An ultrasonic energy and pressure synergistically coupled ultrasonic bonding process

Abstract

The present application relates to the field of high-precision manufacturing technology, and more particularly to an ultrasonic bonding process with ultrasonic energy and pressure synergistic coupling, comprising: presetting a bonding pressure target value and an ultrasonic power range based on the material matching of the lead; making the piezoelectric ceramic vibrate within the ultrasonic power range, transmitting the vibration to the wedge after amplification by the amplitude transformer; positioning the welding point coordinates according to the visual system, controlling the voice coil motor to drive the wedge to move to the target pad; measuring the bonding pressure applied value applied by the wedge in real time, and generating a driving force correction instruction for the piezoelectric ceramic according to the deviation between the bonding pressure applied value and the bonding pressure target value; executing the driving force correction instruction and generating a compensation displacement through the piezoelectric ceramic cooperating with the elastic amplification mechanism; in the moving and compensation process of the wedge, the magnetic rheological damper dynamically absorbs the kinetic impact energy until the bonding is completed. Through the present application, the problem of welding point microstructure defects caused by dynamic instability of the mechanical energy transmission link in the existing bonding technology is effectively solved.

Description

Technical Field

[0001] This invention relates to the field of high-precision manufacturing technology, and in particular to an ultrasonic bonding process that synergistically couples ultrasonic energy and pressure. Background Technology

[0002] In the field of microelectronic packaging manufacturing, wire bonding technology is a key process for achieving electrical connections between chips and external circuits. Among them, ultrasonic bonding technology transfers mechanical vibration energy to the metal wire interface through bonding pins, and utilizes the local frictional heat effect to achieve atomic diffusion bonding between materials.

[0003] Existing technologies generally rely on ultrasonic energy output systems with fixed parameters. However, when bonding machines operate under high-frequency conditions, the elastic deformation, thermal drift effect, and multidimensional stress coupling of the mechanical transmission link can cause dynamic fluctuations in the energy transfer path. Especially when processing high-hardness alloy materials or micro-pads, the mechanical energy flow will experience axial component attenuation and radial phase shift during three-dimensional spatial transmission. This mechanical instability causes abnormal distribution of effective power density at the welding interface, resulting in non-uniform variations in the metal atom diffusion process. Ultimately, two typical defect modes are formed within the weld joint: one is excessive lattice distortion caused by energy accumulation in local micro-regions, forming a brittle intermetallic compound segregation layer; the other is insufficient diffusion in the energy flow attenuation region, resulting in a microscale unbonded interface.

[0004] These microstructural defects not only directly affect the electrical continuity of the solder joint, but also become crack initiation sources under subsequent thermal cycling loads, leading to reliability issues such as device hermeticity failure. The industry has attempted to improve this by increasing manufacturing precision or optimizing control algorithms, but none of these efforts have fundamentally solved the dynamic instability problem of the mechanical energy transfer link, particularly failing to eliminate energy temporal fluctuations caused by the hysteresis of the mechanical structure. Therefore, there is an urgent need for a stability compensation mechanism directly coupled to the bonding machine tool body to directly correct dynamic deviations in the mechanical energy conduction process, thereby ensuring the uniform and dense microstructure of the solder joint.

[0005] The information disclosed in this background section is intended only to enhance the understanding of the general background of this disclosure and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] This invention provides an ultrasonic bonding process that synergistically couples ultrasonic energy and pressure, which can effectively solve the problems in the prior art.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] An ultrasonic bonding process involving the synergistic coupling of ultrasonic energy and pressure, the process comprising:

[0009] Based on the lead wire material matching the preset bonding pressure target value and ultrasonic power range;

[0010] The piezoelectric ceramic is driven to vibrate within the ultrasonic power range by an ultrasonic generator, and the vibration is amplified by an amplitude transformer and transmitted to the chopping tool.

[0011] Based on the coordinates of the solder joint located by the subpixel-level vision system, the voice coil motor is controlled to drive the cutting tool to move to the target solder pad;

[0012] The bonding pressure applied by the chopping blade is measured in real time, and a driving force correction command for the piezoelectric ceramic is generated based on the deviation between the bonding pressure applied value and the target bonding pressure value.

[0013] The piezoelectric ceramic, in conjunction with the elastic amplification mechanism, executes the driving force correction command and generates a micron-level compensation displacement.

[0014] During the movement and compensation process of the cutting blade, the impact energy of the motion is dynamically absorbed by the magnetorheological damper until the bonding is completed.

[0015] Furthermore, based on the lead material matching the preset bonding pressure target value and ultrasonic power range, including:

[0016] The material category of the lead is determined using the elemental spectral characteristics of the lead.

[0017] Based on the elastic-plasticity index and surface hardness parameter of the material type, the upper and lower limits of the target bonding pressure value are obtained from the pre-stored parameter relationship table.

[0018] Based on the energy absorption characteristics and thermal conductivity parameters of the material type, and combined with the transmission efficiency of the amplitude transformer, a power reference value that guarantees minimum bonding strength within the ultrasonic power range is calculated.

[0019] Furthermore, the power reference value is calculated by combining the transmission efficiency of the amplitude transformer, including:

[0020] The actual energy transfer efficiency curve under normal working conditions is established by combining the output power monitoring data of the ultrasonic generator with the vibration sensor data of the amplitude transformer mounting base.

[0021] The theoretical boundary value of the effective energy required for the lead wire to absorb is calculated based on the energy absorption characteristics and the thermal conductivity parameter, and then the actual environmental loss compensation of the amplitude rod mounting base is superimposed.

[0022] After performing energy adaptation mapping on the theoretical boundary value acceleration point based on the actual efficiency curve, a power reference value allocation scheme covering the ultrasonic power range is generated.

[0023] Further, generating the driving force correction command for the piezoelectric ceramic includes:

[0024] The bonding pressure applied value is continuously collected and the real-time deviation from the target bonding pressure value is calculated. At the same time, the direction change and duration characteristics of the real-time deviation are recorded.

[0025] A reverse proportional adjustment component is generated based on the numerical proportion of the real-time deviation.

[0026] The cumulative effect generated by combining the real-time deviation with the duration characteristic generates an integral compensation component.

[0027] The real-time deviation is fused with the differential suppression component generated by the direction change rate;

[0028] The inverse proportional adjustment component, the integral compensation component, and the differential suppression component are linearly superimposed and converted into the electric field strength command value of the corresponding piezoelectric ceramic as the driving force correction command.

[0029] Further, generating a reverse proportional adjustment component based on the numerical proportion of the real-time deviation includes:

[0030] Calculate the ratio of the absolute value of the real-time deviation to the pressure system range defined by the target value of the bonding pressure;

[0031] Based on the elastic modulus characteristics of the piezoelectric ceramic during operation, the ratio is converted into the required driving force correction range for the piezoelectric ceramic.

[0032] The inverse proportional adjustment component is generated by reversing the direction of the real-time deviation and combining it with the driving force correction amplitude.

[0033] Furthermore, an integral compensation component is generated by combining the cumulative effect produced by the duration feature, including:

[0034] For the continuous time period corresponding to the duration feature, the real-time deviation is continuously accumulated to obtain the cumulative integral deviation;

[0035] Based on the displacement response capability of the piezoelectric ceramic and the elastic amplification mechanism, the cumulative integral deviation is converted into the intensity of the integral compensation component.

[0036] Furthermore, the differential suppression component generated by fusing the rate of change of direction includes:

[0037] The instantaneous switching characteristics of the direction of the real-time deviation are measured, and the rate of change of direction is converted into the change of viscous resistance based on the viscous fluid elastoplastic response characteristics of the magnetorheological damper.

[0038] The linear displacement reverse drive mechanism controlling the voice coil motor generates a reverse damping compensation displacement based on the change in viscous resistance, and maps the reverse damping compensation displacement to the amplitude characteristics of the differential suppression component.

[0039] Further, executing the driving force correction command and generating micron-level compensated displacement includes:

[0040] The driving force correction command is converted into a potential field adjustment signal for the piezoelectric ceramic, and the crystal lattice of the piezoelectric ceramic is deformed based on the inverse piezoelectric effect to generate an initial displacement.

[0041] The initial displacement generated by the piezoelectric ceramic is amplified step by step through the mechanical lever amplification structure of the elastic amplification mechanism to obtain the amplified mechanical compensation displacement.

[0042] The amplified mechanical compensation displacement is transmitted to the base connection of the chopping blade as the micron-level compensation displacement to correct the pressure application position of the chopping blade.

[0043] Furthermore, dynamic absorption of motion impact energy is achieved through magnetorheological dampers, including:

[0044] The kinetic impact energy generated by the chopping blade during movement and compensation is transmitted to the piston drive shaft of the magnetorheological damper through a rigid connection assembly;

[0045] The magnetorheological fluid in the cavity of the magnetorheological damper is driven to generate shear deformation flow under piston extrusion, and viscous resistance is formed in the opposite direction of impact based on the magnetic field-dependent viscosity change characteristics of the magnetorheological fluid.

[0046] The impact kinetic energy is continuously dissipated by the viscous resistance and converted into molecular frictional heat energy until the motion returns to static equilibrium when the chopping blade completes the bonding operation.

[0047] Furthermore, it also includes:

[0048] After the cutting tool moves to the target pad, the sub-pixel vision system performs positioning verification and outputs a position ready signal to the control system to initiate an ultrasonic bonding trigger command.

[0049] Driven by the ultrasonic bonding trigger command, the measurement process of the vibration output of the ultrasonic generator and the applied bonding pressure value is started synchronously.

[0050] When the magnetorheological damper detects the ultrasteady oscillation characteristics of the motion impact energy, it autonomously increases the magnetic field strength and the viscous drag value until the impact dissipates.

[0051] The technical solution of this invention can achieve the following technical effects:

[0052] By presetting the target bonding pressure value and the ultrasonic power reference range, and then transmitting vibration energy to the wedge through the piezoelectric ceramic after amplification by the amplitude transformer, the wedge is then precisely moved to the target pad through subpixel visual positioning and voice coil motor coordinated control. Subsequently, the applied bonding pressure value is measured in real time, and the driving force correction command is generated based on the PID control component to execute the compensation displacement. Finally, the impact energy is dynamically dissipated through the viscous resistance of the magnetorheological damper until static equilibrium is reached. This effectively solves the problem of microstructural defects of the solder joint caused by the dynamic instability of the mechanical energy transmission link in the existing bonding technology.

[0053] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 This is a process flow diagram of an ultrasonic bonding process that uses the synergistic coupling of ultrasonic energy and pressure.

[0056] Figure 2 This is a flowchart illustrating the process of calculating the ultrasonic power reference value;

[0057] Figure 3 A flowchart illustrating the process of generating the driving force correction command for piezoelectric ceramics;

[0058] Figure 4 This is a schematic diagram of the bond displacement compensation process. Detailed Implementation

[0059] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0060] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0061] Example 1;

[0062] like Figure 1 As shown, this application provides an ultrasonic bonding process that synergistically couples ultrasonic energy and pressure, the process including:

[0063] S10: Based on the lead wire material matching the preset bonding pressure target value and ultrasonic power range;

[0064] S20: The piezoelectric ceramic is driven to vibrate within the ultrasonic power range by an ultrasonic generator, and the vibration is amplified by an amplitude transformer and transmitted to the chopping tool.

[0065] S30: Based on the sub-pixel level vision system, locate the solder joint coordinates and control the voice coil motor to drive the cutting tool to move to the target solder pad;

[0066] S40: Real-time measurement of the bonding pressure applied by the cleaver, and generation of a piezoelectric ceramic driving force correction command based on the deviation between the bonding pressure applied value and the target bonding pressure value;

[0067] S50: Through piezoelectric ceramics and elastic amplification mechanism, it executes driving force correction commands and generates micron-level compensation displacement;

[0068] S60: During the movement and compensation of the cleaver, the impact energy of the motion is dynamically absorbed by the magnetorheological damper until the bonding is completed.

[0069] Specifically, firstly, based on the material and diameter of the wire to be bonded, the type and surface condition of the pad metal layer, a material matching table is pre-established in the process formulation to obtain the target bonding pressure value and the corresponding ultrasonic power range. For example, for gold wire with good ductility, a relatively low target pressure value can be selected and matched with a medium ultrasonic power range to avoid damaging the pad; for copper wire with higher hardness and oxidation sensitivity, a higher target pressure value and a relatively high ultrasonic power range are preferred to ensure frictional film breaking and intermetallic diffusion; for aluminum wire, a wider ultrasonic power range can be matched with a medium target pressure value to improve the process window. The above target values ​​and ranges can be solidified from historical yield data or determined after the first test soldering. Fine-tuning is performed by writing back the tensile or shear test results to ensure a reliable calibration reference for subsequent real-time coupling control. Then, an ultrasonic generator drives a piezoelectric ceramic to produce high-frequency mechanical vibration within the ultrasonic power range. The piezoelectric ceramic preferably uses a stacked structure with impedance matching to the generator to improve energy conversion efficiency, and a pre-tightening method ensures stable vibration output under applied pressure. This vibration is amplified by an amplitude transformer and transmitted to the cutting tool. The amplitude transformer is preferably made of high-strength metal and has structural dimensions adapted to the operating frequency to reduce heat drift and energy loss. Simultaneously, a high coaxiality fit and anti-loosening structure are preferably used at the connection interface between the cutting tool and the amplitude transformer to avoid uneven weld points caused by vibration eccentricity. During the solder joint positioning stage, a subpixel-level vision system acquires images and performs subpixel fitting calculations on the solder pads or solder joint marks to output solder joint coordinates. The vision system preferably performs distortion correction and coordinate system calibration before bonding, and incorporates adaptive settings of the light source angle and exposure parameters to enhance the contrast of the solder pad edges, thereby improving subpixel positioning stability. Then, a voice coil motor drives the cutting tool to move rapidly along a predetermined trajectory to the target solder pad and press down for contact. The voice coil motor preferably employs closed-loop displacement control and acceleration / deceleration curve limitations to balance speed and jitter suppression. When the cutting tool contacts the solder pad and enters the pressure application stage, the bonding pressure applied by the cutting tool is measured in real time. Preferably, a high-bandwidth force sensor is placed along the cutting tool's loading path. For example, strain gauge or piezoelectric force measurement structures are used, and zero-point drift compensation is performed so that they can still output pressure signals that can be used for control under the background of superimposed ultrasonic vibration. The controller generates a driving force correction command for the piezoelectric ceramic based on the deviation between the applied bonding pressure value and the target bonding pressure value. The correction is preferably manifested as dynamic adjustment of the driving voltage or driving waveform amplitude of the piezoelectric ceramic, so that the ultrasonic vibration energy and the actual pressure state are coordinated: when the pressure is detected to be lower than the target value, the driving force correction command preferably makes the piezoelectric ceramic increase the effective vibration output within the allowable ultrasonic power range and cooperate with the elastic amplification mechanism to generate a downward micron-level compensation displacement, thereby making up for the contact pressure without significantly increasing the macroscopic downward pressing speed.When the detected pressure exceeds the target value, it is preferable to reduce the effective drive of the piezoelectric ceramic and release part of the displacement through an elastic amplification mechanism to reduce the instantaneous pressure, thus preventing the solder pad metal layer from being squeezed or the lead wire from necking and breaking. The elastic amplification mechanism preferably uses a combination of an elastic element with linear recovery characteristics and a displacement transmission structure, enabling the small-stroke output of the piezoelectric ceramic to be stably amplified into a controllable micron-level compensated displacement. Pre-tightening and limiting designs ensure that no gap impact or plastic deformation occurs during repeated bonding cycles. Simultaneously, during the movement and compensation of the piezoelectric ceramic, measures are taken to suppress rapid positioning and downward contact of the voice coil motor and dynamic compensation of the piezoelectric ceramic. To mitigate the impact energy, a magnetorheological damper is preferably installed in the motion transmission path, and its damping level is dynamically adjusted according to the motion state: at the moment of rapid approach and initial contact of the blade, the damping is increased to absorb the impact and prevent rebound; during the stable ultrasonic bonding stage, the damping is reduced to minimize the impact on vibration transmission and maintain energy efficiency; at the end of the bonding and blade lifting stage, the damping is increased again to suppress blade lifting ringing, thereby achieving the effect of dynamically absorbing the motion impact energy until bonding is complete. For example, when ball bonding or wedge bonding gold wires with a diameter of approximately 20 to 30 micrometers, the target pressure value can be set to a lower level and the ultrasonic power range can be adjusted. The pressure setting is set to medium, and sub-pixel positioning ensures the solder joint falls in the center of the pad. If slight contamination on the pad surface leads to insufficient initial friction and a temporary low pressure feedback, the controller immediately outputs a drive force correction command. This increases the piezoelectric ceramic drive and provides micron-level compensation displacement via an elastic amplification mechanism, allowing the contact pressure to quickly return to the target value. Simultaneously, the magnetorheological damper absorbs the instantaneous mechanical impact caused by the compensation, ultimately resulting in a solder joint with uniform morphology and sufficient tensile strength. Similarly, when bonding copper wire, because the material is harder and more prone to oxidation, the target pressure value can be set to a higher setting, and the ultrasonic power range can be set to a higher setting. If a fault is detected during the pressing process... To mitigate pressure overshoot, the system promptly suppresses overpressure by reducing the piezoelectric ceramic drive and releasing compensating displacement. Simultaneously, a magnetorheological damper increases damping to suppress structural vibrations caused by overshoot, thereby reducing the risk of pad damage while ensuring both film rupture and metallurgical bonding. Through this method, a synergistic link is established: real-time pressure measurement, deviation correction command generation, piezoelectric ceramic combined with an elastic amplification mechanism forming micron-level compensating displacement, and magnetorheological damping dynamically absorbing impact. This couples and controls ultrasonic energy and pressure throughout the bonding process, resulting in more stable bonding consistency, a wider process window, and higher yields regardless of lead material and pad condition.

[0070] The technical solution of this invention presets a target bonding pressure value and an ultrasonic power reference range. Then, the piezoelectric ceramic transmits vibration energy to the wedge after amplification by the amplitude transformer. Subsequently, the wedge is precisely moved to the target pad by subpixel visual positioning and voice coil motor coordinated control. Then, the bonding pressure applied value is measured in real time, and a driving force correction command is generated based on the PID control component to execute the compensation displacement. Finally, the impact energy is dynamically dissipated by the viscous resistance of the magnetorheological damper until static equilibrium is reached. This effectively solves the problem of microstructural defects of the solder joint caused by the dynamic instability of the mechanical energy transmission link in the existing bonding technology.

[0071] Furthermore, based on the lead material matching the preset bonding pressure target value and ultrasonic power range, including:

[0072] The material type of the lead is determined using the elemental spectral characteristics of the lead.

[0073] Based on the elastic-plastic properties and surface hardness parameters of the material type, the upper and lower limits of the target bonding pressure are obtained from the pre-stored parameter relationship table.

[0074] Based on the energy absorption characteristics and thermal conductivity parameters of the material type, and combined with the transmission efficiency of the amplitude transformer, the power reference value that guarantees the minimum bonding strength within the ultrasonic power range is calculated.

[0075] As a preferred embodiment of the above, firstly, after feeding or changing the wire and before formal bonding, elemental spectral characteristics of the wire to be bonded are collected to determine the material category. Preferably, a spectral acquisition method with high integration with the wire bonding equipment and good adaptability to fine metal wires is used. The acquisition location is preferably selected from the straight exposed section of the wire or a stable region near the clamping point. Drift caused by differences in ambient light and reflection is suppressed by setting a fixed acquisition distance, incident angle, and integration time. Regarding spectral feature extraction, a set of characteristic peaks that characterize the elemental composition of the material is preferably extracted and matched with a pre-established material spectral fingerprint database to determine whether the wire is a gold wire, copper wire, aluminum wire, or a common alloy category. The spectral fingerprint database preferably comes from... Results are collected from standard samples from the same batch or supply system, and normalization is allowed for peak intensity differences of leads with different diameters to improve identification robustness. After determining the material type, the process proceeds to the mapping step of the pressure target value range. Preferably, a parameter relationship table of material type, elastoplastic index, surface hardness parameter, and pressure target range is pre-stored in the equipment formula. The elastoplastic index preferably adopts a combination of indicators that can reflect the yield and rebound characteristics of the material under micro-pressure contact, such as graded parameters characterizing yield tendency and ductility. The surface hardness parameter preferably adopts a graded parameter that can distinguish between softness and hardness and indentation sensitivity. All of the above parameters are obtained from the material specifications, supplier data, or files cured after standard sample testing in an engineering-feasible manner. In terms of query logic, it is preferable to use material category as the primary key and elasticity index and surface hardness parameter as secondary indexes. The upper and lower limits of the target bonding pressure value are directly obtained from the parameter relationship table. Furthermore, recommended landing points within the range are set according to the pressure resistance sensitivity of the target pad metal layer. This ensures that for materials with higher ductility and softer surfaces, a lower pressure range is preferred to reduce the risk of pad damage, while for materials with higher hardness and more pronounced rebound, a higher pressure range is preferred to ensure effective contact and frictional film breakage. Subsequently, to obtain the ultrasonic power range matching the material category and the power reference value that ensures minimum bonding strength, it is preferable to pre-store material category, energy absorption characteristics, and thermal conductivity parameters in the same formulation system. The material's thermal properties are categorized, with energy absorption characteristics preferably used to characterize the material's ability to dissipate input energy during ultrasonic friction and plastic deformation, and thermal conductivity parameters preferably used to characterize the diffusion rate of heat in the contact area between the lead and the pad. These two parameters are then linked to the transmission efficiency of the amplitude transformer to form a power reference determination rule executable by the equipment. Specifically, the transmission efficiency of the amplitude transformer is preferably obtained through factory calibration or periodic self-testing and written into the equipment parameters in the form of the available energy transmission ratio. When calculating the power reference value, the effective ultrasonic input required to achieve the minimum bonding strength is preferably taken as the target: when the material has stronger energy absorption or higher thermal conductivity, resulting in more easily dissipated effective energy in the solder joint area, the power reference value is preferably increased accordingly.When the amplitude transformer's transmission efficiency decreases due to assembly conditions or frequency deviations, the power reference value is preferably increased to compensate for transmission losses. This ensures that the final determined power reference value falls within a preset ultrasonic power range and serves as the lower limit reference for that range. Furthermore, the upper limit of the ultrasonic power range is preferably set based on pad tolerance and lead necking risk to avoid excessive ultrasonic waves causing pad metal layer peeling or lead damage.

[0076] Furthermore, such as Figure 2 As shown, the power reference value is calculated based on the transmission efficiency of the amplitude transformer, including:

[0077] The actual energy transfer efficiency curve under normal working conditions is established by combining the output power monitoring data of the ultrasonic generator with the vibration sensor data of the amplitude transformer mounting base.

[0078] The theoretical boundary value of the effective energy required for lead wire absorption is calculated based on the energy absorption characteristics and thermal conductivity parameters, and then the actual environmental loss compensation of the amplitude rod mounting base is superimposed.

[0079] After completing the energy adaptation mapping of the theoretical boundary value acceleration point based on the actual efficiency curve, a power reference value allocation scheme covering the ultrasonic power range is generated.

[0080] As a preferred embodiment of the above, firstly, an actual energy transfer efficiency curve under normal operating conditions is established by combining the output power monitoring data of the ultrasonic generator with the vibration sensor data of the amplitude transformer mounting base. Preferably, a rapid calibration is performed after the equipment is factory calibrated, after transducer or amplitude transformer maintenance, or before process switching. During calibration, the cutting tool is controlled to be in a non-bonded state or lightly touching a standard load to avoid pad disturbance. At the same time, the frequency or power is swept stepwise according to preset levels within the ultrasonic power range. The output power monitoring data of the ultrasonic generator is recorded in real time, preferably including instantaneous power, average power, and settling time. A vibration sensor matching the operating frequency band is preferably arranged at the amplitude transformer mounting base. By collecting the vibration response of the mounting base, the amplitude, effective value, or characteristic quantities representing vibration intensity related to energy coupling are preferably extracted. The generator output power at the same level is correlated with the base vibration response to obtain the actual energy transfer efficiency characterization value at this assembly state and frequency point. To ensure the reliability of the curve, the data is preferably denoised and steady-state filtered, that is, the start-up transient, significant temperature rise drift, and obviously abnormal vibration peak segments are removed. The actual efficiency curve of output power versus transfer efficiency is constructed by the mean or median of multiple repeated power scans. At the same time, the curve is accompanied by the recording of the installation information and environmental information, such as the base fastening status, temperature range, and equipment running time. The corresponding hot or cold state allows it to serve as the equipment-side basis for subsequent power reference value calculations. Secondly, after calculating the theoretical boundary value of the effective energy required for lead wire absorption based on the energy absorption characteristics and thermal conductivity parameters of the material type, the actual environmental loss compensation of the amplitude transformer mounting base is superimposed. Preferably, the energy absorption characteristics are used to characterize the effective absorption tendency of the lead wire in ultrasonic friction and plastic dissipation, and the thermal conductivity parameters are used to characterize the energy leakage tendency caused by heat diffusion in the bonding contact area. Both jointly determine the lower limit of the effective energy required to achieve the minimum bonding strength. This lower limit is preferably implemented by writing it into the process formula in the form of graded boundaries or graded thresholds, for example, for gold wire, copper wire, aluminum wire, or their composites. Several energy level boundaries are set for each gold, and the corresponding level can be selected according to the wire diameter or the sensitivity of the pad metal layer, thereby avoiding reliance on complex calculations in the manual and equipment implementation; when superimposing actual environmental loss compensation, it is preferable to extract the environmental loss characterization quantity from the low-frequency background vibration of the amplitude rod mounting base vibration sensor, the response attenuation caused by temperature rise, and the energy loss trend caused by changes in fastening state, and convert it into the compensation amount for the theoretical boundary value. This ensures that even if the material parameters are the same, the required input energy on the generator side can still be automatically adjusted up or down when the equipment is in different environments and assembly states, so as to avoid the situation where the material is correctly matched but the equipment transmits attenuation, resulting in insufficient bonding strength.Finally, based on the actual efficiency curve, after completing the energy adaptation mapping of the acceleration point at the theoretical boundary value, a power reference value allocation scheme covering the ultrasonic power range is generated. The acceleration point is preferably understood as the critical interval near the theoretical boundary value that is most sensitive to the improvement of bond strength and most easily affected by efficiency fluctuations. Preferably, this critical interval is set as several discrete mapping points, and the corresponding equivalent input power position is found in the actual efficiency curve. That is, using the effective energy at the theoretical boundary as the target, the required generator output power level is determined in reverse under the actual efficiency curve, and the allocation scheme is expanded upwards and downwards around this level to form a power reference value allocation scheme covering the ultrasonic power range. This allocation scheme is preferably not a single value, but a set of reference power and allowable upward and downward fluctuations that are updated according to changes in material type, wire diameter level, and amplitude transformer efficiency. Specifically, the power range is divided into several sub-intervals, each with a corresponding reference power value and its recommended dwell time or ramp-up strategy, so that subsequent real-time pressure and ultrasonic coupling control can be fine-tuned above this reference without exceeding the limits or becoming unstable.

[0081] Furthermore, such as Figure 3 As shown, the driving force correction command for generating piezoelectric ceramics includes:

[0082] The bonding pressure applied is continuously collected and the real-time deviation from the target bonding pressure value is calculated. At the same time, the direction change and duration characteristics of the real-time deviation are recorded.

[0083] A reverse proportional adjustment component is generated based on the numerical proportion of the real-time deviation.

[0084] The cumulative effect generated by combining real-time deviation with duration characteristics produces an integral compensation component.

[0085] The differential suppression component generated by the rate of change of the fusion direction of the real-time deviation;

[0086] The inverse proportional adjustment component, integral compensation component, and differential suppression component are linearly superimposed and converted into the electric field strength command value of the corresponding piezoelectric ceramic as the driving force correction command.

[0087] As a preferred embodiment of the above, after the cleaver contacts the bonding pad and applies pressure, the controller continuously acquires the bonding pressure applied value. The acquisition frequency is preferably higher than the dominant frequency of pressure fluctuations during the bonding process, enabling it to capture short-term fluctuations caused by ultrasonic coupling. Simultaneously, the acquired pressure signal is preferably first processed to suppress high-frequency ultrasonic components, for example, by using a low-pass or band-stop filter that is offset from the ultrasonic operating frequency band, or by using steady-state window sampling to extract the pressure envelope characteristics, in order to obtain the effective pressure applied value for control. Based on this, the real-time deviation between the effective pressure applied value and the target bonding pressure value is calculated, and the sign change of this real-time deviation is recorded simultaneously in each sampling period to characterize the direction change and the duration of the deviation. The duration of continuous movement in the same direction is used to characterize the duration feature. Preferably, the number of times the deviation changes from positive to negative or vice versa, along with the timing of these changes, is recorded as an event to quickly identify oscillation trends during pressure rebound, sudden changes in pad surface condition, or repeated boundary violations caused by micro-vibrations of the cutting tool. Subsequently, a reverse proportional adjustment component is generated based on the numerical proportion of the real-time deviation. Preferably, this component is set to an instantaneous correction amount where the larger the deviation, the stronger the reverse adjustment, used to quickly pull back when the pressure suddenly deviates from the target value. This component preferably includes a limiting strategy to prevent excessive correction during instantaneous deviation anomalies, such as spikes at the moment of contact. Furthermore, an integral compensation component is generated by combining the real-time deviation with the cumulative effect of the duration feature. Preferably, this is understood as a slow variable correction where the longer the deviation persists, the more significant the compensation. This is used to offset long-term small deviations caused by lead material springback, pad surface micro-roughness, or thermal drift of the mechanism. To avoid overshoot during the recovery phase due to excessive accumulation when the piezoelectric ceramic stroke or power range is limited, the integral compensation component preferably has anti-accumulation saturation constraints. For example, when the electric field strength command value has reached the allowable boundary or the deviation direction has reversed, the accumulation rate is reduced or the accumulation is frozen, preventing the integral term from accumulating further in the irreparable stage. Simultaneously, the real-time deviation quantity is fused with the direction change rate to generate a differential suppression component, preferably used as an advance suppression measure against excessively rapid deviation changes that are about to cause overshoot or oscillation. When deviation is detected... When the difference increases rapidly in a short period of time or frequent sign flips occur, the differential suppression component is preferably increased to weaken the impact of the correction action, thereby reducing mechanical ringing and pressure rebound of the contact system between the cutter and the pad. After obtaining the inverse proportional adjustment component, integral compensation component and differential suppression component, the three are linearly superimposed to obtain a comprehensive adjustment amount, and the comprehensive adjustment amount is converted into the electric field intensity command value of the corresponding piezoelectric ceramic as the driving force correction command. The conversion method preferably adopts the electric field application logic consistent with the polarization direction of the piezoelectric ceramic, and combines the safe working boundary of the piezoelectric ceramic to impose upper and lower limit constraints and slope restrictions, so that the electric field intensity command value can achieve micron-level compensation displacement, and will not cause over-excitation of the piezoelectric ceramic or cause structural impact due to sudden change in command.

[0088] Furthermore, an inverse proportional adjustment component is generated based on the numerical proportion of the real-time deviation, including:

[0089] Calculate the ratio of the absolute value of the real-time deviation to the pressure system range defined by the target value of the bonding pressure;

[0090] Based on the elastic modulus characteristics of piezoelectric ceramics during operation, the ratio is converted into the required driving force correction range for the piezoelectric ceramics.

[0091] The direction of the real-time deviation is reversed and combined with the driving force correction amplitude to generate a reverse proportional adjustment component.

[0092] As a preferred embodiment of the above, firstly, after the cleaver contacts the bonding pad and enters the effective pressure application stage, the controller obtains the bonding pressure applied value based on the force sensor output and calculates the real-time deviation between the applied value and the target bonding pressure value. To avoid high-frequency jitter in the deviation caused by the superposition of ultrasonic vibrations, which could lead to false triggering of the proportional term, it is preferable to perform envelope extraction or steady-state window sampling on the deviation without changing the deviation direction determination. This makes the obtained deviation more representative of the macroscopic contact pressure rather than the transient ultrasonic fluctuations. Subsequently, the absolute value of the real-time deviation is taken to characterize the current degree of deviation, and this absolute value is ratioized to the pressure system range defined by the target bonding pressure value to obtain a dimensionless normalized deviation ratio. The pressure system range is preferably understood as the range around the target bonding pressure value. The allowable pressure variation range of the target value and the pressure span that the control system can effectively adjust can be pre-defined by the process formula and associated with the material type, wire diameter, and pad tolerance. For example, a narrower range can be set for soft materials or sensitive pads to improve fine control resolution, while a wider range can be set for hard materials or scenarios with stronger requirements for breaking the film to retain sufficient correction margin. After obtaining the normalized deviation ratio, in order to further convert this ratio into the driving force correction amplitude required by the piezoelectric ceramic, it is preferable to introduce the elastic modulus characteristics of the piezoelectric ceramic during operation and combine it with the assembly state to complete the executable mapping between the ratio and the driving force amplitude. Specifically, the equivalent stiffness and elastic response of the piezoelectric ceramic differ under different preloads, temperature rises, and operating frequencies, therefore it is preferable to... After factory calibration or maintenance assembly, the correspondence between electric field strength change, output displacement, and equivalent force is obtained through micro-displacement response testing under standard load. This correspondence is then solidified into mapping rules using a lookup table or piecewise linear method, enabling the controller to directly look up or deduce the required driving force correction amplitude based on the normalized deviation ratio. Furthermore, to ensure the intuitive principle that a larger proportional deviation leads to stronger correction while avoiding over-excitation and impact, it is preferable to superimpose amplitude and slope limitations on the driving force correction amplitude: when the normalized deviation ratio is in a small deviation range, a smaller and smoother correction amplitude is output to avoid unnecessary disturbance to ultrasonic energy transfer; when the normalized deviation ratio enters a large deviation range, the correction amplitude can increase rapidly but not exceed [a certain value]. The piezoelectric ceramic's safe operating boundary and the elastic amplification mechanism's available stroke boundary ensure that proportional correction is both fast and does not exceed the limits. After obtaining the driving force correction amplitude, the controller reverses the direction of the real-time deviation to determine the correction direction. That is, when the applied pressure value is lower than the target value, the reverse proportional adjustment component preferably points in the direction of enhancing the piezoelectric ceramic drive to provide downward compensation displacement or equivalent pressure increase trend. When the applied pressure value is higher than the target value, the reverse proportional adjustment component preferably points in the direction of weakening the piezoelectric ceramic drive to release compensation displacement or equivalent pressure reduction trend. This reverse direction is combined with the aforementioned driving force correction amplitude to generate a reverse proportional adjustment component, so that it can participate in the linear superposition with a clear sign and an executable amplitude in each control cycle.

[0093] Furthermore, the cumulative effect generated by combining duration characteristics is used to generate integral compensation components, including:

[0094] For consecutive time periods corresponding to the duration feature, the real-time deviation is continuously accumulated to obtain the cumulative integral deviation;

[0095] Based on the displacement response capability of the piezoelectric ceramic and the elastic amplification mechanism, the cumulative integral deviation is converted into the intensity of the integral compensation component.

[0096] As a preferred embodiment of the above, the controller, based on continuously acquiring the bonding pressure applied value and obtaining the real-time deviation, first defines and updates the continuous time period using duration characteristics. The continuous time period preferably refers to the time interval during which the real-time deviation is continuously maintained in the same direction, that is, the continuous interval during which the deviation always shows either low pressure or high pressure. To avoid the superposition of ultrasonic vibrations causing the deviation sign to frequently jitter near the zero point and thus missegmenting the time period, a direction maintenance judgment condition is preferably introduced. For example, the time period is not triggered to reset when the absolute value of the deviation does not reach the controllable threshold, or it is still considered as the same continuous time period when the deviation direction briefly flips but does not exceed the set duration. This allows the integral term to focus on material springback and the microstructure of the solder pad surface. The cumulative integral deviation is caused by factors such as roughness variations and thermal drift of the mechanism, rather than by instantaneous zero crossing caused by high-frequency ultrasonic fluctuations. After determining a continuous time period, the controller continuously accumulates the real-time deviation within that period to obtain the cumulative integral deviation. Preferably, the accumulation process is limited to effective accumulation only within the continuous time period, and decays or resets to zero according to a strategy at the end of the time period or when the direction reverses. The accumulation input preferably uses the deviation characterization value after envelope extraction or steady-state window processing to ensure that the accumulated object represents the macroscopic contact pressure deviation rather than ultrasonic transient disturbances. At the same time, to ensure that the cumulative integral deviation does not cause overcompensation in engineering implementation, it is preferable to set a constraint strategy related to the bonding stage during the accumulation process, such as allowing faster bonding during the initial contact establishment stage. To overcome the persistently low level caused by the surface film, the cumulative acceleration rate is reduced during the stable bonding stage to avoid subsequent overshoot due to material softening or enhanced energy coupling. When the piezoelectric ceramic command is detected to be approaching the safety boundary or the elastic amplification mechanism is approaching the usable stroke boundary, the cumulative integral deviation is frozen or slowed down to prevent the cumulative effect from accumulating further when compensation is no longer possible. After obtaining the cumulative integral deviation, it is converted into the intensity of the integral compensation component based on the displacement response capabilities of the piezoelectric ceramic and the elastic amplification mechanism. Preferably, the calibrable displacement response capability of the equipment is used as the mapping basis: on the one hand, the equivalent displacement output capability of the piezoelectric ceramic varies under different preloads, temperature rises, and driving frequencies; on the other hand... The amplification factor and rebound characteristics of the elastic amplification mechanism on the micro-stroke of the piezoelectric ceramic also affect the micron-level compensation displacement ultimately formed at the cutting edge. Therefore, it is preferable to establish the correspondence between the electric field strength change, the piezoelectric ceramic output displacement, and the equivalent displacement or equivalent pressure increase trend of the cutting edge after elastic amplification through micro-displacement response testing under standard load after factory calibration or maintenance assembly. This correspondence should be solidified in a lookup table or piecewise linear manner so that the controller can map the cumulative integral deviation into the effect strength of the integral compensation component. This effect strength is preferably reflected as a slow variable correction amount of the electric field strength command value in linear superposition, and is also limited by the safe working boundary of the piezoelectric ceramic, the stroke boundary of the elastic amplification mechanism, and the pressure system range boundary allowed by the bonding process.

[0097] Furthermore, the differential suppression component generated by the rate of change of the fusion direction includes:

[0098] The instantaneous switching characteristics of the direction of the real-time deviation are measured, and the viscous fluid elastoplastic response characteristics of the magnetorheological damper are used to convert the rate of change of direction into the change of viscous resistance.

[0099] The linear displacement reverse drive mechanism controlling the voice coil motor generates a reverse damping compensation displacement based on the change in viscous resistance, and maps the reverse damping compensation displacement to the amplitude characteristics of the differential suppression component.

[0100] As a preferred embodiment of the above, after the cutting blade contacts the pad and enters effective pressure and ultrasonic vibration is superimposed, the controller continuously obtains the real-time deviation and measures its instantaneous direction switching characteristics. The instantaneous direction switching characteristics preferably include the time of deviation sign reversal, the slope of the deviation amplitude change before and after the reversal, and the degree of aggregation of the number of reversals per unit time, which are used to characterize whether the pressure system has entered a ringing state of rapid rebound, re-pressing, and re-rebound. To avoid misidentifying ultrasonic high-frequency perturbations as direction switching, it is preferable to still use the deviation characterization value of the deviation envelope or steady-state window as the criterion, and set a direction reversal confirmation condition with a minimum duration or minimum amplitude threshold, that is, only when the deviation crosses zero. A valid switch is determined only when the direction remains the same for a short period and reaches a controllable threshold after the initial point, thus ensuring that the rate of directional change truly reflects the dynamic instability of the mechanism and the contact interface, rather than the transient jitter caused by ultrasonic carrier waves. After obtaining the rate of directional change, the viscous fluid elastoplastic response characteristics of the magnetorheological damper are introduced as the physical mapping basis from rate to resistance: the magnetorheological damper exhibits adjustable viscous resistance output under different excitation conditions, and its resistance change corresponds to external velocity disturbances, exhibiting a stronger energy dissipation trend during impact or rapid reverse motion. Therefore, it is preferable to apply controlled small-amplitude reciprocating excitation to the axial direction of the cutter after factory calibration or maintenance assembly. The damper's response is recorded step by step, establishing a correspondence between the directional change rate setting and the viscous resistance change setting. This correspondence is then fixed using a lookup table or segmented rules, enabling the controller to quickly calculate the required viscous resistance change when an increase in the directional change rate is detected in real time. The viscous resistance change is preferably understood as the resistance correction trend that needs to be increased or decreased relative to the current damping level, so as to align with the damping adjustment action for dynamically absorbing kinetic impact energy. Subsequently, to ensure that this resistance change truly acts on the contact system between the cutter and the pad and forms a controllable vibration-damping displacement, a linear displacement reverse drive mechanism of a voice coil motor is preferably used for reverse damping compensation displacement output, i.e., control... The controller uses the change in viscous resistance as the reverse drive target. It compensates for the small displacement in the axial direction by outputting a voice coil motor that is opposite to the current motion trend. This causes a reverse damping effect when the cutting blade shows a tendency to rebound and a reverse damping effect when the cutting blade shows a tendency to overpressure and downward. In this way, the energy that should be released by the free vibration of the structure is converted into controllable damping dissipation. In terms of implementation details, the reverse drive compensation displacement is preferably set to a fast short-range action at the micrometer level or smaller, with amplitude limiting, slope limiting and interlocking strategy with the ultrasonic working stage. For example, in the stable bonding stage, it is preferred to use a small displacement and a high response speed to suppress vibration and avoid large displacement reverse drive to destroy the weld formation process.After obtaining the reverse damping compensation displacement, this compensation displacement is further mapped to the amplitude characteristics of the differential suppression component. That is, the reverse drive displacement strength required to offset the rate of directional change is converted into a differential suppression quantity in a linear superposition and incorporated into the generation of the electric field strength command value. This allows the piezoelectric ceramic drive to coordinate with the voice coil motor's reverse drive in a trend-based manner. When the rate of directional change indicates an impending overshoot, the differential suppression component preferably manifests as a weakening or smoothing of the overall drive, reducing the abruptness of the piezoelectric ceramic compensation action. Thus, oscillations are simultaneously suppressed on the mechanical side (magnetorheological damper and reverse drive displacement) and the execution side (piezoelectric ceramic electric field strength).

[0101] Furthermore, such as Figure 4 As shown, executing the driving force correction command and generating micron-level compensated displacement includes:

[0102] The driving force correction command is converted into a potential field adjustment signal for the piezoelectric ceramic, and the crystal lattice of the piezoelectric ceramic is deformed based on the inverse piezoelectric effect to generate an initial displacement.

[0103] The initial displacement generated by the piezoelectric ceramic is amplified step by step through the mechanical lever amplification structure of the elastic amplification mechanism to obtain the amplified mechanical compensation displacement.

[0104] The amplified mechanical compensation displacement is transmitted to the base connection of the chopping blade as a micron-level compensation displacement to correct the pressure application position of the chopping blade.

[0105] In a preferred embodiment of the above, the driving force correction command output by the controller is preferably first converted into a potential field adjustment signal for the piezoelectric ceramic. This potential field adjustment signal preferably manifests as a controlled adjustment of the amplitude, polarity, and rate of change of the driving voltage of the piezoelectric ceramic, and is combined with the polarization direction of the piezoelectric ceramic to set the application direction, causing the piezoelectric ceramic to undergo lattice deformation along a predetermined axis under the inverse piezoelectric effect, thereby outputting an initial displacement. To ensure that this initial displacement remains controllable and consistent under the background of high-frequency bonding vibration, it is preferable to set an upper limit constraint and a change slope constraint on the potential field adjustment signal, ensuring that it does not exceed the safe electric field operating boundary of the piezoelectric ceramic and also avoids excessive electric field deviation. The sudden change in piezoelectricity triggers structural impact, which can be adapted and corrected by considering the temperature rise state to offset the displacement drift caused by changes in the output capability of the piezoelectric ceramic under hot conditions. After the piezoelectric ceramic generates initial displacement, the initial displacement is preferably amplified stepwise through a mechanical lever amplification structure of an elastic amplification mechanism to obtain the amplified mechanically compensated displacement. The elastic amplification mechanism preferably adopts a lever amplification path composed of flexible hinges or elastic beams, so that the minute expansion and contraction of the piezoelectric ceramic can be converted into a larger end displacement output in a low-gap, low-friction, and repeatable manner. In order to meet the requirements of fast response, resilience, and gapless impact during bonding, this amplification mechanism is preferred. The structure incorporates pre-tightening and limiting designs: on the one hand, pre-tightening eliminates assembly gaps and loosening during reverse return, ensuring continuous force transmission during forward and reverse switching in the amplification process; on the other hand, mechanical limits are set to prevent the amplified structure from entering an irreversible overload deformation zone under abnormal deviations or over-excitation commands. Simultaneously, the amplification factor is preferably optimized to meet micron-level compensation requirements without excessive amplification, to avoid the amplification mechanism's natural frequency being too low and thus resonating due to ultrasonic or motion impact. Furthermore, the force transmission direction of the amplification mechanism is preferably coaxial or nearly coaxial with the wedge pressure axis to reduce the impact of lateral force components on weld morphology and positioning accuracy. After obtaining the amplified machine... After mechanical compensation displacement, it is preferable to directly transmit it to the base connection of the chopper as a micron-level compensation displacement to correct the pressure application position of the chopper. The base connection is preferably set as a force node that can effectively superimpose the compensation displacement onto the downward stroke of the chopper, so that the compensation displacement is mechanically manifested as a slight advance or slight retraction of the pressure contact position at the end of the chopper, thereby achieving fine pull-back of the pressure application value in closed-loop control. To ensure the stiffness and response speed of the transmission, it is preferable to use a structure with high stiffness and reliable connection on the compensation displacement transmission path, and to avoid displacement loss or phase lag during cyclic bonding through coaxiality control and anti-loosening structure.

[0106] Furthermore, the dynamic absorption of kinetic impact energy through magnetorheological dampers includes:

[0107] The kinetic impact energy generated by the chopping blade during movement and compensation is transmitted to the piston drive shaft of the magnetorheological damper through a rigid connection component;

[0108] The magnetorheological fluid in the drive magnetorheological damper cavity generates shear deformation flow under piston extrusion, and viscous resistance is formed in the opposite direction of impact based on the magnetic field-dependent viscosity change characteristics of the magnetorheological fluid.

[0109] The impact kinetic energy is continuously dissipated by viscous resistance and converted into molecular frictional heat energy until the motion returns to static equilibrium when the cleaver completes the bonding operation.

[0110] As a preferred embodiment, when the wedge, driven by a voice coil motor, completes rapid approach, downward contact at the target pad, and minor compensation displacement during the bonding stage, axial inertial impact and transient velocity change will occur in the wedge base and its connecting structure. Preferably, this impact energy is transferred to the piston drive shaft of the magnetorheological damper through a rigid connection assembly. The rigid connection assembly is preferably arranged along the wedge's pressure axis to reduce energy bypass loss caused by lateral force. A high-rigidity, low-clearance connection method is used to ensure that the impact kinetic energy can be quickly and lag-free coupled to the piston drive shaft. At the same time, to avoid the rigid connection introducing assembly stress or cumulative eccentricity, Preferably, a coaxial positioning structure is set at the connection end, along with an anti-loosening fastening method, to maintain stable force transmission even under cyclic bonding. When the piston drive shaft drives the magnetorheological fluid in the magnetorheological damper cavity to flow relative to each other under impact, the magnetorheological fluid forms a shear deformation flow under piston compression and throttling channel constraint. Preferably, a shear energy dissipation method that is more sensitive to high-frequency reciprocating motion with small strokes is selected, so that it can still output significant damping in scenarios where piston displacement is small but velocity changes rapidly, thereby covering typical impact modes such as contact transient rebound, compensation for sudden displacement changes, and blade lifting ringing. Furthermore, based on the magnetic field-dependent viscosity change characteristics of the magnetorheological fluid, the damping... The device, under an applied magnetic field or a preset magnetization state, causes the equivalent viscosity of the magnetorheological fluid to change with the working conditions, thereby forming a viscous resistance opposite to the impact direction during the piston shear flow. This viscous resistance preferably suppresses sudden changes in piston speed and clamps the rebound tendency, ensuring that the impact energy at the cutting edge is not repeatedly exchanged within the elasticity of the mechanism but is instead absorbed by the damping element. During the energy dissipation stage, the continuous action of the viscous resistance on the piston converts the impact kinetic energy into molecular frictional heat energy within the magnetorheological fluid, which diffuses within the cavity structure, thus achieving continuous dissipation of impact energy. Preferably, this dissipation is achieved by controlling the operating state of the damper. The dissipation process covers the entire critical time period from the contact of the wedge with the pad to the completion of bonding, until the wedge motion is detected to return to static equilibrium, that is, the velocity and acceleration of the wedge end tend to stabilize, the applied pressure value stabilizes near the target and no longer shows significant rebound. In order to ensure that this energy dissipation process can suppress impact without weakening the ultrasonic energy transmission, it is preferable that the magnetorheological damper mainly dissipates the low-frequency impact, rebound and structural ringing caused by the macroscopic movement and micro-compensation of the wedge, while minimizing the significant obstruction of the high-frequency ultrasonic vibration transmitted from the piezoelectric ceramic to the wedge through the amplitude transformer, so as to balance the absorption of impact energy and maintain the consistency of the effective ultrasonic vibration input required for bonding.

[0111] Furthermore, it also includes:

[0112] After the cutting tool moves to the target pad, the sub-pixel vision system performs positioning verification and outputs a position ready signal to the control system to initiate the ultrasonic bonding trigger command.

[0113] Driven by the ultrasonic bonding trigger command, the process of measuring the vibration output of the ultrasonic generator and the applied bonding pressure value is started synchronously.

[0114] When the magnetorheological damper detects the ultrasteady oscillation characteristics of the motion impact energy, it autonomously increases the magnetic field strength and the viscous drag value until the impact dissipates.

[0115] As a preferred embodiment of the above, after the voice coil motor drives the cutting tool to move to the nominal position of the target pad, the sub-pixel vision system preferably immediately performs a positioning verification: under the condition that the bonding light source remains unchanged or is switched to a verification light source that is more conducive to edge contrast, it re-images the pad edge, existing marker points, or pad corner points, and outputs the verification coordinates using the same sub-pixel fitting strategy as the previous positioning. At the same time, the verification coordinates are compared with the target pad coordinates in the control system with tolerance. To avoid verification failure due to cutting tool occlusion, it is preferable to insert the verification action into the brief hovering height before the cutting tool makes contact, or to complete the verification using visible features in the side view / coaxial view, so that the verification can still reflect the relative position of the cutting tool center and the pad center. The relationship is as follows: When the verification result meets the preset tolerance, the vision system outputs a position ready signal to the control system. The position ready signal preferably includes a status bit indicating that the verification has passed and an optional indication of a small deviation, so that the control system can make final fine adjustments if necessary before initiating the ultrasonic bonding trigger command. This prevents landing point deviations caused by visual errors, platform thermal drift, or fixture loosening from occurring in advance. Driven by the ultrasonic bonding trigger command, this embodiment preferably synchronizes the vibration output of the ultrasonic generator with the measurement process of the applied bonding pressure. Specifically, the synchronization method preferably involves turning on the generator's energy output at the same trigger moment to allow the piezoelectric ceramic to enter a stable vibration establishment process, while simultaneously activating or enhancing the sampling and effective pressure extraction states of the force sensor. This approach ensures that usable pressure data is obtained from the very first moment of ultrasonic action, avoiding initial pressure gaps caused by vibration before force measurement or trigger timing mismatches caused by force measurement before vibration. To guarantee that the data after synchronous startup can be used for deviation analysis and correction command generation, it is preferable to extract the steady-state window or envelope of the pressure signal during the extremely short setup phase after triggering, and mark this setup phase as a control sub-phase that allows for rapid convergence, enabling the system to complete the first effective pullback of pressure deviation in the early stages of bonding. Simultaneously, to further suppress the coupling impact caused by synchronous startup, this embodiment preferably places the magnetorheological damper at an initial damping level that matches the trigger transient, i.e., maintaining high damping for a short period before and after triggering to absorb the impact of the bonding process. The impact generated by the superposition of touch and ultrasound then falls back to a damping level that does not significantly weaken the transmission of ultrasonic vibrations after entering a stable bonding stage. Furthermore, when the magnetorheological damper detects the ultrasteady-state oscillation characteristics of the motion impact energy, it is preferable for the damper to autonomously enhance the magnetic field strength and increase the viscous resistance value until the impact dissipates. The ultrasteady-state oscillation characteristics preferably refer to the ringing phenomenon that occurs in the force link of the blade, base, and damper, in which the amplitude does not decay rapidly or even shows a trend of maintenance or amplification within multiple vibration cycles. It can be identified by the piston drive shaft speed or displacement change trend that can be obtained by the damper itself, the continuous energy accumulation of the vibration sensor in a specific frequency band, or the pressure deviation direction repeatedly flipping in a short period of time with a high flipping frequency.Once this characteristic is detected, the damper preferably does not wait for instructions from the upper controller, but directly increases the magnetic field strength to raise the equivalent viscosity of the magnetorheological fluid. This rapidly increases the viscous drag and creates a stronger clamping effect on the piston's reverse movement, converting the energy that was originally exchanged in the structural elasticity into viscous dissipation more quickly. Only when a significant decrease in oscillation amplitude is detected, the piston movement returns to a controllable small range, or the frequency of pressure deviation reversals drops below a threshold, does the magnetic field strength gradually decrease to avoid over-damping and affecting the energy efficiency of the stable bonding stage.

[0116] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of the application as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. An ultrasonic bonding process that synergistically couples ultrasonic energy and pressure, characterized in that, The process includes: Based on the lead wire material matching the preset bonding pressure target value and ultrasonic power range; The piezoelectric ceramic is driven to vibrate within the ultrasonic power range by an ultrasonic generator, and the vibration is amplified by an amplitude transformer and transmitted to the chopping tool. Based on the coordinates of the solder joints located by the subpixel-level vision system, the voice coil motor is controlled to drive the cutting tool to move to the target solder pad. The bonding pressure applied by the chopping blade is measured in real time, and a driving force correction command for the piezoelectric ceramic is generated based on the deviation between the bonding pressure applied value and the target bonding pressure value. The piezoelectric ceramic, in conjunction with the elastic amplification mechanism, executes the driving force correction command and generates a micron-level compensation displacement. During the movement and compensation process of the cutting blade, the impact energy of the motion is dynamically absorbed by the magnetorheological damper until the bonding is completed; Generating the driving force correction command for the piezoelectric ceramic includes: The bonding pressure applied value is continuously collected and the real-time deviation from the target bonding pressure value is calculated. At the same time, the direction change and duration characteristics of the real-time deviation are recorded. A reverse proportional adjustment component is generated based on the numerical proportion of the real-time deviation. The cumulative effect generated by combining the real-time deviation with the duration characteristic generates an integral compensation component. The real-time deviation is fused with the differential suppression component generated by the direction change rate; The inverse proportional adjustment component, the integral compensation component, and the differential suppression component are linearly superimposed and converted into the electric field strength command value of the corresponding piezoelectric ceramic as the driving force correction command. Executing the driving force correction command and generating micron-level compensated displacement includes: The driving force correction command is converted into a potential field adjustment signal for the piezoelectric ceramic, and the crystal lattice of the piezoelectric ceramic is deformed based on the inverse piezoelectric effect to generate an initial displacement. The initial displacement generated by the piezoelectric ceramic is amplified step by step through the mechanical lever amplification structure of the elastic amplification mechanism to obtain the amplified mechanical compensation displacement. The amplified mechanical compensation displacement is transmitted to the base connection of the chopping blade as the micron-level compensation displacement to correct the pressure application position of the chopping blade.

2. The ultrasonic bonding process of ultrasonic energy and pressure synergistic coupling according to claim 1, characterized in that, Based on the lead material matching, a preset target value for bonding pressure and ultrasonic power range are defined, including: The material category of the lead is determined using the elemental spectral characteristics of the lead. Based on the elastic-plasticity index and surface hardness parameter of the material type, the upper and lower limits of the target bonding pressure value are obtained from the pre-stored parameter relationship table. Based on the energy absorption characteristics and thermal conductivity parameters of the material type, and combined with the transmission efficiency of the amplitude transformer, a power reference value that guarantees minimum bonding strength within the ultrasonic power range is calculated.

3. The ultrasonic bonding process of ultrasonic energy and pressure synergistic coupling according to claim 2, characterized in that, The power reference value is calculated based on the transmission efficiency of the amplitude transformer, including: The actual energy transfer efficiency curve under normal working conditions is established by combining the output power monitoring data of the ultrasonic generator with the vibration sensor data of the amplitude transformer mounting base. The theoretical boundary value of the effective energy required for the lead wire to absorb is calculated based on the energy absorption characteristics and the thermal conductivity parameter, and then the actual environmental loss compensation of the amplitude rod mounting base is superimposed. After performing energy adaptation mapping on the theoretical boundary value acceleration point based on the actual efficiency curve, a power reference value allocation scheme covering the ultrasonic power range is generated.

4. The ultrasonic bonding process of ultrasonic energy and pressure synergistic coupling according to claim 1, characterized in that, Generate a reverse proportional adjustment component based on the numerical proportion of the real-time deviation, including: Calculate the ratio of the absolute value of the real-time deviation to the pressure system range defined by the target value of the bonding pressure; Based on the elastic modulus characteristics of the piezoelectric ceramic during operation, the ratio is converted into the required driving force correction range for the piezoelectric ceramic. The inverse proportional adjustment component is generated by reversing the direction of the real-time deviation and combining it with the driving force correction amplitude.

5. The ultrasonic bonding process of ultrasonic energy and pressure synergistic coupling according to claim 1, characterized in that, Integral compensation components are generated by combining the cumulative effect produced by the duration characteristics, including: For the continuous time period corresponding to the duration feature, the real-time deviation is continuously accumulated to obtain the cumulative integral deviation; Based on the displacement response capability of the piezoelectric ceramic and the elastic amplification mechanism, the cumulative integral deviation is converted into the intensity of the integral compensation component.

6. The ultrasonic bonding process of ultrasonic energy and pressure synergistic coupling according to claim 1, characterized in that, The differential suppression component generated by fusing the rate of change of direction includes: The instantaneous switching characteristics of the direction of the real-time deviation are measured, and the rate of change of direction is converted into the change of viscous resistance based on the viscous fluid elastoplastic response characteristics of the magnetorheological damper. The linear displacement reverse drive mechanism controlling the voice coil motor generates a reverse damping compensation displacement based on the change in viscous resistance, and maps the reverse damping compensation displacement to the amplitude characteristics of the differential suppression component.

7. The ultrasonic bonding process of ultrasonic energy and pressure synergistic coupling according to claim 1, characterized in that, Dynamically absorbing motion impact energy through magnetorheological dampers, including: The kinetic impact energy generated by the chopping blade during movement and compensation is transmitted to the piston drive shaft of the magnetorheological damper through a rigid connection assembly; The magnetorheological fluid in the cavity of the magnetorheological damper is driven to generate shear deformation flow under piston extrusion, and viscous resistance is formed in the opposite direction of impact based on the magnetic field-dependent viscosity change characteristics of the magnetorheological fluid. The impact kinetic energy is continuously dissipated by the viscous resistance and converted into molecular frictional heat energy until the motion returns to static equilibrium when the chopping blade completes the bonding operation.

8. The ultrasonic bonding process of ultrasonic energy and pressure synergistic coupling according to claim 1, characterized in that, Also includes: After the cutting tool moves to the target pad, the sub-pixel vision system performs positioning verification and outputs a position ready signal to the control system to initiate an ultrasonic bonding trigger command. Driven by the ultrasonic bonding trigger command, the measurement process of the vibration output of the ultrasonic generator and the applied bonding pressure value is started synchronously. When the magnetorheological damper detects the ultrasteady oscillation characteristics of the motion impact energy, it autonomously increases the magnetic field strength and the viscous drag value until the impact dissipates.

Images