JP2007104797A - Pulse power supply device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reduce the burden on a circuit element and the size and cost of equipment. <P>SOLUTION: A pulse power supply device is so constructed that the following is implemented: an initial stage capacitor C0 is initially charged by a charger 2; energy is transferred from the capacitor C0 to a capacitor C1 by controlling turn-on of a switch S1; and a compressed narrow-pulse current is obtained from the capacitor C1 by a saturable reactor SI2 that is saturated with the voltage of the capacitor C1 charged through the energy transfer and is supplied to a load 4. The pulse power supply device includes an energy regeneration circuit. When the capacitor C0 is charged in polarity opposite the initial charging polarity by a kickback current that is not absorbed by the load and is reflected to the capacitor C0, the energy regeneration circuit charges the capacitor C0 in polarity reversed to its initial charging polarity to regenerate kickback energy. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a pulse power supply device that combines a pulse generation circuit using a power semiconductor switch and a magnetic pulse compression circuit to generate a large current pulse with a narrow width and high repetition, and particularly uses an excimer laser device or glow discharge. The present invention relates to a circuit for supplying a pulse current to a discharge load such as a thin film forming apparatus.

  The configuration of this type of pulse power supply device is shown in FIGS. 12 and 13 (see, for example, Patent Document 1 and Patent Document 2).

12, the pulse generating circuit 1 is to initially charge the capacitor C0 through the saturable reactor saturable reactor SI0 from the charger 2, reversing the charging voltage of the capacitor C0 by oscillating current I ₀ by turning the semiconductor switch SW after this When the saturable reactor SI1 is saturated, the pulse current I ₁ is supplied from the capacitor C0 to the step-up / magnetic pulse compression circuit 3 through the diode D1 and the saturable reactor SI1, and the pulse current I ₁ is step-up / magnetic pulse compressed. The circuit 3 boosts and compresses the magnetic pulse to charge the peaking capacitor Cp of the load device 4 with a high voltage, and generates a high voltage and large pulse current on the electrode (laser head or the like) with this charging voltage.

In the discharge in the load device 4, a part of the energy is returned to the boosting / magnetic pulse compression circuit side as a kickback current without consuming all of the energy. This kickback current returns as the charging current of the capacitor C0 in the same polarity as the current I ₁ through the booster-magnetic pulse compression circuit 3 to obtain the regeneration of kickback energy by charging a capacitor C0 in the same polarity as the initial charging polarity .

  In FIG. 12, the boosting / magnetic pulse compression circuit 3 will be described. However, there is a configuration in which the boosting circuit (insulating boosting transformer) is omitted. However, at this time, the lower side of the peaking capacitor Cp becomes the ground potential.

In the configuration of FIG. 13, the pulse generation circuit 2 is configured as an LC inversion charging circuit, and this circuit is connected to the high voltage power supply device 1 that provides only half of the high voltage (HV) required for gas discharge. The first capacitor C1 of the LC inversion charging circuit is directly arranged in parallel with the DC voltage source. After the second capacitor C2 is charged at the same time, the first capacitor C1 is saturable to reversely recharge the first capacitor C1. Via the charge reversal inductor Lvi, it contributes to the reversal recharge of the first capacitor C1 using the triggered switch S1, and the complete high voltage required for the discharge is obtained in the capacitor bank (C1-C2). The preliminary ionization unit 5 applies a voltage necessary for preliminary ionization from the standby power supply voltage source to the load device 4. The trigger unit 6 turns on the switch S <b> 1 in the pulse current supply cycle to the load device 4.
Japanese Patent Laid-Open No. 11-262278 JP 2005-185092 A

  The conventional pulse power supply device has the following problems.

(1) In the configuration of FIG. 12, as shown in the equivalent circuit configuration of FIG. 14 (a), the voltage of the capacitor C0 is charged in reverse polarity by turning on the switch SW, so that the current peak when the switch SW is turned on i _peak and current pulse width W _p are increased, and both current peak i _peak and current pulse width W _p are compared to the general capacitor-to-capacitor energy transfer circuit shown in FIG. √2 times, and the power loss of the switch SW increases.

  Further, since the current pulse width is increased, the VT product (voltage time product) of the saturable reactor of the subsequent magnetic pulse compression circuit 3 is required to be √2 times, and the apparatus becomes large and expensive.

Incidentally, saturable reactor SI0 is with magnetic assist function to reduce the switching loss when the switch SW is turned on, a current peak i _peak after the switch SW is turned on, does not reduce the current pulse width W _p.

  (2) In the configuration of FIG. 13, compared to a general capacitor-to-capacitor energy transfer circuit, the switch S1 is responsible for a current peak of 1 / √2, but a current pulse width of √2 times, The loss of switch S1 does not change much. Moreover, the total capacitance of the capacitors C1 and C2 needs to be four times that of the capacitors C1 and C2, and the capacitor C1 on the side whose voltage is reversed by the switch S1 has a large current duty (general capacitors). In the case of an energy transfer circuit from a capacitor to a capacitor, “charging → polarity inversion → discharging” with respect to “charging → discharging”), the apparatus becomes large and expensive.

  12 or 13, when a negative voltage is applied to the load without using an insulating transformer, the control power source of the switch SW or S1 or the pulse generation circuit 1 and the magnetic pulse compression circuit 3 are floated (insulated). The circuit configuration becomes complicated. On the contrary, if the circuit configuration includes an insulating transformer, the size and cost of the device will increase.

  The objective of this invention is providing the pulse power supply device which solved said subject.

  The present invention for solving the above-described problems is characterized by the following configuration.

(1) A pulse power supply device that obtains a high voltage / narrow pulse current and supplies it to a load at a constant repetition cycle,
Energy for performing energy transfer from the capacitor C0 to the capacitor C1 by the turn-on control of the switch S1 connected in series with the first stage capacitor C0 initially charged by the charger, the saturable reactor SI1, and the capacitor C1. A transition circuit;
And a magnetic pulse compression circuit for obtaining a narrow pulse current compressed from the capacitor C1 by a saturable reactor SI2 saturated with the voltage of the capacitor C1 charged by the energy transfer.

  (2) When the capacitor C0 is charged with the opposite polarity to the initial charge polarity by the kickback current reflected to the capacitor C0 without being absorbed by the load, the capacitor C0 is inverted to the initial charge polarity. An energy regeneration circuit for regenerating kickback energy by charging is provided.

(3) The energy regeneration circuit includes:
A configuration in which an LC resonance reactor L1 and a diode D2 in a direction of conducting in a reverse charged state of the capacitor C0 are connected in series with the capacitor C0;
A configuration in which a saturable reactor is provided in parallel to the switch S1 from a connection point between the capacitor C0 and the saturable reactor SI1;
A configuration in which a series circuit of a diode D3 and a reactor L2 is provided in parallel to the switch S1 from a connection point between the capacitor C0 and the saturable reactor SI1;
A configuration in which a series circuit of a switch S2 and a reactor L2 is provided in parallel to the switch S1 from a connection point between the capacitor C0 and the saturable reactor SI1;
It is characterized by having set it as any one of these.

  (4) The switch S1 is turned on, and after that, the capacitor C0 is charged with a reverse polarity to the initial charge polarity by a kickback current, and then the capacitor C0 is inverted and charged to the initial charge polarity. It is characterized by comprising a control circuit for turning off the switch or for turning off the front switch S1 and the switch S2.

  (5) The magnetic pulse compression circuit is composed of two or more stages.

  (6) The switch S1 is configured as a high-voltage circuit by connecting a plurality of switches in series, or configured to have a large current capacity by connecting a plurality of switches in parallel.

  As described above, according to the present invention, since it is constituted by the capacitor-capacitor energy transfer circuit, the voltage duty of the capacitor is reduced, the current duty of the switch S1 is reduced, the current pulse width is shortened, and the magnetic pulse compression circuit of the subsequent stage is reduced. The VT product of the saturable reactor can be reduced.

  Further, a negative voltage can be supplied to the load without using a transformer.

  In addition, one end of the load and charger is connected to the ground, one end (source) of the switch S1 is connected to the ground, and the power source of the on / off control circuit of the switch S1 becomes the ground potential, eliminating the need for insulation between them. This simplifies the circuit configuration.

  In addition, energy that is reflected without being absorbed by the load is accumulated in the first stage capacitor C0, and energy regeneration that can be reused during the next charging can also be performed.

(Embodiment 1)
FIG. 1 is a main circuit configuration diagram showing the present embodiment. An initial stage capacitor C0, a saturable reactor SI1, and a capacitor C1 are connected in series, and a series circuit of a switch S1 and an antiparallel diode D1 is connected in parallel to the series circuit to constitute an energy transfer circuit. The output terminal of the charger 2 is connected to the input stage of the energy transfer circuit (both ends of the switch S1). The capacitor C1 forms a one-stage magnetic pulse compression circuit with the saturable reactor SI2 and supplies a high voltage / narrow pulse current to the load 4. The saturable reactor SI0 is provided in parallel with the load 4 and is reset in a direction of high inductance with respect to the polarity of the high voltage applied to the load 4.

  The operation of the pulse power supply device having the above main circuit configuration will be described. First, the charger 2 generates a high voltage output at the repetition period of the pulse current supplied to the load 4 and initially charges the capacitor C0. The charging current of the capacitor C0 due to the output of the charger 2 flows in a loop of charger → C0 → SI1 → SI2 → SI0. At this time, all the saturable reactors (SI0 to SI2) are saturated (reset) in a direction where the inductance becomes low with respect to the charging current. This reset may be a method of supplying a DC reset current from an external power source to a control winding provided in the saturable reactor, or a method of using the charging current itself of the capacitor C0.

  Next, the switch S1 is turned on, current is passed through the loop of C0 → S1 → C1 → SI1, and the charge is transferred from the capacitor C0 to the capacitor C1 (energy is transferred). At this time, immediately after the switch S1 is turned on, the saturable reactor SI1 is not saturated (high inductance). Therefore, no transition current flows until the saturable reactor SI1 is saturated, and the switch S1 is turned on to reduce the switching loss. . Further, since the saturable reactor SI2 has a high inductance with respect to the negative voltage of the capacitor C1, no voltage is applied to the load 4.

  Next, at the timing when the energy transfer from the capacitor C0 to the capacitor C1 is completed, the saturable reactor SI2 is saturated by the voltage of the capacitor C1, and a compressed narrow pulse current is applied to the load 4 side from the capacitor C1. A narrow pulse current is supplied to the load 4 in a loop of load → SI2. At this time, since the saturable reactor SI0 has a high inductance, almost all of the current from the capacitor C1 flows to the load 4. The energy that is not absorbed by the load 4 appears in the capacitor C1 as a positive voltage of the capacitor C1. The positive voltage of the capacitor C1 flows as a kickback current in a loop of C1, SI1, C0, and S1, and returns as a reverse charge of the capacitor C0.

Therefore, according to this embodiment, the capacitor-to-capacitor energy transfer circuit without voltage polarity reversal is obtained, and both the current peak i _peak and the current pulse width W _p are smaller than in the conventional configuration, and the switch S1 Current and power duties are reduced, and voltage duties and required capacity of capacitors C0 and C1 are reduced. Further, it is possible to from the current pulse width W _p becomes narrow, the miniaturization by reducing the VT product required saturable reactor SI2 of the magnetic pulse compression circuit in the subsequent stage.

  Further, it is possible to supply a negative voltage to the load without using an insulating transformer, and to make a circuit configuration that does not require the control power supply of the switch S1 or the floating (insulation) of the magnetic pulse compression circuit. That is, one end (source) of the switch S1 is connected to the ground, the power source of the signal circuit for driving the switch S1 becomes the ground potential, and insulation between devices becomes unnecessary.

  Although the present embodiment shows a case where the magnetic pulse compression circuit is one stage, it can be a two-stage or more magnetic pulse compression circuit in order to enhance the magnetic pulse compression effect. FIG. 2 shows a case where a three-stage magnetic pulse compression circuit is formed by capacitors C1 to C3 and saturable reactors SI2 to SI4, and similar effects can be obtained while enhancing the magnetic pulse compression effect.

  In addition, this embodiment shows a case where the switch S1 is configured by one semiconductor switch. However, when a higher voltage is required, or when the withstand voltage of the switch is insufficient, the switch S1 is connected in series. can do. FIG. 3 shows a configuration in which three switches S1a to S1c are connected in series. Further, when a larger current is controlled or when the control current capacity of the switch is insufficient, each of the switches S1a to S1c can be configured by connecting a plurality of switches in parallel.

(Embodiment 2)
In the first embodiment, the kickback current reflected to the capacitor C0 without being absorbed by the load 4 charges the capacitor C0 with a polarity opposite to the initial charge polarity, and this kickback energy is restored. There are no processing functions to use or consume. In this case, extra charging energy by the charger 2 is required, and the reflected energy is released to the load 4 again, which may cause deterioration in performance of the load 4 and deterioration of the discharge electrode.

  In this embodiment, the kickback current reflected to the capacitor C0 can be reused (regenerated) as the charging power of the capacitor C0 for the next pulse generation.

  FIG. 4 is a main circuit configuration diagram and a control circuit diagram showing this embodiment. The main circuit of FIG. 6 differs from FIG. 1 in that a kickback current regeneration circuit 7 is provided in parallel with the first stage capacitor C0. The kickback current regeneration circuit 7 has a configuration in which a backflow prevention diode D2 is connected in series to a parallel circuit of an LC resonance reactor L1 and a surge absorbing resistor R1, and the polarity of the connection of the diode D2 is in the reverse charge state of the capacitor C0. The direction to conduct.

  In the above main circuit configuration, the energy that has not been absorbed by the load 4 remains in the capacitor C1 as the positive voltage of the capacitor C1, this energy is transferred to the capacitor C0, and the charging voltage of the capacitor C0 has a negative polarity (in the figure). The right side becomes +). At this time, a half cycle LC oscillation current flows in the capacitor C0 in the loop of C0 → D1 → L1, the voltage of the capacitor C0 is inverted (the voltage of the capacitor C0 is positive: the left side in the figure is +) and kicked back. Energy is regenerated in the capacitor C0. In this regeneration period, the saturable reactor SI1 is not saturated (high inductance), so that current flows only through the regeneration circuit 7. Further, since the switch S1 is turned off, the kickback energy remains accumulated in the capacitor C0, and this energy is reused at the next charging from the charger 2, and the power conversion efficiency as the power supply device. Is improved.

  The control circuit of the switch S1 for enabling the above regenerative operation includes a logic circuit 8, a current detector (CT) 9, a current edge detection circuit 10, and a logic circuit 11 with a timer. When the turn-on signal of the switch S1 is applied, the logic circuit 8 turns it on, and turns off the switch S1 at the timing when the turn-off signal is given. The current detector 9 detects the current of the capacitor C0 at the switch S1, detects the rising timing of this detected current with the current edge detection circuit 10, and gives a certain time delay from this current edge detection to the logic circuit 11 Generates a turn-off signal.

  FIG. 5 shows a time chart of the control circuit. The switch S1 is turned on by the turn-on signal (t1), an energy transfer current starts to flow through the initially charged capacitor C0, and the voltage of the capacitor C0 also starts to decrease. The current edge detection circuit 10 detects that the current at this time has reached a certain level (t2), and further detects that the current has reached a certain level or less (t3). After the energy transfer and the magnetic pulse compression are performed and the pulse current is supplied to the load 4, the kickback current from the load 4 starts to flow into the capacitor C0, and the current edge detection circuit 10 indicates that this current has reached a certain level. It is detected (t4), and it is further detected that the level has reached a certain level (t5). Until the kickback current becomes zero (t6), the voltage of the capacitor C0 continues to be charged with a reverse polarity, and thereafter, regeneration starts by the regenerative circuit 7 and regenerative current starts to flow through the diode D2 (t7). Is reversed to the initial charge voltage polarity, and the regenerative current is also zero (t8). In this regenerative operation, the switch S1 may be turned off between times t6 and t8. Therefore, the logic circuit 11 generates a turn-off signal with a certain delay from the second falling edge detection timing (t5), and the logic circuit 8 turns off the switch S1.

  As described above, in this embodiment, when energy that has not been absorbed by the load is accumulated in the capacitor C0, it is reused at the next pulse generation by the regenerative operation. As a result, the power conversion efficiency of the apparatus is increased and unnecessary pulse voltage is not supplied to the load, and problems such as load performance deterioration and discharge electrode deterioration are solved.

(Embodiment 3)
This embodiment is a kickback current regeneration system different from that of the second embodiment. FIG. 6 shows a main circuit configuration and a control circuit configuration, and FIG. 7 shows a time chart of the control circuit.

  The main circuit configuration in FIG. 6 is that the saturable reactor SI0 in FIG. 1 is provided in parallel to the switch S1 from the connection point between the capacitor C0 and the saturable reactor SI1, or another saturable reactor is provided, thereby providing a kickback current. It is a method to get the regeneration. In this configuration, the charging current from the charger 2 to the capacitor C0 flows in a loop of charger → C0 → SI0. Then, a negative voltage (energy reflected from the load) of the capacitor C0 is passed through a loop of C0 → SI0 → D1, and the voltage of the capacitor C0 is inverted (the voltage of the capacitor C0 becomes positive: FIG. + On the left side.

  At this time, since the saturable reactor SI1 is not saturated (high inductance), current flows only through the regenerative circuit. Further, since the switch S1 is off, the energy of the capacitor C0 remains stored in the capacitor C0, and this energy is reused at the next charging from the charger, thereby improving the efficiency.

  The control circuit in FIG. 6 includes a logic circuit 8, a current detector (CT) 9, and a current edge detection circuit 10. The on / off control of the switch S1 is a time chart shown in FIG. The switch S1 is turned on by the turn-on signal (t1), whereby an energy transfer current flows from the initially charged capacitor C0 to the capacitor C1, further energy transfer and magnetic pulse compression are performed, and a pulse current is applied to the load 4. Supplied. Thereafter, the kickback current from the load 4 starts to flow from the capacitor C1 to the capacitor C0, the voltage of the capacitor C0 is charged with the opposite polarity, and the kickback current eventually becomes zero (t6). Thereafter, a regenerative current begins to flow through the saturable reactor SI0 (t7), and when this current reaches a certain level, the current edge detection circuit 10 generates a turn-off signal (t9). Thereafter, the voltage of the capacitor C0 is inverted to the initial charge voltage polarity, and the regenerative current is also zero (t8).

(Embodiment 4)
This embodiment is a kickback current regeneration system different from those of the second and third embodiments. FIG. 8 shows a main circuit configuration and a control circuit configuration, and FIG. 9 shows a time chart of the control circuit.

  The main circuit configuration in FIG. 8 is a circuit configuration in which a diode D3 and a reactor L2 are connected in series instead of the saturable reactor SI0 in FIG. 6, and surge voltage absorbing resistors R2 and R3 are provided in parallel to each other. Similarly to FIG. 4, the kickback current according to this configuration flows as a vibration current of the reactor L2 and the capacitor C0 through the diode D3, but flows through the diode D1 in parallel with the switch S1.

  The control circuit detects the current of the diode D1 with the current detector 9, and detects the current edge with the current edge detection circuit 10, thereby performing the turn-off control of the switch S1.

(Embodiment 5)
This embodiment uses a kickback current regeneration system different from those in the second to fourth embodiments. FIG. 10 shows a main circuit configuration and a control circuit configuration, and FIG. 11 shows a time chart of the control circuit.

  The main circuit configuration in FIG. 10 is a circuit configuration in which a switch S2 is provided instead of the diode D3 in FIG. 8, and a reverse voltage protection diode D4 is provided in parallel thereto. In the regenerative operation with this configuration, the switch S2 is turned on during the initial charging of the capacitor C0, and the switch S2 is turned off before the switch S1 is turned on during energy transfer. During this energy transfer, the current flowing through the diode D4 is blocked by the reactor L2. The regeneration of the kickback current inverts the negative voltage of the capacitor C0 (energy reflected by the load) in a loop of C0 → L2 → S2 → D1.

  The control circuit detects the turn-on current of the switch S1 with the positive current edge detection circuit 10A, detects the regenerative current of the diode D1 with the negative current edge detection circuit 10B, and detects the positive current by switching the switch S2 with a delay by the timer circuit 11. The switch S1 is turned off by detecting the negative current.

  In Embodiments 2 to 5 described above, the main circuit configuration can be applied to the configuration shown in FIG. 2 or FIG.

The main circuit block diagram which shows Embodiment 1 of this invention. The block diagram of the magnetic pulse compression circuit of 3 steps | paragraphs structure. The switch block diagram of series connection. The main circuit structure and control circuit diagram which show Embodiment 2 of this invention. 7 is a time chart of a control circuit (second embodiment). The main circuit structure and control circuit diagram which show Embodiment 3 of this invention. 9 is a time chart of a control circuit (third embodiment). The main circuit structure and control circuit diagram which show Embodiment 4 of this invention. 10 is a time chart of a control circuit (Embodiment 4). The main circuit structure and control circuit diagram which show Embodiment 5 of this invention. 10 is a time chart of a control circuit (Embodiment 5). The block diagram of the conventional pulse power supply device. The block diagram of the conventional pulse power supply device. Explanatory drawing of an energy transfer circuit.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Pulse generation circuit 2 Charger 3 Magnetic pulse compression circuit 4 Load 7 Kickback current regeneration circuit 8 Logic circuit 9 Voltage detector 10, 10A, 10B Current edge detection circuit 11 Timer circuit S1, S2 switch SI0, SI1, SI2 Saturability Reactor C0 First stage capacitor C1 Capacitor L1, L2 Reactor D1, D2, D3, D4 Diode

Claims (6)

A pulse power supply device that obtains a high voltage / narrow pulse current at a constant repetition period and supplies it to a load,
Energy for performing energy transfer from the capacitor C0 to the capacitor C1 by the turn-on control of the switch S1 connected in series with the first stage capacitor C0 initially charged by the charger, the saturable reactor SI1, and the capacitor C1. A transition circuit;
And a magnetic pulse compression circuit for obtaining a narrow pulse current compressed from the capacitor C1 by a saturable reactor SI2 saturated with the voltage of the capacitor C1 charged by the energy transfer.   When the capacitor C0 is charged with the opposite polarity to the initial charge polarity by the kickback current reflected to the capacitor C0 without being absorbed by the load, the capacitor C0 is inverted and charged to the initial charge polarity. The pulse power supply device according to claim 1, further comprising an energy regeneration circuit that regenerates kickback energy. The energy regeneration circuit includes:
A configuration in which an LC resonance reactor L1 and a diode D2 in a direction of conducting in a reverse charged state of the capacitor C0 are connected in series with the capacitor C0;
A configuration in which a saturable reactor is provided in parallel to the switch S1 from a connection point between the capacitor C0 and the saturable reactor SI1;
A configuration in which a series circuit of a diode D3 and a reactor L2 is provided in parallel to the switch S1 from a connection point between the capacitor C0 and the saturable reactor SI1;
A configuration in which a series circuit of a switch S2 and a reactor L2 is provided in parallel to the switch S1 from a connection point between the capacitor C0 and the saturable reactor SI1;
The pulse power supply device according to claim 2, wherein any one of the configurations is adopted.   The switch S1 is turned on, and after that, the capacitor C0 is charged with a kickback current in the opposite polarity to the initial charge polarity, and then the switch is turned off at the timing when the capacitor C0 is reversely charged to the initial charge polarity. The pulse power supply device according to any one of claims 1 to 3, further comprising a control circuit for controlling or turning off the front switch S1 and the switch S2.   5. The pulse power supply device according to claim 1, wherein the magnetic pulse compression circuit includes two or more stages. The said switch S1 was comprised in the high voltage | pressure-resistant circuit by the serial connection of several switches, or comprised by the large current capacity by the parallel connection of several switches, The switch of any one of Claims 1-5 characterized by the above-mentioned. Pulse power supply.

Images