JP5550648B2 - High frequency power supply - Google Patents

Description

  The present invention relates to a high frequency power supply device that outputs voltage pulses.

  As a high-frequency power supply device that outputs voltage pulses, for example, there is a high-frequency power supply device that is used when laser oscillation is performed using AC discharge. In recent years, as a laser apparatus for microfabrication represented by printed circuit board drilling, there is an increasing demand for a laser apparatus that can perform pulse oscillation with a pulse width of a laser pulse output of 1 microsecond to several hundred microseconds.

  In the high frequency power supply device used in the laser device as described above (for example, a carbon dioxide laser device), in order to excite the laser medium, the discharge voltage applied to the discharge electrode of the laser oscillator is several kV, The flowing discharge current needs to have a peak of several tens of amperes and an AC frequency (discharge frequency) during discharge of several hundred kHz to several MHz or more. Hereinafter, the power for causing the discharge in this way is referred to as discharge power.

  In order to supply the discharge power as described above to the laser device, a high frequency power supply device using an inverter has been conventionally used. By performing on / off control of the switching element of the inverter, a substantially sinusoidal current is supplied to the laser oscillator, and electric power is supplied to the discharge space filled with the laser medium to generate a discharge. Then, the laser medium excited by the discharge oscillates and is output as a laser pulse (see, for example, Patent Document 1 below).

  By the way, in a laser apparatus for microfabrication such as a printed circuit board, there is an optimum energy value of a laser depending on a material of a workpiece and a processing method. Therefore, various control parameters for controlling the laser output of the laser oscillator are set.

  As such control parameters, for example, as shown in FIG. 20, (i) a peak output indicating the peak value of the discharge power of the laser pulse, (ii) a repetitive pulse frequency indicating the output frequency of the laser pulse, and (iii) There is a pulse width of the laser pulse.

  In order to obtain a laser pulse suitable for the above control parameters, for example, the following control is performed in the high frequency power supply device.

(I) The peak output of the discharge power of the laser pulse is mainly determined by the discharge voltage applied to the electrodes of the laser oscillator and the peak of the discharge current flowing by the discharge generated between the electrodes. Therefore, in order to control the peak output of the discharge power of the laser pulse, the discharge voltage is controlled by adjusting the voltage of the DC power supply to the inverter, or the duty of the inverter is set by PWM (Pulse Width Modulation). A method of controlling the peak of the discharge current by controlling is adopted. For example, as shown in FIG. 21, in the conditions 1 and 2 of FIG. 20, when the peak output is p1> p2, the voltage of the DC power supply of the inverter is set to E1> E2.

(Ii) The repetition pulse frequency is an output frequency of a laser pulse corresponding to the number of laser pulses irradiated per unit time. By increasing the repetition pulse frequency, the processing speed (the number of processing) at the time of miniaturization processing can be improved. And in order to control this repetition pulse frequency, the number per unit time of a series of control pulse groups which carry out on / off control of the switching element of an inverter is controlled. For example, as shown in FIG. 21, when the repetitive pulse frequency is f1> f2 in conditions 1 and 2 of FIG. 20, per unit time T of a series of control pulse groups for controlling on / off of the switching elements of the inverter. Is set to N1> N2. In FIG. 21, N1 = 3 and N2 = 2 are set.

(Iii) The pulse width of the laser pulse is proportional to the number of a series of control pulses for ON / OFF control of the switching element of the inverter, that is, the number of switching times n of the switching element. For this reason, in order to control the pulse width of the laser, the switching frequency n of the switching element is controlled. For example, as shown in FIG. 21, when the pulse width is t1 <t2 under the conditions 1 and 2 in FIG. 20, the switching frequency of the switching element is set to n1 <n2. In FIG. 21, 2 · n1 = n2, and the laser pulse width can be doubled by doubling the series of switching times (2n).

  By the way, in recent years, with the spread of IT-related equipment, a laser device for fine processing with higher productivity is demanded. In addition, processing materials and types of processing in fine processing are diversified. Therefore, it is necessary to set appropriate control parameters in order to increase the processing speed (the number of times of processing) and inject processing energy suitable for the processing material and the type of processing.

  In this case, in order to improve the processing speed (the number of processing), as described above, it is necessary to increase the repetition pulse frequency of laser oscillation as a control parameter. Also, in order to inject processing energy suitable for the processing material and type of processing, the laser pulse width should be changed significantly as a control parameter (for example, from 1 microsecond or less to several hundred microseconds). Is required. For example, in the case of a polyimide resin, high quality processing can be obtained when the peak output is high but the laser irradiation time is short (that is, the pulse width is small). In addition, a glass epoxy material containing glass fibers has a smaller peak output but a relatively long laser irradiation time (that is, a longer pulse width) can provide better processing.

  As described above, it is important to increase the repetitive pulse frequency of the pulse laser, which is a control parameter, or to change the pulse width of the laser in consideration of productivity and characteristics of the processed material. However, when the repetition pulse frequency of the pulse laser is increased or the pulse width of the laser is increased, the number of times of switching per unit time of the switching element inevitably increases as described above. Loss also increases and heat generation of the high frequency power supply device increases.

  In the prior art, as a measure for increasing the laser pulse width while suppressing the switching loss of the inverter, a series of control pulses for on / off control of the switching elements of the inverter are regularly thinned out to obtain a desired laser pulse. A technique has been proposed in which the switching frequency of the switching element is relatively reduced and the switching loss is suppressed while ensuring the width (for example, see Patent Document 2 below).

JP 2000-4059 A JP 2004-22696 A

  As described above, if the repetition pulse frequency of the laser pulse, which is a control parameter, or the pulse width is increased in consideration of productivity and characteristics of the processed material, switching per unit time of the switching element is inevitably required. The number of times increases. As a result, the switching loss of the switching element also increases and the heat generation of the power supply device increases.

Switching loss associated with ON / OFF of the switching element of the inverter includes a turn-on loss and a turn-off loss caused by a product of a voltage and a current applied to the channel of the switching element during switching, and when the switching element is turned on. Further, it is represented by the sum of charge and discharge losses due to the stray capacitance generated when the charge charged in the stray capacitance of the switching element is discharged through the channel of the switching element.
Among the above, the charge / discharge loss W due to the stray capacitance is expressed by the following equation (1).
W = (1/2) CV ² f (1)

  Here, C is the stray capacitance, V is the charge voltage of the stray capacitance, and f is the number of switching times of the switching element when loss due to discharge of the stray capacitance occurs. The stray capacitance C includes a stray capacitance that is parasitic between the drain and source of each switching element, a stray capacitance that is floating between the drain and the source of the switching element by floating on the gate power supply line, and a series of the switching elements. This is the entire stray capacitance including the capacitance of an external capacitor to equalize the partial pressure.

  As can be seen from the above equation (1), the charge / discharge loss W due to the stray capacitance is a discharge loss mainly resulting from the charge / discharge of the stray capacitance. Therefore, in a high-frequency inverter circuit in which the charge voltage of the stray capacitance is high and the switching frequency of the switching element is large, the charge / discharge loss due to the stray capacitance is significantly increased. Therefore, in order to operate the inverter at a high frequency, it is necessary to reduce the charge / discharge loss W, that is, to reduce the charge voltage V of the stray capacitance or to reduce the stray capacitance itself.

  In the prior art (Patent Document 2), a series of control pulses for on / off control of the switching elements of the inverter are regularly thinned out to ensure a desired laser pulse width while relatively switching the number of switching times of the switching elements. This reduces the heat generated by the power supply. However, when a series of control pulses are regularly thinned out, the number of repeated operations of stopping / starting the inverter increases, and accordingly, the switching loss of the switching element due to charging / discharging of the stray capacitance increases remarkably. As a result, there is still a problem that heat generation of the power supply device increases.

  In other words, when the switching element is continuously turned on / off without thinning out a series of control pulses for on / off control of the switching element, the charge charged in the stray capacitance is discharged by the regenerative operation. However, when a series of control pulses are regularly thinned out, the operation of stopping / starting the inverter is frequently repeated, so that after the inverter is stopped, the inverter is next discharged in a state where the charge charged in the stray capacitance is not sufficiently discharged. When activated, the resonance condition is not sufficiently satisfied, and a large current flows immediately after the inverter is activated, resulting in a large loss. For this reason, the effect of reducing the switching loss due to thinning is not so remarkable.

  In addition, in order to increase the repetition pulse frequency of the laser pulse in order to improve the processing speed (the number of processing), it is not possible to cope with it by thinning out a series of control pulses for on / off control of the switching element. Therefore, the number of switching operations per unit time of the switching element still increases, and as a result, the heat generation of the high frequency power supply device increases.

  The present invention has been made to solve the above-described problem, and in a high-frequency power supply apparatus that realizes high-frequency power supply by on / off control of switching elements of an inverter, an increase in the number of switching times of each switching element constituting the inverter An object of the present invention is to provide a high-frequency power supply device that can significantly reduce the switching loss due to.

The high frequency power supply device of the present invention outputs a voltage pulse to a load, and is connected to an inverter for power conversion having a switching element, a first reactor connected to the load, and an output terminal of the inverter. And a charge recovery circuit for recovering the charge of the stray capacitance of the switching element, and a control means for controlling the operation of the inverter and the charge recovery circuit. The charge recovery circuit is a switching circuit for starting the circuit. A series of inverter control pulses for controlling on / off of the switching element of the inverter, the second reactor for performing LC resonance with the stray capacitance, and a capacitor for charge recovery. of starting the switching elements of the group and the charge recovery circuit dead-time period of the switching element of the inverter It is characterized by outputting a that charge collection control pulses.

  According to the present invention, the switching element constituting the charge recovery circuit is turned on / off at a necessary timing, and LC resonance is caused between the second reactor of the circuit and the stray capacitance of the switching element of the inverter, The charge charged in the stray capacitance of each switching element constituting the inverter is input / output to / from the capacitor of the charge recovery circuit. Thereby, since the charging voltage of the stray capacitance when the switching element of the inverter is switched can be reduced, the switching loss of the inverter can be greatly reduced even if the number of times of switching is increased. Thereby, the cooling mechanism of the apparatus is simplified, and the apparatus can be reduced in size, cost, installation space, and the like.

It is a block diagram which shows the outline of a gas laser apparatus. It is a block diagram of the high frequency power supply device by Embodiment 1 of this invention. It is a block diagram which shows the structure of the pulse control circuit of the power supply device. It is a block diagram which shows an example of the electric charge collection circuit of the power supply device. It is a block diagram which shows another example of the electric charge collection circuit of the power supply device. It is a timing chart which shows an example of control operation of the power supply device. 4 is a timing chart showing an inverter control pulse and a charge recovery control pulse when the power supply device is driven, and a current flowing through each switching element by these pulses. It is explanatory drawing which shows the current pathway which flows into each part of an inverter at the time of the drive of the power supply device. It is explanatory drawing which shows the current pathway which flows into each part of an inverter at the time of the drive of the power supply device. It is explanatory drawing which shows the current pathway which flows into each part of an inverter at the time of the drive of the power supply device. It is explanatory drawing which shows the current pathway which flows into each part of an inverter at the time of the drive of the power supply device. It is explanatory drawing which shows the current pathway which flows into each part of an inverter at the time of the drive of the power supply device. It is explanatory drawing which shows the current pathway which flows into each part of an inverter at the time of the drive of the power supply device. It is a timing chart which shows an example in the case of performing the thinning-out control operation of the power supply device. It is a timing chart which shows another example in the case of performing the thinning-out control operation of the same power supply device. It is a timing chart which shows another example in the case of performing the thinning-out control operation of the same power supply device. It is a block diagram of the high frequency power supply device in Embodiment 2 of this invention. It is a block diagram which shows an example of the electric charge collection circuit of the power supply device. It is a block diagram which shows another example of the electric charge collection circuit of the power supply device. It is explanatory drawing of the control parameter set in order to control the laser output of a laser oscillator. It is explanatory drawing which shows the relationship between the drive state of the switching element of the inverter accompanying the setting of the control parameter of FIG. 20, and a laser pulse output.

Embodiment 1 FIG.
FIG. 1 is a configuration diagram showing an outline of a gas laser device to which a high frequency power supply device according to Embodiment 1 of the present invention is applied, and FIG. 2 is a configuration diagram of the high frequency power supply device according to Embodiment 1 of the present invention.

  The high frequency power supply device 1 according to the first embodiment is connected to a laser oscillator 2 constituting a gas laser device. As shown in FIG. 1, the laser oscillator 2 includes a pair of electrodes 3 arranged opposite to each other, a dielectric 4 formed on the opposite surface side of each electrode 3, a partial reflection mirror 5 for performing laser oscillation, A resonance mirror including the total reflection mirror 6. The discharge space 7 formed between the dielectrics 4 is filled with a laser medium (mixed gas). Then, electric power is supplied to the discharge space 7 to generate discharge, and laser emission is generated between the mirrors 5 and 6 by the stimulated emission of the laser medium excited by the discharge, and is output as laser light. In this case, as shown in FIG. 2, the dielectric 4 and the discharge space 7 are represented by a series circuit of a discharge resistor 191 and a dielectric capacitor 192 as a capacitive load 19 for the high frequency power supply device 1 in an equivalent circuit. The

  As shown in FIG. 2, the high-frequency power device 1 includes a full-bridge inverter 11, a DC power source 12, a step-up transformer 13, a first reactor 14, an inverter control circuit 15, a thinning circuit 16, a pulse control circuit 17, and The charge recovery circuit 18 is mainly configured. In the high-frequency power supply device 1, the output side of the inverter 11 is connected to a series circuit of a discharge resistor 191 and a dielectric capacitor 192 of a gas laser device serving as a capacitive load 19 via the step-up transformer 13 and the first reactor 14. ing.

  The first reactor 14 often uses the leakage inductance of a high-frequency transformer and is not necessarily attached to the outside. Further, when the inverter 11 can output a high voltage, it can be directly connected to the load 19 via the first reactor 14 without using the step-up transformer 13. Further, the high frequency power supply device 1 is not applied only to the gas laser device having the configuration as shown in FIG. 1, but a semiconductor manufacturing device, a liquid crystal manufacturing device, an electric discharge machining device, an ozone generator, and a microwave generator. Plasma generation applications such as equipment, laser generation applications such as laser equipment for sheet metal processing and printed circuit board drilling, high frequency welding machines for welding thermoplastic resins, induction heating press for metal heating, and IH cooking heaters for cooking It can be applied to all devices having a load driven by high-frequency AC power, such as high-frequency induction overheating.

  The inverter 11 includes switching elements S1 to S4 made of MOSFET, IGBT, and the like, and gate drive circuits 111 to 114 that individually drive the switching elements S1 to S4. In addition, the diodes D11 to D14 and the stray capacitances C11 to C14 are connected in parallel to the switching elements S1 to S4.

  Each of the diodes D11 to D14 is used when the inverter 11 performs regeneration or reflux, and can be omitted when a diode is parasitic inside each of the switching elements S1 to S4. In addition, each of the stray capacitances C11 to C14 here is a stray capacitance parasitic between the drain and the source of the switching elements S1 to S4, and a stray capacitance that floats on the gate power supply line and equivalently occurs between the drain and the source of the switching element. , And the total stray capacitance including the capacitance of a capacitor externally attached to equalize the series voltage division of each of the switching elements S1 to S4.

  On the other hand, the inverter control circuit 15 outputs inverter control command pulses A1 to A4 suitable for setting various control parameters as shown in FIG.

  The thinning circuit 16 is a circuit that thins out the command pulses A1 to A4 at regular intervals according to the thinning setting for the inverter 11. The thinning circuit 16 can freely set the thinning interval and the number of thinnings, such as thinning one pulse every other pulse or thinning two pulses every two pulses. Further, it is possible not to perform thinning, in which case the operation of the thinning circuit 16 is stopped or the thinning circuit 16 is omitted and the command pulses A1 to A4 of the inverter control circuit 15 are directly applied. This can be given to the pulse control circuit 17.

  The pulse control circuit 17 corresponds to the control means in the claims, and includes an inverter control pulse generation circuit 171 and a charge recovery control pulse generation circuit 172 as shown in FIG. The inverter control pulse generation circuit 171 generates inverter control pulses G1 to G4 based on the command pulses B1 to B4 after thinning after passing through the thinning circuit 16 and outputs the inverter control pulses G1 to G4 to the gate drive circuits 111 to 114 of the inverter 11. . The charge recovery control pulse generation circuit 172 generates charge recovery control pulses G11 to G44 having a predetermined pulse width tw (described later) based on the command pulses B1 to B4 after thinning after passing through the thinning circuit 16. The charge is output to the charge recovery circuit 18.

  In this case, the output timing of the inverter control pulses G1 to G4 is delayed by a fixed time Δt from the output timing of the charge recovery control pulses G11 to G44. Therefore, the charge recovery circuit 18 is activated by the charge recovery control pulses G11 to G44 immediately before the switching elements S1 to S4 constituting the inverter 11 are turned on by the inverter control pulses G1 to G4. This delay time Δt is the maximum charge recovery time of the stray capacitance charge by the charge recovery circuit 18.

  The pulse control circuit 17 for creating such inverter control pulses G1 to G4 and charge recovery control pulses G11 to G44 may be created using an analog IC circuit as shown in FIG. For example, a necessary pulse pattern may be registered in advance in a storage device such as a ROM, FPGA, or microcomputer, and each pulse may be output according to each setting.

  The charge recovery circuit 18 has a configuration as shown in FIG. 4 or FIG. 5, for example. The charge recovery circuit 18 is activated by the charge recovery control pulses G11 to G44 immediately before the switching elements S1 to S4 constituting the inverter 11 are turned on by the inverter control pulses G1 to G4. Then, the charge recovery circuit 18 discharges the charges of the stray capacitances C11 to C14 existing in the switching elements S1 to S4 to the charge recovery circuit 18 or charges the stray capacitances C11 to C14 from the charge recovery circuit 18. I do. The configuration and operation of the charge recovery circuit 18 will be described in detail later.

  FIG. 6 shows an example of a timing chart corresponding to each signal shown in FIG. 2. Here, each signal waveform in the output period Tp corresponding to one pulse of the laser pulse (voltage pulse) is shown. .

  The inverter control circuit 15 outputs command pulses A1 to A4 corresponding to control parameters (for example, pulse width command) to the thinning circuit 16. The thinning circuit 16 thins out the command pulses A1 to A4 to a state that requires them, and outputs the thinned command pulses B1 to B4. The thinned command pulses B1 to B4 are input to the pulse control circuit 17.

  The thinned command pulses B1 to B4 input to the charge recovery control pulse generation circuit 172 of the pulse control circuit 17 are charge recovery control pulses G11 having a pulse width tw shorter than the thinned command pulses B1 to B4. To G44 and output. Further, the thinned command pulses B1 to B4 input to the inverter control pulse generation circuit 171 are output as inverter control pulses G1 to G4 in a state delayed by a fixed time Δt from the charge recovery control pulses G11 to G44. Is done.

  The inverter control pulses G1 to G4 output from the pulse control circuit 17 are input to gate drive circuits 111 to 114 provided corresponding to the switching elements S1 to S4 of the inverter 11, respectively. The gate drive circuits 111 to 114 output gate drive pulses necessary for driving the switching elements S1 to S4 according to the control pulses G1 to G4, and drive the switching elements S1 to S4 on / off.

  As a result, a rectangular high-frequency voltage is generated at the output of the inverter 11, and the output is boosted by the high-frequency transformer 13 and supplied to the load 19 through the first reactor 14. In that case, the high frequency component of the output current is eliminated by the LC circuit including the first reactor 14 and the dielectric capacitor 192, and a substantially sinusoidal current flows through the load 19. As a result, electric power is supplied to the discharge space 7 filled with the laser medium (mixed gas) as shown in FIG. 1 to generate a discharge, and the laser oscillator 2 oscillates by stimulated emission of the laser medium excited by this discharge. A laser pulse is output. When the inverter 11 can output a high voltage, it can be directly connected to the load 19 via the first reactor 14 without using the step-up transformer 13.

Next, the configuration and operation of the above-described charge recovery circuit 18 will be described in detail.
In the first embodiment, for example, the charge recovery circuit 18 having the configuration shown in FIG. 4 is employed. The charge recovery circuit 18 has a role of charging and discharging charges of the stray capacitances C11 to C14 existing in the switching elements S1 to S4 of the inverter 11 using LC resonance. For this purpose, the charge recovery circuit 18 includes two charge / discharge circuits 18a and 18b. Each of the charge / discharge circuits 18a and 18b includes switching elements S11 and S44 and S22 and S33 that start the circuits 18a and 18b, and an inverter 11. The second reactor 182 for performing LC resonance with the stray capacitances C11 to C14 and the capacitor 183 for charge collection are provided.

  Further, when MOSFETs are used as the switching elements S11 to S44 constituting the charge / discharge circuits 18a and 18b, since the parasitic diode 184 exists in each MOSFET, in order to prevent forward conduction of the parasitic diode 184. The diode 185 is connected in series so that the polarity is opposite to that of the parasitic diode 184. In addition, since it is necessary to perform a bidirectional switch operation for charging / discharging, each charging / discharging circuit 18a, 18b is connected in antiparallel to a series circuit in which each switching element S11-S44 and diode 185 are connected. And the 2nd reactor 182 is connected to the one end side, and the capacitor | condenser 183 is connected to the other end side.

  The charging / discharging circuit 18a on one side (for example, the upper side in the figure) has the second reactor 182 side on one output terminal (the output terminal X in FIG. 2) of the inverter 11 and the other (for example, the lower side in the figure). In the circuit 18b, the second reactor 182 side is connected to the other output terminal (the output terminal Y in FIG. 2) of the inverter 11, and the capacitor 183 side of each of these charge / discharge circuits 18a and 18b is connected to the negative side of the inverter 11. Connected to the (N side) terminal.

  In the charge recovery circuit 18, at least a voltage that is 1/2 of the bus voltage E of the inverter 11 is applied to the capacitor 183, the switching elements S 11 to S 44, and the diode 185. Therefore, in the charge recovery circuit 18, an element having a rating of 1/2 or more of the bus voltage E is selected, or the necessary number of elements are serially connected so that the entire circuit has a withstand voltage of 1/2 of the bus voltage. Need to connect. The capacitor 183 receives and inputs charges from the stray capacitances C11 to C14 of the inverter 11. For this reason, it is preferable to connect a capacitor 183 having a larger capacity than the stray capacitances C11 to C14 so that the voltage of the capacitor 183 does not fluctuate greatly due to the input and output of the charges. For example, when the capacitances of the stray capacitances C11 to C14 are several hundreds pF, it is preferable to select a capacitor having a capacitance of several tens of nF or more.

  In an ideal case where the parasitic diode 184 does not exist in each of the switching elements S11 to S44, a configuration as shown in FIG. That is, the charge recovery circuit 18 includes two charge / discharge circuits 18a and 18b. Each charge / discharge circuit 18a and 18b includes a second reactor 182, a switching element 186, and a capacitor 183 sequentially connected in series. Yes. The charging / discharging circuit 18a on one side (for example, the upper side in the figure) has the second reactor 182 side on one output terminal (the output terminal X in FIG. 2) of the inverter 11 and the other (for example, the lower side in the figure). In the circuit 18b, the second reactor 182 side is connected to the other output terminal (the output terminal Y in FIG. 2) of the inverter 11, respectively. Further, the capacitor 183 side of each of these charge / discharge circuits 18 a and 18 b is connected to the negative side (N side) terminal of the inverter 11.

The charge recovery circuit 18 shown in FIG. 4 or FIG. 5 charges and discharges charges in the stray capacitances C11 to C14 of the inverter 11 using LC resonance. Therefore, as shown in the following equation (2), 1/2 of the resonance period determined by the inductance value L of the second reactor 182 of the charge recovery circuit 18 and the capacitance C of the total stray capacitances C11 to C14 is the maximum charge. It is most ideal to set the inductance value L of the second reactor 182 so as to be shorter than the recovery time Δt. And the switching loss resulting from the charge / discharge of the stray capacitances C11 to C14 occurring at this time is minimized.
Δt ≧ π√LC
∴L ≦ (Δt / π) ² × 1 / C (2)

However, the current peak I flowing in the charge recovery circuit 18 is determined by the following equation (3) under ideal conditions where the circuit impedance when the switching elements S1 to S4 of the inverter 11 and the diode 185 are energized is very small.
I = V / √ (L / C) (3)

  For this reason, if the charge recovery time Δt is shortened, the inductance value L of the second reactor 182 decreases, but the current peak value I increases, and the switching elements S11 to S44 (which constitute the charge / discharge circuits 18a and 18b) For 186) and diode 185, it is necessary to select an element capable of flowing a relatively high peak current. Conversely, if the charge recovery time Δt is increased, the inductance value L of the second reactor 182 needs to be increased. However, since the current peak I can be reduced, the element must be capable of flowing a high peak current. Will disappear. Thus, since the current flowing through the charge recovery circuit 18 is determined by the charge recovery time Δt and the inductance value L of the second reactor 182, an appropriate reactor is selected according to the rating of the selected element and the circuit size.

  The series resonance frequency determined by the second reactor 182 and the stray capacitances C11 to C14 is set so that the inductance value of the sum of the inductance value of the second reactor 182 and the inductance value of the first reactor 14 and the dielectric capacitor 192 are in series. If it is set sufficiently higher than the resonance frequency, the inflow current from the charge recovery circuit 18 to the load 19 can be reduced. According to this, the effect of collecting charges from the stray capacitances C11 to C14 is increased, and the loss of the switching elements S11 to S44 (186) and the diode 185 constituting the charge / discharge circuits 18a and 18b can be reduced.

  Further, even if the inductance value of the second reactor 182 is set to be smaller than the inductance value of the first reactor 14, the charge recovery effect from the stray capacitances C11 to C14 is increased, and the charge / discharge circuits 18a and 18b are connected to each other. Loss of the switching elements S11 to S44 (186) and the diode 185 to be configured can be reduced. Alternatively, the inductance value of the first reactor 14 is set so that the impedance of the stray capacitances C11 to C14 is smaller than the series impedance of the first reactor 14 and the load 19, and the inductance value of the first reactor 14 Even if the frequency of the inverter 11 is set so as to resonate with the dielectric capacitor 192 of the load 19, the same effect as described above can be obtained.

Note that the charge recovery of the stray capacitances C11 to C14 of the switching elements S1 to S4 constituting the inverter 11 is a dead time period determined so that the upper and lower switching elements (S1 and S4, S2 and S3) are not turned on simultaneously. It needs to be done in a very short time. For example, when the inverter 11 is driven at several hundred kHz to several MHz, the dead time is about several tens of nanoseconds to several hundred nanoseconds, and the charge recovery needs to be performed in a time shorter than that. Therefore, it has a high-speed switching (10 ⁴ V / microsecond to 10 ⁵ V / microsecond) of one digit or more than the switching speed (10 ³ V / microsecond) of the switching element using the conventional Si (silicon) semiconductor. By using a wide gap semiconductor, the charge recovery effect is further increased, and the loss of each of the switching elements S1 to S4 of the inverter 11 can be further reduced. Examples of the wide band gap semiconductor in this case include SiC (silicon carbide), a gallium nitride material, and diamond.

  Next, charging / discharging operations for the stray capacitances C11 to C14 of the inverter 11 when the charge recovery circuit 18 having the configuration shown in FIG. 4 is used will be described with reference to FIGS. 7 and 8 to 13. FIG. FIG. 7 shows operation waveforms of the switching elements S1 to S4 and the charge recovery circuit 18 when the inverter 11 is continuously operated based on the inverter control pulses G1 to G4 without thinning out the command pulses A1 to A4. The voltages Viv and Iiv output from the accompanying inverter 11 are shown, and FIGS. 8 to 13 show current paths (broken lines in the figure) flowing through the inverter 11 in the case of FIG.

  As shown in FIG. 7, t0 <t1 <t2 from when switching elements S3 and S4 of inverter 11 are completely turned off at time t0 to when switching elements S1 and S2 of inverter 11 are turned on at time t2. At time t1, the switching elements S11 and S22 of the charge recovery circuit 18 are turned on. Further, the switching elements S11 and S22 of the charge recovery circuit 18 are turned off at the timing of time t3 where t2 <t3 <t4 until the switching elements S1 and S2 of the inverter 11 are turned on at time t2 and turned off at time t4. To do.

  As described above, the timing time t1 when the switching elements S11 and S22 of the charge recovery circuit 18 are turned on is t0 <t1 <t2, and the timing time t3 when the switching elements S11 and S22 are turned off is t2 <t3 <t4. The reason is as follows. That is, the switching elements S11 of the charge recovery circuit 18 and the switching element S4 of the inverter 11 and the switching elements S22 of the charge recovery circuit 18 and the switching element S3 of the inverter 11 are simultaneously turned on, so that each switching element fails. Is to prevent. An appropriate output timing of these pulses is determined by the pulse control circuit 17.

  As described above, since the charge recovery control pulses G11 and G22 are output at time t1, the switching elements S11 and S22 of the charge recovery circuit 18 are simultaneously turned on by the pulses G11 and G22 as shown in FIG. As a result, during the period of (t1 to t2), the charge is transferred from the capacitor 183 of the charge recovery circuit 18 to the stray capacitance C14 of the switching element S4 of the inverter 11 and charged, and the stray capacitance of the switching element S1 of the inverter 11 is charged. The accumulated charge of C11 is discharged. At the same time, the charge accumulated in the stray capacitance C12 of the switching element S2 of the inverter 11 is discharged and transferred to the capacitor 183 of the charge recovery circuit 18, and the charge is charged to the stray capacitance C13 of the switching element S3 of the inverter 11. The

In this way, in a state where almost all the charges stored in the stray capacitances C11 and C12 of the switching elements S1 and S2 of the inverter 11 are discharged, the inverter control pulses G1 and G2 are output at the next time t2, and the inverter 11 Both switching elements S1 and S2 are turned on. For this reason, there is no charge from the stray capacitances C11 and C12 passing through the channels of the switching elements S1 and S2, and the switching loss of the switching elements S1 and S2 is greatly reduced.
Then, during the period of (t2 to t4), when both the switching elements S1 and S2 of the inverter 11 are turned on, a current flows through the switching elements S1 and S2, as shown in FIG.

  Next, a period of (t4 to t6) from when switching elements S1 and S2 of inverter 11 are completely turned off at time t4 to when switching elements S3 and S4 of inverter 11 are turned on at time t6 It is a dead time period. As will be described later, the switching elements S33 and S44 of the charge recovery circuit 18 are turned on at the time t5 in the dead time period. As shown in FIG. 10, during the period (t4 to t5) until that time. At the same time, the current of the inverter 11 starts to be regenerated, and at the same time, the stray capacitances C13 and C14 of the switching elements S3 and S4 start to be discharged, while the stray capacitances C11 and C12 of the switching elements S1 and S2 start to be charged. When the charges of the stray capacitances C13 and C14 are discharged, the charging voltage gradually decreases, but depending on the length of the regeneration time, the discharge becomes insufficient, so that the stray capacitances C13 and C14 remain in a state where charges remain.

  Next, the switching elements S33 and S44 of the charge recovery circuit 18 are turned on at time t5 when t4 <t5 <t6 during the dead time period in which the switching elements S1 to S4 of the inverter 11 are both turned off. Further, the switching elements S33 and S44 of the charge recovery circuit 18 are turned off at the timing of time t7 where t6 <t7 <t8 until the switching elements S3 and S4 of the inverter 11 are turned on at time t6 and turned off at time t8. To do.

  As described above, the timing time t5 when the switching elements S33 and S44 of the charge recovery circuit 18 are turned on is set to a period of t4 <t5 <t6, and the timing time t7 when the switching elements S33 and S44 are turned off is set to t6 <t7 <t8. The period is as follows. That is, each switching element breaks down due to an arm short circuit caused by simultaneously turning on the switching element S33 of the charge recovery circuit 18 and the switching element S2 of the inverter 11, and the switching element S44 of the charge recovery circuit 18 and the switching element S1 of the inverter 11. Is to prevent. An appropriate output timing of these pulses is determined by the pulse control circuit 17.

  As described above, since the charge recovery control pulses G33 and G44 are output at time t5, the switching elements S33 and S44 of the charge recovery circuit 18 are simultaneously turned on by the pulses G33 and G44 as shown in FIG. Therefore, during the period (t5 to t6), the charge is transferred from the capacitor 183 of the charge recovery circuit 18 to the stray capacitance C12 of the switching element S2 of the inverter 11 and charged, and the stray capacitance C13 of the switching element S3 of the inverter 11 is charged. The electric charge is discharged from. At the same time, the stray capacitance C14 of the switching element S4 of the inverter 11 is discharged, the charge is transferred to the capacitor 183 of the charge recovery circuit 18, and the charge is charged to the stray capacitance C11 of the switching element S1 of the inverter 11.

Thus, in a state where almost all the charges stored in the stray capacitances C13 and C14 of the switching elements S3 and S4 of the inverter 11 are discharged, the inverter control pulses G3 and G4 are output at time t6 and both switching elements S3 of the inverter 11 are output. , S4 is turned on. For this reason, there is no charge from the stray capacitances C13 and C14 passing through the channels of both switching elements S3 and S4, and the switching loss of these switching elements S3 and S4 is greatly reduced.
Then, during the period of (t6 to t8), when the switching elements S3 and S4 of the inverter 11 are both turned on, a current flows through the switching elements S3 and S4 as shown in FIG.

Next, a period of t8 <t <t10 from when switching elements S3 and S4 of inverter 11 are completely turned off at time t8 to when switching elements S1 and S2 of inverter 11 are turned on at time t10 It is a dead time period. The switching elements S11 and S22 of the charge recovery circuit 18 are turned on at time t9 during the dead time period. In the period of (t8 to t9), as shown in FIG. 13, the current of the inverter 11 starts to be regenerated, and at the same time, the charges of the stray capacitances C11 and C12 of the switching elements S1 and S2 start to be discharged. The stray capacitances C13 and C14 of the elements S3 and S4 start to be charged. When the charges of the stray capacitances C11 and C12 are discharged, the charging voltage gradually decreases. However, depending on the length of the regeneration time, the discharge becomes insufficient, so that the charges remain in the stray capacitances C11 and C12.
Thereafter, the same operation continues.

  In this way, the switching elements S11 to S44 constituting the charge recovery circuit 18 are turned on / off at a necessary timing, and the reactor 182 of the charge recovery circuit 18 and the stray capacitances C11 to C14 of the switching elements S1 to S4 of the inverter 11 , The charges charged in the stray capacitances C11 to C14 of the switching elements S1 to S4 of the inverter 11 can be input to and output from the capacitor 183 of the charge recovery circuit 18. Thus, in particular, the charging voltage of the stray capacitances C11 to C14 of the switching elements S1 to S4 can be reduced immediately before the switching elements S1 to S4 of the inverter 11 are turned on. For this reason, even when the number of times of switching increases when increasing the repetition frequency and pulse width of the laser pulse, the switching loss can be greatly reduced.

  Further, when the charge recovery circuit 18 is used as in the first embodiment, as shown in FIG. 7, when the switching elements S1 to S4 of the inverter 11 are turned on / off, a pulsed current is generated. The flow causes a slight current loss in the switching elements S11 to S44 and the diode 185 constituting the charge recovery circuit 18. However, most of the charges are reused for charging and discharging the stray capacitances C11 to C14 of the switching elements S1 to S4 of the inverter 11. Therefore, the loss generated in the charge recovery circuit 18 itself is small and the loss of the entire high-frequency power supply device 1 is reduced, so that a very high efficiency power source can be configured. As a result, when the capacity of the apparatus is increased, it is possible to avoid an increase in the number of cooling mechanisms of the apparatus and an increase in the structure of the apparatus itself, eliminating the need for paralleling many elements and circuits. A power supply apparatus that is very advantageous also in terms of installation space of the machine can be configured.

  Further, when such a charge recovery circuit 18 is used, the charge stored in the stray capacitances C11 to C14 is charged and discharged to the charge recovery capacitor 183 before the operation of the inverter 11, and therefore the switching element S1 of the inverter 11 is used. ~ S4 switching speed can be increased, and as a result, as shown in each period of (t1 to t2), (t5 to t6), (t9 to t10) in FIG. 7, the duty of the inverter output is increased by Δt, The inverter current Iiv for each pulse increases and the input power increases. When the laser is pulse-driven, it is difficult for current to flow at the beginning of discharge when discharge is started in the discharge space 7, and it is difficult to increase the laser output. However, by using this method, the inverter current Iiv can be increased from the beginning of the pulse, so that it is possible to increase the laser output of one pulse and one pulse, and to produce a laser pulse output with a very sharp rise. It becomes possible to obtain.

  In the example shown in FIG. 7, the diagonal switching elements S1, S2, and S3, S4 of the inverter 11 are simultaneously turned on / off. However, the switching elements S1, S2, and Even when S3 and S4 are not turned on / off at the same time, the switching elements S11 to S44 of the charge recovery circuit 18 corresponding to the switching elements S1 to S4 are turned on / off at the same timing as described above. An effect can be obtained.

  In the example shown in FIG. 7, the inverter 11 is continuously operated based on the inverter control pulses G1 to G4 without thinning out the command pulses A1 to A4. Even when the inverter control pulses G1 to G4 and the charge recovery control pulses G11 to G44 are generated by thinning out the above, the same effect as when the inverter 11 is operated continuously can be obtained.

  FIG. 14 shows the operation when the thinning circuit 16 thins out the command pulses A1 to A4 to generate the inverter control pulses G1 to G4 and the charge recovery control pulses G11 to G44. That is, in FIG. 14, after the inverter control pulses G1 to G4 and the charge recovery control pulses G11 to G44 are output during the period (t1 to t8), the command pulse is thinned out during the period (t8 to t9). Therefore, these control pulses are not output, and the inverter control pulses G1 to G4 and the charge recovery control pulses G11 to G44 are output again after the next time t9. In this case, since the discharge current peak and the output peak of the laser pulse change depending on the interval at which the command pulses A1 to A4 are thinned, the output peak from the laser oscillator 2 is taken into consideration, and how many pulses are thinned every other pulse is determined. Is done.

  When the command pulse A1 to A4 is thinned and operated by the thinning circuit 16, current does not flow through the inverter 11 during the thinned period (t8 to t9) (unlike continuous operation, stray capacitance charges are discharged by regeneration) Is not) period. In the case of the prior art, although the charging voltage is somewhat reduced by the resonance vibration immediately after the inverter is stopped, the stray capacitances C11 to C14 are maintained at the charging voltage immediately before the inverter is stopped (approximately the bus voltage of the inverter 11). Therefore, the switching loss is much larger than that in the case of continuous operation, and it is very difficult to widen the pulse width of the laser pulse under the condition that the pulses are thinned out.

  On the other hand, in the present invention, when the operation is performed by thinning out the pulses, the switching loss can be greatly reduced by using the charge recovery circuit 18, so that the repetition frequency and the pulse width of the laser pulse are greatly increased. It becomes possible to make it. In addition, since the switching loss can be greatly reduced, the cooling mechanism of the high-frequency power supply device is simplified, and it is possible to reduce the size of the device, the cost, and the installation space.

  In the example shown in FIG. 14, the case has been described in which the thinning circuit 16 thins out the command pulses A1 to A4 and simply generates the inverter control pulses G1 to G4 and the charge recovery control pulses G11 to G44. However, the present invention is not limited to this, and as shown in FIG. 15, among the inverter control pulses G <b> 1 to G <b> 4 output from the pulse control circuit 17, for example, inverters output corresponding to the two switching elements S <b> 3 and S <b> 4 of the inverter 11. The output of the control pulses G3 and G4 may be continued during the thinning period (t8 to t9) so that the ON state of the switching elements S3 and S4 may be continued.

  Thus, by continuing the ON state of the switching elements S3 and S4 during the thinning-out period (t8 to t9), the charging voltage of the stray capacitances C13 and C14 parasitic on the switching elements S3 and S4 becomes zero. On the other hand, the stray capacitances C11 and C12 of the other switching elements S1 and S2 are maintained in a state of being substantially charged with the bus voltage E. At this time, the capacitor 183 of the charge recovery circuit 18 is in a state of being charged to a voltage (1 / 2E) that is 1/2 of the bus voltage, and the voltage 1 / 2E that is the difference is a voltage at the time of LC resonance. In the charge recovery circuit 18, this voltage is the maximum voltage of the LC resonance, and when the inverter 11 is driven in a pulse manner, the most effective charge recovery can be obtained.

  As a result, the laser pulse width and the repetition frequency of the laser pulse can be greatly increased. In addition, since the switching loss that occurs can be greatly reduced, the cooling mechanism of the high-frequency power supply device is simplified, and it is possible to reduce the size of the device, the cost, and the installation space. .

  In addition to the operation of continuously turning on the switching elements S3 and S4 of the inverter 11, the peak of the inverter current may increase when the charge recovery circuit 18 is operated. In such a case, the on-duty of the switching elements S1 and S4 of the inverter 11 may be lowered. By reducing the on-duty, the peak of the inverter current can be lowered, the discharge power can be adjusted, and the laser output can be adjusted more finely.

  Further, FIGS. 7, 14, and 15 show examples in which the charge recovery control pulses G11 to G44 are always output before the output timing of the inverter control pulses G1 to G4. However, when the dead time period for the upper and lower switching elements S1, S4 and S2, S3 of the inverter 11 cannot be sufficiently obtained, the charge recovery corresponding to the inverter control pulses G1, G2 is performed as shown in FIG. It is also possible to output only the control pulses G11 and G22 and not output the charge recovery control pulses G33 and G44 corresponding to the inverter control pulses G3 and G4.

  As described above, by outputting the charge recovery control pulses G11 and G22 only immediately before the output timing of the first inverter control pulses G1 and G2 when the inverter 11 shifts from the stop state to the operation, a great effect on the switching loss can be obtained. Can be obtained.

  As described above, when the inverter 11 operates continuously, the electric charge is discharged during the regeneration period and the loss is reduced. However, when the inverter 11 is shifted from the stopped state to the operation, the floating capacitance This is because switching starts in a state where C11 and C12 are charged with a high voltage, resulting in a very large switching loss. Accordingly, even with the switching pattern as shown in FIG. 16, it is possible to greatly reduce the switching loss when the repetition frequency and pulse width of the laser pulse are greatly increased, and the cooling mechanism of the high frequency power supply device is simplified. It is possible to reduce the size of the apparatus, reduce the cost, reduce the installation space, and the like.

Embodiment 2. FIG.
FIG. 17 shows the configuration of the high-frequency power supply device according to the second embodiment of the present invention, and the same reference numerals are given to components corresponding to or corresponding to those of the first embodiment shown in FIGS.

  In the first embodiment, a single DC power supply 12 is attached to the inverter 11 and this is used as the bus voltage E of the inverter 11. In the second embodiment, the DC power source input to the inverter 11 is divided into two, and the midpoint of both DC power sources 12a and 12b is installed at the ground so that a voltage of ± E / 2 is input to the inverter 11. ing. Capacitors C1 and C2 are connected between the DC power supplies 12a and 12b and the inverter 11 in parallel with the DC power supplies 12a and 12b. A charge recovery circuit 18 is connected to the inverter 11.

In this case, the charge recovery circuit 18 can employ a configuration as shown in FIG. 18 or FIG. Compared with the configurations of FIGS. 4 and 5 of the first embodiment, the charge recovery capacitors 183 are omitted in the charge / discharge circuits 18a and 18b, and the reactor 182 side is connected to the output terminals X and Y of the inverter 11, respectively. The switching elements S11 to S44 (186) are connected to the ground, which is the midpoint of the DC power supplies 12a and 12b.
Since other configurations are the same as those in the first embodiment, detailed description thereof is omitted here.

  In such a configuration, since the midpoint of the two separated DC power supplies 12a and 12b is grounded, the maximum voltage with respect to the ground of each part of the circuit is equal to the voltage of the DC power supply 12 in the first embodiment. Suppressed in half. As a result, the withstand voltage design of the inverter 11 is facilitated, and the circuit can be miniaturized. In the first embodiment, the charge recovery capacitor 183 needs to be charged with a voltage half the bus voltage. In the second embodiment, the DC power supplies 12a and 12b are connected to each other. Since the capacitor C2 connected to the middle point and the negative side (N side) of the inverter 11 can be substituted, the charge recovery circuit 18 does not need a capacitor, and the charge recovery circuit 18 can be further downsized.

  The control operations of the inverter 11 and the charge recovery circuit 18 are the same as those in the first embodiment, which makes it possible to greatly increase the repetition frequency and pulse width of the laser pulse. In addition, since the switching loss can be greatly reduced, the cooling mechanism of the high-frequency power supply apparatus is simplified, and the apparatus can be downsized, the cost can be reduced, and the installation space can be reduced.

Claims (11)

A high frequency power supply device that outputs voltage pulses to a load,
An inverter for power conversion having a switching element; a first reactor connected to the load; and a charge recovery circuit connected to the output terminal of the inverter for recovering the charge of the stray capacitance of the switching element; A control means for controlling the operation of the inverter and the charge recovery circuit;
The charge recovery circuit has a switching element for starting the circuit, a second reactor for LC resonance with the stray capacitance, and a capacitor for charge recovery,
The control means outputs a series of inverter control pulses for controlling on / off of the switching element of the inverter and a charge recovery control pulse for starting the switching element of the charge recovery circuit in a dead time period of the switching element of the inverter. A high frequency power supply device characterized by that . A high frequency power supply device that outputs voltage pulses to a load,
An inverter for power conversion having a switching element, a first reactor connected to the load, a DC power supply for supplying DC power to the inverter, and a stray capacitance of the switching element connected to the output terminal of the inverter A charge recovery circuit for recovering the electric charge, and a control means for controlling the operation of the inverter and the charge recovery circuit,
The DC power source is divided into two, and the middle point is grounded, and a capacitor is connected in parallel to each DC power source between the DC power source and the inverter.
The charge recovery circuit includes a switching element for starting the circuit and a second reactor for LC resonance with the stray capacitance,
The control means outputs a series of inverter control pulses for controlling on / off of the switching element of the inverter and a charge recovery control pulse for starting the switching element of the charge recovery circuit in a dead time period of the switching element of the inverter. A high frequency power supply device characterized by that . 1/2 of the period of LC resonance determined by the stray capacitance of the second reactor and the inverter switching element, so that the charge collection time switching elements of the electrical charge recovery circuit is turned on, the second reactor The high frequency power supply device according to claim 1 or 2, wherein a value is set. The high frequency power supply device according to claim 1, wherein the capacitance value of the capacitor of the charge recovery circuit is set to be larger than the capacitance value of the stray capacitance of the switching element of the inverter. The series resonance frequency determined by the second reactor and the stray capacitance of the switching element of the inverter is the series resonance frequency determined by the inductance value of the sum of the first reactor and the second reactor and the dielectric capacitor of the load. The high frequency power supply device according to claim 1, wherein the high frequency power supply device is set higher than the first frequency. The high frequency power supply device according to claim 1 or 2, wherein an inductance value of the second reactor is set smaller than an inductance value of the first reactor. The inductance value of the first reactor is set so that the impedance of the stray capacitance of the switching element of the inverter is smaller than the series impedance of the first reactor and the load, and the inductance value of the first reactor and the The high frequency power supply device according to claim 1 or 2, wherein the frequency of the inverter is set so as to cause series resonance with a dielectric capacitor of a load. A thinning circuit is provided that thins out the inverter control pulse and the charge recovery control pulse to change the number of switching of each switching element of the inverter and the charge recovery circuit according to a control parameter based on an output characteristic of the voltage pulse. The high frequency power supply device according to any one of claims 1 to 7. The control means outputs the inverter control pulse so that the conduction period of the other diagonal switching element operating thereafter is longer than the conduction period of the diagonal switching element operating first of the inverter. The high frequency power supply device according to any one of claims 1 to 8, wherein the high frequency power supply device is one. The high frequency power supply device according to claim 1, wherein the switching element of the charge recovery circuit is formed of a wide band gap semiconductor. The high-frequency power supply device according to claim 10, wherein the wide band gap semiconductor is silicon carbide, a gallium nitride-based material, or diamond.

Images