JP2013191888A - Damage index prediction system and damage prediction method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a damage index prediction system of an electronic apparatus capable of highly accurately performing damage prediction of mounting components or the like on a circuit board.SOLUTION: A damage index prediction system is a damage prediction system that predicts an index associated with damage of a joining portion of an electronic apparatus having a joining portion for electrically connecting electronic components to a mounting circuit board and a detection joining portion designed to have a life shorter than that of the joining portion. The damage index prediction system comprises: a damage detection section for acquiring information associated with damage of the detection joining portion; and an arithmetic section for calculating a prediction value of the index associated with the damage of the joining portion from the information associated with the damage of the detection joining portion acquired by the damage detection section and a relation between an index associated with the damage of the detection joining portion and the index associated with the damage of the joining portion.

Description

  Embodiments described herein relate generally to a damage index prediction system and a damage prediction method, and more particularly, to a damage index prediction system and a damage prediction method for an electronic device.

  In recent years, various devices equipped with semiconductor devices have become more sophisticated and multifunctional, and with this, semiconductor chips have become more highly integrated and larger in scale. There is a tendency that the number of electrical joints connecting the two, generally solder joints, is greatly increased. As a result, thermal stress is repeatedly generated in the joint portion of the semiconductor package, which causes a problem of thermal fatigue failure.

  Thus, in order to prevent thermal fatigue failure at the joint, a method for detecting that the thermal fatigue failure at the joint is approaching has been proposed (see, for example, Patent Document 1). In addition to the electrical connection bumps, sensor bumps are provided to electrically connect the semiconductor package side and the circuit board side, and the electrical resistance value of the connection path including the sensor bumps is automatically detected to a predetermined level. It is determined that the thermal fatigue failure is progressing when the value exceeds.

  In the thermal fatigue failure detection method proposed in Patent Document 1 described above, it is detected that the thermal fatigue failure of the joint portion of the semiconductor package is approaching, but it is determined how much the remaining life of the joint portion is. There is a problem that you can not.

  In addition, in the case where damage is predicted before using an electronic device, or when damage is predicted from parameters measured during use of the electronic device, the joint of the mounted component on the circuit board is not recommended. There is an individual difference in each component such as a part. In addition, since there is also a variation in the fatigue characteristics of the joint, there is a large variation in the actual joint damage, which may be significantly different from the damage prediction value.

Japanese Patent No. 3265197

  The problem to be solved by the present invention is to provide a damage index predicting system capable of more surely predicting damage at a joint portion of mounted components on a mounted circuit board.

  The damage index prediction system according to the embodiment includes the joint of an electronic device having a joint that electrically connects an electronic component to a circuit board for mounting, and a joint for detection that is designed to have a lower life than the joint. A damage prediction system for predicting an index related to damage of a part, a damage detection unit for acquiring information on damage of the detection joint, and information on damage of the detection joint obtained by the damage detection unit, A calculation unit that calculates a predicted value of the index related to damage of the joint from the relationship between the index related to damage of the detection joint and the index related to damage of the joint;

The damage index predicting method according to the embodiment includes the joint part of an electronic device having a joint part for electrically connecting an electronic component to a circuit board for mounting, and a joint part for detection designed to have a lower life than the joint part. A damage prediction method for predicting a damage index of
Obtaining information relating to damage of the detection joint, an index relating to the damage of the joint for detection obtained by this damage detection, an index relating to damage to the joint for detection, and an indicator relating to damage to the joint; From this relationship, the predicted value of the index related to the damage of the joint is calculated.

It is a block diagram which shows schematic structure of the damage parameter | index prediction system regarding embodiment. It is sectional drawing of the electronic component regarding embodiment, and its periphery. It is a block diagram which shows the structure outline of the database construction of a life ratio. It is a figure which shows the flow of the damage parameter | index prediction method based on the damage of the junction part of an electronic component. It is a figure which shows the flow of the damage parameter | index prediction method of the junction part of the electronic component in the modification 1. FIG. 11 is a cross-sectional view of an electronic component according to Modification 2 and its surroundings. It is a figure which shows the flow of the damage parameter | index prediction method of a junction part. It is a figure which shows the flow of the damage parameter | index prediction method of the junction part of the electronic component in the modification 3. It is a block diagram which shows the structure outline of the database construction of the life ratio of a junction part and a junction part for a detection from a phenomenon analysis. FIG. 10 is a cross-sectional view of an electronic component according to Modification Example 5 and its periphery. It is a figure explaining the electronic component which concerns on the modification 6, and its periphery, (a) is sectional drawing of an electronic component and its periphery, (b) shows arrangement | positioning of the junction part and detection junction part of an electronic component ing. It is a figure explaining the electronic component which concerns on this modification 7, and its periphery, (a) is sectional drawing of the electronic component 2 and its periphery, (b) is the junction part 6 and detection junction part of the electronic component 2 7 is shown. It is a figure explaining the electronic component which concerns on this modification 8, and its periphery, (a) is sectional drawing of the electronic component 2 and its periphery, (b) is the junction part 6 and detection junction part of the electronic component 2 7 is shown. It is a block diagram which shows schematic structure of the damage parameter | index prediction system in 2nd Embodiment. It is a block diagram which shows the structure outline of the database construction of the distortion range of a junction part and a junction part for a detection from a phenomenon analysis. The flow of the lifetime prediction method of the junction part 6 of the electronic component in 2nd Embodiment is shown. It is a schematic diagram for demonstrating the correction method of the database of the fatigue characteristic of a junction part. It is a schematic diagram for demonstrating the correction method of a fatigue characteristic database. It is a block diagram which shows schematic structure for constructing | assembling the database of the distortion range of the junction part in a modification, and the junction part for a detection.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In the drawings, the same portions are denoted by the same reference numerals, and redundant description is omitted.

(First embodiment)
FIG. 1 is a block diagram illustrating a schematic configuration of a damage index prediction system according to the first embodiment, which uses damage information of a joint portion of an electronic component on a mounted circuit board. This damage index prediction system can be incorporated inside the electronic device, or can be arranged and connected outside the electronic device.

  As shown in FIG. 1, the damage index prediction system 100 includes a database 11 regarding the relationship between the joint and the detection joint damage, a damage detection unit 12 that detects the damage of the detection joint, and an index about the joint damage. Is provided with a calculation unit 13 and a display output unit 14. However, the display output unit 14 may not be provided.

  In the database 11, for example, an acceleration test such as a temperature cycle test and a phenomenon analysis such as stress analysis are performed on the electronic device in advance, and from the result, the joint part of the electronic component and the joint part for detection (described later) Accumulate the results of calculating the relationship for damage.

  In addition, the relationship regarding the damage between the joint of the electronic component and the joint for detection is expressed by, for example, a function expression of the index regarding the damage between the two, and the index regarding the damage of the joint of the electronic component from the index regarding the damage of the joint for detection Can be calculated.

  Here, the index regarding damage will be described. In general, an electronic device is turned on every time it is used, and is turned off when the use is finished. Thermal stress is generated in the mounted components inside the electronic device due to such repeated power ON / OFF. In addition, when used in a moving vehicle or when an electronic device is dropped, vibration is also applied to the electronic components inside the electronic device. Thermal stress and vibration generated in such an electronic component will affect the damage or life, and what is expressed as an index is used as an index for damage. Examples of the damage index include a breakage life, a damage value, a function using the breakage life, a function using the damage value, and the like.

  Here, the breakage life represents a period until breakage, and is represented by, for example, the number of cycles until breakage or breakage time.

  The damage value can be defined as follows. The damage value when one cyclic load is applied is represented by the reciprocal of the number of life cycles when the same cyclic load is applied. The damage value when the load is repeatedly generated is the cumulative damage value generated in each cycle. It is a thing. When the cumulative damage value reaches 1, it indicates that the joint has been damaged.

  Assume that the damage index is the failure life, the relationship of the damage index is the life ratio, and the detection joint is designed and manufactured with a damage life against temperature fluctuation of half of that of the electronic component joint. If the life ratio is defined as the ratio of the predicted life of the joint for detection when the predicted life of the joint of the electronic component is 1, the life ratio of the joint for detection at this time is 0.5. At this time, if the detection joint can withstand repeated use 2000 times, it can be predicted that the joint of the electronic component can withstand 4000 repeated use. Note that the predicted life is not limited to the number of repeated cycles of temperature fluctuation. That is, if it is grasped by the broken life and the usage time of the detection joint, the predicted life can be predicted as the time until the life of the electronic device.

  The damage detection unit 12 of the detection joint is for detecting damage of the detection joint. The detection method will be described later. For example, the damage detection unit 12 can be configured as an electric circuit.

  In addition, the detection joint is structurally similar to the joint of the electronic component, i.e., an electrical signal for invoking the function of the electronic component does not pass through. You may form as a part and these detection parts are comprised, for example with a solder material.

  First, the index calculation unit 13 relating to the damage of the joint part inputs the information on the joint part for detection from the damage detection unit 12, while taking in the relation regarding the damage accumulated from the database 11, A predicted value of an index related to damage, for example, a predicted value of the number of cycles until failure is calculated. This calculating part 13 can be comprised by CPU, for example. Note that the predicted value of the index related to the damage of the joint portion of the electronic component may be displayed on the display output unit 14.

  Since individual electronic devices are used in different environments and have individual differences, uniform prediction before use is not sufficient from the perspective of accuracy, and the use of devices By reflecting the progress and individual differences in the prediction, an appropriate damage index can be predicted.

  The display output unit 14 is for presenting information on the lifetime to the operator of the electronic device, and it is preferable to share the display of the electronic device. For example, “The number of times that it can be used is 4000 times, and the number of times it can be used is 2000 times.” Is displayed. However, the display output unit 14 may not be provided.

  Next, the configuration of the detection joint will be described. FIG. 2 is a cross-sectional view of the electronic component and its surroundings according to the embodiment. As shown in FIG. 2, the electronic component 2 is mounted on the circuit board 1, and the detection device 3 is mounted on the back surface side of the circuit board 1. The detection device 3 is electrically connected to the circuit board 1 by a detection joint 7. The joint 7 for detection has a life shorter than that of the joint 6 of the electronic device 2. For example, when the joint 6 has a lifetime that can withstand repeated use 10,000 times, the joint 7 for detection has half the life of the joint 6. It is designed to have a lifetime that can withstand 5000 repeated uses.

  There are a variety of design methods that may cause differences in the service life. For example, the detection device 3 is arranged immediately below the corner of the electronic component 2 where the curvature of the circuit board 1 changes characteristically or when the electronic component 2 is a semiconductor package, on the back side of the circuit board near the chip corner. To do. With such an arrangement, the load on the detection junction 7 becomes larger than the load on the junction 6, so the life is shortened.

  The circuit board 1, the package substrate 4 of the electronic component 2, and each detection device substrate 5 are made of materials having different linear expansion coefficients, and the linear expansion coefficients are the circuit board 1, the package substrate 4, and the detection device substrate 5. The difference in linear expansion coefficient between the package substrate 4 and the circuit board 1 is made larger than the difference in linear expansion coefficient between the package substrate 4 and the circuit board 1. For example, the circuit substrate 1 and the package substrate 4 are resin substrates, and the detection device substrate 5 is a ceramic substrate. As a result, when the shape and boundary conditions of the joint are the same, the greater the difference in linear expansion coefficient, the greater the thermal stress generated at the joint, so that the detection joint 7 is the joint 6 of the electronic component 2. Compared to, it can be made to have a shorter life. Further, by using materials with different linear expansion coefficients for the detection device substrate 5, the lifetime can be set in stages, and a device having a larger difference in linear expansion coefficient from the circuit board 1 has a shorter lifetime.

Next, the detection of damage of the detection joint 7 which is indispensable for grasping the life of the detection joint will be described. Here, for example, a change in electrical characteristic value is used. Assuming that electrical resistance is used as the electrical characteristic value, continuity cannot be obtained and the resistance value becomes infinite, or if the resistance value has changed significantly from the past, it can be understood that the detection joint has been damaged. it can. As another example of the electric characteristic value, the capacitance can be increased, and the detection junction can detect damage due to the change. In this embodiment, the detection junction 7 can be provided at a plurality of locations. This is possible, and a more reliable damage index can be predicted as compared with the case where one joint 7 for detection is provided.

  FIG. 3 shows an outline of the construction of a database construction of the life ratio of the joint 6 and the detection joint 7 from an accelerated test such as a temperature cycle test when the damage index is the failure life and the damage relation is the life ratio. FIG. Further, the life ratio is assumed to be a ratio of the predicted life of the joint for detection when the predicted life of the joint of the electronic component is 1. The database includes an acceleration test result database 31, a life ratio calculation unit 32 between the joint 6 and the detection joint 7, and a life ratio database 33 between the joint 6 and the detection joint 7. An accelerated test such as a temperature cycle test is performed in advance, and the damage life of the joint 6 and the detection joint 7 in the accelerated test is stored in the accelerated test result database 31. As described above, the failure life is, for example, a test time or the number of cycles. From the damage life in the accelerated test of the joint portion 6 and the detection joint portion 7 of the acceleration test result 31, the life ratio of the joint portion and the detection joint portion 7 is calculated by the calculation portion 32, and the joint portion and the detection joint portion 7 are calculated. It accumulates in the life ratio database 33.

  FIG. 4 shows the flow of the damage index prediction method based on the damage of the joint 6 of the electronic component when the number of detection joints 7 is one. As shown in FIG. 4, the electrical characteristic value of the detection joint 7 is measured (S401). If the electrical characteristic value in the detection joint 7 is not within the specified range, it is determined that the detection joint 7 is damaged (S402). When the joint 7 for detection detects breakage (S403), the actual life of the joint 7 for detection and the life ratio data of the joint 6 and the joint 7 for detection from the database 33 constructed in advance are obtained. The life expectancy value of the joint 6 is calculated (S404), and is appropriately output as the life expectancy value of the joint 6 of the electronic component 2 (S405).

  The system may operate constantly during use, operate at a certain number of times of use such as when the device is started up, or operate at regular intervals such as once a week.

  According to this embodiment, during use of an electronic device, by predicting the life of the joint of the electronic component based on the damage life of the joint for detection, the environment and usage situation around the device, the individual of the device It is possible to predict damage index with high accuracy considering the difference.

  Next, Modification 1 of the first embodiment will be described. FIG. 5 shows a flow of a technique for predicting a damage index of the joint 6 of the electronic component in the first modification. As shown in FIG. 5, in the first modification, the first embodiment is the same as the first embodiment except for a method for calculating a predicted value of an index related to damage to the joint 6 of the electronic component 2 every time each detection joint is damaged. It is the same. In this modification, an index relating to damage of the joint 6 of the electronic component 2 is described as a damage value, and a relationship regarding damage between the joint 6 and the detection joint 7 is described as a life ratio. In the first modification, the electrical characteristic value of the detection joint 7 is measured (S501). If the electrical characteristic value in the detection joint 7 is not within the specified value range, it is determined that the detection joint 7 is damaged (S502). When the detection joint 7 is damaged (S503), the life ratio data of the joint 6 and the detection joint 7 from the database 33 is taken in, and the damage value of the joint 6 of the electronic component 2 is calculated ( In step S504, the predicted damage value of the joint 6 of the electronic component 2 is output as appropriate (S505). If the life ratio between the detection joint 7 and the joint 6 is 0.5, the damage value of the joint 6 is predicted to be 0.5 when the detection joint 7 is damaged. Can do. When the display output unit 14 is provided, for example, “50% of the currently available time” can be displayed.

  According to the first modification example, when the detection joint is damaged without using the failure life of the detection joint, information on the damage of the detection joint, the electronic component joint 6 and the detection joint Since the predicted damage value of the joint 6 of the electronic component 2 is calculated based on the life ratio, which is a relationship related to the breakage of the part 7, it is not necessary to measure the cycle number and usage time of the detection joint. Therefore, it is possible to predict the damage index without requiring a device for measuring the number of cycles and usage time and a device for storing the number of cycles and usage time. Although this modification 1 is a case where the number of detection joints is one, it is possible to use a plurality of detection joints, and it is possible to predict the life with higher accuracy.

  Next, Modification 2 of the first embodiment will be described. FIG. 6 is a cross-sectional view of an electronic component and its periphery in Modification 2. The structure of the second modification is the same as that of the first embodiment except that there are a plurality of detection joints 7. Any of the detection joints 7 is designed to have a shorter lifetime than the joint 6 of the electronic component 2.

FIG. 7 shows the flow of the damage index prediction method for the joint 6. As shown in FIG. 7, in the second modification, there are n detection joints 7, and a method other than the method of correcting the predicted value of the index related to the damage of the joints 6 every time each detection joint 7 is damaged. This is the same as in the first embodiment. x indicates the state of the detection joint, and is 0 when normal and 1 when broken. Moreover, k shows the order of the damaged junction part. The initial values of k and x of all joints are set to 0 (S701). The value of k is incremented by 1 every time the detection joint is broken, and the value of x is set to be 1 when the breakage of the detection joint is detected (S706). Electrical characteristic values are measured for each of the n detection joints 7 (S702 to S708). It is determined whether or not the detection joint 7 is damaged (S703), and the electrical characteristic value of the detection joint that is not damaged is measured (S704). If the electrical characteristic value of the detection joint 7 is not within the specified range, it is determined that the detection joint 7 is damaged (S705). When the breakage is detected (S706), the breakage life of the detection joint 7 where the breakage is detected and the joint 6 taken from the database 33 of the life ratio of the joint 6 and the detection joint 7 and the breakage detection are detected. The life of the joint 6 is predicted from the life ratio of the joint 7 for use, and the value is set to N _fk (S707). Statistical values of N _{f1 to} N _fk , for example, in the second modification, an average value is calculated, and is used as a predicted life value N _f (S710), and is displayed as a corrected predicted life value (S711). If none of the detection joints are damaged, the predicted life value is not displayed (S709). Thereafter, each time the detection joint 7 is damaged, the joint calculated from the damage life of each of the detection joints 7 that has been damaged so far and the life ratio of the joint 6 and the detection joint 7 that has been damaged. the average value of each life prediction value of 6, will modify the life prediction value N _f of the joint 6.

  In the second modification, the failure life is used as an index related to damage, but the damage value described in the first modification may be used instead.

  According to the second modification, more accurate prediction can be performed by correcting damage index prediction step by step by detecting breakage of a plurality of detection joints.

Next, Modification 3 of the first embodiment will be described. FIG. 8 shows a flow of a technique for predicting a damage index of the joint 6 in the third modification. As shown in FIG. 8, in the third modification, except for the processing method for determining the predicted life of the joint from the predicted life of the plurality of joints calculated from the failure lives of the plurality of joints for detection, This is the same as the second modification of the first embodiment. In order to detect the breakage of the n detection joints 7, electrical characteristic values are measured for each of them (S802 to S808). When the damage is detected (S806), from the actually obtained damage life of the detection joint 7 and the life ratio data (33) between the joint 6 and the detection joint 7 that has been damaged, The predicted life value N _fk is calculated (S807). The minimum value of N _{1 to} N _fk is set as the predicted life value N _f of the joint 6 (S810), and is displayed as a corrected predicted life value (S811). Thereafter, the minimum value of each life prediction value of the joint 6 calculated from the failure life of each of the joints for detection 7 damaged so far and the life ratio of the joint 6 to the joints for detection 7 damaged is calculated. Then, the life expectancy value N _f of the joint 6 is corrected.

  According to the third modification, since the data of the minimum life is reflected among the plurality of pieces of damage life data, it is possible to predict the damage index on the safer side.

  The processing method for determining the life prediction value of the joint from the life prediction value of the plurality of joints calculated from the failure life of the plurality of detection joints is not limited to the above. In some cases, the median value is used, or in other cases, a life prediction value calculated from the failure life of the last damaged joint for detection is used. The damage life of the detection joint that has been damaged last reflects the usage history from the start of use of the electronic device to the time of breakage. In the case of prediction based on this, the damage index prediction can be corrected step by step according to the usage history.

  In the third modification, the failure life is used as an index related to damage, but the damage value described in the first modification may be used instead.

  Next, a modified example 4 in which the method of constructing the life ratio database of the joint 6 and the detection joint 7 in the first embodiment will be described. FIG. 9 is a schematic configuration diagram of database construction of the life ratio of the joint 6 and the detection joint 7 from a phenomenon analysis such as FEM analysis when the damage index is a failure life and the relationship of the damage index is a life ratio. FIG. Design information, material database 91, phenomenon analysis module 92, state parameter database 93 of joint 6 and detection joint 7, fatigue characteristic database 94, life ratio calculation unit 95 of joint 6 and detection joint 7 The life ratio database 33 of the joint 6 and the detection joint 7 is provided. In the design information and material database 91, information necessary for the phenomenon analysis, for example, the size and arrangement of each part, the material property value of each part, and the like are accumulated. The phenomenon analysis module 92 performs a phenomenon analysis simulating the use state of the electronic device from the design information and material information of the database 91, calculates state parameters for the joint part of the electronic component and the joint part for detection, and the database 93 To accumulate. Assuming that the state parameter is the strain range of the joint, the strain range data generated in the electronic component joint 6 and the detection joint 7 stored in the database 93 and the joint stored in the fatigue characteristic database 94 are stored. The life ratio between the joint 6 and the detection joint 7 is calculated by the calculation unit 95 using the data representing the relationship between the strain range and the life, and stored in the database 33.

  According to the fourth modification, a life ratio database of the joint 6 and the detection joint 7 can be easily constructed in a relatively short time without requiring an experimental device.

  Next, in the first embodiment, a description will be given of a modified example 5 in which the arrangement of the detection device 3 for predicting an index related to damage of the joint 6 that electrically connects the electronic component 2 and the circuit board 1 is changed. . FIG. 10 is a cross-sectional view of an electronic component and its surroundings in Modification 5. As shown in FIG. 10, the electronic component 2 is mounted on the circuit board 1, and a plurality of detection devices 3 are mounted in the vicinity of the electronic component 2 on the same surface as the electronic component 2 of the circuit board 1. The plurality of detection devices 3 are electrically connected to the circuit board 1 through detection joints 7. Each of the detection joints 7 is designed to have a lower lifetime than the junction 6 of the electronic device 2 and to have a difference in the respective lifetimes. According to the fifth modification, since the detection device 3 is arranged on the same plane as the electronic component 2 of the circuit board 1, the thickness of the entire electronic device can be reduced. In addition, the damage index can be predicted with higher accuracy by using the index regarding damage at the joints of the plurality of detection devices having the index regarding damage in stages.

  Next, a description will be given of a further modified example 6 in which the arrangement of the detection units for predicting the lifetime of the joint is changed. 11A and 11B are diagrams for explaining an electronic component and its surroundings according to the modified example 6. FIG. 11A is a cross-sectional view of the electronic component 2 and its surroundings, and FIG. The arrangement | positioning of the junction part 7 is shown.

  As shown in FIG. 11, the electronic component 2 mounted on the package substrate 4 is a BGA type semiconductor package and is mounted on the circuit substrate 1. Between the circuit board 1 and the package substrate 4, a joint portion 6 that electrically connects the electronic component 2 and the circuit board 1 is disposed, and the joint portion 6 is positioned in a lattice shape. In addition, a detection junction 7 for predicting the lifetime of the junction 6 is disposed at the package corner. Detecting one or both of the two joints 7 for detection by damaging one or both of them by monitoring the electrical characteristics of the circuit. It is configured to be able to.

  Each of the detection joints 7 is designed so as to have a shorter life than the joints 6, and the life of each of the detection joints 7 is different. In Modification 6 shown in FIG. 11, the detection joint 7 is arranged around the corner of the electronic component 2 that generally has a large curvature change of the circuit board 1 and a large load. In addition, the detection joint 7 is designed to have a shorter life than the joint 6 by making the diameter of the joint 7 to be smaller than the diameter of the joint 6. Moreover, it is possible to make a difference in the lifetime of each pair of the detection joints 7 by making the sizes of the respective pairs of the detection joints 7 different.

  According to the sixth modification, the outer size including the detection joint 7 can be made compact.

  Further, the arrangement or the like of the detection joint 7 can be modified. FIG. 12 is a view for explaining the electronic component and its periphery according to the modified example 7, (a) is a cross-sectional view of the electronic component 2 and its periphery, and (b) is a joint portion 6 of the electronic component 2. The arrangement of the detection joint 7 is shown. As shown in FIG. 12, the detection joint 7 is designed to have a shorter life than the joint 6 by making the diameter of the joint 7 smaller than the diameter of the joint 6. Generally, the curvature change of the circuit board 1 is large around the corner of the electronic device 2, and the load is large. Therefore, in the present modification example 7, by disposing the pairs of detection joints 7 of the sets so that the distances from the package corners are different, the lifetimes of the pairs of the detection joints 7 are made different. is there.

  According to the seventh modification, it is easy to ensure the degree of freedom in layout design on the package substrate 4.

  Further, the arrangement or the like of the detection joint 7 can be modified. FIG. 13 is a view for explaining an electronic component and its periphery according to the modified example 8, (a) is a cross-sectional view of the electronic component 2 and its periphery, and (b) is a joint portion 6 of the electronic component 2. The arrangement | positioning of the junction part 7 for a detection is shown. Also in the present modification 8, the detection joints 7 are all designed to have a lower life than the joints 6 and have different lifetimes. In the present modification 8, the joint portion for detection 7 is arranged near the corner of the electronic component 2 that is generally heavily loaded, and the diameter of the joint portion 7 for detection is made smaller than the diameter of the joint portion 6, thereby Designed to have a lifespan lower than 6. Generally, the periphery of the fixing portion 8 for fixing the circuit board 1 is greatly deformed as compared to other regions. Therefore, by disposing the detection joints 7 so that the distances from the positions of the fixed portions 8 of the respective sets are different, it is possible to make a difference in the life of each set of the detection joints 7.

  This modification 8 also makes it easy to ensure the degree of freedom in layout design on the package substrate.

(Second Embodiment)
Next, a second embodiment will be described.

  In the present embodiment, when predicting a damage index of a joint in an electronic component, a phenomenon analysis or the like that simulates the usage state of the electronic device is used. FIG. 14 is a block diagram illustrating a schematic configuration of a damage index prediction system according to the second embodiment. This damage index prediction system can be incorporated inside the electronic device, or can be arranged and connected outside the electronic device. As shown in FIG. 14, the damage index prediction system 200 includes a damage detection unit 141 for a detection joint, a correction calculation unit 142 for correcting a fatigue characteristic database, a fatigue characteristic database 143, a joint 6, and a detection joint 7. A state parameter database 144 that stores state parameters, a calculation unit 145 of a predicted value of an index related to damage of the joint 6, and a display output unit 146 are provided. However, the display output unit 146 may not be provided.

  The damage detection unit 141 is for detecting damage to the detection joint 7. In the state parameter database 144, state parameters of the joint 6 and the detection joint 7 such as temperature, load, stress, displacement, and strain are stored in advance. When the damage detection information of the detection joint 7 is input from the damage detection unit 141, the correction calculation unit 142 receives index data relating to damage of the detection joint 7 and the detection joint 7 captured from the state parameter database 144. The fatigue characteristic database 143 is corrected based on the state parameters (details will be described later). In the fatigue characteristic database 143, fatigue characteristic data representing the relationship between the state parameters of the joint 6 and the indexes related to damage is accumulated. The fatigue parameter data representing the relationship between the state parameter of the joint 6 and the detection joint 7 from the state parameter database 144 and the relationship between the state parameter of the joint and the index related to damage is taken from the database 143, and the damage of the joint 6 is calculated by the calculation unit 145. Calculate the predicted value of the index. When the display output unit 146 is provided, a predicted value of an index related to damage to the joint 6 may be displayed. As a result, it is possible to predict a damage index that is more certain than the originally predicted value.

  FIG. 15 is a diagram illustrating a case in which, in the second embodiment, the state parameter of the joint is set to a strain range generated in the joint, for example, and the joint 6 is obtained from a phenomenon analysis such as FEM analysis when the index related to damage of the joint is a failure life. It is a block diagram which shows the outline of a structure at the time of constructing | assembling the state parameter database regarding the distortion range of the junction part 7 for a detection.

  A design information, material database 151, a phenomenon analysis module 152, and a strain range database (state parameter database) 153 of the joint 6 and the detection joint 7 are provided. In the design information and material database 151, information necessary for the phenomenon analysis, for example, the size and arrangement of each part, the material property value of each part, and the like are stored. The phenomenon analysis module 152 takes in information necessary for the phenomenon analysis from the database 151, performs a phenomenon analysis such as an FEM analysis simulating the use environment or an acceleration test, and calculates the strain range of each joint, and stores it in the database 153. accumulate.

  FIG. 16 shows a flow of a method for predicting the lifetime of the joint part 6 of the electronic component in the second embodiment. Here, the strain range database 153 of the joint 6 and the detection joint 7 shown in FIG. 15 is used as the state parameter database.

  The electrical characteristic value of the detection joint 7 is measured (S1601), and the breakage is detected when the electrical characteristic value of the detection joint 7 is outside the specified range (S1602). Here, as described above, an electrical resistance value, a capacitance value, or the like is used as the electrical characteristic value. When breakage of the detection joint 7 is detected (S1603), based on the actual damage life of the detection joint 7 and the strain range data of the detection joint 7 taken from the database 153, the fatigue characteristic database 143 Is corrected (S1604). Details of the correction method will be described later. Data on the strain range of the joint 6 from the database 153 and data on the relationship between the strain range and life of the joint from the fatigue property database 143 are calculated, and life prediction values of the joint 6 and the detection joint 7 are calculated ( S1605), the predicted life value of the joint 6 is displayed (S1606).

  According to the second embodiment, in addition to the same effects as those of the first embodiment, more accurate damage can be achieved by correcting a pre-constructed fatigue property database according to the use situation. Prediction becomes possible.

  In the present embodiment, the number of detection joints 7 is one, but it is also possible to provide a plurality of detection joints, and a more reliable damage index prediction is possible as compared with the case where one detection joint 7 is provided. It becomes possible.

FIG. 17 is a schematic diagram for explaining a method for correcting a database of fatigue characteristics of joints. The state parameter is the strain range Δε of the joint, the damage index is the failure life cycle number N _f , and the fatigue characteristic database data is the relational expression between the strain range Δε and the failure life cycle number N _f . The vertical axis in FIG. 17 represents the joint strain range Δε, and the horizontal axis represents the predicted value of the failure life cycle number N _f . Further, the black plot indicates data of the strain range Δε and the actual failure life cycle number N _f predicted from the information on the phenomenon analysis and the state of the circuit board, etc. of each of the broken detection joints 7. For example, fatigue characteristics of general metal materials include Coffin-Manson law for low cycle fatigue, Baskin law for high cycle fatigue, and a combination of Coffin-Manson law and Baskin law for both low cycle fatigue and high cycle fatigue. It can be expressed as In the present embodiment, it is assumed that the fatigue characteristic database has a relational expression between a strain range and a failure life that can be represented by one straight line as in the graph of FIG. In FIG. 17, f _initial is a straight line indicating fatigue characteristic data at the time of designing an electronic device, for example.

The straight line is horizontal for all plots so that each plot of strain range Δε and failure life N _f of broken joint 7 is on a straight line showing the relationship between strain range Δε and failure life cycle number N _f. Translate in the direction to obtain straight lines f ₁ to f _n as many as the number of plots. For the n straight lines f ₁ to f _n after the movement, the predicted failure life value N _{f1 of the} joint 6 estimated from the predicted strain range Δε _p of the joint 6 and the straight lines f ₁ to f _n at that time, respectively. Perform statistical processing on ~ N _fn . Calculate the failure life prediction value N _f of the joint 6 and show a straight line indicating the relationship between the strain range Δε and the failure life cycle number N _f so that the failure life prediction value becomes N _f when the strain range is Δε _p. The fatigue property database is modified by shifting in parallel.

For statistical processing, when the failure life prediction values N _f1 to N _fn are lognormally distributed, it is appropriate to set the average value as the failure life prediction value N _f . When the failure life prediction values N _f1 to N _fn have a Weibull distribution, it is suitable to set the median value as the failure life prediction value N _f . FIG. 17 is an example in which the median value of N _f1 to N _fn is N _f , and since N _f3 is the median value, the failure life prediction value N _f = N _f3 , the damage prediction model-based straight line is f ₃ In the above description, the straight line for predicting the damage index is translated in the horizontal direction of the graph shown in FIG. 17, but it is also possible to translate it in the vertical direction.
Next, a modification of the database correction method for fatigue characteristics will be described. Also in this modification, the damage index prediction database is expressed by a straight line. FIG. 18 is a schematic diagram for explaining a technique for correcting a fatigue characteristic database in the present modification.

The state parameter is the strain range Δε of the joint, the damage index is the failure life cycle number N _f , and the fatigue characteristic database data is a relational expression between the strain range Δε and the failure life N _f . One straight line f _initial as shown in the graph of FIG. 18 is used as fatigue characteristic data at the time of designing an electronic device, for example. Obtains a straight line f, which is determined by the least squares method of damaged each plot of the actual damage life N _f a strain range [Delta] [epsilon] of the detection joint 7 (black plot), the predicted strain range [Delta] [epsilon] _p and the straight junction 6 From f, the life expectancy value N _f of the joint 6 is calculated. Each time a failure of the detection joint 7 is detected, the data of the strain range Δε and the failure life N _f are expanded, the straight line f changes, and the fatigue characteristic database is corrected.

  Next, a modification of the second embodiment will be described. In this modification, information on the state measured during use of the electronic device is used when predicting the damage index of the joint in the electronic component. FIG. 19 shows the detection of the joint 6 in this modification. It is a block diagram which shows schematic structure for construction | assembly of the database 153 of the distortion | strain range of the junction part 7 for an object. This modification is the same as that of the second embodiment except for the method of constructing the strain range database 153 of the joint 6 and the detection joint 7. As shown in FIG. 19, a measurement unit 191 that measures information on the state of the circuit board, a design information database 192, a response surface database 193, a calculation unit 194 that calculates the strain range of the joint 6 and the detection joint 7; A strain range database 153 of the joint 6 and the detection joint 7 is provided, and data is accumulated in the database 153 in real time while the electronic device is used. The information measurement unit 191 relating to the state of the circuit board measures, for example, the temperature, strain, stress, acceleration, electrical resistance value, and the like of the circuit board. In the design information database 192, for example, the size and arrangement of each component are stored. The response surface database 193 stores response surface data for calculating the strain range of each joint from information related to the state of the circuit board and design information. The calculation unit 194 obtains information on the state of the circuit board from the measurement unit 191, design information from the design information database 192, information on the state of the circuit board from the response surface database 193, design information, and the relationship between the strain ranges of the joints. Response surface data to be expressed is taken in, the strain range of the joint 6 and the detection joint 7 is calculated, and stored in the database 153. This makes it possible to predict the lifetime more reliably than the initially predicted value.

The range of strain generated at the joint during use of an electronic device varies with each use because the use time and the surrounding environment during use are not the same. Therefore, the repetition of the strain range Δε that fluctuates according to the repeated use of the electronic device is converted into a constant strain range Δε _eq so that the damage for each use becomes equivalent, and Δε _eq is called the cumulative equivalent strain range. I will decide. In this modification, the fatigue characteristic data in the fatigue characteristic database is a relational expression between the cumulative equivalent strain range Δε _eq of the joint and the failure life, and the cumulative equivalent strain Δε _eq of the joint 6 and the cumulative equivalent strain range Δε _{eq of the} joint are Calculate the life prediction value from the relational expression of the broken life.

  According to this modification, the damage index prediction is performed in real time based on the data measured during use, and the database used for damage index prediction is updated as needed. It is possible to predict the damage index.

  According to this embodiment, during use of an electronic device, by predicting the life of the joint part of the electronic component based on the damage index of the joint part for detection, the environment and usage situation around the device, the individual device It is possible to predict damage index with high accuracy considering the difference.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

  DESCRIPTION OF SYMBOLS 1 ... Circuit board, 2 ... Electronic component, 3 ... Detection device, 4 ... Package substrate, 5 ... Detection device substrate, 6 ... Joint part, 7 ... Detection For joints, 11... Database about the relationship between the joints and the indicators for damage of the joints for detection, 12... Damage detection unit, 13... Arithmetic unit, 14. ..Damage index prediction system.

Claims (20)

Damage that predicts an index related to damage of an electronic device having a joint that electrically connects an electronic component to a circuit board for mounting, and a detection joint that is designed to have a lower lifetime than the joint. A prediction system,
A damage detection unit for obtaining information on damage of the detection joint;
From the relationship between the information on the detection joint damage obtained by the damage detection unit, the index on the detection joint damage, and the index on the joint damage, the predicted value of the index on the joint damage A computing unit for computing
A damage index prediction system characterized by comprising:   The damage index prediction system according to claim 1, further comprising a display output unit that displays the predicted value calculated by the calculation unit.   The damage index prediction system according to claim 1, wherein an index related to damage of the joint is a failure life, and the calculation unit calculates a life prediction value until the joint is damaged.   The index regarding the damage of the joint is a damage value represented by the reciprocal of the number of life cycles when the same repeated load is applied when a repeated load is applied for one cycle, or each cycle when a load is repeatedly generated. 3. The damage value obtained by accumulating the damage values generated, wherein the calculation unit calculates a predicted damage value of the joint at the time when the detection joint is broken. The described damage index prediction system. The detection joints are arranged at a plurality of locations,
3. The damage index according to claim 1, wherein the calculation unit calculates an average value of an index related to damage of the joint every time the joint for detection breaks, and corrects the predicted value. Prediction system. The detection joints are arranged at a plurality of locations,
The said calculating part calculates the lifetime prediction value of the lowest lifetime among the said lifetime prediction values, as the lifetime prediction value of the said junction, whenever the said junction for a detection breaks. Damage index prediction system.   7. The damage index prediction system according to claim 5, wherein the detection joints are formed so that loads applied thereto are different and strengths against repeated use of the electronic device are different. . Damage that predicts an index related to damage of an electronic device having a joint that electrically connects an electronic component to a circuit board for mounting, and a detection joint that is designed to have a lower lifetime than the joint. A prediction system,
A damage detection unit for obtaining information on damage of the detection joint;
Based on an index relating to damage of the detection joint from the information relating to damage of the detection joint acquired by the damage detection unit and a state parameter of the detection joint, the state parameter of the joint and the joint A correction calculation unit for correcting material property data representing a relationship with an index related to damage of
A calculation unit that takes in the state parameters of the joint and the modified material property data, and calculates a predicted value of an index related to damage of the joint;
A damage index prediction system characterized by comprising:   The damage index prediction system according to claim 8, further comprising a display output unit that displays an index related to damage of the joint calculated by the calculation unit.   10. The damage index prediction system according to claim 8, wherein the state parameter is at least one of temperature, load, stress, displacement, and strain.   The material characteristic data of the joint is corrected in accordance with the strain range of the joint predicted from information on the phenomenon analysis and information on the state of the circuit board and the index regarding the damage of the joint actually obtained. The damage index prediction system according to claim 8 or 9.   Near the corner of the electronic component, or when the electronic component is a semiconductor package, the detection junction is disposed on the back side of the circuit board near the chip corner, and the load on the detection junction is The damage index prediction system according to claim 1, wherein the damage index prediction system is larger than a load applied to the joint.   A plurality of the detection joints are provided so as to electrically connect the electronic component side and the circuit board side, and at least two are electrically connected via wiring provided on the circuit board. The damage index prediction system according to any one of claims 1, 2, 8, and 9, wherein the damage index prediction system is connected in series.   10. The detection joint portion is provided on the same plane as the electronic component placed on the circuit board via a package substrate. The damage index prediction system according to claim 1.   The two detection joints are arranged at a package corner portion as a set, and it is determined that one or both of the two joints are damaged, and it is determined that the detection joint is damaged. The damage index prediction system according to any one of claims 1, 2, 8, and 9.   The detection joint portion is smaller than the diameter of the joint portion, and is arranged so that the distance from the package corner portion of each detection joint portion is different. The described damage index prediction system.   The fixing part for fixing the circuit board is formed, and the distance from the position of the fixing part of each of the joints for detection of each set is different, respectively. Damage index prediction system. Damage prediction for predicting an index related to damage of an electronic device having a joint that electrically connects an electronic component to a circuit board for mounting, and a detection joint that is designed to have a lower life than the joint A method,
Obtain information on damage to the detection joint,
Prediction of the index relating to the damage of the joint from the relationship between the index relating to the damage of the joint for detection acquired by this damage detection, the index relating to the damage of the joint for detection and the index relating to the damage of the joint A damage index predicting method comprising calculating a value.   18. The damage index prediction system according to claim 1, wherein the information related to damage of the detection joint is detection information of breakage of the detection joint.   18. The damage index is any one of a failure life, a damage value, a function using the failure life, and a function using the damage value. The described damage index prediction system.

Images