JPH07114601A - Manufacturing defect analysis system, method, and database generation method related thereto - Google Patents

Abstract

(57) [Summary] [Purpose] Collectively manages manufacturing conditions in the manufacturing process, design conditions for each product, and defect records, and presents defect occurrence cause candidates based on the record data at the time of manufacturing defective products.
A process and a problematic manufacturing condition can be presented as the defect cause candidate. In addition, we will easily construct a database on the causal relationship between the cause and result of defects. [Structure] In the manufacturing failure analysis system, manufacturing conditions,
Perform a failure analysis using a check item database consisting of design conditions or failure tendency. As a user know-how registration function, the check item database is
Use the judgment history used for judgment when viewing the data of defective records so that it can be updated. Further, the candidates for the cause of the defect are narrowed down by using the non-defective product manufacturing conditions judged from the past defect rate.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a manufacturing defect analysis system and method, and more particularly to a system for analyzing the cause of manufacturing defects caused by improper manufacturing conditions or design conditions.

[0002]

2. Description of the Related Art Conventionally, when a manufacturing defect that requires investigation of the cause has occurred, the manufacturing conditions of the defective product have been interviewed or investigated based on documents throughout the manufacturing process, and the cause has been analyzed and analogized. .

Further, in the defective process extraction method of the LSI manufacturing process disclosed in Japanese Patent Laid-Open No. 2-234451, which analyzes the cause of a defect occurring in the LSI manufacturing process, troubles occurring in each process and the troubles are caused. There is described a method of creating a trouble database showing the correspondence of defects (wafers) that have occurred, and extracting defective processes based on the database.

[0004]

In the above-mentioned prior art, no consideration is given to the effective utilization of manufacturing conditions as data when a defective product is produced. Therefore, over the entire manufacturing process, we interviewed the manufacturing conditions of applicable defective products or investigated based on documents,
Since the cause was analyzed and analogized, it took a lot of time for the investigation, and there was a problem that the result of the investigation and the analysis was not utilized for the next problem solving.

Further, in the defective process extraction method of the LSI manufacturing process described in Japanese Patent Laid-Open No. 2-234451, which extracts a defective process based on a trouble database, the trouble in each manufacturing process and the defect are dealt with. Although a process in which a trouble has occurred can be selected as a process in which a defect has occurred, it is not possible to specify what manufacturing condition of the process caused the trouble in the process. Further, when no trouble occurs in the process, even if a defect is found in the subsequent inspection process, it is not possible to trace back the defect occurrence process. Further, there is a problem that it usually takes a lot of time to build a database regarding a causal relationship that stores the cause and result of a defect.

The present invention has been made in order to solve the above-mentioned problems of the prior art, and its purpose is to collectively manage the manufacturing conditions in the manufacturing process, the design conditions of each product, and the defect record, and to manage the defects. An object of the present invention is to provide a manufacturing defect analysis system that presents defect occurrence cause candidates based on performance data during product manufacturing. In particular, it is to provide a manufacturing defect analysis system capable of presenting a process and a manufacturing condition causing a problem as a candidate for the cause of the defect.

Another object of the present invention is to provide a manufacturing defect analysis system having means for easily constructing a database relating to a causal relationship which stores the correlation between the cause and the result of a defect used for defect analysis. Still another object of the present invention is to provide a manufacturing defect analysis system capable of quickly narrowing down candidates of defect causes in defect analysis.

[0008]

In order to achieve the above-mentioned object, the structure of the first invention relating to the manufacturing defect analysis system of the present invention is such that the cause of defect occurrence during the manufacture of a product is In a manufacturing failure analysis system that investigates from product design conditions, an analysis function that analyzes the cause of the failure, the cause of the failure, and the manufacturing condition management items or design specification items that may cause the failure or the tendency of the failure A check item database storing set check items, and the analysis function uses the check item database to check the manufacturing condition management item or design specification item or the tendency of the defect that may cause the defect from the defect occurrence record. Is searched and presented.

According to another configuration, the manufacturing defect analysis system has a defect record database and a manufacturing record database, and the analysis function uses the check item database to check the manufacturing condition management item or the manufacturing condition management item from the defect occurrence record. After searching the design specification item or its failure tendency, by comparing the searched check item with the defect record data stored in the defect record database and the manufacturing record stored in the manufacturing record database The candidate for the cause of the defect is presented.

Further, in the manufacturing defect analysis system, the cause of the defect and the manufacturing condition management item or design specification item or the tendency of the defect which may cause the defect are detected in the check item database. It is set for each process.

Further, in the manufacturing defect analysis system, the check item has a correlation value between a cause of the defect and a manufacturing condition management item or a design specification item or a tendency of the defect which may cause the defect. It was done like this.

In order to achieve the above object, the configuration of the second invention relating to the manufacturing failure analysis system of the present invention has a statistical display function for displaying statistical data consisting of manufacturing results or defective results instructed by a user, The type of statistical display screen that has been requested to be displayed, the judgment history management function that stores the contents or display conditions of the statistical data specified to perform the display, the tendency that the statistical display screen can show, and its tendency. It has a screen-specific reference content database that stores manufacturing condition management items and design specification items that may cause defects when existing, and the statistical display function is based on the screen-specific reference content database. A display unit that presents the tendency of the display contents shown in the judgment history recorded by using the judgment history management function when operating the, and the judgment result and judgment of the user. To have a user know-how registration function consisting of a judgment input section for selecting and inputting available items from the tendency and a check item updating section for updating a check item database related to the input based on the input. It was done.

In order to achieve the above object, the configuration of the first invention relating to the manufacturing failure analysis method of the present invention is the manufacturing condition management described in the check item database in any one of the manufacturing failure analysis systems. For items or design specifications, compare the standard value with the state when manufacturing a defective product, or compare the state when manufacturing a defective product with the good product manufacturing condition data that is the actual data when the same product was manufactured in the past, or manufacturing conditions, design The transition of the defect rate due to the change of the condition is checked, and the defect occurrence cause candidate is presented based on the check result.

More specifically, in the manufacturing failure analysis method, the defective rate of the product is obtained for each of the manufacturing conditions or the design conditions. If the defective rate is less than a certain value, the product manufactured under the manufacturing condition or the design condition. Is a product which is determined to be manufactured as a non-defective product and uses the non-defective product condition data.

In order to achieve the above-mentioned object, the configuration of the first invention relating to the database generation method related to the manufacturing defect analysis system of the present invention is such that in any one of the manufacturing defect analysis systems, the non-defective product manufacturing condition data is stored. A method of generating a good product manufacturing condition database to be stored, which is a manufacturing condition record database storing manufacturing condition change time and condition value of each manufacturing condition management item, and product manufacturing storing working time for each manufacturing process for a product. It has a result database and an assembly information database storing design specifications of each product, and from the non-defective product manufacturing condition database, the manufacturing condition result database and the assembly information database, manufacturing conditions for each manufacturing condition management item or design specification Alternatively, for each design condition, the number of products manufactured under that condition and the occurrence of defects Create a defect list for each manufacturing condition that stores the number of manufactured products, calculate the defect rate for each manufacturing condition or design condition from the list, and use the manufacturing condition that the defect ratio is within the allowable range as the good product manufacturing condition By storing the condition data, a non-defective product manufacturing condition database is generated.

[0016]

[Function] Since the defect analysis is performed using the check item database that sets the occurrence of defects and the manufacturing condition management items that can cause the defects, design specification items, and the tendency of the defects, actual data at the time of manufacturing defective products Based on, it is possible to present candidates for the cause of failure. In addition, the analysis function can present a process and a manufacturing condition that poses a problem as a defect cause candidate. Furthermore, the correlation value between the cause of the defect, the manufacturing condition management item, the design specification item and the tendency of the defect can be quantitatively evaluated by the correlation value of the check items.

A causal relationship in which a correlation between a defect cause and a result used for defect analysis is stored by a user know-how registration function including a judgment history management function, a reference content database for each screen, a judgment input section, and a check item updating section. Database construction becomes easy.

Further, the defect rate is calculated for each manufacturing condition by the non-defective item manufacturing condition database, and the defect rate can be removed from the candidates of the cause of the defect, which causes the defect. It is easy to narrow down the manufacturing conditions.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Each embodiment of the manufacturing defect analysis apparatus and method according to the present invention will be described below with reference to FIGS.

[I] First, referring to FIG. 2, an outline of the function and configuration of the apparatus as a system will be described. Figure 2
It is a block diagram showing the function and composition of a system.

The user know-how registration function 203 includes a judgment history display process, a valid reference content determination process, a check item registration advisability determination process, and a check item update process. The routine analysis execution function 204 includes a check item reading process, a check execution process, a cause candidate scoring process based on the check result, a result display process, and a check item database updating process.

The statistical display function 201 performs statistical processing and display of defect record data, manufacturing record data, work standard data, design standard data, and good product manufacturing condition data. The judgment history management function 202 stores the contents displayed by the statistical display function in the judgment history database. In the user know-how registration function, the screen used as a reference for registering check items by the user based on the judgment history data and the reference content data for each screen and the tendency that can be judged from the screen are presented, and the check items are judged according to the user's judgment. Update the database.

The analysis function executes a check on the designated failure according to the method registered in the check item database by using the defect record data, the manufacturing record data, the work standard data, the design standard data, and the good manufacturing condition data. From the result, the defect cause candidates are presented in descending order of possibility.

[II] Now, the concept of the present invention will be described with reference to FIG. 2 in accordance with the function and configuration of the present system outlined above.

First, let's focus on the check item database 207 shown in FIG. The check items to be stored in the check item database 207 are literally items that should be checked in the analysis of manufacturing defects. Specifically, as shown later in FIG. It is a pair of a cause, a process related to the defect cause, and a management item in the process, a design specification item related to the defect cause, and a tendency of defects related to the defect cause. Further, there is an item called a correlation value in order to numerically evaluate the strength of the connection between the process related to the defect cause and the management item, the design specification item related to the defect cause, or the tendency of the defect.

In the present invention, the check item database 207 plays a central role in the system, and the cause of the defect, the process related to the cause of the defect, and the management item or design specification item in the process are linked. Because
Since it becomes possible to point out the control items for each process for investigating the cause of the defect, it becomes possible to carry out the defect analysis effectively. Also,
You will be able to investigate the cause of defects and evaluate their tendency. Furthermore, the correlation value enables quantitative and accurate failure analysis.

The user know-how registration function 203 and the non-defective item manufacturing condition database 212 described below are also based on this check item database.

Next, the update of the check item database and the user know-how registration function will be described. The check item database 207 described above is used when analyzing the cause of the defect, but it is desirable that the check item database 207 be constantly updated to the latest one in order to match the actual condition of the manufacturing process or the manufacturing apparatus. Therefore, if there is a function that supports the update of the check item database by a rational means, it is easy for the user to use, and the diagnostic accuracy of failure analysis is improved. The user know-how registration function 203 is based on such a background, and its concept is as follows.

This system stores various data necessary for defect analysis as databases such as the defect record database 208 and the manufacturing record database 209.
Then, these can be referred to by the statistical display function 201 as necessary.

Further, the contents that the user wants to update the check item database are highly likely to be related to the contents judged by displaying them by the statistical display function 201. For example, when the information of the defect record database 208 regarding the position information of the defective part is displayed and judged, it is often the case that the check item regarding the position of the defective part is desired to be updated. Therefore, by accumulating this information in the judgment history database 205 and the reference database for each screen and presenting the check items related to the check item update in an easy-to-update form, the user's input, in a broader sense, the user It supports the accumulation of user know-how for defect analysis. Further, in the correlation value of the items in the above-mentioned check items, a function of increasing the correlation value when the user definitively selects a candidate of the cause of failure is also supported.

Finally, the non-defective product manufacturing condition database will be described. In general, the ratio of defective products to manufactured components, that is, the defective rate can be calculated under a certain manufacturing process and certain manufacturing conditions. When a defect occurs and which defect is actually caused by which manufacturing process or manufacturing condition, the product with a high defect rate is actually a candidate for the defect of the product.

However, when the defect rate is extremely low, it can be considered that the process and the manufacturing conditions are not defective, that is, the manufacturing conditions are good products. Therefore, when this defect rate is extremely low, in other words, when there is a high possibility that it is a non-defective product manufacturing condition, it is possible to reduce the number of defect cause candidates in advance by eliminating it from the defect cause candidates in order to clarify the defect cause early. It is a thing.

[III] A specific hardware configuration example of the manufacturing defect analysis system according to the present invention will be described below with reference to FIG. FIG. 3 is a diagram showing a specific hardware configuration example of the manufacturing defect analysis system according to the present invention.

The magnetic disk device 301 stores data such as a check item database, a defect record, and a manufacturing record. The processing device 302 performs various calculations for implementing the present invention.

[IV] This embodiment is a case where the manufacturing failure analysis apparatus and method according to the present invention is applied to analysis of a failure that occurs in printed circuit board mounting. First, a manufacturing process for mounting a printed board that is a target of failure analysis will be described with reference to FIG. FIG. 4 is a process diagram showing an example of a printed board mounting line.

The printed board mounting line includes a manufacturing process such as a printing process, a mounting process, an inserting process, a reflow soldering process, and a flow soldering process, and an inspection process such as an appearance inspection process, an in-circuit test, a function test, and solder correction. It consists of manual processes such as process, visual inspection process, and post-correction process.

The board is put in from the first step, the solder paste is applied on the board by the printing machine, the component is mounted by the mounting machine, the solder paste is melted by the reflow soldering device, and the component and the board are electrically connected. .

The appearance inspection device inspects the state of the solder, and has a defect in soldering such as insufficient solder, solder bridge or excess solder, and mounting defect such as component mounting position deviation and polarity difference (reverse attachment). To detect. In the solder repair process, a repair work person repairs a defect found in the visual inspection device or a defect found in a visual inspection process described later, that is, replacement of parts, remounting of solder, and the like. The inserter mounts the insert parts and is soldered by a flow soldering device. In the visual inspection process, an inspector inspects for soldering defects and assembling defects. In the post-installation-correction process, defects found in the visual inspection process are corrected, and manually mounted components are mounted and soldered. Furthermore, the in-circuit tester determines solder bridge, short open, and quality of mounted parts, and the function tester tests each function.

[V] In the manufacturing defect analysis apparatus and method according to the present invention, the defect results (the defective board, the defective part, the defective position, the defective content, etc.) are collected from the appearance inspection apparatus or the visual inspection process, The candidate for the cause of the defect is presented. Here, FIG. 5 shows an example of a defect phenomenon and a cause of defect in mounting a printed board according to an embodiment of the present invention. FIG. 5 is an explanatory diagram showing an example of a defect phenomenon and a cause of the defect in mounting a printed board according to an embodiment of the present invention.

Manufacturing conditions of each device when a defect occurs,
It is necessary to identify the cause of failure and take countermeasures by examining the matching of the board and its manufacturing conditions, what position the defect has occurred, and what causes each defect in the past.

[VI] In the following, FIG. 1, FIG. 6 to FIG.
4, the processing contents of the manufacturing failure analysis method according to the embodiment of the present invention will be described in detail. FIG. 1 is a diagram showing a flowchart of processing contents of a manufacturing failure analysis method according to an embodiment of the present invention. FIG. 6 is an example of the statistical display screen, which is a defective position display screen. FIG. 7 is an example of the statistical display screen, and is a Pareto chart for each board of the number of defects. FIG. 8 is an example of the statistical display screen, and is a transition diagram of the number of defective cases. Figure 9
In the example of the statistical display screen, it is a substrate history display screen showing when each substrate passed through which process. FIG. 10 is an example of the statistic display screen, which is a work record display screen showing a change history of manufacturing conditions of each device. FIG. 11 is a diagram showing an example of the judgment history database. FIG. 12 is a diagram showing an example of the reference content database for each screen. FIG. 13 is a diagram showing an example of the check item database. FIG. 14 is a diagram showing a flowchart of the processing contents of update of the check item database.

(A) System startup

[Step 1] Start the system.

[Step 2] The manufacturing defect analysis apparatus and method according to the present invention has a statistical display function and an analysis function. In this step, the user selects any one of the above functions. Step 3 to Step 10 when the statistics display function is selected, Step 11 when the analysis function is selected
To step 16 are performed.

(B) Statistics display function

[Step 3] The statistics display screen is selected. Specifically, it is as shown in FIGS.
Note that FIG. 6 shows the board, the mounted components, the defect occurrence position, and the defect occurrence number. The defect occurrence number may be indicated by changing the size of the circle indicating the defect position or by changing the color. it can.

[Step 4] The display contents and display range of the screen selected in step 3 are designated. In the defective position display screen of FIG. 6, the assembly drawing number, which is the type of the board, the defective generation period,
Specify the type of defect. The screen is displayed with the specified contents and range. In the Pareto diagram for each substrate in FIG. 7, the period of occurrence of defects and the type of defects are designated. In the transition diagram of the number of defects generated in FIG. 8, the assembly drawing number, which is the type of the board, the period of occurrence of the defect, and the type of the defect are designated. On the board history display screen of FIG. 9, a serial number for identifying each board is designated.
In the work result display screen of FIG. 10, a device (process) and management items of each device (process) are designated.

[Step 5] The screen number selected in step 3, the display contents and display range specified in step 4 are registered in the judgment history database. As shown in FIG. 11, the contents of the judgment history database correspond to the screen displayed by the user and the display contents such as the defect type and the assembly drawing number, the assembly drawing number, the manufacturing period, and the defect detection period. It is stored.

[Step 6] The screen is displayed according to the designations in steps 3 and 4. Steps 3 to 6 are repeated according to the user's designation. Only when the end of the statistical display function is selected, proceed to step 7.

(C) User know-how registration function

[Step 7] The user know-how registration function is activated.

[Step 8] A judgment history database,
The tendency estimated to be used by the user is displayed as a list from the reference content database for each screen in which the tendency that can be judged from the screen is registered for each display screen number. FIG. 12 shows an example of the reference content database for each screen.

[Step 9] Each screen (data) is displayed using the statistical display function, and the user inputs the result of examining the cause of the defect. The items to be input are what kind of defects were examined, that is, the types of defects, the boards, the defective parts, the estimated causes, and the items that could be used for estimating the cause of the defects. Items that can be used for estimating the cause of the defect are selected from the list of tendencies displayed in step 8.

[Step 10] The check item database is updated from the items input in step 9. The check item database will be described with reference to FIG.

The check items are set for each defect type, each mounting line, and each mounting process of the defective parts. This is because different types of defects have different causes of occurrence, and different mounting lines and different steps of mounting defective parts cause different devices or the like as the causes of occurrence.

The contents to be set are items to be checked,
Candidates for the cause of defects related to the check result,
It is the correlation value between them. Here, the items to be checked are manufacturing conditions in the process, design specifications, or a defect occurrence tendency such as a defect occurrence position.

If the item to be checked is the manufacturing condition in the process, the process number and the management item number indicating which manufacturing condition is the target are designated as the check contents. In FIG. 13, item numbers 1 to 3 correspond to this.

If the item to be checked is the design specification,
Specify the items to be checked from the design specification check items prepared by the system in advance. In FIG. 13, item number 6 corresponds to this. Therefore, what is specified as the check contents is the design specification code attached to each design specification check item.

If the item to be checked is a tendency, the item to be checked is designated from the tendency check items prepared in advance by the system. In FIG. 13, the item numbers 4 and 5 correspond to this. Therefore, the tendency code assigned to each tendency check item is designated as the check content.

Check item database is step 11
It is used in the analysis process of steps 16 to 16. The usage will be described in detail in step 13.

(C-1) Update of Check Item Database The procedure for updating the check item database will be described with reference to FIG.

[Step 10-1] From the name of the board (assembly drawing number) input in step 9, the assembly information of the board is searched, and the mounting line information of the board described in the assembly information is read.

[Step 10-2] From the assembly information,
Information on the mounting process of the defect-related component input in step 9 is read.

[Step 10-3] The check item when the defect designated in step 9 occurs in the mounting line and the mounting process is read from the existing check item database.

[Step 10-4] It is determined whether or not the combination of the item that can be used for estimating the cause of failure specified in step 9 and the estimated cause is the check item. If the check item already exists, the process of step 10-5 is performed. If the item is not included in the check items, the process of step 10-6 is performed.

[Step 10-5] The correlation value of the check item is increased.

[Step 10-6] It is determined whether or not the combination of the item used for the defect cause estimation designated in step 9 and the estimated cause is inconsistent with the existing check item.

If there is a contradiction, the user is requested to judge whether to add a new combination of the item that can be used for the defect cause estimation and the estimated cause as a check item.
Register as a check item only when it is input when added.

If there is no contradiction, it is unconditionally registered as a check item.

(D) Analysis function

[Step 11] If the analysis function is selected in step 2, the processing from this step is performed. In this step, the user specifies the assembly drawing number, defective part, and defective type as the analysis target.

[Step 12] The corresponding check item is read from the check item database on the basis of the assembly drawing number, defective part, and defect type designated in step 11.

[Step 13] For each of the check items, if the manufacturing condition in the process is designated as the check item, the manufacturing condition and the work standard of the substrate in which the defect occurs are manufactured under the condition. Comparison, comparison with conditions when the same substrate is manufactured in the past, and checking whether the manufacturing conditions have been changed immediately before the occurrence of a defect are checked. If, as a result, results such as those that were out of the standard, different from the past manufacturing conditions, or that the manufacturing conditions were changed immediately before the occurrence of a defect were obtained, it was set as related to the check item. Increase the score of the candidate for the cause of failure. The higher this score is, the higher the possibility of being a candidate for the cause of the defect occurrence.

If the tendency is designated as a check item, it is judged whether or not the tendency exists. If there is a tendency,
The score of the candidate of the cause of failure set as related to the check item is increased. The score is increased by the value set as the correlation value.

[Step 14] At the time when all the checks are completed, the scores of the candidates for the cause of each defect are compared and presented to the user.

[Step 15] The user determines the cause of the defect from the candidates and inputs the determination result.

[Step 16] The correlation value in the check item database is updated based on the user's judgment result.

[Step 17] Based on the analysis result, the content of the countermeasure corresponding to the cause of the defect is presented. Create a measure instruction according to the measure selected by the user.

[VII] In the following, the manufacturing failure analysis method according to one embodiment of the present invention will be specifically described with reference to FIGS.
Will be explained. 15 to 18 are flowcharts showing specific processing contents of the manufacturing failure analysis method according to the embodiment of the present invention. FIG. 19 is an example of the statistical display screen, which is a defective position display screen. FIG. 20 is an example of the statistic display screen, and is a board-based chart of the number of occurrences of defects. FIG. 21 is an example of a statistic display screen, which is a transition diagram of the number of defective cases. FIG. 22 is a diagram showing an example of the judgment history database.

(A) Narrowing down the cause of defects by the analysis function First, the steps of inputting a defect record and reading check items will be described with reference to FIG.

After the system is activated, the analysis function is selected, and as the defect record, the items “K1” of the substrate and the types 1501 and 1502 of the defect type “unsoldered” are selected.

Then, the system checks the "print speed of process 1" 1503, "reflow temperature of process 3" 1504, and "the defective parts are the same" which are the check items related to the defective performance item from the selected check item database. A tendency "1505 to concentrate on part symbols" is read into the system.

Next, the steps of the work standard check will be described with reference to FIG. This is to check whether the environment at the time of actual work was within the standard according to the check items read in the previous step.
Then, when it is out of the work standard, the user picks up the work environment of the process as a candidate of the cause of the defect.

In this example, according to the check item "printing speed of step 1" 1503, whether the printing speed of step 1 when the solder paste is printed on the defective board is within the predetermined work standard or not. Check.

As a procedure, the reference 1601 of "printing speed of step 1" is searched from the work standard data, and the date and time 1602 of solder paste printing on the corresponding board is searched from the board manufacturing record data. Furthermore, from the manufacturing condition data, 1
The manufacturing condition 1603 corresponding to the date and time of 602 is searched.

Thereafter, the actual value of the manufacturing condition 1602 and the reference value of the work standard 1601 are compared.

In this example, the standard 1601 of "print speed of step 1" is minimum = 13 and maximum = 15, while
Date and time of manufacture of the corresponding board (09:03, October 11, 1992)
The actual value of the manufacturing performance of 0 minutes) is 13, which is within the work standard range.

Similarly, "reflow temperature of step 3" is 150
The work standard of 4 is checked.

Next, referring to FIG. 17, the steps of the non-defective product manufacturing condition check and tendency check will be described.

The non-defective product manufacturing condition check is to check whether or not the actually performed manufacturing condition is a condition in which a defect is likely to occur in view of the non-defective product manufacturing condition data. Then, when it is out of the non-defective product manufacturing conditions, the user picks up the work environment of the process as a candidate for the cause of the defect. Here, the non-defective product manufacturing condition data is data on the condition when the same substrate was manufactured in the past and the defect occurrence rate for each condition.

In this example, a non-defective manufacturing condition check is performed on the print speed check item "print speed of step 1" 1503 of step 1 when solder paste printing is performed on a board having a defect.

As a procedure, data relating to the check item "print speed of step 1" 1503 is searched from the non-defective product manufacturing condition data. Then, an appropriate work condition is extracted from the defect occurrence rate for each condition of the non-defective product manufacturing condition data and a separately set allowable range, and is compared with the actual value obtained at the step of the work standard check of the previous step. Here, the allowable range is a target value for suppressing the defect occurrence rate within this range.

In this case, the minimum is 13.3 and the maximum is 1.
It is 4.7, which is out of the range when compared with the actual value = 13 obtained at the step of the work standard check of the previous step,
Since the actual manufacturing conditions do not apply to the non-defective product manufacturing conditions, this process is a candidate for the cause of the defect.

Similarly, regarding the check item "reflow temperature of step 3" 1504 when the defective board is applied to the solder reflow device, it is determined whether or not it is an appropriate working condition extracted from the good product manufacturing condition data. To check.

Next, the tendency is checked. In this example, the check item “probability that defective parts concentrate on the same part symbol” 1505 corresponds to this.

As a procedure, in order to check whether the defective parts are concentrated on the same part symbol, the defect ratios of all the parts mounted on the board are calculated for each part symbol, and the parts having the same difference in each defect ratio are obtained. Check if the difference is significant enough to be focused on the symbol. In this case, there is no significant difference, and it is determined that "the defective parts do not tend to be concentrated on the same part symbol".

Finally, referring to FIG. 18, from the above check results, candidates for defect causes are set and the analysis results are displayed.
The steps for registering will be described.

The evaluation points are used to narrow down the candidates for the cause of failure from the results of various checks. In other words, for each check item, if the actual value actually manufactured in each of the work standard check, non-defective product manufacturing check, and tendency check is outside the range of the work standard value, it is actually manufactured in that process. If the actual value is out of the allowable range of good product manufacturing conditions, or if it is determined that there is a tendency to be pointed out, the check item is added as it is likely to be the cause of the defect. The one with a high evaluation point is a candidate for the cause of failure.

The correlation value between the process and the defect cause candidate is evaluated in advance for each of the check items, and the process and the defect cause candidate having a high correlation value are held as the correlation value. By adding as an evaluation point, more reasonable candidates can be narrowed down.

In this example, the evaluation points are "Inadequate reflow temperature in step 3" is 20 points and "Print speed in step 1" is 5 points, and "Inadequate reflow temperature in step 3" is the cause of failure. It turns out that there is a high possibility.

As described above, the analyzed result can be stored in the system and displayed on the screen. The analysis result is fed back to the system by increasing the above-mentioned correlation value of the check items of the items which are candidates for the cause of the defect.

(B) User's Judgment and Judgment History Database Next, with reference to FIGS. 19 to 22, how to analyze the cause of the defect will be described with reference to the above analysis results.

The user can see the screen shown in FIG. 22 by displaying the defect position screen and inputting the board "K1" in question, the defect occurrence period, and the defect type "unsoldered". You can In this example, it may be determined that the defect occurrence position is biased to the right end surface of the substrate.

Similarly, it is judged from the Pareto diagram display screen for each substrate shown in FIG. 20 that the defective substrates are biased, and the transition diagram display screen for the number of defects shown in FIG. It will be judged that the number of defects is gradually increasing over time.

The above display contents are stored in the user judgment history database as shown in FIG. 23, as they are used for the judgment by the user.

(C) User know-how registration function In this system, the check item database can be updated by utilizing the know-how possessed by the user regarding failure analysis. This is the user know-how registration function.

To use this function, the user first displays the know-how data input screen. On the know-how data input screen, using the judgment history database of (B), the tendency that can be read from the statistical display screen used by the user is displayed.

For example, in the defect position display screen shown in FIG. 22, "the defect occurrence position is biased to the center (not biased)",
Displays a list of trends such as "Defective occurrence position is biased toward the edge (not biased)", "Defective occurrence part type is biased (not biased)", "Defective occurrence part symbol is biased (not biased)", etc. To be done.

At this time, for example, the fact that the squeegee in step 1 is worn is found, and "if the position where the defect is generated is biased to one end of the substrate, the squeegee in step 1 is worn. If you newly gained the knowledge that "There is a problem", from the know-how data input screen, defect type "unsoldered", board "K1", related component mounting process "process 1, estimated cause" squeegee wear ", defect cause estimation Input the item "Position of defect occurrence is biased toward the edge" that was available for the screen.

In this way, the check item database is updated according to the know-how of the user. The system checks the consistency between the input item and the contents of the item registered in the check item database in advance. In the case of inconsistency, the logical consistency of the check item database is maintained by inquiring which item should be adopted as the check item.

(D) Countermeasure Instructions for Cause of System Defects When the analysis result or user judgment is input, the contents of countermeasures according to the cause of the defects are displayed on the screen. The user can take countermeasures against the defect accordingly.

[VIII] The procedure for creating the non-defective manufacturing condition database will be described below with reference to FIGS. 23 and 24. FIG. 23 is a flowchart showing an outline of the procedure for creating the non-defective product manufacturing condition database. FIG. 24 shows non-defective products by manufacturing condition, which is the basis for creating a good condition manufacturing condition database.
It is the figure which showed the defective item number list data.

The non-defective product manufacturing condition database is a database in which the manufacturing conditions at the time of manufacturing without defects are accumulated and edited, and is created for each product type. The creation procedure is as follows.

[Step 1] The last update date of the non-defective product manufacturing condition database is searched.

[Step 2] The number of non-defective products and the number of defective products manufactured under each manufacturing condition are calculated based on the manufacturing results after the last update date and the assembly information composed of the mounting position, mounting direction, and part shape of each product. Count and update the list of non-defective / defective product numbers by manufacturing conditions.

At this time, according to the user's judgment, if the cause of the defect is clearly unrelated to the manufacturing conditions of the current target among the defective products, even if the defective product is found, the product is defective. Can be counted as a non-defective item.

[Step 3] The defect rate under each manufacturing condition is calculated based on the list data of the number of non-defective products / defective products by manufacturing condition.

[Step 4] Search for manufacturing conditions in which the defective rate is within the allowable range.

[Step 5] A range in which the manufacturing conditions in which the defective rate is less than or equal to the allowable range and which is continuous over the widest range is registered as non-defective product manufacturing conditions.

[IX] Another embodiment of the present invention will be described with reference to FIG. FIG. 25 is a diagram showing an outline of a correction / countermeasure work instruction system in mounting a printed board.

The present embodiment is an example in which the manufacturing defect analysis system according to the present invention is used as one function of a correction / countermeasure work instruction system for mounting printed boards.

The repair / countermeasure work instruction system includes a defect record collection function 2501 for collecting defect records from results of inspection in an inspection process by an inspection device or manually, and a correction work instruction for issuing a defect board correction work instruction based on the defect record. A function 2502, a work analysis support function 2503 that supports failure cause analysis based on the defect record, and a countermeasure work instruction function 2504 that issues a countermeasure work instruction for the defect cause found using the work analysis support function. .

When a defect is found in the correction / countermeasure work instruction system, a correction instruction is given to the defect, and the cause of the defect occurrence is pursued and countermeasures are taken.

A record of implementation of the countermeasures is accumulated and is useful for preventing recurrence of defects.

The means for realizing the defect record collecting function in this embodiment has, for example, a work instruction for the visual inspection process and a record collecting function, displays a component mounting diagram on the screen, and if a defect is found, a touch panel is used. There is a method of inputting the defective position via the.

Further, the correction work instruction function 2 in this embodiment
As a means for realizing 504, there is a method of displaying a component mounting diagram showing the correction position on the screen.

The manufacturing defect analysis system according to the present invention is used as the work analysis support function 2503.

[0126]

According to the present invention, the manufacturing condition in the manufacturing process, the design condition of each product, and the defect record are collectively managed, and the defect occurrence cause candidate is presented based on the record data at the time of manufacturing the defective product. A failure analysis device can be provided. In particular, it is possible to provide a manufacturing defect analysis device capable of presenting a process and a manufacturing condition causing a problem as the defect occurrence cause candidate.

Further, according to the present invention, it is possible to easily construct a database relating to a causal relationship that stores the correlation between the cause of failure and the result used in failure analysis.

Furthermore, in the failure analysis, it is possible to quickly narrow down the cause of the failure.

[Brief description of drawings]

FIG. 1 is a diagram showing a flowchart of processing contents of a manufacturing failure analysis method according to an embodiment of the present invention.

FIG. 2 is a block diagram showing the function and configuration of the system.

FIG. 3 is a diagram showing a specific hardware configuration example of a manufacturing defect analysis system according to the present invention.

FIG. 4 is a process chart showing an example of a printed board mounting line.

FIG. 5 is an explanatory diagram showing an example of a defect phenomenon and a cause of the defect in mounting a printed board according to an embodiment of the present invention.

FIG. 6 is an example of a statistical display screen, which is a defective position display screen.

FIG. 7 is an example of a statistical display screen, and is a Pareto chart for each board showing the number of defects.

FIG. 8 is an example of a statistical display screen and is a transition diagram of the number of defective occurrences.

FIG. 9 is an example of a statistical display screen, which is a substrate history display screen showing when each substrate passed through which process.

FIG. 10 is an example of a statistic display screen, and is a work record display screen showing a change history of manufacturing conditions of each device.

FIG. 11 is a diagram showing an example of a judgment history database.

FIG. 12 is a diagram showing an example of a reference content database for each screen.

FIG. 13 is a diagram showing an example of a check item database.

FIG. 14 is a diagram showing a flowchart of processing contents of update of a check item database.

FIG. 15 is a view showing a flowchart (part one) of specific processing contents of the manufacturing failure analysis method according to the embodiment of the present invention.

FIG. 16 is a view showing a flowchart (part 2) of specific processing contents of the manufacturing failure analysis method according to the embodiment of the present invention.

FIG. 17 is a view showing a flowchart (part 3) of specific processing contents of the manufacturing failure analysis method according to the embodiment of the present invention.

FIG. 18 is a view showing a flowchart (part 4) of specific processing contents of the manufacturing failure analysis method according to the embodiment of the present invention.

FIG. 19 is an example of a statistical display screen, which is a defective position display screen.

FIG. 20 is an example of a statistic display screen and is a board-by-board palette diagram of the number of defects.

FIG. 21 is a transition diagram of the number of defective occurrences in an example of a statistical display screen.

FIG. 22 is a diagram showing an example of a judgment history database.

FIG. 23 is a flowchart showing an outline of a procedure for creating a non-defective product manufacturing condition database.

FIG. 24 is a view showing data on the number of non-defective products / defective products by manufacturing condition, which is the basis for creating a non-defective product manufacturing condition database.

FIG. 25 is a diagram showing an outline of a correction / countermeasure work instruction system in mounting a printed board.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Tatsu Nomoto 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Kanagawa Manufacturing Technology Research Institute

Claims (8)

[Claims] 1. The cause of defects in manufacturing a product,
In a manufacturing failure analysis system that investigates from manufacturing conditions or product design conditions in the manufacturing process, an analysis function that analyzes the cause of the failure, the cause of the failure, and the manufacturing condition management items or design specification items that may cause the failure. Or with a check item database that stores check items that set the tendency of the defect, the analysis function by the check item database,
A manufacturing defect analysis system characterized by searching a manufacturing condition management item or a design specification item or a tendency of the defect, which may be a cause of the defect, from a defect occurrence record and presenting it. 2. A defect record database and a manufacturing record database, wherein the analysis function uses the check item database,
After searching the manufacturing condition management item or design specification item or the tendency of the defect from the defect occurrence record, the searched check item and the defect record data stored in the defect record database and the manufacturing record database are stored. By comparing the manufacturing results
The manufacturing defect analysis system according to claim 1, wherein candidates of defect causes are presented. 3. In the check item database, a cause of a defect and a manufacturing condition management item or a design specification item or a tendency of the defect that may cause the defect are set for each process in which the defect is found. The manufacturing defect analysis system according to claim 1 or 2, characterized in that 4. The check item has a correlation value between a cause of a defect and a manufacturing condition management item or a design specification item or a tendency of the defect that may cause the defect. 4. The manufacturing defect analysis system according to claim 3. 5. A statistical display function for displaying statistical data consisting of manufacturing results or defective results instructed by a user, a type of statistical display screen requested to be displayed, and the statistical data specified to perform the display. The judgment history management function that stores the contents or display conditions, the trends that can be displayed on the statistical display screen, and the manufacturing condition management items and design specification items that may cause defects when the trends exist. It has a stored reference content database for each screen, and based on the reference content database for each screen, the display shown in the judgment history recorded by using the judgment history management function when the statistical display function is operated. A display section that presents the tendency that shows the content, a judgment input section that selects and inputs the judgment result of the user and items that can be used for judgment from the tendency, and , Claims 1 to 4 either production failure analysis system according to characterized in that it has a user expertise registration function consisting of a check item updating unit that updates the check item database associated with the input. 6. The manufacturing defect analysis system according to claim 1, wherein the manufacturing condition management item or the design specification item described in the check item database has a reference value and a defective product manufacturing time. Comparison of conditions or comparison of defective product manufacturing condition data, which is the actual data when the same product was manufactured in the past, and the manufacturing condition data
A manufacturing defect analysis method characterized by checking the transition of the defect rate due to a change in design conditions and presenting defect occurrence cause candidates based on the check result. 7. A defective rate of a product is obtained for each manufacturing condition or design condition, and if the defective rate is less than a certain value, it is judged that the product manufactured under the manufacturing condition or design condition is manufactured as a good product. 7. The manufacturing failure analysis method according to claim 6, wherein the non-defective condition data is used. 8. The manufacturing failure analysis system according to claim 1, which is a method for generating a non-defective product manufacturing condition database storing the non-defective product manufacturing condition data, wherein each manufacturing condition management item is manufactured. A condition change time and a manufacturing condition record database storing condition values, a product manufacture record database storing work times for each manufacturing process for a product, and an assembly information database storing design specifications of each product, From the non-defective manufacturing condition database, the manufacturing condition record database, and the assembly information database, the number of products manufactured under each manufacturing condition or design condition for each manufacturing condition management item or design specification, and the product in which a defect occurs Create a defect list for each manufacturing condition that stores the number of products, and use that list to A non-defective item manufacturing condition database generating method, which calculates a defective rate for each total condition, and stores the non-defective item manufacturing condition data with a manufacturing condition in which the defective rate is within an allowable range as a non-defective item manufacturing condition.