Piezoelectric MEMS intelligent loudspeaker preparation method

Abstract

The invention particularly relates to a preparation method of a piezoelectric MEMS intelligent loudspeaker. The preparation method comprises the following steps: preliminarily calculating the vibration frequency of a film by using a theoretical formula of the vibration frequency of the film according to a target frequency range; carrying out analogue simulation by using finite element analysis software, and predicting the resonant frequency of each vibrating diaphragm to determine each layer of material; selecting a piezoelectric material; selecting a substrate material; sputtering a first electrode material on the substrate, and etching to form a required electrode pattern; depositing a piezoelectric layer material to ensure uniform thickness; sputtering a second electrode material on the piezoelectric layer, and etching to form a required electrode pattern; testing the actual performance frequency of the loudspeaker; adjusting the thickness or the shape of the piezoelectric layer according to a test result; adjusting the thickness or the shape of the piezoelectric layer by adopting a dynamic algorithm; re-depositing the piezoelectric layer and adjusting the piezoelectric layer; and the resonant frequencies of all the vibrating diaphragms are within the target range.

Description

Technical Field

[0001] The present invention belongs to the field of smart speakers, and in particular relates to a method for preparing a piezoelectric MEMS smart speaker. Background Art

[0002] With the development of portable electronic devices and Internet of Things technology, the demand for miniaturized and high-performance speakers is growing. Piezoelectric MEMS (micro-electromechanical system) speakers have attracted much attention due to their small size, low power consumption, and fast response speed. However, when designing a piezoelectric MEMS speaker with a double-ring surround circular diaphragm structure, there is a technical challenge of how to accurately control the resonant frequencies of different diaphragms (circular and annular) to achieve full audio range (20Hz-20kHz) coverage. This challenge mainly stems from the need to accurately control the thickness, shape, and material properties of the piezoelectric layer, and existing manufacturing technologies and design methods often find it difficult to achieve this high-precision requirement.

[0003] The existing piezoelectric MEMS speaker design with a double-ring surround circular diaphragm structure includes a substrate with a back chamber on the back of the substrate. The first electrode, the piezoelectric layer, and the second electrode are deposited in sequence on the front of the substrate. The vibration membrane layer is composed of these three layers and the actuation layer of the back chamber, including a central circular diaphragm, an inner ring diaphragm, and an outer ring diaphragm. These diaphragms are separated by groove gaps and connected by support beams. The full audio range of 20Hz-20kHz is covered by the different resonant frequencies of the three diaphragms (central circular diaphragm: 10kHz-20kHz; center + inner ring diaphragm: 1kHz-10kHz; all three diaphragms: 20Hz-1kHz).

[0004] With existing technologies, it is difficult to control the resonance frequency: since the resonance frequencies of three vibration membranes need to be controlled simultaneously, precise adjustment becomes a problem.

[0005] For existing technologies, material and structure sensitivity: The thickness, shape and material properties of the piezoelectric layer have a direct impact on the resonant frequency of the vibration membrane and require extremely precise control.

[0006] For existing technologies, manufacturing process limitations: Current manufacturing processes make it difficult to achieve the required high-precision control, especially in adjusting the thickness and shape of the piezoelectric layer. Summary of the invention

[0007] The purpose of the present invention is to provide a method for preparing a piezoelectric MEMS smart speaker to solve the problems raised in the above background technology.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions: The method for preparing a piezoelectric MEMS speaker with a double-annular surrounding circular vibration membrane structure comprises: Preliminarily calculate the film vibration frequency using the theoretical formula of the film vibration frequency according to the target frequency range; use finite element analysis software to simulate and predict the resonance frequency of each vibration film to determine the material of each layer, or directly determine the material of each layer based on empirical data; select piezoelectric material; select substrate material; sputter the first electrode material on the substrate, and etch to form the required electrode pattern; deposit the piezoelectric layer material to ensure uniform thickness; Sputter the second electrode material on the piezoelectric layer and etch to form the required electrode pattern; etch to form the first groove gap, the second groove gap and the third groove gap, and use a cross-shaped vertical beam, a three-cross beam or a four-cross beam as the support beam; etch to form a back cavity on the back of the substrate; test the actual performance frequency of the speaker; adjust the thickness or shape of the piezoelectric layer according to the test results: Assuming that the test results show that the resonant frequency of the circular diaphragm is too high, the diameter of the circular diaphragm structure is increased; Assuming that the test results show that the resonant frequency of the circular diaphragm is too low, the diameter of the circular diaphragm structure is reduced; Assuming that the test result shows that the resonance frequency of the first annular vibration membrane is too high, the ring width of the first annular vibration membrane structure is increased; Assuming that the test result shows that the resonance frequency of the first annular vibration membrane is too low, the ring width of the first annular vibration membrane structure is reduced; Assuming that the test results show that the resonance frequency of the second annular vibration membrane is too high, the ring width of the second annular vibration membrane structure is increased; Assuming that the test results show that the resonance frequency of the second annular vibration membrane is too low, the ring width of the second annular vibration membrane structure is reduced; and also including using a dynamic algorithm to adjust the thickness or shape of the piezoelectric layer; The piezoelectric layer is redeposited and adjusted until the resonant frequencies of all diaphragms are within the target range.

[0009] Furthermore, finite element analysis software was used to simulate and predict the resonant frequency of each diaphragm to determine the material of each layer, specifically: Based on the designed double-ring surrounding circular diaphragm structure, a geometric model was created using CAD software; Input the physical and mechanical properties of each layer material, such as elastic modulus, Poisson's ratio, density; Define the fixed endpoints of the support beam as boundary conditions, as well as the pressure conditions for the back chamber; For the diaphragm area, a finer mesh is used to improve the simulation accuracy; Use a coarser grid in non-critical areas; Applying a voltage between the first electrode and the second electrode to simulate the excitation of the speaker in a working state; Conduct frequency response analysis, gradually increase the excitation frequency to find the resonant frequency of each diaphragm; Perform modal analysis to extract the first few natural frequencies and corresponding vibration modes; Furthermore, a frequency response analysis is performed to confirm the stability of the resonant frequency and its corresponding vibration mode; Compare the simulation results with the theoretical calculation results to verify the effectiveness of the model; Based on the simulation results, adjust the thickness, shape, or material properties of the piezoelectric layer to achieve the target resonant frequency range.

[0010] Furthermore, the piezoelectric material includes PZT, ZnO, and AlN; the substrate material includes SOI wafer, PDMS, PE, and PI.

[0011] Further, sputtering the first electrode material on the substrate comprises: Depositing electrode material on the substrate using a sputtering process in a physical vapor deposition technique; Controlling the sputtering time or deposition rate to achieve the desired electrode thickness; Monitor the quality of thin films during deposition; Measure the thickness of electrode materials; Etching to form the desired electrode pattern: Evenly coating a layer of photoresist on the surface of the deposited electrode material; Spin coater is used for coating to ensure uniform distribution of photoresist; exposing the photoresist using a mask to transfer the desired electrode pattern; Using a developer to remove the photoresist in the exposed area, leaving the photoresist in the unexposed area as a mask for subsequent etching; Using dry etching or wet etching to remove the electrode material not masked by the photoresist to form a desired electrode pattern; Removing residual photoresist; Check the quality of the electrode pattern after cleaning to ensure there is no residual photoresist or other impurities; Check the integrity, edge quality, and dimensional accuracy of the electrode pattern using an optical microscope or scanning electron microscope; Confirm that the electrode pattern is exactly the same as the design file.

[0012] Further, the process of depositing the piezoelectric layer material includes: Prepare the deposition chamber: Clean the deposition chamber to ensure it is free of impurities; Load the substrate and set the position of the support beams; Select ALD or PLD deposition method and set deposition parameters such as temperature and gas flow rate; Start deposition and monitor film thickness via an online monitoring system; Ensure that the film thickness is uniform and the deviation is within the allowable range; The crystalline quality of the film was examined using XRD; Use AFM to examine the surface roughness of the film; Check the uniformity of film thickness using ellipsometer; If the film thickness is not uniform or does not meet the design requirements, adjust the deposition parameters or re-deposit.

[0013] Furthermore, the steps of sputtering a second electrode material on the piezoelectric layer and etching to form a desired electrode pattern are the same as the steps of sputtering the first electrode material on the substrate.

[0014] Further, etching is performed to form a first trench gap, a second trench gap and a third trench gap. Specifically: uniformly coating a layer of photoresist on the completed electrode and piezoelectric layer structure; Use a spin coater for coating to ensure uniform distribution of the photoresist; exposing the photoresist using a mask to transfer a desired trench gap pattern; Using a developer to remove the photoresist in the exposed area, leaving the photoresist in the unexposed area as a mask for subsequent etching; Using dry etching to remove materials not masked by the photoresist to form the desired trench gap; The first groove gap is located between the central circular diaphragm and the inner annular diaphragm; The second groove gap is located between the inner annular diaphragm and the outer annular diaphragm; The third groove gap is used to separate the outer annular vibration membrane from other structures.

[0015] Further, etching the back side of the substrate to form a back cavity comprises: Evenly coating a layer of photoresist on the back side of the substrate; Spin coater is used for coating to ensure uniform distribution of photoresist; exposing the photoresist using a mask to transfer the pattern of the back cavity; A developer is used to remove the photoresist in the exposed areas, leaving the photoresist in the unexposed areas as a mask for subsequent etching.

[0016] Further, test the actual performance frequency of the speaker: Set up the test bench and microphone positions, making sure the distance between the speaker and microphone meets the test requirements; Connect to the acoustic test system and configure the corresponding hardware and software settings; Set the parameters of the test signal, including the sweep frequency range and signal amplitude; Start the test signal and start the frequency response test; Record the output signal of the speaker at different frequencies; Process the recorded data using the acoustic test system's software to generate a frequency response curve; Analyze the frequency response curve to determine the actual resonant frequency of each diaphragm; Compare the actual resonant frequency with the design target to assess whether the expected frequency range is achieved; Evaluate whether the actual performance frequency of the speaker meets the design requirements based on the test results.

[0017] Furthermore, the modification process data of the vibration membrane structure is obtained, and the modification amount of the vibration membrane structure is adjusted by using a particle swarm optimization algorithm, including: Based on the previous tests, the actual resonance frequency data of each diaphragm was sorted out; Determine the parameters that need to be optimized, including the thickness of the piezoelectric layer and the diameter of the vibration membrane; Particle swarm initialization: Set the size of the particle swarm; Randomly initialize the position and velocity of each particle to ensure that the particles are distributed in the parameter space; The fitness function is defined as the inverse of the sum of squares of deviations from the target frequency range, that is, the closer to the target frequency range, the higher the fitness; Use the standard PSO update rule, taking into account the particle’s current position, personal optimal position, and global optimal position; When updating particle speed and position, make sure the parameter values ​​remain within a reasonable range; Iterative optimization process: In each iteration, the fitness value of each particle is evaluated; Update the particle's position and velocity; Repeat the iteration until the preset maximum number of iterations is reached or the fitness value converges; According to the optimal parameter values ​​obtained through optimization, the structure of the vibration membrane is adjusted.

[0018] Compared with the prior art, the beneficial effects of this application are: Finite element analysis software is used to perform simulations and predict the resonant frequency of each diaphragm to ensure the accuracy of the design.

[0019] Fine etching and photolithography processes have been implemented to form high-precision electrode patterns, groove gaps and support beam structures, which enhance the structural stability and acoustic performance of the speaker. Effective coverage of the full audio range of 20Hz-20kHz is achieved by precisely controlling the thickness, shape and material properties of the piezoelectric layer and dynamically adjusting the design parameters.

[0020] By precisely controlling the resonance frequencies of different vibration membranes (circular, annular), the speaker is ensured to have a smooth frequency response curve in the full audio range, thereby improving the sound quality.

[0021] The particle swarm optimization algorithm (PSO) is used to automatically adjust the parameters of the vibration membrane structure, such as the thickness of the piezoelectric layer and the diameter of the vibration membrane, which improves the accuracy and efficiency of the design.

[0022] Through the iterative optimization process, the best parameter combination can be found quickly, reducing the time and effort required for manual adjustments and accelerating the product development cycle.

[0023] Materials with suitable piezoelectric properties (such as PZT, ZnO, AlN), and substrate materials with good mechanical strength and chemical stability (such as SOI wafer, PDMS, PE or PI) are selected.

[0024] The sputtering process and deposition parameters were optimized to ensure high-quality preparation of electrodes and piezoelectric layers, improving the performance and reliability of the loudspeaker. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a flow chart of the method of this application. DETAILED DESCRIPTION

[0026] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0027] A method for preparing a piezoelectric MEMS speaker with a double-ring surrounding circular diaphragm structure, such as Figure 1 As shown, Includes steps: The film vibration frequency is preliminarily calculated using the theoretical formula of the film vibration frequency according to the target frequency range.

[0028] The theoretical formula of film vibration frequency is used to preliminarily calculate the film vibration frequency f: , Where T is the tension on the film, ρ is the surface density of the film, and A is the effective vibration area of ​​the film; Use finite element analysis software to simulate and predict the resonance frequency of each vibration membrane to determine the material of each layer, or directly determine the material of each layer based on empirical data. Specifically: 1. Prepare the model Establishing geometric model: Based on the designed double-ring surround circular vibration membrane structure, a three-dimensional model was created using CAD software.

[0029] Material property definition: Input the physical and mechanical properties of each layer material (substrate, electrode, piezoelectric layer), such as elastic modulus, Poisson's ratio, density.

[0030] Boundary condition settings: Define the fixed end points of the support beam as boundary conditions, and the pressure condition of the back chamber.

[0031] 2. Model meshing Fine Mesh: For the diaphragm area, a finer mesh is used to improve the simulation accuracy.

[0032] Coarser mesh: Use a coarser mesh in non-critical areas to reduce computing time and resource consumption.

[0033] 3. Provide incentives Apply voltage: Apply voltage between the first electrode and the second electrode to simulate the excitation of the speaker in working state.

[0034] Frequency sweep: Perform a frequency response analysis by gradually increasing the excitation frequency to find the resonant frequency of each diaphragm.

[0035] 4. Perform modal analysis Modal Extraction: Perform modal analysis to extract the first few natural frequencies and corresponding vibration modes.

[0036] Frequency response analysis: Further frequency response analysis is performed to confirm the stability of the resonant frequency and its corresponding vibration mode.

[0037] 5. Analyze the results Comparison with theoretical calculations: Compare the simulation results with the results of theoretical calculations to verify the effectiveness of the model.

[0038] Optimize design parameters: Adjust the thickness, shape, or material properties of the piezoelectric layer based on simulation results to achieve a target resonant frequency range.

[0039] 6. Repeat Optimize design: If the resonance frequency of some diaphragms is found to deviate from the target range, adjust the relevant design parameters and repeat the simulation until the resonance frequency of all diaphragms meets the design requirements; Choose a suitable piezoelectric material (e.g. PZT, ZnO, AlN).

[0040] Select the substrate material (e.g. SOI wafer, PDMS, PE, or PI).

[0041] PZT (Lead Zirconate Titanate): has a high piezoelectric coefficient and good temperature stability, suitable for high-frequency applications.

[0042] ZnO (zinc oxide): has good chemical stability and a high piezoelectric coefficient, and is suitable for high and low frequency applications.

[0043] AlN (aluminum nitride): has good thermal stability and dielectric properties, is suitable for high-frequency applications, and can still maintain good performance at high temperatures.

[0044] Selection criteria Piezoelectric coefficient: A higher piezoelectric coefficient can improve conversion efficiency.

[0045] Mechanical quality factor: A higher mechanical quality factor helps improve the clarity of the resonance peaks.

[0046] Temperature stability: Good temperature stability can ensure consistent performance at different temperatures.

[0047] Chemical stability: Good chemical stability can extend the service life of the device; SOI (Silicon on Insulator): has good mechanical strength and chemical stability, suitable for the manufacture of MEMS devices.

[0048] PDMS (polydimethylsiloxane): It is a soft polymer with good flexibility and biocompatibility, suitable for structures that require flexibility or deformability.

[0049] PE (Polyethylene): has good mechanical strength and chemical stability and low cost.

[0050] PI (polyimide): has excellent thermal and chemical stability and is suitable for high temperature environments.

[0051] Selection criteria Mechanical strength: Sufficient mechanical strength to withstand the stress during processing.

[0052] Chemical stability: Good chemical stability can prevent the material from degrading during use.

[0053] Thermal stability: Good thermal stability can maintain performance in high temperature environments.

[0054] Compatibility with piezoelectric materials: The substrate material should have good adhesion and matching with the piezoelectric material.

[0055] Sputtering the first electrode material (e.g. Pt, Au, Cr, Al) on the substrate: The appropriate electrode material is selected according to the piezoelectric layer material and design requirements. For example, Pt (platinum), Au (gold), Cr (chromium) or Al (aluminum) metal can be used as the electrode material.

[0056] Pt and Au have good electrical conductivity and chemical stability and are suitable for use as electrode materials.

[0057] Cr and Al are low cost and have good conductivity, but may require additional protective layers to prevent oxidation or corrosion.

[0058] Sputtering process: The electrode material is deposited on the substrate using a sputtering process in a physical vapor deposition (PVD) technique.

[0059] Ensure that the sputtering conditions (such as working gas pressure, sputtering power, and target-to-substrate distance) are appropriate to obtain uniform and well-adhered films.

[0060] The sputtering time or deposition rate is controlled to achieve the desired electrode thickness, typically tens to hundreds of nanometers.

[0061] Quality Control: Monitor the quality of thin films during deposition to ensure there are no defects such as pinholes or inhomogeneities.

[0062] Measure the thickness of the electrode material to ensure it is within the designed range.

[0063] Etching to form the desired electrode pattern: Photoresist coating: A layer of photoresist is evenly coated on the surface of the deposited electrode material.

[0064] Use a spin coater for coating to ensure uniform distribution of the photoresist.

[0065] Exposure and Development: The photoresist is exposed using a mask to transfer the desired electrode pattern.

[0066] A developer is used to remove the photoresist in the exposed areas, leaving the photoresist in the unexposed areas as a mask for subsequent etching.

[0067] Dry or wet etching: Dry etching (such as reactive ion etching RIE) or wet etching (such as chemical etching) is used to remove the electrode material not masked by the photoresist to form a desired electrode pattern.

[0068] Cleaning and degumming: Residual photoresist is removed, usually using ion cleaning or chemical solvent cleaning.

[0069] Check the quality of the electrode pattern after cleaning to ensure there is no residual photoresist or other impurities.

[0070] Quality Check: Check the integrity, edge quality, and dimensional accuracy of the electrode patterns using an optical microscope or a scanning electron microscope (SEM).

[0071] Confirm that the electrode pattern is exactly the same as the design file.

[0072] Deposit the piezoelectric layer material, ensuring uniform thickness.

[0073] Deposition process of piezoelectric layer material: Prepare the deposition chamber: Clean the deposition chamber to ensure it is free of impurities.

[0074] Load the base and set the position of the support beams.

[0075] Deposition of piezoelectric layer: Select ALD or PLD deposition method.

[0076] Set deposition parameters such as temperature and gas flow rate.

[0077] Deposition was started and the film thickness was monitored by an online monitoring system.

[0078] Ensure that the film thickness is uniform and the deviation is within the allowable range.

[0079] Quality Check: The crystalline quality of the films was examined using XRD.

[0080] The surface roughness of the films was examined using AFM.

[0081] The film thickness uniformity was checked using an ellipsometer.

[0082] Adjustments and Optimizations: If the film thickness is not uniform or does not meet the design requirements, adjust the deposition parameters or re-deposit.

[0083] Preparation of the second electrode: sputtering a second electrode material on the piezoelectric layer, Etching to form the required electrode pattern. Sputtering the second electrode material on the piezoelectric layer, etching to form the required electrode pattern is consistent with the process of sputtering the first electrode material on the substrate.

[0084] Forming the groove gap: Etching is performed to form a first trench gap, a second trench gap and a third trench gap. Specifically: Photoresist coating: A layer of photoresist is uniformly coated on the completed electrode and piezoelectric layer structure.

[0085] Use a spin coater for coating to ensure uniform distribution of the photoresist.

[0086] Exposure and Development: The photoresist is exposed using a mask to transfer the desired trench gap pattern.

[0087] A developer is used to remove the photoresist in the exposed areas, leaving the photoresist in the unexposed areas as a mask for subsequent etching.

[0088] Dry etching: Dry etching (such as reactive ion etching (RIE)) is used to remove materials not masked by the photoresist to form the desired trench gap.

[0089] The first groove gap is located between the central circular diaphragm and the inner annular diaphragm.

[0090] The second groove gap is located between the inner annular diaphragm and the outer annular diaphragm.

[0091] The third groove gap is used to separate the outer annular vibration membrane from other structures.

[0092] Quality Check: Use an optical microscope or scanning electron microscope (SEM) to inspect the integrity of the groove gap, edge quality, and dimensional accuracy.

[0093] Confirm that the groove gap is exactly the same as the design file to ensure isolation between the diaphragms.

[0094] Etching forms a supporting beam structure to enhance structural stability.

[0095] The supporting beams are cross-shaped vertical beams, three cross-shaped beams or four cross-shaped beams.

[0096] A back cavity is formed by etching on the back side of the substrate.

[0097] Etching a back cavity on the back side of the substrate includes: A layer of photoresist is uniformly coated on the back side of the substrate.

[0098] Use a spin coater for coating to ensure uniform distribution of the photoresist.

[0099] Exposure and Development: The photoresist is exposed using a mask to transfer the pattern of the back cavity.

[0100] A developer is used to remove the photoresist in the exposed areas, leaving the photoresist in the unexposed areas as a mask for subsequent etching.

[0101] Test the actual frequency performance of the speaker.

[0102] Test the actual performance frequency of the speaker: Use a professional acoustic test system, such as a B&K (Brüel & Kjær) or other brand of acoustic analyzer.

[0103] Ensure that the test system has a broadband response, covering the frequency range of 20Hz to 20kHz.

[0104] The tests were conducted in an anechoic chamber to eliminate external noise interference.

[0105] Keep the temperature and humidity of the test environment stable to ensure the accuracy of the test results.

[0106] The test was performed using a swept frequency signal, gradually increasing from the lowest frequency (20 Hz) to the highest frequency (20 kHz).

[0107] The signal can be generated by a speaker driving circuit to ensure signal purity and stability.

[0108] Place the speaker on the test bench, making sure it is a fixed distance from the microphone.

[0109] The sound from the loudspeaker is captured using a microphone and the sound signal is recorded by an acoustic test system.

[0110] Use the acoustic test system's software to analyze the frequency response curve of the speaker output.

[0111] The curves are analyzed to determine the actual resonant frequency of each diaphragm.

[0112] Compare with the design target to evaluate whether the predetermined frequency range is achieved.

[0113] Specific test steps: Prepare the test environment: Make sure the temperature and humidity in the anechoic chamber are at stable levels.

[0114] Set up the test bench and microphone positions, making sure the distance between the speaker and microphone meets the test requirements.

[0115] Setting up the test system: Connect the acoustic test system and configure the appropriate hardware and software settings.

[0116] Set the test signal parameters, such as sweep frequency range and signal amplitude.

[0117] To test: Start the test signal and begin the frequency response test.

[0118] Record the loudspeaker output signal at different frequencies.

[0119] Data Analysis: The recorded data was processed using the acoustic test system's software to generate a frequency response curve.

[0120] Analyze the frequency response curves to determine the actual resonant frequency of each diaphragm.

[0121] Compare the actual resonant frequency with the design target to assess whether the expected frequency range is achieved.

[0122] Result evaluation: Evaluate whether the actual performance frequency of the speaker meets the design requirements based on the test results.

[0123] Adjust the thickness or shape of the piezoelectric layer based on the test results: Assuming that the test results show that the resonant frequency of the circular diaphragm is too high, the diameter of the circular diaphragm structure is increased; Assuming that the test results show that the resonant frequency of the circular diaphragm is too low, the diameter of the circular diaphragm structure is reduced; Assuming that the test result shows that the resonance frequency of the first annular vibration membrane is too high, the ring width of the first annular vibration membrane structure is increased; Assuming that the test result shows that the resonance frequency of the first annular vibration membrane is too low, the ring width of the first annular vibration membrane structure is reduced; Assuming that the test results show that the resonance frequency of the second annular vibration membrane is too high, the ring width of the second annular vibration membrane structure is increased; Assuming that the test results show that the resonance frequency of the second annular vibration membrane is too low, the ring width of the second annular vibration membrane structure is reduced; Preferably, adjusting the thickness or shape of the piezoelectric layer according to the test results also includes obtaining modification process data of the vibration membrane structure and using a particle swarm optimization algorithm to adjust the modification amount of the vibration membrane structure; including: After each test, record the actual frequency data of the loudspeaker, especially the resonant frequency of each diaphragm; Collect and test other parameters, including piezoelectric layer thickness and vibration membrane diameter; Organize the test data into tables or databases to facilitate subsequent analysis and optimization; The particle swarm optimization algorithm is used to adjust the modification amount of the vibration membrane structure: Set the size of the particle group, where each particle represents a set of possible parameter values ​​(including piezoelectric layer thickness and vibration membrane diameter); Initialize the position and velocity of each particle. The position represents a specific set of parameter values, and the velocity represents the trend of parameter change. Evaluate particle fitness: Conduct simulation or experimental testing on the parameter combination represented by each particle to obtain the corresponding fitness value; The fitness function can be defined based on the closeness to the target frequency range, for example minimizing the sum of squared deviations from the target frequency range; Update particle positions and velocities: According to the rules of the particle swarm optimization algorithm, the position and velocity of each particle are updated; The update rule considers the influence of the global optimal solution and the personal optimal solution to guide the particles to move towards a better solution; Iterative Optimization: Repeat the evaluation and update steps until the termination condition is met (such as reaching the maximum number of iterations or the fitness value converges); Select the best particle position as the optimized parameter value; The structural parameters of the vibration membrane, such as the thickness of the piezoelectric layer and the diameter of the vibration membrane, are adjusted according to the optimization results.

[0124] Specific implementation steps Data preparation: Based on the previous tests, the actual resonance frequency data of each diaphragm was sorted out; Determine the parameters that need to be optimized, such as piezoelectric layer thickness and diaphragm diameter; Particle swarm initialization: Set the size of the particle group, for example, 50 particles; Randomly initialize the position and velocity of each particle to ensure that the particles are distributed in the parameter space; Fitness function definition: The fitness function is defined as the inverse of the sum of squares of deviations from the target frequency range, that is, the closer to the target frequency range, the higher the fitness; Particle update rules: Use the standard PSO update rule, considering the particle's current position, personal optimal position, and global optimal position; When updating particle speed and position, make sure the parameter values ​​remain within a reasonable range; Iterative optimization process: In each iteration, the fitness value of each particle is evaluated; Update the particle's position and velocity; Repeat the iteration until the preset maximum number of iterations is reached or the fitness value converges; Optimization results application: Adjust the structure of the vibration membrane according to the optimal parameter value obtained by optimization; In this way, the particle swarm optimization algorithm can be effectively used to automatically adjust the parameters of the vibration membrane structure and improve the accuracy and efficiency of the design; this method is particularly suitable for complex design processes that require multiple rounds of iteration and optimization; The piezoelectric layer is redeposited and adjusted.

[0125] Re-deposit the piezoelectric layer and ensure uniform thickness.

[0126] Repeat the previous etching steps to form the desired electrode pattern, trench gap and support beam structure.

[0127] Retest the actual frequency performance of the speaker to verify the effect of the adjustment.

[0128] Until the resonant frequencies of all diaphragms are within the target range.

[0129] The present invention proposes a method for preparing a piezoelectric MEMS loudspeaker with a double-annular surround circular vibration membrane structure, uses finite element analysis software for simulation, predicts the resonance frequency of each vibration membrane, and ensures the accuracy of the design.

[0130] Fine etching and photolithography processes have been implemented to form high-precision electrode patterns, groove gaps and support beam structures, which enhance the structural stability and acoustic performance of the speaker. Effective coverage of the full audio range of 20Hz-20kHz is achieved by precisely controlling the thickness, shape and material properties of the piezoelectric layer and dynamically adjusting the design parameters.

[0131] By precisely controlling the resonance frequencies of different vibration membranes (circular, annular), the speaker is ensured to have a smooth frequency response curve in the full audio range, thereby improving the sound quality.

[0132] The particle swarm optimization algorithm (PSO) is used to automatically adjust the parameters of the vibration membrane structure, such as the thickness of the piezoelectric layer and the diameter of the vibration membrane, which improves the accuracy and efficiency of the design.

[0133] Through the iterative optimization process, the best parameter combination can be found quickly, reducing the time and effort required for manual adjustments and accelerating the product development cycle.

[0134] Materials with suitable piezoelectric properties (such as PZT, ZnO, AlN), and substrate materials with good mechanical strength and chemical stability (such as SOI wafer, PDMS, PE or PI) are selected.

[0135] The sputtering process and deposition parameters were optimized to ensure high-quality preparation of electrodes and piezoelectric layers, improving the performance and reliability of the loudspeaker.

Claims

1. A method for preparing a piezoelectric MEMS smart speaker, characterized in that: include: The film vibration frequency is preliminarily calculated using the theoretical formula of the film vibration frequency according to the target frequency range; Use finite element analysis software to simulate and predict the resonance frequency of each vibration membrane to determine the material of each layer, or directly determine the material of each layer based on empirical data; select piezoelectric materials; Selecting a substrate material; sputtering a first electrode material on the substrate, and etching to form a desired electrode pattern; Deposit the piezoelectric layer material to ensure uniform thickness; Sputter the second electrode material on the piezoelectric layer and etch to form the required electrode pattern; etch to form the first groove gap, the second groove gap and the third groove gap, and use a cross-shaped vertical beam, a three-cross beam or a four-cross beam as the support beam; etch to form a back cavity on the back of the substrate; test the actual performance frequency of the speaker; adjust the thickness or shape of the piezoelectric layer according to the test results: Assuming that the test results show that the resonant frequency of the circular diaphragm is too high, the diameter of the circular diaphragm structure is increased; Assuming that the test results show that the resonant frequency of the circular diaphragm is too low, the diameter of the circular diaphragm structure is reduced; Assuming that the test result shows that the resonance frequency of the first annular vibration membrane is too high, the ring width of the first annular vibration membrane structure is increased; Assuming that the test result shows that the resonance frequency of the first annular vibration membrane is too low, the ring width of the first annular vibration membrane structure is reduced; Assuming that the test results show that the resonance frequency of the second annular vibration membrane is too high, the ring width of the second annular vibration membrane structure is increased; Assuming that the test results show that the resonance frequency of the second annular vibration membrane is too low, the ring width of the second annular vibration membrane structure is reduced; and also including using a dynamic algorithm to adjust the thickness or shape of the piezoelectric layer; The piezoelectric layer is redeposited and adjusted until the resonant frequencies of all diaphragms are within the target range.

2. The method for preparing a piezoelectric MEMS smart speaker according to claim 1, characterized in that: Finite element analysis software is used to simulate and predict the resonant frequency of each diaphragm to determine the material of each layer, specifically: Based on the designed double-ring surrounding circular diaphragm structure, a geometric model was created using CAD software; Input the physical and mechanical properties of each layer material; Define the fixed endpoints of the support beam as boundary conditions, as well as the pressure conditions for the back chamber; For the diaphragm area, a finer mesh is used to improve the simulation accuracy; Use a coarser grid in non-critical areas; Applying a voltage between the first electrode and the second electrode to simulate the excitation of the speaker in a working state; Conduct frequency response analysis, gradually increase the excitation frequency to find the resonant frequency of each diaphragm; Perform modal analysis to extract the first few natural frequencies and corresponding vibration modes; Further frequency response analysis is performed to confirm the stability of the resonant frequency and its corresponding vibration mode; Compare the simulation results with the theoretical calculation results to verify the effectiveness of the model; Based on the simulation results, adjust the thickness, shape, or material properties of the piezoelectric layer to achieve the target resonant frequency range.

3. The method for preparing a piezoelectric MEMS smart speaker according to claim 1, characterized in that: Piezoelectric materials include PZT, ZnO, and AlN; substrate materials include SOI wafer, PDMS, PE, and PI.

4. The method for preparing a piezoelectric MEMS smart speaker according to claim 1, characterized in that: Sputtering a first electrode material on a substrate includes: Depositing electrode material on the substrate using a sputtering process in a physical vapor deposition technique; Controlling the sputtering time or deposition rate to achieve the desired electrode thickness; Monitor the quality of thin films during deposition; Measure the thickness of electrode materials; Etching to form the desired electrode pattern: Evenly coating a layer of photoresist on the surface of the deposited electrode material; Spin coater is used for coating to ensure uniform distribution of photoresist; exposing the photoresist using a mask to transfer the desired electrode pattern; Using a developer to remove the photoresist in the exposed area, leaving the photoresist in the unexposed area as a mask for subsequent etching; Using dry etching or wet etching to remove the electrode material not masked by the photoresist to form a desired electrode pattern; Removing residual photoresist; Check the quality of the electrode pattern after cleaning to ensure there is no residual photoresist or other impurities; Check the integrity, edge quality, and dimensional accuracy of the electrode pattern using an optical microscope or scanning electron microscope; Confirm that the electrode pattern is exactly the same as the design file.

5. The method for preparing a piezoelectric MEMS smart speaker according to claim 1, characterized in that: The process of depositing piezoelectric layer materials includes: Prepare the deposition chamber: Clean the deposition chamber to ensure it is free of impurities; Load the substrate and set the position of the support beams; Select ALD or PLD deposition method and set deposition parameters; Start deposition and monitor film thickness via an online monitoring system; Ensure that the film thickness is uniform and the deviation is within the allowable range; The crystalline quality of the film was checked using XRD; Use AFM to examine the surface roughness of the film; Check the uniformity of film thickness using ellipsometer; If the film thickness is not uniform or does not meet the design requirements, adjust the deposition parameters or re-deposit.

6. The method for preparing a piezoelectric MEMS smart speaker according to claim 1, characterized in that: The steps of sputtering the second electrode material on the piezoelectric layer and etching to form the required electrode pattern are the same as the steps of sputtering the first electrode material on the substrate.

7. The method for preparing a piezoelectric MEMS smart speaker according to claim 1, characterized in that: Etching to form the first trench gap, the second trench gap and the third trench gap specifically: uniformly coating a layer of photoresist on the completed electrode and piezoelectric layer structure; Use a spin coater for coating to ensure uniform distribution of the photoresist; exposing the photoresist using a mask to transfer a desired trench gap pattern; Using a developer to remove the photoresist in the exposed area, leaving the photoresist in the unexposed area as a mask for subsequent etching; Using dry etching to remove materials not masked by the photoresist to form the desired trench gap; The first groove gap is located between the central circular diaphragm and the inner annular diaphragm; The second groove gap is located between the inner annular diaphragm and the outer annular diaphragm; The third groove gap is used to separate the outer annular vibration membrane from other structures.

8. The method for preparing a piezoelectric MEMS smart speaker according to claim 1, characterized in that: Etching a back cavity on the back side of the substrate includes: Evenly coating a layer of photoresist on the back side of the substrate; Spin coater is used for coating to ensure uniform distribution of photoresist; exposing the photoresist using a mask to transfer the pattern of the back cavity; A developer is used to remove the photoresist in the exposed areas, leaving the photoresist in the unexposed areas as a mask for subsequent etching.

9. The method for preparing a piezoelectric MEMS smart speaker according to claim 1, characterized in that: Test the actual performance frequency of the speaker: Set up the test bench and microphone positions, making sure the distance between the speaker and microphone meets the test requirements; Connect to the acoustic test system and configure the corresponding hardware and software settings; Set the parameters of the test signal, including the sweep frequency range and signal amplitude; Start the test signal and start the frequency response test; Record the output signal of the speaker at different frequencies; Process the recorded data using the acoustic test system's software to generate a frequency response curve; Analyze the frequency response curve to determine the actual resonant frequency of each diaphragm; Compare the actual resonant frequency with the design target to assess whether the expected frequency range is achieved; Evaluate whether the actual performance frequency of the speaker meets the design requirements based on the test results.

10. The method for preparing a piezoelectric MEMS smart speaker according to claim 1, characterized in that: Obtain the modification process data of the vibration membrane structure, and use the particle swarm optimization algorithm to adjust the modification amount of the vibration membrane structure, including: Based on the previous tests, the actual resonance frequency data of each diaphragm was sorted out; Determine the parameters that need to be optimized, including the thickness of the piezoelectric layer and the diameter of the vibration membrane; Particle swarm initialization: Set the size of the particle swarm; Randomly initialize the position and velocity of each particle to ensure that the particles are distributed in the parameter space; The fitness function is defined as the inverse of the sum of squares of deviations from the target frequency range, that is, the closer to the target frequency range, the higher the fitness; Use the standard PSO update rule, taking into account the particle’s current position, personal optimal position, and global optimal position; When updating particle speed and position, make sure the parameter values ​​remain within a reasonable range; Iterative optimization process: In each iteration, the fitness value of each particle is evaluated; Update the particle's position and velocity; Repeat the iteration until the preset maximum number of iterations is reached or the fitness value converges; According to the optimal parameter values ​​obtained through optimization, the structure of the vibration membrane is adjusted.

Images