WO2010022749A1

WO2010022749A1 - Plasma power supply

Info

Publication number: WO2010022749A1
Application number: PCT/EP2008/006993
Authority: WO
Inventors: Thomas Kirchmeier
Original assignee: Hüttinger Elektronik Gmbh + Co. Kg
Priority date: 2008-08-27
Filing date: 2008-08-27
Publication date: 2010-03-04
Also published as: EP2319167A1; CN102132481B; CN102132481A

Abstract

In the case of an RF power amplifier arrangement (11) having a DC electrical power supply and having a switching loop (12) which is connected thereto and contains two switching elements (S1, S2), which are at least indirectly connected in series, and whose centre point (M) forms the switching link output, wherein the first switching element (S1) is controlled via an active control signal generator, an auxiliary circuit is provided in order to produce a control signal for the second switching element (S2) as a function of a change in the operating state of the first switching element (S1).

Description

Plasma power supply

The invention relates to a plasma power supply for generating an RF signal having a frequency> 3 MHz and a power> 500 W, comprising an RF power amplifier arrangement with a DC power supply and a switching bridge connected thereto, which contains two switching elements connected at least indirectly in series and their center forms the switching bridge output, wherein the first switching element is driven via an active drive signal generator.

Power amplifiers for exciting plasma processes (eg for RF sputtering, etching or for exciting gas lasers) in a frequency range from 1 to 50 MHz, in particular at the industrial frequencies of 13.56, 27.12 and 40.68 MHz, are generally known. Such power amplifiers are available in different power classes from approx. 1 kW up to several 100 kW. For the smaller powers in the range of 1 to 20 kW, amplifiers based on semiconductor modules (semiconductor amplifiers) are preferably used. For larger power tube amplifiers are often used. The tube amplifiers have one Amplifier tube, which in turn is driven by a power amplifier, which in turn is based on semiconductor modules, that corresponds to the amplifiers at low power. Since tubes require more space, it is desirable to also build larger power amplifiers from amplifiers based on semiconductor modules. For this purpose, semiconductor amplifiers of lower power are interconnected with suitable power couplers.

An important aspect of the power amplifiers based on semiconductor modules is the minimization of the power loss, especially in the semiconductor modules themselves. Therefore, in order to excite plasma processes in particular RF generators are used with such power amplifiers, which have one or more switching elements and class D - or class E operation are operated.

For class D operation usually two series switching elements, z. As MOSFETs used. This circuit arrangement is referred to as a switching bridge or half bridge. MOSFETs have two power connections (drain and source) and a control connection (gate). The driving of the MOSFETs takes place via the gate-source voltage, wherein the source connection can be regarded as the voltage reference point of this component. Preferably, n-channel MOSFETs are used because they can be turned on and off faster and generate less power dissipation than p-channel MOSFETs.

The usual circuit is as follows: The upper MOSFET (High Side Switch HSS) is connected with its first power connection (drain connection in n-channel MOSFETs) to the positive DC supply voltage and with its second power connection (source connection at n-channel MOSFETs) to the drain of the lower MOSFET (Low Side Switch LSS, usually an n-channel MOSFET). The source terminal of the lower MOSFET is connected to the negative DC supply voltage. The output signal of the half-bridge is tapped between the two switching elements (MOSFETs) (midpoint of the switching bridge). Both MOSFETs are controlled via their gate connection (control connection). The source terminal of the LSS is at the quiet negative potential of the DC supply voltage, while when using an n-channel MOSFET for the HSS whose source terminal carries the RF output signal. Therefore, the control voltage (gate-source voltage) of the HSS must be relative to the RF output signal with its fast changing (floating) high potential difference, which in a drive signal generator, which is at a quiet potential, a complicated potential separation in the transmission of Control voltage required and by the small compared with the RF output voltage control voltage carries the risk of incorrect control of the HSS.

In the case of voltage-fed half-bridges, it is essential that both switching elements are never in the conductive state at the same time, otherwise the supply voltage would be short-circuited. This is becoming increasingly difficult as the frequency increases. Time delays when switching on and off the switching elements arise here mainly from the inductive effect of the leads and transhipment of the internal capacitances of the switching elements, which can take different lengths of the individual switching elements beyond.

In the Class E amplifier, the HSS is replaced by a choke, which provides a DC current flow to the output of the circuit (corresponding to the midpoint of the class D amplifier). A closed switching element (equivalent to LSS in the class D amplifier) causes a low voltage at the output of the circuit, causing current to flow through the inductor; this current flow is maintained after opening of the switching element by the self-inductance of the inductor, which generates a high voltage at the center.

Class D operation has the advantage over class E operation that the voltages on the switching elements formed as transistors are limited to the DC supply voltage, while in class E operation the reverse voltages on the transistors are limited to the DC supply voltage Three times the supply voltages can increase. However, class D operation has the disadvantage over class E operation of requiring very precise synchronization of the cooperative switching elements, which becomes increasingly difficult as the switching frequency increases, and that the reference point for driving the HSS relative to a rapidly changing potential must occur.

Examples of class D and class E amplifiers can be found in US Pat. No. 7,180,758 B2.

It is therefore an object of the present invention to provide a circuit arrangement which reduces or avoids the mentioned disadvantages of Class D or Class E operation and combines the advantages.

This object is achieved according to the invention by a plasma power supply of the type mentioned, wherein an auxiliary circuit for generating a drive signal for the second switching element is provided as a result of a change in operating state of the first switching element. The terms "switching element" and "switching element" are synonymous and interchangeable. According to the invention, the second switching element is activated automatically, passively, not actively, ie without one, or independently of a device control. Compared to the conventional class E amplifier, no large voltage surges occur. Since no active control (and thus no active drive signal generator) of the second switching element is necessary, components can be saved, so that the arrangement can be constructed more cost-effectively. In addition, neither a driver nor a driver control is necessary for the second switching element. Problems that occur in the prior art because the driver or the driver drive are at a floating potential of the second switching element are avoided. The inventive arrangement continues to be an automatic synchronization of the two switching elements. As a result, in contrast to the classic class D method, there is no risk that both switching elements switch on at the same time and a short circuit occurs. In summary, like a class E amplifier, the RF power amplifier arrangement includes only one actively driven switching element but, like a class D amplifier, essentially limits its operating voltage to the supply voltage. The first and / or second switching element may be formed as a bipolar transistor, field effect transistor, in particular MOSFET, or IGBT. Particular advantages are the design as MOSFET ₁ because these components are inexpensive and have a high efficiency.

The auxiliary circuit may have a signal processing device which is connected directly or indirectly to the midpoint of the switching bridge. By means of the signal processing device, the drive signal for the second switching element can be generated on the basis of signals which are related to the current or the voltage in the center. The signal processing device need not be controlled by an external controller or an external signal generator.

In a particularly simple embodiment, the signal processing device can have only one conductor section, wherein the conductor section can connect the control connection of the second switching element to the center.

The auxiliary circuit may include a capacitor connected between the second switching element and the midpoint of the switching bridge or the second switching element and a potential, in particular the positive potential, of the DC power supply. At the end of a conducting phase of the first switching element, this blocks and initiates a swinging of the output voltage (voltage at the center). As soon as the output voltage rises above the positive DC power supply voltage, the body diode of the second switching element becomes conductive. The reverse current through the body diode of the second switching element now charges the capacitor in series therewith. The voltage of this capacitor is also the gate drive voltage of the second switching element, so that this is now conductive and remains conductive until after Stromkommutierung the capacitor is discharged back to less than the threshold voltage required for switching. The second switching element is thus self-switching and thus automatically synchronized with the first switching element. The capacitance of the capacitor may be at least twice as large, in particular at least five times as large, preferably at least ten times as large as the capacitance between the power terminals (drain and source) of the first switching element. An excessive voltage at the center is thereby reliably derived by the second switching element to the positive connection of the DC power supply.

It can be provided in parallel with the second switching element, a throttle. In particular, a DC current flow is ensured even when the second switching element is switched off. The throttle may be configured such that the current through the inductor changes less than 20% during a period of the fundamental frequency. As a result, a switching behavior similar to the class E principle can be realized. The charge of the induction current of the reactor may be stored in a capacitor arranged in series with the second switching element.

The auxiliary circuit may comprise a coil or a coil section, one end of which is connected to the center and the other end to the signal processing device. By such a coil, when the first switching element is switched off, a voltage can be generated which is sufficient to switch on the second switching element. The coil may be provided as a separate coil in series with a primary winding of an output transformer and be magnetically coupled thereto or be part of a primary winding of a Ausgangsübertragers as coil section, wherein the center is connected to a tap of the primary winding. This means that the primary winding of the output transformer is extended compared to the conventional design and has a tap.

Alternatively, the coil may be part of a throttle as a coil section, which is connected in parallel to the second switching element, wherein the center point is connected to a tap of the throttle. However, as well as being arranged in series with a primary winding of the output transformer, the coil could be arranged as a separate coil in series with and magnetically coupled to the inductor. The signal processing device may comprise a filter, in particular a high, low or bandpass filter, or a resonant circuit. Thereby, the waveform of the drive signal of the second switching element can be adjusted. The signal processing device may also contain active components. However, it does not need to be connected to a controller, in particular an active signal generator, which specifies a signal.

Alternatively or additionally, the signal processing device may further comprise a voltage divider and / or an attenuator. The voltage divider can be realized by an amplifier arrangement having a gain factor <1. An attenuator, such as a damped series resonant circuit, may be used to shape the drive signal of the second switching element.

The signal processing device may comprise one or more amplifiers. In particular, input signals of the signal processing device can be amplified.

If the signal processing device has a transformer, galvanic isolation can take place in the signal processing device.

The signal processing device may be connected to a plurality of elements of the auxiliary circuit. This means that the drive signal for the second switching element is generated in consideration of a plurality of signals of the auxiliary circuit. The signal processing device can therefore comprise a plurality of inputs, for example for a measured value of the midpoint voltage or a current or a voltage which are applied to one end of a coil.

A capacitor may be connected in parallel with the first switching element. By this capacitor, the switching of the second switching element can be delayed after switching off the first switching element.

It can be provided an output network which is connected to the center and at its output terminal an RF output signal can be tapped. Through the output network, current and voltage can be shifted relative to one another in order to obtain a suitable voltage at the center point relative to the source connection of the second switching element, so that a reliable switching on and off of the second switching element is ensured.

The output network can be tuned to the fundamental frequency. This means that a signal which has the fundamental frequency and which is desired at the output is allowed to pass through. Other frequencies, in particular harmonics of the fundamental frequency, are filtered out. The output network can also deliberately detuned from the fundamental frequency, so be tuned to one of the fundamental frequency slightly different frequency. This achieves certain waveforms and time intervals of the voltage at the midpoint of the jumper and the current through the output network.

The scope of the invention also includes a method for operating a plasma power supply comprising an RF power amplifier arrangement with a DC power supply and a switching bridge connected thereto, which contains two switching elements connected at least indirectly in series and whose center forms the switching bridge output, wherein the first switching element is controlled via an active drive signal generator. A drive signal for the second switching element is generated as a result of an operating state change of the first switching element. As a result, the operating voltage of the first switching element can be essentially limited to the supply voltage. The first switching element is thereby protected. There is no active control of the second switching element necessary. It can be saved components. The circuit works safely and reliably.

As a drive signal for the second switching element, a positive voltage relative to the midpoint potential of the switching bridge can be generated. Thereby, the second switching element can be turned on. The drive signal can be generated without active signal generator.

The second switching element can be controlled in such a way that it initially passes current flow of charge in a first direction and current flow in reversed direction locks only when a part of the flow in the first direction has drained in the reverse direction.

The voltage profile and current profile at the first switching element can be shifted relative to one another such that the voltage across the first switching element at the time of its switching on is <30%, preferably <20% of the potential difference between the potentials of the DC power supply. As a result, low switching losses can be realized.

A choke dimensioned such that the current through the choke changes by less than 20% during a period of the fundamental frequency may be used in parallel with the second switching element. As a result, a constant power supply of the first switching element is ensured.

Further features and advantages of the invention will become apparent from the following description of embodiments of the invention, with reference to the figures of the drawing, which show details essential to the invention, and from the claims. The individual features can be realized individually for themselves or for several in any combination in a variant of the invention.

Preferred embodiments of the invention are shown schematically in the drawing and are explained below with reference to the figures of the drawing. Show it:

Fig. 1 is a schematic representation of an RF power amplifier arrangement;

2 is a first more detailed representation of an RF

Power amplifier arrangement;

3 shows an alternative embodiment of an RF power amplifier arrangement;

4 shows a further embodiment of an RF power amplifier arrangement;

Fig. 5 shows a modification of the embodiment of Fig. 4; 6a-6g different possible embodiments of a signal processing device;

7a shows the voltage curve across the switching element of a class E

amplifier;

Fig. 7b, the voltage waveform across the lower switching element of a

Jumper operated in class D operation;

Fig. 7c, the voltage waveform across the first switching element of the RF power amplifier arrangement according to the invention.

1 shows a schematic representation of an HF power generator 10 having an RF power amplifier arrangement 11, which has a switching bridge 12. The switching bridge 12 comprises the two series-connected switching elements S1 and S2. The switching bridge 12 is connected to both the positive potential 13 and the negative potential 14 of a DC power supply. M is the center of the series circuit of the switching elements S1 and S2. The center M represents the output terminal of the switching bridge 12. Connected to the center M is an output network 15, which in the exemplary embodiment has an output transformer 16 to which a plasma load 17 is connected. The output network 15 may have in addition to the output transformer 16 series capacitors, series inductors, resonant circuits, taps of the output transformer, etc. Output networks 15 without output transformer 16 are conceivable. In addition, the output network 15 may include an autotransformer.

The plasma load 17 is shown only as impedance. However, the load may also include an impedance matching network. At the point X, another switching bridge may be connected. Alternatively, the point X may be connected to ground AC-connected, as in the embodiment shown by the capacitors 18.1, 18.2. The first switching element S1 is driven by a driver 19, which in turn is connected to an active Ansteuersignalgenerator, not shown, for example, a device control. The drive signal of the switching element S1 is therefore actively generated by a signal generator. The drive signal of the second switching element S2 is generated by a signal processing device 20. With the signal processing device 20 different components 21, 22, 23 are connected. These components and the component 26 together with the signal processing device 20 form an auxiliary circuit.

In order to switch on the second switching element S2, the signal processing device 20 together with the components 21, 22, 23 must be so-that the potential at the control terminal 24 of the switching element S2 is above the potential at the (power) terminal 25 of the second switching element S2, which thus also the reference potential of the signal processing device 20 is. Except for voltage drops that can be generated in the device 21, this potential is essentially the potential of the midpoint M. The activation of the second switching element S2 by the signal processing device 20 is passive, i. without signal from the device control.

In the embodiment in FIG. 2, the signal processing device 20 is designed merely as a line section 30 which connects the center M to the control connection 24 of the second switching element S2. The component 21 is designed as a capacitor and the component 23 is designed as a throttle. The output network 15 comprises a series resonant circuit. The capacitor 31 in parallel to the first switching element S1 is optional. Therefore, the connecting lines are shown in dashed lines.

In the embodiment of Figure 2, the RF power amplifier assembly 11 is formed as a class E amplifier with additional switching element S2. By designed as a throttle device 23, the current flow to the center M is held approximately constant. When the first switching element S1 is turned off, the induced current flow through the device 23 causes an increase in the potential at the center M over the positive potential 13 of the DC power supply. The through the device 21 and the second Switching element S2 flowing current (initially, current flows only through the body diode or the parasitic capacitance of the switching element S2) causes the control terminal 24 and the potential at the control terminal 24 is more positive than the potential at the terminal 25. The second switching element S2 begins to lead. The charge originating from the current flow through the component 23 is stored in the component 21 or strengthens the current flow through the output network 15 after the current direction reversal. The component 21 and the conductive switching element S2 limit the voltage at the center M to a value which is only slightly above the positive potential 13 of the DC power supply. With the capacitor (component 21), the component 23 and the series resonant circuit in the output network 15, the waveform of the voltage curve in the center M can be controlled. The optional capacitor 31 in parallel with the switching element S1 delays the switching on of the second switching element S2 after switching off the first switching element S1 and otherwise also influences the waveform.

In the embodiment of an RF power amplifier arrangement shown in FIG. 3, the output network 15 has an output transformer 16, which comprises a primary coil 35. The primary coil 35 is extended by a coil section 36 compared to the conventional embodiment. The center M is connected to the one end of the coil portion 36. This means that the primary winding 35 has a tap 37. The other end of the coil section 36 is connected to the signal processing device 20, which generates a drive signal for the second switching element S2. The coil section 36 is thus connected both to the center M and to the signal processing device 20, and the signal processing device 20 is thereby indirectly connected to the center M.

When the switching element S1 is turned off, the voltage induced in the primary winding 35 causes a voltage increase at M. At the end of the coil section 36, which is connected to the signal processing device 20, the voltage overshoot is even higher. This voltage overshoot is applied via the signal processing device 20 to the control terminal 24 of the switching element S2. The potential at the control terminal 24 is higher than the potential at the center M, at which the terminal 25 of the switching element S2 is located. As a result, the switching element S2 turns on.

In this embodiment, the components 21 and 23 are missing. This means that essentially a class D amplifier is realized. The waveform of the voltage in the center M can be adjusted by further impedances in the output network 15.

For the first switching on of the switching element S1, a choke between the midpoint M and the positive potential 13 of the DC power supply can be provided, which, however, has no influence on the further operation.

The voltage overshoot that is caused by the coil section 36 may also be provided in the embodiment according to FIG.

In the embodiment of Figure 4, in turn, designed as a throttle device 23 is provided, wherein a tap 40 defines a coil portion 41 through which in turn takes place a voltage increase. When the switching element S1 is switched off, an excessively high voltage is generated by the current flowing through the component 23 and in particular by its extension through the coil section 41, which voltage is applied to the control terminal 24 via the signal processing device 20. Thereby, the potential at the control terminal 24 is raised above the potential at the center M, which is the reference potential of the switching element S2, so that the switching element S2 turns on. In this embodiment, the device 26 is formed as a capacitor. Instead of the component 26, however, it is also possible for a component 21, which is designed as a capacitor, to be provided between the switching element S2 and the center M (FIG. 5).

If only the device 26 is provided, the switching elements S1 and S2 are directly connected to each other, if instead of the device 26, the device 21 between the switching element S1 and the switching Element S2 is provided, the switching elements S1, S2 indirectly connected to each other.

In addition, it can be seen in FIG. 5 that the signal processing device 20 is not only supplied with the signal present at the end of the coil section 41, but also connected to the connection 25 of the switching element S2, so that the signal processing device 20 has the reference potential of the switching element Elements S2 is provided. The signal processing device 20 is indirectly connected to the center M both via the coil section 41 and via the component 21. Taking into account these signals, a drive signal for the switching element S2 is determined in the signal processing device 20.

In addition, it could also be provided that the primary coil 35 is extended by a coil section in order to continue to contribute to the voltage increase.

Different embodiments of the signal processing device 20 are shown in FIGS. 6a to 6g. The connections on the left side represent the input connections of the signal processing device 20 and the right connection represent the output connection which is connected to the control connection 24. With this signal processing device 20 it is possible to influence the amplitude, curve shape and time behavior of the activation signal for S2.

In FIG. 6 a, the signal processing device 20 is formed only by a conductor section 30. By this, the signal input 45 and the output terminal 80 are connected.

In the embodiment according to FIG. 6b, an RC low-pass filter is formed in the signal processing device 20, by means of which the drive voltage of the second switching element S2 is formed. The terminal 46 is connected to a reference potential, which lies in the region of the reference potential of the second switching element S2. For example, the terminal 46 may be connected to the source terminal (terminal 25) of the switching element S2 or to the midpoint M. In the embodiment according to FIG. 6 c, the signal processing device 20 has a transformer 47, which is connected on the primary side to the connection 45 and on the secondary side to the connection 80. The terminals 46.1 and 46.2 are in turn connected to a reference potential, which is in the range of the reference potential of the second switching element S2 and may be the potential of the center point M, for example.

In the embodiment according to FIG. 6d, the signal processing device 20 has two signal inputs 45.1, 45.2, which are each connected to an amplifier 48, 49. The output signals of the amplifiers 48, 49 are added in an adder 50. Here also a constant voltage U can be additionally added. The gain of the amplifiers 48, 49 may also be <1. The output signal of the adder 50 is supplied via the terminal 80 to the control terminal 24. If the amplifiers 48, 49 have a gain <1, they can be realized by a voltage divider.

FIG. 6e shows an embodiment of a signal processing device 20 which has a diode-shaped rectifier 51 with a subsequent low-pass filter 52 comprising two resistors 53, 54 and a capacitor 55. Diode 51 and low pass filter 52 result in a rapid increase and a slow drop of the drive signal.

FIG. 6f shows a signal processing device 20 which has a damped series resonant circuit 56. The series resonant circuit has a series connection of a coil 57 and a capacitor 58 and, in parallel thereto, a resistor 59.

FIG. 6 g shows a high-pass filter 60, consisting of a capacitor 61 and a resistor 62, which represent the signal processing device 20.

In Figure 7a, the voltage waveform 70 is shown at the top terminal of the switching element S1 when a conventional class E operation is performed. In the conventional class E-operation is usually the second Switching element S2 does not exist, but only a throttle, through which the switching element S1 is connected to the positive potential of the DC power supply. As can be seen from Figure 7a, the maximum voltage is significantly higher than the positive potential 13 of the DC power supply. Thus, situations may occur in which a very high voltage is applied across the switching element S1.

FIG. 7b shows the voltage curve 71 at the center M or above the lower switching element S1 of a switching bridge which is operated in class D operation. Except for a small overshoot 72, the voltage is limited to the positive potential 13 of the DC power supply.

In the figure 7c, the voltage curve 72 is shown, which adjusts in the RF power amplifier arrangement according to the invention. It can be seen that the voltage 73 is substantially limited to the positive potential 13 of the DC power supply.

Claims

claims

A plasma power supply for generating an RF signal having a frequency> 3 MHz and a power> 500 W, comprising an RF power amplifier arrangement (11) having a DC power supply and a switching bridge (12) connected thereto, the two switching elements connected at least indirectly in series (S1, S2) and whose center (M) forms the switching bridge output, wherein the first switching element (S1) is driven by an active drive signal generator, characterized in that an auxiliary circuit for generating a drive signal for the second switching element (S2) as a result Operating state change of the first switching element (S1) is provided.

2. plasma power supply according to claim 1, characterized in that the auxiliary circuit comprises a signal processing means (20) which is directly or indirectly connected to the center (M).

3. Plasma power supply according to claim 2, characterized in that the signal processing device (20) has only one conductor section (30).

4. Plasma power supply according to one of the preceding claims, characterized in that the auxiliary circuit comprises a capacitor between the second switching element (S2) and the midpoint (M) or the second switching element (S2) and a potential (13) of the DC power supply is switched.

5. plasma power supply according to claim 4, characterized in that the capacitance of the capacitor is at least twice as large, in particular at least 5 times as large, preferably at least 10 times as large as the capacitance between the power terminals of the first switching element (S1).

6. Plasma power supply according to one of the preceding claims, characterized in that the auxiliary circuit comprises a coil or a coil section (36, 41), one end of which is connected to the center (M) and the other end to the signal processing means (20) ,

7. Plasma power supply according to claim 6, characterized in that the coil section (36) is part of a primary winding (35) of an output transformer (16) and the center (M) with a tap (37) of the primary winding (35) is connected.

8. plasma power supply according to claim 6, characterized in that the coil section (41) is part of a throttle (23) which is connected in parallel to the second switching element (S2), wherein the midpoint (M) with a tap (40) of the throttle ( 23) is connected.

9. plasma power supply according to one of the preceding claims, characterized in that a throttle (23) is provided parallel to the second switching element (S2).

10. Plasma power supply according to claim 9, characterized in that the throttle (23) is designed such that the current through the inductor (23) during a period of the fundamental frequency changes by less than 20%.

11. Plasma power supply according to one of the preceding claims 2 to 10, characterized in that the signal processing device (20) comprises a filter, in particular a high, low or band pass filter, or a resonant circuit.

12. Plasma power supply according to one of the preceding claims 2 to 11, characterized in that the signal processing device (20) comprises a voltage divider and / or an attenuator.

13. Plasma power supply according to one of the preceding claims 2 to 12, characterized in that the signal processing device (20) comprises one or more amplifiers (48, 49).

14. Plasma power supply according to one of the preceding claims 2 to 13, characterized in that the signal processing device (20) comprises a transformer (47).

15. Plasma power supply according to one of the preceding claims 2 to 14, characterized in that the signal processing device (20) is connected to a plurality of elements (21, 22, 23) of the auxiliary circuit.

16. Plasma power supply according to one of the preceding claims, characterized in that a capacitor (31) is connected in parallel to the first switching element (S1).

17. Plasma power supply according to one of the preceding claims, characterized in that an output network (15) is provided, which is connected to the midpoint (M) and at the output terminal of an RF output signal can be tapped.

18. Plasma power supply according to claim 17, characterized in that the output network (15) is tuned to the fundamental frequency.

19. Plasma power supply according to claim 17, characterized in that the output network (15) is tuned to a deviating from the fundamental frequency.

20. Plasma power supply according to one of the preceding claims 2 to 19, characterized in that the signal processing device (20) is not connected to any active signal generator.

21. A method for operating a plasma power supply comprising an RF power amplifier arrangement (11) with a DC power supply Power supply and a switching bridge (12) connected thereto, the two switching elements (S1, S2) connected at least indirectly in series and whose center (M) forms the switching bridge output, wherein the first switching element (S1) is controlled by an active drive signal generator, characterized in that a drive signal for the second switching element (S2) is generated as a result of an operating state change of the first switching element (S1).

22. The method according to claim 21, characterized in that as the drive signal for the second switching element (S2), a positive voltage relative to the reference potential (terminal 25) of the second switching element (S2) is generated.

23. The method according to claim 22, characterized in that the drive signal is generated without active signal generator.

24. The method according to any one of the preceding claims 21 to 23, characterized in that the second switching element (S2) is driven such that it first passes current flow of charge in a first direction and current flow in the reverse direction only blocks when a part of in first direction flowed charge has flowed in the reverse direction.

25. The method according to any one of the preceding claims 21 to 24, characterized in that the voltage waveform and current waveform are shifted to each other at the first switching element, that the voltage across the first switching element (S1) at the time of switching it <30%, preferably <20% the potential difference between the potentials (13, 14) of the DC power supply is.

26. The method according to any one of the preceding claims 21 to 25, characterized in that parallel to the second switching element (S2), a throttle is used, which is dimensioned so that the current through the Throttle changes by less than 20% during a period of the fundamental frequency.