Driving device for piezoelectric actuator

This application is based on and incorporates herein by reference Japanese Patent Application No. 2006-173883 filed on Jun. 23, 2006.

The present invention relates to a driving device for charging and discharging a piezoelectric actuator to expand and contract the piezoelectric actuator.

A conventional driving device for charging and discharging a piezoelectric actuator to expand and contract the piezoelectric actuator includes a charging circuit and a discharging circuit, as disclosed in JP 2005-16431A for example. In the charging circuit, electric power is supplied from a DC power source through a charging switch having MOSFET to a series circuit of an inductor and the piezoelectric actuator. The discharging circuit is connected to the series circuit in parallel. In the discharging circuit, electric charges charged in the piezoelectric actuator are discharged through a discharging switch having MOSFET.

In this driving device, when a driving signal is input from an external device to the driving device, the charging switch is repetitively turned on/off while the discharging switch is in the turned-off state, thereby charging the piezoelectric actuator and thus expanding the piezoelectric actuator. Then, when no driving signal is input, the discharging switch is repetitively turned on/off while the charging switch is in the turned-off state, thereby discharging the piezoelectric actuator and thus contracting the piezoelectric actuator.

That is, during the charging period, the piezoelectric actuator is charged in stepwise manner according to the following procedure. That is, the charging switch is turned on while the discharging switch is turned off, whereby charging current flows from the DC power source through the charging circuit into the piezoelectric actuator. Thereafter, the charging switch is turned off so that the charging current (flywheel current) flowing by means of the energy accumulated in an inductor is made to flow from the negative side of the piezoelectric actuator to the positive side thereof through a parasitic diode of MOSFET used as the discharging switch.

On the other hand, during the discharging period, the piezoelectric actuator is discharged in stepwise manner according to the following procedure. That is, the discharging switch is turned on while the charging switch is turned off to make the discharging current flow from the positive side of the piezoelectric actuator to the discharging circuit. Thereafter, the discharging switch is turned off so that charging power is regenerated to the DC power source by a flyback voltage generated by the energy accumulated in the inductor.

In this type of driving device, a desired charge amount is charged into the piezoelectric actuator by controlling the amount of charges per unit time which are supplied from the DC power source.

That is, when the driving signal is input to this driving device, the charging switch is turned on, and then turned off at the time point when the integration value of the charging current flowing into the piezoelectric actuator reaches a predetermined target value. Thereafter, the charging switch is turned on when the charging current reaches a smaller predetermined value than the target value concerned. This turn-on/off operation is repeated, thereby charging the piezoelectric actuator.

Since the electrostatic capacitance of the piezoelectric actuator varies in accordance with temperature, the current amount of charging current flowing into the piezoelectric actuator at the charging time varies in accordance with surrounding conditions under which the piezoelectric actuator is used.

When the electrostatic capacitance of the piezoelectric actuator is varied, the amount of current flowing from the DC power source into the piezoelectric actuator when the charging switch is turned on is varied, and thus the energy accumulated in the inductor at that time also varies. Therefore, the time required until the discharging of the energy from the inductor is completed under the state that the charging switch is turned off is also varied.

Therefore, during the charging period, the timing at which the charging switch is turned on is varied in accordance with the variation of the electrostatic capacitance of the piezoelectric actuator, and thus the variation amount of charges accumulated in the piezoelectric actuator (that is, the expansion rate at which the piezoelectric actuator expands per unit time) is varied in some cases.

For example, in a case where the conventional driving device is used to control the fuel injection of an injector so that fuel injection is started at the time when the piezoelectric actuator expands by a predetermined amount, the timing of the fuel injection of the injector is varied, and thus it is impossible to perform high-precision fuel injection.

The present invention therefore has an object to provide a driving device for a piezoelectric actuator that reduces variation of the expansion rate of a piezoelectric actuator due to variation of an electrostatic capacitance.

In order to attain the above object, a driving device for a piezoelectric actuator has: an inductor connected to the piezoelectric actuator in series so that an inductor and the piezoelectric actuator forms a series circuit; a charging circuit for supplying power from a DC power source through a charging switch to the series circuit; and a discharging circuit that is connected to the series circuit in parallel and discharging charges charged in the piezoelectric actuator through a discharging switch.

When a driving instruction is input from an external device, a charge/discharge control unit repeats the turn-on/off operation of the charging switch under the state that the discharging switch is turned off, thereby charging and expanding the piezoelectric actuator. Thereafter, when a driving stop instruction is input from the external device, the charge/discharge control unit repeats the turn-on/off operation of the discharging switch under the state that the charging switch is turned off, thereby discharging and contracting the piezoelectric actuator.

The charge/discharge control unit includes a target value setting section, a current detecting section, an estimating section and a charging switch driving section. The target value setting section sets a target value of charge energy to be charged in the piezoelectric actuator during ON-period of the charging switch on the basis of the charge energy to be charged in the piezoelectric actuator during one charging period by the driving device, and a power supply voltage of the DC power source.

When the driving instruction is input, the charging switch driving section turns on the charging switch at a preset period, and turns off the charging switch every time an estimation value estimated by the estimating section reaches the target value set by the target value setting section.

The period when the charging switch driving section turns on the charging switch, and the charging amount of the piezoelectric actuator in one turn-on operation of the charging switch can be fixed. Therefore, the expansion rate of the piezoelectric actuator in the charging period can be set to a desired expansion rate.

Accordingly, even when the electrostatic capacitance of the piezoelectric actuator is varied due to temperature variation or the like, the expansion rate of the piezoelectric actuator can be prevented from being varied.

Furthermore, when the driving device is applied to fuel injection control for a fuel injector that starts the fuel injection at the time when the piezoelectric actuator expands by a predetermined amount, the variation of the fuel injection timing of the injector is reduced even when the electrostatic capacitance varies.

The above and other objects, features and advantages of the present invention will become more apparent from the following detailed description made with reference to the accompanying drawings. In the drawings:

FIG. 1 is a circuit diagram showing a driving device for a piezoelectric actuator according to a first embodiment of the present invention;

FIG. 2 is a circuit diagram showing a part of a switch controller which controls a charging switch;

FIG. 3 is a flowchart showing the processing executed by CPU of a charging switch controller of the driving device according to the first embodiment;

FIG. 4 is a time chart showing the operation of each part in a charging operation in the driving device according to the first embodiment;

FIG. 5 is a circuit diagram showing a driving device according to a second embodiment of the present invention;

FIG. 6 is a time chart showing the operation of each part in the charging operation in the driving device according to the second embodiment;

FIG. 7 is a flowchart showing the processing executed by CPU of a charging switch controller of a driving device according to a third embodiment of the present invention;

FIG. 8 is a time chart showing the operation of each part in the charging operation in the driving device of the third embodiment;

FIG. 9 is a flowchart showing the processing executed by CPU of a charging switch controller of a driving device according to a fourth embodiment of the present invention;

FIG. 10 is a time chart showing the operation of each part in the charging operation in the driving device according to the fourth embodiment;

FIG. 11 is a time chart showing the operation of each part in the charging operation in a driving device according to a fifth embodiment of the present invention;

FIG. 12 is a flowchart showing the processing executed by CPU of a charging switch controller of a driving device according to a sixth embodiment of the present invention;

FIG. 13 is a time chart showing the operation of each part in the charging operation in the driving device according to the sixth embodiment; and

FIG. 14 is a circuit diagram showing a driving device of a modification of the fourth and fifth embodiments.

First Embodiment

Referring first to FIG. 1 , a driving device 1 is designed to charge/discharge four piezoelectric actuators P 1 to P 4 mounted in fuel injectors for injecting fuel in an internal combustion engine of a vehicle by expanding/contracting the piezoelectric actuators, thereby making each fuel injector provided on each cylinder of the engine start/stop the fuel injection.

The driving device 1 includes an inductor L 1 that is connected in series to a parallel circuit 10 having the four piezoelectric actuators P 1 to P 4 connected to one another in parallel, a power supply circuit 12 for boosting a battery voltage from a storage battery 14 to generate a power supply voltage, a charging circuit 30 supplying electric power from the power supply circuit 12 through a charging switch SW 1 to a series circuit 20 having the inductor L 1 and the parallel circuit 10 , a discharging circuit 40 that is connected to the series circuit in parallel to discharge charges charged in the piezoelectric actuators P 1 to P 4 through a discharging switch SW 2 , a current detecting resistor R 1 for detecting current supplied from the power supply circuit 12 to the piezoelectric actuators P 1 to P 4 at the charging time, and a switch controller 50 for controlling the turn-on/off operations of the charging switch SW 1 and the discharging switch SW 2 .

The parallel circuit 10 includes a first series circuit having the piezoelectric actuator P 1 and a cylinder selecting switch SWa, a second series circuit having the piezoelectric actuator P 2 and a cylinder selecting switch SWb, a third series circuit having the piezoelectric actuator P 3 and a cylinder selecting switch SWc, and a fourth series circuit having the piezoelectric actuator P 4 and a cylinder selecting switch SWd. These series circuits are connected to one another in parallel.

One end portion of the parallel circuit 10 at the piezoelectric actuator (P 1 to P 4 ) side is connected to one end of the inductor L 1 , and the other end portion of the parallel circuit 10 at the cylinder selecting switch (SWa to SWd) side is connected to a ground line which is connected to the negative-side of the power supply circuit 12 .

Each of the cylinder selecting switches SWa to SWd is an n-channel MOSFET; The source of each n-channel MOSFET is connected to the ground line, the drain thereof is connected to one end of the corresponding piezoelectric actuator (P 1 to P 4 ), and the gate thereof is connected to an output terminal of the switch controller 50 .

A diode Da is connected to the cylinder selecting switch SWa in parallel, a diode Db is connected to the cylinder selecting switch SWb in parallel, a diode Dc is connected to the cylinder selecting switch SWc in parallel, and a diode Dd is connected to the cylinder selecting switch SWd in parallel.

The anode of each of the diodes Da to Dd is connected to the ground line, and the cathode thereof is connected to the connection point between each piezoelectric actuator (P 1 to P 4 ) and the drain of each cylinder selecting switch (SWa to SWd). By using MOSFETs for the cylinder selecting switch SWa to SWd, the parasitic diodes of the MOSFETs may be used as the diodes Da to Dd.

One end portion of the series circuit 20 at the inductor L 1 side is connected to the positive-side of the power supply circuit 12 through the charging switch SW 1 . The charging switch SW 1 is an n-channel MOSFET. The source of the n-channel MOSFET is connected to one end of the inductor L 1 , the drain thereof is connected to the positive-side of the power supply circuit 12 , and the gate thereof is connected to an output terminal of the switch controller 50 .

Furthermore, a diode D 1 is connected to the charging switch SW 1 in parallel, and the diode D 1 serves to allow regenerated current to flow into the power supply circuit 12 , when the discharging switch SW 2 is turned off at the discharging time of the piezoelectric actuators P 1 to P 4 . By using MOSFET as the charging switch SW 1 , the parasitic diode of the MOSFET may be used as the diode D 1 .

The charging switch SW 2 is an n-channel MOSFET The source of the n-channel MOSFET is connected to the ground line, the drain thereof is connected to the connection point between the charging switch SW 1 and the inductor L 1 , and the gate thereof is connected to an output terminal of the switch controller 50 .

Furthermore, a diode D 2 is connected to the discharging switch SW 2 in parallel, and the diode D 2 serves to allow flywheel current to flow, when the charging switch SW 1 is turned off at the charging time of the piezoelectric actuators P 1 to P 4 . By using MOSFET for the discharging switch SW 2 , the parasitic diode of the MOSFET may be used as the diode D 2 .

The power supply circuit 12 has a boosting circuit 16 for boosting the battery voltage (for example, 12V) from the battery 14 . The negative-side terminal of the battery 14 is connected to the ground line.

The boosting circuit 16 has a boosting inductor L 2 whose one end is connected to the positive-side terminal of the battery 14 , a diode D 3 whose anode is connected to the other end of the boosting inductor L 2 and whose cathode is connected to the drain of the charging switch SW 1 , a boosting switch SW 3 which is connected to the connection point between the diode D 3 and the boosting inductor L 2 at one end thereof and also connected to the ground line at the other end thereof, a boosting control circuit 18 for turning on/off the boosting switch SW 3 , and a buffer capacitor C 1 which is connected to the connection point between the diode D 3 and the charging switch SW 1 at one end thereof and also connected to the ground line through the current detecting resistor R 1 at the other end thereof. The diode D 3 is used to prevent discharge of the buffer capacitor C 1 .

The boosting control circuit 18 repetitively turns on/off the boosting switch SW 3 to charge a voltage of several tens to several hundreds volts into the buffer capacitor C 1 through the diode D 3 . The boosting control circuit 18 is designed to detect the voltage value V of the buffer capacitor C 1 , and outputs a voltage value signal S representing the voltage value V to the switch controller 50 .

When the charging switch SW 1 and the cylinder selecting switch SWa to SWd are turned on, the buffer capacitor C 1 makes current flow into the charging circuit 30 to charge the piezoelectric actuator (P 1 to P 4 ) corresponding to the turned-on cylinder selecting switch (SWa to SWd).

Here, the current flowing from the buffer capacitor C 1 into the series circuit 20 when the piezoelectric actuator P 1 to P 4 is charged is called as a buffer capacitor current Ic 1 . Furthermore, the buffer capacitor C 1 is designed to have a relatively large electrostatic capacitance, and it keeps a substantially fixed voltage value when power is supplied to the piezoelectric actuators P 1 to P 4 .

When a driving signal Sd output from an external ECU (not shown) for performing the engine control of the vehicle is input to the switch controller 50 , the switch controller 50 turns on the cylinder selecting switch SWa to SWd which is specified by the driving signal Sd, and also repetitively turns on/off the charging switch SW 1 for a predetermined time under the state that the discharging switch SW 2 is turned off.

At this time, the switch controller 50 periodically turns on the charging switch SW 1 , and also integrates the buffer capacitor current Ic 1 with time. When the integration value reaches a predetermined value, the switch controller 50 turns off the charging switch SW 1 .

When no driving signal Sd is input, the switch controller 50 repetitively turns on/off the discharging switch SW 2 under the state that the charging switch SW 1 is turned off.

Next, a part of the switch controller 50 which controls the charging switch SW 1 will be described in detail with reference to FIG. 2 . The detailed description on a part of the switch controller 50 which controls the discharging switch SW 2 is omitted because it is not directly associated with the present invention.

The switch controller 50 has a current detecting circuit 60 for detecting the buffer capacitor current Ic 1 , a current integrating circuit 70 for integrating the buffer capacitor current detected by the current detecting circuit 60 with time, a determining circuit 80 for determining whether the integration result calculated by the current integrating circuit 70 reaches a target charge amount Q, and a charging switch controller 90 for turning on the charging switch SW 1 periodically when the driving signal Sd is input and turning off the charging switch SW 1 on the basis of the determination result of the determining circuit 80 .

The current detecting circuit 60 has resistors R 2 , R 3 , R 4 and R 5 and an operational amplifier OP 1 . The connection point between the current detecting resistor R 1 and the buffer capacitor C 1 is connected through the resistor R 2 to the non-inverting input terminal (+) of the operational amplifier OP 1 , the connection point between the current detecting resistor R 1 and the ground line is connected through the resistor R 3 to the inverting input terminal (−) of the operational amplifier OP 1 .

Thus, a current detection signal Sc is output from the output terminal of the operational amplifier OP 1 . The current detection signal Sc represents a voltage value acquired by amplifying the potential difference between both the ends of the current detecting resistor R 1 at a predetermined gain. A reference voltage Vref (for example, 2.5[V]) for setting the offset of the buffer capacitor current is input through the resistor R 5 to the non-inverting input terminal (+) of the operational amplifier OP 1 .

The current integrating circuit 70 has an integrating circuit 72 for integrating the current detection signal Sc with time, and an integration reset switch SW 4 for resetting the integration result of the integrating circuit 72 . The integration circuit 72 has a resistor R 6 , a capacitor C 2 and an operational amplifier OP 2 . The reference voltage Vref is input to the non-inverting input terminal (+) of the operational amplifier OP 2 , and the current detection signal Sc is input through the resistor R 6 to the inverting input terminal (−) of the operational amplifier OP 2 . Thus, a signal acquired by integrating the current detection signal Sc with time (hereinafter referred to as integration signal Si) is output from the output terminal of the operational amplifier OP 2 . The integration reset switch SW 4 is connected to the capacitor C 2 in parallel, and it is turned on to discharge the charges accumulated in the capacitor C 2 .

The determining circuit 80 has resistors R 7 , R 8 , R 9 and a comparator CMP 1 . The integration signal Si is input through the resistor R 7 to the inverting input terminal (−) of the comparator CMP 1 , and the target charge amount Q (specifically, a target charge amount setting signal So) output from the charging switch controller 90 is input through the resistor R 8 to the non-inverting input terminal (+) of the comparator CMP 1 .

Furthermore, the determining circuit 80 is constructed as a comparator CMP 1 with hysteresis (Schmidt trigger) for changing over the magnitude of the target charge amount Q between two values in accordance with the output signal of the comparator CMP 1 . That is, the comparator CMP 1 adds the output signal from the output terminal to the target charge amount Q by the resistor R 9 connected to the output terminal and the non-inverting input terminal (+), and switches over the target charge amount input to the non-inverting input terminal (+), that is, to a value larger than the target charge amount Q and a value smaller than the target charge amount Q.

In the determining circuit 80 , where the signal level of the output signal of the comparator CMP 1 (determination signal Sh) is high, the signal level of the determination signal Sh becomes low when the charge amount represented by the integration signal Si exceeds the value by adding the target charge amount Q with the amount corresponding to the hysteresis. Conversely, the signal level of the determination signal Sh becomes high when the charge amount represented by the integration signal Si is lower than the value acquired by subtracting the amount corresponding to the hysteresis from the target charge amount Q.

The charging switch controller 90 is mainly constructed by CPU, RAM, ROM, etc., and the driving signal Sd, the determination signal Sh, the voltage value signal Sv, etc. are input to the charging switch controller 90 .

Furthermore, ROM of the charging switch controller 90 stores one charging period Tc, an energy instruction value E representing the charge energy to be charged to the piezoelectric actuators P 1 to P 4 during the charging period Tc, a switching period Ts representing the turn-on period of the charging switch SW 1 in the charging period Tc, a switching frequency N representing the frequency at which the charging switch SW 1 is turned on during the charging period Tc, etc. The switching period Ts is fixed.

The processing executed by CPU of the charging switch controller 90 will be described with reference to the flowchart of FIG. 3 .

When CPU of the charging switch controller 90 is started upon turn-on of an ignition switch of a vehicle, as shown in FIG. 3 , the energy instruction value E, the charging period Tc, the switching period Ts and the switching frequency N are first read in from ROM in S 110 , and also the voltage value V of the buffer capacitor C 1 is acquired from the boosting control circuit 18 .

Subsequently, in S 120 , the energy instruction value E is divided by the voltage value V to calculate the total charge amount Qt to be discharged from the buffer capacitor C 1 within the charging period Tc (Qt=E/V). Then, in S 130 , the total charge amount Qt calculated in S 120 is divided by the switching frequency N, thereby calculating a target charge amount Q to be discharged from the buffer capacitor C 1 during one ON-period of the charging switch SW 1 (Q=Qt/N). In S 140 , an integration reset signal Sr for turning on the integration reset switch SW 4 is output for a predetermined time. Subsequently, in S 150 , a target charge amount setting signal So representing the target charge amount Q calculated in S 130 is output to the non-inverting input terminal (+) of the determining circuit 80 (the comparator CMP 1 ).

Subsequently, it is checked in S 160 whether the signal level of the determination signal Sh is high (Sh=high) or not. If it is determined in S 160 that the signal level of the determination signal Sh is not high, the processing of S 160 is repetitively executed, and the processing proceeds to S 170 if it is determined that the signal level of the determination signal Sh is high.

It is checked in S 170 whether the driving signal Sd is input or not. If it is determined in S 170 that no driving signal Sd is input, the processing of S 170 is repetitively executed. Conversely, if it is determined that the driving signal Sd is input, the processing proceeds to S 180 .

In S 180 , the measurement of the charging period Tc is started, and also a cylinder selecting switching signal is output to the gate of the cylinder selecting switch (SWa to SWd) which is specified by the driving signal Sd so that the specified cylinder selecting switch (SWa to SWd) is turned on.

Then, in the subsequent step S 190 , the measurement of the switching period Ts is started. Subsequently, in S 200 , a switching period St representing the timing at which the charging switch SW 1 should be turned on is output for a predetermined time.

In S 210 , a charging switching signal Ss for turning on the charging switch SW 1 is output to the gate of the charging switch SW 1 . Then, it is checked in S 220 whether the signal level of the determination signal Sh is low or not.

If it is determined in S 220 that the signal level of the determination signal Sh is not low, the processing of S 220 is repetitively executed. Conversely, if it is determined that the signal level of the determination signal Sh is low, the processing proceeds to S 230 .

In S 230 , the output of the charging switch Ss is stopped, and in the subsequent step S 240 , the integration reset signal Sr is output for a predetermined time. Then, in S 250 , it is checked whether the signal level of the determination signal Sh is high or not.

If it is determined in S 250 that the signal level of the determination signal Sh is not high, the processing of S 250 is repetitively executed. Conversely, if it is determined that the signal level of the determination signal Sh is high, the processing proceeds to S 260 .

In S 260 , it is checked whether the switching period Ts has elapsed from the start of the measurement of the switching period Ts in S 190 . If it is determined in S 260 that the switching period Ts has not yet elapsed, the processing of S 260 is repetitively executed. Conversely, if it is determined that the switching period Ts has elapsed, the processing proceeds to S 270 .

In S 270 , it is checked whether the charging period Tc has elapsed (been finished) from the start of the measurement of the charging period Tc in S 180 . If it is determined in S 270 that the charging period Tc has not yet been finished, the processing proceeds to S 190 . Conversely, if it is determined that the charging period Tc has been finished, the output of the cylinder selecting switching signal is stopped, and the processing proceeds to S 110 .

Next, the operation of each part when the driving device 1 charges the piezoelectric actuators P 1 to P 4 will be described with reference to FIG. 4 . All the piezoelectric actuators P 1 to P 4 execute substantially the same operation, and thus the piezoelectric actuator P 1 will be representatively described in the following description.

As shown in FIG. 4 , in the driving device 1 , when the driving signal Sd is input to the switch controller 50 at time ti (S 17 O: YES), the switching period signal St and the charging switching signal Ss are output (S 200 , S 210 ), and also the cylinder selecting switching signal is output to the cylinder selecting switch SWa.

In this case, the charging switch SW 1 is turned on to make the charging circuit 30 conductive, so that buffer capacitor current Ic 1 flows from the buffer capacitor C 1 through the charging circuit 30 to the piezoelectric actuator P 1 .

The buffer capacitor current Ic 1 gradually increases on the basis of the time constant determined by the inductor L 1 and the capacitance of the piezoelectric actuator P 1 , and thus the charge amount represented by the integration signal Si also increases.

When at time t 2 the charge amount represented by the integration signal Si reaches the target charge amount Q (the value acquired by adding the target charge amount Q with the amount corresponding to the hysteresis), the output level of the determination signal Sh is switched to low level (S 220 : YES), and the output of the charging switching signal Ss is stopped (S 230 ).

At this time, the charging switch SW 1 is turned off to block the charging circuit 30 , and a flywheel current flows to the piezoelectric actuator P 1 through the parasitic diode D 2 by the energy accumulated through the inductor L 1 .

Then, the flywheel current is gradually reduced through the discharge of the energy accumulated in the inductor L 1 .

Furthermore, at the time t 2 , the charging switch controller 90 outputs the integration reset signal Sr for a predetermined time, so that the integration reset switch SW 4 is turned on and the charges accumulated in the capacitor C 2 is discharged (that is, the integration signal Si is reset).

When the switching period signal St is output again at time t 3 at which the fixed switching period Ts has elapsed from the time t 1 , the same operation as the time t 1 is performed. When the charge amount represented by the integration time Si reaches the target charge amount Q at time t 4 , the same operation as the time t 2 is performed.

The operation that the charging switch SW 1 is turned on at the period corresponding to each switching period Ts and the charging switch SW 1 is turned off every time the integration signal Si reaches the target charge amount Q is repeated until a time t 5 at which the charging period Tc has elapsed from the time t 1 .

One-dotted chain line of FIG. 4 represents a case where in place of the piezoelectric actuators P 1 to P 4 , piezoelectric actuators having a medium capacitance Cm smaller than the electrostatic capacitance (large capacitance C 1 represented by a solid line) of the piezoelectric actuators P 1 to P 4 are used as driving targets in the driving device 1 . Two-dotted chain line of FIG. 4 represents a case where piezoelectric actuators having further smaller electrostatic capacitance Cs than the electrostatic capacitance of the piezoelectric actuators used in the case indicated by the one-dotted chain line of FIG. 4 are used as driving targets in the driving device 1 .

That is, in FIG. 4 , the case where the piezoelectric actuators P 1 to P 4 having large-level electrostatic capacitance C 1 are used in the driving device 1 is represented by the solid line, the case where the piezoelectric actuators P 1 to P 4 having medium electrostatic capacitance Cm are used in the driving device is represented by one-dotted chain line, and the case where the piezoelectric actuators P 1 to P 4 having small-level electrostatic capacitance Cs are used in the driving device is represented by two-dotted chain line.

In FIG. 4 , the signals Ss, Sh, Sr when the electrostatic capacitance of the piezoelectric actuators P 1 to P 4 is in the medium level and in the small level are omitted from the illustration because the time chart of FIG. 4 would be difficult to be visually understood if all the signals are illustrated.

Even when the piezoelectric actuators P 1 to P 4 which are different in electrostatic capacitance from one another are used, the output start timing of the charging period signal St is not varied. Therefore, with respect to the signals Ss, Sh, Sr when the electrostatic capacitance is in the medium level or small level, the output stop timing of the charging switching signal Ss and the variation patterns of the signals Sh, Sr are delayed as compared with those when the electrostatic capacitance is in the large level, as in the case of the buffer capacitance current Ic 1 and the variation pattern of the integration signal Si.

As described above, in the driving device 1 , during the charging period Tc from the time at which the driving signal Sd is input till the time when the driving signal Sd is not input, when the driving signal Sd is input, the charging switch SW 1 is turned on at a fixed period, and also the charging switch SW 1 is turned off every time the charge amount represented by the integration signal Si reaches the target charge amount Q.

Therefore, according to the driving device 1 , the switching period Ts when the charging switch SW 1 is turned on during the charging period Tc is fixed, and the charge energy in one turn-on operation of the charging switch SW 1 can be also fixed irrespective of temperature variation. Therefore, the expansion rate of the piezoelectric actuators P 1 to P 4 (that is, the expanding behavior of the piezoelectric actuators P 1 to P 4 through one charging operation) during the charging period Tc can be set to a desired expansion rate.

Accordingly, even when the electrostatic capacitance of the piezoelectric actuators P 1 to P 4 is varied like Cl, Cm, Cs due to temperature variation or the like and, as a result, the piezoelectric actuator current Ip, the charge energy Ep and the piezoelectric actuator voltage Vp vary as shown in FIG. 4 , the expansion rate of the piezoelectric actuators P 1 to P 4 can be prevented from being varied.

Furthermore, according to the driving device 1 , the timing of the fuel injection of the injector can be prevented from being varied due to the variation of the electrostatic capacitance.

Still furthermore, if the integration value of the buffer capacitor current Ic 1 flowing from the buffer capacitor C 1 to the piezoelectric actuators P 1 to P 4 is calculated, the charge energy Ep to be charged in the piezoelectric actuators P 1 to P 4 per unit time can be derived. Therefore, in the driving device 1 of the first embodiment, when the integration signal Si reaches the target charge amount Q, the charging switch SW 1 is turned off.

Therefore, the charge energy Ep can be estimated by merely deriving the integration value of the buffer capacitor current Ic 1 , and thus the construction of the driving device 1 can be simplified.

Furthermore, according to the driving device 1 , the target value Q is determined on the basis of the voltage value of the buffer capacitor C 1 and the energy instruction value E, and thus the expansion rate of the piezoelectric actuators P 1 to P 4 can be set to be identical even when the voltage value V of the buffer capacitor C 1 varies.

In the first embodiment, the power supply circuit 12 operates as a DC power source, the parasitic diode D 1 operates as a first diode, and the parasitic diode D 2 operates as a second diode. The input of the driving signal Sd is a driving instruction, and the stop of the driving signal Sd is a driving stop instruction. The switch controller 50 operates as a charge/discharge control section, the current detecting resistor R 1 operates as a current detecting section, the current integration circuit 70 operates as an estimating section, and the processing of S 170 to S 270 operates as a charge switch driving section. The processing of 5120 and S 130 operates as a target value setting section.

Second Embodiment

The driving device 1 of the second embodiment is different from the first embodiment in the connecting position of the current detecting resistor R 1 and the stop timing of the output of the integration reset signal Sr.

Specifically, the current detecting resistor R 1 is connected between the end portion of the parallel circuit 10 at the cylinder selecting switch (SWa to SWd) side and the ground line.

In such a case, not only the buffer capacitor current, but also the flywheel current flowing at the turn-off time of the charging switch SW 1 in the charging period Tc flow through the current detecting resistor R 1 , and thus the current integration circuit 70 integrates not only the buffer capacitor current, but also the flywheel current.

Therefore, according to the second embodiment, in order to prevent the flywheel current from being integrated, the output stop timing of the integration reset signal Sr is set to the time when the switching period signal St is next output, at time t 3 for example, as shown in FIG. 6 .

Specifically, the output of the integration reset signal Sr is stopped when the charging switch controller 90 finishes the processing of S 200 .

According to the driving device 1 of the second embodiment, irrespective of the temperature variation, the charge energy in one turn-on operation of the charging switch SW 1 can be fixed while the switching period Ts when the charging switch SW 1 is turned on is fixed, and thus the same advantage as the first embodiment can be acquired.

Third Embodiment

The driving device of the third embodiment is different from the driving device 1 of the first embodiment so that the target charge amount Q is not fixed among the switching periods Ts in the charging period Ts. In this case, the target charge amount Q 1 in the first ON-period of the charging switch SW 1 is smaller than the target charge amount Q of the second and subsequent ON-periods.

That is, the target charge amount Q 1 is stored in ROM of the charging switch controller 90 , and the charging switch controller 90 calculates the target charge amount Q of the second and subsequent ON-periods of the charging switch SW 1 on the basis of the target charge amount Q 1 and the total charge amount Qt.

The charging switch controller 90 controls the turn-off of the charging switch SW 1 by using the target charge amount Q 1 in the first ON-period of the charging switch SW 1 , and also controls the turn-off of the charging switch SW 1 by using the target charge amount Q larger than the target charge amount Q 1 in the second and subsequent ON-periods of the charging switch SW 1 .

Next, the processing executed by the charging switch controller 90 of the driving device of the third embodiment will be described with reference to FIG. 7 .

When the charging switch controller of the driving device of the third embodiment starts the processing, the energy instruction value E, the charging period Tc, the switching period Ts and the switching frequency N are first read from ROM in S 310 , and also the voltage value V of the buffer capacitor C 1 is acquired from the boosting control circuit 18 .

Subsequently, in S 320 , the energy instruction value E is divided by the voltage value V to calculate the total charge amount Qt (Qt=E/V). In S 330 , the target charge amount Q 1 in the first ON-period of the charging switch SW 1 is read from ROM. Subsequently, in S 340 , as is apparent from the following equation (1), the value (Qt−Q 1 ) acquired by subtracting the target charge amount Q 1 from the total charge amount Qt is divided by the value (N−1) acquired by subtracting 1 from the switching frequency N, thereby calculating the target charge value Q in each of the second and subsequent ON-periods of the charging switch SW 1 .

Q= ( Qt−Q 1)/( N− 1)  (1)

In S 350 , the integration reset signal Sr is output for a predetermined time. Subsequently, in S 360 , a target charge amount signal So representing the target charge amount Q 1 in the first ON-period of the charging switching Sw 1 is output.

Subsequently, it is checked in S 370 whether the signal level of the determination signal Sh is high or not. If it is determined in S 370 that the signal level of the determination signal Sh is not high, the processing of S 370 is repetitively executed. Conversely, if it is determined in S 370 that the signal level of the determination signal Sh is high, the processing proceeds to S 380 .

In S 380 , it is checked whether the driving signal Sd is input or not. Then, if it is determined in S 380 that no driving signal Sd is input, the processing of S 380 is repetitively executed. Conversely, if it is determined that the driving signal Sd is input, the processing proceeds to S 390 . In S 390 , the measurement of the charging period Ts is started, and the cylinder selecting switching signal is output. In the subsequent S 400 , the measurement of the switching period Ts is started. Subsequently, in S 410 , the switching period signal St is output for a predetermined time, and in the subsequent S 420 , the charging switching signal Ss is output.

Subsequently, in S 430 , it is checked whether the signal level of the determination signal Sh is low or not. Then, if it is determined in S 430 that the signal level of the determination signal Sh is not low, the processing of S 430 is repetitively executed. Conversely, if it is determined in S 430 that the signal level of the determination signal Sh is low, the processing proceeds to S 440 .

In S 440 , the output of the charging switching signal Ss is stopped, and in the subsequent S 450 , the integration reset signal Sr is output for a predetermined time. Then, in S 460 , the target charge amount setting signal So representing the target charge amount Q in each of the second and subsequent ON-periods of the charging switch SW 1 is output.

In S 470 , it is checked whether the signal level of the determination signal Sh is high or not. If it is determined in S 470 that the signal level of the determination signal Sh is not high, the processing of S 470 is repetitively executed. Conversely, if it is determined in S 470 that the signal level of the determination signal Sh is high, the processing proceeds to S 480 .

In S 480 , it is checked whether the switching period Ts has elapsed or not from the time when the measurement of the switching period Ts is started in S 400 . If it is determined in S 480 that the switching period Ts has not yet elapsed, the processing of S 480 is repetitively executed. Conversely, if it is determined in S 480 that the switching period has elapsed, the processing proceeds to S 490 .

In S 490 , it is checked whether the charging period Tc has elapsed (finished) or not from the time when the measurement of the charging period Tc is started in S 390 . If it is determined in S 490 that the charging period Tc has not yet elapsed, the processing returns to S 400 . Conversely, if it is determined in S 490 that the charging period Tc has elapsed (finished), the output of the cylinder selecting switching signal is stopped, and the processing proceeds to S 310 .

Next, the operation of each part when the driving device of the third embodiment charges the piezoelectric actuators P 1 to P 4 will be described with reference to FIG. 8 .

In the driving device of the third embodiment, first, the target charge amount Q 1 is input to the non-inverting input terminal (+) of the determining circuit 80 (S 360 ), and when the driving signal Sd is input to the switch controller 50 at time t 6 as shown in FIG. 8 (S 380 : YES), the switching period signal St and the charging switching signal Ss are output (S 410 , S 420 ), and the cylinder selecting switching signal is output to the cylinder selecting switch SWa.

As described above, the buffer capacitor current flows from the buffer capacitor C 1 through the charging circuit 30 to the piezoelectric actuator P 1 . Then, when at time t 7 the charge amount represented by the integration signal Si reaches the target charge amount Q 1 , the output level of the determination signal Sh is switched to low (S 430 : YES), the output of the charging switch signal Ss is stopped (S 440 ).

At this time, the flywheel current flows through the parasitic diode D 2 into the piezoelectric actuator P 1 as described above. Furthermore, at time t 7 , the charging switching controller 90 outputs the integration reset signal Sr for a predetermined time, and the target charge amount Q larger than the target charge amount Q 1 is input to the non-inverting input terminal (+) of the determining circuit 80 (S 460 ).

Then, when the switching period signal St is input again at time t 8 , the same operation as the time t 6 is carried out, and when the charge amount represented by the integration signal Si reaches the target charge amount Q at time t 9 , the same operation as the time t 7 is carried out.

Thereafter, the operation of turning on the charging switch SW 1 every switching period Ts and also turning off the charging switch SW 1 every time the integration signal Si reaches the target charge amount Q is repeated until a time t 10 .

According to the driving device of the third embodiment described above, the switching period Ts when the charging switch SW 1 is turned on in the charging period Tc is fixed, and the charge energy in the first turn-on operation of the charging switch SW 1 can be fixed irrespective of temperature variation, so that the same advantage as the first embodiment can be acquired.

Furthermore, in the driving device of the third embodiment, the target charge amount every switching period Ts in the charging period Tc is set so that the target charge amount Q 1 in the first ON-period of the charging switch SW 1 is smaller than the target charge amount Q of each of the second and subsequent ON-periods. Therefore, the charging energy to be accumulated into the piezoelectric actuators P 1 to P 4 at the charging start time (first ON-period of the charging switch SW 1 ) can be reduced to be less than the charging energy when the target charge amount Q of each switching period Ts is set to be identical among the respective switching periods Ts (one-dotted chain line of FIG. 8 ).

Therefore, the load imposed on the piezoelectric actuators P 1 to P 4 at the charging start time can be reduced, and thus the lifetime and reliability of the piezoelectric actuators can be enhanced. In addition, occurrence of operating sounds can be suppressed.

In the driving device of the third embodiment, the switching period Ts is set so as to become shorter every time the charging switch SW 1 is turned on. However, the switching period Ts may be set to be fixed as in the case of the first and second embodiments.

Furthermore, the target charge amount Q 1 in the first ON-period of the charging switch SW 1 may be set to any value insofar as it is smaller than the target charge amount Q in the second and subsequent ON-periods.

In the third embodiment, the target charge amount Q 1 in the first ON-period of the charging switch SW 1 is stored in ROM. However, the target charge amount Q 1 may be set and changed by the user or the like.

Fourth Embodiment

Next, a driving device according to a fourth embodiment will be described with reference to FIGS. 9 and 10 .

The driving device of the fourth embodiment is different from the driving device 1 of the first embodiment in that the switching period Ts is not fixed, but it is set to be shorter every time the charging switch SW 1 is once turned on.

Specifically, the switching period Ts(n) of the fourth embodiment is set on the basis of a value which is acquired by an experiment or the like in advance so that no flywheel has been already prevented from flowing in the discharge circuit 40 at the output time of the switching period signal St (the turn-on time of the charging switch SW 1 ) even when the electrostatic capacitance of the piezoelectric actuators P 1 to P 4 varies).

In the case where the piezoelectric actuators P 1 to P 4 having large electrostatic capacitance are used in the driving device, the period at which the flywheel current flows in the discharge circuit 40 during charging operation is measured by the experiment or the like in advance, and the switching period Ts(n) is set on the basis of the measurement result.

The processing executed by the charging switch controller 90 of the driving device according to the fourth embodiment will be described hereunder with reference to FIG. 9 .

When the charging switch controller 90 of the driving device starts the processing of FIG. 9 , in S 510 , the energy instruction value E, the charging period Tc, the switching period Ts(n) and the switching frequency N are read from ROM, and also the voltage V of the buffer capacitor C 1 is acquired from the boosting control circuit 18 .

Subsequently, in S 520 , the energy instruction value E is divided by the voltage value V to calculate the total charge amount Qt (Qt+E/V). In S 530 , the total charge amount Qt calculated in S 120 is divided by the switching frequency N to calculate the target charge amount (Q=Qt/N). Subsequently, in S 540 , n is set to 1. Then, in S 550 , the integration reset signal Sr is output for a predetermined time. In the subsequent S 560 , the target charge amount setting signal So representing the target charge amount Q is output.

Subsequently, in S 570 , it is checked whether the signal level of the determination signal Sh is high or not. If it is determined in S 570 that the signal level of the determination signal Sh is not high, the processing of S 570 is repetitively executed. Conversely, if it is determined in S 570 that the signal level of the determination signal Sh is high, the processing proceeds to S 580 .

In S 580 , it is checked whether the driving signal Sd is input. If it is determined in S 580 that no driving signal Sd is input, the processing of S 580 is repetitively executed. Conversely, if it is determined in S 580 that the driving signal Sd is input, the processing proceeds to S 590 .

In S 590 , the measurement of the charging period Tc is started, and the cylinder selecting switching signal is output. Then, in S 600 , the measurement of the switching period Ts(n) is started. Subsequently, in S 610 , the switching period signal St is output a predetermined time, and in the subsequent S 620 , the charging switching signal Ss is output. Subsequently, it is checked in S 630 whether the signal level of the determination signal Sh is low or not.

If it is determined in S 630 that the signal level of the determination signal Sh is not low, the processing of S 630 is repetitively executed. Conversely, if it is determined in S 630 that the signal level of the determination signal is low, the processing proceeds to S 640 .

In S 640 , the output of the charging switching signal Ss is stopped, and in the subsequent S 650 , the integration reset signal Sr is output for a predetermined time. Then, it is checked in S 660 whether the signal level of the determination signal Sh is high or not.

If it is determined in S 660 that the signal level of the determination signal Sh is not high, the processing of S 660 is repetitively executed. Conversely, if it is determined in S 660 that the signal level of the determination signal Sh is high, the processing proceeds to S 670 .

In S 670 , it is checked whether the switching period Ts(n) has elapsed from the start time of the measurement of the switching period Ts(n) in S 600 . If it is determined in S 670 that the switching period Ts(n) has not yet elapsed, the processing of S 670 is repetitively executed. Conversely, if it is determined that the switching period Ts(n) has elapsed, the processing proceeds to S 680 .

In S 680 , n is incremented by adding 1 to n. Then, it is determined in S 690 whether the charging period Tc has elapsed (finished) from the time when the measurement of the charging period Tc is started in S 590 .

If it is determined in S 590 that the charging period Tc has not finished, the processing proceeds to S 600 . Conversely, if it is determined in S 590 that the charging period Tc has finished, the output of the cylinder selecting switching signal is stopped, and the processing proceeds to S 510 .

As described above, according to the driving device of the fourth embodiment, irrespective of the temperature variation, the charge energy in one turn-on operation of the charging switch SW 1 can be also fixed while the switching period Ts at the ON-time of the charging switch SW 1 in the charging period Tc is fixed. Therefore, the same advantage as the first embodiment can be acquired. Accordingly, as compared with the first embodiment, the expansion rate of the piezoelectric actuators P 1 to P 4 can be more efficiently suppressed from being varied due to variation of the electrostatic capacitance.

In the first embodiment, the charging switch SW 1 is turned on during charging operation before flow of the flywheel current is stopped. Therefore, all the energy accumulated in the inductor L 1 is not necessarily discharged, and thus a residual amount of energy is carried over to the next operation. Therefore, when the electrostatic capacitance of the piezoelectric actuators P 1 to P 4 varies, the energy to be carried over is also varied. Accordingly, the charge amount every OFF-period of the charging switch SW 1 is varied, and the expansion rate every electrostatic capacitance is slightly dispersed.

On the other hand, in the driving device according to the fourth embodiment, as shown in FIG. 10 , the switching period Ts(n) is set so that no flywheel current has been already prevented from flowing to the discharge circuit 40 at the output time of the switching period signal St even when the electrostatic capacitance varies.

Therefore, when charging switch SW 1 is next turned on, all the energy accumulated in the inductor L 1 is discharged, so that the charge amount can be suppressed from being varied every OFF-period of the charging switch SW 1 due to variation of the electrostatic capacitance of the piezoelectric actuators P 1 to P 4 .

Accordingly, according to the fourth embodiment, the variation of the expansion rate of the piezoelectric actuators P 1 to P 4 due to the variation of the electrostatic capacitance can be more efficiently suppressed as compared with the first embodiment.

Furthermore, according to the fourth embodiment, since the switching period Ts(n) is set so that flywheel current has been already prevented from flowing to the discharge circuit 40 at the time when the switching period signal St is output, current (through current) is prevented from flowing in the opposite direction flows to the parasitic diode D 2 of the discharge switch SW 2 at the time when the charging switch SW 1 is turned on. Accordingly, occurrence of noise and loss can be suppressed.

Fifth Embodiment

As compared with the driving device of the fourth embodiment, according to the driving device according to the fifth embodiment, when the electrostatic capacitance of the piezoelectric actuators P 1 to P 4 is small, a period at which the flow of flywheel current to the discharging circuit 40 is stopped in the charging operation is measured by an experiment or the like. The switching period Ts(n) is set on the basis of the measurement result.

That is, in the fourth embodiment, the switching period Ts(n) is set on the basis of the case where the electrostatic capacitance of the piezoelectric actuators P 1 to P 4 is large. However, in the fifth embodiment, the switching period Ts(n) is set on the basis of the case where the electrostatic capacitance of the piezoelectric actuators P 1 to P 4 is small.

Therefore, according to the driving device of the fifth embodiment, irrespective of the temperature variation, the charge energy in one turn-on operation of the charging switch SW 1 can be also fixed while the switching period Ts at the ON-time of the charging switch SW 1 in the charging period Tc is fixed. Therefore, the same advantage as the first embodiment can be acquired.

Furthermore, as shown in FIG. 11 , at least when the electrostatic capacitance of the piezoelectric actuators P 1 to P 4 is small, flywheel current is prevented from flowing when the charging switch SW 1 is turned on. Therefore, as compared with the first embodiment, the expansion rate of the piezoelectric actuators P 1 to P 4 can be more efficiently suppressed from being varied due to the variation of the electrostatic capacitance.

Still furthermore, according to the fifth embodiment, the switching period Ts is set on the basis of the case where the electrostatic. capacitance of the piezoelectric actuators P 1 to P 4 is small, and thus the switching period Ts can be set to a smaller value as compared with the fourth embodiment.

Accordingly, as compared with the fourth embodiment, the switching frequency N can be set to a larger value, and thus the piezoelectric actuators P 1 to P 4 can be charged so as to expand more smoothly.

Sixth Embodiment

The driving device of the sixth embodiment is different from the driving device 1 of the first embodiment in that the target charge amount Q is not fixed and the target charge amount Qn is set so that the variation amount G of the voltage per unit time applied to the piezoelectric actuators P 1 to P 4 is equal among the respective switching periods Ts(n).

Specifically, in the driving device of the sixth embodiment, the target charge amount Qn is calculated by using the following equation (2) so that the variation amount G of the voltage is equal among the respective switching periods Ts(n).

The target charge amount Qn can be calculated as above for the following reasons.

First, the total charge amount Qt to be discharged from the buffer capacitor C 1 within the charging period Tc corresponds to the value acquired by dividing the energy instruction value E by the voltage value V as defined in the following equation (3).

Q t =E/V   (3)

The variation amount G of the voltage corresponds to the value acquired by dividing the voltage value Va by the charging period Tc as defined in the following equation (4). Here, Va represents the voltage value applied to the piezoelectric actuators P 1 to P 4 when the charging is finished.

G =V a /T c   (4)

Here, assuming that V 1 represents the voltage value applied to the piezoelectric actuators P 1 to P 4 when the charging switch SW 1 is first turned off in the charging period Tc, the variation amount G 1 of the voltage when the charging switch SW 1 is first turned off corresponds to the value acquired by dividing the voltage value V 1 by the switching period Ts( 1 ) (that is, the period Ts( 1 ) when the charging switch SW 1 is first turned on) as indicated by the following equation (5).

G 1 =V 1 /T s(1)   (5)

On the other hand, the charge energy to be charged in the piezoelectric actuators P 1 to P 4 during the charging period Tc (that is, the energy instruction value E) can be represented as E={Cx(Va) 2 }/2, when the electrostatic capacitance of the piezoelectric actuators P 1 to P 4 is represented by C, and thus the voltage value Va can be represented by the following equation (6) by modifying the above equation.

Furthermore, the charge energy to be charged in the piezoelectric actuators in the charging period Tc (that is, the energy instruction value) can be also represented as E=QtV by modifying the above equation (3), and thus the following equation (7) can be derived by substituting the right-hand side of this equation into E of the above equation (6).

Accordingly, by substituting the right-hand side of the above equation (7) into Va of the above equation (4), the following equation (8) can be derived.

Likewise, the charge energy E 1 to be charged in the piezoelectric actuators P 1 to. P 4 until the charging switch SW 1 is first turned off can be represented as E 1 ={CX(V 1 ) 2 }/2, and thus the voltage value V 1 can be represented by the following equation (9) by modifying the above equation.

Accordingly, the variation amount G 1 of the voltage can be represented by the following equation (10) by substituting the right-hand side of the equation (9) into V 1 of the equation (5).

In the sixth embodiment, the variation amount G of the voltage and the variation amount G 1 of the voltage may be set to be equal to each other (G=G 1 ) so that the variation amount G of the voltage is equal among the respective switching periods Ts(n).

Therefore, if the target charge amount Q 1 is derived by substituting the right-hand side of the equation (8) and the right-hand side of the equation (10) into G=G 1 , the target charge amount Q 1 can be represented by the following equation (11).

Accordingly, the target charge amount Q 2 when the charging switch SW 1 is secondly turned on can be likewise derived.

Specifically, the variation amount G 2 of the voltage can be represented by G 2 =V 2 /{Ts( 1 )+Ts( 2 )}.

Next, the charge energy E 2 to be charged in the piezoelectric actuators P 1 to P 4 until the charging switch SW 1 is secondly turned off can be represented as E 2 ={C(V 2 ) 2 }/2=(Q 1 +Q 2 )V, and thus the voltage value V 2 applied to the piezoelectric actuators P 1 to P 4 when the charging switch SW 1 is secondly turned off can be represented by the following equation (12).

Therefore, the variation amount G 2 of the voltage can be represented by the following equation (13) by substituting the right-hand side of the equation (12) into V 2 of G 2 =V 2 /{Ts( 1 )+Ts( 2 )}.

The variation amount G 2 of the voltage may be set to be equal to the variation amount G of the voltage, and thus the target charge amount Q 2 can be represented by the following equation (14) on the basis of the equations (8) and (13).

Likewise, the target charge amount Q 3 when the charging switch SW 1 is thirdly turned on can be represented by the following equation (15).

Accordingly, the target charge amount Qn when the charging switch SW 1 is turned on at the n-th time can be represented by the equation (2) from the equations (11), (14) and (15).

Next, the processing executed by the charging switch controller 90 of the driving device of the sixth embodiment will be described hereunder with reference to FIG. 12 .

When the charging switch controller 90 of the driving device according to the sixth embodiment starts the processing of FIG. 12 , in S 710 the energy instruction value E, the charging period Tc, the switching period Ts(n) and the switching frequency N are read out from ROM, and also the voltage value V of the buffer capacitor C 1 is acquired from the boosting control circuit 18 .

Subsequently, in S 720 , the energy instruction value E is divided by the voltage value V to calculate the total charge amount Qt (Qt=E/V). In S 730 , 1 is substituted into n, and in S 740 , the target charge amount Qn (that is, the first target charge amount Q 1 ) is calculated on the basis of the equation (2). Subsequently, the integration reset signal Sr is output for a predetermined time in S 750 , and the target charge amount setting signal So representing the target charge amount Q 1 is output in the subsequent S 760 . Then, it is checked in S 770 whether the signal level of the determination signal Sh is high or not.

If it is determined in S 770 that the signal level of the determination signal Sh is not high, the processing of S 770 is repetitively executed. Conversely, if it is determined in S 770 that the signal level of the determination signal Sh is high, the processing proceeds to S 780 .

In S 780 , it is checked whether the driving signal Sd is input. If it is determined in S 780 that no driving signal Sd is input, the processing of S 780 is repetitively executed. Conversely, if it is determined in S 780 that the driving signal Sd is input, the processing proceeds to S 790 .

In S 790 , the measurement of the charging period Tc is started, and also the cylinder selecting switching signal is output. In the subsequent S 800 , the measurement of the switching period Ts(n) is started, and the switching period signal St is output for a predetermined time in the subsequent S 810 . Then, in S 820 , the charging switching signal Ss is output. It is checked in the subsequent S 830 whether the signal level of the determination signal Sh is low or not.

If it is determined in S 830 that the signal level of the determination signal Sh is not low, the processing of S 830 is repetitively executed. Conversely, if it is determined in S 830 that the signal level of the determination signal Sh is low, the processing proceeds to S 840 .

In S 840 , the output of the charging switching signal Ss is stopped, and in the subsequent S 850 , the integration reset signal Sr is output for a predetermined time. Then, it is checked in S 860 whether the value of n is not less than the switching frequency N. If it is determined in S 860 that the value of n is not less than the switching frequency N, the processing proceeds to S 710 . Conversely, if it is determined in S 860 that the value of n is less than the switching frequency N, the processing proceeds to S 870 .

In S 870 , n is incremented by added 1 to the last value. In the subsequent S 880 , the target charge amount Qn is calculated on the basis of the equation (2). In S 890 , the target charge amount setting signal So representing the target charge amount Qn calculated in S 880 is output. It is checked in the subsequent S 900 whether the signal level of the determination signal Sh is high.

If it is determined in S 900 that the signal level of the determination signal Sh is not high, the processing of S 900 is repetitively executed. Conversely, if it is determined in S 900 that the signal level of the determination signal Sh is high, the processing proceeds to S 910 .

In S 910 , it is checked whether the switching period Ts(n−1) has elapsed from the start of the measurement in S 800 . If it is determined in S 910 that the switching period Ts(n−1) has not elapsed, the processing of S 910 is repetitively executed. Conversely, it is checked in S 910 that the switching period Ts(n−1) has elapsed, the processing proceeds to S 920 .

In S 920 , it is checked whether the charging period Tc has elapsed (finished) from the start of the measurement in S 790 . If it is determined in S 920 that the charging period Tc has not finished, the processing proceeds to S 800 . Conversely, if it is determined in S 920 that the charging period Tc has finished, the output of the cylinder selecting switching signal is stopped, and the processing proceeds to S 710 .

Next, the operation of each part when the driving device of the sixth embodiment charges the piezoelectric actuators P 1 to P 4 will be described with reference to FIG. 13 .

In the driving device of the sixth embodiment, the target charge amount Q 1 is input to the non-inverting input terminal(+) of the determining circuit 80 (S 760 ). When the driving signal Sd is input to the switch controller 50 at time til as shown in FIG. 13 (S 780 : YES), both the switching period signal St and the charging switching signal Ss are output (S 810 , S 820 ), and the cylinder selecting switching signal is output to the cylinder selecting switch SWa.

At this time, the buffer capacitor current Ic 1 flows from the buffer capacitor C 1 through the charging circuit 30 to the piezoelectric actuator P 1 as described above.

When the charge amount represented by the integration signal Si reaches the target charge amount Q 1 at time t 12 , the output level of the determination signal Sh is switched to low level (S 830 : YES), and the output of the charging switching signal Ss is stopped (S 840 ).

At this time, the flywheel current flows through the parasitic diode D 2 to the piezoelectric actuator P 1 as described above.

At the time t 12 , the charging switch controller 90 outputs the integration reset signal Sr for a predetermined time, and the target charge amount Q 2 is calculated and input to the non-inverting input terminal (+) of the determining circuit 80 (S 880 , S 890 ).

When the switching period signal St is input again at time t 13 , the same operation as the time t 11 is carried out, and when the charge amount represented by the integration signal Si reaches the target charge amount Q 2 at time t 14 , the same operation as the time t 12 is carried out.

Thereafter, the operation of turning on the charging switch SW 1 every switching period Ts(n) and also turning off the switch SW 1 every time the integration signal Si reaches the target charge amount Qn is repeated until a time t 15 .

According to the driving device of the sixth embodiment, irrespective of the temperature variation, the charge energy in one turn-on operation of the charging switch SW 1 can be fixed while the switching period Ts when the charging switch SW 1 is turned on is fixed. Thus the same advantage as the first embodiment can be acquired.

Furthermore, according to the sixth embodiment, the variation amount Gn of the voltage every ON-period of the charging switch SW 1 is set to be equal, so that the expansion rate of the piezoelectric actuators P 1 to P 4 every ON-period of the charging switch SW 1 can be set to be equal.

Seventh Embodiment

In the fourth embodiment and the fifth embodiment, the switching period Ts(n) is set so that the flywheel current has been already prevented from flowing through the discharge circuit 40 at the output time of the switching period St. However, by constructing the driving device as shown in FIG. 14 , the switching period Ts(n) as described above can be set.

As shown in FIG. 14 , the driving device 1 is designed so that the current detecting resistor R 10 for detecting the charge current flowing through the piezoelectric actuators P 1 to P 4 under charging operation, a current detecting circuit 100 for detecting the charge current and a determining circuit 110 for determining whether the flow of the charge current detected by the current detecting circuit 100 is stopped are additionally provided to the driving device 1 of the first embodiment ( FIG. 2 ). The current detecting resistor R 10 is connected between an end portion at the cylinder selecting switch (SWa to SWd) side of the parallel circuit 10 and the ground line.

The current detecting circuit 100 has resistors R 11 , R 12 , R 13 , R 14 and an operational amplifier OP 3 . The connection point between the current detecting resistor R 10 and the parallel circuit 10 is connected through the resistor R 11 to the non-inverting input terminal (+) of the operational amplifier OP 3 . The connection point between the current detecting resistor R 10 and the ground line is connected through the R 12 to the inverting input terminal (−) of the operational amplifier OP 3 .

Therefore, a current detecting signal Sc 2 represented by the voltage value acquired by amplifying the potential between both terminals of the current detecting resistor R 10 at a predetermined gain is output from the output terminal of the operational amplifier OP 3 .

The reference voltage Vref is input through the resistor R 14 to the non-inverting input terminal (+) of the operational amplifier OP 3 . The determination circuit 110 has resistors R 15 , R 16 , R 17 and a comparator CMP 2 . The current detection signal Sc 2 is input through the resistor R 15 to the inverting input terminal (−) of the comparator CMP 2 , and a current zero threshold value signal Sz output from the charging switch controller 90 is input through the resistor R 16 to the non-inverting input terminal (+) of the comparator CMP 2 . The current zero threshold value signal Sz is used as a threshold value with which the determining circuit 110 checks whether no charge current (more specifically, the flywheel current) has just flowed to the piezoelectric actuators P 1 to P 4 , Thus, it is input to the non-inverting input terminal (+).

Furthermore, the determining circuit 110 is constructed as a comparator CMP 2 having hysteresis (Schmidt trigger) for switching the magnitude of the threshold value represented by the current zero threshold value signal in two stages in accordance with the output signal of the comparator CMP 2 .

As described above, in the driving device of this embodiment, when the signal level of the output signal from the determining circuit 110 is switched from low level to high level during charging operation, the flow of the charge current (more specifically, flywheel current) to the piezoelectric actuators P 1 to P 4 has been just stopped.

Accordingly, the time when the flow of the flywheel current has been just stopped can be known, and thus the switching period Ts(n) can be set so that the flow of the flywheel current to the discharge circuit 40 has been already stopped at the output time of the switching period signal St.

The setting of the switching period Ts(n) may be carried out at the shipping time of a vehicle in which the driving device is mounted, or carried out periodically.

The above embodiments may be modified in various ways. For example, although the boosting control circuit 18 turns on/off the boosting switch SW 3 in the above embodiments, the switch controller 90 may turn on/off the boosting switch SW 3 .