one neutral pole 93 are present. The neutral pole 93 is connected to an AC power supply terminal 94, while a second AC power supply terminal 95 is connected to the positive pole 91 and the negative pole 92 of the direct current source 90 via two oppositely polarized rectifier diodes 96 and 97. Two charging capacitors 98 and 99 are connected between the positive pole 91 and the neutral pole 93 or between the latter and the negative pole 92. The neutral pole 93 of the direct current source 90 is connected via a diagonal of a rectifier bridge 100, via the measuring resistor 50 of the current sensor device 33 and the induction coil 23 to the one connection end 21a of the glow electrode 21 of the fluorescent lamp 20. Instead of power transistors or thyristors in the exemplary embodiment according to FIG. 8, the switching device 25 contains two gate turn-off semiconductor switches 101 and 102, of which one 101 lies between the positive pole 91 of the direct current source 90 and the connection end 22a of the second glow electrode 22 of the fluorescent lamp 20 while the other 102 is between the same terminal end 22a and the negative pole 92 of the DC power source. Gate-turn-off semiconductor switches are only recently launched electronic components that combine the properties of a thyristor and a high-voltage switching transistor and can be brought into a conductive state by a positive control pulse and into a non-conductive state by a negative control pulse (cf. "Electronic Components and Applications ", Vol. 2, No. 4, Aug. 1980, pages 194ff).

The emitter-collector path of the power transistor 45 in the switch arrangement 32 used for current regulation is connected to the second diagonal of the rectifier bridge 100 mentioned. In the present case, the latter has only a single control signal input 321, which is connected to the emitter of the power transistor 45 and the base of a further transistor 46. The emitter of transistor 46 is connected to the base of power transistor 45 via a resistor 48, while the collector of transistor 46 is connected to an auxiliary voltage source (not shown). The configuration of the switch arrangement 32 is thus the same as in FIG. 4, with the Exception that the second control signal input 322 and the associated transistor 47 are unnecessary. The current sensor device 32 is of exactly the same design as in the example according to FIG. 4, and it supplies the same control signal U33 to the control signal input 331 of the switch arrangement 32. The one freewheeling diode 341 is on the one hand at the positive pole 91 of the direct current source 90 and on the other hand at the connecting conductor 112 connected between the rectifier bridge 100 and the measuring resistor 50 of the current sensor device 33, while the other freewheeling diode 342 is connected between the mentioned connecting conductor 112 and the negative pole 92 of the direct current source.

The control device 28 used to control the switching device 25 is configured as follows: An astable multivibrator 105 is connected to the trigger input of a flip-flop 106, which is followed by an amplifier 107. The output of amplifier 107 feeds an input winding 108 of a pulse transformer 109 with two galvanically isolated output windings 110 and 111. One output winding 110 is connected between the cathode and the control electrode of the first gate turn-off semiconductor switch 101, while the other output winding 111 is on the cathode and the control electrode of the second gate turn-off semiconductor switch 102 are connected. The astable multivibrator 105 can be synchronized with a trigger signal which e.g. Via at least one capacitor 113 is derived from the line AC voltage supplied to the connection terminals 94 and 95.

Analogously to the first exemplary embodiment according to FIG. 1, a switch is connected between the connection ends 21b and 22b of the two glow electrodes 21 and 22 of the fluorescent lamp 20 Order 26 available, which can be electronically controlled by an associated control device 27. The switch arrangement 26 according to FIG. 8 has a rectifier bridge 115, one diagonal of which is connected to the connection ends 21b and 22b of the glow electrodes 21 and 22. The anode-cathode path of a further gate-turn-off semiconductor switch is connected to the other diagonal of the rectifier bridge 115. The control arrangement 27 contains an astable multivibrator 117, which is followed by an amplifier 118. The output of the amplifier 118 feeds the primary winding 119 of a pulse transformer 120, the secondary winding 121 of which is connected to the Kahtode and the control electrode of the gate turn-off semiconductor switch 116 in the switch arrangement 26. The astable multivibrator 117 is provided with a device 122, for example in the form of a manually variable setting resistor, by means of which the ratio of pulse duration to pulse pause can be changed for the purpose of controlling the brightness of the fluorescent lamp 20.

The use and mode of operation of the device according to FIG. 8 is explained below with reference to FIGS. 9 and 10: It is initially assumed that in the switching device 25 the gate turn-off semiconductor switch 101 is in the conductive state and the other gate turn-off semiconductor switch 102 is non-conductive. It is further assumed that the fluorescent lamp 20 is ignited. Under these conditions, an electric current I _{L flows} from the positive pole 91 of the direct current source 90 via the gate turn-off semiconductor switch 101, through the fluorescent lamp 20, through the induction coil 23, through the measuring resistor 50 of the current sensor device 33 and via the power transistor 45 neutral pole 93 of the direct current source 90. In dependence on the control by the current sensor device 33 generated control signal U33, the switch arrangement 32 automatically ensures that the current intensity oscillates only to a relatively small extent around a predetermined desired value, as has been explained in more detail with reference to FIGS. 1 and 2. Each time the power transistor 45 is in a non-conductive state, the freewheeling diode 341 enables the further current flow on the basis of the electrical energy stored in the induction coil 23.

The astable multivibrator 117 in the control device 27 generates a clock signal U117, the temporal course of which is shown in the top diagram in FIG. 9. The clock signal U117 is amplified in the amplifier 118 and then fed to the primary winding 119 of the pulse transformer 120. Needle-shaped voltage pulses U121 are induced in the secondary winding 121 of the pulse transformer 120, as shown in the middle diagram in FIG. 9. It can be seen that a positive needle pulse U121 arises in each case on the rising edges of each pulse of the clock signal 117 and a negative needle pulse arises on the falling edges of each clock signal pulse. Each positive needle pulse U121 controls the gate turn-off semiconductor switch 116 in the switch arrangement 26 into the conductive state, which is maintained until the subsequent negative needle pulse U121 occurs. Each negative needle pulse U121 thus controls the gate turn-off semiconductor switch back to the non-conductive state. If and as long as the gate turn-off semiconductor switch 116 is in the conductive state, the current flow I _L through the fluorescent lamp 20 is interrupted, because then the current flows through the low-resistance shunt that flows through the glow electrode 22, part of the rectifier bridge 115 , the gate turn-off semiconductor switch 116, another part of the rectifier bridge 115 and the glow electrode 21 is formed. The resulting one The time course of the current I _L flowing through the fluorescent lamp 20 is shown in the bottom diagram in FIG. 9. Analogously to the first exemplary embodiment, the brightness of the fluorescent lamp 20 can be arbitrarily controlled by changing the ratio of the time intervals t1 and t2 in which the gate turn-off semiconductor switch 116 is conductive or nonconductive. The clock frequency of the signal U117 generated by the multivibrator 117 is chosen to be considerably higher than 50 Hz, so that the human eye does not notice any flickering of the fluorescent lamp.

The astable multivibrator 105 contained in the control device 28 generates a voltage pulse sequence U105, as is illustrated in the second uppermost diagram in FIG. 10. The pulses of the signal U105 are synchronized with the AC line voltage U _W , which is supplied to the connecting terminals 94 and 95 of the direct current source 90. In the uppermost diagram in FIG. 10, the time profile of the AC line voltage U _{W is} indicated with a dash-dotted line. It can be seen that the pulses of the signal U105 generated by the multivibrator 105 are triggered shortly after the zero crossing of the mains AC voltage, and that the repetition frequency of the pulses is twice the mains frequency. The falling edge of each pulse of signal 105 controls flip-flop 106, which in turn generates signal U106 shown in the third diagram in FIG. 10, which is amplified in amplifier 107 and fed to input winding 108 of pulse transformer 109. In the two output windings 110 and 111 of the pulse transformer 109, needle-shaped voltage pulses U251 and U252 are induced, which are fed as control signals to the two gate turn-off semiconductor switches 101 and 102 of the switching device 25. In the third lowest and second lowest diagram of FIG. 10, the resulting needle pulses U252 and 252 is also shown, and it can be seen that they occur in each case on the rising and falling edges of the signal U106 generated by the flip-flop 106. Each positive needle pulse U251 controls the first gate turn-off semiconductor switch 101 into the conductive state, while at the same time a negative needle pulse U252 controls the second gate turn-off semiconductor switch 102 into the non-conductive state. Conversely, each negative needle pulse U251 controls the first gate turn-off semiconductor switch in the non-conductive state, while at the same time a positive needle pulse U252 controls the second gate turn-off semiconductor switch 102 in the conductive state. By reversing the two gate turn-off semiconductor switches 101 and 102, the direction of flow of the current flowing through the fluorescent lamp 20 is periodically reversed, as illustrated in the bottom diagram in FIG. 10.

Each time the first gate turn-off semiconductor switch 101 is non-conductive and the other 102 is conductive, the current flows from the neutral pole 93 of the direct current source 90 via the power transistor 45 provided for current regulation, through the measuring resistor 50 of the current sensor device 32, through the induction coil 23 _"by the fluorescent lamp 20 and above the gate-turn-off semiconductor switch 102 to the negative pole 92 of the DC power source 90. in this case, the free-wheeling diode 342 allows WeiterfHessen of the current in the brief time intervals in which the power transistor 45 in control function of the control signal U33 is always non-conductive.

It can be seen from the above description that the electrical energy used to feed the fluorescent lamp 20 is always only from one or the other half of the Direct current source 90 is supplied, with the half of the direct current source used for the supply being changed simultaneously with the reversal of the switching device 25. Through the aforementioned synchronization of the pulses of the signal U105 generated by the astable multivibrator 105 with the AC line voltage U _{W, it} can be achieved that the fluorescent lamp 20 is supplied in each case by that half of the direct current source 90, the charging capacitor 98 or 99 of which, at this moment, is supplied by a AC half wave is charged. In addition to the AC line voltage U _W , the uppermost diagram in FIG. 10 also shows the time profile of the positive DC voltage U98 or negative DC voltage U99 lying across the charging capacitors 98 and 99. Comparing the top and bottom diagram of FIG. 10, it can be seen that the current I _L flowing through the fluorescent lamp 20 becomes positive (according to the arrow in FIG. 8) when the positive voltage U98 across the charging capacitor 98 increases towards the peak value of a positive mains voltage half-wave, and that the current I _L reverses its direction each time the voltage U98 has dropped to approximately the same value as at the beginning. During this same period of time, the voltage U99 at the other charging capacitor 99 remains practically constant because then the relevant half of the direct current source 90 is unloaded. Conversely, the current I _L is negative (opposite to the arrow in FIG. 8) when the negative DC voltage U99 rises against the peak value of a negative AC voltage half-wave, and the current I _L then reverses its direction when the voltage U99 is raised again about the initial value has dropped. This is achieved by a suitable choice of the pulse duration t5 of each voltage pulse of the signal U105 generated by the multivibrator. This results in the lowest possible ripple in the DC voltages U98 and U99. FIG. 11 shows a further exemplary embodiment in which the same direct current source 90 as in the example according to FIG. 8 is used. The neutral pole 93 of the direct current source is only connected via the measuring resistor 50 of the current sensor device 33 and via the induction coil 23 to the one connecting end 21a of the glow electrode 21 of the fluorescent lamp 20. The corresponding connection end 22a of the other glow electrode 22 is connected to the positive pole 91 of the direct current source via an electronically controllable switch arrangement 32 for regulating the current intensity and to the negative pole 92 of the direct current source 90 via a further similar switch arrangement 32 'for regulating the current intensity. The configuration of the switch arrangement 32 completely corresponds to that of the switch arrangement 32 in FIG. 4, and there are again two mutually independent control signal inputs 321 and 322. The other switch arrangement 32 'is analog, but is equipped with transistors 45', 46 'and 47', which are complementary to the corresponding transistors 45, 46 and 47 of the switch arrangement 32. The switch arrangement 32 'also has two control signal inputs, which are designated 321' and 322 '.

The current sensor device 33 contains amplifiers 51 and 51 ', the inputs of which are connected to the ends of the measuring resistor 50. The output of the amplifier 51 is connected by means of an optocoupler 52 to the one control signal input 321 of the switch arrangement 32 in order to supply the latter with a control signal U33 for the automatic current regulation. In an analogous manner, the output of the other amplifier 51 'is connected to the one control signal input 321' of the switch arrangement 32 'by means of a second optocoupler 52', in order to also provide a control signal U33 'for the current intensity Instead of the freewheeling diodes 341 and 342 of the exemplary embodiment according to FIG. 8, in the example according to FIG. 11 there are two complementary transistors 343 and 344 serving the same purpose, the collector-emitter paths of which have the opposite polarity between the neutral pole 93 of the direct current source 90 and the glow electrode 22 of the fluorescent lamp 20 are switched on. The collector-emitter path of a control transistor 345 and 346 is connected in parallel to the base-emitter path of these transistors 343 and 344. The output signal of the amplifier 127 in the control device 28 is fed to the base of each control transistor 345 or 346. The collector of one control transistor 345 is connected to a positive auxiliary potential via a resistor 347, while in an analogous manner the collector of the other control transistor 346 is connected to a negative auxiliary potential via a resistor 348.

Between the terminal ends 21b and 22b of the two annealing electrodes 21 and 22 of the fluorescent lamp 20 is also in this embodiment, an electronically controllable ^S chalteranordnung 26, which is associated with a control device 27. The configuration of the switch arrangement 26 and the associated control device 27 is the same as in one of the preceding examples and therefore need not be explained any longer.

Finally, the device according to FIG. 11 also has a control device 28 which has an astable multivibrator 125, a flip-flop 126 and an amplifier 127 which are connected in series. The output of amplifier 117 is connected to control signal inputs 322 and 322 'of the two switch arrangements 32 and 32'. The astable multivibrator 125 is the same as in the exemplary embodiment according to FIG. 8 due to an external signal

- can be synchronized and for this purpose, for example by means of a capacitor 113 or the like, coupled to the mains connection terminals 94, 95.

The operation and mode of operation of the device shown in FIG. 11 is brief as follows: It is assumed that the fluorescent lamp 20 is lit. The astable multivibrator 125 in the control device 28 generates a voltage pulse sequence U125, as shown in the second uppermost diagram in FIG. 12. The pulses of this voltage U125 are synchronized with the AC line voltage U _W , in such a way that a pulse of voltage U125 begins immediately after each zero crossing of the AC line voltage U _W , as can be seen by comparing the top and second-top diagrams of FIG. 12.

The flip-flop 126, which generates a signal U125 according to the third, is controlled by the rising edges of the pulses U125

Diagram of FIG. 12 provides the input of amplifier 127. At the output of amplifier 127, a signal U127 appears, which is that shown in the second bottom diagram of FIG

Has course and is fed as a control signal to the inputs 322 and 322 'of the two switch assemblies 32 and 32'.

Each time the signal U127 becomes negative, the emitter-collector path of the transistor 47 'in the switch arrangement 32' becomes conductive, as a result of which the power transistor 45 'is controlled in the non-conductive state. Then the connection between the connection end 22a of the glow electrode of the fluorescent lamp 20 and the negative pole 92 of the direct current source 90 is interrupted. At the same time, the signal U127 by means of the control transistor 346 also causes the transistor 344 to become non-conductive and the control transistor 345 to switch the transistor 343 to Controlled conductive state, so that the latter is the same as the freewheeling diode 341 in the example of FIG. 8 is effective. Continue at the same time the signal U127 controls the transistor 47 in the switch arrangement 32 into the non-conductive state, with the result that the power transistor 45 becomes conductive. As a result, the energy required for the operation of the fluorescent lamp 20 is thus supplied solely by the upper half of the DC power source 90 in FIG. 11, the charging capacitor 98 being charged in this phase by a positive half-wave of the AC mains voltage U _W. The current I _L flowing through the fluorescent lamp 20 is then positive in the sense of the arrow in FIG. 11.

Each time the signal U127 changes from negative to positive potential, the power transistor 45 in the switch arrangement 32 and also the transistor 343 are controlled into the non-conductive state, while at the same time the power transistor 45 'of the other switch arrangement 32' and subsequently as a freewheeling diode acting transistor 344 are controlled in the conductive state. As a result, the connection between the positive pole 91 of the direct current source 90 and the connection end 22a of the glow electrode 22 of the fluorescent

Zenzlampe interrupted and thus the upper half of the DC power source 90 in Fig. 11 switched off. In its place, the other half of the direct current source 90 is now effective for feeding the fluorescent lamp 20, which is why the direction of the current I _L flowing through the fluorescent lamp 20 is now reversed (contrary to the arrow in FIG. 11). In this operating phase, the charging capacitor 99 of the loaded lower half of the direct current source 90 is charged by a negative half wave of the mains alternating voltage U _W , as can be seen from the top diagram in FIG. 12. As in the exemplary embodiment described with reference to FIGS. 8 and 10, the respective duration t5 of the pulses of the signal U125 generated by the multivibrator 125 is selected in such a way that the lowest possible ripple of the direct voltages U98 and U99 lying across the charging capacitors 98 and 99 results.

It can be seen that in the last-described exemplary embodiment of the device according to the invention, the periodic reversal of the flow direction of the current I _L flowing through the fluorescent lamp 20 is advantageously carried out by switching the two switch arrangements 32 and 32 'on and off by means of the control signal U127, so that the switching devices 24 and 25 of the previous exemplary embodiments are eliminated. However, the two switch arrangements 32 and 32 'not only serve to reverse the direction of current flow through the fluorescent lamp 20, but also ensure, depending on the control signals U33 and U33', which are generated by the current sensor device 33, that the current intensity is automatically maintained a predetermined setpoint is adjusted. The mode of operation of the current regulation is in principle the same as has been explained with reference to FIGS. 1 and 2, which is why further words are unnecessary. The switch arrangement 26 still shown in FIG. 11 and the associated control device 27 permit brightness control of the fluorescent lamp 20 in the same manner as has been explained with reference to the preceding exemplary embodiments. The bottom diagram of FIG. 12 shows the time profile of the current I _L flowing through the fluorescent lamp 20, the interruptions in the current flow causing the brightness control also being indicated. It is understandable that the repetition frequency of these interruptions is considerably higher than the frequency of the AC line voltage U _W , with which the periodic reversal of the direction of flow of the current I _{L is} synchronized.

While in the exemplary embodiments according to FIGS. 4, 8 and 11 the switch arrangements 32 and 32 'used for current regulation have a power transistor 45 or 45' for switching the current path on and off, it is of course also possible to use a thyristor instead of the power transistor or to use a gate turn-off semiconductor switch. Likewise, the switch arrangement 26 provided for the brightness control can optionally contain a power transistor, a thyristor or a gate turn-off semiconductor switch. The same can of course also be said of the switching devices 24 and 25, as can already be seen from the various exemplary embodiments described.

In certain cases it may be expedient to connect two fluorescent lamps 20 in series, as is known per se. A switch arrangement 26 is then to be provided for each of these fluorescent lamps connected in series, which is arranged in the same way as described here. The control device 27 for generating the control signal U27 for controlling the switch arrangement 26 can - in deviation from the described exemplary embodiments - be designed such that a change in the ratio of the pulse duration t1 to the pulse pause t2 takes place automatically each time the fluorescent lamp 20 is started up, in such a way that the brightness of the fluorescent lamp is continuously increased from zero to the nominal value or an arbitrarily set intermediate value during a time-limited starting phase which begins when the fluorescent lamp is switched on. This is achieved if, for example, the adjusting resistor 44 or 122 shown in FIGS. 4 and 8 is replaced by an electronic circuit arrangement known per se, by means of which the pulse width t1 is controlled as a function of the gradual charging or discharging of a capacitor.

Finally, it should also be mentioned that in all the exemplary embodiments described, the switch arrangement 32 and the associated current sensor device 33 can be replaced by other devices known per se for automatically regulating the current intensity to a predetermined desired value.

Claims

Patent claims

1. Device for feeding at least one fluorescent lamp with glow electrodes, each having two connection ends, one of which can be connected to a power source and the other can be connected to one another by means of a switch arrangement, and with devices having an induction coil for limiting the strength of the power source current flowing to the lamp, characterized in that the switch arrangement (26) is electronically controllable and is connected to a control device (27) which, at least during a time-limited starting phase which begins when the fluorescent lamp (20) is switched on, switches the switch arrangement ( 26) with a clock frequency of at least 50 Hz periodically alternately in a conductive state during a first time interval (t1) and in a non-conductive state during a second time interval (t2), and that at least one electronic circuit arrangement run to limit the current strength g (32, 33) for automatically regulating the current strength ( _IR ) to a predetermined nominal value is present.

2. Device according to claim 1, characterized in that the circuit arrangement (32, 33) for regulating the current level in the supply circuit of the fluorescent lamp (20) lying second electronically controllable switch arrangement (32), which by a of the respective current current of the Lamp (20) supplied current (I _R ) dependent signal (U33) can be switched on and off, so that the fluorescent lamp (20) is fed with current pulses during pauses in between and the pulse / pause ratio depends on the respective current intensity, whereby the again recovery frequency of the current pulses is higher than the clock frequency for controlling the first-mentioned switch arrangement (26), and that at least one free-wheeling diode (34; 341, 342) is arranged in such a way that in each case the second switch arrangement (32) is switched off during the breaks Energy produced in the induction coil (23) and current flowing on via the free-wheeling diode (34; 341, 342) and the fluorescent lamp (20).

3. Device according to claim 1 or 2, characterized in that the current source is a direct current source (31; 90) and at least one electronically controlled switching device (24, 25) for alternately reversing the direction of flow of the current flowing through the fluorescent lamp (20) (I _L ) as well as a control device (28) for periodically reversing the switching device (24, 25) and that the switching frequency is lower than the clock frequency for controlling the first switch arrangement (26).

4. Device according to claims 2 and 3, characterized in that the second switch arrangement (32) for regulating the current intensity is also connected to the control device (28) for periodically switching the switching device (24, 25) and this control device (28) immediately before the reversal of the switching device (24, 25) delivers a signal (U60) to the switch arrangement (32) for regulating the current intensity, by means of which this switch arrangement (32) is switched off independently of the instantaneous current intensity until the switching device ( 24, 25) is reversed without current.

5. Device according to claim 3 or 4, characterized in that the direct current source (90) two one-way rectifier arrangements fed from an alternating current network (96, 97), each with at least one charging capacitor (98, 99), the charging capacitor (98) of the one rectifier arrangement (96) each having the positive half-waves of the AC voltage (U _W ) and the charging capacitor (99) of the other rectifier arrangement ( 97) can be charged by the negative half-waves of the alternating voltage, and that the switching device (25) is controlled synchronously with the alternating current frequency, so that the fluorescent lamp (20) is connected in each case to the charging capacitor (98 or 99) which is connected at the same time by the corresponding half-wave of the AC voltage is charged.

6. Device according to one of claims 3 and 5, characterized in that the direct current source (90) for supplying a positive and a negative direct voltage (U98 or U99) is designed with respect to a neutral potential and accordingly a positive pole (91) , a negative pole (92) and a neutral pole (93) that the. Circuit arrangement (32, 33) for regulating the current between the neutral pole (93) and the one glow electrode (21) of the fluorescent lamp (20) and that a single switching device (25) with two controlled switch arrangements (101, 102) is present one of which is connected between the positive pole (91) and the other glow electrode (22) of the fluorescent lamp (20) and the second between the negative pole (92) and the same glow electrode (22).

7. Device according to claim 1, characterized in that the current source (90) is a direct current source for supplying a positive and a negative direct voltage (U98 or U99) with respect to a neutral potential and accordingly a positive pole (91), a negative one Pole (92) and a neutral pole (93) has that the neutral pole (93) is connected to the one glow electrode (21) of the fluorescent lamp (20) and there are two circuit arrangements (32, 33; 32 ', 33) for regulating the current intensity, one of which is between the positive pole (91) and the other glow electrode (22) of the fluorescent lamp (20) and the second between the negative pole (92) of the same glow electrode (22) is switched on, and that each of the two circuit arrangements (32, 33; 32 ', 33) has a common control device (28) is connected, which provides at least one control signal (U127) for alternately switching off the switch arrangements (32, 33; 32 ', 33) present in the circuit arrangements (32, 32') independently of the respective current current intensity, so that the direction of flow of the the fluorescent lamp (20) flowing current (I _L ) is periodically reversed, the frequency of the current direction changes being lower than the clock frequency for controlling the first switch arrangement (26).

8. Device according to claim 7, characterized in that the direct current source (90) has two one-way rectifier arrangements (96, 97) fed from an alternating current network, each with at least one charging capacitor (98, 99), the charging capacitor (98) of the one rectifier arrangement (96) is arranged between the neutral pole (93) and the positive pole (91) of the direct current source (90) and can be charged in each case by the positive half-waves of the alternating voltage (U ^.), Whereas the charging capacitor (99) of the other rectifier arrangement (97 ) between the neutral pole (93) and the negative pole (92) of the direct current source (90) and can be charged in each case by the negative half-waves of the alternating voltage (U _W ), and that the control signal (U127) of the common control device (28) with the alternating current frequency is synchronized, so that the fluorescent lamp (20) in each case is controlled by means of the two circuit arrangements (32, 33; 32 ', 33) for regulating the current intensity is connected to the charging capacitor (98 or 99) which is charged at the same time by the corresponding half-wave of the AC voltage.

9. Device according to one of claims 1 to 8, characterized in that the first switch arrangement (26) has at least one power transistor, at least one thyristor or at least one gate turn-off semiconductor switch.

10. Device according to one of claims 2 to 9, characterized in that the second switch arrangement (32) in the circuit arrangement (32, 33) for regulating the current strength at least one power transistor, at least one thyristor or at least one gate turn-off semiconductor switch having.

11. Device according to one of claims 3 to 6, 9 and 10, characterized in that the switching device (24, 25) for reversing the direction of flow of the current through the fluorescent lamp (20) each have at least one power transistor, at least one thyristor or at least one gate -turn-off semiconductor switch.

12. Device according to one of claims 1 to 11, characterized in that the control device (27) for controlling the first switch arrangement (26) with devices (44; 122) for changing the ratio of the time intervals (t1, t2) with conductive and is provided in the non-conductive state of the first switch arrangement (26) for the purpose of controlling the resulting average brightness of the fluorescent lamp (20).

13. The device according to claim 12, characterized in that the ratio of the time intervals (t1, t2) with conductive and non-conductive state of the first switch arrangement (26) is arbitrarily adjustable.

14. Device according to claim 12, characterized in that the devices for changing the ratio of the time intervals (t1, t2) are designed such that the time intervals (t1) with the conductive state of the first switch arrangement in each case in the starting phase which begins when the fluorescent lamp is switched on decrease automatically towards zero.