EP1866977A1

EP1866977A1 - Method for determining the power line frequency of a piezoelectric actuator

Info

Publication number: EP1866977A1
Application number: EP06727397A
Authority: EP
Inventors: Lionel Petit
Original assignee: Somfy SA
Current assignee: Somfy SA
Priority date: 2005-04-07
Filing date: 2006-03-31
Publication date: 2007-12-19
Also published as: FR2884369B1; WO2006106397A1; FR2884369A1

Abstract

The invention concerns a method for determining the normal power line frequency of a piezoelectric actuator comprising generators characterized in that it comprises a phase of determining the frequencies of the generators for which the coupling coefficient of the channels when the power line frequency causes the actuator to resonate and a step of selecting among those frequencies, the normal power line frequency of the actuator by both generators. The invention also concerns a piezoelectric actuator whereof the common frequency of the generators is such that the coupling coefficient of the channels when the actuator is powered at said frequency is less than 40 % of the coupling coefficient of the channels when the power line frequency causes the actuator to resonate.

Description

Method for determining the supply frequency of a piezoelectric actuator

The invention relates to a method for determining the normal supply frequency of a piezoelectric waveguide actuator comprising a stator, a rotor biased in contact with the stator and prestressed piezoelectric transducers, with a multilayer structure, for to deform the stator and distributed along two power channels each connected to a generator. The invention also relates to a piezoelectric actuator with a traveling wave.

Piezoelectric actuators are known. They comprise a stator, whose vibratory state is ensured by piezoelectric transducers and a rotor pressed against the stator and driven by the vibrations of the latter.

Preferably, the vibratory movement of a point situated on the surface of the stator corresponds to an elliptical trajectory whose axis is perpendicular to the plane of the stator-rotor contact surface when the motor is not powered and whose other axis is directed according to the orthoradial direction of rotor drive by the action of the stator. Such a movement is caused by a stationary wave applied to the stator or preferably caused by a progressive wave.

The stator generally has an annular shape of revolution.

To create a progressive wave, at least two pairs of piezoelectric elements are used to combine two standing waves. If the two pairs of elements are shifted in the space of 90 ° and the feeds, of the same amplitude and the same frequency, are themselves shifted in the time of a quarter of period (phase shift of 90 °), then the combination of the two vibratory movements gives a pure progressive wave: all the points of the stator describe the same alternative trajectory around their rest position with a temporal offset which depends only on their angular offset on the stator ring.

According to the geometrical structure of the stator and according to the nature and positioning of the piezoelectric elements with respect to this structure, the vibration amplitude of the structure can considerably exceed the amplitude of vibration of the piezoelectric transducers alone. : this is the resonance phenomenon, which is generally used to amplify the amplitude of vibration, therefore the speed of training.

US Pat. No. 4,562,374 describes such a structure in the case where the piezoelectric transducers are not separated but consist of the same annular ceramic of very small thickness. Piezoelectric transducers are not prestressed. The annular ceramic is bonded to the stator, and is excited in a transverse mode, also called d31 mode. It is this type of structure that has given rise to most of the studies on piezoelectric motors.

However, patents EP 0 930 660 and US 6,388,365 also describe a traveling wave structure, but this time with separate and prestressed piezoelectric transducers. Each piezoelectric transducer is in the form of a column.

The polarization of the piezoelectric transducer is performed along the axis of the column, and its main displacement also, which characterizes the longitudinal excitation mode, also called d33 mode. The invention applies to such progressive wave structures provided with separate piezoelectric transducers preloaded and excited in a longitudinal mode.

The coupling phenomenon is explained in FIG. 4. In advance, FIGS. 1 to 3 describe a piezoelectric motor according to the prior art.

A motor 1 known from the prior art is shown schematically in section in Figure 1. It comprises a shaft 11 secured to a rotor 12 biased in a position of contact with a stator 13 by a static force FO. For more legibility, the rotor is shown here slightly detached from the stator. Friction material is usually glued to the stator to provide better contact between the two parts. The shaft 11 is guided by a ball bearing 14. The axis of this shaft 11 constitutes a cylindrical axis of symmetry for the motor, with the exception of piezoelectric transducers 15, 16 which are separated. These transducers have for example the shape of cylindrical columns or the shape of rectangular columns and are distributed on the stator, as represented for example in patent EP 0 930 660. These piezoelectric transducers are electrically grouped, generally according to a parallel connection, to form an excitation path (in a standing wave motor) or two excitation pathways (in a traveling wave motor). In a traveling wave motor of this type, two diametrically opposed exciters generally belong to the same excitation path and are therefore connected in parallel with alternating polarizations. A motor comprising four exciters B1-B4 would therefore have the two exciters B1 and B3, belonging to the first excitation channel, as represented by the transducers 15 and 16 of FIG. 1, while the axes of the other two exciters B2 and B4. , belonging to the second way, would be contained in a plane perpendicular to the plane of Figure 1. However, for simplicity, consider thereafter that the transducer 15 belongs to a first path, while the transducer 16 belongs to the second path of excitation.

The stator and the rotor are generally metallic. For simplicity, it is assumed that the stator 13 is non-conducting on the surface. This is the case when it comes to an anodized aluminum alloy.

Each piezoelectric transducer is powered by two electrodes, represented by a thick black line and referenced 151 and 152 for the piezoelectric transducer 15. The piezoelectric material is represented by the hatched surface between the electrodes 151 and 152.

A problem posed by these structures concerns the difficulty of controlling the operation of the actuator.

It is known from US 5,554,905 a piezoelectric moving wave actuator comprising a stator, a rotor biased in contact against the stator and prestressed piezoelectric transducers, multilayer structure and for deforming the stator. According to this patent, the technical effect obtained by the multilayer structure of the transducers is to allow greater deformation thereof. Due to the multilayer structure of the transducers, the currents of the supply currents are high, which poses problems, in particular switching problems. In addition, the difficulty of controlling the operation of the actuator is increased.

US Pat. No. 5,767,609 discloses a piezoelectric motor comprising monolithic piezoelectric transducers fed at an anti-resonance frequency. This choice of an anti-frequency resonance is justified only by considerations of efficiency and engine operating stability.

The object of the invention is to provide a piezoelectric actuator and a method for determining the supply frequency of a piezoelectric actuator improving actuators and methods known from the prior art and making it possible to overcome the disadvantages supra. In particular, the invention provides a method for determining a supply frequency for improving control of the actuator. It also relates to a piezoelectric actuator with a gradual wave powered at a frequency to improve its control.

The determination method according to the invention is characterized by the characterizing part of claim 1.

Different embodiments of the method are defined by the dependent claims 2 to 6.

The actuator according to the invention is defined by claim 7.

Different embodiments of the actuator are defined by the dependent claims 8 to 11.

The appended drawing represents, by way of example, a piezoelectric actuator according to the invention.

FIG. 1 schematically represents a piezoelectric actuator of known structure of the prior art. FIG. 2 represents a characteristic curve of the admittance of a piezoelectric transducer, as a function of the supply frequency, when it is fitted to an actuator.

Figure 3 shows an equivalent circuit diagram of a path of a piezoelectric actuator.

Figure 4 shows an equivalent circuit diagram of two paths of a piezoelectric actuator.

FIG. 5 schematically represents a piezoelectric actuator according to the invention.

FIG. 6 diagrammatically represents the electrical part of a piezoelectric actuator according to the invention.

FIG. 7 represents a flowchart of a method for determining a supply frequency of a piezoelectric actuator making it possible to minimize the coupling coefficient between the channels.

FIG. 8 represents a first variant of the method for determining a supply frequency.

FIG. 9 represents a second variant of the method for determining a supply frequency.

FIG. 10 represents a third variant of the method for determining a supply frequency.

The admittance Y of the first excitation channel, comprising the transducer 15, is represented in FIG. 2 over a wide range of frequencies. For a constant value of the amplitude of the excitation voltage U1, the admittance is the image of the intensity of the absorbed current IA. This figure shows an overall relationship of proportionality between admittance and frequency, which is characteristic of capacitive behavior. This is due to the capacitance CO of the capacitor formed by the electrodes 151, 152 opposite, separated by the highly insulating piezoelectric material.

There is also the appearance of several resonance modes, sufficiently distant so that each mode can be the subject of a separate analysis and a representation in the form of an equivalent diagram as in Figure 3.

In the equivalent diagram of FIG. 3, the mode is characterized by an "emotional branch" comprising at least three equivalent components: a capacitor Cn, an inductance Ln, a resistor Rn and, optionally, a generator En. The capacity of the capacitor Cn and the inductance Ln correspond to the stiffness and to the mass of the parts undergoing the deformations. The resistance Rn corresponds to the power dissipated in these mechanical deformations, while the counter-electromotive force of the generator En gives the power converted into useful mechanical power, when it is multiplied by the current Imn circulating in the emotional branch. For a vacuum engine, we have En =

0.

In general, the motional capacity Cn of mode n is very weak in front of the capacity CO.

The electromechanical behavior of the emotional branch depends on the excitation frequency. The current IMn absorbed is indeed maximum when the impedances of the capacitor Cn and the inductance Ln present the same module. This resonance frequency of the n-mode, called "series resonance", is denoted SRFn.

When the frequency FREQ is substantially lower than this value, then the motional branch becomes equivalent to the capacitance of the capacitor Cn only. Since the capacity of the capacitor Cn is small compared to the capacitance of the capacitor CO, the current Imn circulating in the motional branch is negligible. in front of ICO flowing through the capacitor CO.

When the frequency FREQ is greater than this value SRFn, then the motional branch becomes substantially equivalent to the inductance Ln. The assembly formed by the capacitor CO and the inductance Ln then constitutes a parallel resonant circuit, which is known that the absorbed current IAn passes through a minimum. The term PRFn denotes this particular value of the frequency, called "parallel resonance" or anti-resonance frequency.

In fact, the equivalent diagram of FIG. 3 represents what is seen by the generator of one excitation channel when the generator supplying the other channel is disconnected. Figure 4 now shows simultaneously the two generators G1 and G2 feeding each of the channels. Three impedance components Z1, Z2 and ZC represent the greater or lesser facility of the structure to vibrate under the effect of the vibrations of one or the other piezoelectric transducer. But the vibrations on one of the piezoelectric transducers necessarily have an induced effect on the neighboring piezoelectric transducer, which belongs to the other path. Only a perfect articulation at the level of the connection between piezoelectric transducer and stator would avoid such coupling. The object of patent EP 0 930 660 is to reduce at most this coupling, but it remains and is represented here by the impedance ZC. Impedances Z1 and Z2 include elements representative of stiffness and mass in the n mode. In a symmetrical structure, the value of the impedance Z1 is equal to the value of the impedance Z2. The ideal would be to have the impedance ZC worth 0, to be reduced to two independent schemes, one for each channel. When the structure is poorly coupled, the value of the impedance ZC is much lower than the value of the impedances Z1 or Z2.

The generators G1 and G2 are not perfect, and have at least an internal resistance RG1 and RG2. Thus, their output voltage is affected by the current output. If, for example, at the moment when the voltage U1 is maximum and the voltage U2 should be zero, the generator G2 is replaced by its equivalent resistance RG2, a voltage U2 is measured from the voltage across ZC and divided by the potentiometric arrangement formed by the impedances Z2 and RG2 in parallel on C02. It is therefore clear that the desired phase relationship (phase shift of 90 °) is not verified.

In a piezoelectric motor with a fine annular ceramic, this effect is very weak, in particular because of the small thickness of the ceramic, which leads to important values for the blocked capacitor CO, and therefore to a low impedance of the capacitor CO at the frequency considered.

On the other hand, in an engine comprising piezoelectric transducers of stick or cylinder type, the two electrodes are separated by several millimeters and the capacitance CO is very low, hence a high impedance of the capacitor CO at the frequency considered. For these same reasons, it is not possible, except to use inductances of high value, and therefore expensive, to compensate by an inductance additional in series with each generator the presence of the capacity CO.

Such an inductance, which would then constitute the impedance equivalent to the generator G2, may have the advantage of constituting with CO 2 a tuned capping circuit, of infinite impedance at the working frequency, returning to cancel the current in an emotional branch Z2 which is no longer powered.

The addition of capabilities in the form of discrete elements would of course penalize the structure: the gain made in the fight against the coupling would be canceled by the need to provide greater currents or voltage.

The piezoelectric actuator shown in Figure 5 has substantially the same structure as that shown in Figure 1. All similar components have a reference beginning with the number 2 instead of the number 1 in Figure 1. However, it differs of the actuator of FIG. 1 in that it comprises multilayer piezoelectric transducers 25, 26.

This piezoelectric actuator has a structure enabling it to generate a progressive wave. It thus comprises, for example, two channels each comprising two piezoelectric transducers, these two channels being powered by two generators. The transducers used are of multilayer type. Two piezoelectric transducers of this type, belonging to each channel, are represented under the references 25 and 26.

The piezoelectric material is interposed between conductive planes alternately connected to a first electrode 251 and to a second electrode 252. These conductive planes delimit layers 263 of piezoelectric material whose direction of polarization is alternated from one layer to another. Each electrode is connected to a terminal of a voltage generator.

For the sake of clarity, only a few layers of piezoelectric material have been shown, however, the piezoelectric transducers may have several tens or even hundreds of layers. The transducers used for the development of the invention comprise, for example, 20 layers per millimeter.

However, the use of piezoelectric transducers with a multilayer structure considerably increases the series resistance of the transducers, which amounts to increasing by the same the internal resistance RG 1 or RG2 generators G1 and G2. Taking this effect into account leads to departing even further from the series resonance frequency. To this increase of the internal resistance is added the problem of the increase of the absorbed current, resulting from the greater value of the blocking capacitance C01 and CO2 for a multilayer exciter in comparison with a massive exciter.

Ultimately, it is therefore necessary, in order to solve the problem of control difficulty of the actuator found in the documents of the prior art, to choose a supply frequency which makes it possible to safely minimize the coupling coefficient between the channels. .

To do this, this frequency can be determined by the method described with reference to the figure in FIG. 7.

In a first step 100, one of the two generators G2 is replaced by a resistor RG2 whose value is equal to the internal resistance of the generator G2. In the case where the generator G2 has a ZG2 complex output impedance, the generator G2 is replaced by an element present this impedance. Alternatively, one can leave the generator G2 in place and simply ensure that it delivers a zero voltage amplitude.

In a step 110, a supply frequency Fa of the generator G1 is set. This frequency is for example a low end of the frequency range corresponding to the vibration mode n chosen to allow a good drive of the rotor by the deformations of the stator.

In a step 120, a test is performed on the value of Fa and on the incrementation of F as long as this value belongs to the interval considered. This step allows you to scan the entire interval.

In a step 130, the actuator is supplied with the generator G1 at the frequency fa and the voltage of the generator is adjusted so as to obtain a given deformation of the stator, preferably a deformation close to that obtained during the operation of FIG. the actuator.

In a step 140, different electrical values are measured, in particular the voltage across the generator G1 and the voltage across the resistor RG2. Optionally, the current supplied by the voltage generator G1 and / or the current absorbed by the resistor RG2 can also be measured.

In a step 150, the coupling coefficient between the channels is calculated, for example by calculating the ratio of the amplitude of the voltage measured across the resistor RG2 and the amplitude of the voltage. measured at the terminals of the first generator G1 when it feeds the first channel.

At the end of step 150, the method loops on step 110 at which the frequency of the generator G1 is incremented and the steps 120 to 150 are repeated.

When a sufficient frequency range (for example equal to twice the difference between PRFn and SRFn) has been scanned, proceed to step 160 in a second phase of the method in which the frequencies are determined for which the coefficient of coupling is below a certain threshold, for example less than 40%, or preferably 20% of the coupling coefficient when the actuator is powered at a frequency causing its resonance.

In the final step 170, an operating frequency FN of the motor is chosen within the frequency zone defined in step 160. This step 170 may correspond to an arbitrary choice, or preferably be made according to the variants represented. by Figures 8 to 10.

In the simplest case, represented in FIG. 8 in the form of a single substep 170-1, step 170 consists of taking as value FN the value corresponding to a minimum value of the coupling coefficient.

In the variant of FIG. 9, also comprising a single sub-step 170-2, step 170 consists of taking as value FN the value corresponding to a minimum value of the current 11 absorbed on the first excitation channel. This value of FN generally corresponds to the anti-resonance frequency PRFn, if it is in the zone adopted during step 160. In the variant of FIG. 10, two sub-steps are used to carry out step 170.

In sub-step 170-3, a quantity C comprising the coupling coefficient is defined, but also enriched by taking into account the intensity 11 absorbed by the energized excitation path. It is indeed desired to reduce both the coupling coefficient and the intensity 11. Increasing expression of C with these two parameters is appropriate. The simplest size C is therefore written:

C = K1 x U2 x l1

In this expression, K1 is a coefficient of proportionality, U2 is the voltage measured across the resistor RG2 and 11 is the current absorbed on the channel 1, thus provided by the generator G1 while the generator G2 has been replaced by its equivalent resistance RG2.

Since the resistance RG2 is fixed, it is clear that the coupling coefficient can also be expressed from the current 12 in this resistor, and that the quantity C can also be written using a second coefficient of proportionality. K2:

C = K2 x l2 x l1

Finally, while remaining in the spirit of the invention, it is possible to choose a more elaborate criterion, depending on whether one prefers more or less the influence of the coupling or the absorbed current. It is advantageous to integrate the vibration amplitude and / or the vibratory speed in the size C. In sub-step 170-4, the frequency value corresponding to a minimum of this value is taken as the operating frequency value FN of the motor.

The electromechanical simulation means of the structures make it possible to determine the coupling coefficient and the frequency zone in which this coefficient is lower than a given threshold without having to carry out real tests, but the approach is equivalent to that which has just been described.

Finally, it is clear that the normal operating frequency FN of the complete actuator when it is powered by the two generators which is thus determined gives a nominal value. It is known that the actual frequency of the supply frequency of piezoelectric motors can and must be self-adjusted around this nominal value, in particular to take account of variations in temperature or load.

FIG. 6 shows the electrical part of a piezoelectric actuator, with two feed channels and two transducers per channel, in the form of an equivalent diagram. It is preferentially on such a complete actuator that the previous measurements of coupling coefficient and absorbed current are carried out. In this electrical part, the components schematizing the coupling of the channels have disappeared (or have become negligible) because of the use of piezoelectric transducers with multilayer structure B1 and B3 on the first channel, B2 and B4 on the second channel and a frequency that minimizes the coupling between the channels. The electrical characteristics of the channel supply voltages U1 and U2 are such that their excitation frequencies are identical (with a desired phase shift, for example 90 °) and are chosen according to the method described above. The diagram equivalent to such an assembly is that of FIG. 4, in which the impedance ZC is replaced by a short-circuit: in fact, the invention makes it possible to make the impedance ZC negligible compared with the emotional impedances Z1 or Z2.

The generators G1 and G2 may include additional series inductors so as to compensate for the blocked capacitances of the piezoelectric transducers, which have become important. These series inductances are represented in dotted lines and are referenced by Ls 1 and Ls2.

The advantage of the inductance Ls2 is that it can constitute, with the blocked capacitor C02, a parallel resonant circuit, or a plug circuit, on a channel whose voltage is normally zero at a given instant. At the tuning frequency, this amounts to considerably increasing the motional impedance Z2 of this channel in comparison with the coupling impedance ZC. Which amounts to making the coupling term even more negligible.

On the other hand, the inductance Ls1, as already applied in the prior art, constitutes a series resonant circuit with the blocked capacitor C01 for the first generator G1. In summary, the first generator "does not see" the second channel, thanks to with the coupling C02 and Ls2, while the effects of a strong blocked capacity C01 are compensated by Ls1. Conversely, the second generator does not see the first feed path, thanks to the coupling C01 and Ls1, while the effects of a strong blocked capacity C02 are compensated by Ls2. The additional inductance values are chosen such that the tuning frequency is equal to the motor supply frequency.

Claims

Claims:

A method for determining the normal supply frequency (FN) of a piezoelectric waveguide actuator (2) comprising a stator (23), a rotor (22) biased in contact with the stator (23), and prestressed piezoelectric transducers (25, 26) having a multilayer structure, intended to deform the stator (23) and distributed along two supply paths each connected to a generator (G1, G2), characterized in that it comprises : at. a phase for determining the frequencies of the generators (G1, G2) for which the coupling coefficient of the channels is less than a given threshold at most equal to 40% of the channel coupling coefficient when the supply frequency causes the resonance of the actuator and b. a step of selecting among these frequencies, the frequency (FN) of normal supply of the actuator by the two generators.

2. Method according to claim 1, characterized in that the coupling coefficient is determined by the measurement of electrical quantities (U 1, U 2, M, 12) when a channel is powered by the first generator (G1) while the second generator (G2) is replaced by its internal resistance (RG2) and in that the coupling coefficient is calculated by the ratio of two of these magnitudes.

3. Method according to one of the preceding claims, characterized in that the step of selecting the normal supply frequency (FN) of the actuator consists in choosing among the frequencies determined in the first phase, the frequency value giving the lowest value for the coupling coefficient.

4. Method according to one of claims 1 to 2, characterized in that the step of selecting the normal supply frequency (FN) of the actuator consists in choosing, among the frequencies determined in the first phase, the frequency value giving the lowest value for the intensity (11) supplied by the first generator.

5. Method according to one of claims 1 to 2, characterized in that the step of selecting the normal supply frequency (FN) of the actuator comprises a substep of defining a magnitude (C) both increasing with the coupling coefficient and with the absorbed intensity, and a substep of choosing, among the frequencies determined in the first phase, the frequency value giving the lowest value for this quantity.

6. Method according to one of the preceding claims, characterized in that in the frequency determination phase, the coupling coefficient threshold is 20% of the coupling coefficient of the channels when the supply frequency causes the resonance of the actuator.

7. Piezoelectric actuator (2) with a progressive wave comprising a stator (23), a rotor (22) biased in contact with the stator (23) and piezoelectric transducers (25, 26) prestressed, multilayer structure and for deforming the stator (23), distributed along two supply paths each connected to a generator (G1, G2), characterized in that the common frequency of the generators is such that the coupling coefficient of the channels when the actuator is fed at this frequency is less than 40% of the coupling coefficient of the channels when the supply frequency causes the resonance of the actuator.

8. Piezoelectric actuator (2) according to claim 7, characterized in that the common frequency of the generators is determined by the method according to one of claims 1 to 5.

9. piezoelectric actuator (2) according to claim 7, characterized in that the common frequency of the generators is such that the coupling coefficient of the channels when the actuator is powered at this frequency is less than 20% of the coupling coefficient channels when the supply frequency causes resonance of the actuator.

10. Piezoelectric actuator (2) according to claim 9, characterized in that the common frequency of the generators is determined by the method according to claim 6.

11. Piezoelectric actuator (2) according to one of claims 7 to 10, characterized in that each voltage generator is connected in series to an inductive element (Ls1, Ls2) and in that each of these inductive elements constitutes , at the supply frequency, a series resonant circuit with the transducers of the powered channel.