A PCB circuit board defect detection system

Abstract

The application relates to the technical field of material physical property detection, and discloses a PCB (Printed Circuit Board) defect detection system, which comprises a quality analysis unit, an atmosphere control unit and a data processing unit. The data processing unit is used for acquiring a first measurement signal of the quality analysis unit, acquiring a process control signal of the atmosphere control unit and integrating the process control signal to obtain a second measurement signal. The data processing unit monitors the noise of the first measurement signal, and when it is determined that the first measurement signal is interfered, the second measurement signal is used for judgment. The second measurement signal based on flow is introduced, and the first measurement signal based on quality is used for orthogonal verification. When the first measurement signal is interfered by mechanical vibration, the second measurement signal is used for judgment, so that the detection reliability of the system in an industrial environment is ensured.

Description

Technical Field

[0001] This invention relates to a PCB circuit board defect detection system, belonging to the field of material physical property testing technology. Background Technology

[0002] In the manufacturing of printed circuit boards and their assemblies, quality control generally relies on a combination of automated optical inspection (AOI) and electrical testing (ICT / FPT). The former focuses on the structural appearance, such as solder joint morphology and component placement, while the latter focuses on the immediate electrical performance, such as circuit continuity, open circuits, and short circuits. This combination plays a core role in ensuring the functional qualification of products upon leaving the factory. As the industry generally shifts towards high-density packaging, lead-free processes, and no-clean processes, the narrow gaps at the bottom of high-density components make it physically difficult to completely remove active residues such as flux. At the same time, the application of no-clean processes and higher reflow soldering temperatures make the physical properties of the substrate material itself, such as the degree of resin curing or porosity, sensitive to process fluctuations.

[0003] These residues or substrate pores, which cannot be detected by optical or electrical testing, manifest as an anomaly in the material's physical properties. The core characteristic is an alteration in the hygroscopicity or hydrophilicity of the material's surface and bulk phase. A clean, cured substrate exhibits hydrophobicity, while a substrate with the aforementioned defects displays the ability to physically adsorb moisture. This abnormal hygroscopicity, undetectable by electrical testing at the factory, accumulates ambient moisture in the actual service environment, activating residual ions and triggering electrochemical migration (ECM) or conductive anode filament (CAF) growth, leading to long-term failure. Existing optical testing methods cannot penetrate components or substrates to detect such physicochemical anomalies, while existing electrical testing methods can determine current electrical connectivity but cannot effectively assess the potential failure risk determined by physical adsorption capacity. To overcome the limitations of traditional optical testing, this field has already... Several improved solutions based on 3D imaging and data analysis have been developed, but the technical approach still mainly focuses on characterizing macroscopic or microscopic morphology. There are still shortcomings in detecting the above-mentioned abnormal physical and chemical properties. For example, Chinese invention patent CN120446164B discloses a PCB circuit board defect detection method and system. Although this solution constructs a 3D surface topology matrix through multispectral imaging and combines thermal stress gradient and layer recognition model to improve the ability to identify structural defects, it is essentially still a combination of optics and model analysis. This method still focuses on detecting morphological defects such as open circuits, voids or dielectric layer separation. The technical solution itself cannot characterize the key abnormal physical and chemical property of hygroscopicity caused by residues or porosity. In addition, when such high-precision optical 3D modeling systems are deployed on industrial production lines, the measurement accuracy is easily affected by environmental interference such as mechanical vibration, which leads to a decrease in detection reliability.

[0004] Therefore, how to provide a system that can quickly and quantitatively characterize the physical adsorption properties of PCB circuit board materials in order to assess their process consistency and potential long-term failure risks has become the technical problem to be solved by this invention. Summary of the Invention

[0005] This invention provides a PCB circuit board defect detection system, the main purpose of which is to solve the problem that the existing technology lacks the ability to quickly and quantitatively characterize the physical adsorption properties of PCB circuit board materials in order to assess their process consistency and potential long-term failure risks.

[0006] To achieve the above objectives, the present invention provides a PCB circuit board defect detection system, the system comprising:

[0007] The quality analysis unit is used to carry the PCB circuit board and measure its quality in real time to obtain the first measurement signal characterizing the quality.

[0008] A sealed test chamber coupled to the quality analysis unit;

[0009] An atmosphere control unit is used to controllably supply at least one dry baseline atmosphere and at least one analytical atmosphere to a sealed test chamber. The atmosphere control unit includes a flow control device for supplying a specific component in the analytical atmosphere and outputting a process control signal characterizing the supply rate of the specific component.

[0010] The data processing unit is configured to: Step a, determine the drying baseline mass of the PCB circuit board in a drying baseline atmosphere; Step b, after switching the atmosphere from the drying baseline atmosphere to the analysis atmosphere, monitor the first measurement signal in real time, and calculate the first mass increment curve based on the first measurement signal and the drying baseline mass; Step c, during the execution of Step b, acquire the process control signal and perform time integration on the process control signal to obtain the cumulative mass increment based on the flow rate, which serves as the second mass increment curve; Step d, monitor the noise level of the first measurement signal, and when the noise level exceeds a preset vibration threshold, determine that the mass analysis unit is subject to mechanical vibration interference; Step e, when determined to be uninterrupted, determine whether the PCB circuit board has abnormal physical and chemical properties based on the first mass increment curve; and when determined to be interfered with, determine whether the PCB circuit board has abnormal physical and chemical properties based on the second mass increment curve.

[0011] Preferably, before performing step a to determine the drying baseline mass, the data processing unit further performs the following: monitoring the mass change of the PCB circuit board within a preset time in a drying baseline atmosphere to determine the baseline mass drift rate; and step e of the data processing unit further includes: compensating the first mass increment curve based on the baseline mass drift rate to obtain a compensated first mass increment curve; when it is determined that there is no interference, the compensated first mass increment curve determines whether there is any abnormality in the physical and chemical properties of the PCB circuit board.

[0012] Preferably, the calculation rule used by the data processing unit to compensate the first mass increment curve based on the baseline mass drift rate is as follows: ,in This is the first mass increment curve after compensation. This is the first mass increment curve. The baseline mass drift rate, This represents the time since step b began.

[0013] Preferably, the data processing unit is further configured to: extract the kinetic characteristics of the first mass increment curve or the second mass increment curve within an initial time window to determine whether the abnormal physicochemical properties originate from surface adsorption; and extract the kinetic characteristics of the first mass increment curve or the second mass increment curve within a subsequent time window to determine whether the abnormal physicochemical properties originate from bulk diffusion.

[0014] Preferably, the dynamic characteristics include the initial slope of the first mass increment curve or the second mass increment curve within the initial time window, and the asymptotic slope of the first mass increment curve or the second mass increment curve within subsequent time windows.

[0015] Preferably, the atmosphere control unit is further configured to, after step b, switch the atmosphere from the analytical atmosphere back to the dry baseline atmosphere and execute the desorption process; the data processing unit is further configured to: in step f, monitor the mass change of the PCB circuit board in the desorption process in real time and calculate the mass decay curve; in step g, extract the adsorption kinetic characteristics of the first mass increment curve or the second mass increment curve and extract the desorption kinetic characteristics of the mass decay curve; in step h, determine whether there is an adsorption hysteresis effect in the abnormal physicochemical properties by comparing the adsorption kinetic characteristics and the desorption kinetic characteristics.

[0016] Preferably, the adsorption kinetics are characterized at the adsorption half-life, and the desorption kinetics are characterized at the desorption half-life; the data processing unit is further used to confirm that the abnormal physicochemical properties originate from the bulk pore structure by judging the existence of the adsorption hysteresis effect.

[0017] Preferably, the system includes a temperature control unit for setting the analysis temperature of the sealed test chamber; the data processing unit is further configured to: control the system to execute steps a to e at a first analysis temperature to obtain a first discrimination result characterizing the abnormal physicochemical properties; control the system to repeat steps a to e at a second analysis temperature different from the first analysis temperature to obtain a second discrimination result characterizing the abnormal physicochemical properties; and, based on the comparison of the first discrimination result and the second discrimination result, determine the thermodynamic characteristics of the abnormal physicochemical properties.

[0018] Preferably, the system further includes: a non-contact temperature sensor for monitoring the surface temperature of the PCB circuit board; a data processing unit for further monitoring the transient changes in surface temperature caused by the adsorption and exothermic reaction of the analytical atmosphere during step b in step f; and step g for verifying the discrimination result based on the first mass increment curve or the second mass increment curve in step e based on the transient changes in surface temperature, or using it as a substitute for discrimination when the first measurement signal is interfered with.

[0019] Preferably, the analytical atmosphere is a constant humidity atmosphere with a preset relative humidity.

[0020] Compared with the prior art, the beneficial effects of the present invention are:

[0021] 1. Through the collaborative work of the quality analysis unit, atmosphere control unit and data processing unit, a quality response curve of the circuit board under test is established under controlled atmosphere switching. The physical characteristics of this curve, such as the mass increment or mass change rate within a specific time window, are used to directly characterize the physical adsorption properties of the circuit board material surface and interior. This physicochemical property is directly related to the potential long-term failure risk, and this property cannot be measured by traditional optical detection or electrical testing methods.

[0022] 2. The data processing unit not only analyzes the total amount of adsorption, but also the dynamic shape of the mass increment curve. By analyzing the instantaneous change characteristics of the curve within the initial time window and the gradual change characteristics within subsequent time windows, it distinguishes physical adsorption processes from different sources, such as rapid surface adsorption and slow bulk diffusion. This distinction enables the system to provide diagnostic information about the causes of defects, such as distinguishing between surface residue problems and substrate bulk porosity problems, providing more in-depth guidance for process control.

[0023] 3. This system can also be configured to execute a desorption process. After switching the atmosphere to the analytical atmosphere and recording the adsorption process, the atmosphere is switched back to the dry baseline atmosphere. The data processing unit determines whether there is an adsorption hysteresis effect by comparing the mass increment kinetic characteristics of the adsorption stage with the mass decay kinetic characteristics of the desorption stage. Attached Figure Description

[0024] Figure 1 This is a data processing and interference arbitration logic diagram of a PCB defect detection system according to the present invention;

[0025] Figure 2 This is a comparison curve of the dynamic characteristics of the quality increment of PCBs with different defect types according to the present invention;

[0026] Figure 3 This is a system interaction timing diagram of the defect identification process of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. The embodiments are only some embodiments of this invention, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0028] This invention provides a PCB circuit board defect detection system, including a quality analysis unit, a sealed test chamber, an atmosphere control unit, and a data processing unit. The data processing unit serves as the control and analysis core of the system, coordinating the operation of other units. The atmosphere control unit is used to establish and switch the environmental atmosphere of the PCB circuit board within the sealed test chamber. The quality analysis unit is used to measure the mass change of the PCB circuit board in real time during the atmosphere switching process. The data processing unit collects and analyzes the measurement signals provided by the quality analysis unit and the atmosphere control unit to determine whether the PCB circuit board has abnormal physical adsorption characteristics caused by factors such as residues or internal pores. In a specific embodiment, the quality analysis unit can employ a high-sensitivity mass measurement device, such as a microbalance based on the electromagnetic force compensation principle or a quartz crystal microbalance. This unit has the sensitivity to capture instantaneous mass changes at the microgram or even nanogram level. The sample rate is provided with a tooling fixture for carrying the PCB circuit board under test, which is wholly or partially housed in a sealed test chamber. The chamber is airtight and has a gas inlet and outlet connected to an atmosphere control unit. The atmosphere control unit is used to supply at least two atmospheres according to program instructions: one is a dry baseline atmosphere, which can be dry nitrogen or dry air with a relative humidity of less than 1%RH; the other is an analytical atmosphere, which is a constant humidity atmosphere with a preset relative humidity, such as humid air or nitrogen with 85%RH. This unit is used to precisely control the mass flow controller for the carrier gas flow rate and a steam generator for generating water vapor, which can be a temperature-controlled bubbler or a direct evaporation source. The atmosphere control unit includes at least a flow control device for supplying specific components in the analytical atmosphere. In the closed-loop control that maintains constant humidity in the chamber, the output process control signal represents the instantaneous supply rate that needs to be replenished to compensate for the water vapor consumed by the PCB adsorption.

[0029] The data processing unit, which can be an embedded microprocessor or an external computer, is connected to the quality analysis unit and the atmosphere control unit to execute the following analysis process: The data processing unit performs step a, determining the dry baseline mass; before performing this step, to eliminate interference from sample volatilization or instrument drift on subsequent measurements, the data processing unit can perform a baseline diagnostic process, introducing a dry baseline atmosphere into the system, monitoring the mass change of the PCB circuit board within a preset time, such as 300 seconds, performing linear fitting on the mass data within this time window, and determining the baseline mass drift rate. ,Should The value is stored, and at the end of the preset time, the mass reading or rate of change of mass falls below a certain stable threshold, such as 0.1. The mass reading at g / min is determined as the drying baseline mass; the data processing unit executes step b, controlling the atmosphere control unit to quickly switch the atmosphere from the drying baseline atmosphere to the analytical atmosphere, and performs real-time high-frequency monitoring during the analysis phase after the switch, for example, within 600 seconds, with a sampling rate of 1Hz. The first measurement signal output by the mass analysis unit is... Based on this signal and the existing drying baseline quality, the first mass increment curve is calculated. During the entire analysis phase of step b, the data processing unit, according to step c, acquires process control signals from the flow control device in the atmosphere control unit at a high frequency, with a sampling rate of up to 1 Hz. These signals characterize the instantaneous replenishment rate of water vapor. The data processing unit integrates this rate signal over time to obtain the cumulative mass increment based on the flow rate, which serves as the second mass increment curve. This curve constitutes an independent characterization of the same physical adsorption event based on fluid measurement principles. To establish the measurement benchmark for the second mass increment curve, the data processing unit further performs a system pseudo-adsorption calibration process before online detection. The process includes: placing an inert standard piece, confirmed to have no adsorption properties and with a thermal mass similar to the PCB circuit board under test, on the mass analysis unit; executing steps a to c in their entirety; after atmosphere switching, acquiring and integrating the process control signal to obtain a system pseudo-adsorption curve characterizing adsorption or minor leakage on the wall of the sealed test chamber; storing the system pseudo-adsorption curve in the data processing unit; and subtracting the system pseudo-adsorption curve from the obtained second mass increment curve in the subsequent step e of testing the PCB circuit board to obtain a compensated second mass increment curve; and identifying physicochemical property anomalies based on the compensated second mass increment curve.

[0030] During data acquisition, the data processing unit performs real-time signal quality monitoring according to step d, specifically by continuously calculating the first measurement signal. Within a short sliding window, such as 1 second, the variance or standard deviation of the high-frequency components is used as the noise level characterizing mechanical vibration. This noise level is compared with a preset vibration threshold. The preset vibration threshold is determined by multiplying the maximum mass signal noise standard deviation measured during the system calibration phase (using a non-adhesive inert standard to replace the PCB and activating the production line background vibration source) by a safety factor, such as 1.5. When the noise level exceeds the preset vibration threshold, the data processing unit determines that the quality analysis unit is interfered with by mechanical vibration. Finally, the data processing unit executes step e to perform defect identification. The identification criterion for this step is dynamically selected: when the system is determined to be undisturbed, it defaults to using the first mass increment curve based on quality measurement. As a criterion; specifically, to avoid the influence of baseline drift, the system may preferentially use the first mass increment curve calculated according to the rules after compensation. The calculation rules are as follows: ,in, This is the first mass increment curve after compensation. This is the first mass increment curve. The baseline mass drift rate was previously determined. The system is based on the time since step b. Determine if there are any abnormal physical or chemical properties on the PCB circuit board, and when it is determined to be due to interference, i.e. If the signal becomes unreliable, the system automatically switches to a second mass increment curve, based on the curve obtained from the flow integral, to determine whether there are any abnormalities in the physical and chemical properties of the PCB circuit board. Since the flow measurement device is far less sensitive to mechanical vibration than a high-precision balance, this orthogonal verification and substitution mechanism ensures the system's reliability in industrial environments. When determining abnormalities in physical and chemical properties, the data processing unit can further refine the selected mass increment curve according to regulations. (Or the second mass increment curve) for dynamic characteristic analysis, the system extracts the curve within the initial time window, with Instant The initial slope, taken as an example, characterizes the degree of rapid surface adsorption; and the curve is extracted for subsequent time windows, with... Instant The asymptotic slope, taking seconds as an example, characterizes the degree of slow bulk diffusion. By comparing these kinetic characteristics with the benchmark of a standard good product, the system can not only determine the presence or absence of defects, but also provide diagnostic information on the physical sources of defects, such as surface residues or bulk porosity.

[0031] To further confirm the existence of the bulk pore structure, the system can execute the desorption process according to regulations. After the adsorption process in step b is completed, the atmosphere control unit switches back to the dry baseline atmosphere, and the data processing unit monitors the mass decay curve in real time; the kinetic characteristics of the adsorption stage are extracted, which can be used to determine the adsorption half-life. Extracting the kinetic characteristics of the desorption phase (i.e., the time to reach 50% of the total adsorption capacity) can provide insights into the desorption half-life. (That is, the time it takes for the mass to decay by 50% from the saturation state); by comparing the two, if it is found that... Greater than If this is detected, an adsorption hysteresis effect is identified. This phenomenon is a physical characteristic of capillary condensation and is used to confirm that the abnormal physicochemical properties originate from the bulk pore structure. Furthermore, the system includes a temperature control unit, which can be a Peltier temperature controller or a circulating water bath, used to set the analytical temperature of the sealed test chamber. The data processing unit can control the system at the first analytical temperature. For example, 25 The system executes steps a to e to obtain the first discrimination result, and controls the system at the second analysis temperature. For example, 40 The process repeats steps a to e to obtain a second discrimination result. By comparing the adsorption amount or adsorption rate at different temperatures, the thermodynamic characteristics of the adsorption process can be determined, providing a basis for identifying the chemical nature of defects or residues. Simultaneously, the system includes a non-contact temperature sensor, which can be an infrared thermopile sensor. Its field of view is focused on the PCB surface to monitor the surface temperature of the PCB circuit board. Since physical adsorption is an exothermic process, a transient temperature rise occurs the instant the analytical atmosphere contacts the dry PCB at the beginning of step b. The data processing unit simultaneously monitors this transient change in surface temperature. Based on the amplitude or integral area of ​​this transient temperature rise peak, the amplitude... The value or integral area is related to the initial adsorption rate. Cross-validation is performed on the discrimination results based on the mass increment curve. Alternatively, when the first measurement signal (i.e., the mass signal) is subjected to severe mechanical vibration interference, this thermal signal is used as a substitute for discrimination to enhance the system's anti-interference capability. In embodiments where the system includes a non-contact temperature sensor, to ensure the accuracy of the transient change measurement of surface temperature, the data processing unit further performs a sensor calibration and field-of-view alignment procedure before executing the analysis process. This procedure includes: using the temperature control unit to stabilize the sealed test chamber and a standard PCB sample with known infrared emissivity together at a first calibration temperature. The data processing unit reads the measured value from the non-contact temperature sensor and adjusts its internal emissivity compensation coefficient until the measured value matches the first calibrated temperature. The deviation is less than the preset calibration tolerance; the temperature control unit stabilizes the standard PCB sample to the second calibration temperature. The measurement value of the non-contact temperature sensor can be verified to accurately track the temperature change, thus confirming the effectiveness of emissivity compensation. Furthermore, by aligning the sensor's measurement spot with a calibration target smaller than the PCB board and monitoring the attenuation of the measurement signal when the target moves out of the measurement spot, the sensor's field of view is confirmed to be limited to the effective surface area of ​​the PCB board.

[0032] Example 1: In a high-density PCBA automated production line environment, the PCB defect detection system of the present invention is deployed after the cleaning and drying process to evaluate the residual level of no-clean flux online. The challenge of this deployment location is that the system's quality analysis unit is adjacent to a high-speed pick-and-place machine. The operation of the pick-and-place machine generates continuous and strong mechanical vibrations, making it impossible to obtain effective measurement signals using the measurement methods of the quality analysis unit in such an environment. When a freshly dried PCB to be tested is sent into the system's sealed test chamber and placed on the quality analysis unit, the data processing unit executes step a, controlling the atmosphere control unit to introduce a dry baseline atmosphere. During this period, the data processing unit performs baseline diagnostics and monitors -0.05... Baseline mass drift rate in g / sec This rate is caused by the evaporation of trace amounts of solvent remaining in the PCB substrate, and the system records this. The data processing unit executes steps b and c, and the atmosphere control unit switches to an analytical atmosphere with a relative humidity of 85%. The data processing unit begins parallel monitoring of the first measurement signal from the quality analysis unit and acquires the process control signal from the flow control device. Approximately 10 seconds after the start of the test, the adjacent pick-and-place machine begins high-frequency vibration. When the data processing unit performs noise monitoring in step d, it finds that the noise level of the first measurement signal momentarily exceeds the preset vibration threshold. At this time, the first mass increment curve is distorted due to noise. Based on the logic of step e, the system automatically determines that it is in a state of interference and uses the second mass increment curve obtained by time integration of the process control signal as the basis for judgment. This second mass increment curve is not affected by noise. Mechanical vibration affects the adsorption kinetics. Based on this curve, the data processing unit calculates the total mass increment within a 60-second time window as 12.8 μg. This value exceeds the 8.0 μg threshold of the good product benchmark. The system determines that the PCB circuit board has abnormal physicochemical properties. In this scenario, in addition to supplying the analytical atmosphere for analysis, the flow control device of the atmosphere control unit also provides a measurement dimension independent of mechanical vibration through the output process control signal. The data processing unit arbitrates between the precision measurement that relies on the first measurement signal and the production line reliability that relies on the second measurement signal through noise monitoring in step d and dynamic switching logic in step e. This preserves the material physical property analysis capability while solving the limitation that high-precision mass measurement units cannot work reliably in industrial vibration environments.

[0033] Example 2: This example verifies the reliability of the system of the present invention in dealing with mechanical vibration interference during material physical property testing. Specifically, it verifies the synergistic effect of the data processing unit monitoring the noise level of the first measurement signal and making judgments based on the second mass increment curve. The experiment uses the PCB circuit board defect detection system of the present invention, which is additionally connected to a controllable mechanical vibration table. This vibration table is used to simulate mechanical vibration in an industrial environment and is synchronously controlled by the data processing unit. The mass analysis unit used in the experiment is a unit with an accuracy of 0.1. A microbalance of g is used. The flow control device in the atmosphere control unit is used to generate a constant humidity atmosphere. The process control signal is acquired by the data processing unit to calculate the second mass increment curve. The preset vibration threshold in the data processing unit is determined to be 1.0 through calibration procedures. Two sets of samples were prepared for the experiment: one set consisted of good PCB samples, which underwent rigorous cleaning and drying to ensure no organic residue remained. These samples exhibited hydrophobic physical adsorption characteristics and were expected to have low moisture absorption under the analytical atmosphere. The other set consisted of defective PCB samples, which were prepared by quantitatively depositing 5.0 μL of the material into the gap at the bottom of the BGA devices using a micro-injection pump, based on the good PCB samples. The sample contains a residue of a commercial no-clean flux, which is hygroscopic. The test procedure is as follows: For the two groups of samples, the test procedure of this invention is performed under both low and high vibration conditions, i.e., steps a to e are performed, with a total test time of 120 seconds. Under low vibration conditions, the amplitude of the vibration table is set so that the noise level (standard deviation) of the first measurement signal of the quality analysis unit is maintained between 0.2 μg and 0.3 μg, which is lower than the preset vibration threshold of 1.0 μg. Under high vibration conditions, the amplitude of the vibration table is set so that the noise level (standard deviation) of the first measurement signal increases to above 5.0 μg, which is higher than the preset vibration threshold of 1.0 μg. The data processing unit records two comparison results: the result without switching logic (i.e., regardless of whether there is interference, only the reading of the first measurement signal is used) and the result of the system of this invention. The test data are summarized in Table 1.

[0034] Table 1: Comparison of Measurement Data of the System under Different Vibration Conditions

[0035]

[0036] Referring to Table 1, the data analysis is as follows: Under low vibration conditions, the noise level of the first measurement signal is lower than the preset vibration threshold. At this time, the data processing unit of the present invention determines that it is not interfered with, and its output (1.8 μg and 12.4 μg) is consistent with the output of the system without switching logic, and matches the readings of the second mass increment curve (1.9 μg and 12.6 μg), indicating accurate measurement results. Under high vibration conditions, the first measurement signal (15.3 μg and 28.7 μg) is distorted due to vibration interference, and deviates from the true values ​​of physical adsorption (2.0 μg and 12.5 μg measured by the second mass increment curve). At this time, the system without switching logic, lacking an arbitration mechanism, outputs incorrect measurement results (15.3 μg and 28.7 μg). Relatively speaking, According to step d, the data processing unit of the present invention detects that the noise level (5.5 μg and 5.8 μg) has exceeded the preset vibration threshold of 1.0 μg, and therefore determines that it is interfered with. According to the logic of step e, it automatically switches and uses the readings of the second mass increment curve (2.0 μg and 12.5 μg) as the final output. The experimental results show that the present invention, by introducing the second mass increment curve based on flow rate and orthogonally verifying it with the first measurement signal based on mass, and utilizing the noise monitoring and discrimination switching logic of the data processing unit (steps d and e), can use the second mass increment curve for discrimination when the first measurement signal fails due to mechanical vibration interference. Thus, even in industrial vibration environments, it can still obtain accurate characterization of material physical properties and ensure the reliability of detection.

[0037] Example 3: This example combines Figures 1 to 3 A description of a PCB circuit board defect detection system, such as... Figure 1 As shown, the quality analysis unit provides a first measurement signal, and the atmosphere control unit provides a process control signal. The first measurement signal is used to determine the baseline to identify D1: the dry baseline mass and D2: the baseline mass drift rate. The first measurement signal and the process control signal are used to generate incremental curves, calculating the first mass increment curve and the second mass increment curve respectively. The first measurement signal is monitored for vibration interference to output an interference discrimination signal. In the arbitration and defect discrimination stage, the system integrates D1: the dry baseline mass, D2: the baseline mass drift rate, the first mass increment curve, the second mass increment curve, and the interference discrimination signal. When interference is identified, discrimination is preferably based on the second mass increment curve. When no interference is identified, discrimination is based on the first mass increment curve, which can be compensated by the baseline mass drift rate. Finally, the system outputs the physicochemical property anomaly discrimination result or the discrimination result.

[0038] like Figure 2As shown, the horizontal axis represents time (s), and the vertical axis represents mass increment (μg). The figure illustrates three PCB samples with different characteristics: the good PCB exhibits the lowest mass increment throughout the entire 120s test cycle; the PCB with surface defects shows the largest mass increment slope in the initial stage, such as 0-10s, quickly reaching the adsorption plateau; while the PCB with bulk defects adsorbs more slowly in the initial stage, but its mass increment continues to increase and surpasses that of the PCB with surface defects after a subsequent time, such as 60s, exhibiting bulk diffusion characteristics. Figure 3 As shown, the process involves four main components: a data processing unit, a quality analysis unit, an atmosphere control unit, and a sealed test chamber. In step a: determining the drying baseline quality, the data processing unit instructs the atmosphere control unit to introduce the drying baseline atmosphere and monitors the PCB quality from the quality analysis unit to store the drying baseline quality. In step b: switching to the analysis atmosphere, the data processing unit instructs the atmosphere control unit to switch and supply a constant humidity atmosphere, and the system enters the real-time monitoring stage. The data processing unit cyclically acquires the real-time quality data of the first measurement signal and calculates the first quality increment curve. Finally, in step e: identifying abnormal physicochemical properties, the data processing unit makes a judgment based on the first quality increment curve and outputs the detection result.

[0039] Example 4: This example illustrates the specific procedure for the data processing unit to perform kinetic characteristic analysis. This procedure is used to further distinguish whether the physical source of the anomaly is surface adsorption or bulk diffusion after identifying the physicochemical property anomaly. The experiment uses the PCB circuit board defect detection system of this invention. This system operates in a stable environment without mechanical vibration. Therefore, the data processing unit is set to always use the first mass increment curve as the discrimination criterion. Three characteristic sample groups were prepared for the experiment: Sample group A, surface defect type sample group, taking a PCB substrate that has been confirmed to be qualified, and uniformly adsorbing 15.0 on its surface through atomization deposition. Sample A, containing no-clean flux residue, exhibits defects primarily characterized by surface adsorption. Sample B, a bulk defect type, consists of a PCB substrate with incomplete curing and micropores during the lamination process. This substrate underwent plasma surface cleaning before testing to remove surface residues. The main defect in this sample group is its bulk diffusion characteristics. Sample C, serving as a benchmark, is a PCB substrate from the same batch as Samples A and B, but with a qualified process and rigorous cleaning. All three sample groups were sequentially placed into the system's sealed test chamber and subjected to the same analytical procedure: Step a, stabilization to the dry baseline quality in a <1%RH dry baseline atmosphere; Step b, in… The analysis atmosphere is constantly switched to 85%RH and monitored continuously for 120 seconds. The data processing unit records the first mass increment curve of this process. Based on logic, the data processing unit performs kinetic feature extraction on the acquired first mass increment curve. The specific algorithmic procedure is as follows: First, the initial time window is set to... Instant Calculate the average slope of the mass increment curve within this window, using it as the initial slope characterizing the surface adsorption rate. Second, set the subsequent time window as Instant Calculate the average slope of the mass increment curve within this window, using it as the asymptotic slope characterizing the bulk diffusion rate. Based on the test data of sample group C, i.e. =0.22 g / s, =0.04 g / s, the threshold for determining the cause of defects in the system: =0.5 g / s; =0.1 g / s, the built-in discrimination logic of the data processing unit is: if > and < It is determined to be a surface residual defect; if < and > It is determined to be a bulk porosity defect; if < and < If the result is positive, it is considered qualified. The results of dynamic feature extraction and discrimination for sample group A and sample group B are summarized in Table 2.

[0040] Table 2: Dynamic Feature Extraction and Defect Cause Judgment Table

[0041]

[0042] Referring to Table 2, the first mass increment curve of sample group A shows a value of 2.18 within the initial time window. high g / s Value, in subsequent time windows Value (0.06) The g / s ratio is below the good product level, which is consistent with the physical characteristic of rapid saturation of surface adsorption. Based on the discrimination logic, it is identified as a surface residual defect; Sample B Value (0.25) (g / s) is close to that of good quality, and in the subsequent time window Value (0.72) If the mass increment curve (g / s) is higher than the threshold, it is consistent with the physical characteristics of slow and continuous bulk diffusion, and is identified as a bulk porosity defect. The results of this embodiment show that by quantitatively analyzing the initial slope and the progressive slope of the mass increment curve, the physical source of abnormal physicochemical properties can be distinguished, providing a diagnostic basis for subsequent process tracing, such as cleaning or lamination processes.

[0043] Example 5: This example illustrates the specific procedure by which the data processing unit executes the desorption process and compares the adsorption and desorption kinetics to confirm bulk porosity defects. Building upon the experimental background, in Example 4, the data processing unit had already used a gradual slope... Analysis initially determined that sample group B (bulk pores) exhibited abnormal physicochemical properties. To further confirm, using physical principles, that this abnormality originated from the bulk pore structure, a supplementary verification process for the adsorption hysteresis effect was initiated for sample group A (surface residue) and sample group B (bulk pores). At step b (adsorption for 120 seconds), the adsorption plateau was reached, and the mass increment at this point was recorded as [value missing]. Then execution begins, and the data processing unit calculates the mass increment reached based on the first mass increment curve of the adsorption stage. The required time is denoted as the adsorption half-life. The data processing unit executes step f, controlling the atmosphere control unit to switch the atmosphere back to the drying baseline atmosphere and monitoring the mass decay curve in real time until the mass returns to the drying baseline mass. The data processing unit then calculates the mass decay from... decay to The time required is denoted as the desorption half-life. The system's built-in hysteresis detection threshold is: if If the adsorption hysteresis effect is present, it is determined that the adsorption hysteresis effect exists; the experiment measured that: sample group A It lasted 11.2 seconds. The time was 13.5 seconds, and the ratio was 1.2, so the system determined it to be hysteresis-free; Sample B... It lasted 28.5 seconds. The duration was 215.8 seconds, with a ratio of 7.6; given that sample group B... The ratio (7.6) is greater than 2.0. The data processing unit determines the existence of adsorption hysteresis effect based on step h, and confirms that the abnormal physicochemical properties of the PCB circuit board are caused by the bulk pore structure based on this physical characterization.

[0044] Example 6: This example illustrates a standardized engineering calibration procedure used by the data processing unit to determine key discrimination parameters before performing analysis. This procedure aims to eliminate inconsistencies in measurement benchmarks caused by differences in PCB board type, material batch, or production line environment. Specifically, it includes the determination of dynamic characteristic thresholds (…). and The system includes a time window and calibration of the preset vibration threshold. When the system is first deployed or the PCB board type under test is changed, dynamic calibration is performed. This procedure requires the preparation of three types of characteristic samples: sample group A (surface defect type), sample group B (bulk defect type), and sample group C (good product benchmark). The data processing unit sequentially performs the analysis process on these three types of samples under vibration-free conditions, recording their complete first mass increment curves. By analyzing the curve shape, the initial time window is determined (within...). Instant (e.g., seconds) and subsequent time windows (e.g., seconds) Instant (Taking seconds as an example), the system automatically calculates and stores the initial slope of each template. (respectively) , , and asymptotic slope (respectively) , , The data processing unit, based on its built-in algorithm, performs... and Set surface defect discrimination thresholds between and in and Set a threshold for bulk defect discrimination. A feasible algorithm is to set the threshold at the average measurement value between the good product benchmark and the corresponding defective sample, for example... The result and Values ​​(such as 0.5 in Example 4) g / s and 0.1 (g / s) is fixed as the criterion for judging this type of PCB.

[0045] The system performs vibration threshold calibration to determine the preset vibration threshold in step d. During operation, a metal standard component with inert characteristics and a mass similar to the PCB under test is placed on the mass analysis unit. All foreseeable external mechanical vibration sources (such as nearby equipment) are turned off. The data processing unit executes a complete test procedure (including atmosphere switching). During this period, the first measurement signal is continuously monitored, and its noise level baseline throughout the entire process is calculated and recorded as the baseline standard deviation. (e.g. 0.2) g) Next, turn on all external mechanical vibration sources to their typical operating conditions, repeat the above test procedure, calculate the noise level at this time, and record it as the standard deviation of interference. (e.g. 5.5) g) The data processing unit sets the preset vibration threshold to a value between the two that can effectively distinguish the disturbance state, for example, it can be set to... Five times (i.e., 1.0) (g) This threshold is stored for online noise monitoring in step d. At this point, both the dynamic parameters and the vibration threshold have been calibrated, and the system has the ability to make online judgments.

[0046] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0047] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A PCB circuit board defect detection system, characterized in that, The system includes: The quality analysis unit is used to carry the PCB circuit board and measure its quality in real time to obtain the first measurement signal characterizing the quality. A sealed test chamber coupled to the quality analysis unit; An atmosphere control unit is used to controllably supply at least one dry baseline atmosphere and at least one analytical atmosphere to a sealed test chamber. The atmosphere control unit includes a flow control device for supplying a specific component in the analytical atmosphere and outputting a process control signal characterizing the supply rate of the specific component. The data processing unit is configured to: Step a, determine the drying baseline mass of the PCB circuit board in a drying baseline atmosphere; Step b, after switching the atmosphere from the drying baseline atmosphere to the analysis atmosphere, monitor the first measurement signal in real time, and calculate the first mass increment curve based on the first measurement signal and the drying baseline mass; Step c, during the execution of Step b, acquire the process control signal and perform time integration on the process control signal to obtain the cumulative mass increment based on the flow rate, which serves as the second mass increment curve; Step d, monitor the noise level of the first measurement signal, and when the noise level exceeds a preset vibration threshold, determine that the mass analysis unit is subject to mechanical vibration interference; Step e, when determined to be uninterrupted, determine whether the PCB circuit board has abnormal physical and chemical properties based on the first mass increment curve; and when determined to be interfered with, determine whether the PCB circuit board has abnormal physical and chemical properties based on the second mass increment curve. The data processing unit is further used to: extract the kinetic characteristics of the first mass increment curve or the second mass increment curve within the initial time window to determine whether the abnormal physicochemical properties originate from surface adsorption; and extract the kinetic characteristics of the first mass increment curve or the second mass increment curve within the subsequent time window to determine whether the abnormal physicochemical properties originate from bulk diffusion. The dynamic characteristics include the initial slope of the first mass increment curve or the second mass increment curve within the initial time window, and the asymptotic slope of the first mass increment curve or the second mass increment curve within subsequent time windows. Furthermore, the atmosphere control unit includes at least a flow control device for supplying specific components in the analytical atmosphere. In closed-loop control that maintains constant humidity within the chamber, the output process control signal represents the instantaneous supply rate required to compensate for the water vapor consumed by PCB adsorption.

2. The PCB circuit board defect detection system according to claim 1, characterized in that, Before performing step a to determine the drying baseline mass, the data processing unit further performs the following steps: monitoring the mass change of the PCB circuit board within a preset time in a drying baseline atmosphere to determine the baseline mass drift rate; and step e of the data processing unit further includes: compensating the first mass increment curve based on the baseline mass drift rate to obtain a compensated first mass increment curve; when it is determined that there is no interference, the compensated first mass increment curve determines whether there is any abnormality in the physical and chemical properties of the PCB circuit board.

3. The PCB circuit board defect detection system according to claim 2, characterized in that, The calculation rule used by the data processing unit to compensate for the first mass increment curve based on the baseline mass drift rate is as follows: ,in, This is the first mass increment curve after compensation. This is the first mass increment curve. The baseline mass drift rate, This represents the time since step b began.

4. The PCB circuit board defect detection system according to claim 1, characterized in that, The atmosphere control unit is also used to switch the atmosphere from the analytical atmosphere back to the dry baseline atmosphere after step b, and execute the desorption process; the data processing unit is also used to: in step f, monitor the mass change of the PCB circuit board in real time during the desorption process and calculate the mass decay curve; in step g, extract the adsorption kinetic characteristics of the first mass increment curve or the second mass increment curve, and extract the desorption kinetic characteristics of the mass decay curve; in step h, determine whether there is an adsorption hysteresis effect in the abnormal physicochemical properties by comparing the adsorption kinetic characteristics and the desorption kinetic characteristics.

5. A PCB circuit board defect detection system according to claim 1, characterized in that, The adsorption kinetics are characterized by the adsorption half-life, and the desorption kinetics are characterized by the desorption half-life. The data processing unit is further used to confirm that the abnormal physicochemical properties originate from the bulk pore structure by judging the existence of the adsorption hysteresis effect.

6. The PCB circuit board defect detection system according to claim 1, characterized in that, The system includes a temperature control unit for setting the analysis temperature of the sealed test chamber; the data processing unit is further used to: control the system to execute steps a to e at the first analysis temperature to obtain a first discrimination result characterizing the abnormal physicochemical properties; and control the system to repeat steps a to e at a second analysis temperature different from the first analysis temperature to obtain a second discrimination result characterizing the abnormal physicochemical properties. Based on the comparison of the first and second discrimination results, the thermodynamic properties of abnormal physicochemical properties are determined.

7. A PCB circuit board defect detection system according to claim 1, characterized in that, The system also includes: a non-contact temperature sensor for monitoring the surface temperature of the PCB circuit board; a data processing unit for further monitoring the transient changes in surface temperature caused by the adsorption and exothermic reaction of the analytical atmosphere during step b in step f; and step g for verifying the discrimination result based on the first mass increment curve or the second mass increment curve in step e based on the transient changes in surface temperature, or using it as a substitute for discrimination when the first measurement signal is interfered with.

8. The PCB circuit board defect detection system according to claim 1, characterized in that, The analytical atmosphere is a constant humidity atmosphere with a preset relative humidity.

Images