Chip welding method and welding device thereof

Abstract

The invention provides a chip welding method and a welding device thereof. The device comprises a 3D printing device, a chip welding device and a welding device, wherein the 3D printing device is used for accurately printing welding flux to the connecting part of a chip and a substrate according to a preset path; the welding device is used for welding the chip and the substrate; the ultrasonic detection module is used for detecting the quality of the welding spots; the signal receiving and processing unit is connected with the ultrasonic detection module and is used for analyzing signals fed back by the ultrasonic detection module and judging the quality of the welding spots; the control system is used for adjusting welding parameters and printing parameters in real time according to an ultrasonic detection result, so that serious defects can be repaired in time; the welding spots are monitored in real time, the printing parameters and the welding parameters are adjusted in real time according to the welding spot defects, or after the parameters are adjusted, the printing spray head and the welding machine of the 3D printing equipment are controlled to conduct secondary printing repair on the welding spots with the defect positions, and the situation that due to problems of the welding parameters and the like, the whole batch has the welding problem, and the working efficiency is low is avoided.

Description

Technical Field

[0001] The present application relates to the technical field of chip welding, and in particular to a chip welding method and a welding device thereof. Background Art

[0002] Chip welding is a key link in modern electronic manufacturing. It connects the chip to carriers such as circuit boards to achieve electrical and mechanical connections. The welding processes in the existing technology are mainly divided into the following categories: 1. Surface Mount Technology (SMT): The process flow of surface mount technology is: printing solder paste onto the pad of the circuit board through a steel mesh. The thickness and position accuracy of the solder paste have a significant impact on the welding quality. The placement machine uses a vacuum nozzle to accurately place the chip on the pad coated with solder paste, and sends the circuit board into the reflow oven. After preheating, heating, reflow and cooling stages, the solder paste is melted and then solidified to complete the welding of the chip and the circuit board. It is widely used in various electronic products, such as mobile phones, computer motherboards, etc. It is suitable for chip packaging with small pin spacing and small size, and can achieve high-density assembly.

[0003] 2. Through-hole technology (THT): The process flow of through-hole technology is: first insert the pins of the chip into the corresponding through holes of the circuit board, and then fix them by wave soldering or manual soldering. During wave soldering, the circuit board passes through the molten solder wave on the conveyor belt, and the pins come into contact with the solder to complete the welding. Manual soldering uses an electric soldering iron to solder each pin one by one. It is often used for chips with high mechanical strength requirements and thicker pins, such as some power chips, sockets, etc., and is still used in traditional electronic equipment and some special electronic modules.

[0004] 3. Flip chip technology (FC): The process flow of flip chip technology is as follows: the active side of the chip faces downward and is directly connected to the circuit board through solder balls. First, metal bumps (such as copper pillars, tin-lead or lead-free solder balls, etc.) are made on the chip's pads, and then the chip is flipped and aligned with the corresponding pads on the circuit board. The solder balls are melted by heating to achieve connection. Sometimes, bottom filler is filled between the chip and the circuit board to enhance connection reliability. It is widely used in high-performance processors, FPGAs and other chips. It can provide a shorter electrical connection path, reduce signal transmission delay, improve chip performance, and achieve higher packaging density.

[0005] After the chip welding is completed, quality monitoring is carried out. The detection methods usually include appearance inspection, electrical inspection, X-ray inspection, etc. Among them, appearance inspection often uses the naked eye, magnifying glass or microscope to observe the appearance of the solder joint to check whether there are defects such as cold soldering, short circuit, tin deficiency, tin beads, etc.; electrical inspection uses multimeters, oscilloscopes and other equipment to detect the electrical connection performance of the chip and the circuit board after welding. By measuring parameters such as resistance, voltage, and current between pins, it is determined whether there are electrical faults such as open circuits and short circuits to ensure that the chip can work normally; X-ray inspection uses X-rays to penetrate circuit boards and chips, and forms images on imaging boards or detectors according to the different degrees of X-ray absorption by different materials. The internal structure of the solder joint can be observed to detect whether there are internal defects such as voids and cold solder joints.

[0006] According to the above, in the prior art, chip inspection is usually carried out after welding, and the welding parameters in the welding process are consistent. If the selection is inappropriate, it may cause problems in the welding of the entire batch of chips. This application intends to monitor the welding quality in real time during the welding process, to achieve timely response and repair of welding defects, to ensure welding quality, and to avoid welding problems in the entire batch due to problems with welding parameters, resulting in low work efficiency.

[0007] After a detailed search, the applicant did not find any relevant technical solutions and needed to provide a new technical solution to solve the above technical problems. Summary of the invention

[0008] The present application provides a chip welding method, comprising the following steps: S1: pre-process the chip and substrate; S2: Set the printing parameters of the 3D printing device and the welding parameters of the welding machine. The printing parameters include the moving speed of the print head, the amount of solder extruded, and the temperature of the print head; the welding parameters include the welding temperature, welding time, welding power, welding pressure, and welding frequency; S3: calibrate the ultrasonic detection module; S4: The print head of the 3D printing device prints solder, and the chip is welded by the welding machine; S5: During the welding process, the ultrasonic detection module detects the welding point in real time to determine whether the welding point has defects. If not, welding continues. If yes, the process goes to S6; S6: The control system controls the print head to stop printing solder, and resets the printing parameters of the 3D printing device and / or the welding parameters of the welding machine. Alternatively, after resetting the printing parameters of the 3D printing device and / or the welding parameters of the welding machine, the control system controls the print head and the welding machine to print and repair the solder joints at the defective positions. S7: During the printing and repairing process, the ultrasonic detection module detects the solder joints in real time to determine whether there are defects in the solder joints. If not, continue welding. If yes, enter S6.

[0009] As a preferred solution, the method further includes S8: using an ultrasonic detection module to comprehensively detect the quality of the solder joints of the soldered chips.

[0010] As a preferred solution, in S5, the ultrasonic detection module completes the detection and feedback of whether the weld has defects before the weld solidifies.

[0011] As a preferred solution, in S5, during the welding process, the ultrasonic detection module detects the welding point in real time, including the following steps: S51: In the system of the 3D printing device, an information module is added that can provide real-time feedback on the print head position, print speed, and estimated solder joint forming time; S52: Building a communication link between the 3D printing device and the ultrasonic transducer of the ultrasonic detection module to ensure that data between the two can be transmitted in real time; S53: Setting a unified time reference in the 3D printing device and the ultrasonic transducer; S54: Predict the forming time of each solder joint in advance based on the speed and path planning of the 3D printing equipment; S55: Before the solder joint is expected to be formed, the ultrasonic transducer is triggered in advance for preheating and preparation; S56: The ultrasonic transducer transmits ultrasonic waves to the weld before the weld solidifies, and sends a signal to the signal receiving and processing unit before the weld solidifies.

[0012] As a preferred solution, it also includes S57: dynamically adjusting the detection time point of the ultrasonic transducer in real time according to the fluctuation of the actual speed of the 3D printing device.

[0013] As a preferred solution, in S5, the solder joint defects include void defects, crack defects, inclusion defects, and unfused defects.

[0014] As a preferred solution, in S6, the print head pauses printing solder, and the printing parameters of the 3D printing device and / or the welding parameters of the welding are reset, including: when the solder joint defect is a void defect, adjusting at least one parameter of the solder extrusion amount, the print head movement speed, the welding temperature, the welding time, the welding power, the welding pressure, and the welding frequency; when the solder joint defect is a crack defect, adjusting at least one parameter of the print head temperature, the print head movement speed, the welding temperature, the welding time, the welding power, the welding pressure, and the welding frequency; when the solder joint defect is an inclusion defect, adjusting at least one parameter of the print head movement speed, the welding temperature, the welding time, the welding pressure, and the welding frequency; when the defect is an unfused defect, adjusting at least one parameter of the print head movement speed, the welding temperature, the welding time, the welding pressure, the welding power, and the welding frequency.

[0015] As a preferred solution, when the solder joint defect is a crack defect, the moving speed of the print head is adjusted according to the length and direction of the crack, and the temperature of the print head is adjusted according to the depth of the crack.

[0016] As a preferred solution, when the solder joint defect is a void defect, the calculation formula for the solder extrusion amount is: , M represents the adjusted solder extrusion amount; Indicates the solder extrusion amount set in S2; Indicates the amount of solder added, = Indicates the density of solder, It indicates the spreading and loss coefficient of solder when filling voids. The value range of S is 1.2-1.5. V indicates the volume of the void. , r is the radius of the cavity.

[0017] As a preferred solution, when the solder joint defect is a void defect, the calculation formula for the solder extrusion amount is: , M represents the adjusted solder extrusion amount; represents the solder extrusion amount set in S2; d represents the distance between the void position and the center of the solder joint, , the midline coordinates of the solder joint are ( , , ), the coordinates of the void position are (X, Y, Z); j represents the solder extrusion coefficient, d and j are linearly related, and j increases by 0.05 for every 1 mm increase in d.

[0018] As a preferred solution, when the solder joint defect is a crack defect, the calculation formula for the moving speed of the printing nozzle is: ;in, Indicates the moving speed of the print head after adjustment; Indicates the moving speed of the print head set in S2; represents the crack length; D represents the velocity change corresponding to the unit crack length. For every 1 mm crack length, the change of D is ; represents the angle correction coefficient, , It is the angle between the crack direction and the direction of print head movement.

[0019] As a preferred solution, when the solder joint defect is a crack defect, the calculation formula for the temperature of the print head is: ;in, Indicates the adjusted print head temperature; represents the temperature of the print head set in S2; h represents the crack depth; Indicates the temperature change corresponding to the unit crack depth. For every 1μm crack depth, The change is 1.5.

[0020] A chip welding device, comprising: 3D printing equipment: accurately prints solder to the connection between the chip and the substrate according to the preset path; Welding device: weld the chip and the substrate; Ultrasonic detection module: detect the quality of solder joints; Signal receiving and processing unit: connected to the ultrasonic detection module, the signal receiving and processing unit analyzes the signal fed back by the ultrasonic detection module to determine the quality of the solder joint; Control system: reset the printing parameters of the 3D printing device and / or the welding parameters of the welder according to the ultrasonic detection results, or after reset the printing parameters of the 3D printing device and / or the welding parameters of the welder, the control system controls the print head and the welder to print and repair the welds at defective locations.

[0021] As a preferred solution, the 3D printing device includes a 3D printing nozzle, a solder delivery system, and a motion control platform: Solder delivery system: connected to the 3D printing nozzle to provide solder to the 3D printing nozzle; 3D printing nozzle: used to accurately print the solder delivered by the solder delivery system to the connection part between the chip and the substrate; Motion control platform: connected to the 3D printing nozzle to control the movement of the 3D printing nozzle.

[0022] As a preferred solution, the ultrasonic detection module includes an ultrasonic generator and an ultrasonic transducer: Ultrasonic generator: generates high-frequency electrical signals; Ultrasonic transducer: connected to the ultrasonic generator, converts the high-frequency electrical signal of the ultrasonic generator into ultrasonic waves, transmits them to the welding point that needs to be inspected, receives the ultrasonic signals reflected and scattered back by the welding point, converts them into electrical signals, and transmits the electrical signals to the signal receiving and processing unit.

[0023] This application uses 3D printing equipment and ultrasonic detection equipment to weld photovoltaic panels, which has the following advantages: (1) Use ultrasound to monitor the internal quality of the solder joint in real time without contacting the solder joint and ensure the accuracy of defect detection; (2) Using 3D printing equipment to precisely control the print head movement speed, solder extrusion volume, and print head temperature, timely response and repair of welding defects can be achieved to ensure welding quality; (3) Adjust the printing parameters of the 3D printing equipment and the welding parameters of the welding machine in real time according to the defects, or control the printing nozzle and welding machine of the 3D printing equipment to perform secondary printing repair on the welds at the defective positions after adjusting the parameters; avoid welding problems in the entire batch due to problems with welding parameters, etc., resulting in low work efficiency. DETAILED DESCRIPTION

[0024] The specific implementation of the present invention is described in detail below. It should be noted that the specific implementation described here is only used to illustrate and explain the present invention, and is not used to limit the present invention. Embodiment 1

[0025] This embodiment provides a chip welding method, comprising the following steps: S1: pre-process the chip and substrate; The pretreatment of the chip mainly includes: carefully check whether the chip has any damaged or loose parts, and repair or replace it if necessary; gently place the inspected chip into the cleaning basket to avoid direct contact between the chip and the bottom of the cleaning tank during the cleaning process to reduce potential damage, select a suitable cleaning solution according to the type and degree of contamination of the chip, and control the concentration and pH value of the cleaning solution according to the specific situation; set the appropriate cleaning time and temperature, start ultrasonic cleaning, usually the cleaning solution temperature should be controlled between 40℃-60℃, under the action of ultrasonic waves, the solvent molecules in the cleaning solution can more effectively penetrate into the tiny gaps on the surface of the chip to remove dirt and grease, etc. For some difficult-to-remove stains, high-pressure spray cleaning can be used to assist cleaning, but care should be taken to avoid direct impact on the chip to avoid damage, and airflow can also be used to blow away insoluble pollutants such as dust and fibers on the surface of the chip; for the chips after the above cleaning, rinse the chips thoroughly with pure water to ensure that the cleaning agent residue is removed; place the chips in a dry environment or use drying equipment to dry them to ensure that the chip surface is dry.

[0026] The pretreatment of the substrate mainly includes: rust removal, oil removal, grinding and other treatments on the substrate surface to remove rust, oil, oxide scale and other impurities on the surface, so as to make the substrate surface clean and smooth to improve the bonding strength with the chip. Sandblasting, pickling and other methods can be used to clean the surface; check whether the size of the substrate meets the requirements, cut and correct the substrate that does not meet the requirements to ensure the dimensional accuracy of the substrate; use brushing, micro-etching and other methods to appropriately roughen the substrate surface, in order to increase the roughness of the substrate surface, so as to facilitate better bonding with the chip. Cleaning liquid and micro-etching liquid are usually used for treatment; Use solvents, ultrasonic cleaning or distilled water to clean the substrate to remove dust, oil and other impurities on the surface, ensure that the substrate surface is clean, and create good conditions for subsequent processing steps; dry the cleaned substrate by air drying or drying equipment to ensure that the substrate surface is dry; determine whether chemical treatment is required based on specific needs. If necessary, the substrate can be chemically treated, such as acid washing, alkali washing, etching and oxidation, to further improve the surface properties of the substrate, remove residual impurities, increase surface roughness or provide special chemical activity.

[0027] S2: Set the printing parameters of the 3D printing equipment and the welding parameters of the welding machine. The welding parameters of the welding machine include welding temperature, welding time, welding power, welding pressure, and welding frequency. The setting of welding temperature mainly depends on the melting point of the solder. Usually, the welding temperature needs to be set within the range of 40℃-70℃ above the melting point of the solder to ensure that the solder can be fully melted and achieve a good welding effect. At the same time, considering the thermal sensitivity of the chip and the solder joint, the temperature should not be too high to avoid damage to the chip; the setting of welding time needs to ensure that the solder has enough time to melt, flow and fully combine with the weldment, but too long time will cause heat accumulation, which may damage the chip or the structure around the solder joint. It needs to be determined in combination with the solder joint size, welding temperature and welding process requirements. For example, for small chip solder joints, the welding time can be initially set at 0.3 - 0.5 seconds; The setting of welding power is closely related to welding temperature and welding time. It determines the energy transferred to the solder joint per unit time. The power size needs to be determined according to the characteristics of the welding equipment, the heat capacity of the solder joint and the required welding speed; For example, for ordinary small welding equipment used for chip solder joint welding, the power can be initially set at 50W-100W. If it is later found that the welding temperature rises too slowly or the welding effect is not good, the power can be adjusted according to the actual situation; The welding pressure is mainly used to ensure close contact between the weldment and the solder and promote metallurgical bonding during the welding process. The pressure size needs to consider the size and shape of the solder joint and the bearing capacity of the chip. Excessive pressure may damage the chip or solder joint. For example, in chip solder joint welding, the welding pressure can be initially set at 0.2 - 0.5N; welding frequency, for thicker chip solder joint materials, or strict requirements on heat input, when it is necessary to avoid overheating and damaging the surrounding structure of the chip, low-frequency welding can be used. For example, for solder joints with a thickness of more than 1mm, in order to prevent excessive heat accumulation and chip damage, a lower frequency can be selected, generally between 10Hz-50Hz. At low frequency, the laser pulse interval is longer, and the heat input is relatively dispersed, which is conducive to controlling the temperature of the welding area. For example, when welding larger and heat-sensitive chip solder joints, a frequency of 15Hz can ensure the quality of solder joint welding and reduce the thermal impact on the chip; for thinner solder joint materials, or when welding needs to be completed quickly to improve production efficiency, high-frequency welding is more suitable. For example, for solder joints with a thickness of less than 0.3mm, high-frequency welding can quickly achieve connection and ensure the continuity of the weld. It is usually between 100Hz-500Hz. At high frequency, the laser pulses are dense, which can quickly melt the solder and achieve efficient welding. For example, in the large-scale production of chips, for small and thin solder joints, a frequency of 300Hz is used, which can greatly increase the welding speed and meet production needs.

[0028] Printing parameters include: print head moving speed, solder extrusion volume, and print head temperature. When setting the printing parameters of 3D printing equipment for chip welding, factors such as chip type, substrate material, and solder properties need to be considered comprehensively, as follows: S21: Determine the basic parameter range: Different solders have different melting points, viscosities and other characteristics, which are the basis for setting parameters. For example, the commonly used tin-silver-copper (SAC) alloy solder has a melting point of about 217°C - 220°C. Based on this, the lower limit of the print head temperature is determined to be slightly higher than the melting point. Different 3D printing equipment has different performances. Please refer to the parameter range provided in the equipment manual for details.

[0029] S22: Setting the print head temperature: Based on the melting point of the solder, the print head temperature needs to make the solder fully melted and have good fluidity. For example, for tin-silver-copper (SAC) alloy solder, the initial setting temperature of the print head temperature can be set to 250°C-280°C, which is 30°C-60°C higher than the highest melting point to ensure that the solder is melted. Preferably, a small amount of solder can be printed for testing to observe the melting state. For example, if the solder is difficult to extrude and wire drawing occurs, it means that the temperature is low and the temperature needs to be gradually increased (5°C-10°C each time). If the solder flows too fast and the molding is poor, it means that the temperature is high and the temperature should be appropriately lowered. The temperature needs to be gradually lowered (5°C-10°C each time). For example, when the solder is first tested at 250°C, it is not extruded smoothly. After rising to 260°C, the state is good. It is determined that 260°C is the initial setting temperature of the print head temperature.

[0030] S23: Set the solder extrusion amount: Estimate the solder amount (solder extrusion amount) required for each solder joint according to the size and shape of the solder joint. For example, when soldering chip pins, the diameter of the small pin solder joint is 0.5mm. Assuming that the solder joint is hemispherical, according to the sphere volume formula: , the calculated solder joint volume is about 0.033mm 3 Considering that the actual welding requires slightly more solder, the initial setting of each solder extrusion amount is 0.05mm 3 ; Preferably, you can print a series of solder joints to observe the solder coverage and filling. If the solder joint is too small or not filled, increase the extrusion amount; if there is too much solder or it overflows, reduce the extrusion amount; For example, after the first print, if it is found that the solder joint does not completely cover the connection between the pin and the substrate, increase the extrusion amount appropriately and test again until the solder joint is full and well formed.

[0031] S24: Set the print head moving speed: Combine the solder extrusion volume and the print head temperature to set the initial moving speed of the print head. Take tin-silver-copper (SAC) alloy solder and tiny chips as an example. If the extrusion volume is small (less than 0.01mm 3 ), the temperature is appropriate, the print head can move slightly faster; if the extrusion volume is large (greater than 0.05mm 3), the temperature is appropriate, the print head movement speed should be slowed down to ensure uniform distribution of the solder; the standard for appropriate temperature is set according to the specific situation, such as the commonly used tin-silver-copper (SAC) alloy solder, the appropriate temperature is 250℃-280℃, but the melting points of tin-silver-copper (SAC) alloy solders are different due to different component ratios, and need to be adjusted accordingly. For example, the melting point of tin-silver-copper (SAC) alloy solder with a high silver content is slightly higher, and the appropriate temperature is 260℃-290℃. Other types of solder, such as tin-lead solder, have low melting points and the appropriate temperature range is around 183℃-230℃℃. The technicians accurately set it according to the characteristics of the solder.

[0032] When the temperature meets the solder characteristics, the calculation formula for the print head moving speed is: , in, Indicates the print head moving speed. For the initial solder extrusion amount set above, is the cross-sectional area of ​​the solder joint, It is an empirical coefficient, with a value between 0.5 and 2. The empirical coefficient is affected by the viscosity of the solder and the temperature of the print head (the viscosity of the solder affects the fluidity, high viscosity solder requires a slow speed, and the Y value decreases accordingly; the temperature of the print head affects the state of the solder, the fluidity is good when the temperature is high, and the Y value can be appropriately increased); for example, when welding tiny solder joints, Value 1, solder point A=0.2mm 2 , =0.05mm 3 , calculated 0.25mm / s, during the printing test, solder accumulation was found, indicating that the speed was slow. Increase the Y value to 1.5 and recalculate to get 0.375mm / s, observe the forming effect again; the above The calculation method is to emit ultrasonic waves to the standard solder joints, calculate the geometric parameters of the solder joints according to the reflection echo time and signal strength of the ultrasonic waves at different interfaces of the solder joints, and deduce the thickness and other parameters of the solder joints by measuring the propagation speed of the ultrasonic waves in the solder joints and the time delay of the reflection echoes, combined with the shape model of the solder joints. The cross-sectional area A can be indirectly calculated by combining other measurement methods (such as industrial cameras measuring the planar dimensions of the solder joints). For example, if the horizontal dimensions of the solder joints are known, the thickness of the solder joints can be obtained by ultrasonic measurement, and the two can be multiplied to obtain the cross-sectional area. This method does not require direct contact with the solder joints.

[0033] Preferably, after the above parameters are initially set, an overall chip welding test is performed to observe the quality of multiple solder joints and the firmness of the connection between the chip and the substrate. The parameters can be fine-tuned according to the overall test results. If some solder joints are still found to be cold, the extrusion amount can be appropriately increased, the welding time can be increased, or the nozzle movement speed can be reduced. If there are signs of overheating after chip welding, the print nozzle temperature and welding temperature can be fine-tuned to lower them. After multiple tests and fine-tuning, a set of optimal parameters suitable for specific chip welding can be determined.

[0034] S3: Calibrate the ultrasonic detection module; it can ensure the accuracy, stability and reliability of the ultrasonic detection module and the welding process in the subsequent chip welding; the ultrasonic detection module in this embodiment includes an ultrasonic generator, an ultrasonic transducer, and a signal receiving and processing unit.

[0035] Calibrate the ultrasonic generator, including frequency calibration, amplitude calibration, power calibration, and pulse parameter calibration. Specifically: Frequency calibration: Use a frequency meter connected to the output of the ultrasonic generator to measure its actual output frequency, and compare the measured value with the frequency set by the ultrasonic generator. If there is a deviation, adjust it through the frequency adjustment knob of the ultrasonic generator or software settings to make its output frequency reach the set value; Amplitude calibration: Connect the output of the ultrasonic generator to the input of the oscilloscope through a suitable probe, and ensure that the attenuation coefficient of the probe is set correctly to accurately measure the signal amplitude; Set a known frequency and the initial value of the expected output amplitude on the ultrasonic generator, such as setting the frequency to 20kHz and the amplitude to 10Vpp (peak-to-peak value), observe the waveform displayed on the oscilloscope, and read the actual measured signal amplitude, such as If the actual amplitude does not match the set value, it means that adjustment is needed. For example, if the measured value is 8Vpp, it means that adjustment is needed. Find the amplitude adjustment knob on the ultrasonic generator or use the amplitude setting option in the software interface to gradually adjust the amplitude and observe the change in the amplitude of the oscilloscope waveform until the amplitude displayed by the oscilloscope reaches the expected 10Vpp; Power calibration: Connect the power meter to the output end of the ultrasonic generator and measure its output power. According to the specifications of the equipment, adjust the power adjustment knob or related parameters of the ultrasonic generator to make the output power reach the specified range or set value; Pulse parameter calibration: For ultrasonic generators with pulse output function, it is necessary to calibrate parameters such as pulse width and pulse repetition frequency; Use an oscilloscope to observe the pulse signal output by the generator and compare it with the set parameters. If there is any deviation, adjust it through the corresponding settings.

[0036] Calibrate the ultrasonic transducer, including sensitivity calibration, frequency response calibration, and directivity calibration. Specifically: Sensitivity calibration: Place the ultrasonic transducer in a standard environment with a known sound pressure field, such as a calibration water tank or a standard test block. Use a standard hydrophone or other receiving device with known sensitivity to receive the ultrasonic signal emitted by the transducer and measure its sound pressure. Calculate the sensitivity of the transducer based on the measurement results and the known sound pressure field strength. If the sensitivity deviates from the nominal value, it can be calibrated by adjusting the installation position and angle of the transducer or making necessary parameter compensation. Frequency response calibration: Use a signal generator to generate electrical signals of different frequencies and input them into the ultrasonic transducer to make it emit ultrasonic waves. At the same time, use a receiving device to measure the output response of the ultrasonic transducer at different frequencies, such as the sound pressure amplitude or the strength of the received signal, draw a frequency response curve, and observe the response of the transducer at different frequencies. If the frequency response is abnormal or does not meet the specifications, it can be calibrated by adjusting the matching circuit, resonant frequency and other parameters of the transducer. Directivity calibration: Fix the ultrasonic transducer on a rotating bracket so that it can emit ultrasonic waves at different angles. Place a receiving device at a certain distance, measure the transmitted sound pressure or received signal strength of the ultrasonic transducer at different angles, draw a directivity pattern, and check whether the directivity of the ultrasonic transducer meets the design requirements; if there is a deviation in the directivity, it can be calibrated by adjusting the structure of the transducer, reflector or damping material, etc.

[0037] The calibration of the signal receiving and processing unit includes gain calibration, filter calibration, and time delay calibration. Specifically: Gain calibration: Input a standard signal of known amplitude into the signal receiving and processing unit, adjust the unit's gain knob or parameter to make the amplitude of the output signal reach the expected value, calculate the gain value by comparing the amplitude of the input and output signals, and compare it with the set gain. If there is a deviation, adjust it until the gain is accurate. Filter calibration: Use signals of different frequencies to input into the unit, and check whether the filter's cutoff frequency, bandwidth and other parameters are accurate by observing the spectrum or filtering effect of the output signal. If the filtering effect is not ideal or the parameters do not match the design value, you can adjust the filter's circuit parameters or software settings for calibration. Time delay calibration: For units involving time measurement or signal processing, it is necessary to calibrate the time delay parameters. Use a standard time signal or a signal with a known delay to input into the unit and measure the time delay of the output signal. Compare it with the set delay value. If there is a deviation, calibrate it by adjusting the relevant time parameters or circuit delay elements.

[0038] S4: The print head of the 3D printing device prints solder, and the chip is welded by a welding machine; the welding machine uses a welding machine that does not need to transport solder in the prior art such as laser welding, and this embodiment takes laser welding as an example; S5: During the welding process, the ultrasonic detection module detects the welding point in real time to determine whether there is a defect in the welding point. If there is no defect, welding continues. If there is a defect, enter S6; The following methods are mainly used to judge whether the solder joint has defects: 1. Based on signal amplitude analysis: Normal amplitude range setting: Before formal welding, ultrasonic detection is performed on solder joints of the same type and specification that are determined to be defect-free, and a series of ultrasonic reflection signal amplitude data are obtained. By statistically analyzing these data, the fluctuation range of the normal solder joint signal amplitude is determined; for example, after multiple detections of a certain type of solder joint, it is found that the normal signal amplitude fluctuates between 80-120mV; the above methods for determining defect-free solder joints include but are not limited to: electrical performance testing, X-ray detection, metallographic analysis, etc. The above are conventional solder joint detection methods, and this application will not be described in detail here; for example, X-ray detection is to use X-ray detection equipment to perform perspective imaging of solder joints. X-rays can penetrate solder joints and circuit boards to show solder joints. By observing the X-ray image, we can find the internal voids, inclusions and the connection between the pin and the pad. Real-time detection comparison: Real-time detection of solder joints during the welding process. If the amplitude of the ultrasonic reflection signal is significantly lower than the lower limit of the normal range, it may indicate that there are defects such as voids in the solder joints, because these defects will aggravate the energy attenuation during the ultrasonic propagation process, resulting in a weaker reflection signal. On the contrary, if the amplitude is higher than the upper limit of the normal range, it may mean that the internal structure density of the solder joint is abnormal, such as the presence of defects such as inclusions, which affect the reflection characteristics of the ultrasonic wave. For example, if the signal amplitude of a certain solder joint is detected to be 50mV, which is far lower than the normal range of 80-120mV, it can be preliminarily determined that the solder joint may have a void defect.

[0039] 2. Based on the signal reflection time analysis: Calculate the normal propagation time: According to the material properties of the solder joint (the propagation speed of ultrasonic waves in the material is known) and the geometric dimensions of the solder joint, calculate the theoretical time for ultrasonic waves to propagate to different positions in the normal solder joint and reflect back; for example, if the solder joint thickness d is known to be 5mm and the ultrasonic wave propagation speed v in the welding material is 5000m / s, then the theoretical time from the emission of ultrasonic waves to the reflection from the bottom of the solder joint is , d=5mm=5 10 -3 m, v=5000m / s, calculated ; Compare the actual reflection time: During real-time detection, if there is a large deviation between the actual arrival time of the ultrasonic reflection signal and the theoretical time, it means that the internal structure of the solder joint may have changed and there are defects. If the reflection time is advanced, there may be defects such as voids and unfused solder joints inside the solder joint, which shortens the ultrasonic propagation path; if the reflection time is delayed, there may be inclusions in the solder joint, which lengthens the propagation path; if the actual detected reflection time is , which is significantly smaller than the theoretical time, indicating that there may be void defects in the solder joints; the actual arrival time of the ultrasonic reflection signal mentioned above has a large deviation from the theoretical time. The large deviation standard is not a fixed value. It is affected by many factors and needs to be determined comprehensively based on the specific welding scenario, material properties and detection accuracy requirements. For example, small solder joints (electronic chip pin solder joints, with a size below the millimeter level) are sensitive to time deviations, and the deviation standard is generally set at 5%-10%. Due to their small size and short ultrasonic propagation path, tiny defects may significantly change the propagation time. Assuming that the theoretical time for a small solder joint is 2μs, the actual time exceeds 2.1-2.2μs, which may indicate the presence of defects.

[0040] 3. Based on the analysis of signal waveform morphology: Determination of normal waveform characteristics: Through the detection of normal solder joints, the typical waveform characteristics of its ultrasonic reflection signal are summarized, including the shape of the waveform, the number and position of peaks and troughs, the smoothness of the waveform, etc.: For example, the waveform of a normal solder joint usually has a regular shape, the peaks and troughs are evenly distributed, and the waveform is relatively smooth; Abnormal waveform recognition: During real-time detection, if the waveform is distorted, such as the peak becomes wider, the trough becomes shallower, and additional small peaks or troughs appear, it may indicate that the solder joint has defects. Irregular waveform changes are often related to ultrasonic scattering and uneven reflection caused by defects inside the solder joint; for example, the normal waveform is single-peaked and smooth, while the detected waveform has double peaks and irregular peak shape, which may mean that the solder joint has defects such as cracks, resulting in complex reflection and scattering of ultrasonic waves at the cracks.

[0041] 4. Based on multi-parameter comprehensive analysis: Establish a defect judgment model: Consider multiple parameters such as signal amplitude, reflection time, waveform shape, etc., and use machine learning, statistical analysis and other methods to establish a defect judgment model; collect a large amount of data on known defect types and defect-free solder joints as training samples, and train the model so that it can accurately identify the multi-parameter feature combinations corresponding to different defect types; real-time judgment application: detect solder joints in real time during the welding process, and input the acquired multi-parameter data into the established model. The model determines whether there are defects in the solder joints and the type of defects based on the rules and experience obtained from the training; for example, through model analysis, when the signal amplitude is lower than the lower limit of the normal range, the reflection time is advanced, and the waveform is distorted, it is determined that the solder joint has a mixed defect of voids and cracks. This multi-parameter comprehensive analysis method can improve the accuracy and reliability of defect judgment, especially for complex welding structures and the coexistence of multiple defects. It is more effective.

[0042] More specifically: defects include void defects, crack defects, inclusion defects, and unfused defects. Based on the above detection methods, various types of defects are explained through specific examples: Void defect: In a chip production workshop, an ultrasonic detection module with a frequency of 5MHz was used to detect the solder joints of a chip. The size of the solder joint is 0.8mm long, 0.6mm wide and 0.4mm high. Under normal circumstances, the propagation speed of ultrasonic waves in the solder joint material is about 3200m / s. (1) Change in the amplitude of the reflected signal of the void defect: After multiple measurements of normal solder joints, the average amplitude of the reflected signal is 100mV. When a solder joint with a void defect is detected, the amplitude of the reflected signal instantly rises to 180mV. This is because the acoustic impedance difference between the void and the solder is large, and the ultrasonic wave is strongly reflected at the void interface, resulting in a significant increase in the amplitude of the reflected signal. (2) Shortened propagation time: The round-trip propagation distance of ultrasonic waves in a normal solder joint is about 2x0.4mm=0.8x10 -3 m, the normal propagation time is calculated to be 0.8x10 -3 ÷3200=2.5x10 -7 s; when a void solder joint is detected, the actual measured propagation time is shortened to 2x10 -7 s, which indicates that the ultrasonic propagation path is shortened due to the presence of the cavity, further confirming the existence of the cavity; (3) Waveform distortion: The ultrasonic reflection waveform of a normal solder joint presents a regular single peak with a smooth peak shape, while the waveform of a solder joint with a cavity has an obvious small peak after the main reflection peak. This is because the ultrasonic wave is reflected multiple times in the cavity, forming an additional reflection signal superimposed on the original waveform; the detection changes corresponding to the above-mentioned cavity defects can be used in combination, as long as it can be clearly identified as a cavity defect, and this application does not make any specific limitations.

[0043] Crack defect: Also in the chip production workshop, an ultrasonic detection module with a frequency of 5MHz was used to detect the solder joints of another chip; the size of the solder joint is 0.8mm long, 0.6mm wide and 0.4mm high. Under normal circumstances, the propagation speed of ultrasonic waves in the solder joint material is about 3200m / s; (1) Change in the amplitude of the reflected signal of the crack defect: Through multiple measurements of normal solder joints, the average amplitude of the reflected signal of normal solder joints is 105mV. When a solder joint with a crack is detected, the amplitude of the reflected signal rises to 125mV. Although the crack is narrow in width, it can still cause ultrasonic waves to reflect. However, due to its geometric shape and distribution characteristics, the increase in the amplitude of the reflected signal is not as obvious as that of the void; (2) The uncertainty of the propagation time of the crack defect: Assuming that the crack is a crack that runs vertically through the thickness of the solder joint, the original ultrasonic propagation distance is 2x0.4mm=0.8x10 -3 m, after the crack appears, the propagation distance is shortened to 2x0.3mm=0.6x10 -3 m (assuming the crack depth is 0.1 mm); based on the propagation speed of 3200 m / s, the normal propagation time is 0.8x10 -3 ÷3200=2.5x10-7 s, when there is a crack, the propagation time is 0.6x10 -3 ÷3200=1.875x10 -7 s, the propagation time is significantly shortened; however, if the crack is inside the weld in the horizontal direction, the propagation time may remain basically unchanged because the ultrasonic propagation path has not changed significantly. (3) Waveform characteristics of crack defects: The waveform of a normal weld is a single, smooth peak. The waveform of a weld with a crack has multiple small reflection peaks near the main reflection peak. This is because the ultrasonic wave is reflected multiple times on the crack surface, and the wave signals of different reflection paths are superimposed, making the waveform complex; the detection changes corresponding to the above crack defects can be used in combination, as long as it can clearly identify the crack defect, and this application does not make specific restrictions.

[0044] Inclusion defects: In the chip production workshop, the solder joint material is tin-silver-copper alloy. The chip solder joints are tested using an ultrasonic detection module with a frequency of 4MHz. The solder joint dimensions are 0.8mm long, 0.6mm wide, and 0.4mm high. Under normal circumstances, the propagation speed of ultrasonic waves in the solder joint material is about 3200m / s. (1) Changes in the amplitude and propagation time of inclusion defects: Through multiple measurements of normal solder joints, the average amplitude of the reflected signal of normal solder joints is 95mV, and the propagation time is 2.6x10 -7 s. When a solder joint containing metal oxide inclusions is detected, the amplitude of the reflected signal is reduced to 70mV. This is because the acoustic impedance of the metal oxide inclusions is different from that of the solder, which weakens the ultrasonic reflection. At the same time, due to the presence of inclusions, the ultrasonic propagation path is changed and the propagation time is extended to 3x10 -7 s. (2) Waveform change of inclusion defects: The waveform of a normal solder joint presents a regular sine wave shape, while the waveform of a solder joint containing inclusions has an obvious trough before the main reflection peak, and a smaller peak after the main reflection peak. This is because the ultrasonic wave is reflected and refracted multiple times at the interface between the inclusion and the solder, changing the original shape of the waveform; the detection changes corresponding to the above inclusion defects can be used in combination, as long as it can be clearly identified as an inclusion defect, and this application does not make specific limitations.

[0045] Unfused defects: The chip solder joints were inspected using an ultrasonic detection module with a frequency of 6 MHz. The solder joint dimensions were 0.7 mm long, 0.5 mm wide, and 0.35 mm high. Under normal circumstances, the propagation speed of ultrasonic waves in the solder joint material was about 3200 m / s. (1) Changes in the amplitude and propagation time of unfused defects: After measuring a large number of normal solder joints, the average amplitude of the reflected signal was 110 mV and the propagation time was 2.2 x 10 -7s. When a solder joint with an unfused defect is detected, the amplitude of the reflected signal rises sharply to 150mV. This is because the acoustic properties of the materials on both sides of the unfused interface are quite different, resulting in strong reflection of the ultrasonic wave. At the same time, the unfused area hinders the normal propagation of the ultrasonic wave, and the propagation time is shortened to 1.8x10 -7 s, indicating that the ultrasonic wave failed to propagate to the expected depth along the normal path; (2) Waveform characteristics of unfused defects: The waveform of a normal solder joint is continuous and smooth, while the waveform of a solder joint with an unfused defect has an obvious interruption at the expected reflected wave position, followed by an irregular reflected wave peak. This is because the ultrasonic wave is reflected at the unfused interface and cannot continue to propagate along the normal path. The subsequent reflected wave signal is superimposed on the original signal, resulting in a discontinuous and irregular waveform. The detection changes corresponding to the above-mentioned unfused defects can be used in combination, as long as it can be clearly identified as an unfused defect. This application does not make any specific limitations.

[0046] S6: The print head stops printing solder, and the printing parameters of the 3D printing device and the welding parameters of the welding machine are reset. Alternatively, after the printing parameters of the 3D printing device and the welding parameters of the welding machine are reset, the control system controls the print head and the welding machine to print and repair the solder joints at the defective positions; The specific criteria for resetting the printing parameters and / or welding parameters of the 3D printing equipment are: 1. The signal amplitude deviation is small and regular: When the deviation of the ultrasonic reflection signal amplitude is between ±10% and ±20% compared with the normal solder joint amplitude, consider resetting the parameters; for example, the average amplitude of the normal solder joint signal is 100mV. If the solder joint amplitude fluctuates between 80-120mV, it may be that the slight change in the welding parameters has caused the internal structure of the solder joint to change, but has not yet seriously affected its quality; this situation may be caused by slight fluctuations in the solder extrusion amount, the print head temperature, or the print head movement speed: For example, a slightly smaller amount of solder extrusion may result in a slightly lower density inside the solder joint, which reduces the amplitude of the ultrasonic reflection signal. At this time, these parameters can be appropriately fine-tuned, such as increasing the solder extrusion amount by 5%-10%, and / or increasing the print head temperature by 10-20℃, and observing whether the ultrasonic detection signal of the subsequent solder joint returns to normal.

[0047] 2. The reflection time deviation is small and consistent: If the deviation between the actual arrival time of the ultrasonic reflection signal and the theoretical time is within ±10%, preferably, when welding multiple solder joints, the deviation direction and degree of the multiple solder joints are relatively consistent, which can be solved by resetting the parameters; for example, the theoretical reflection time is 10μs, and the actual time is between 9-11μs, indicating that although the internal structure of the solder joint has changed, it is still within an acceptable range as a whole; this may be due to slight changes in material properties during the welding process, such as the effect of welding temperature on the solidification speed of the solder, which causes a slight change in the ultrasonic propagation speed. You can try to adjust the welding environment parameters, such as controlling the welding temperature of the welding area, or fine-tuning the movement speed of the print head to compensate for changes in the propagation time.

[0048] 3. The waveform morphology is slightly abnormal but similar: When the ultrasonic detection waveform of the weld shows a slight deformation, such as a slight change in the shape of the crest or trough, but the overall waveform is still similar to the normal waveform, preferably, when welding multiple welds, when multiple welds show the same slight abnormality, consider resetting the parameters; for example, the crest of the normal waveform is sharper, while the detected crest is slightly wider, but no additional crests or troughs appear; this slight waveform abnormality may be caused by a slight uneven distribution of welding energy. The welding power or welding time can be appropriately adjusted to make the welding energy more evenly distributed and improve the internal structure of the weld, so that the waveform returns to normal.

[0049] After resetting the printing parameters of the 3D printing equipment and the welding parameters of the welding machine, the specific standards for the control system to control the print head and the welding machine to print and repair the welds at the defective positions are: 1. Excessive signal amplitude deviation: If the amplitude of the ultrasonic reflection signal is lower than the amplitude of the normal solder joint by more than 30% or higher than 50%, it indicates that the solder joint has serious defects and needs to be repaired by printing. For example, if the normal amplitude is 100mV and the detected amplitude is lower than 70mV or higher than 150mV, it means that there may be a large number of cavities, looseness, and serious problems such as removal of inclusions inside the solder joint, which affect the propagation and reflection of the ultrasonic wave. Such defects will seriously affect the electrical and mechanical properties of the solder joint, and printing repair must be performed to fill the cavities, remove inclusions, etc., to restore the normal structure and performance of the solder joint. In this embodiment, the processing method for removing inclusions can be adopted: using a high-energy-density laser beam to melt the filling material provided by the 3D printing device and clad it on the inclusion part of the solder joint. During the laser cladding process, the high temperature can fully melt the impurities included and react physically and chemically with the filling material. Some impurities may be evaporated or floated to the surface and removed. At the same time, laser cladding can accurately control the repair area to ensure the dimensional accuracy and performance of the repaired solder joint.

[0050] 2. Excessive deviation in reflection time: When the deviation between the actual arrival time of the ultrasonic reflection signal and the theoretical time exceeds ±20%, it means that the internal structure of the solder joint has undergone significant changes, and there may be serious defects that require printing repair. For example, if the theoretical reflection time is 10μs, and the actual time is less than 8μs or greater than 12μs, there may be large voids, cracks, or unfused areas inside the solder joint, resulting in abnormal ultrasonic propagation paths. These defects will seriously weaken the connection strength and stability of the solder joint, and the defective position must be printed and repaired, and the solder joint structure must be rebuilt to ensure that its performance meets the requirements.

[0051] 3. Severe waveform distortion: If the ultrasonic detection waveform is severely distorted, such as multiple additional peaks and troughs, or the waveform becomes chaotic and very different from the normal waveform, it indicates that the solder joint has complex and serious defects and should be repaired by printing. For example, the normal waveform is a regular single peak, while the detected waveform has more than three peaks, and the spacing and amplitude between the peaks are irregular; this severe waveform distortion reflects the complex defects inside the solder joint, such as multiple cracks intertwined, large areas of unfused, etc., which need to be repaired by printing to accurately control the filling and distribution of solder to repair the defects and restore the solder joint to normal.

[0052] The following are the specific repair methods for each defect: Repair of void defects: The location and size information of the void is obtained through ultrasonic testing. If it is detected that the void causes the propagation time to be shortened, the size of the void can be calculated. The 3D printing equipment control system calculates the volume of solder required to be filled based on this. After the print head is restarted, the corresponding volume of solder is accurately extruded to fill the void. For example, if the void volume is calculated to be 0.03m³, the print head extrude an equal amount of solder with extremely high precision to ensure that the void is completely filled.

[0053] Repair of crack defects: According to the direction and length of the crack, the print head plans the path and extrude solder along the crack to completely cover the crack. For cracks with narrow width, the print head extrude solder in fine lines to ensure that the crack is filled: for example, for cracks with a width of 0.02mm, the print head extrude solder lines with a width of 0.03mm to cover them. At the same time, the amount of solder extrusion can be appropriately increased to improve the strength of the repaired area.

[0054] Repair of inclusion defects: Use a high-energy-density laser beam to melt the filler material provided by the 3D printing equipment and clad it on the inclusion part of the solder joint. During the laser cladding process, the high temperature can fully melt the impurities included and react physically and chemically with the filler material. Some impurities may be evaporated or float to the surface and removed. At the same time, laser cladding can accurately control the repair area to ensure the dimensional accuracy and performance of the repaired solder joint. Inclusions can also be sucked out through a tiny suction nozzle device to form a cavity, and the print nozzle can accurately extrude solder for filling. For example, if the volume of the cavity formed after sucking out the inclusions is 0.05m³, the print nozzle accurately extrude the corresponding volume of solder to ensure the accuracy of the filling.

[0055] Repair of unfused defects: By detecting the unfused position and size information, the size of the unfused part can be inferred. The control system calculates the volume of solder required to be filled based on this. After the print head is restarted, the corresponding volume of solder is accurately squeezed out to fill the unfused area, and the welder performs secondary welding on the unfused area.

[0056] S7: During the printing and repairing process, the ultrasonic detection module detects the solder joints in real time to determine whether there are defects in the solder joints. Otherwise, welding continues. If so, the process goes to S6. The solder joint detection refers to S5 and will not be described in detail here.

[0057] Preferably, in order to further ensure the quality of chip welding, this embodiment also includes S8: using an ultrasonic transducer to comprehensively detect the quality of the solder joints of the soldered chips, thereby saving the detection equipment specifically used to detect the chips. In this step, because the solder joints have solidified, when problems are detected in the solder joints, they are treated as problematic chips for subsequent processing. Embodiment 2

[0058] This embodiment specifically defines S5: In the S5, the ultrasonic detection module completes the detection and feedback of whether there are defects in the solder joint before the solder joint solidifies, can timely discover defects, immediately adjust printing parameters or take repair measures, avoid defects after the solder joint solidifies, save time and cost, improve production efficiency and product quality. For example, in the production of electronic products, timely discovering and handling unsolidified solder joint defects can reduce subsequent detection and rework processes; that is, it is necessary to control the ultrasonic detection time and the printing speed of the 3D printing equipment to ensure that the detection and feedback are completed before the solder joint solidifies, so as to buy time for defect repair; it should be noted that in the process of using this embodiment, due to real-time detection, high temperature, strong electromagnetic field, splashing and other factors will interfere with the ultrasonic signal, and metal splashing may hit the ultrasonic signal. The ultrasonic transducer of the wave detection module affects the detection, and shielding measures can be taken to wrap the ultrasonic transducer with a metal shielding cover to reduce electromagnetic field interference; select a suitable ultrasonic transducer position and angle to avoid metal splashing directly impacting the ultrasonic transducer, and a protective device can also be set to block splashing, etc.; in addition, the ultrasonic detection module needs to meet the requirements of high temperature resistance and harsh environment resistance, and a special high temperature resistant ultrasonic detection module is required. Its internal structure and materials can withstand the high temperature of the solder joint to prevent the ultrasonic transducer from being damaged or its performance degraded due to high temperature. At the same time, the circuits and electronic components must also adapt to harsh environments such as high temperature and humidity to ensure stable operation. The specific structure is not specifically limited in this application, and the technical personnel can make corresponding settings according to the specific situation.

[0059] In S5, during the welding process, the ultrasonic detection module detects the welding point in real time, which specifically includes the following steps: S51: In the system of the 3D printing device, add an information module that can provide real-time feedback on the print head position, print head speed and expected solder joint forming time; for example, by installing a high-precision displacement sensor on the 3D printing nozzle, the real-time position of the print head can be accurately obtained; the speed sensor is used to monitor the printing speed of the print head, and the expected forming time of each solder joint is calculated based on the estimated size of the solder joint, the set printing path and the current speed; select an ultrasonic detection module that can respond quickly and has high-precision detection capabilities. Its detection frequency and sensitivity should be optimized according to the solder joint material, size and possible defect types. The probe of the ultrasonic detection module is integrated near the printing area of ​​the 3D printing device to ensure that the probe can detect immediately after the solder joint is formed; for example, for tiny chip 3D printed solder joints, a high-frequency (such as 5-10MHz) ultrasonic detection probe can be selected to improve the detection resolution of tiny defects.

[0060] S52: Build a communication link between the 3D printing device and the ultrasonic transducer to ensure that data between the two can be transmitted in real time; the communication link uses a high-speed and reliable communication link, such as high-speed Ethernet or optical fiber communication technology, to achieve rapid exchange of data between the two. At the same time, the control system can receive and analyze the data from the 3D printing device and the ultrasonic detection module in real time, and make decisions according to a preset algorithm; S53: In the 3D printing device and the ultrasonic transducer, a unified time reference is set; for example, a high-precision clock module (atomic clock or high-precision crystal oscillator) is used to provide an accurate time signal for the 3D printing device and the ultrasonic transducer to ensure the time consistency between the two. S54: Predict the forming time of each solder joint in advance based on the speed and path planning of the 3D printing equipment; S55: before the solder joint is expected to be formed, the ultrasonic transducer is triggered in advance to perform preheating and preparation work; for example, when the 3D printing device calculates that there are 1 to 2 seconds before the next solder joint is formed, a pre-trigger signal is sent to the ultrasonic detection module, so that the ultrasonic detection module adjusts parameters in advance and prepares for detection; S56: The ultrasonic transducer transmits ultrasonic waves to the weld before the weld solidifies, and sends a signal to the signal receiving and processing unit before the weld solidifies.

[0061] Preferably, it also includes S57: according to the fluctuation of the actual speed of the 3D printing device, dynamically adjusting the detection time point of the ultrasonic transducer in real time; for example, if the printing speed of the print head changes due to some reason (such as changes in material supply, equipment vibration, etc.), the 3D printing device promptly sends the new speed information and the estimated welding point forming time to the control system, and the control system recalculates and adjusts the detection time of the ultrasonic detection module to ensure that the detection is completed before the welding point solidifies.

[0062] Optimize the control system's data processing and feedback process for the ultrasonic detection module to shorten the time from detection to feedback as much as possible; use high-speed data acquisition cards and efficient signal processing algorithms to quickly analyze and process the detected ultrasonic signals. For example, use a real-time signal processing module based on FPGA (field programmable gate array) to complete feature extraction and defect judgment of ultrasonic signals within a few milliseconds, and feed back the results to the 3D printing equipment. Embodiment 3

[0063] This embodiment specifically describes S6: In S6, before the control system controls the print head and the welding machine to print and repair the solder joint with the defective position, the signal processing algorithm and the ultrasonic detection module data fusion are first used to accurately determine the position of the defect in the solder joint, and provide accurate coordinates for 3D printing repair, which specifically includes the following steps: S61: Arrange multiple ultrasonic transducers reasonably around each solder joint to ensure that ultrasonic signals can be obtained from different angles; for example, 3 to 4 ultrasonic transducers can be evenly distributed above and on the sides of the solder joint to fully cover the solder joint area; S62: The ultrasonic transducer transmits ultrasonic waves to the welding point, receives reflected signals, and records information such as the time and amplitude of the signals received by each ultrasonic transducer; S63: The signal receiving and processing unit pre-processes the received electrical signal, removes noise from the ultrasonic signal, amplifies the filtered signal, enhances the signal amplitude, and amplifies the signal amplitude to a suitable range for subsequent analysis, while avoiding signal distortion; S64: Extract time-related features from the preprocessed signal, such as the propagation time and reflection time of the signal, and calculate the distance of the ultrasonic wave from the sensor to the defect position and then back by measuring the time difference between the transmitted signal and the reflected signal and combining the propagation speed of the ultrasonic wave in the solder joint material. For example, it is known that the propagation speed of the ultrasonic wave in the solder joint material is , the measured time difference is , then the propagation distance (divided by 2 because the distance is round trip); Analyze the changes in signal amplitude, signal reflection time, and signal waveform from the preprocessed signal. Defects of different locations and types will cause the amplitude, reflection time, and signal waveform of the ultrasonic reflection signal to be different. For example, a larger defect may significantly reduce the amplitude of the reflection signal. By establishing a relationship model between the amplitude and the defect location and size, the defect location can be determined. S65: The coordinates of the defect location are calculated based on the data fusion of triangulation positioning. The data collected by the ultrasonic transducer are fused using the principle of triangulation positioning. Assuming that three ultrasonic transducers , , The coordinates of are ( , )、( , )、( , ),from , , The distances to the defects are , , , with ultrasonic transducers as , , The center of the circle , , Draw a circle with radius , the intersection of the three circles is the location of the defect. The coordinates of the defect can be obtained by solving the following method (1) (2) (3) ( , ); (1) (2) (3) Preferably, in addition to triangulation positioning, data fusion algorithms such as weighted averaging method and Kalman filtering method can also be used; the weighted averaging method assigns different weights according to the reliability of the data of each ultrasonic actuator, and then calculates the weighted average to determine the defect location; the Kalman filtering method uses the state equation and observation equation of the system to optimally estimate the defect location, which can effectively deal with noise and uncertainty in the data. It belongs to the conventional prior art and the applicant will not elaborate on it here. The technicians can make corresponding choices according to the specific circumstances.

[0064] S66: Convert the defect position coordinates calculated by the above method into coordinates in a coordinate system that can be recognized by the 3D printing device; the conversion relationship between the ultrasonic detection module coordinate system and the 3D printing device coordinate system, converting the coordinates through transformation matrices such as translation and rotation; preferably, in order to improve the accuracy of the converted coordinates, it is necessary to calibrate the entire detection and calculation process, and use standard defect samples with known positions for multiple detections and calculations, compare the calculated defect position with the actual position, and adjust and optimize the calculation method and parameters according to the deviation to ensure that the defect position coordinates finally provided to the 3D printing device are accurate and reliable.

[0065] In S6, the printing nozzle pauses printing solder, and the welding parameters of the 3D printing device are reset, including: When the defect is a void, adjust at least one of the solder extrusion amount, print head moving speed, welding temperature, welding time, welding power, welding pressure, and welding frequency of the 3D printing device; reduce the print head moving speed by 20%-50% of the print head moving speed set in S2 to ensure that the solder fully fills the void. The lower speed can allow the solder to have enough time to spread and solidify, reduce the problem of insufficiency caused by rapid filling, avoid the generation of void defects, and when the void defect needs to be repaired, the solder also has enough time to spread and solidify in the void; increase the welding temperature by 20°C - based on the welding temperature set in S2. 30°C, enhance the fluidity of the solder, avoid the generation of void defects, and make it better fill the void when the void defects need to be repaired; extend the welding time, extend it by 0.3-0.5 seconds on the basis of the welding time set in S2, ensure that the solder is fully integrated with the surrounding solder joints, avoid the generation of void defects, and fill the voids and fuse with the surrounding solder joints when the void defects need to be repaired; increase the welding power, increase it by 20%-30% on the basis of the original power of S2, quickly increase the temperature, accelerate the flow of the solder, avoid the generation of void defects, and fill the void when the void defects need to be repaired; adjust the welding pressure, the welding pressure for the void is preferably 0.5N-1N, which is higher than the normally set welding pressure, ensure that the solder is closely combined with the surrounding of the void, avoid the generation of void defects, and help the solder fill the void when the void defects need to be repaired. The appropriate pressure can make the solder better enter every corner of the void and improve the filling effect; increase the welding frequency, increase the welding frequency set in S2 by 20%- 30%, to avoid void defects, and when void defects occur and need to be repaired, the increase in frequency helps fill the voids.

[0066] The calculation formula for the adjusted solder extrusion amount is: , M represents the adjusted solder extrusion amount; Indicates the solder extrusion amount set in S2; Indicates the amount of solder added, , Indicates the density of solder, It indicates the spreading and loss coefficient of solder when filling voids. The value range of S is 1.2-1.5. V indicates the volume of the void. , r is the radius of the cavity. It is necessary to analyze the approximate volume of the cavity through ultrasonic detection signals. Assuming that the cavity is approximately regarded as a sphere, the cavity radius can be estimated based on information such as the intensity and time difference of the ultrasonic reflection signal.

[0067] Based on the above description, in 3D printing equipment, the extrusion amount of solder is usually adjusted by the motor speed of the solder conveyor. The relationship between the motor speed and the extrusion amount of solder is set to be linear. It is known that the extrusion amount of solder corresponding to the motor speed n per unit time is hl (calibrated by experiment). In order to extrude the additional , the increase in motor speed .

[0068] Another solution is to calculate the solder extrusion amount as follows: , M represents the adjusted solder extrusion amount; represents the solder extrusion amount set in S2; d represents the distance between the void position and the center of the solder joint, , the midline coordinates of the solder joint are ( , , ), the coordinates of the void position are (X, Y, Z); j represents the solder extrusion coefficient, d and j are linearly related, and j increases by 0.05 for every 1 mm increase in d.

[0069] When the defect is a crack, at least one of the print head temperature, print head moving speed, welding temperature, welding time, welding power, welding pressure, and welding frequency is adjusted; the welding temperature is increased by 10°C-20°C on the basis of the welding temperature set in S2, which helps to enhance the fluidity of the solder and avoid the occurrence of crack defects, and when the crack defect needs to be repaired, the solder can flow into the crack better; the welding time is extended by 0.2-0.4 seconds on the basis of the welding time set in S2, so as to avoid the occurrence of crack defects, and when the crack defect needs to be repaired, the solder can be fully fused with the material at the crack; the welding power is increased by 10%-20% on the basis of the original power of S2, so as to avoid the occurrence of crack defects, and when the crack defect needs to be repaired, increasing the power can promote the infiltration and fusion of the solder at the crack; the welding pressure is adjusted, and the welding pressure for the crack is preferably 0.3-0.4 seconds. 0.8N pressure is used to avoid the occurrence of crack defects, and when crack defects need to be repaired, it helps to promote the flow of solder into the crack and enhance the bonding strength between the solder and the materials on both sides of the crack; the welding frequency is increased by 10% - 20% based on the welding frequency set in S2 to avoid the occurrence of crack defects, and when crack defects need to be repaired, the increase in frequency helps to fill the cracks.

[0070] More specifically, the moving speed of the print head is adjusted according to the length and direction of the crack, and the temperature of the print head is adjusted according to the depth of the crack.

[0071] The calculation formula of the moving speed of the print head is: ;in, Indicates the moving speed of the print head after adjustment; Indicates the moving speed of the print head set in S2; represents the crack length; D represents the velocity change corresponding to the unit crack length. For every 1 mm crack length, the change of D is ; represents the angle correction coefficient, , It is the angle between the crack direction and the moving direction of the print head. When the crack direction is perpendicular to the moving direction of the print head, , , do not adjust the speed; when the directions are inconsistent, , ; The adjusted moving speed of the print head is .

[0072] The calculation formula for the temperature of the print head is: ;in, Indicates the adjusted print head temperature; represents the temperature of the print head set in S2; h represents the crack depth; Indicates the temperature change corresponding to the unit crack depth. The crack depth of 1 μm is determined experimentally. The change is 1.5.

[0073] When the defect is an inclusion defect, adjust at least one of the print head moving speed, welding temperature, welding time, welding pressure, and welding frequency; reduce the print head moving speed, and reduce the print speed by 10%-20% of the print head temperature set in S2, so that the solder has more time to fill the solder joint area, reduce the problem of uneven solder distribution caused by inclusions, avoid the occurrence of inclusion defects, and facilitate repair of inclusion defects when they need to be repaired; increase the welding temperature, increase it by 10-20°C on the basis of S2 setting, which helps the solder to melt and flow better, so that inclusions can be better fused with the solder or squeezed out, avoid the occurrence of inclusion defects, and facilitate repair of inclusion defects when they need to be repaired; extend the welding time, usually by adding 0 on the basis of the welding time set in S2. .5-1 second, so that the solder has more time to interact with the inclusions, improve the welding quality, avoid the generation of inclusion defects, and facilitate the repair of inclusion defects when they need to be repaired; increase the welding pressure, increase 0.1-0.2N on the basis of the welding pressure set by S2, avoid the generation of inclusion defects, and when inclusion defects need to be repaired, it helps to squeeze the inclusions out of the weld or make them better combined with the solder; the frequency adjustment is specifically set according to the situation of the inclusions. For example, when encountering small and easily dispersed inclusions, the welding frequency can be appropriately increased. For example, when the frequency is in the range of 200Hz-500Hz, select the higher value of this range, 400Hz-500Hz; when the inclusions are large or hard, a lower frequency (laser welding) should be used, and the frequency can be adjusted to the range of 10Hz-30Hz.

[0074] When the defect is an unfused defect, adjust at least one of the print head moving speed, welding temperature, welding time, welding pressure, welding power, and welding frequency; reduce the print head moving speed, and reduce the print head speed to 30% - 50% of the print head temperature in S2, which can allow the solder to have more sufficient time to spread on the solder joint, reduce the occurrence rate of unfused defects, and when repairing unfused defects, the newly extruded solder can better contact with the original material and fill the gap; increase the welding temperature, increase it by 30°C - 40°C on the basis of the welding temperature set in S2, provide sufficient heat to fully melt the solder, avoid the occurrence of unfused defects, and promote the fusion of the solder in the unmelted area when repairing unfused defects; extend the welding time, and extend it by 0.4 - 0.6 seconds on the basis of the welding time set in S2. The extension of the time is to ensure that the solder has sufficient time to melt and fuse under sufficient heat, avoid the occurrence of unfused defects, and promote the fusion of the solder in the unmelted area when repairing unfused defects; increase the welding power, and increase it by 30% on the basis of the welding power set in S2. - 40%, which can speed up the heat transfer and make the solder reach the melting state faster to avoid the unfusion defect; adjust the welding pressure. When the unfusion defect occurs, the welding pressure is preferably 0.6N - 1.2N, which can make the unmelted interface in close contact, help the solder to better fill the gap and fuse, avoid the unfusion defect, and ensure that the solder can better fill the unmelted area under the action of pressure when repairing the unfusion defect, thereby enhancing the bonding strength of the solder joint; increase the welding frequency, which is 10-30% higher than the welding frequency set in the original S2. The frequency can speed up the fusion speed of the unfused interface, avoid the unfusion defect, and flow and fuse more quickly when repairing the unfusion defect, thereby improving the quality of the solder joint. Embodiment 4

[0075] This embodiment provides a chip bonding device, including: 3D printing equipment: accurately prints solder to the connection part between chip and substrate according to a preset path; specifically, the 3D printing equipment includes a 3D printing nozzle, a solder delivery system, and a motion control platform: the solder delivery system is connected to the 3D printing nozzle and is used to provide solder to the 3D printing nozzle; the 3D printing nozzle is used to accurately print the solder delivered by the solder delivery system to the connection part between chip and substrate; the motion control platform is connected to the 3D printing nozzle and is used to control the movement of the 3D printing nozzle.

[0076] Welding device: weld the chip and the substrate; Ultrasonic detection module: detects the quality of the solder joint; the ultrasonic detection module includes an ultrasonic generator and an ultrasonic transducer; the ultrasonic generator generates a high-frequency electrical signal; the ultrasonic transducer is connected to the ultrasonic generator, converts the high-frequency electrical signal of the ultrasonic generator into ultrasonic waves, transmits them to the solder joint to be detected, receives the ultrasonic signals reflected and scattered by the solder joint, and converts them into electrical signals; Signal receiving and processing unit: connected to the ultrasonic detection module, the signal receiving and processing unit analyzes the signal fed back by the ultrasonic detection module to determine the quality of the weld; more specifically, connected to the ultrasonic transducer, the ultrasonic transducer transmits the electrical signal to the signal receiving and processing unit, and the signal receiving and processing unit determines the quality of the weld based on the received electrical signal.

[0077] Control system: After resetting the printing parameters of the 3D printing device and / or the welding parameters of the welding machine, or after resetting the printing parameters of the 3D printing device and / or the welding parameters of the welding machine, the control system controls the print head and the welding machine to print and repair the welds at defective locations; and can calculate data such as the volume of the void, the depth and length of the crack, etc.

[0078] This application uses 3D printing equipment and ultrasonic detection equipment to weld photovoltaic panels, which has the following advantages: (1) Use ultrasound to monitor the internal quality of the solder joint in real time without contacting the solder joint and ensure the accuracy of defect detection; (2) Using 3D printing equipment to precisely control the print head movement speed, solder extrusion volume, and print head temperature, timely response and repair of welding defects can be achieved to ensure welding quality; (3) Adjust the printing parameters of the 3D printing equipment and the welding parameters of the welding machine in real time according to the defects, or control the printing nozzle and welding machine of the 3D printing equipment to perform secondary printing and repair on the welds at the defective positions after adjusting the parameters; avoid welding problems in the entire batch due to problems with welding parameters, etc., resulting in low work efficiency.

[0079] The connection relationships in the present invention are all well-known connection methods in the prior art, such as electrical connection and other connection methods. The technicians can choose according to the specific situation. Simple changes in the connection method all fall within the protection scope of the present invention; however, the present application is not limited to the specific details in the above-mentioned embodiments. Within the technical concept of the present application, the technical scheme of the present application can be subjected to a variety of simple modifications, and these simple modifications all fall within the protection scope of the present application; in addition, it should be noted that the various specific technical features described in the above-mentioned specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the various possible combinations of the present application will not be described separately.

[0080] In addition, the various implementation modes of the present application may be arbitrarily combined, and as long as they do not violate the concept of the present application, they should also be regarded as the contents disclosed in the present application.

Claims

1. A chip welding method, characterized in that: The steps include: S1: pre-process the chip and substrate; S2: Set the printing parameters of the 3D printing device and the welding parameters of the welding machine. The printing parameters include the moving speed of the print head, the amount of solder extruded, and the temperature of the print head; the welding parameters include the welding temperature, welding time, welding power, welding pressure, and welding frequency; S3: calibrate the ultrasonic detection module; S4: The print head of the 3D printing device prints solder, and the chip is welded by the welding machine; S5: During the welding process, the ultrasonic detection module detects the welding point in real time to determine whether the welding point has defects. If not, welding continues. If yes, the process goes to S6; S6: The control system controls the print head to stop printing solder, and resets the printing parameters of the 3D printing device and / or the welding parameters of the welding machine. Alternatively, after resetting the printing parameters of the 3D printing device and / or the welding parameters of the welding machine, the control system controls the print head and the welding machine to print and repair the solder joints at the defective positions. S7: During the printing and repairing process, the ultrasonic detection module detects the solder joints in real time to determine whether there are defects in the solder joints. If not, continue welding. If yes, enter S6.

2. A chip bonding method according to claim 1, characterized in that: In S5, the ultrasonic detection module completes the detection and feedback of whether the solder joint has defects before the solder joint solidifies.

3. A chip bonding method according to claim 1, characterized in that: In S5, during the welding process, the ultrasonic detection module detects the welding point in real time, including the following steps: S51: In the 3D printing equipment system, an information module is added that can provide real-time feedback on the print head position, print speed, and estimated solder joint forming time; S52: Building a communication link between the 3D printing device and the ultrasonic transducer of the ultrasonic detection module to ensure that data between the two can be transmitted in real time; S53: Setting a unified time reference in the 3D printing device and the ultrasonic transducer; S54: Predict the forming time of each solder joint in advance based on the speed and path planning of the 3D printing equipment; S55: Before the solder joint is expected to be formed, the ultrasonic transducer is triggered in advance for preheating and preparation; S56: The ultrasonic transducer transmits ultrasonic waves to the weld before the weld solidifies, and sends a signal to the signal receiving and processing unit before the weld solidifies.

4. A chip bonding method according to claim 1, characterized in that: In S5, the solder joint defects include void defects, crack defects, inclusion defects, and unfused defects.

5. A chip bonding method according to claim 4, characterized in that: When the solder joint defect is a void defect, the calculation formula for the solder extrusion amount is: , M represents the adjusted solder extrusion amount; Indicates the solder extrusion amount set in S2; Indicates the amount of solder added, , Indicates the density of solder, It represents the spreading and loss coefficient of solder when filling voids. The value range of S is 1.2-1.5, and V represents the volume of the void.

6. A chip bonding method according to claim 4, characterized in that: When the solder joint defect is a void defect, the calculation formula for the solder extrusion amount is: , Indicates the adjusted solder extrusion amount; represents the solder extrusion amount set in S2; d represents the distance between the void position and the center of the solder joint, , the midline coordinates of the solder joint are , the coordinates of the hole position are ; j represents the solder extrusion coefficient, d and j are in a linear relationship, and j increases by 0.05 for every 1mm increase in d.

7. A chip bonding method according to claim 4, characterized in that: When the solder joint defect is a crack defect, the calculation formula for the moving speed of the print head is: ;in, Indicates the moving speed of the print head after adjustment; Indicates the moving speed of the print head set in S2; represents the crack length; D represents the velocity change corresponding to the unit crack length. For every 1 mm crack length, the change of D is ; represents the angle correction coefficient, , It is the angle between the crack direction and the direction of print head movement.

8. A chip bonding method according to claim 4, characterized in that: When the solder joint defect is a crack defect, the calculation formula for the temperature of the print head is: ;in, Indicates the adjusted print head temperature; represents the temperature of the print head set in S2; h represents the crack depth; Indicates the temperature change corresponding to the unit crack depth. For every 1μm crack depth, The change is 1.

5.

9. A chip bonding device, characterized in that: include: 3D printing equipment: accurately prints solder to the connection between the chip and the substrate according to the preset path; Welding device: weld the chip and the substrate; Ultrasonic detection module: detect the quality of solder joints; Signal receiving and processing unit: connected to the ultrasonic detection module, the signal receiving and processing unit analyzes the signal fed back by the ultrasonic detection module to determine the quality of the solder joint; Control system: adjust the welding parameters of the 3D printing equipment and / or repair problematic welds in real time based on the ultrasonic detection results.

10. The chip bonding device according to claim 9, characterized in that: The ultrasonic detection module includes an ultrasonic generator and an ultrasonic transducer: Ultrasonic generator: generates high-frequency electrical signals; Ultrasonic transducer: connected to the ultrasonic generator, converts the high-frequency electrical signal of the ultrasonic generator into ultrasonic waves, transmits them to the welding point that needs to be inspected, receives the ultrasonic signals reflected and scattered back by the welding point, converts them into electrical signals, and transmits the electrical signals to the signal receiving and processing unit.