Measuring machine and method for detecting defects in welded joints

Abstract

An exemplary embodiment relates to an inspection method for training a measuring machine to accurately measure the side joint length and detect defects among a plurality of welded joints. The method includes receiving first data and second data, the first data representing the side joint lengths of the plurality of welded joints measured by a first measuring machine, and the second data representing the side joint lengths measured by a second measuring machine. Further, the method includes determining a correlation value based on a statistical analysis of the relationship between the first data and the second data. The method further includes updating an algorithm used by the first measuring machine to measure the side joint length based on the correlation value to reduce the deviation between the first data and the second data. Subsequently, the updated algorithm is used as a dimensional measurement in the first measuring machine for detecting defects in the welded joint.

Description

Technical Field

[0001] The present disclosure relates to the field of defect measurement, and more particularly to a measuring machine and method for detecting defects of a welding joint. Background Art

[0002] Electronic devices (e.g., computing devices) may include electronic components and electronic boards, wherein the electronic components are typically connected to the electronic boards by soldering. Typically, electronic components such as sockets include a plurality of solder leads that are connected to an electronic board such as a printed circuit board (PCB) for power and signal distribution therebetween. Generally, plated through hole technology or surface mount technology is used to connect the electronic components to the electronic boards. Summary of the invention

[0003] An inspection method is provided for training a measuring machine to accurately measure the side joint length and detect defects among a plurality of welded joints. The method includes receiving first data and second data, the first data representing the side joint lengths of the plurality of welded joints measured by a first measuring machine, and the second data representing the side joint lengths measured by a second measuring machine. Further, the method includes determining a correlation value based on a statistical analysis of the relationship between the first data and the second data. The method further includes updating an algorithm used by the first measuring machine to measure the side joint length based on the correlation value to reduce the deviation between the first data and the second data. Subsequently, the updated algorithm is used as a dimensional measurement in the first measuring machine for detecting defects in the welded joint. BRIEF DESCRIPTION OF THE DRAWINGS

[0004] Various examples will be described below with reference to the following drawings.

[0005] Figure 1A A perspective view of a portion of an electronic device having electronic components and an electronic board according to an exemplary embodiment of the present disclosure is illustrated.

[0006] Figure 1B The diagram shows an exemplary embodiment according to the present disclosure. Figure 1A A cross-sectional side view of a portion of an electronic device taken along line 1 - 1 ′ in FIG.

[0007] Figure 2A is a block diagram depicting a cross-sectional front view of a portion of an electronic device according to an exemplary embodiment of the present disclosure.

[0008] Figure 2B is a diagram illustrating an exemplary embodiment according to the present disclosure Figure 2A Block diagram of a cross-sectional side view of a solder joint in an electronic device.

[0009] Figure 2Cis a diagram illustrating an exemplary embodiment according to the present disclosure Figure 2A A block diagram of a cross-sectional side view of another solder joint in an electronic device.

[0010] Figure 3A A perspective side view of a weld joint according to an exemplary embodiment of the present disclosure is illustrated.

[0011] Figure 3B A perspective front view of a weld joint according to an exemplary embodiment of the present disclosure is illustrated.

[0012] Figure 4 is a flow chart depicting a method of training a measuring machine for measuring a side joint length of a weld joint according to an exemplary embodiment of the present disclosure.

[0013] Figure 5A A graph representing a correlation based on a statistical analysis of a relationship between a set of first data and second data respectively obtained from a first measuring machine and a second measuring machine according to an exemplary embodiment of the present disclosure is illustrated.

[0014] Figure 5B A graph representing a new correlation based on a statistical analysis of a relationship between another set of first data and second data respectively obtained from a first measuring machine and a second measuring machine according to an exemplary embodiment of the present disclosure is illustrated.

[0015] Figure 6 is a flow chart depicting a method of detecting defects of a weld joint based on a measured side joint length of the weld joint by using a measuring machine according to an exemplary embodiment of the present disclosure.

[0016] Figure 7 is a block diagram depicting a measurement machine having a processing resource (or processor) operably coupled to a machine-readable medium storing executable program instructions (algorithms) for detecting defects in weld joints according to an exemplary embodiment of the present disclosure.

[0017] In all drawings, the same reference numerals may represent similar but not necessarily identical elements. The index number "N" attached to some reference numerals indicates a plurality, and may not necessarily represent the same quantity for each reference numeral with such an index number "N". In addition, in the case where the reference numerals are mentioned with the index numbers elsewhere herein, the use of such reference numerals without the index numbers may collectively or individually refer to the corresponding plurality of elements. In another example, the index numbers "I", "M", etc. may be used in place of the index number N. The drawings are not necessarily proportional, and the sizes of some parts may be enlarged to more clearly illustrate the examples shown. In addition, the drawings provide examples and / or embodiments consistent with the description; however, the description is not limited to the examples and / or embodiments provided in the drawings. DETAILED DESCRIPTION

[0018] The following detailed description refers to the accompanying drawings. Where possible, the same reference numerals are used in the drawings and in the following description to refer to the same or similar parts. However, it should be clearly understood that the drawings are only for the purpose of illustration and description. Although several examples are described in this document, modifications, adaptations and other embodiments are possible. Therefore, the following specific embodiments do not limit the disclosed examples. On the contrary, the correct scope of the disclosed examples can be defined by the appended claims.

[0019] The terms used herein are only used for the purpose of describing examples, and are not intended to be restrictive. As used herein, the singular forms "a", "an" and "the" are intended to also include plural forms, unless the context clearly indicates otherwise. As used herein, the term "multiple" is defined as two or more. As used herein, the term "another" is defined as at least the second or more. As used herein, unless otherwise indicated, the term "connected" is defined as connected, whether directly connected without any intervening element or indirectly connected by means of at least one intervening element. Two elements can be mechanically connected, electrically connected or communicated through a communication channel, path, network or system. The term "and / or" used herein refers to and encompasses any and all possible combinations of one or more items in the associated enumerated items. It should also be understood that although the terms first, second, third, etc. can be used to describe various elements in this article, these elements should not be limited by these terms, because these terms are only used to distinguish one element from another element, unless otherwise specified or the context indicates otherwise. As used herein, the term "including" refers to including but not limited to. The term "based on" refers to "based at least in part on".

[0020] As used herein, the term "weld joint" may refer to a thermally and / or electrically conductive interconnection element formed between a weld lead of an electronic component and the surface of an electronic board. As used herein, the term "side joint" may refer to a portion of a weld joint extending along the length of a weld lead leg. As used herein, the term "side joint length" or "weld joint side length" may refer to the length of a weld joint measured along the length of a weld lead leg. As used herein, the term "dimensional metrology" may refer to the measurement science used in a measuring machine by calibrating (or training) a measuring machine to determine at least one of a physical dimension, a distance to an object, or a length of an object / element. The term "calibrated measuring machine" may refer to a measuring device with an updated algorithm for accurately measuring the side joint length of a weld joint and detecting defects in the weld joint based on the measured side joint length.

[0021] For the purpose of explaining the present disclosure, reference is made to FIGS. Figure 7 The components illustrated in the figure describe certain examples. However, the functionality of the components illustrated may overlap and may exist in a fewer or greater number of elements and components. In addition, all or part of the functionality of the components illustrated may be co-existing or distributed in several geographically dispersed locations. Furthermore, the disclosed examples may be implemented in a variety of environments and are not limited to the examples illustrated. In addition, in conjunction with FIGS. 5 to 6, Figure 7 The described operation sequence is an example and is not intended to be limiting. Additional or fewer operations or combinations of operations may be used or changed without departing from the scope of the disclosed examples. Therefore, the present disclosure merely describes examples of implementations, and many changes and modifications may be made to the described examples. Such modifications and variations are intended to be included within the scope of the present disclosure and are protected by the appended claims.

[0022] An electronic device includes electronic components and electronic boards, wherein the electronic components are typically connected to the electronic boards by soldering to form interconnected electronic parts of the electronic device. For example, the electronic device may be a computing device, and the electronic components may include integrated circuit chips, chip packages, chip capacitors, sockets, etc. Similarly, the electronic board may include a printed circuit board (PCB), a substrate, a circuit board, etc., and the interconnected electronic parts may include a printed circuit assembly (PCA), a main logic board (MLB) assembly, etc.

[0023] Interconnection technologies such as plated through hole technology can be used to connect electronic components to electronic boards to form interconnected electronic parts of electronic devices. For example, in plated through hole technology, multiple solder leads of electronic components can extend through multiple through holes in the electronic board, and then the multiple solder leads can be connected to the electronic board by soldering.

[0024] Due to the change of open standards of interconnection technology, surface mount technology can be used to connect electronic components to electronic boards instead of plated through hole technology. For example, in surface mount technology, electronic components can be directly mounted on the surface of the electronic board, and the plurality of solder leads can be connected to the surface of the electronic board by soldering. Surface mount technology can include the following steps: i) flux application, ii) die placement, iii) reflow soldering, and iv) bottom filling process for connecting the electronic components to the surface of the electronic board.

[0025] Sometimes, due to the warping of the electronic board during the heating and cooling cycle in the reflow soldering process of surface mount technology, defects may occur in the solder joints of the interconnected electronic parts. For example, a defect can be the necking of the solder joints between at least some of the solder leads of the electronic component and the electronic board. Typically, the necking of the solder joint may reduce (or reduce) the side joint length of the solder joint, and thus cause the premature rupture of the solder joint. Therefore, electronic equipment with such interconnected electronic parts may fail, resulting in unplanned downtime of one or more workloads carried by the electronic equipment, and the costs associated with replacement and / or maintenance. Therefore, a method of inspection is needed, which can detect defective solder joints (that is, the side joint length that does not meet the acceptable standards such as predefined side joint length thresholds) to separate defective interconnected electronic parts in a quality test environment.

[0026] One inspection method used in industry to detect defective welded joints is based on the judgment of the operator. For example, an image (e.g., obtained from an imaging machine) of a welded joint belonging to an interconnected electronic part is visually analyzed by an operator to identify defective welded joints. This is a non-destructive inspection method because it uses an image of the interconnected electronic part for visual inspection. However, this method relies on the operator's training and knowledge to identify defective welded joints. Therefore, non-destructive inspection methods can be interpreted as subjective inspection methods that lack objective metrics for identifying defective welded joints. Therefore, non-destructive inspection methods may allow defective welded joints to bypass the quality testing environment.

[0027] Another inspection method used in industry uses an optical microscope to detect defective solder joints. For example, at least one sample interconnect electronic part is damaged so that an optical microscope can be used to inspect the cross section of the solder joint in the sample interconnect electronic part to identify defective solder joints. This method is a destructive inspection method because it requires the interconnect electronic part to be damaged so that the solder joint is visible to measure the length of the solder joint using an optical microscope. Therefore, the method can provide an objective metric to identify defective solder joints and prevent defective solder joints from bypassing the quality testing environment. However, destructive inspection methods for detecting defective solder joints are not feasible inspection methods because this approach is unreliable due to sample selection and is costly because multiple interconnect electronic parts need to be damaged for each inspection. Therefore, a non-destructive, feasible inspection method is needed that provides the ability to objectively detect defective solder joints.

[0028] A feasible technical solution to the above-mentioned problem includes training a measuring machine (e.g., an X-ray machine) to accurately measure the side joint length of a weld joint, and detecting defective weld joints based on the measured side joint length. In other words, the measuring machine is first trained using training data to accurately measure the side joint length of an actual weld joint. In one or more examples, the training data represents the side joint length measured by another measuring machine (e.g., an optical microscope). For example, an optical microscope can analyze a cross-section of a weld joint to accurately measure the side joint length of a set of weld joints, thereby generating training data. Therefore, the training data can be used as a benchmark or reference data for training an X-ray machine to accurately measure the side joint length of a weld joint. Subsequently, the X-ray machine is used to detect defective weld joints based on a comparison of the measured side joint length of each weld joint with a predefined side joint length threshold.

[0029] In one or more examples, the X-ray machine is trained based on an analysis of a correlation value obtained by a statistical analysis of a relationship between first data obtained from the X-ray machine and second data obtained from the optical microscope. In some examples, training the X-ray machine may include updating an algorithm used by the X-ray machine to accurately measure the side joint length of the weld joint. In some examples, after analyzing the correction value, if it is determined that the correction value is outside a predefined value (i.e., equal to or lower than the predefined value), the algorithm used in the X-ray machine is updated based on the second data to reduce the deviation between the first data and the second data. In some other examples, after analyzing the correction value, if it is determined that the correction value is within a predefined value (i.e., higher than the predefined value), a new correction value is obtained based on a statistical analysis of the relationship between another set of first data and second data. In such an example, after analyzing the new correlation value, if it is determined that the new correction value is outside the predefined value, the algorithm used in the X-ray machine can be further updated based on the second data in another set of first data and second data to reduce the deviation between the first data and the second data. However, after analyzing the new correlation value, if it is determined that the new correction value is within the predefined value, the algorithm can be updated using the predefined side joint length threshold, thereby enabling the updated algorithm to detect defects in weld joints based on a comparison of the measured side joint length of each weld joint with the predefined side joint length threshold.

[0030] In one or more examples, an X-ray machine with an updated algorithm can be deployed in a quality testing environment for detecting defective weld joints. For example, the X-ray machine can non-destructively detect at least one quality feature of the weld joint and analyze the at least one quality feature to measure a side joint length of the weld joint. In such an example, the X-ray machine can determine whether to generate an alarm signal indicating a defect detected in the weld joint based on a comparison of the measured side joint length of the weld joint with a predefined side joint length threshold. In one or more examples, the predefined side joint length threshold is determined based on an analysis of raw material measurements of the leg length and lead width of the weld lead.

[0031] Therefore, the present disclosure describes an exemplary embodiment of an inspection method for i) training a measuring machine to accurately measure the side joint length of multiple welded joints, and ii) detecting defects among multiple welded joints based on the side joint length measured by the measuring machine. In one or more examples, the method includes receiving first data, the first data representing the side joint length of multiple welded joints measured by a first measuring machine, wherein the multiple welded joints are formed between an electronic component (particularly, such as a socket or a memory socket) and an electronic board (e.g., a circuit board) to form an interconnected electronic part of an electronic device. The method further includes receiving second data, the second data representing the side joint length of the multiple welded joints measured by a second measuring machine. Further, the method includes determining a correlation value based on a statistical analysis of the relationship between the first data and the second data. The method further includes training the first measuring machine by updating an algorithm used by the first measuring machine to measure the side joint length based on an analysis of the correlation value to reduce the deviation between the first data and the second data. In one or more examples, the algorithm can be implemented as a dimensional measurement in the first measuring machine to detect defects among the multiple welded joints.

[0032] Figure 1A A perspective view of an electronic device 100 having one or more electronic modules, such as an electronic component 102 and an electronic board 104, is illustrated. In some examples, the electronic device 100 may be a computing device, a workstation, a server, etc., the electronic component 102 may include an integrated circuit chip, a chip package, a chip capacitor, a socket, etc., and the electronic board 104 may include a printed circuit board (PCB), a substrate, a circuit board, etc. Figure 1A In the example illustrated in , electronic device 100 is computing device 100A, electronic component 102 is socket 102A, and electronic board 104 is circuit board 104A. In some examples, socket 102A may be a double data rate (DDRx) socket (or memory socket), and circuit board 104A may be a main logic board. In one or more examples, socket 102A is mounted on a surface of circuit board 104A and is connected to circuit board 104A (e.g., via soldering 106). Figure 1B ).

[0033] In addition, the electronic device 100 may include one or more other electronic modules (such as a processor (not shown)) connected to the memory system of the computing device 100A, and one or more power supplies (not shown) connected to the memory system via the circuit board 104A. In such an example, the socket 102A may be a portion of the memory system that stores data for the computing device 100A. The power supply may be electrically connected to the traces of the circuit board 104A, which supply power to the socket 102A. In some examples, the memory system may include a circuit card (not shown) connected to the circuit board 104A via the socket 102A. The circuit card may constitute a synchronous dynamic random access memory (SDRAM) module or a dual in-line memory module (DIMM module). In the illustrated example, the socket 102A is configured to receive a circuit card to form a memory system and orient the circuit card at right angles relative to the circuit board 104A. In some examples, the socket 102A and the circuit card are electrically connected to one or more data devices (not shown) for sending data thereto and / or receiving data therefrom. In such an example, the circuit card may store data generated by the data device and / or send the stored data to the data device.

[0034] The socket 102A includes a housing 108 having a base end 110 mounted to the circuit board 104A, and a mating end 112 generally opposite the base end 110 for mating with a circuit card. The housing 108 includes: a longitudinally extending slot 114 at the mating end 112 for receiving the circuit card; and a plurality of solder tails 116 at the base end 110 for connecting with the circuit board 104A through soldering 106. In some examples, the socket 102A further includes a pair of latches 118 disposed at opposite ends of the housing 108, wherein the pair of latches 118 are configured to bend outward to receive the circuit card and bend inward to retain the circuit card in the socket 102A. The plurality of solder tails 116 can transmit power and signals from the circuit board 104A to the circuit card, and vice versa. In some examples, the socket 102A has 288 solder leads 116 arranged in 2 rows that are arranged parallel to each other laterally 10 along the width of the housing 108. In other words, each row has 144 solder leads 116 arranged longitudinally 20 along the length of the housing 108. In such an example, the first solder leads "P1", "P2" and "P3" of the plurality of solder leads 116 in each row are arranged parallel to each other laterally 10 along the width of the housing 108. 145 " are arranged laterally opposite to each other, and the last welded lead "P 144 ”, “P 288” are also disposed laterally relative to each other. The term “laterally” as used herein may refer to a direction perpendicular to the direction along which the circuit card is inserted into the slot 114 of the socket 102A. Similarly, the term “longitudinally” may refer to a direction parallel to the direction along which the circuit card is inserted into the slot 114 of the socket 102A.

[0035] Figure 1B is along Figure 1A 1-1' in FIG. 1-1' is a cross-sectional side view of a portion of the electronic device 100. For example, the cross-sectional side view is taken so that Figure 1B The first welding wire "P1", "P 145 In one or more examples, a solder joint 106 is formed between each of the plurality of solder leads 116 and a corresponding solder pad 120 in the circuit board 104A to connect the socket 102A to the circuit board 104A. For example, the first solder leads "P1", "P 145 ” are connected to pads 120A, 120B in the circuit board 104A via corresponding solder joints 106A, 106B, respectively. In one or more examples, the socket 102A is connected to the circuit board 104A using surface mount technology. As discussed above, surface mount technology typically can include the following steps for connecting (i.e., soldering) the socket 102A to the circuit board 104A, for example, i) flux application, ii) die placement, iii) reflow soldering, and iv) bottom filling processing. Because surface mount technology is an industry standard method for connecting electronic modules, it is not discussed in detail herein. In some examples, each pair of welding leads of the plurality of welding leads 116 that are opposite to each other (e.g., the first welding leads "P1", "P2" 145 ” etc.) to form a gull-wing shaped solder tail of the socket 102A. Further, each of the plurality of solder tails 116 has a foot 122 and a body 124 connected to the foot 122. In such an example, the foot 122 has a flat structure, or is a flat gull-wing solder tail. In one or more examples, each solder joint 106 is formed between the foot 122 of its corresponding solder tail 116 and a solder pad 120 corresponding to the corresponding solder tail on the circuit board.

[0036] Figure 2A 1 is a block diagram depicting a cross-sectional front view of a portion of an electronic device 100 according to an exemplary embodiment of the present disclosure. Figure 1A When the line 2-2' is intercepted, Figure 2AThe cross-sectional front view of the electronic device 100 shown in FIG. 1 may be similar to the cross-sectional front view of the electronic device 100. As discussed herein, the electronic device 100 may be a computing device 100A, the electronic component 102 may be a socket 102A, and the electronic board 104 may be a circuit board 104A. In such an example, the socket 102A may be connected to the circuit board 104A via a solder joint 106A.

[0037] exist Figure 2A In the example depicted in FIG. 1 , some of the solder joints 106A are defective due to warping of the circuit board 104A during the heating and cooling cycles of the reflow soldering process of the surface mount technology. The defect may be necking of the solder joints 106A between at least some of the solder leads 116 of the socket 102A and the circuit board 104A. For example, the solder joints 106A1 formed between the first solder lead “P1” of the plurality of solder leads 116 and the first solder pad 120A1 of the circuit board 104A, and the solder joints 106A1 formed between the last solder lead “P2” and the first solder pad 120A1 of the circuit board 104A may be formed. 144 ” and the last pad 120A of the circuit board 104A 144 Welded joints between 106A 144 It can be elongated (necked). Typically, the weld joint 106A1 (or 106A 144 ) may reduce or decrease the side length of the weld joint or the side joint length "dim-D1" (e.g. Figure 2B ) or "dim-D 144 ”, thereby causing premature rupture of the weld joints 106A, 106B. However, the weld lead “P 78 ” (or “P 79 ”) with the pad 120A substantially centered on the circuit board 104A 78 (or 120A 79 ) between the substantially central welded joints 106A 78 (or 106A 79 ) may not be defective (ie, not elongated (or necked)). Therefore, weld joint 106A 78 or 106A 79 Can have appropriate side joint length "dim-D 78 ” or “dim-D 79 "(like Figure 2C ).

[0038] Figure 2B FIG. 1 is a diagram depicting a weld joint 106A (eg, Figure 2A A block diagram of a cross-sectional side view of a first weld joint 106A1) in an electronic device 100. Figure 2BIn the example of FIG. 1 , the solder joint 106A1 is elongated (necked), resulting in a side joint length “dim-D1” formed at a radial joint length “L1” between the solder pad 120A1 and the first lead “P1”. Figure 2B As shown in , the side joint length "dim-D1" has a reduced or decreased thickness at the center of the weld joint 106A1, thereby causing the weld joint 106A1 to break prematurely. Therefore, the weld joint 106A1 with the reduced side joint length "dim-D1" can be classified as a defective weld joint.

[0039] Figure 2C 1 is a diagram depicting another weld joint 106 (eg, Figure 2A The central solder joint 106A in the electronic device 100 78 ) is a block diagram of a cross-sectional side view. Figure 2C In the example, weld joint 106A 78 With side joint length "dim-D 78 ”, which is formed on the pad 120A 78 With center lead "P 78 "Radial joint length between" L 78 ". Side joint length "dim-D 78 " has a thickness much greater than the side joint length "dim-D1", and the radial joint length "L 78 "Specific radial joint length" L 78 " is much smaller, making the weld joint 106A 78 There is a significantly better bond / connection with the circuit board 104A. Therefore, with a side joint length "dim-D 78 "Welded joints 106A 78 The welded joints can be classified as defect-free.

[0040] As discussed above, operator judgment can be used as a non-destructive inspection method to detect the aforementioned conditions (i.e., defective welds and / or non-defective welds). This non-destructive inspection method relies on the operator's training and knowledge to identify defective welds (such as welds 106A1, 106A2, etc.). 144 ) and / or defect-free welded joints (such as 106A 78 As a result, the operator's judgment is subjective and lacks an objective metric that may allow defective weld joints 106A1, 106A 144 Bypass the quality testing environment. In addition, as discussed above, another inspection method can use optical microscopy to detect defective weld joints. However, optical microscopy uses a destructive inspection method to detect defects. For example, refer to Figure 2A , the electronic device 100 may be damaged, such that, for example, Figure 2B The cross section of the weld joint 106A1 shown in FIG. 1 can be measured using an optical microscope to accurately determine the side joint length "dim-D1" (and / or "dim-D 78 ” and / or “dim-D 144 Later, the side joint length "dim-D1" (and / or "dim-D 78 ” and / or “dim-D 144 ”) to identify defective weld joints 106A1. Therefore, this destructive inspection method can provide an objective metric for identifying defective weld joints so as to prevent defective weld joints 106 from bypassing the quality testing environment. However, the destructive inspection method for measuring the side joint length “dim-D” and detecting defective weld joints 106 based on the measured side joint length “dim-D” is not a feasible inspection method because it requires damaging the electronic device 100 to measure the side joint length “dim-D”.

[0041] Therefore, in some examples of the present disclosure, a measurement machine (e.g., an X-ray machine) based on a non-destructive inspection method is trained to accurately measure the side joint length “dim-D1” of the weld joint 106. In some examples, the X-ray machine is trained based on measurement data obtained through an optical microscope to accurately measure the side joint length “dim-D1” of the weld joint 106 (through an optical microscope). In such an example, the measurement data of the side joint length “dim-D2” obtained from the optical microscope can be used as a benchmark or reference data for training the X-ray machine to accurately measure the side joint length “dim-D1” of the weld joint. In one or more examples, after training the X-ray machine, the machine can be used to accurately measure the side joint length “dim-D1” of the weld joint based on the measured side joint length “dim-D1” (through the X-ray machine) and a predefined side joint length threshold “dim-D3” ( Figure 3A The method of training the X-ray machine to measure the side joint length "dim-D1" and detecting defective weld joints will be discussed in more detail below.

[0042] Figure 3A is a perspective side view of a solder joint 106 formed between a solder wire 116 and a solder pad 120 according to an exemplary embodiment of the present disclosure. Figure 3B FIG. 1 is a perspective front view of a weld joint 106 formed between a weld wire 116 and a weld pad 120 according to an exemplary embodiment of the present disclosure. In some examples, the weld wire 116 may be, for example, a center weld wire “P 78”, and the weld joint “106” may be, for example, a central weld joint “106 78 As discussed above, the solder lead 116 has a foot 122 and a body 124 connecting the foot 122. The foot 122 has a flat structure or a flat gull-wing solder lead. In some examples, the foot 122 has a length "L" and a width "W", which can be obtained by direct measurement and / or cross-sectional measurement from a randomly selected subset of sockets 102A from the same batch used in the correlation study (as discussed in detail below).

[0043] In one or more examples, the weld joint 106 is formed between the foot 122 and the pad 120. In such examples, the weld joint 106 has a side joint length “dim-D” extending along the length “L” of the foot 122. In such examples, the measurement machine is trained to accurately measure the side joint length “dim-D” of the weld joint 106.

[0044] In one or more examples, a predefined side joint length threshold "dim-D3" is determined based on raw material measurements of the leg length "L" and the lead width "W" of each of the plurality of weld leads 116. In such examples, the predefined side joint length threshold "dim-D3" can be used as a comparison criterion for deciding whether a particular weld joint is acceptable (i.e., non-defective) or unacceptable (i.e., defective). In other words, the side joint length of a particular weld joint can be compared to the criteria set by the side joint length threshold to determine whether the particular weld joint is defective or non-defective. In some examples, the determined predefined side joint length threshold "dim-D3" meets (i.e., is acceptable) the Printed Circuit Association-A-610 (IPC-A-610) standard requirements for printed circuit joints. It can be noted here that "IPC-A-610" is a separate grade certification based on a standard published by the IPC entitled "Acceptability of Electronic Assemblies." In one or more examples, the predefined side joint length threshold "dim-D3" indicates that the side joint length of a particular weld joint should be greater than or equal to three times the lead width "3W" of the weld lead corresponding to the particular weld joint in order to obtain a defect-free weld joint. For example, a weld joint with a side joint length of 3.5W, 4W, or 4.25W can be considered a defect-free weld joint. However, a weld joint with a side joint length of 2.8W, 2.5W, or 2W may be considered a defective weld joint. In some other examples, the predefined side joint length threshold "dim-D3" indicates that the side joint length of a particular weld joint should be greater than or equal to seventy-five percent of the leg length "L" of the weld lead 116 corresponding to the particular weld joint in order to obtain a defect-free weld joint. For example, a weld joint with a side joint length of 78%, 80%, or 85% of the leg length "L" can be considered a defect-free weld joint. However, a weld joint having a side joint length of 73%, 67%, or 60% of the foot length "L" may be considered a defective weld joint.

[0045] In some examples, the leg length “L” and the lead width “W” of each of the plurality of welding leads 116 are measured in order to determine a predefined side joint length threshold “dim-D3”. For example, the leg length “L” and the lead width “W” of a plurality of welding leads obtained from a plurality of batches (and suppliers) are measured to determine an average leg length “AL” and an average lead width “AW” of the welding leads 116. Later, the predefined side joint length threshold “dim-D3” is determined based on the average leg length “AL” and the average lead width “AW” of the welding leads 116. In such an example, the predefined side joint length threshold “dim-D3” can be used as a comparison criterion for deciding whether a particular welding joint is acceptable (i.e., non-defective) or unacceptable (i.e., defective).

[0046] Figure 4 4 is a flow chart depicting a method 400 for training a measuring machine (or a first measuring machine) to measure the side joint length of a weld joint according to an exemplary embodiment of the present disclosure. It should be noted here that the method 400 is combined with Figure 1A and Figure 1B , FIG. 2A to FIG. 2C as well as Figure 3A and Figure 3B In some examples, the first measuring machine is an X-ray machine, such as a 3-dimensional X-ray machine or a 5-dimensional X-ray machine.

[0047] In one or more examples, the various steps discussed herein are performed by a computing system. In some examples, the computing system may include processing resources, a machine-readable medium storing executable program instructions, and a memory. The processing resources may be a physical processor. In some examples, the physical processor may be a processor suitable for executing the instructions. Figure 4 In some examples, the machine-readable medium is non-transitory and is alternatively referred to as a non-transitory machine-readable medium. The processing resource may execute one or more program instructions (e.g., the processing resource may execute program instructions) to perform Figure 4 The memory may store an algorithm that a first measuring machine (i.e., an X-ray machine) may use to measure the side joint length "dim-D". In such an example, after the X-ray machine is trained to measure the side joint length, an updated algorithm (i.e., based on the training) may be implemented in the first measuring machine to detect defects among a plurality of weld joints, such as in Figure 6 and Figure 7 discussed in more detail in .

[0048] The method 400 starts at block 402 and continues to block 404. At block 404, the method 400 includes receiving first data representing side joint lengths "dim-D1" of a plurality of solder joints measured by a first measuring machine. As discussed herein, the plurality of solder joints may be formed between solder leads in an electronic component and an electronic board using surface mount technology to form interconnected electronic portions of an electronic device.

[0049] In some examples, the computing system may receive the first data after the first measuring machine performs the following sub-steps: a) measuring the side joint length of the weld joint, b) generating first data representing the side joint length, and c) transmitting the measured side joint length to the computing system. Therefore, the first measuring machine (or X-ray machine) may also include a processor for executing one or more executable program instructions according to an algorithm stored in a machine-readable medium of the first measuring machine to perform the above sub-steps.

[0050] In some examples, a sample interconnect electronic part of an electronic device having a plurality of solder joints for connecting a socket to a circuit board can be selected from a batch of interconnect electronic parts. In such an example, the sample interconnect electronic part can be inspected using a first measuring machine to measure the side joint length "dim-D1". During the inspection, the first measuring machine can include a processor that can execute an algorithm to first detect at least one quality feature of the solder joint and analyze the at least one quality feature to measure the side joint length "dim-D1" of the solder joint. The at least one quality feature can include a gray value gradient and / or an average gray value, which can be determined by the first measuring machine in an inspection area, such as a solder joint in the sample interconnect electronic part. The processor of the first measuring machine can further execute an algorithm to analyze the gray value gradient and / or the average gray value later to measure the side joint length "dim-D1" of the plurality of solder joints. In some examples, the first measuring machine can use image analysis software to determine the side joint length "dim-D1" based on the gray value gradient and / or the average gray value. The first measuring machine can then generate first data representing the side joint length "dim-D1" of the weld joint. Later, the first measuring machine can transmit the first data to the computing system. In some examples, the data transmission can occur in an online mode (i.e., via a network) or an offline mode (i.e., a flash drive), etc.

[0051] Method 400 continues to step 406. At step 406, method 400 includes receiving second data, which represents the side joint length "dim-D2" of the multiple welded joints measured by the second measuring machine. In some examples, the second measuring machine can be an optical machine, such as an optical microscope. As discussed herein, the optical microscope can perform cross-sectional analysis on the multiple welded joints in the same sample interconnect used by the first measuring machine to generate the first data, so as to generate the second data. In other words, the second data is generated based on a destructive method, in which the sample interconnect electronic part inspected by the first measuring machine is destroyed so that the cross section of the welded joint can be inspected using an optical microscope to measure the side joint length "dim-D2". Then, the second measuring machine can generate second data representing the side joint length "dim-D2" of the welded joint. Later, the second measuring machine can transmit the second data to the computing system. In some examples, data transmission can occur in an online mode (i.e., via a network) or an offline mode (i.e., a flash drive). In one or more examples, the second data can be used as a baseline or reference data for training the first measurement machine to accurately measure the side joint length “dim- D1 ” of the weld joint. Method 400 continues to block 408 .

[0052] At block 408, method 400 includes determining a correlation value based on a statistical analysis of the relationship between the first data and the second data. In one or more examples, the correction is a statistical analysis that measures the extent to which two data (ie, the first data and the second data) move relative to each other. Figure 5A (and Figure 4 ), which depicts a graph 500 representing a correlation based on a statistical analysis of a relationship between first data and second data. The first data representing the side joint length "dim-D1" is plotted with reference to a y-axis 502 on the graph 500, and the second data representing the side joint length "dim-D2" is plotted with reference to an x-axis 504 on the graph 500 to measure the statistical relationship between the first data and the second data. In one or more examples, the correction value may be a correlation coefficient value and / or a determination coefficient value.

[0053] In some examples, a correlation coefficient (CC) value is used to measure how strong the relationship between the first data and the second data is. In other words, the CC value can be used to measure the strength of the linear relationship between the first data and the second data. In some examples, a Pearson correlation (such as Minitab) can be used to determine the CC value based on the first data and the second data. In such an example, the CC value can range from "-1" to "1", where a negative value of "-1" indicates a strong negative relationship and a positive value of "1" indicates a strong positive relationship. In particular, if the CC value ranges from "-0.65" to "0.65", it indicates that there is a meaningful correlation between the first data and the second data. More specifically, CC values ​​with a range higher than "0.65" (and can be close to "1") are classified as a predefined correlation coefficient range (or predefined value), which indicates that there is a meaningful or strong relationship between the first data and the second data, or the deviation between the first data and the second data is reduced, or the deviation of the first data relative to the second data is reduced. Refer again to Figure 5A , the side joint lengths “dim-D1” and “dim-D2” are dispersed in the graph 500, and thus the CC value based on the first data and the second data is determined to be about “0.0209.” In other words, the CC value “0.0209” indicates that the first data and the second data are not close to each other.

[0054] Coefficient of determination (R 2 ) values ​​can be used to analyze how differences in the first data can be explained by differences in the second data. More specifically, R 2 The value provides an idea of ​​how many data points fall within the results of the straight line 506 formed by the regression equation. 2 The value may range from "0" to "1" (i.e., "0%" to "100%"). Specifically, R values ​​that range above "0.8" ("80%" and can be close to "100%") are 2 The values ​​are categorized into predefined coefficient of determination percentages (or predefined ranges), which indicates that a greater percentage of the first data falls within / is closer to the straight line 506, or that the deviation between the first data and the second data decreases, or that the deviation of the first data relative to the second data decreases. Figure 5A , the side joint lengths "dim-D1" and "dim-D2" are located away from the straight line 506, so the R based on the first data and the second data 2 The value is determined to be about "0.044" ("4.4%"). In other words, about four percent of R 2 The value indicates that a relatively small percentage of the first data is close to straight line 506 .

[0055] The method continues to block 410. At block 410, the method 400 includes determining a correlation value (ie, a correlation coefficient (CC) value) and / or a determination coefficient (R 2 ) value is outside (i.e., equal to or below) a predefined value (i.e., a predefined correlation coefficient range and / or a predefined determination coefficient percentage). Figure 5A , the CC value is "0.0209", which is outside the predefined value "0.65". Similarly, R 2 The value is "4.4%", which is outside the predefined value "80%".

[0056] Thus, at block 410, the CC value and / or R is determined in response to the processing resources of the computing system. 2 The value is outside the predefined value (ie, equal to or below the predefined value), that is, "yes" at block 410, the method 400 continues to block 412. In some examples, the algorithm may be responsive to determining the CC value and R 2 In some other examples, the algorithm may be updated in response to determining that the CC value or R 2 The value is updated outside the predefined value. At box 412, the method 400 includes processing resources continuing to update the algorithm stored in the memory (the algorithm can be used by the first measuring machine) to measure the side joint length "dim-D1" so as to reduce the deviation between the first data and the second data. In such an example, the algorithm can be used as a dimensional metric in the first measuring machine to detect defects among the multiple welded joints. In some examples, updating the algorithm based on the second data in response to determining that the correlation value is outside the predefined correlation coefficient range and / or the predefined determination coefficient percentage can include fine-tuning the algorithm used by the measuring machine. In one or more examples, fine-tuning the algorithm can include updating the algorithm to change one or more measurement points in the gray value gradient and / or the average gray value to accurately measure the side joint length "dim-D1" of the welded joint. In some examples, fine-tuning the algorithm can help the measuring machine accurately detect the at least one quality feature, such as the gray value gradient and / or the average gray value, so as to accurately measure the side joint length "dim-D1" of the welded joint in the sample interconnected electronic part. The method 400 continues to box 414.

[0057] Further, referring back to block 410, in response to the processing resources of the computing system determining the CC value and / or R 2 If the value is within the predefined value (ie, above the predefined value), that is, "No" at block 410, the method 400 continues to block 414. In other words, in response to determining the CC value and / or R 2 If the value is within the predefined value, the algorithm may not be updated. In some examples, the algorithm may respond to determining the CC value and R 2In some other examples, the algorithm may respond to determining that the CC value or R 2 The value is within the predefined value and is not updated. However, to perform metrology R&R (repeatability and reproducibility) on another set of first data and second data, the method 400 continues to block 414 .

[0058] At block 414, the method 400 includes receiving another set of first data and second data from the first measurement machine and the second measurement machine, respectively, and determining another correlation value (new correlation value) based on a statistical analysis of the relationship between the another set of first data and the second data. Figure 5B (and Figure 4 ), which depicts a chart 550 representing another correlation based on a statistical analysis of the relationship between another set of first data and second data. As discussed above, the other set of first data and second data may be received from a first measuring machine and a second measuring machine, respectively. Specifically, the other set of first data and second data is generated from another sample interconnected electronic part of an electronic device having a plurality of welded joints selected from another batch. In such an example, the first measuring machine may inspect the other sample interconnected electronic part to generate the first data, as discussed above. Further, the second measuring machine may perform a cross-sectional analysis on the multiple welded joints in the same other sample interconnected electronic part used by the first measuring machine to generate the first data to generate the second data, as discussed above.

[0059] In such an example, the first data representing the side joint length (new dim-D1) is plotted with reference to the y-axis 552 on the graph 550, and the second data related to the side joint length (new dim-D2) is plotted with reference to the x-axis 554 on the graph 550 to measure the statistical relationship between the other set of first data and the second data. Later, another correlation value (or new correlation value) can be determined using a statistical model. The other correction value can be a new correlation coefficient (new CC) value and / or a new determination coefficient (new R 2 )value.

[0060] Reference again Figure 5B , the side joint lengths (new dim-D1 and new dim-D2) are clustered in the graph 550, and thus the new CC value based on the other set of first data and second data is determined to be about "0.933". In other words, the CC value "0.933" indicates that the first data and the second data are close to each other. Further, referring to Figure 5B , the side joint lengths (new dim-D1 and new dim-D2) are located close to the straight line 556, so R based on the other set of first data and second data 2The value is determined to be about "0.872" ("87.2%"). In other words, about eighty-seven percent of R 2 The value indicates that a significant percentage of the first data from the other group is close to straight line 556.

[0061] Return to reference Figure 4 , the method 400 continues to block 416. At block 416, the method 400 includes determining a new correlation value (ie, a new correlation coefficient (CC) value and / or a new determination coefficient (R 2 ) value) is outside of a predefined value (i.e., a predefined correlation coefficient range and / or a predefined coefficient of determination percentage). Return to reference Figure 5B , the CC value is "0.93", which is within the predefined value "0.65" (i.e., higher than the predefined value). Similarly, R 2 The value is "87.2%", which is within the predefined value of "80%" (ie, above the predefined value). Therefore, at block 416, the CC value and / or R is determined in response to the processing resources of the computing system. 2 The value is within the predefined value, ie, “no” at block 416 , the method 400 continues to block 418 .

[0062] At box 418, method 400 includes processing resources continuing to update the algorithm with a predefined side joint length threshold value "dim-D3". In some examples, the updated algorithm can be used in a measurement machine (X-ray machine) to detect defects among the plurality of weld joints based on a comparison of the measured side joint length "dim-D1" of each weld joint with the predefined side joint length threshold value "dim-D3". In some examples, the predefined side joint length threshold value "dim-D3" is determined based on raw material measurements of the leg length and the lead width of each of the plurality of weld leads, such as Figure 3A and Figure 3B In order to obtain a defect-free solder joint, the measured side joint length "dim-D1" may need to be greater than or equal to three times the lead width. In some other examples, in order to obtain a defect-free solder joint, the measured side joint length "dim-D1" may need to be greater than or equal to seventy-five percent of the leg length of the solder lead in the socket.

[0063] However, returning to block 416, in response to the processing resources of the computing system determining the CC value and / or R 2 If the value is outside the predefined value (i.e., equal to or below the predefined value), that is, "yes" at block 416, the method 400 loops back to block 412. In such an example, the method 400 includes repeating the steps in blocks 412, 414, and 416 using another set of first data and second data until the processing resources of the computing system determine the CC value and / or R 2The value is within the predefined value (i.e., above the predefined value), i.e., "No" at box 416. Therefore, the first measuring machine is trained based on the side joint length "dim-D2" of the weld joint measured by the optical microscope to accurately measure the side joint length "dim-D1" of the weld joint. Therefore, the updated algorithm (trained) is used as a dimensional metric in the first measuring machine to detect defects among the plurality of weld joints.

[0064] In some other examples, the first measuring machine (e.g., an X-ray machine) may include a processor to perform one or more steps discussed in steps 406 to 418. For example, the processor of the X-ray machine may first execute an algorithm (default algorithm) to measure the side joint lengths of multiple weld joints to generate first data. Further, the processor may receive second data representing the side joint lengths of the multiple weld joints measured by the second measuring machine. Further, the processor may determine a correlation value based on a statistical analysis of the relationship between the first data and the second data. Later, the processor may determine whether the correlation value is outside or within a predefined value. In response to determining that the correlation value is outside the predefined value, the processor may update the default algorithm to improve the accuracy of measuring the side joint lengths of multiple weld joints. However, in response to determining that the correlation value is within the predefined value, the processor may not update the default algorithm for measuring the side joint lengths of multiple weld joints. Therefore, the X-ray machine (or the default algorithm) may be trained to accurately measure the side joint lengths of multiple weld joints. In one or more examples, a trained X-ray machine can be used to determine defective weld joints based on a comparison of a measured side joint length to a predefined side joint length threshold. Figure 6 The steps involved in determining a defective weld joint are discussed in more detail in the example of FIG.

[0065] Figure 6 6 is a flowchart depicting a method 600 for detecting defects in a weld joint according to an exemplary embodiment of the present disclosure. In one or more examples, the various steps discussed herein can be performed by a measurement machine (or a first measurement machine). In some examples, the first measurement machine is an X-ray machine. In one or more examples, the measurement machine may include processing resources and a machine-readable medium that stores executable program instructions (or an updated algorithm or a trained algorithm) received from a computing system, as described above with reference to Figure 4 as well as Figure 5A and Figure 5B In some examples, the processing resource may be a physical processor. In some other examples, the physical processor may be a microprocessor suitable for performing the functions described herein. In some examples, the machine-readable medium is non-transitory and is alternatively referred to as non-transitory machine-readable medium.

[0066] Method 600 starts from block 602 and continues to block 604. At block 604, method 600 includes detecting at least one quality feature of a welded joint in a plurality of welded joints formed between a welded lead in an electronic component (such as a socket) and an electronic board (such as a circuit board). The at least one quality feature may include a gray value gradient and / or an average gray value, which may be determined by a measuring machine in an inspection area, such as a welded joint in an interconnected electronic portion of an electronic device. In some examples, the measuring machine may obtain an image of the welded joint in the interconnected electronic portion. Method 600 continues to block 606.

[0067] At box 606, method 600 includes analyzing the at least one quality feature to measure the side joint length "dim-D1" of each weld joint. In particular, the measuring machine can analyze the gray value gradient and / or the average gray value to measure the side joint length "dim-D1" of the plurality of weld joints. In some examples, the measuring machine can use image analysis software to determine the side joint length "dim-D1" based on the gray value gradient and / or the average gray value. Method 600 continues to box 608.

[0068] At box 608, method 600 includes determining whether to generate an alarm signal indicating a defect detected in the weld joint based on a comparison of the measured side joint length "dim-D1" of each weld joint with a predefined side joint length threshold "dim-D3". As discussed herein, the predefined side joint length threshold "dim-D3" is determined based on raw material measurements of the leg length and the lead width of the welded lead. In some examples, the predefined side joint length threshold "dim-D3" is greater than or equal to three times the lead width. In some other examples, the predefined side joint length threshold "dim-D3" is greater than or equal to seventy-five percent of the leg length of the welded lead.

[0069] In some examples, in response to determining that the measured side joint length "dim-D1" is less than three times the lead width or less than seventy-five percent of the leg length of the welding lead, the measuring machine can generate an alarm signal indicating a defect in the weld joint. In other words, based on the comparison of the measured side joint length "dim-D1" of the weld joint with the predefined side joint length threshold "dim-D3", it is determined whether the weld joint is defective or non-defective. In addition, the measuring machine can define the severity of the defect based on the deviation range of the measured side joint length "dim-D1" and the predefined side joint length threshold "dim-D3". For example, in response to determining that the measured side joint length "dim-D1" of each weld joint is lower than the predefined side joint length threshold "dim-D3" by more than fifteen percent, the measuring machine can define the severity as "critical". In some examples, the measured side joint length "dim-D1" of the weld joint can be 2.5 mm, the predefined side joint length threshold "dim-D3" can be 3 mm, and the value that is fifteen percent lower than the predefined side joint length threshold "dim-D3" is 2.55 mm. In such an example, the measuring machine can mark the weld joint as a defective weld joint because the measured side joint length "dim-D1" (2.5 mm) is lower than the side joint length threshold "dim-D3" (3 mm). Further, the measuring machine can define the severity of the weld joint as severe because the measured side joint length "dim-D1" (2.5 mm) is less than the value (2.55 mm) that is fifteen percent lower than the predefined side joint length threshold "dim-D3". In some examples, the severity defined as "severe" can initiate a stopover of a production environment of an interconnected electronic part having multiple weld joints. In some other examples, in response to determining that the measured side joint length "dim-D1" of each of the multiple weld joints is less than or equal to fifteen percent below the predefined side joint length threshold "dim-D3", the measurement machine may define the severity as "major". In some examples, the measured side joint length "dim-D1" of the weld joint may be 2.85 millimeters, the predefined side joint length threshold "dim-D3" may be 3 millimeters, and the value that is fifteen percent lower than the predefined side joint length threshold "dim-D3" is 2.55 millimeters. In such an example, the measurement machine may mark the weld joint as a defective weld joint because the measured side joint length "dim-D1" (2.85 millimeters) is lower than the side joint length threshold "dim-D3" (3 millimeters). Further, the measuring machine may define the severity of the weld joint as major because the measured side joint length "dim-D1" (2.85 mm) is greater than a value (2.55 mm) that is fifteen percent lower than a predefined side joint length threshold "dim-D3".In some examples, a severity defined as "major" may include an indicator for initiating process improvements in the manufacture of interconnected electronic parts having multiple solder joints.

[0070] In response to the measuring machine determining that the measured side joint length "dim-D1" is greater than or equal to three times the lead width "dim-D3" (e.g., 3.5W, 3.8W, 4W, etc.) or greater than or equal to seventy-five percent (e.g., 75%, 76%, 86.5%, 95%, etc.) of the leg length "dim-D3" of the soldered lead, the measuring machine may not generate an alarm signal. In such an example, the production environment can continue to manufacture interconnected electronic parts with multiple solder joints without defects. Method 600 ends at block 610.

[0071] Figure 7 A block diagram 700 is depicted of a measurement machine 702 (or a first measurement machine or an x-ray machine) having a processing resource 704 and a machine-readable medium 706 storing executable program instructions (an updated algorithm or a trained algorithm) for detecting defects in a plurality of weld joints. In some examples, the processing resource 704 is operably coupled to the machine-readable medium 706. The processing resource 704 may be a physical processor. In some examples, the physical processor may be a processor adapted to perform operations related to the detection of defects in the weld joints. Figure 7 The machine-readable medium 706 is non-transitory and may alternatively be referred to as a non-transitory machine-readable medium. In some examples, the machine-readable medium 706 may be accessed by the processing resource 704. In some examples, the machine-readable medium 706 stores program instructions corresponding to the functions of the measurement machine 702. In such examples, the processing resource 704 executes one or more program instructions (e.g., the processing resource may execute program instructions) to perform Figure 6 One or more of the functions described in .

[0072] The processing resource 704 can execute program instructions to detect at least one quality feature of a solder joint in a plurality of solder joints formed between a solder lead in a socket and a circuit board, such as Figure 6 As discussed in . In one or more examples, each of the plurality of solder joints has a flat gull-wing solder tail. In such an example, the solder joint is formed between the flat gull-wing solder tail and the circuit board. The measuring machine can obtain an image of the solder joint to capture the gray value gradient and / or average gray value in the solder joint.

[0073] Further, the processing resource 704 can execute one or more program instructions for analyzing the at least one quality feature to measure the side joint length "dim-D1" of the weld joint, such as Figure 6In some examples, the measuring machine may include image analysis software to determine the side joint length "dim-D1" based on the gray value gradient and / or the average gray value.

[0074] The processing resource 704 can further execute the one or more program instructions to determine whether the measured side joint length "dim-D1" is higher than a predefined side joint length threshold "dim-D3". In some examples, the predefined side joint length threshold "dim-D3" is determined based on raw material measurements of the leg length and the lead width of each of the plurality of welding leads, such as Figure 3A and Figure 3B In some examples, to obtain a defect-free solder joint, the measured side joint length "dim-D1" may need to be greater than or equal to three times the lead width. In some other examples, to obtain a defect-free solder joint, the measured side joint length "dim-D1" may need to be greater than or equal to seventy-five percent of the leg length of the solder lead.

[0075] In response to the processing resource 704 determining that the measured side joint length "dim-D1" is less than three times the lead width or less than seventy-five percent of the leg length of the welded lead, the processing resource 704 may generate an alarm signal indicating a defect of the welded joint based on a comparison of the measured side joint length "dim-D1" of the welded joint with a predefined side joint length threshold "dim-D3". In response to the processing resource 704 determining that the measured side joint length "dim-D1" is greater than or equal to three times the lead width "dim-D3" or greater than or equal to seventy-five percent of the leg length "dim-D3" of the welded lead, the processing resource 704 may not generate an alarm signal.

[0076] The various features illustrated in the examples described herein can be implemented to prevent defective solder joints from escaping a quality inspection environment. Therefore, embodiments of the present disclosure can provide a higher confidence level for a socket connector (e.g., a DDRX SMT connector) because the solder joint meets the IPC-A-610 requirements to meet a predefined warranty period. Further, the inspection method is a non-destructive quantitative method that can improve operational work efficiency and reduce failure analysis cycle time. In addition, the inspection method can define quantitative metrics for X-ray inspections with objective threshold acceptance criteria rather than subjective operator judgment, thereby reducing destructive analysis of interconnected electronic parts.

[0077] In the foregoing description, many details are set forth to provide an understanding of the subject matter disclosed herein. However, embodiments may be practiced without some or all of these details. Other embodiments may include modifications, combinations, and variations of the details discussed above. The appended claims are intended to cover such modifications and variations.

Claims

1. A method for training a measuring machine for measuring the side joint length of a weld joint to detect defects, comprising: i) a processing resource of a computing device receives first data, the first data representing side joint lengths of a plurality of weld joints measured by a first measuring machine, wherein the plurality of weld joints are formed between welding leads in an electronic component and an electronic board, the first measuring machine using a non-destructive inspection method to measure the side joint lengths of the plurality of weld joints; ii) the processing resource receives second data representing side joint lengths of the plurality of weld joints measured by a second measuring machine, the second measuring machine being capable of determining the side joint lengths of the plurality of weld joints more accurately than the first measuring machine; iii) the processing resource determines a correlation value based on a statistical analysis of a relationship between the first data and the second data; and iv) the processing resource updates an algorithm used by the first measuring machine to measure the side joint length based on the correlation value to reduce a deviation between the first data and the second data, wherein the algorithm is used as a dimensional metric in the first measuring machine to detect defects among the plurality of weld joints, Wherein, updating the algorithm comprises training the algorithm for measuring the side joint length based on the second data generated by the second measuring machine.

2. The method of claim 1, wherein: Updating the algorithm includes: The processing resource determines whether the correlation value is outside a predefined correlation coefficient range; and The processing resource updates the algorithm based on the second data in response to determining that the correlation value is outside the predefined correlation coefficient range, wherein the correlation value is a correlation coefficient value.

3. The method of claim 1, wherein: Updating the algorithm includes: The processing resource determines whether the correlation value is outside a predefined coefficient of determination percentage; and The processing resource updates the algorithm based on the second data in response to determining that the correlation value is outside the predefined coefficient of determination percentage, wherein the correlation value is a coefficient of determination value.

4. The method of claim 1, wherein: Updating the algorithm includes: The processing resource determines whether the correlation value is within a predefined value, wherein the correlation value comprises a correlation coefficient value and a determination coefficient value, and wherein the predefined value comprises a predefined correlation coefficient range and a predefined determination coefficient percentage; and The processing resource repeats steps i) to iii) based on another set of the first and second data in response to determining that the correlation value is within the predefined value.

5. The method of claim 4, further comprising: The processing resource determines whether a correlation value of another set of the first data and the second data is outside the predefined value; as well as The processing resource updates the algorithm based on second data in another set of the first data and the second data in response to determining that the correlation value is outside the predefined value.

6. The method of claim 4, further comprising: The processing resource determines whether a correlation value of another set of the first data and the second data is within the predefined value; as well as The processing resource updates the algorithm using a predefined side joint length threshold in response to determining that the correlation value is within the predefined value, wherein the predefined side joint length threshold is determined based on raw material measurements of leg length and lead width of the weld wire.

7. The method of claim 6, further comprising: a processor of the first measuring machine processing the algorithm to measure a side joint length of the weld joint; as well as The processor determines whether to generate an alarm signal indicative of a defect detected in the weld joint based on a comparison of a measured side joint length of the weld joint with the predefined side joint length threshold.

8. The method of claim 1, wherein: The first measuring machine is an X-ray machine and wherein the second measuring machine is an optical machine.

9. A method for detecting defects in a weld joint, comprising: detecting, by a processing resource of a measuring machine using the method of claim 1, at least one quality feature of a solder joint among a plurality of solder joints formed between solder leads in an electronic component and an electronic board; The processing resource analyzes the at least one quality feature to measure a side joint length of the weld joint; and The processing resource determines whether to generate an alarm signal indicating a defect detected in the weld joint based on a comparison of a measured side joint length of the weld joint with a predefined side joint length threshold, wherein the predefined side joint length threshold is determined based on raw material measurements of a leg length and a lead width of the weld lead.

10. The method of claim 9, wherein: The predefined side joint length threshold is greater than or equal to three times a lead width of the soldering wire in the electronic component.

11. The method of claim 9, wherein: The predefined side joint length threshold is greater than or equal to seventy-five percent of a leg length of the bond wire in the electronic component.

12. The method of claim 9, further comprising: The processing resource, in response to detecting a defect in the weld joint, defines a severity of the defect based on a range of deviation of the measured side joint length from the predefined side joint length threshold.

13. The method of claim 12, wherein: The severity includes an indicator of process improvement in the manufacture of the plurality of weld joints, or an interruption of a production environment for the plurality of weld joints.

14. The method of claim 9, wherein: The measuring machine is an X-ray machine.

15. A measuring machine using the method according to claim 1, comprising: a machine-readable medium storing program instructions; as well as A processing resource operably coupled to the machine-readable medium, wherein the processing resource executes the program instructions to: detecting at least one quality characteristic of a weld joint among a plurality of weld joints formed between a weld lead in the electronic component and an electronic board; analyzing the at least one quality feature to measure a side joint length of the weld joint; and A determination is made whether to generate an alarm signal indicating a defect detected in the weld joint based on a comparison of a measured side joint length of the weld joint with a predefined side joint length threshold, wherein the predefined side joint length threshold is determined based on raw material measurements of a leg length and a lead width of the weld lead.

16. The measuring machine according to claim 15, wherein: Each of the plurality of weld joints has a flat gull-wing weld tail.

17. The measuring machine of claim 15, wherein: The measuring machine is an X-ray machine.

18. The measuring machine of claim 15, wherein: The predefined side joint length threshold is greater than or equal to three times a lead width of the soldering wire in the electronic component.

19. The measuring machine of claim 15, wherein: The predefined side joint length threshold is greater than or equal to seventy-five percent of a leg length of the bond wire in the electronic component.

Images