DE102007052669B4 - Ballast for a gas discharge lamp, e.g. an HID lamp - Google Patents

Abstract

Method for controlling the control input of a switching element (8, M1), in particular a transistor, wherein - the control input is galvanically decoupled by a transformer, and - switching the control input on or off by transformer transmission of current or voltage pulses from a control unit (15, V1) takes place as a clock signal source, the clock signal source outputting three different states, a current or voltage pulse of the first polarity being passed to the control input via a first control element (S1) and charging the control input, this charge also after the current or voltage pulse of the first polarity and thus the switching element (8, M1) remains switched on and a current or voltage pulse with reversed polarity discharges the charge at the control input via a second control element (S2) in such a way that the switching element (8, M1) is switched off.

Description

The invention relates to the control of a high-potential ("highside") transistor, which can be used, for example, in an operating device for lamps, for example a ballast for a gas discharge lamp or an LED, and more precisely, for example, in a switching regulator such as an inverter .

A ballast with a half-bridge circuit (inverter) is known. FETs, MOSFETs or IGBTs are used as switching elements of the inverter, in particular when high power levels are to be switched. These components are characterized not only by the fact that they switch with little power, but also by the fact that they switch very quickly. However, the components mentioned are provided with very high-impedance inputs (gates), which also require a driver circuit with a very high-impedance output. The driver circuit must even ensure that at least the high-potential (“high side”) switch element is galvanically isolated from ground. This is possible because a driver circuit with a transformer is used, the primary winding and secondary winding of which are only inductively coupled but are galvanically isolated.

The pamphlet US 56 68 712 A discloses a circuit for controlling a DC/AC converter in a power supply circuit for a discharge lamp of an automobile headlight. The pamphlet DE 40 09 267 A1 discloses a high frequency ballast. The pamphlet U.S. 55 46 300 A discloses a zero voltage switching control of a resonant mode converter and an electronic ballast with such control. The pamphlet DE 94 08 734 U1 discloses a high voltage power supply circuit for a gas discharge lamp. The pamphlet U.S. 5,798,619 A discloses techniques for controlling remote lamp loads. The pamphlet DE 44 25 679 A1 discloses a device for operating a high-pressure gas discharge lamp. The pamphlet EP 1 147 686 B1 discloses a flyback converter as an LED driver.

The invention is based on the object of proposing a technique for driving a switching element which is designed in particular for the conditions in ballasts with full-bridge or half-bridge inverters or another switching regulator.

According to the invention, this object is achieved by the features of the independent claims. The dependent claims develop the central idea of the invention in a particularly advantageous manner.

A first aspect of the invention relates to a method for driving the control input, for example a switching element which is at high potential, in particular a transistor, the control input being transformer-electrically decoupled. In this case, the control input is switched on or off by means of a transformer transmission of current or voltage pulses from a control unit as a clock signal source.

Switching on and off takes place, for example, by means of bipolar voltage pulses. These voltage pulses are switching pulses with which current pulses, for example, can be functionally equated.

The invention can be used both to control the control input of a switching element that is at high potential, for which galvanic isolation may be required, and for galvanically decoupled control of the control input of a switching element that is not at high potential, but whose control has one galvanic isolation required.

A voltage pulse of the first polarity charges the control input, for example via a first control element, this charge remains even after the voltage pulse of the first polarity and thus the switching element is switched on, and a voltage pulse with the opposite polarity discharges the charge at the control input via a second control element in such a way that the switching element is switched off. The term “circuit component” or “circuit element” can also be used instead of the term “control element”.

A diode, preferably a zener diode, can be used as the first control element.

A diode and/or a transistor can be used as the second control element.

The current or voltage pulses are preferably shorter than the switch-on or switch-off time of the activated switching element. In this case, sequences of several pulses are also possible in order to keep the switching element in a desired state.

As a clock signal source, the control unit can output voltage pulses in a predetermined cycle, regardless of whether the state of the switching element is to be changed.

In this case, the control unit can respond to a first voltage pulse in the event that the state of the switching element is to be changed emit a second pulse with the opposite polarity to the first voltage pulse.

If the state of the switching element is to be retained, the control unit can send a voltage pulse of the same polarity again in response to a voltage pulse of the first polarity.

The invention also relates to a control unit, in particular an integrated circuit such as an ASIC, which is designed to carry out a method according to any one of the preceding claims.

The invention further relates to a half-bridge or full-bridge inverter or other switching regulator with at least one switch. In this case, at least one switch is controlled by means of a method according to one of the preceding claims explained above.

A ballast for gas discharge lamps, in particular HID lamps, can have such an inverter.

• 1 zeigt ein Vorschaltgerät zum Betreiben einer Gasentladungslampe,
• 2 zeigt Signalverläufe in dem Vorschaltgerät,
• 3 bis 5 zeigen Ausgestaltungen von Transformatoren als Potentialtrennung eines Highside-Transistors,
• 6 bis 8 zeigen Ausführungsbeispiele von schematischen Schaltungen zur Ansteuerung eines Highside-Transistors,
• 9 zeigt Signalverläufe in der Schaltung gemäss 7b,
• 10 zeigt eine Schaltung zum Ansteuern einer LED, und
• 11 zeigt ein Vorschaltgerät zum Betreiben einer HID-Lampe.

An embodiment of the invention is described below with reference to the drawings.

• 1 shows a ballast for operating a gas discharge lamp,
• 2 shows signal curves in the ballast,
• 3 until 5 show configurations of transformers as potential isolation of a high-side transistor,
• 6 until 8th show exemplary embodiments of schematic circuits for controlling a high-side transistor,
• 9 shows signal curves in the circuit according to 7b ,
• 10 shows a circuit for driving an LED, and
• 11 shows a ballast for operating an HID lamp.

1 shows a ballast for operating a gas discharge lamp 2 in schematic form. The ballast consists of a load circuit 1, an inverter 7, a driver circuit 11, a control unit 15 and a supply voltage unit 16. The load circuit 1, as usual, contains a series resonant circuit consisting of a choke 4 and a charging capacitor 3. The connection point between the choke 4 and the charging capacitor 3 is connected to an electrode of the gas discharge lamp 2 via a coupling capacitor 5 . The other electrode of the gas discharge lamp 2 is grounded. The choke 4 consists of a core (not shown) and a choke winding D. The terminal of the choke 4 facing away from the charging capacitor 3 is also connected to a terminal of a damping capacitor (snubber cap), the other terminal of which is connected to ground.

The inverter 7 is a half-bridge circuit consisting of two series-connected MOSFETs 8, 9, which work as electronic switching elements. The series connection of the two MOSFETs 8, 9 is connected to a high DC voltage potential V _bus on the one hand and to ground on the other hand. Connection point 10 of the half-bridge leads to load circuit 1.

(The invention can of course also be applied to full-bridge circuits.)

The inputs (gates) of the two MOSFETs have a very high resistance to ground. This makes it necessary for the corresponding outputs of driver circuit 11 to also have very high impedance. In the case of the MOSFET 8 lying on top, there is even a requirement for complete electrical isolation from ground. For this purpose, the driver circuit 11 contains a transformer 14 with a core (not shown) and a primary winding T _p and a secondary winding T _s . One connection of the secondary winding T _s is connected to the input of the overhead driver 12 whose output is at the gate of the MOSFET 8 . The other connection of the secondary winding T _s is at the bridge point 10 of the half-bridge circuit forming the inverter 7 . The gate of the MOSFET 9 lying underneath is also controlled by a driver 13 which leads directly to the control unit 15 . The primary winding T _p of the transformer 14 is also supplied by the control unit 15 .

The control unit 15 supplies switching pulses (on/off) to the driver circuit 11. The resonance between the winding capacitance of the primary winding and its inductance is used here. When the switching pulse is "on", the winding capacitance is charged and the transformer is magnetized in one direction. The MOSFET 8 lying at the top is then switched to conduction, while the MOSFET 9 lying at the bottom is blocked. When the "off" switching pulse then follows, the transformer continues to drive the current and is remagnetized in the other direction. As a result, the MOSFET 8 located at the top is blocked and the MOSFET 9 located at the bottom is switched to conduction.

The snubber capacitor 6 in the load circuit enables the MOSFETs 8, 9 to be switched at zero potential and also acts as a filter against electromagnetic interference frequencies which are a consequence of the switching operations.

In 2(a) it can be seen how the two MOSFETs 8, 9 are alternately switched on.

2 B) shows the voltage vmp(t) at the bridge point 10 of the inverter 7. It can be seen that trapezoidal switching pulses occur.

2(c) shows the course of the current iL(t) through the inductor 4. It can be seen that the current is almost sinusoidal, which is an indication that almost no harmonics and thus interference are generated.

2(d) shows the course of the current through the damping capacitor 6. These are small current pulses which occur with changing polarity in the switching pauses of the MOSFETs 8, 9. In this context it should be noted that the 2(a) apparent switching pauses have zero potential. The actual invention is in the 3 until 5 shown.

The transformer 14 and the choke 4 have a common core 16, which consists of E-partial cores 17 and 18 placed against one another. The core 16 has two outer legs 19 and 20 and a middle leg 21. The middle leg 21 is interrupted by an air gap 24 and is thereby divided into two partial legs 21a, 21b. The three legs 19, 20, 21 are each bridged at their two ends by a transverse yoke 22, 23.

The primary winding T _p of transformer 14 is split, with one half wound on leg 19 and the other half wound on leg 20 . Likewise, the secondary winding T _s is divided, one partial winding also being wound onto the leg 19 and the other also being wound onto the leg 20 of the common core 16 . The sense of winding will be explained later.

The choke 4 consists of three choke windings D1, D2 and D3. These three choke windings can optionally be connected in series to change the inductance of the choke. In any case, all three choke windings D1, D2 and D3 are wound on the middle leg 21 of the common core 16.

Now, based on the 4 and 5 explained how the winding direction of the primary windings of the transformer 14 must be chosen so that the desired effect is achieved. In 4 only two windings are shown, the top one of which is one of the two transformer windings in 3 corresponds, and the lower of which one of the choke windings in 3 corresponds.

With reference to 4 the first step is to apply a voltage U _D to the winding on the middle leg, which generates a current i _D in the winding. The current i _D generates a magnetic flux ϕ _D in the middle limb, which is divided into two partial fluxes ϕ _D1 and ϕ _D2 , which flow back through the outer limb. The sense of winding of the two partial windings of the upper winding is selected in such a way that the current induced in the two partial windings by the two partial fluxes ϕ _D1 and ϕ _D2 leads to voltages _UR1 and _UR2 , which cancel each other out, so that at the input of the upper Winding U _R = 0 arises. So the result is: the upper winding is decoupled from the lower one. It is crucial that the partial windings are attached in such a way that the partial voltages resulting from the currents induced in them cancel out.

In 5 a voltage U _T is applied to the upper winding, which results in a current I _T . Due to the current, a flux ϕ _T is generated in the core, which only runs through the two outer legs of the core and the two transverse yokes, but not through the middle leg. The reason is that the middle leg has a much higher magnetic reluctance than the two outer legs due to its air gap. Due to the fact that no current flows through the middle leg, no current is induced in the lower winding sitting on this middle leg. This also decouples the lower winding from the upper winding.

That for both of them 4 and 5 stated principle is in 3 realized with two transformer windings and three choke windings. The sense of winding is the same as in the 4 and 5 . Here, too, complete decoupling between the transformer windings on the one hand and the choke windings on the other is guaranteed. However, the primary winding T _p and the secondary winding T _s of the transformer 14 are not decoupled - which should not be the case either. Because the transformer windings T _p and T _s and the choke windings D1, D2 and D3 are all on the same core 16, the amount of circuitry around the second core that is normally present has been reduced to a considerable extent.

The principle described above can also - with reference to 3 - be modified as follows: D ₁ and D ₂ can be transformer windings. T _p can be a choke winding. One thinks away D ₃ and T _s . Even with such a modified version, the transformer windings on the one hand and the choke winding on the other hand are decoupled from one another, while the two transformer windings are coupled to one another.

In 6 shows a circuit according to the invention for driving a transistor M1, in particular a MOSFET. The transistor M1 is preferably a transistor which is at a high potential (high side) and is supplied, for example, with a DC bus voltage of a ballast of the type explained above. Therefore, the gate of the transistor M1 should be driven in a floating manner. According to the 6 In the embodiment shown, a control unit is provided as a clock signal source V1, which supplies the primary-side winding of the transformer T1.

The control unit V1 is designed in such a way that it can preferably output bipolar pulses of +/-12V, for example.

These bipolar pulses are transmitted through transformer T1. A first control element S1 is provided on the secondary side, via which the switching element (transistor) M1 can be switched on selectively with a first pulse from the control unit V1. For this purpose, the control input (gate) of the transistor M1 can be charged, for example, with a first pulse via the control element S1. According to the invention, the switch then remains in this state (for example switched on), even when the pulse has decayed again and the signal 0V is therefore present on the primary side of the transistor T1.

Another pulse from the control unit V1 is used to switch off the switching element (transistor) M1. This can preferably be a pulse of reversed polarity (e.g. -12V). In particular, a further control element S2 can be provided on the secondary side of the transformer T1 in such a way that the second pulse causes the control input (gate) of the transistor M1 to be discharged to ground through the second control element S2.

A charge or energy store C1 (for example a capacitor) can optionally be connected in parallel with the switching element M1 (transistor).

7a shows a first circuit design of the general scheme of the circuit of 6 . The first control element, which is used to turn on the transistor by charging the control input (gate), is a diode D1.

To discharge and thus turn off the transistor M1, a circuit is provided which has a zener diode D3 and a transistor Q1.

In 8th shows a circuit in which the discharging of the control input of the transistor M1 is not effected by an active element (like the transistor Q1 according to the embodiment of 7a) takes place, but by another passive component, namely a zener diode D2.

In the 7b , which is essentially the 7a corresponds to various voltages indicated in the waveform diagram of FIG 9 are shown.

It can be seen there that the gate of transistor M1 is charged to a turn-on voltage V _Gate by a first pulse DRV+, which is also transformed as voltage V _T1 by the secondary winding of transformer T1.

This switched-on state of the transistor M1 is retained even with a second positive pulse DRV+.

According to this exemplary embodiment, switching off by discharging the gate is effected by a pulse of reverse polarity, namely with a voltage DRV-, which is transformed as a negative voltage pulse onto the secondary side of transistor T1 as voltage V_T1. The leading edge of this pulse then causes the control input of transistor M1 to be discharged via a passive or an active component.

According to the invention, three different states can be generated on the input side, namely a pulse with a first polarity, a pulse with a second, inverse polarity, and the '0V' state.

The first control element S1, which is preferably in the form of a diode D1 (normal diode or zener diode), conducts the positive pulse, i.e. the transition from, for example, 0V to +12V from the transformer T1 to the gate of the transistor M1.

On a transition from 0V to the negative polarity (e.g. -12V), the zener diode D3 turns on, which in turn turns on the transistor Q1, thereby discharging the gate of the transistor (e.g. field effect transistor) M1.

According to the invention, a clocked signal is thus provided by a source V1, with this clocked signal then being transmitted to the secondary side via the transformer T1. There, the control elements S1 and S2 are provided, through which the clock signal can optionally be routed to the gate of the MOSFET M1 in order to charge it, or the charged gate of the MOSFET M1 can be discharged again.

A particular advantage is that the operation shown can also take place with a very low pulse repetition frequency of, for example, less than 1 kHz, preferably less than 100 Hz, or even almost DC operation. When switch M1 is part of a full-bridge or half-bridge inverter circuit, this very low frequency operation of the inverter is particularly advantageous for HID lamp ballasts, for example.

According to the invention, see FIG 9 That after, for example, after switching on the transistor M1 by a pulse of the first polarity, this pulse is repeated, namely after a defined time interval (clocking of the signal source), which allows a 'refresh' the charge state of the gate of the transistor M1.

The control unit as a clock signal source can therefore be designed to output pulses in a predetermined clock, regardless of whether the state of the transistor M1 is to be changed. In the event that the state is to be changed, a second pulse with reversed polarity is sent out after a first pulse. In the event that the state is to be retained, a pulse of the first polarity is again sent a pulse of the same polarity, which confirms ('refreshes') the circuit state of the transistor M1.

The three different states that can be generated on the input side do not have to have a fixed, defined voltage level, but can lie in three different value ranges. A pulse with a first polarity can range from 10V to 15V, a pulse with a second, inverse polarity can range from -10V to -15V, and the state '0V' can range from 5V to -5V. A clock signal with discrete values is therefore not necessary, but an analog clock signal with a corresponding signal form in the three voltage ranges can also be used.

11 shows an operating circuit of an HID lamp with a full bridge, the two switching elements S1 and S3, which are at high potential, being controlled via a control according to the invention. (The general function of the circuit for operating HID lamps is in the EP1114571B1 described.)

10 shows a switching regulator for operating an LED, this is another possible application of the invention. The switching regulator is designed as a so-called step-down converter or buck converter and has a switching element controlled according to the invention.

A smoothed and rectified intermediate circuit voltage V1 (which can be provided by an upstream circuit unit that feeds a storage capacitor) is present as a result of a voltage supply.

After the application of a suitably dimensioned control voltage U_G to the gate of a first semiconductor power switch M1 implemented, for example, as a self-locking n-channel MOS field effect transistor, this electronically controllable switching stage becomes electrically conductive, so that a drain current begins to flow, which as a result of the acting as an energy store Storage inductor L10 increases continuously and flows through the light-emitting diodes D10 (LED) as a diode current I_D. The increase in this diode current I_D when the storage inductor L10 is charged is detected by a low-impedance measuring resistor R10, which is also arranged in the load circuit of the first power switch M10 and is connected to the ground node. When the diode current I_D has reached a certain value, the power switch M10 is opened.

The consequence of this is that the diode current I_D built up via the storage inductor L10 is derived from the inductive reactance XL10 formed by the storage inductor L10 and the low-impedance measuring resistor R10 through a freewheeling diode DF in the parallel branch to the series connection of the light-emitting diodes D10. With the help of this relatively simple circuitry measure, the first semiconductor power transistor M1 is not endangered by the induction voltage U_L10 dropping at the inductive reactance XL10 when the drain current I_D is switched off (when M10 is blocked), which can be a multiple of the operating voltage. The voltage U_R10 dropping across the low-impedance measuring resistor R10 serves to detect the diode current I_D flowing through the light-emitting diodes D10 in the free-running current path. The switching stage M10 remains locked until the current flow has fallen below a certain threshold. After the switching stage M10 has started to conduct again, the process described above is continued in a periodically recurring sequence. In the method according to the invention, both the charging and the discharging current I_L10 of the inductive reactance XL10 flow as a diode current I_D through the arrangement of the series-connected luminescence diodes D10 of the LED lighting module according to the invention, so that a triangular current that periodically oscillates around a mean value results through the LED .

By arranging the power switch M10 in front of the storage choke L10, the LED D10 can be arranged so that both the charging and the discharge current can also be measured via a measuring resistor to ground. Thus, a simple current measurement is possible for both currents. The arrangement of the circuit breaker M10 in front of the storage inductor L10, however, requires switching elements that are at a high potential to be driven. This task can be solved by using the control method according to the invention.

The current flow through the measuring resistor R10 is monitored by the control circuit 50, which transmits the control signal for switching the circuit breaker M10 on and off to the control circuit according to the invention for the switching element (M10) at high potential.

Application-Specific Integrated Circuits (ASICs) with a comparatively small space requirement for implementing the control unit 50 are also conceivable, the measured-value detection part of which does not have to have a high dielectric strength.

Claims (17)

Method for driving the control input of a switching element (8, M1), in particular a transistor, where - the control input is galvanically decoupled from a transformer, and - the control input is switched on or off by transformer transmission of current or voltage pulses from a control unit (15, V1) as a clock signal source, where the clock signal source outputs three different states, a current or voltage pulse of the first polarity being passed via a first control element (S1) to the control input and charging the control input, this charge also remains switched on after the current or voltage pulse of the first polarity and thus the switching element (8, M1) and a current - or voltage pulse with reverse polarity discharges the charge at the control input via a second control element (S2) in such a way that the switching element (8, M1) is switched off. procedure after claim 1 , whereby switching on and off is effected by bipolar current or voltage pulses. procedure after claim 1 or 2 , A passive component, preferably a diode, being used as the first control element (S1). Method according to one of the preceding claims, in which a passive component, preferably a diode, particularly preferably a zener diode and/or an active component, such as a transistor, are used as the second control element (S2). Method according to one of the preceding claims, in which the current or voltage pulses are shorter than the switch-on or switch-off time of the driven switching element. Method according to one of the preceding claims, in which the control unit (15, V1) as a clock signal source outputs current or voltage pulses in a predetermined cycle, regardless of whether the state of the switching element (8, M1) is to be changed. procedure after claim 6 , In the event that the state of the switching element (8, M1) is to be changed, the control unit (15, V1) generates a current or voltage pulse which has a reversed polarity compared to the previous pulse. procedure after claim 6 or 7 , wherein the control unit (15, V1) in the event that the state of the switching element (8, M1) is to be retained, sends a current or voltage pulse of the first polarity again a current or voltage pulse of the same polarity. Control unit (15, V1), in particular an integrated circuit, which is designed to carry out a method according to one of the preceding claims. Half-bridge or full-bridge inverter with at least two switches, one switch using a method according to one of Claims 1 until 8th is controlled. Ballast for gas discharge lamps, having an inverter claim 10 . Ballast for HID lamps, comprising an inverter claim 10 . Switching regulator with at least one switch as a switching element, the switch using a method according to one of Claims 1 until 8th is controlled. Flyback converter with at least one switch as a switching element, the switch using a method according to one of Claims 1 until 8th is controlled. Ballast for LED, having a flyback converter Claim 14 . Method for operating HID lamps with a full-bridge inverter with at least four switches, the control inputs of the two switching elements at high potential, in particular transistors, being activated by switching the control inputs on or off by transformer transmission of current or voltage pulses from a control unit as the clock signal source, and the control inputs being galvanically transformer-decoupled, and wherein the control inputs of the two switching elements that are at low potential, in particular of transistors, are not galvanically decoupled by transformers, wherein when the control input is driven, each switching element that is at high potential - the clock signal source outputs three different states, and - A current or voltage pulse of the first polarity is passed via a first control element to the control input and charges the control input, this charge remains switched on even after the current or voltage pulse of the first polarity and thus the switching element, which is at high potential, and a current or voltage pulse with reversed polarity, the charge at the control input is discharged via a second control element in such a way that the switching element, which is at a high potential, is switched off. procedure after Claim 16 , the clock signal pulses being shorter than the switch-on or switch-off time of the driven switching elements which are at high potential.

Images