JPH11298200A - Method and apparatus for automatic quality control in process of mounting components on printed board - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method and apparatus for automatic quality control in process of mounting components on a printed board, enabling an optimum system operation in case a plurality of processes are executed one after another. SOLUTION: In the printed board components mounting execution process, independent solder print process, components mounting process and soldering process are made one after another and the quality control is made every process, measured data in each process which has dealt with defects are previously stored in a detabase, and the detabase is referenced, utilizing the measured data, if a defect appears, thereby determining corresponding dealing with the defect and dealing with it. Thus the automatic quality control suppressing the appearance of defects is enabled, without taking care of the correlation exceeding regulations and separation among processes.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic quality control method and apparatus in a component mounting process of a printed circuit board.

[0002]

2. Description of the Related Art Various types of quality control systems have been conventionally proposed.
Japanese Unexamined Patent Application Publication No. 2 and JP-A-9-51200 are known. The invention disclosed in the former one is
The graph is output so that quality control can be carried out as soon as possible when an upward trend or a downward trend is observed even within the control limits.The invention disclosed in the latter is based on inspection result data and repair result data. Then, the correlation between the defect rates is obtained, and the inspection standard is tuned.

[0003]

In any of the above-mentioned conventional quality control systems, the processing result in one process simply falls within a certain reference range. On the other hand, in a normal operation, a plurality of processes are sequentially performed. In such a case, each process can be managed by the conventional inter-process quality control method. However, even if a slight shift in an upstream process may cause a major problem in a later process, such a problem cannot be located.
In addition, if the quality of the upstream process is excessively controlled under such a situation, the yield is deteriorated.

The present invention has been made in view of the above problems, and has an automatic quality control method in a printed circuit board component mounting process capable of performing optimal system operation when a plurality of processes are sequentially performed. And the provision of such a device.

[0005]

In order to achieve the above object, the invention according to claim 1 is a component mounting method for a printed circuit board when a plurality of processes whose quality is individually judged based on actually measured values are sequentially executed. This is an automatic quality control method for the process, where the quality judgment of the downstream process is correlated with the quality judgment of the upstream process and the correspondence between the cause-inherited location when quality degradation is judged by the quality judgment of the downstream process. A database is prepared as a relationship, and based on the quality judgment of each process, the same database is referred to to find a cause internal location of a correlated upstream process and feed it back.

[0006] In quality control of a single process, a processing result is actually measured in the process, and a processing step in the process is fed back based on the measured value. Although the correlation is complicated but has a causal relationship with each other, it can be said that feedback is possible. However, in the case of a separate and unique process, it is difficult to find a causal relationship only from measured data. However, in the invention according to claim 1 configured as described above, the correspondence between the quality determination in the downstream process, the quality determination in the upstream process, and the location of the cause is previously stored in a database as a correlation. In other words, even if it is impossible to derive the correlation from the quality judgment itself, it is sufficient to store the actual situation if the association of events is stored in a database. Then, the database is referred to using the quality judgment of each process as a key. Since this database stores the quality judgment of each process when the quality deterioration is judged by the quality judgment,
If there is a match with reference to each process based on the quality judgment, it means a situation in which quality deterioration is judged in a downstream process. Therefore, a cause inherent location in that case is also identified. If the location within the cause can be ascertained, feedback processing will be performed there.

That is, by referring to a database prepared in advance using actual measured values as keys,
Regardless of the difficulty of estimating the correlation based on the causal relationship, it is possible to feed back to the problem internal part having the actual correlation. The process referred to here means some kind of processing step, and any process may be used as long as an actually measured numerical value is measured for the result of this processing step and feedback is possible based on the measurement result. This feedback can preferably be tuned automatically in the process, but does not prevent human intervention or the like.

[0008] The database stores the correspondence of each quality judgment when quality deterioration is judged by the quality judgment of the downstream process, but this judgment of quality deterioration does not necessarily mean only when the quality is poor. Not relative. That is, even if the quality judgment is lowered, it naturally includes that the quality is extremely excellent in the original situation. Therefore, in other words, it means that it is used for improving the quality. For example, the feedback may be made to the location inherent in the cause so as not to reduce the quality, and the feedback may be made to the location inherent in the cause so as to improve the quality comprehensively.

It is sufficient that the database itself is capable of storing a certain correspondence in association with each other, and it is appropriate to construct the database from a computer or the like from the necessity of judgment.
Of course, not only database searches that match exactly, but also those that use fuzzy control and neural network control to derive reasonable results can be used. It is also possible to use a computer.

[0010] The correspondence between the quality judgment in the downstream process, the quality judgment in the upstream process, and the cause-inherited location is stored in a database. Is not necessarily determined.
On the other hand, the cause-inherited location may be any location that can provide information necessary for feedback. As a preferred example in consideration of the specific response to the feedback as described above, the invention according to claim 2 is a method for classifying items in the automatic quality control method according to claim 1 when a defect is processed in a downstream process. Then, the tendency of the actually measured numerical value in each process upstream of the defect target for each defect treatment is configured as a database.

In the invention according to claim 2 configured as described above, attention is paid to a case where a defective process is performed in a downstream process. It is not an exaggeration to say that feedback cannot be provided unless a defective treatment is performed, so the itemization of the defect treatment corresponds to the itemization of the feedback. In addition, when a defective treatment is performed, the situation related to such an event should surely appear. Therefore, by referring to the database using the tendency of the actually measured numerical value before the defect treatment, it is possible to determine the item in which the corresponding tendency is seen, and to locate the problem-inherited part assuming that such a defect treatment is taken. Then, feedback corresponding to the defective treatment is performed.

According to a third aspect of the present invention, in the automatic quality control method according to the second aspect, the tendency of the measured values used as the database is determined by comparing the distribution of the measured values and the standard measured values with respect to the defect. It is configured as a distribution shift. In the invention according to claim 3 configured as described above, even if the measured values are within the reference range, the distribution of some measured values may deviate from the distribution of the standard measured values. . Then, it is possible to determine that a situation in which it is necessary to take a fault treatment eventually may occur only when there is a deviation in a certain correspondence relationship as stored in the database.

Needless to say, such deviations are various, and the average value of the distribution may be shifted upward or shifted downward, the width of the distribution may be changed, or the period of the variation may vary. Various tendencies can be adopted, for example, such as It is sufficient to provide feedback so that the quality is not degraded if the problem internal part is located. As an example of the feedback, the invention according to Claim 4 is based on the automatic quality management method according to any one of Claims 1 to 3. When feeding back to the upstream process, the processing steps are tuned in the upstream process.

[0014] In the invention according to claim 4 configured as described above, in response to a case where the target of the processing step itself is wrong, the tuning of the processing step is performed so that a problem does not occur in a downstream process. I do.

The automatic quality control method can be applied to various fields and is not particularly limited, but as an example of the application, the invention according to claim 5 includes a plurality of manufacturing steps and includes An automatic quality control device in a component mounting process of a printed circuit board, comprising: an inspection step of measuring a processing result corresponding to a manufacturing process and determining the quality of the processing result; and a quality determination result in each inspection step. The feedback mode is stored in a database together with the measurement data in the downstream and upstream inspection processes, which are determined to be defective, and the upstream for problems that occur downstream by referring to the same database based on the measurement data in each inspection process In this configuration, the cause of the problem in the manufacturing process is located and fed back.

In the invention according to claim 5 configured as described above, the component mounting process of the printed board includes a plurality of manufacturing processes, and the inspection process corresponding to each manufacturing process measures the processing result and performs the same process. Quality control is performed by judging the quality of the result. In this case, since the feedback mode is stored together with the measurement data in the downstream and upstream inspection processes in a database, although the quality is determined to be defective in the quality determination result in each inspection process, based on the measurement data obtained in each inspection process. By referring to the database, it is possible to identify the location of the cause in the upstream manufacturing process. Therefore, the problem occurrence can be prevented by feeding back the problem occurrence in the downstream in advance.

[0017]

As described above, according to the present invention, when a plurality of processes are sequentially performed, it is possible to achieve a good quality control by eliminating the problems inherent in the processes which cannot be prevented only by managing each process. Can provide an automatic quality control method capable of performing the above. Claim 2
According to the invention, feedback is facilitated because the item corresponds to an item that is actually treated as a defect. further,
According to the third aspect of the present invention, since the tendency of the deviation of the actually measured value when the defective treatment is taken is stored in the database, it is possible to relatively easily search even if the actually measured value fluctuates by matching the tendency. it can.

Further, according to the invention according to claim 4,
Feedback is realized as tuning, which improves not only the cause but also the workability. Furthermore, according to the invention according to claim 5, it is possible to perform good quality control in a component mounting process of a printed board composed of a plurality of processes.

[0019]

Embodiments of the present invention will be described below with reference to the drawings. <Schematic Description of Overall Configuration> FIG. 1 is a schematic block diagram showing a configuration when an automatic quality control method according to an embodiment of the present invention is applied to a printed circuit board component mounting process. In this component mounting process, a solder printing process (process) for printing cream solder on a substrate and a component mounting process (process) for mounting electronic components at predetermined positions on the substrate.
And a soldering step of heating in an oven to melt the cream solder and solder.

In FIG. 1, a substrate PB on which a circuit pattern is printed is a solder printing device 11 corresponding to a solder printing process.
(Corresponding to the manufacturing process of claim 5), and the solder printing apparatus 11 prints a predetermined amount of cream solder on the pad positions on the substrate PB. Solder printing process is solder print inspection device 1
2 is provided to check whether or not the printing operation of the cream solder in the solder printing apparatus 11 is correctly performed.
Although the details of the inspection performed by the solder printing inspection device 12 will be described later, the predetermined measurement is performed and the measurement result is output to the management personal computer (PC1) 41.

The board PB on which the cream solder is printed is supplied to a component mounting device 21 corresponding to a component mounting process, and the component mounting device 21 converts each electronic component so that the electronic component can be soldered with the cream solder printed on the pad. Mount the parts at the designated positions. Component mounting process is mounted component inspection equipment 2
2 for inspecting whether or not the electronic component mounting work in the component mounting apparatus 21 is performed accurately. Although the details of the inspection performed by the mounted component inspection device 22 will be described later, predetermined measurement is performed and the measurement result is output to the management personal computer (PC2) 42.

Finally, the substrate PB is supplied to a soldering device 31 constituting a soldering process.
Heats the substrate PB in an oven to melt the cream solder and solder the pads and the leads and electrodes of the electronic component. In the soldering step, a soldering inspection device 32 is provided, and it is inspected whether or not the soldering operation by the soldering device 31 is performed accurately. The details of the inspection performed by the soldering inspection device 32 will be described later.
C3) is output to 43.

The management personal computers 41 to 43 are connected to the inspection devices 12 and 2 via a network or a parallel port.
2, 32, and converts the measurement data obtained by the communication into a predetermined format and stores it. Each of the management personal computers 41 to 43 is connected to a LAN line 44,
The above measurement data can be referred to from other personal computers. A process control device 45 is also connected to the LAN line 44, and the above-mentioned management personal computers 41 to 4 are connected.
3 to read the measured data and comprehensively and automatically control the quality of the printed circuit board component mounting process. This quality control not only provides various visual displays such as a Pareto chart to the administrator, but also the solder printing apparatus 1.
1 and the component mounting device 21 and the soldering device 31 are actively tuned.

FIG. 2 shows a specific configuration of the process control device 45, in which a control unit 45a for performing arithmetic processing and a bus 4 are provided.
5b, a program memory 45c accessible from the control unit 45a via the bus 45b,
And a data memory 45d accessible from the control unit 45a via the LAN 45 and the bus 45b to allow the control unit 45a to access the management personal computers 41 to 43 via the LAN 45 and the bus 45b. And an output data interface unit 45f connected to the bus 45b and enabling the control unit 45a to access the solder printing device 11, the component mounting device 21, and the soldering device 31.
And

The program memory 45c stores a program to be executed by the control unit 45a.
Not only M but also a RAM for temporarily storing an execution program read from an external auxiliary storage device or the like as appropriate. The contents of the program include a basic operating system, a quality management program described later, various drivers, and the like. Data memory 4
Reference numeral 5d stores various data used in the quality management program, and more specifically stores a measurement data file, a history data file, a correlation data file, and the like. The input data interface unit 45e is an interface adapter for substantially connecting to the LAN line 44. By executing the above-described driver program, the control unit 45a allows the management personal computer 4
Measurement data files and the like stored in auxiliary storage devices 1 to 43 and the like can be read. The output data interface unit 45f includes a network port and a parallel port that can be connected to the solder printing device 11, the component mounting device 21, and the soldering device 31, and outputs predetermined control data so that fine tuning can be performed. It has become.

Control data for performing tuning is generated by the quality management program, and FIGS. 3 to 5 show schematic flowcharts thereof. FIG. 3 is a main flowchart. During production, a program for preventing the occurrence of defects in each step shown in FIG. 4 is executed in step ST1, and a program for reducing the defect rate by step improvement shown in FIG. Repeat what you are doing. When it is determined in step ST3 that the production has ended, the execution of the program ends.

Here, the program for preventing the occurrence of defects in each step executed in step ST1 controls the production and inspection in one process to obtain a certain quality by executing the manufacturing and inspection. This is the conventional concept. More specifically, based on the measurement data obtained by inspecting whether or not the printing operation of the cream solder in the solder printing device 11 is correctly performed by the solder printing inspection device 12, the same solder printing device 11 is used. The component mounting apparatus 2 is tuned, and whether the electronic component mounting work in the component mounting apparatus 21 is correctly performed or not is determined by the mounted component inspection apparatus 22 based on measurement data obtained by the inspection.
1 and tunes the soldering device 31 based on measurement data obtained by inspecting whether the soldering operation by the soldering device 31 is performed correctly by the soldering inspection device 32. Thus, quality control is performed individually for each process.

On the other hand, the defect rate reduction program by the process improvement executed in step ST2 is to comprehensively perform quality control based on the inherent correlation when a plurality of processes (processes) are sequentially performed. . In the case of the above example, although each operation in the solder printing apparatus 11, the component mounting apparatus 21, and the soldering apparatus 31 is quality-controlled in each case, it should not be able to affect the operation in the first place. In practice, however, a correlation often occurs. It is intended to respond to such a fact from the front and make improvements.

<Detailed Description of Each Configuration> FIG. 6 is a model diagram showing the inspection principle of the solder print inspection apparatus 12. This inspection principle is called light-section three-dimensional measurement, in which the substrate PB is irradiated with laser light to perform main scanning in a certain direction (L1, L2).
2... Ln) and a sub-scan is performed at a constant sweep pitch, and the laser scan is photographed by a measuring camera. Substrate P
On B, the height of the material surface is used as the height reference, and the change in height of the protrusion on the substrate PB can be measured based on the trajectory of the scattered light from the portion irradiated with the laser light. Cross-sectional area S1 based on height change
It can be understood that the volume V
(= ΣS1) is also required. In addition, it is possible to determine the positional deviation of the protrusion as well as the area S2 of the portion having a certain height or more.

The solder printing inspection device 12 is a solder printing device 11
Since the printed result is inspected, the protrusion on the substrate PB is cream solder. Then, the measurement data represents the volume, area, and displacement of the cream solder. Of course, these are output to the management personal computer 41 as measurement data. Next, FIG. 7 is a model diagram showing the inspection principle of the mounted component inspection device 22. Each part having the part mounted at an accurate position is photographed as a model, and the image is stored as digital data in a memory. On the other hand, when the electronic component is mounted on each substrate PB by the component mounting apparatus 21, the electronic component is photographed as a work, and the image of the work is compared with the image of the model stored in the memory to determine the degree of coincidence.

In comparing images as digital data, the degree of coincidence is determined by a numerical value. If the degree of coincidence as a whole is high, it can be determined that the product is good, and if it is low, it can be determined that there is some defect. Further, it is possible to determine the component deviation in the (X, Y) coordinate system by the gray scale pattern matching method. As described above, in the mounted component inspection apparatus 22, the component deviation amount (and the missing item not shown) shown in FIG. 8 is output to the management personal computer 42 as the measurement data.

FIGS. 9 to 11 show the inspection principle of the soldering inspection device 32 by a model diagram. Above the XY stage, a light-receiving cell is arranged, a box-shaped sensor having an opening on the lower surface is arranged to face each other, and a laser beam is emitted vertically downward from the center of the sensor. The XY stage can be precisely moved in the two-dimensional direction, and scans the laser beam in a certain direction on the substrate PB as shown in FIG. When the laser light is applied to a material such as a lead or a solder on the substrate PB, the laser light is reflected with a high reflectance, and irradiates the inner surface of the box-shaped sensor that opens upward and faces downward. Independent light receiving cells are arranged on the inner surface of the sensor depending on the position, so that it is possible to detect which light receiving cell is irradiated with the reflected light. The change in the light receiving cell to be irradiated can very accurately represent the shape of the solder. FIG. 11 illustrates this clearly.

FIG. 11 (a) shows the shape of a good solder. The leads are arranged in close contact with the pads, and if the amount of cream solder is appropriate, the solder spreads over two orthogonal metal surfaces due to wettability. Attempts are made to form the concave surface of the solder. On the other hand, as shown in FIG. 11B, in the case of a defective product in which the lead is lifted off the pad, the lead becomes a polka dot due to the low wettability of the periphery of the pad and the surface tension of the solder itself. In this case, the surface shape is convex. When laser scanning is performed on the premise of this, the locus of the reflected light greatly changes depending on whether the surface is concave or convex. If the identification number of the light receiving cell shown in FIG. 10 is used as it is, the former becomes “6-5-4-3-
2 ", and the latter changes to" 2-3-4-5-6 ". In more detail, it can be seen that it follows "665544433322" or "662223444455566". Here, it can be easily understood that the height of the solder can be obtained from a change in a series of numbers sampled at a constant pitch representing the inclination of the solder surface.

Using this principle, the soldering inspection device 3
2 measures the shape of the solder (S) that joins the lead (L) and the pad (P) as shown in FIG. Here, it is preferable to determine whether the soldering is good or not, for example, the height of the fillet or the height of the solder, and these are output to the management personal computer 43 as measurement data. As described above, the measurement data obtained by the inspection device for each process is stored in the management personal computers 41 to 43, and the process control device 45 stores the quality management programs stored in the program memory 45c shown in FIGS. In the process of executing the above, the above-mentioned measurement data is read via the input data interface unit 45e, and the following quality control is executed while expanding the data in the data memory 45d.

The program for preventing the occurrence of defects in each process executed in step ST1 performs the following processing. In the case of the solder printing process (process No. 1), the management personal computer 41 as described above in step ST11.
, The average value and the variation are calculated and stored in a history file in step ST12. In step ST13, a time-series change regarding such an average value and variation is monitored. This is generally the same as monitoring a control chart. In step ST14, it is determined whether or not the transition of the average value or variation is about to exceed the control limit.

If it is determined that the control limit is about to be exceeded, the parameters of the manufacturing apparatus (solder printing apparatus 11) are tuned in step ST15. Hereinafter, the process is executed in a loop until a series of measurement data for the measurement items in the solder printing process is determined. Steps ST14 and ST15 provide feedback in the sense that parameters of the manufacturing apparatus are changed based on the measurement data, thereby preventing the occurrence of defects. However, the feedback here is for those with a clear causal relationship.

FIG. 13 shows the solder volume in the form of a histogram for the purpose of explanation. In FIGS. 13A and 13B, the variation exceeds the upper and lower limits of the standard. In this case, in the R control chart showing the variation, the data of the variation changes near the control limit or toward the control limit, and the average value of the X-bar control chart (for convenience of notation, expressed as follows) is obtained. By determining that the print amount has increased or decreased, for example, the print speed is tuned to adjust the print amount. Similarly, in the component mounting process (process No. 2) and the soldering process (process No. 3), in order to prevent the occurrence of defects, before the average value or variation of the measured data is likely to exceed the control limit. Adjust the parameters of the manufacturing equipment to provide feedback.

Originally, if quality control is performed for each process as described above, the occurrence of defects should be suppressed beforehand, and high quality products should always be manufactured. However, in reality, defects often occur continuously, and in such a case, the problem is often solved by tuning an upstream process. This can clearly be said to be a proof that the processes have a correlation. However, as shown in FIG. 12 (b), when the quality is controlled within the control limits with little variation, the problem is inherent. The location is not readily apparent.

However, to give an easy-to-understand example, the volume of the cream solder is within the control limit but large, and the component mounting position is within the control limit but in a certain direction (A in FIG. 12). Direction). In this case, the upper lead, which has a lot of cream solder, enlarges the pad, so that it can be assumed that the solder is likely to overflow, and as a result, it is understood that a defect of the bridge is easily found in the soldering inspection process.

On the other hand, even if such a causal relationship is not known, tuning is performed to reduce the volume of cream solder as long as it is a bridge. That is, tuning such as "reduction of squeegee pressure" is actually performed as a "defect treatment" for a defect called "bridge". In the present embodiment, it is assumed that various defect treatments are performed in advance, and the measurement data at that time is stored in a database. In addition, since the analysis is simply difficult when the measurement data is simply arranged, the tendency of the measurement data when compared with a standard distribution for defining the control limit is stored in a database. More specifically, assuming that the standard distribution is called a standard,
Is it a "larger tendency" or a "smaller tendency" than the standard? Is it a "larger tendency" or a "central tendency" as a larger tendency? Is it “a tendency to have a large variation” or “a tendency to have a small variation” compared to the variation of the standard? Also, is there a tendency to see periodicity in the variation?
Grouping like that.

FIG. 14 shows a tendency when the measurement data in each process corresponding to each item is compared with the standard while the defect judgment is classified into items according to the defect treatment. According to the database, when a certain defect treatment is performed, “average solder amount ≧ standard” in the “solder printing process”, “average position deviation ≧ standard” in the component mounting process, and “solder height ≧ standard” in the soldering process. Standard) and "fillet height ≤ standard"
It is assumed that there was a tendency. Therefore, the case where this defect treatment is required is recorded in the database as the defect determination “No. 001”. In addition, when another defect treatment is performed, only in the solder printing process,
The other tendency is No. Assuming that this is the same as the case of 001, the case where this defect treatment is required is recorded in the database as the defect determination “No. 002”. Hereinafter, similarly, the comparison between the measurement data in each case and the standard is recorded for each defect treatment.

If such a database is prepared, quality control is performed for each process, and if a defect occurs despite the fact that the quality is maintained within the control limit, the tendency of the respective measurement data is considered. Refers to the same database using as a key. If there is a corresponding defect determination, it can be said that the problem is inherent therein, and furthermore, since the corresponding defect treatment is known, tuning can be easily performed. The defect rate reduction program by process improvement actually executes the above processing. In step ST21, the measurement data recorded in the previous step ST1 is read, and used in step ST22 to calculate a defect and a defect rate. Then, the calculated failure and the failure rate are stored in a history file.

In the following step ST23, a time-series change relating to the defect and the defect rate is monitored, and it is determined in a step ST24 whether the defect continuously appears and the change in the defect rate is in a process unmanaged state. I do. In other words, this is a case where the failure rate occurs continuously and the failure rate is not correlated even though the tuning process is performed in each process. In such a case, it can be said that the correlation beyond the process is caused, and therefore, step ST2
In step 5, the causal relationship is specified by analyzing the measurement data of the previous process corresponding to the defective portion. Of course, here the causal relationship is not identified by numerical analysis, but by analyzing the measured data and analyzing its tendency, and using the analysis results to refer to the database described above, This means determining defective treatment while locating it.

If the defect treatment is found, the parameters of the manufacturing apparatus are changed in step ST26. After changing the parameters, the process stands by in step ST27 until measurement data of the next new lot is collected, and in step ST28, it is determined whether or not the defect rate has decreased compared to the immediately preceding lot. If it has decreased, it is only necessary to end the process assuming that the defect treatment is effective, and if not, the process returns to step ST26 to perform the defect treatment again. This is because it is possible that the first failure treatment alone does not reach an effect. In any case, when the defect rate is reduced, the defect processing is finished and the defect rate reduction program by process improvement is ended. Note that the above-described defect rate reduction program includes a variation suppression program including correlation data relating to an increase in variation. Further, in this example, the component mounting process has been described as one continuous line. However, each step may be arranged independently, and the relay (supply and storage) of the substrate may be executed by the substrate stocker.

Further, in this example, the tendency of comparison between the measured data and the standard is uniquely determined and compiled into a database, but a more fuzzy determination may be made. FIG.
Reference numeral 5 indicates that the number of analysis items of the measurement data is increased, and when this database is referred to, the total point is calculated by using more information to perform more optimal defect judgment. In this case, each evaluation may be weighted to take into account analog computer processing.
FIG. 16 shows a correlation between a defect factor in reflow soldering and specific unnecessary contents. For example, the causes of the defect content of the unstable amount of solder include “alignment accuracy of the printing apparatus”, “squeegee angle, printing pressure”, and “printing (squeegee)”.
Although there are various types such as "speed", a database will be created for each defect treatment based on such correspondence.

As described above, in the printed circuit board component mounting process, the independent solder printing process, component mounting process, and soldering process are sequentially performed, and quality control is performed for each process. Database of measurement data in each process when is taken,
When a defect occurs, the database can be referred to using the measurement data at that time, and the corresponding defect treatment can be determined and treated, so care must be taken of the cross-process correlation. It is possible to perform automatic quality control that suppresses the occurrence of defects without any problem.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram when an inter-process quality control method according to an embodiment of the present invention is applied to a printed circuit board mounting line.

FIG. 2 is a block diagram of a process control device.

FIG. 3 is a flowchart of an automatic quality management program.

FIG. 4 is a flowchart of a program for preventing occurrence of a defect in each process.

FIG. 5 is a flowchart of a defect rate reduction program by process improvement.

FIG. 6 is a model diagram showing the inspection principle of the solder printing inspection device.

FIG. 7 is a model diagram showing the inspection principle of the mounted component inspection device.

FIG. 8 is a diagram showing an actual example of position shift detection.

FIG. 9 is a model diagram showing the inspection principle of the soldering inspection device.

FIG. 10 is a cross-sectional model diagram showing the inspection principle of the soldering inspection device.

FIG. 11 is a model diagram showing an actual inspection example of the soldering inspection apparatus.

FIG. 12 is an explanatory diagram showing a measurement target of the soldering inspection device.

FIG. 13 is a histogram showing the transition of the solder volume measured by the solder print inspection device.

FIG. 14 is a diagram showing a part of the contents of a database.

FIG. 15 is a diagram showing a part of the contents of a database according to a modification.

FIG. 16 is a diagram showing a cause of failure in reflow soldering and specific unnecessary contents.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 ... Solder printing apparatus 12 ... Solder print inspection apparatus 21 ... Component mounting apparatus 22 ... Mounted component inspection apparatus 31 ... Soldering apparatus 32 ... Soldering inspection apparatus 41-43 ... Management personal computer 44 ... LAN line 45 ... Process control apparatus 45a ... Control unit 45b ... Bus 45c ... Program memory 45d ... Data memory 45e ... Input data interface unit 45f ... Output data interface unit PB ... Substrate

Claims (5)

[Claims] 1. An automatic quality control method in a component mounting process of a printed circuit board in which a plurality of processes whose quality is individually determined based on actually measured values are sequentially executed, wherein quality is reduced by quality determination of a downstream process. If the quality is determined in the downstream process, the correspondence between the quality determination in the downstream process, the quality determination in the upstream process, and the location within the cause is stored in a database as a correlation, and the database is referred to based on the quality determination in each process An automatic quality control method in a component mounting process of a printed circuit board, wherein a cause internal location of an upstream process having a correlation is identified and fed back. 2. The automatic quality control method according to claim 1, wherein a case where a defect is processed in a downstream process is classified into items, and a defect target for each defect treatment is measured in each upstream process. An automatic quality control method in a component mounting process of a printed circuit board, wherein a numerical trend is used as a database. 3. The automatic quality control method according to claim 2, wherein the tendency of the measured values used as the database is a difference between the distribution of the measured values of the defect target and the distribution of the standard measured values. An automatic quality control method in the component mounting process of printed circuit boards. 4. The automatic quality control method according to any one of claims 1 to 3, wherein when feeding back to an upstream process, a processing step is tuned in the upstream process. An automatic quality control method in the component mounting process. 5. An automatic quality control in a component mounting process of a printed circuit board, comprising: a plurality of manufacturing steps; and an inspection step of measuring a processing result corresponding to each manufacturing step and judging the quality of the processing result. The apparatus, in which the feedback mode is stored in a database together with the measurement data in the downstream and upstream inspection processes, which are determined to be defective in the quality determination result in each inspection process, and based on the measurement data in each inspection process. An automatic quality control device in a component mounting process of a printed circuit board, characterized by referring to the database to find out the cause of a problem occurring downstream in an upstream manufacturing process and feed it back.