EP1333707B1 - Electronic ballast for a discharge lamp - Google Patents

Abstract

An electronic starter for a gas discharge, preferably fluorescent, lamp (LA) has a dc source, inverter with half-bridges (Q1,2), a load of lamp and series resonant circuit. A control (1) operates on the switches during ignition to give a stepwise increase in switching time and monitors the load current, briefly opening the switches if a threshold is exceeded. An Independent claim is also given for a process of operation as above.

Description

The present invention relates to an electronic ballast for at least one gas discharge lamp according to the preamble of claim 1 and a method for operating such a device. In particular, the invention deals with the processes during the ignition of the gas discharge lamp.

The use of electronic ballasts for operating gas discharge lamps leads to significant energy savings compared to the use of conventional ballasts due to lower losses and improved lamp efficiency. The input of a typical electronic ballast is a connected to the power supply network high-frequency filter, which is connected to a rectifier circuit. The rectified by the rectifier circuit supply voltage is supplied to a smoothing circuit for generating an intermediate circuit voltage, a fed with the intermediate circuit voltage inverter finally generates a high-frequency AC voltage which is applied to the load circuit with the gas discharge lamp connected thereto. The operation with the high-frequency alternating voltage results in a reduction of the electrode losses and an increase in the luminous efficacy. It is also possible to control the lamps in a controlled and gentle manner.

In order to ignite the gas discharge lamp, the electrodes are preheated at first at an increased frequency of the inverter. At the end of this preheat period, the frequency generated by the inverter is then lowered so that it approaches the resonant frequency of the load circuit, which is primarily determined by a series resonant circuit in the load circuit, and as a result, the voltage applied to the lamp increases. Finally, at a certain point in time during the lowering of the frequency, the ignition of the gas discharge lamp takes place, with the result that its resistance drops sharply. However, an excessive current flow occurring as a result of this is avoided by virtue of the fact that the inductance of the series resonant circuit is designed as a choke and accordingly has a current-limiting effect.

If a strong current flows through the choke of the load circuit, then at a certain current value, the result is that the choke no longer acts linearly. This is certainly the case if it is not just a pure coil, but a coil with an additional ferromagnetic core. If the throttle is outside its linearity range, this means that it is more inductive Resistance reduced, which in turn can result in that relatively high current and voltage peaks occur, which may have damage or even destruction of the ballast result.

In order to avoid the above-described problems, so far chokes are used which are large enough not to get into a non-linear state under regular conditions. However, such chokes are also relatively expensive due to their larger volume. Furthermore, this limits the possibilities to reduce the overall dimensions of the ballast.

In order to avoid damage to the electronic ballast, it is also known, in the event that a saturation or non-linearity of the throttle is detected to turn off the inverter as soon as possible. The shutdown is then so fast that this is not enough to ensure a reliable ignition of the lamp. The procedure described above for igniting the lamp must then be repeated.

From US 4,060,751 a ballast for a gas discharge lamp is known, which has a connected to a DC voltage source inverter with two switches. Furthermore, the switching arrangement has a load circuit which contains the lamp and a series resonant circuit. The switches are connected to a current sensor. The switching arrangement operates as a resonant starter to provide a high voltage for the lamp ignition.

No. 5,925,990 describes an electronic ballast for a gas discharge lamp, the load circuit of which has a series resonant circuit in addition to the lamp. The ballast further includes a microprocessor-controlled inverter. For the accident, a voltage limiter is provided.

The present invention is therefore based on the object to bring about a reliable ignition of the lamp with a reduced inductance in their volume. The ignition current or the ignition voltage should remain within an acceptable range in order to prevent damage to the ballast can.

This object is achieved by an inventive electronic ballast having the features of claim 1. This is characterized in that a provided in the ballast control circuit, which increases the turn-on of the switches of the inverter during the ignition phase gradually and thus lowers the operating frequency, during the turn-on of one of the two switches, the height of the Monitor the current supplied to the load circuit and in the event that the current exceeds a predetermined reference value, the switch opens prematurely and then regularly turns on again.

The object is further achieved by a method according to the invention.

The premature opening of the switch has the consequence that the current switching period of the two switches of the inverter is shortened, which is equivalent to a higher instantaneous frequency. As a result, with a suitable choice of the reference value, the operation in the non-linear region of the throttle can be mastered.

However, since the switch is subsequently turned back on normally, i. In the subsequent switching periods, the previously set switch-on time for the two switches or their operating frequency is maintained, a controlled operation in the boundary region of the throttle of the load circuit and thus ultimately a safe ignition of the lamp is enabled over a longer period of time.

Further developments of the invention are the subject of the dependent claims.

Thus, the control circuit for monitoring the load circuit current preferably detects the voltage drop across a resistor arranged at the base of the half-bridge of the inverter and compares this with a reference voltage. The monitoring of the current then takes place during the switch-on of the lower switch of the half-bridge. In addition, after switching off one of the two switches and the subsequent switching on of the other switch, a predetermined delay time is provided to preclude a short circuit of the inverter. The switches are preferably MOS field effect transistors whose gates are controlled by the control circuit by means of pulse width modulated signals. In addition, the ballast may include an additional safety circuit that completely disables the inverter for a temporary period of time in the event of too late opening of the switch and thus uncontrollable voltage spikes. For this purpose, the control circuit additionally monitors the lamp voltage and compares it with a further predetermined limit value.

Fig. 1

ein erstes Ausführungsbeispiel einer Schaltungsanordnung für das erfindungsgemäße Vorschaltgerät;

Fig. 2a

ein Taktschema der beiden Schalter des Wechselrichters für den Fall, dass die erfindungsgemässe Regelung inaktiv, d.h. ein Referenzwert noch nicht überschritten ist;

Fig. 2b

ein Taktschema der beiden Schalter, für den Fall, dass die erfindungsgemässe Regelung aktiv, d.h. der Referenzwert bereits überschritten ist;

Fig. 3

ein Flussdiagramm des erfindungsgemäßen Verfahrens; und

Fig. 4

eine Erweiterung der in Fig. 1 dargestellten Schaltungsanordnung.

In the following the invention will be explained in more detail with reference to the accompanying drawings. Show it:

Fig. 1

a first embodiment of a circuit arrangement for the ballast according to the invention;

Fig. 2a

a timing diagram of the two switches of the inverter in the event that the inventive control inactive, ie a reference value is not exceeded;

Fig. 2b

a timing diagram of the two switches, in the event that the inventive control active, ie the reference value is already exceeded;

Fig. 3

a flow chart of the method according to the invention; and

Fig. 4

an extension of the circuit arrangement shown in Fig. 1.

The essential components of the present invention are shown in the diagram in FIG. The arranged before the inverter further elements of the ballast - for example, the rectifier and the smoothing circuit - are already well known and will therefore not be explained in detail below.

The inverter is formed by a half-bridge of two series-connected electronic switches Q1 and Q2. These switches Q1, Q2 can be formed, for example, by two MOS field-effect transistors. The base of the half-bridge is connected via a shunt resistor R1 to ground, while at the input of the half-bridge, the DC voltage U _BUS is applied, which can be generated for example by the formation of the mains voltage by a combination of radio interference suppression and rectifier. Alternatively, however, any other DC voltage source may be applied to the half-bridge.

At the common node of the two switches Q1 and Q2 of the gas discharge lamp LA, which is preferably a fluorescent lamp, connected load circuit is connected. This first consists of a series resonant circuit, which is composed of a choke L1 and a resonance capacitor C1. At the connection point between the inductor L1 and the resonance capacitor C1, a series circuit of a coupling capacitor C2 and the gas discharge lamp LA is further connected in such a way that it is parallel to the resonance capacitor C1.

The two switches Q1, Q2 of the inverter are driven by a control circuit 1, which transmits control signals to the gates of the two field effect transistors Q1 and Q2. A clocking scheme for the two switches Q1, Q2 which normally occurs during the activation is shown in FIG. 2a.

According to this illustration, a switching period begins with the switching on and off of the upper switch Q1 of the half-bridge for a specific switch-on time t _on , the term "closing" when using field-effect transistors being understood as being switched through. At the end of this switch-on time t _on , the switch Q1 is opened again or the transistor is switched off and the switch Q2 is closed alternately. Between the opening of the switch Q1 and the subsequent closing of the switch Q2, a predetermined delay time t _{d is} waited for to simultaneously close the two switches Q1, Q2 and thus avoid a short circuit of the inverter in any case. The second switch Q2 is also closed for the switch-on time t _on and then opened again. After further waiting for the delay time t _d , the upper switch Q1 is closed again, thus ending a complete switching period.

The total time T _{p of} a period is thus: $$T_p=2\cdot \left(t_{\mathit{on}}+t_d\right)$$

The frequency of the inverter is calculated accordingly to: $$f=1/T_p=\frac{1}{2\cdot \left(t_{\mathit{on}}+t_d\right)}$$

During a start phase of the ballast, the electrodes of the lamp LA are first preheated, which is done by applying an alternating voltage to the load circuit at a frequency which is significantly above the resonant frequency of the load circuit. The resulting voltage is then too low to cause ignition of the lamp LA.

At the end of the preheating time the ignition of the lamp LA is initiated, which is done by gradually increasing the on-time t _on for the two switches Q1, Q2 of the inverter and accordingly reducing the operating frequency of the inverter. The frequency then approaches ever closer to the resonant frequency of the load circuit until the resulting voltage is so large that it causes an ignition of the gas discharge lamp LA. As already explained at the beginning, however, a high current can already occur in the load circuit before the ignition of the lamp LA, which has the consequence that the inductor L1 no longer acts linearly and consequently high current and voltage peaks occur.

In order to be able to utilize this non-linear region, the control circuit 1 monitors the magnitude of the current supplied to the load circuit during the switch-on phase of the lower switch Q2. For this purpose, it detects the voltage drop across the shunt resistor R1 and _supplies it to a comparator 2, which compares it with a reference _voltage V _ref1 . This voltage V _ref1 is chosen such that the maximum current on the one hand sufficient to ensure a reliable ignition, and on the other hand to high current loads are avoided, especially when the throttle is operated in saturation.

The reaction of the control circuit 1 to a state in which the voltage drop across the resistor R1 _exceeds the reference value V _ref1, that is, the current supplied to the load circuit exceeds the maximum current of the inductor L1, is shown in FIG. 2b. The first half of the switching period corresponds to the clock scheme shown in Fig. 2a for the unregulated case, since in the present example, the monitoring described below takes place only during the switch-on of the lower switch Q2. First, therefore, the upper switch Q1 for the current set by the control circuit 1 on-time t _on, and then closed after the opening of the switch Q1 and waiting for the delay time t _d of the lower switch Q2 is closed.

After opening the first switch Q1, the voltage U _R1 dropping across the shunt resistor _{R1 increases} continuously, as shown in the upper area of FIG. This is monitored by the control circuit 1. As soon as the measuring voltage U _R1 exceeds the reference voltage V _ref1 , a pulse P occurs at the output of the comparator 2, which causes the control circuit 1 to switch off the switch Q2 after a system-related reaction time t _r . The turn-on time t ' _{on of} the lower switch Q2 is thus reduced compared to the turn-on time t _{on of} the upper switch Q1. After waiting for the delay time t _d , the first switch Q 1 is then again normal, ie switched _on and off again for the previously set switch-on time t _on .

der Schaltperiode reduziert und damit die Frequenz $$f^{\prime }=\frac{1}{\left(2\cdot t_d+t_{\mathit{on}}+{\mathit{t}}_{\mathit{on}}^{\mathit{\prime }}\right)}$$

des Wechselrichters kurzfristig erhöht. Hierdurch wird einem unkontrollierten Betrieb im Sättigungsbereich der Drossel L1 und den sich daraus ergebenden negativen Folgen für die Schaltung entgegenwirkt. Die von der Steuerschaltung 1 zuvor eingestellte Einschaltzeit t_on wird allerdings für die darauffolgenden Schaltperioden beibehalten, wobei der untere Schalter Q2 wiederholt regulär ein aber ggf. in der oben beschriebenen Weise vorzeitig abgeschaltet wird. Die Schaltung wird damit nahe der Grenze zum nicht-linearen Bereich der Drossel L1 betrieben und zwar über einen längeren Zeitraum hinweg, so dass letztendlich eine sichere Zündung der Lampe LA ermöglicht wird.Due to the premature opening and the resulting shortening of the switch-on time t ' _on for the switch Q2, the total duration thus becomes $${\mathit{T}}_{\mathit{p}}^{\text{'}}=2\cdot t_d+t_{\mathit{on}}+{\mathit{t}}_{\mathit{on}}^{\text{'}}$$

reduces the switching period and thus the frequency $$f^{\text{'}}=\frac{1}{\left(2\cdot t_d+t_{\mathit{on}}+{\mathit{t}}_{\mathit{on}}^{\mathit{\text{'}}}\right)}$$

of the inverter increased at short notice. This counteracts an uncontrolled operation in the saturation region of the inductor L1 and the resulting negative consequences for the circuit. However, the on-time t _on previously set by the control circuit 1 is maintained for the subsequent switching periods, with the lower switch Q 2 being repeatedly switched off regularly but possibly prematurely in the manner described above. The circuit is thus operated near the boundary to the non-linear region of the inductor L1 and via a longer period of time, so that ultimately a safe ignition of the lamp LA is possible.

The switching behavior according to the invention just described is summarized in the flowchart in FIG. The first two steps 100 and 101 of the method consist - as described above - in that the upper switch Q1 is switched on and normally switched off again after the switch-on time has elapsed. After waiting for the delay time t _d in step 102, the lower switch Q2 is turned on in step 103. During the turn-on time of the switch Q2, it is continuously checked whether the reference _voltage V _{ref1 has been} reached (step 104). If this is not the case, then an internal count R of the control circuit 1 is increased by (step 105) and the switch Q2 is switched off regularly (step 106).

wobei T_cl der Dauer des Basistaktes eines internen Zeitgebers der Steuerschaltung 1 entspricht.Increasing the count R in step 105 results in increasing the on-time t _on for the two switches Q1 and Q2. This is because the on-time is derived from the count value R by, for example, counting an internal counter of the control circuit 1 up to the value of R. Thus: $${\mathrm{t}}_{\mathrm{on}}=\mathrm{R}\cdot {\mathrm{T}}_{\mathrm{cl}}$$

where T _cl corresponds to the duration of the base clock of an internal timer of the control circuit 1.

After the switch Q2 and the waiting for the delay time t _d is switched off in step 108 is a cycle completely through and the switch Q1 is turned on again, but now for a slightly extended on-time.

However, if the query in step 104 shows that the reference _voltage V _{ref1 has been} reached, the switch Q2 is immediately switched off (step 107) and the delay time in step 108 is waited before the switch Q1 is switched on again. In this case, therefore, the count value R and thus the duty cycle for the two switches Q1 and Q2 is no longer increased, which - as explained above - has the consequence that a controlled operation of the throttle is realized.

In addition to the program sequence just described, it should be noted that the count R does not have to be increased on each pass. For example, increasing the count value R can also occur only every third or fifth pass, depending on how fast the frequency of the inverter is to be reduced during the ignition operation.

Although the method according to the invention makes possible reliable ignition of the lamp LA even when using a smaller throttle, it can not be completely ruled out that due to unforeseeable circumstances temporarily increased voltage peaks occur in the load circuit. In Fig. 4, therefore, an extended variant of the circuit according to the invention is shown, in which there is an additional monitoring of the lamp voltage. The voltage dropping at the lamp LA is here supplied to a further comparator 3, which compares it with a further reference value V _ref2 . If the lamp voltage _exceeds this second reference value V _ref2 , the control circuit 1 completely shuts off the inverter for a short period of time in order to avoid damaging the entire circuit arrangement.

By combining the monitoring of load circuit current and lamp voltage can thus be obtained even increased reliability.

Claims (12)

1. Electronic ballast for at least one gas discharge lamp, preferably for a fluorescent lamp (LA), having an inverter, connected with a d.c. voltage source (U_BUS), and a load circuit, which has the lamp (LA) and a series resonance circuit, connected to the inverter,
   wherein the inverter is formed by means of two switches (Q1, Q2) arranged in a half-bridge arrangement, which switches are so alternatingly controllable by a control circuit (1) during an ignition phase that their switch-on time (T_on) is increased step-wise to reduce the control frequency,
   characterized in that,
   the control circuit (1) in the ignition phase, during the switch-on time of one of the two switches (Q1, Q2), monitors the level of the current delivered to the load circuit and in the event that the current exceeds a predetermined reference value, prematurely opens the switch (Q1, Q2) and in the next period of the alternating control of the switches (Q1, Q2) again switches on the switch in regular manner.
2. Electronic ballast according to claim 1,
   characterized in that,
   the predetermined reference value is so set that a choke (L1) contained in the series resonance circuit is operated in saturation.
3. Electronic ballast according to claim 1 or 2,
   characterized in that,
   the control circuit (1), for monitoring the current, detects the voltage (U_R1) dropping at a resistance (R1) arranged at the foot at the half-bridge, and compares this voltage with a reference voltage (V_ref1).
4. Electronic ballast according to claim 3,
   characterized in that,
   the control circuit (1) monitors the level of the current delivered to the load circuit during the switch-on time of the lower switch (Q2) of the half-bridge.
5. Electronic ballast according to any of claim 1 to 4,
   characterized in that,
   the control circuit (1) waits for a predetermined delay time (t_d) between the switching-off of one of the two switches (Q1, Q2) and the subsequent switching-on of the other switch (Q1, Q2).
6. Electronic ballast according to any preceding claim,
   characterized in that,
   the two switches are field-effect transistors (Q1, Q2).
7. Electronic ballast according to any preceding claim,
   characterized in that,
   the control circuit (1) controls the two switches (Q1, Q2) by means of pulse signals.
8. Electronic ballast according to any preceding claim,
   characterized in that,
   the control circuit (1) further monitors also the voltage applied at the lamp (LA), and in the event that this exceeds a predetermined further reference value (V_ref2) temporarily switches off the inverter.
9. Electronic ballast according to any preceding claim,
   characterized in that,
   the control circuit (1), before the ignition phase, controls the load circuit in a pre-heating operation for heating the lamp electrodes.
10. Electronic ballast according to any preceding claim,
    characterized in that,
    the control circuit (1) only increases the switch-on time (t_on) when the switches (Q1, Q2) have carried out the switching scheme a plurality of times.
11. Method for operating an electronic ballast for at least one gas discharge lamp, preferably for a fluorescent lamp (LA), having an inverter, connected with a d.c. voltage source (U_BUS), and a load circuit, which has the lamp (LA) and a series resonance circuit, connected to the inverter,
    wherein the inverter is formed by means of two switches (Q1, Q2) arranged in a half-bridge arrangement, which switches are so alternatingly controlled during an ignition phase that their switch-on time (t_on) is increased step-wise to reduce the control frequency,
    characterized in that,
    in the ignition phase, during the switch-on time of one of the two switches (Q1, Q2), the level of the current delivered to the load circuit is monitored and in the event that the current exceeds a predetermined reference value, the switch (Q1, Q2) prematurely opens and, in the next period of the alternating control of the switches (Q1, Q2), the switch is again switched on in regular manner.
12. Method according to claim 11,
    characterized in that,
    a choke (L1) contained in the series resonance circuit is operated in saturation by means of appropriate selection of the reference value.

Images