A full-optical laser-ultrasonic imaging detection system and method for flip-chip solder joint defects
The all-optical laser ultrasonic imaging inspection system, which combines an optical microphone probe and an ultrasonic lens with a pulsed laser, solves the problem of high-resolution detection of solder joint defects in flip chips, achieving efficient and accurate non-destructive testing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA UNIV OF PETROLEUM (BEIJING)
- Filing Date
- 2025-12-18
- Publication Date
- 2026-07-21
AI Technical Summary
Existing technologies struggle to achieve high-resolution non-destructive testing of solder joint defects on flip-chip devices, especially for the precise detection of tiny solder joints. Furthermore, traditional ultrasonic probes suffer from narrow bandwidth and complex mechanical structures.
An all-optical laser ultrasonic imaging inspection system is adopted, which uses an optical microphone probe and ultrasonic lens combined with a pulsed laser to generate ultrasonic waves through photoacoustic effect. Feature images are extracted by combining polynomial fitting method to achieve high-resolution imaging of weld defects.
It enables efficient, accurate, and non-destructive testing of solder joint defects on flip chips, is suitable for in-situ testing, improves detection resolution and efficiency, provides quantitative characterization methods, and is applicable to complex environments.
Smart Images
Figure CN121612992B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to an all-optical laser ultrasonic imaging detection system and method, belonging to the field of integrated circuit and precision measurement technology. Background Technology
[0002] Flip-chip technology, as a new generation of high-end electronic packaging technology, strongly supports the development of my country's strategic emerging industries, such as artificial intelligence, 5G / 6G communications, and high-end equipment. Flip-chip technology features high processing speed, low loss, and low cost, and has become the mainstream technology in the microelectronics packaging field, with wide applications in computers, electronic communications, military defense, and many other areas. As a crucial connection between the chip and the PCB substrate, the solder joints of flip-chip devices directly affect the mechanical, thermal, and electrical properties of the entire packaged product. However, with the increasing miniaturization of flip-chip solder joints and their encapsulation within the chip and substrate, reliable defect detection of the internal interconnect solder joints of flip-chip devices presents a challenge.
[0003] Currently, X-ray, CT imaging, and ultrasonic microscopy are commonly used methods for detecting solder joint defects in flip-chip devices both domestically and internationally. X-ray and CT imaging both use radioactive rays to inspect samples, offering high accuracy but imposing specific environmental requirements. Ultrasonic microscopy offers advantages such as high sensitivity and accuracy, but it requires a coupling agent and can easily cause irreversible changes to the sample surface, thus limiting its use to batch sampling inspections. Furthermore, both methods struggle to achieve in-situ inspection of samples.
[0004] Laser ultrasonic testing, as a novel non-destructive testing (NDT) technology, integrates laser and ultrasonic technologies. It is commonly used for detecting surface cracks, composite material delamination, stress, and grain size. For example, Lu Minghui et al. from Nanjing University used laser ultrasonic testing to inspect the quality of solder joints with a diameter of 1.2 mm. However, traditional ultrasonic probes often employ air-coupled piezoelectric ultrasonic detectors, which have narrow bandwidth and complex mechanical structures. Their detection resolution is insufficient to meet the demands of increasingly smaller solder joint sizes. Therefore, there is an urgent need to develop a fully optical laser ultrasonic imaging method and system for detecting defects in flip-chip solder joints, enabling non-destructive testing of minute solder joint defects. Summary of the Invention
[0005] To address the problems of low resolution and cumbersome feature extraction in current laser ultrasonic testing of solder joints, this invention proposes a fully optical laser ultrasonic imaging detection system and method for flip-chip solder joint defects.
[0006] The technical solution adopted by the present invention to solve the above problems is as follows: The all-optical laser ultrasonic imaging detection system for flip chip solder joint defects of the present invention includes an optical platform, an X-axis moving stage, a Z-axis moving stage and an optical microphone clamping assembly, a Y-axis moving stage on the X-axis moving stage, a sample clamping assembly on the Y-axis moving stage, a sample to be tested clamped on the sample clamping assembly, an optical microphone probe clamped on the optical microphone clamping assembly, a first optical rod on the Z-axis moving stage, an optical focusing lens on the first optical rod, the optical focusing lens being located above the sample to be tested, and the optical microphone probe being located below the sample to be tested, both the optical focusing lens and the optical microphone probe being connected to a computer component.
[0007] Furthermore, the computer component includes a computer connected to a high-frequency data acquisition card via a fourth signal transmission line. The high-frequency data acquisition card is connected to a pulsed laser module via a third signal transmission line. The pulsed laser module is connected to an optical focusing lens via a pulsed laser fiber. The high-frequency data acquisition card is also connected to an optical microphone host via a second signal transmission line. The optical microphone host is connected to an optical microphone probe via a first signal transmission line.
[0008] Furthermore, the optical microphone clamping assembly includes a third optical rod, the lower end of which is fixed to an optical platform, an optical microphone probe clamp is fixedly connected to the middle of the third optical rod, an optical microphone probe is clamped on the optical microphone probe clamp, and an ultrasonic lens is fixedly connected to the upper end of the third optical rod, the ultrasonic lens being located between the optical microphone probe and the sample to be tested.
[0009] Furthermore, the sample clamping assembly includes a fourth optical rod, the lower end of which is fixed on the Y-axis moving stage, and the upper end of which is fixedly connected to one end of a fifth optical rod. The other end of the fifth optical rod is fixedly connected to a sample holder, which holds the sample to be tested.
[0010] The steps of the all-optical laser ultrasonic imaging detection method for flip chip solder joint defects described in this invention include: Step 1: The Y-axis and Z-axis moving stages are rigidly mounted on the optical platform, and the X-axis moving stage is mounted on the Y-axis moving stage to achieve two-dimensional horizontal movement and vertical movement. Step 2: The sample to be tested is held by the sample holder and mounted on the X-axis moving stage by the fourth and fifth optical rods, so that the Y-axis moving stage and the X-axis moving stage can drive the sample to be tested to perform two-dimensional scanning. Step 3: Install the pulsed laser fiber onto the optical focusing lens and mount it on the Z-axis moving stage via the first optical rod. The optical focusing lens can be driven to move up and down. By adjusting the position of the optical focusing lens, ensure that the light spot is perpendicularly irradiated on the upper surface of the sample under test. By adjusting the up and down position of the Z-axis moving stage, minimize the diameter of the pulsed light spot irradiating the upper surface of the sample under test. Step 4: The third optical rod is rigidly mounted on the optical platform. The ultrasonic lens is fixed to the third optical rod through the second optical rod, and the center of the ultrasonic lens is directly below the optical focusing lens. The optical microphone probe is fixedly mounted on the third optical rod through the optical microphone probe clamp. The horizontal position of the optical microphone probe is adjusted so that the axis of the optical microphone probe coincides with the axis of the optical focusing lens and the ultrasonic lens, and is directly below the ultrasonic lens, 1~4mm away from the lower surface of the sample to be tested. By jointly adjusting the position of the ultrasonic lens and the optical microphone probe, the ultrasonic signal is maximized. Step 5: The ultrasonic signal detected by the optical microphone probe is transmitted to the optical microphone host through the first signal transmission line. The optical microphone host is connected to the input channel of the high-speed data acquisition card through the second signal transmission line. The trigger signal of the high-speed data acquisition card and the clock signal of the pulse laser module are connected through the third signal transmission line. Step 6: The pulsed laser module, controlled by a computer, emits a pulsed laser beam. The beam passes through the laser focusing lens via the fiber optic cable of the pulsed laser and is focused onto the surface of the sample to be tested. This generates an ultrasonic signal, which is focused onto the optical microphone probe by the ultrasonic lens and reaches the microphone host via the first signal transmission line. The signal is then acquired by the high-speed data acquisition card via the second signal transmission line. The clock signal of the pulsed laser module is synchronized to the high-speed data acquisition card. The ultrasonic signal acquired by the high-speed data acquisition card is transmitted to the computer via the fourth signal transmission line. The computer then controls the X-axis or Y-axis moving stage to move to the next position. Step 7: Based on the size of the sample to be tested, scan other positions of the sample by adjusting the Y-axis moving stage 11 and the X-axis moving stage. Repeat step 6 and store the laser ultrasound data in the computer. Step 8: Perform wavelet denoising on each location point of the sample scanning data stored in the computer. The wavelet denoising uses the Dobessi wavelet function, db3, and the "UniversalThreshold" threshold to obtain the wavelet denoised signal. Step 9: Observe the time-domain ultrasonic signal at the center point of the sample, the time points of the first peak and trough of the defect, and use this as the time window data. Then, use this window data and a fifth-order polynomial to fit the signal. The data point at this location is used to obtain the coefficient for that point. p 1. p 2. p 3. p4. p 5 and p 6; Step 10: Repeat step 9 for other scanning locations to obtain 6 coefficient matrices. These 6 coefficient matrices are the feature images of the present invention, which clearly show the defect distribution.
[0011] The beneficial effects of this invention are: 1. This invention employs all-optical laser ultrasonic testing technology for high-resolution imaging detection of solder joint defects in flip chips: A pulsed laser excites the surface of the flip chip, and due to the photoacoustic effect, instantaneous ultrasonic waves propagate from the top surface into the sample's interior and surrounding medium. The ultrasonic signal is detected directly below the sample by an all-optical microphone detector. If a solder joint defect exists (missing components or microcracks, etc.), the defective portion of the medium is air, which severely attenuates the ultrasonic signal, resulting in a reduced ultrasonic signal amplitude and a prolonged arrival time at the detector. Based on this physical process, it can be used for reliable detection of solder joint defects in flip chips.
[0012] 2. In order to enhance the ultrasonic signal received by the detector, the present invention adds an ultrasonic lens above the detector, which can greatly improve the ultrasonic signal and provide a powerful means for accurate and reliable detection of solder joint defects of flip chips. It also provides a quantitative characterization means for in-situ detection and has broad application prospects.
[0013] 3. This invention is based on the principle of all-optical laser ultrasonic testing, which is a non-contact imaging testing method. The test results are intuitive, and the pulsed laser has a high repetition rate (above 1kHz), which greatly improves the testing capability and efficiency.
[0014] 4. This invention proposes a laser ultrasound feature image and data dimensionality reduction method. The three-dimensional sequence matrix is fitted by a polynomial fitting method, and the fitting coefficient is the feature image. Its signal-to-noise ratio is greater than that of feature extraction algorithms such as wavelet denoising, principal component analysis, and independent component analysis. It also has the advantages of fast dimensionality reduction speed and small storage space.
[0015] 5. This invention employs all-optical laser ultrasonic testing technology, which is not only suitable for detecting defects in solder joints of inverted chips, but also for imaging detection and quality evaluation of internal defects of similar thickness and structure.
[0016] 6. The method described in this invention has high detection efficiency and high resolution, and is suitable for detection under complex conditions such as in-situ and extreme environments. Another objective of this invention is to provide a fast feature extraction method that achieves a higher image signal-to-noise ratio compared to currently popular feature extraction methods such as amplitude mapping, wavelet denoising, and principal component analysis. Attached Figure Description
[0017] Figure 1 This is a schematic diagram of the principle of the all-optical laser ultrasonic imaging detection system for flip chip solder joint defects of the present invention; Figure 1 In the middle: 1-Pulsed laser fiber, 2-First optical rod, 3-Optical focusing lens, 4-Focused laser spot, 5-Sample to be tested, 6-Ultrasonic lens, 7-Focused sound beam, 8-Second optical rod, 9-Third optical rod, 10-Y-axis moving stage, 11-X-axis moving stage, 12-Optical platform, 13-Optical microphone probe fixture, 14-Optical microphone probe, 15-First signal transmission line, 16-Optical microphone host, 17-Second signal transmission line, 18-Third signal transmission line, 19-Pulsed laser module, 20-High frequency data acquisition card, 21-Fourth signal transmission line, 22-Computer, 23-Fourth optical rod, 24-Fifth optical rod, 25-Sample holder, 26-Z-axis moving stage; Figure 2 This is a schematic diagram of a flip-chip structure; Figure 3 This is a schematic diagram of the original laser ultrasound signal, the wavelet-denoised signal, and the selection of the polynomial fitting time window. Figure 4 This is a schematic diagram of the fifth-order polynomial fitting curve of ultrasound signal and number of samples; Figure 5 These are the laser ultrasonic testing results and characteristic images of the inverted chip sample; Figure 5 (a) is wavelet denoising with a signal-to-noise ratio (SNR) of 12.74; Figure 5 (b) is the principal component analysis, with an SNR of 37.87; Figure 5 (c) is the independent component analysis, SNR=23.43; Figure 5 (d) are the polynomial fitting coefficients, SNR=51.04. Detailed Implementation
[0018] Specific implementation method one: as follows Figure 1 As shown, a full-optical laser ultrasonic imaging detection system for flip chip solder joint defects includes a pulsed laser fiber 1, a first optical rod 2, an optical focusing lens 3, a focused laser spot 4, a sample under test 5, an ultrasonic lens 6, a focused sound beam 7, a second optical rod 8, a third optical rod 9, a Y-axis moving stage 10, an X-axis moving stage 11, an optical platform 12, an optical microphone probe holder 13, an optical microphone probe 14, a first signal transmission line 15, an optical microphone host 16, a second signal transmission line 17, a third signal transmission line 18, a pulsed laser module 19, a high-frequency data acquisition card 20, a fourth signal transmission line 21, a computer 22, a fourth optical rod 23, a fifth optical rod 24, a sample holder 25, and a Z-axis moving stage 26. The optical platform 12 carries the Y-axis moving stage 11 and the Z-axis moving stage 26 and is fixed by a rigid connection. The optical platform 12 also carries the third optical rod 9 and is rigidly connected to it. The X-axis moving stage 10 is connected to the Y-axis moving stage 11. The X-axis moving stage 10 carries the fourth optical rod 23. The fourth optical rod 23 is connected to the fifth optical rod 24 and the sample holder 25. The sample 5 to be tested is kept horizontal by adjusting the installation position of the sample holder. The sample 5 is moved in two dimensions by the control of the X-axis moving stage 10 and the Y-axis moving stage 11 by the computer 22. The Z-axis moving stage 26 is rigidly connected to the first optical rod 2, which is connected to the optical focusing lens. The height of the Z-axis moving stage 26 is controlled by the computer 22 so that the focused laser spot 4 is perpendicularly irradiated onto the surface of the sample 5 to be tested, and the laser energy at that position is maximized. The optical microphone probe 14 is fixed to the third optical rod 9 by the optical microphone probe clamp 13, which ensures that the optical microphone probe 14 is directly below the focused laser spot 4 and 1-4 mm away from the lower surface of the sample 5 to be tested; The ultrasonic lens 6 is installed directly above the optical microphone probe 14. By jointly adjusting the ultrasonic lens 6 and the optical microphone probe 14, the ultrasonic signal is made to reach its maximum value. The optical microphone host 16 and the optical microphone probe 14 are flexibly connected through the first signal transmission line 15. The ultrasonic signal acquired by the optical microphone host 16 is transmitted to the signal acquisition channel of the high-speed data acquisition card 20 through the second signal transmission line 17. The synchronous trigger signal of the high-speed acquisition card 20 and the clock signal of the pulse laser 19 are connected through the third signal transmission line 18 to realize the synchronization of laser excitation and ultrasonic signal acquisition. Computer 22 is connected to high-speed data acquisition card 20 via fourth signal transmission line 21. Computer 22 can control the X-axis moving stage 10, Y-axis moving stage 11 and Z-axis moving stage 26 to move individually and in combination. It can also adjust the power and repetition rate of pulse laser module 19 and the sampling parameters of high-speed data acquisition card 20. Computer 22, pulsed laser module 19, X-axis moving stage 10, Y-axis moving stage 11, Z-axis moving stage 26, high-speed data acquisition card 20, and optical microphone host 16 are powered on. After setting appropriate detection parameters, the pulsed laser, controlled by the software of computer 22, is focused from the pulsed laser module 19 through the pulsed laser fiber 1 and the laser focusing lens 3 to obtain a focused laser spot 4 that radiates onto the surface of sample 5. The generated ultrasonic signal is focused by ultrasonic lens 6 onto optical microphone probe 14 and reaches the microphone host through the first signal transmission line 18. It is then acquired by high-speed data acquisition card 17 through the second signal transmission line 17. The clock signal of pulsed laser module 19 is synchronized to high-speed data acquisition card 17. The ultrasonic signal acquired by high-speed data acquisition card 17 is transmitted to computer 22 through the fourth signal transmission line 21. Then, computer 22 controls X-axis moving stage 10 or Y-axis moving stage 11 to move to the next position point. This process is repeated to achieve two-dimensional scanning of sample 5 under test. The ultrasonic data is stored in computer 22.
[0019] like Figure 2 As shown, the test sample 5 consists of a chip, an array of solder balls (4×4 solder balls), and a PCB substrate, and the corresponding three-dimensional structure is given. The diameter of a single solder ball is 550 mm and the height is 400 mm. It can be seen from the figure that two solder balls are missing.
[0020] like Figure 3 As shown, the original signal is a signal acquired by a high-speed signal acquisition card at a certain location. After wavelet denoising, a wavelet-denoised signal is obtained. The wavelet denoising uses the Dobesi wavelet function, db3, and the "UniversalThreshold" threshold to obtain the wavelet-denoised signal. The experimental data selected by the marked time window are used for subsequent polynomial fitting.
[0021] like Figure 4 As shown, the selected time window data is processed using a fifth-degree polynomial: The fitting result can be obtained, and the coefficient at that point can be obtained. p 1. p 2. p 3. p 4. p 5 and p 6. The time window varies depending on the sample thickness, but the time window used for the same sample is fixed. 21 data points were selected for fitting this sample.
[0022] like Figure 5 As shown, the feature image obtained by wavelet denoising, principal component analysis, independent component analysis, and the feature extraction method proposed in this invention for sample 5 clearly shows the missing ball problem in the upper right corner; among them, the signal-to-noise ratio (SNR) of wavelet denoising is 12.74, the SNR of principal component analysis is 37.87, the SNR of independent component analysis is 23.43, and the polynomial fitting coefficient (p 3) has an SNR of 51.04.
[0023] Specific Implementation Method Two: A fully optical laser ultrasonic imaging method for detecting solder joint defects on flip chips, comprising the following steps: Step 1: The Y-axis moving stage 11 and the Z-axis moving stage 26 are rigidly mounted on the optical platform 12, and the X-axis moving stage 10 is mounted on the Y-axis moving stage 11 to realize two-dimensional horizontal movement and vertical movement. Step 2: The sample to be tested 5 is held by the sample holder 25 and installed on the X-axis moving stage 10 through the fourth optical rod 23 and the fifth optical rod 24, so that the Y-axis moving stage 11 and the X-axis moving stage 10 can drive the sample to be tested 5 to perform two-dimensional scanning. Step 3: Install the pulsed laser fiber 1 onto the optical focusing lens 3, and install it on the Z-axis moving stage 26 via the first optical rod 2. The optical focusing lens 3 can be driven to move up and down. By adjusting the position of the optical focusing lens 3, the light spot is ensured to be perpendicularly irradiated on the upper surface of the sample 5 under test. By adjusting the up and down position of the Z-axis moving stage 26, the diameter of the pulsed light spot irradiating the upper surface of the sample 5 under test is minimized. Step 4: The third optical rod 9 is rigidly mounted on the optical platform 12. The ultrasonic lens 6 is fixed to the third optical rod 9 through the second optical rod 8, and the center of the ultrasonic lens 6 is directly below the optical focusing lens 3. The optical microphone probe 14 is fixedly mounted on the third optical rod 9 through the optical microphone probe clamp 13. The horizontal position of the optical microphone probe 14 is adjusted so that the axis of the optical microphone probe 14 coincides with the axis of the optical focusing lens 3 and the ultrasonic lens 6, and is directly below the ultrasonic lens 6, 1~4mm away from the lower surface of the sample 5 to be tested. By jointly adjusting the positions of the ultrasonic lens 6 and the optical microphone probe 14, the ultrasonic signal is maximized. Step 5: The ultrasonic signal detected by the optical microphone probe 14 is transmitted to the optical microphone host 16 through the first signal transmission line 15. The optical microphone host 16 is connected to the input channel of the high-speed data acquisition card 20 through the second signal transmission line 17. The trigger signal of the high-speed data acquisition card 20 and the clock signal of the pulse laser module 19 are connected through the third signal transmission line 18. Step 6: The computer 22 controls the pulsed laser module 19 to emit pulsed laser light. The pulsed laser light passes through the laser focusing lens 3 via the laser fiber 1 to obtain a focused laser spot 4, which is radiated onto the surface of the sample 5 under test. The generated ultrasonic signal is focused by the ultrasonic lens 6 onto the optical microphone probe 14 and reaches the microphone host via the first signal transmission line 18. It is then acquired by the high-speed data acquisition card 17 via the second signal transmission line 17. The clock signal of the pulsed laser module 19 is synchronized to the high-speed data acquisition card 17. The ultrasonic signal acquired by the high-speed data acquisition card 17 is transmitted to the computer 22 via the fourth signal transmission line 21. Then, the computer 22 controls the X-axis moving stage 10 or the Y-axis moving stage 11 to move to the next position point. Step 7: Based on the size of the sample 5 to be tested, scan other positions of the sample 5 to be tested by adjusting the Y-axis moving stage 11 and the X-axis moving stage 10. Repeat step 6 and store the laser ultrasonic data in the computer 22. Step 8: Perform wavelet denoising on the data of each location point of the sample scanning data stored in the computer 22. The wavelet denoising uses the Dobessi wavelet function, db3, and the "UniversalThreshold" threshold to obtain the wavelet denoised signal. Step 9: Observe the time-domain ultrasonic signal at the center point of the sample, the time points of the first peak and trough of the defect, and use this as the time window data. Then, use this window data and a fifth-order polynomial to fit the signal. The data point at this location is used to obtain the coefficient for that point. p 1. p 2. p 3. p 4. p 5 and p 6; Step 10: Repeat step 9 for other scanning locations to obtain 6 coefficient matrices. These 6 coefficient matrices are the feature images of the present invention, which clearly show the defect distribution.
[0024] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent substitutions, and improvements made to the above embodiments without departing from the scope of the present invention, based on the technical essence of the present invention and within the spirit and principles of the present invention, shall still fall within the protection scope of the present invention.
Claims
1. A method for detecting solder joint defects in flip-chip chips using all-optical laser ultrasonic imaging, the method being implemented based on an all-optical laser ultrasonic imaging system for detecting solder joint defects in flip-chip chips, the system comprising an optical platform (12), the optical platform (12) being provided with an X-axis moving stage (11), a Z-axis moving stage (26) and an optical microphone clamping assembly, the X-axis moving stage (11) being provided with a Y-axis moving stage (10), the Y-axis moving stage (10) being provided with a sample clamping assembly, the sample clamping assembly clamping a sample to be tested (5), and the optical microphone clamping assembly clamping an optical microphone probe (14). The Z-type mobile stage (26) is equipped with a first optical rod (2), and an optical focusing lens (3) is provided on the first optical rod (2). The optical focusing lens (3) is located above the sample to be tested (5), and the optical microphone probe (14) is located below the sample to be tested (5). Both the optical focusing lens (3) and the optical microphone probe (14) are connected to the computer component. The computer component includes a computer (22), which is connected to a high-frequency data acquisition card (20) through a fourth signal transmission line (21). The high-frequency data acquisition card (20) is connected to a pulsed laser module (3) through a third signal transmission line (18). 19) Connection: The pulsed laser module (19) is connected to the optical focusing lens (3) via the pulsed laser fiber (1). The high-frequency data acquisition card (20) is also connected to the optical microphone host (16) via the second signal transmission line (17). The optical microphone host (16) is connected to the optical microphone probe (14) via the first signal transmission line (15). The optical microphone clamping assembly includes a third optical rod (9). The lower end of the third optical rod (9) is fixed on the optical platform (12). An optical microphone probe clamp (13) is fixedly connected to the middle of the third optical rod (9). An optical microphone probe (14) is held in a clamp on the device (13). An ultrasonic lens (6) is fixedly connected to the upper end of the third optical rod (9). The ultrasonic lens (6) is located between the optical microphone probe (14) and the sample to be tested (5). The sample clamping assembly includes a fourth optical rod (23). The lower end of the fourth optical rod (23) is fixed on the Y-axis moving stage (10). One end of the fifth optical rod (24) is fixedly connected to the upper end of the fourth optical rod (23). The other end of the fifth optical rod (24) is fixedly connected to a sample holder (25). The sample holder (25) holds the sample to be tested (5). Its features are, The steps of the all-optical laser ultrasonic imaging detection method for flip chip solder joint defects include: Step 1: The Y-axis moving stage (11) and the Z-axis moving stage (26) are rigidly mounted on the optical platform (12), and the X-axis moving stage (10) is mounted on the Y-axis moving stage (11) to realize two-dimensional horizontal movement and vertical movement. Step 2: The sample to be tested (5) is held by the sample holder (25) and installed on the X-axis moving stage (10) by the fourth optical rod (23) and the fifth optical rod (24), so that the Y-axis moving stage (11) and the X-axis moving stage (10) can drive the sample to be tested (5) to perform two-dimensional scanning. Step 3: Install the pulsed laser fiber (1) onto the optical focusing lens (3) and install it on the Z-axis moving stage (26) through the first optical rod (2). The optical focusing lens (3) can be driven to move up and down. By adjusting the position of the optical focusing lens (3), the light spot is ensured to be vertically irradiated on the upper surface of the sample to be tested (5). By adjusting the up and down position of the Z-axis moving stage (26), the diameter of the pulsed light spot irradiating the upper surface of the sample to be tested (5) is minimized. Step 4: The third optical rod (9) is rigidly installed on the optical platform (12). The ultrasonic lens (6) is fixed to the third optical rod (9) through the second optical rod (8), and the center of the ultrasonic lens (6) is directly below the optical focusing lens (3). The optical microphone probe (14) is fixedly installed on the third optical rod (9) through the optical microphone probe clamp (13). The horizontal position of the optical microphone probe (14) is adjusted so that the axis of the optical microphone probe (14) coincides with the axis of the optical focusing lens (3) and the ultrasonic lens (6), and is directly below the ultrasonic lens (6), 1~4mm away from the lower surface of the sample (5) to be tested. By jointly adjusting the position of the ultrasonic lens (6) and the optical microphone probe (14), the ultrasonic signal is maximized. Step 5: The ultrasonic signal detected by the optical microphone probe (14) is transmitted to the optical microphone host (16) through the first signal transmission line (15). The optical microphone host (16) is connected to the input channel of the high-speed data acquisition card (20) through the second signal transmission line (17). The trigger signal of the high-speed data acquisition card (20) and the clock signal of the pulse laser module (19) are connected through the third signal transmission line (18). Step 6: The computer (22) controls the pulsed laser module (19) to emit pulsed laser light. The pulsed laser light passes through the laser focusing lens (3) via the pulsed laser fiber 1 to obtain a focused laser spot (4) which is radiated onto the surface of the sample (5) to be tested, generating an ultrasonic signal. The ultrasonic signal is focused onto the optical microphone probe (14) via the ultrasonic lens (6) and reaches the microphone host via the first signal transmission line (18). It is also collected by the high-speed data acquisition card (17) via the second signal transmission line (17). The clock signal of the pulsed laser module (19) is synchronized to the high-speed data acquisition card (17). The ultrasonic signal collected by the high-speed data acquisition card (17) is transmitted to the computer (22) via the fourth signal transmission line (21). Then the computer (22) controls the X-axis moving stage (10) or the Y-axis moving stage (11) to move to the next position point. Step 7: Based on the size of the sample (5) to be tested, scan other positions of the sample (5) to be tested by adjusting the Y-axis moving stage 11 and the X-axis moving stage (10), repeat step 6, and store the laser ultrasound data in the computer (22); Step 8: Perform wavelet denoising on the data of each location point of the sample scanning data stored in the computer (22). The wavelet denoising uses the Dobessi wavelet function, db3, and "UniversalThreshold" threshold to obtain the wavelet denoised signal. Step 9: Observe the time-domain ultrasonic signal at the center point of the sample, the time points of the first peak and trough of the defect, and use this as the time window data. Then, use this window data and a fifth-order polynomial to fit the signal. The data point at this location is used to obtain the coefficient for that point. p 1. p 2. p 3. p 4. p 5 and p 6; Step 10: Repeat step 9 for other scanning locations to obtain 6 coefficient matrices. These 6 coefficient matrices are the feature images of the present invention, which clearly show the defect distribution.