KR100339008B1 - Method for cross sectional inspection of PCB using X-ray - Google Patents

Abstract

The present invention relates to a single-layer inspection method of a printed circuit board (hereinafter, referred to as a PCB) using X-rays, and in particular, in order to automatically check the mounting state and the soldering state of the electronic component after soldering the electronic component to the PCB. After applying the variable mesh configured by varying the pixel size according to the inspection importance of each part of the inspection area, the gradient inspection image of the inspection area of the PCB is obtained, and then restored and analyzed to the cross-sectional or longitudinal section image And it relates to a single layer inspection method of the PCB using the X-ray to determine whether the soldering state defective.

In the present invention, the upper and lower surfaces of the inspection target PCB can be selectively inspected by acquiring an image, thereby maximizing inspection efficiency and enabling high-speed inspection, as well as applying the most appropriate algorithm to each inspection region. Since the inspection is performed, various mounting state defects of the parts are detected, and the release parts having various types of electrodes can be effectively inspected.

Description

Method for cross sectional inspection of PCB using X-ray

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a soldering state inspection method for PCBs for soldering electronic components to printed circuit boards (hereinafter, referred to as PCBs) and inspecting defect states of the soldering. In particular, BGA (Ball Grid Array) and CSP (Chip) To check the mounting and soldering status of the parts where the electrodes are made of lead balls, such as scale packages, flip chips, etc., the oblique perspective image of the lead balls is obtained, restored and analyzed into longitudinal or cross-sectional images, and there are no defects. It relates to a tomographic inspection method of a PCB using the X-ray to determine.

In general, in the manufacturing process of electric and electronic boards, after soldering parts such as reflow after mounting parts, the process of inspecting the soldering state of each component mounted on the board is performed to improve the reliability of the product. Will perform.

One such inspection method is visual inspection or soldering external inspection. The apparatus for performing the above-mentioned soldering external inspection is a shape of soldering in the CCD camera using an illumination device and a CCD camera to inspect the appearance of a part having a lead. Was judged as good or bad according to the shape by processing the video signal.

However, when mounting BGA or CSP parts, the electrode is formed in the form of a lead ball on the lower surface of the part, and thus it is impossible to inspect the soldering external inspection device. Thus, X-ray inspection equipment using X-rays has emerged to enable this.

The X-ray inspection apparatuses basically use image formation based on the transmission X-ray intensity, which is inversely proportional to the thickness and density of the material to be penetrated, so that defects of the solder can be easily inspected. Can be.

These X-ray inspection apparatuses are mainly X-ray fluoroscopy and laminography apparatus, and other similar technologies are referred to as computed tomography (CT), which is mainly used for medical purposes. ) There is a device.

Hereinafter, the brief description of the X-ray inspection equipment as follows.

First, the X-ray perspective inspection apparatuses are described. In KR 89-12109 (inspection method of a soldered portion by X-ray transmission image and the internal structure of the electronic component on the apparatus and the substrate), the X-ray toward the specimen is examined. The X-ray image transmitted through the test object is obtained through a CCD camera, and the image signal converted into an electrical signal by the CCD camera is processed to determine whether soldering is defective.

Although the above-mentioned method is most commonly used, the soldering related information of the test object is all viewed through the X-ray image so that the exact shape of the solder cannot be known. In particular, when the PCB to be inspected is a double-sided board, Since the soldering shape of the part and the lower surface of the soldering part cannot be distinguished, there is a problem that the reliability of the inspection is significantly lowered.

In addition, the above-described method estimates the lead amount by applying the shadow technique for automatic inspection and compares the estimated lead amount with a preset reference value to determine whether there is a defect. Difficulties in inspecting defects such as movement, polarity inversion, etc., as well as difficulties in dealing with new release parts.

Next, U.S. Patent No. 4,516,252 (DEVICE FOR IMAGING LAYERS OF A BODY) separates and fixes two X-ray generators and uses one image acquisition unit, which is expensive in terms of cost and images acquired through two X-rays There is a problem of poor quality.

Next, U.S. Patent No. 2,667,585 (DEVICE FOR PRODUCING SCREENING IMAGES OF BODY SECTIONS) is fixed with an electrostatic deflection function of the electron beam in the X-ray, X-ray generator for changing the path of the X-ray projection, and the X-ray generator And a detector fixed to the opposite side of the X-ray generator to obtain a projected X-ray deflected from the detector, the detector being provided with an electro-optical system for deflecting the electronic image on the detector.

The above structure can overcome various problems due to mechanical movement, but since there is no measure for keeping the focus and energy of the electron beam constant, the size and brightness of X-rays are changed. There is a problem with repeatability and resolution. In addition, the above-described structure has a problem in that only a very small area can be inspected because there is a degree of limitation in the vertical direction with respect to the focal plane.

Next, unlike the X-ray fluoroscopy apparatus as described above, WO Patent No. 89/04477 (AUTOMATED LAMINOGRAPHY SYSTEM FOR INSPECTION OF ELECTRONICS), U.S. Patent No. 4,926,452 (AUTOMATED LAMINOGRAPHY SYSTEM FOR INSPECTION OF ELECTRONICS), U.S. Patent No. 5,081,656 (AUTOMATED LAMINOGRAPHY SYSTEM FOR INSPECTION OF ELECTRONICS), U.S. Pat. Patent No. 5,561,696 (METHOD AND APPARATUS FOR INSPECTING ELECTRICAL CONNECTIONS), U.S. Patent No. 5,621, 811 (LEARNING METHOD AND APPARATUS FOR DETECTING AND CONTROLLING SOLDER DEFECTS).

In the lamination apparatus, an X-ray generator for rotating and radiating X-rays obliquely toward an object to be inspected, and an X-ray image transmitted through the object in synchronization with the X-ray generator is carried out by a CCD camera. Acquire multiple frames of the tilted perspective image of the specimen using a rotating mirror that reflects the X-ray image to be acquired, and then restore the cross-sectional image of the specimen by applying a mean operation algorithm based on the inspection area. , Analyze.

This method has the advantage that the soldering inspection can be processed quickly because the acquisition and restoration of the cross-sectional image is performed at high speed, while the shadow effect caused by an object that is not on the Z-axis focal plane formed in the vertical direction is very large. In this case, the accuracy of the restoration is poor.

As a result, the inspection result of the lead ball deterioration in which the solder balls are separated from the PCB lands at a small distance from the soldering defects is very inaccurate and unreliable, and in order to compensate for this, several inspections are repeatedly performed in the Z-axis direction. There is a huge problem in terms of effort and time.

In addition, the laminating apparatus has an X-ray generator that rotates while radiating X-rays obliquely, and a rotating mirror that is rotated in synchronization with the X-ray generator, which is of course complicated in mechanical structure. This takes a lot of cost, there is a problem that the vibration is generated due to the high-speed rotation of the rotating mirror to weaken the durability of the device.

In addition, the above-mentioned laminating device is typically restored at the same time in the land surface of the PCB by restoring a cross section of a certain height portion by using the reconstructed image by comparing the average value in the predetermined inspection area by inspecting the defects and at the same time selected in advance Bridge failure is checked by comparing the brightness of the marked part.

Such a method is difficult to cope with various new parts because only a specific algorithm is applied to check for defects, and high-speed inspections are not possible, but defects such as parts defects, movements, and polarity inversions of parts are impossible to inspect. . In addition, there is a problem in that the inspection speed is very slow when the image is acquired and restored at various heights in the vertical direction with respect to the PCB according to the component size according to the component type to inspect the component defect.

Finally, U.S. Patent No. Looking at the CT device using 4,852,131 (COMPUTED TOMOGRAPHY INSPECTION OF ELECTRONIC DEVICES), the CT device is rotated 180 degrees around the specimen (1), as shown in Figure 1, the specimen (1) An X-ray generating unit 3 emitting X-rays toward the light source, and a detector 5 installed opposite to the X-ray generating unit 3 and obtaining an X-ray image passing through the specimen 1; It is composed of

Unlike the lamination apparatus described above, the CT device having the above-described structure is an apparatus for inspecting internal defects after acquiring and restoring a longitudinal cross-sectional image of the specimen 1, and a method for inspecting an internal defect of a recent electronic component. As a result, it is being applied to industrial applications.

However, in order to restore the longitudinal cross-sectional image of the inspected object 1, the CT apparatus has a very long time to inspect since the number of pixels of the image secured by the detector 5 must be equal to or greater than that of the inspected object 1. There is a problem that takes time.

That is, for example, assuming that there is a longitudinal section having 256 x 256 pixels, the total number of pixels of this longitudinal section is 65,536, so the number of pixels of the image to be secured by the detector 5 should be 65,536 or more. Therefore, if the number of pixels of the detector 5 is 256, it is necessary to repeatedly perform 256 image recordings in order to obtain sufficient data for inspection, which takes a lot of shooting time, and when restoring the longitudinal section image using the secured data. Again, the amount of data increases the amount of computation, which takes a long time, resulting in a very long total inspection time.

Incidentally, since the image is restored by applying the same level of importance as the entire resolution to the longitudinal section of the specimen 1, the effort, cost, There is a problem in that the test loss is increased due to time.

In addition, since the CT device takes a very long time to determine whether the specimen (1) is defective, it is difficult to apply in-line (In-Line), which is difficult for industrial use There is an inadequate problem.

The present invention has been made to solve the above problems, by being able to selectively inspect the upper and lower surfaces of the inspection target PCB by one image acquisition, the efficiency of the inspection is maximized and high-speed inspection, respectively, It provides a single-layer PCB inspection method using X-rays that detects various defects in mounting state of components and also effectively inspects deformed parts with various types of electrodes by applying the algorithm that is most suitable for the inspection area. There is a purpose.

1 is a schematic view showing the structure of a general computed tomography (CT) device;

2 is a conceptual diagram of imaging of CT,

3 is a view for explaining the imaging principle of the CT,

4 is a view for explaining the X-ray image forming contribution of the CT,

5 is a view for explaining the principle of restoration of CT,

6 is a view for explaining longitudinal section imaging of a sphere using CT;

7 is a view showing the form of the fixed mesh used for CT (a) and the variable mesh (b) used in the present invention,

8 is a block diagram showing a tomographic inspection apparatus of a printed circuit board (hereinafter referred to as a PCB) using an X-ray for implementing the present invention,

9 is a flow chart showing an automatic inspection plan establishment process for the PCB of the single-layer inspection method of the PCB using the X-ray,

10 is a flowchart showing the main inspection process for the PCB of the single-layer inspection method of the PCB using the X-ray,

11 is a view showing the inspection importance of each part of the lead ball to be considered when manufacturing the variable mesh applied to the electrode formed parts of the lead according to the present invention,

12 and 13 are views illustrating various forms of the variable mesh according to the present invention;

14 is a view showing a method for restoring a cross-sectional image of the inspection area of the inspection target PCB in accordance with the present invention;

15 is a flowchart showing a component and solder defect inspection procedure in the PCB main inspection process of the single-layer PCB inspection method using X-rays according to the present invention,

16 is a flowchart showing a component defect item inspection process of a cross-sectional image of the inspection area during the PCB main inspection process of the PCB tomography inspection method according to the present invention,

FIG. 17 is a view illustrating a defective item in a cross section inspected according to a tomographic inspection method of a PCB using X-rays according to the present invention; FIG.

FIG. 18 is a view showing a reconstructed image of an inspection area for explaining an algorithm applied to inspect a component failure item and a solder failure item in the present invention; FIG.

FIG. 19 is a diagram for explaining an algorithm of analyzing and restoring the reconstructed image of FIG. 18 using binarization. FIG.

FIG. 20 is a diagram for explaining an algorithm of analyzing and restoring the reconstructed image of FIG. 18 using edge strength. FIG.

FIG. 21 is a diagram for describing an algorithm for analyzing and restoring a reconstructed image of FIG. 18 using a variance value; FIG.

22 is a flowchart showing a soldering defect item inspection process of a cross-sectional image of the inspection area during the PCB main inspection process of the PCB tomography inspection method according to the present invention,

FIG. 23 is a diagram illustrating a solder failure item in a cross section inspected according to a tomographic inspection method of a PCB using X-rays according to the present invention; FIG.

24 is a flowchart showing a soldering defect item inspection process of a longitudinal cross-sectional image of an inspection area during a PCB main inspection process of the PCB tomography inspection method using X-rays according to the present invention;

25 is a view showing a solder failure item in the longitudinal section inspected according to the tomographic inspection method of the PCB using the X-ray according to the present invention.

<Explanation of symbols for main parts of the drawings>

10: inspection target PCB 20: positioning part

30a, 30b: first and second X-ray generators 50a, 50b: first and second AREA image multipliers

60a, 60b: first and second AREA camera 70: image processor

80: main controller

According to a first aspect of the present invention for achieving the above object, after the soldering of the electronic component on the PCB, the inspection importance of each part of the inspection area in the inspection area of the PCB to automatically check the mounting state and the soldering state of the electronic component X-rays to obtain the inspection area gradient image of PCB by applying variable mesh composed of variable size of pixels according to X-rays In the tomographic inspection method of the PCB using a first step of obtaining a tilted perspective image of the inspection area of the PCB and the cross-sectional image of the upper surface of the PCB by restoring the cross-sectional image of the upper surface of the PCB using the image obtained in the first process The second process of inspecting the parts defects, movement, and polarity reversal, which are the defective parts, on the parts installed in the The cross-sectional image of the lower surface of the PCB is inspected using the third process of inspecting the solderless items such as lead-free, unpaid, over-paid, bridge, and void, and the images obtained in the first process. After the restoration, the fourth process of inspecting the component defects, moving parts, and polarity reversal, which are the defective items, on the components mounted on the lower surface of the PCB, and if the inspection result of the fourth process is good, the components mounted on the lower surface of the PCB The fifth step of performing the inspection on the lead-free, unpaid, over-paid, bridge, and voids, and the inspection results of the third and fifth steps are good and the parts mounted in the inspection area are ball type electrodes. If the part has a component, the restoration of the longitudinal section image of the PCB inspection area using the image obtained in the first step, and then inspection for the poor soldering, voids, voids of the components mounted on the upper and lower surfaces of the PCB The sixth process is to perform.

In addition, an additional feature of the present invention is that before the second, third, fourth, fifth, and sixth steps, the user arbitrarily selects the number of inspection areas, the location of the inspection areas, and the like. It further includes the process of adding, modifying, and deleting an inspection plan that includes size, algorithms applied to each inspection area, and parameters required for each algorithm.

In addition, an additional feature of the present invention is that the algorithm applied to each of the inspection areas includes: an algorithm for detecting defective items by calculating an average value of the image brightness in the inspection area and comparing the average value of the image brightness with a preset reference value; or Algorithm for detecting defective items by binarizing the inspection area to obtain the area of the component or solder area in the binarized image and comparing this area with a preset reference value, or area of the component or solder area in the binarized image by binarizing the inspection area An algorithm to detect defective items by obtaining the ratio of the area of the test area to the total inspection area and comparing the area ratio with a preset reference value, or in the horizontal or vertical direction of the component or solder area in the binarized image by binarizing the test area. Find the maximum length of, then compare this maximum length to the preset By setting the threshold edge strength for the edge intensity which is the difference in brightness between neighboring pixels in the inspection area or the inspection area, the number of pixels with the edge intensity greater than this threshold edge intensity is obtained. An algorithm for detecting a defective item by comparing with a preset reference value or an algorithm for detecting a defective item by calculating a variance value in the inspection area and comparing the variance value with a preset reference value.

In the present invention as described above, the upper and lower surfaces of the inspection target PCB 10 can be selectively inspected by acquiring an image, thereby maximizing inspection efficiency and enabling high-speed inspection, as well as the most suitable for each inspection region. Since the inspection is performed by applying the algorithm, it is possible to detect various mounting state defects of the parts, and there is an advantage that an effective part can also be inspected even with various parts having various electrodes.

Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings.

The single-layer inspection method of the PCB using the X-ray according to the present invention basically by using the principle of CT after the gradient of the junction of the electronic component mounted on the upper or lower surface of the PCB and then restored to the longitudinal section or cross-sectional cross-sectional image PCB This is an automatic inline type test that simultaneously checks the top and bottom surfaces of.

Conventional CT method has high accuracy of inspection but large amount of shooting and calculation, so inspection is not done at high speed, and it takes a lot of time. Unlike the CT method, lamination method is capable of high speed inspection. When inspecting parts formed with lead balls such as BGA, CSP, and flip chips, it is difficult to grasp fatal defects in PCB soldering inspection such as poor lead ball moxibustion, so that the test result is unreliable.

Accordingly, the present invention has been devised as a method capable of eliminating the disadvantages while simultaneously utilizing the advantages of the CT method and the lamination method, that is, a method of inspecting the longitudinal section at high speed while maintaining the accuracy of the inspection.

The principle of the general CT method is as follows. As shown in FIG. 2, when the X-ray image emitted from the X-ray generator 3 and transmitted through the test object 1 is obtained by the detector 5, the detecting energy at this time is Can be displayed as here, Is the distance moved from the center position of the X-ray generator 3, Denotes an angle positioned with respect to the X-ray generator 3. therefore, It is possible to display the energy, that is, the density at the corresponding position of the specimen 1 using.

Therefore, from the X-ray generator 3 As far as the sensing energy, May be defined as the sum of density values of the specimen 1 present on the same angle. In other words, as shown in FIG. Is the density value at the ij position of the specimen (1), The ij position of this specimen (1) is determined by the X-ray image. If the degree of X-ray image forming contribution to the position is indicated, the detecting energy of the position may be defined as in Equation 1.

Even if the X-ray image forming machine, Referring to FIG. 4 to explain the concept of, the rectangular grid means a cross section in which the specimen 1 may be positioned as one pixel. That is, 1, 2, 3,... In the X direction. , m-1, m pixels and 1, 2, 3,... in the Y direction. That is, the specimen 1 may be located at a cross section consisting of n-1, n pixels.

At this time, each of the pixels in the X, Y direction is 1, 2, 3,... Of the detector 5 to form an X-ray image. will affect the pixels k, k-1 and k, respectively. For example, in the pixel 1 of the detector 5, the X and Y pixels at positions correlated with the pixel 1 of the detector 5 among the pixels between the X-ray generator 3 and the detector 5 are affected. Even if the degree of influence at this time is an X-ray image forming contributor, to be.

Therefore, when H is the X-ray image forming device of each pixel, and F is the density of each pixel, the detecting energy, g, can be expressed as in Equation (2).

g = HF

As a simple example, the density of each pixel is f (1,1), f (1,2), f (2,1), f (2,2), as shown in FIG. The detecting energy is g (1,1), g (2,1), and the second detecting energy obtained after moving a certain distance is g (1,2), g (2,2), each pixel to each of the de-tekting energy, g (1,1), g ( 1,2), g X- ray image forming contribution affecting the (2,1), g (2,2) H 11, If it is defined as H ₁₂ , H ₂₁ , H ₂₂ , the detecting energy at this time may be defined by a matrix such as Equation 3.

In the above-described detecting energy, g (1,1), g (1,2), g (2,1), g (2,2) are information obtained by X-ray image, and even X-ray image forming machine Since H ₁₁ , H ₁₂ , H ₂₁ , and H ₂₂ are information that can be known in advance by calculation, substituting the values of H ₁₁ , H ₁₂ , H ₂₁ , and H ₂₂ into the matrix shown in Equation 4 can be arranged. have.

As a result, since the restoration of the image is a process of obtaining density values of each pixel, it can be defined as in Equation 5.

F = H ^-1 g

Therefore, the CT method saves time by obtaining H ⁻¹ in advance and then performing inspection to obtain F, which is a density value of each pixel.

However, in the conventional CT method, since the number of pixels to be inspected is absolutely large, not only the photographing time is required but also the amount of calculation at the time of restoration is increased, so that the restoration time is long. At this time, as the size of the matrix H increases, the time required to obtain the inverse of H increases by three times. Therefore, in order to reduce the calculation amount at the time of restoration, the number of pixels to be inspected must be reduced.

The conventional CT method described above is mainly used for medical purposes, and since the test subject was a human body, all parts of the test subject were equally important. For this reason, a uniform size of pixels was uniformly applied to all parts of the human body to be inspected to obtain and restore a cross-sectional image. However, when used for industrial purposes, not all parts of the inspection object are equally important.

Therefore, in the soldering state inspection of the electronic components mounted on the PCB in advance, it is expected that an important part is predicted in advance, and the pixel of the part is made small and tight, and the pixel of the other non-critical part is made large. That is, image acquisition and reconstruction are made at high speed by reducing the number of pixels in the inspection area as much as possible by using a variable mesh that automatically applies the size of the pixels in accordance with the inspection importance for each part of the inspection area.

For example, when a tomography of a sphere as shown in FIG. 6 is tomographyed, using a conventional CT method, a fixed size pixel is uniformly applied to all portions of the cross section of the inspection area as shown in FIG. You must use a mesh. The fixed mesh is 120 × 80, which is 9600 pixels in total. Therefore, if the number of pixels of the detector 5 is 120, at least 80 images should be taken until the data at the time of restoration is more than or equal to the number of pixels to be inspected. In addition, the size of the matrix F is 9600 and the size of the matrix H is 9600 × 9600 (92,160,000). The size of the matrix H is so large that it occupies about 351 MBytes of memory, and the time required for restoration is about 1 hour.

On the other hand, by applying a variable mesh as shown in Fig. 7 (b) to the cross section of the inspection area according to the inspection importance of each part to form a relatively small pixel in the area around the lead ball densely larger in other non-critical parts Configuring pixels of size reduces the total number of pixels to 640. When the total number of pixels is reduced to 640 as described above, the number of pixels of the detector 5 is reduced to six times, and the size of the matrix H is also greatly reduced to 720 × 640 (406,800). Therefore, if the amount of processing data and the amount of operation restoration when applying the fixed mesh according to the conventional CT method is 100%, respectively, the amount of processing data and the amount of operation restoration when the variable mesh is applied is 0.44% and 0.029%, respectively. It will show a dramatic reduction rate. In addition, the memory occupied by the matrix H is also reduced to 0.8M Bytes, and the most important restoration time is greatly shortened, so that the existing CT method can complete the work to be processed over an hour in about 1 second.

The above comparison is summarized in Table 1 below.

division Conventional CT Method How to apply variable mesh Total pixels 120 × 80 (9600) 640 Video screening 80 (9600/120) 6 (640/120) Size of F 9600 (100% standard) 640 (7%) Size of H 9600 × 9600 720 × 640 Processing data amount 100% standard 0.44% (640/9600) ² Operational Restore 100% standard 0.029% (640/9600) ³ Occupied memory size of H 351M Byte (100% standard) 1.8M Byte (0.6%) Restoration time 1Hr 1sec Usage Medical Industrial

As described above, when the inspection is performed by applying the variable mesh to the cross section of the inspection region, it can be seen that the inspection time is drastically reduced because the total number of pixels is greatly reduced.

Therefore, the present invention maintains the accuracy of the inspection at the level of the conventional CT method by using a variable mesh on the cross section of the inspection area and at least two image acquisition parts for acquiring an image by tilting the inspection area. At the same time, the speed of inspection has been improved to the level of conventional lamination methods, enabling precise high-speed inspection of inspection areas.

Hereinafter, in the present invention, the above terminology is used as a meaning that a pixel is composed of one lattice having a predetermined size, and the variable mesh is configured by varying the size of the pixel according to the inspection importance of each part of the inspection area.

Referring to FIG. 8, the tomography inspection apparatus of the PCB using the X-ray according to the present invention includes a positioning unit 20 for positioning the inspection target PCB 10 and simultaneously moving the inspection target PCB 10. A first image acquisition unit for acquiring an inspection region tilted perspective image of the inspection target PCB 10 using X-rays, and obtaining an inspection region tilted perspective image of the inspection target PCB 10 using X-rays A second image acquisition unit, an image processing unit 70 processing the tilted perspective image obtained by the first image acquisition unit and the second image acquisition unit, and a main controller 80 controlling the operation of each of the above-described components. It is configured to include.

In the above, the positioning unit 20 is composed of an XYZ robot for determining the position by moving the inspection target PCB 10 to each axis of the XYZ, and a robot controller for controlling the operation of the XYZ robot.

In addition, the first image acquisition unit and the first X-RAY generator 30a for moving up and down spaced apart from the inspection target PCB 10 by a predetermined distance to emit an X-ray toward the inspection area of the inspection target PCB 10 and The first X-ray generator 30a is installed to correspond to the first X-ray generator 30a so as to be positioned between the first X-ray generator 30a and passes through the inspection region of the inspection target PCB 10. A first AREA image intensifier tube 50a for converting an X-ray image into a visible light image and a first AREA camera 60a for acquiring an image converted by the first AREA image amplification tube 50a. It is composed of

In addition, the second image acquisition unit is moved up and down spaced apart from the inspection target PCB 10 by a predetermined distance and the second X-RAY generator 30b for emitting an X-ray toward the inspection area of the inspection target PCB 10 and And installed so as to correspond to the second X-RAY generator 30b such that the inspection target PCB 10 is positioned between the second X-RAY generator 30b and penetrating the inspection region of the inspection target PCB 10. And a second AREA image multiplier 50b for converting the X-ray image into a visible light image and a second AREA camera 60b for acquiring the image converted by the second AREA image multiplier 50b. .

Herein, the first and second X-RAY generators 30a and 30b respectively move the X-RAY source and the X-RAY source that move the X-RAY source up and down to generate the X-RAY source and the inspection target PCB 10. Z-robot for adjusting the distance between the X-ray and X-RAY controller for adjusting the amount of light emitted from the X-ray source on / off and mounted on the positioning unit 20.

The first and second X-RAY generators 30a and 30b configured as described above should all have the same conditions such as voltage or current affecting X-ray generation, and the operation of the Z robot may be performed by the positioning unit ( It is controlled by the robot controller of 20).

In addition, the first and second X-RAY generators (30a, 30b) in the displacement measuring unit 40 for measuring the distance between the inspection target PCB 10 and the first and second X-RAY generators (30a, 30b) Is installed, the displacement measuring unit 40 is moved to the inspection area of the inspection target PCB 10 when the inspection target PCB 10 or the first and second X-RAY generators 30a and 30b are moved up and down. It performs a function of recognizing the Z-axis displacement of the inspection target PCB 10 or the first and second X-RAY generators (30a, 30b) so that the appropriate variable mesh of the appropriate shape on the Z-axis accurately.

In addition, the first inspection area and the second inspection area of the inspection target PCB 10 are transmitted between the first and second AREA image multipliers 50a and 50b and the first and second AREA cameras 60a and 60b. Zoom lenses for enlarging or reducing an image are respectively provided.

Referring to the tomographic inspection method of the PCB using the X-ray of the present invention for inspecting the longitudinal section or the cross-section of the inspection target PCB 10 by using the device configured as described above are as follows.

First, an automatic inspection plan is established as a preparation process for automatically inspecting the inspection target PCB 10. Referring to FIG. 9, in operation S110, a worker inputs general data of the inspection target PCB 10, such as the size, model name, and reference point of the inspection target PCB 10, to the main controller 80. Thereafter, in S120, the operator inputs CAD data in which part names, position coordinates, mounting angles, and the like of components mounted on the upper or lower surface of the inspection target PCB 10 are defined.

Thereafter, in S130, the same library name is assigned to the parts defined as the same type in the CAD data and then indexed. In this case, the above-described library is part information classified according to the type and size of the part, and the part size, image restoration method, no part, movement, polarity, lead ball, lead free, unpaid, over lead, bridge, void, lead moist, lead It includes inspection types such as ball radish, inspection area location by inspection type, number of inspection areas by inspection type, algorithm applied to each inspection area, and various parameters used in each algorithm. In addition, an arbitrary mesh size may be specified and modified in an area designated as an inspection area in the library. This process is intended to perform the same type of inspection on parts of the same size, regardless of the capacity of the parts and the type and appearance of the parts.

Then, in S140 using the CAD data and library name data of the inspection target PCB 10 automatically creates an inspection plan. At this time, the inspection plan is made so that all parts mounted on the upper or lower surface of the inspection target PCB 10 can be inspected, and automatically selects an appropriate inspection type according to the size, type, and mounting form of the component. In order to minimize the moving distance and the number of times during the inspection, the time is shortened as much as possible to determine the inspection order between the parts, that is, the inspection objects are included.

In addition, the inspection plan further includes identifying and determining whether the inspection region is suitable for longitudinal tomographic inspection or cross-sectional tomographic inspection according to the type and mounting type of the electronic component mounted in the inspection region of the inspection target PCB 10. Included.

In this case, if the component mounted in the inspection region of the inspection target PCB 10 is a component formed by lead balls such as BGA, CSP, flip chip, and the like, the longitudinal cross-sectional tomographic inspection is mainly performed, and the inspection region of the inspection target PCB 10 is performed. If the mounted part is an IC part or a chip part, the cross-sectional tomography inspection is mainly performed.

Thereafter, in order to optimize the automatic inspection plan in S150, the final automatic inspection plan is determined by teaching the inspection plan to delete, modify, and add data on the inspection target PCB 10.

When the automatic inspection plan of the inspection target PCB 10 is established as described above, the full-scale inspection proceeds according to the automatic inspection plan, which will be described below with reference to FIG. 10.

First, in S310 the positioning unit 20 is operated to position the inspection target PCB 10 at a predetermined inspection position. That is, the main controller 80 outputs a control signal to the XYZ controller and drives the XYZ robot based on the general data on the inspection target PCB 10 input in the automatic inspection plan establishment step. Adjust the inspection position of).

At this time, when the object to be inspected is a micro component or needs to be inspected by enlarging it carefully, in order to enlarge the transmission image of the inspection area during inspection, the inspection object PCB 10 is moved up and down through the positioning unit 20. Alternatively, the Z-axis position can be changed by moving the first and second X-RAY generators 30a and 30b up and down, and the inspection area of the inspection target PCB 10 is the first and second X-RAYs. As the generators 30a and 30b move downward toward the inspection target PCB 10, the inspection objects PCB 10 move upward toward the first and second X-RAY generators 30a and 30b. do. That is, the inspection area of the inspection target PCB 10 is enlarged as the distance between the first and second X-RAY generators 30a and 30b and the inspection target PCB 10 gets closer.

As described above, instead of varying the Z-axis positions of the inspection target PCB 10 or the first and second X-ray generators 30a and 30b, the first and second AREA image multipliers 50a and 50b and the first are as follows. And the transmission image of the inspection area may be enlarged by adjusting the magnification of the transmission image of the inspection area by using a zoom lens provided between the second AREA cameras 60a and 60b, respectively.

Then, in S320 it is checked whether to perform a longitudinal cross-sectional tomographic inspection or a cross-sectional tomographic inspection for the inspection area of the inspection target PCB (10). That is, whether the longitudinal cross-sectional tomographic inspection is suitable for the inspection area of the inspection target PCB 10 or the cross-sectional tomographic inspection is determined in advance in the automatic inspection plan, so the S320 reads the data of the main controller 80 and checks it again. do.

Subsequently, the distance between the first and second X-RAY generators 30a and 30b and the inspection target PCB 10 is measured through the displacement measuring unit 40 in S330, and the first and second X-RAY generators 30a are measured. , 30b) and the Z-axis displacement of the inspection target PCB 10.

Thereafter, the main controller 80 automatically selects an appropriate variable mesh shape according to the type and mounting type of the components mounted in the inspection area of the inspection target PCB 10 in S340. At this time, the variable mesh can be manually modified by the user, but the most suitable for the inspection area through a variable mesh writing tool that receives the height of the inspection target substrate, the size of the electrode, the pixel size at each position of the inspection area, and the like. It can also be written automatically in form.

Thereafter, the inspection target PCB 10 is inspected according to the Z-axis displacement of the first and second X-RAY generators 30a and 30b and the inspection target PCB 10 measured by the displacement measuring instrument 40 in S350. By matching the focus of the area with the variable mesh, the variable mesh is applied exactly on the Z axis of the inspection area longitudinal section.

As described above, the variable mesh is configured by varying the size of the pixel according to the shape of the soldering position of the electronic component in the inspection region longitudinal section or the inspection importance for each part thereof.

For example, when the electrode is formed of a lead ball such as a BGA, a CSP, a flip chip, and the like, the inspection importance for each part of the lead ball may be formed by connecting the lead ball and the PCB land (D) and the lead ball to the outside as shown in FIG. 11. You can rank the outer (B), the inside of the lead ball (A), and then the outside of the lead ball (D). At this time, if the importance of the junction between the lead ball and the PCB land (D) is 100, the remaining parts can have 50, 20, and 10 inspection importance in order. The variable mesh is determined.

Therefore, the variable mesh has a smallest sized pixel at the junction between the lead ball and the PCB land, and a defect such as insufficient lead or excessive lead is determined at the outer part (b) where the lead ball is externally shaped. The pixel is composed of the size of the lead ball, and the inside of the lead ball constitutes the size of the lead balloid defect that can be judged. Most preferred.

However, according to the purpose of performing the component mounting state and the soldering state inspection of the inspection target PCB 10, it is also possible to select the form of the variable mesh that selectively reflects the inspection importance of each part of the lead ball.

That is, when very fine inspection is to be performed in order to obtain ultra high accuracy, as shown in FIG. 12 (a), the shape in which the pixels of the same size are formed on the longitudinal section of the inspection area is selected, In the case of appropriately responding to the defective item, as shown in (b) of FIG. 12, the lead ball and the lead ball are externally formed with a relatively small size of pixels, and the rest of the other part is of a somewhat larger size. You can teach the mesh in the form of pixels.

In addition, in the case of quickly inspecting only an important part during inspection, as shown in (c) of FIG. 12, a relatively small pixel is densely formed only at the junction of the lead ball and the PCB land, and the other pixel is somewhat larger in size. Teaching the mesh in the form of the configuration.

In addition, in the case of inspecting only the lead ball defects at the time of inspection at high speed, as shown in FIG. 12 (d), only pixels of a relatively small size are formed in the junction of the lead ball and the PCB land, and the remaining parts inside the lead ball, The mesh can be taught in a form in which only one pixel is formed in the outer portion where the lead ball is formed on the outside and the outside of the lead ball.

On the other hand, the variable mesh shown in FIG. 13 is a form applied to solder inspection of IC components or chip components, and the pixels need to be configured only at the position where solder is present, and the amount of solder solder is much smaller than that of solder balls. It is possible to reconstruct the longitudinal or cross-sectional image of the inspection area by using the variable mesh having a much smaller number of pixels than the variable mesh used for the state inspection.

More specifically, the variable mesh shown in (a) and (b) of FIG. 13 is a form in which pixels of a relatively small size are densely formed only at the junction between the IC component or the chip component and the PCB. In the variable mesh shown in (d), only one pixel is formed in the longitudinal direction, and pixels are formed at the junction between the IC component or the chip component and the PCB such that a plurality of pixels are formed according to the area unit in the lateral sugar direction.

At this time, the variable mesh shown in (c) and (d) of FIG. 13 includes only one pixel in the longitudinal direction and many pixels in the transverse direction in terms of area, respectively. When the sum is processed, the cross-sectional image of the upper or lower surface of the inspection area of the inspection target PCB 10 can be easily obtained.

After applying the corresponding variable mesh of the type suitable for the inspection area of the inspection target PCB 10 as described above, the X-ray toward the inspection region of the inspection target PCB 10 in the first X-RAY generator 30a in S360a When the X-ray image transmitted through the inspection area of the inspection target PCB 10 in S370a is converted into a visible light image by the first AREA image multiplier 50a. Thereafter, the tilted perspective image converted into the visible light region through the first AREA image multiplier 50a in S380a is obtained by the first AREA camera 60a, and then the analog image is processed by the image processor 70 in S390a. This digital image is converted and stored.

At the same time, when the X-ray is emitted from the second X-RAY generator 30b toward the inspection region of the inspection target PCB 10 in S360b, the X-ray image transmitted through the inspection region of the inspection target PCB 10 in S370b. The second AREA image multiplier 50b is converted into a visible light image. Thereafter, the tilted perspective image converted into the visible light region through the second AREA image multiplier 50b in S380b is obtained by the second AREA camera 60b, and then the analog image is processed by the image processor 70 in S390b. This digital image is converted and stored.

At this time, the first X-RAY generator 30a, the first AREA image multiplier 50a, the first AREA camera 60a and the second X-RAY generator 30b, the second AREA image multiplier 50b ), The process of acquiring the tilted perspective image of the inspection area by the second AREA camera 60b may be performed sequentially to further increase the accuracy of the inspection.

As described above, in order to proceed with the process of acquiring the perspective view image of the inspection region at the same time or sequentially, the first and second X-ray generators 30a and 30b need to repeatedly perform the photographing after moving as well as the simultaneous photographing in the stationary state. . Accordingly, in order to obtain X-ray images emitted from the first and second X-RAY generators 30a and 30b and transmitted through the inspection area of the inspection target PCB 10, the first and second AREA image amplification pipes, respectively. 50a and 50b and the first and second AREA cameras 60a and 60b should be used.

In addition, the image processing unit 70 pre-processes the stored digital image, that is, the tilted perspective image, before the tilted perspective image is restored to the cross-sectional image, so that the restored longitudinal section or the cross-sectional image is clear and clear.

More specifically, the first AREA camera after capturing each tilted perspective image obtained by the first and second AREA camera 60a, 60b in the master DSP (Digital Signal Processor) of the image processing unit 70, The tilted perspective image acquired at 60a is moved to the first slave DSP at about the same time as the tilted perspective image obtained at the second AREA camera 60b. Subsequently, the first and second slave DSPs perform preprocessing on the gradient perspective images, respectively, and move the respective gradient perspective images filtered through the preprocessing back to the master DSP.

Thereafter, in S400, the number of pixels of the variable mesh applied to the inspection area of the inspection target PCB 10 is compared with the number of pixels of the gradient perspective image obtained by the first and second AREA cameras 60a and 60b.

Subsequently, as a result of the S400, when the number of pixels of the gradient perspective image is smaller than the number of pixels of the variable mesh, the positioning unit 20 operates in S410 to move the inspection target PCB 10 by a predetermined amount and then return to the S360a and S360b. Let's do it.

S360a, S360b, S370a, 370b, S380a, S380b, S390a, S390b, S400, and S410 are larger than the number of pixels of the variable mesh applied to the inspection area of the inspection target PCB 10. The first and second AREA cameras 60a, The number of pixels of the oblique perspective image obtained by 60b) is equal to or greater than that, and is repeated until sufficient data for reconstructing the longitudinal or cross-sectional image is obtained.

In this case, when the longitudinal cross-sectional tomography inspection is performed on the inspection area of the inspection target PCB 10, the acquisition process of the inclined perspective image should be repeated about 3 to 4 times, and the oblique perspective when the cross-sectional tomography is performed. Even if the image acquisition process is performed only once, sufficient data for reconstructing the cross-sectional image is secured.

Subsequently, when the number of pixels of the variable mesh applied to the inspection area of the inspection target PCB 10 is greater than the number of pixels of the tilted perspective image as a result of S400, the inspection target PCB (based on the digital image stored in the image processor 70 in S420). Restore the inspection section longitudinal section or cross-sectional image of 10).

At this time, the restoration method of the longitudinal section image is made in the same manner as the conventional CT method described above, and the restoration method of the cross-section image is to separate the images of the upper and lower surfaces when the inspection target PCB (10) is a double-sided board. It is made by way of acquiring.

That is, when the variable mesh shown in (c) and (d) of FIG. 13 is applied as an example, first, as shown in (a) of FIG. 14, the upper and lower parts, that is, the lead ball and the chip part, are shown together. Acquire a perspective image. Subsequently, after restoring the cross-sectional image of the chip component shown in FIG. 14B from the perspective image, the cross-sectional image of the chip component shown in FIG. 14B is illustrated in the perspective image of FIG. 14A. Subtracted, a cross-sectional reconstructed image of the lead ball shown in FIG. 14C can be obtained.

Subsequently, by analyzing the inspection region longitudinal section or cross-sectional image of the inspection target PCB 10 restored in S430, it is automatically determined whether the component mounting state and the soldering state defective.

Referring to FIG. 15, a procedure for inspecting a component and a soldering failure in a longitudinal cross section or a cross-sectional image of an inspection region acquired and restored in the above manner will be described as follows.

First, an oblique perspective image of the inspection area is obtained in S510, and a cross-sectional image of the upper surface of the inspection target PCB 10 is restored in S520 by using the image obtained in S510, and then the upper surface of the inspection object PCB 10 is inspected in S530. The installed parts shall be inspected for parts defects, parts defects, movement, and polarity reversal. If the result of the inspection of S530 is a good product, in S540, the parts mounted on the upper surface of the PCB 10 to be inspected are inspected for lead-free, unpaid, over-paid, bridged, and voids.

When the cross-sectional inspection of the upper surface of the inspection target PCB 10 is completed as described above, after restoring the cross-sectional image of the lower surface of the inspection target PCB 10 in S550 using the image obtained in S510, the inspection target PCB 10 in S560. The parts mounted on the bottom surface are inspected for parts defects, moving parts, and polarity reversal. If the inspection result of S560 is a good product, in S570, the inspection is performed on the parts that are mounted on the lower surface of the PCB 10 to be soldered for the non-soldered, unpaid, overpaid, bridged, void.

Subsequently, if the inspection result of S540 and S560 is good and the component mounted in the inspection region is a component having a ball-type electrode, the longitudinal section image of the inspection region of the inspection target PCB 10 is used in S580 by using the image obtained in S510. After restoring, the S590 performs inspection on the soldering defects of the ball moxibustion and voids on the components mounted on the upper and lower surfaces of the inspection target PCB 10. Thus, the inspection area cross section inspection and the longitudinal section inspection of the inspection object PCB 10 are completed.

Referring to FIG. 16, a process of inspecting a component failure item in S530 and S560 will be first binarized based on a preset image brightness threshold in S710. That is, if the current image brightness value is equal to or greater than the image brightness threshold, the feature is defined as white. If the current image brightness value is less than the image brightness threshold, the feature is defined as black.

Subsequently, the area defined in black in each inspection area is calculated in S720, and the area of black calculated in this way is compared with the area corresponding to each component defect item in S730 to determine whether there is a defect.

Referring to FIG. 17, a comparison and determination process for each of the parts defective items described above will be performed. In the parts free inspection, if a black area is smaller than a preset reference value within the parts free inspection region defined by a dotted line, the parts are defective. If the area of black in the movement inspection area defined by the dotted line is larger than the preset reference value, the movement inspection determines that the movement is bad.In the polarity inversion inspection, the area of black in the polarity inspection area defined by the dotted line is preset If it is larger than the reference value, it is regarded as a polarity inversion failure.

In addition, before S530, S540, S560, S570, and S590, for each inspection item, the user arbitrarily sets the number of inspection regions, the position and size of the inspection regions, the algorithm applied to each inspection region, and the parameters required for each algorithm. It also includes the process of adding, modifying, and deleting inspection plans that include them.

The type of algorithm applied to each of the inspection areas will be described in detail as follows.

First, by calculating the average image brightness in the inspection area by dividing the sum of the image brightness values of each pixel present in the inspection area by the number of pixels in the entire inspection area, and comparing the average image brightness value with a preset reference value to detect a defective item. There is an algorithm.

To illustrate this algorithm, taking the reconstructed image of the inspection area shown in FIG. 18 as an example, the total number of pixels of the reconstructed image of the inspection area is 10x10, and the image brightness value of each pixel is 100 or 200, and the image brightness value is shown. Since the number of pixels having 100 is 70 and the number of pixels having 200 brightness is 30, the average brightness of the image is 130. The result of the inspection in the inspection area is derived by comparing the average value of the image brightness thus obtained with a preset reference value.

The algorithm using the average image brightness is preferably applied when the contrast of the inspection region reconstructed image is clear or when the inspection region has no significant difference compared with the inspection feature.

Next, there is an algorithm for detecting a defective item by binarizing the inspection area to obtain an area of a component or solder in the binarized image and comparing the area with a preset reference value.

More specifically, binarizing the inspection region reconstructed image of FIG. 18 using an arbitrary binarization value larger than 100 and smaller than 200 yields a binarized image as shown in FIG. 19, in which the component or solder feature is a bright region. Assuming that the number of pixels corresponding to the area of the component or solder is 30. The inspection result in the inspection region is derived by comparing the number of pixels in the bright region thus obtained with the preset reference pixels.

Next, there is an algorithm for detecting defective items by binarizing the inspection area to obtain a ratio of the area of the component or solder area to the area of the entire inspection area in the binarized image, and comparing the area ratio with a preset reference value.

More specifically, binarizing the inspection region reconstructed image of FIG. 18 using an arbitrary binarization value larger than 100 and smaller than 200 yields a binarized image as shown in FIG. 19, in which a component or solder feature is a bright region. Assuming that the number of pixels corresponding to the area of the component or solder is 30. Therefore, the ratio of the area of the component or solder to the area of the entire inspection area is equal to the ratio of the total number of pixels to the number of pixels of the bright area, which is 0.3 (30/100). The area ratio thus obtained is compared with a predetermined reference value to derive the inspection result in the inspection area.

Next, there is an algorithm that binarizes the inspection area to find the maximum length of the component or solder area in the binarized image in the horizontal or vertical direction, and compares the maximum length with a preset reference value to detect the defective item.

Next, set the threshold edge intensity for the edge intensity, which is the difference in brightness between neighboring pixels in the inspection area, obtain the number of pixels with edge intensity greater than this threshold edge intensity, and then compare the number of pixels with a preset reference value. There is an algorithm for detecting items.

More specifically, when the inspection region reconstructed image of FIG. 18 is processed using any boundary edge strength in the horizontal or vertical direction, an image in which edge intensity is shown as shown in FIG. 20 is obtained, and the edge intensity at this time is the edge intensity ( C _i , R _j ) = ?? P (C _i , R _j ) -P (C _i-1 , R _j-1 ) ?? Here, the ranges of i and j are 2 ≦ i ≦ 10 and 1 ≦ j ≦ 10, respectively. Subsequently, assuming that the feature of the component or the solder is a bright region in the obtained image, the number of pixels in the bright region is obtained, and the number of pixels is compared with a preset reference value to derive the inspection result in the inspection region.

As described above, the algorithm using edge strength is preferably applied when the image brightness difference between the inspection feature and the inspection background is small in the inspection region or when the overall image brightness is changed.

Next, there is an algorithm that detects a defective item by calculating a variance value in the inspection area and comparing the variance value with a preset reference value.

More specifically, as illustrated in FIG. 21, arbitrary distributed inspection regions V1 and V2 are set in the inspection region reconstructed image of FIG. 18, and variances for the distributed inspection regions V1 and V2 are obtained, respectively. At this time, since the area of V1 is composed of pixels having the same brightness value, a small dispersion value is obtained, and the area of V2 is composed of pixels having two kinds of brightness values, so a large dispersion value is obtained. Subsequently, the inspection results in the inspection region are derived by comparing the dispersion values obtained in each of the dispersion inspection regions V1 and V2 with a preset reference value.

In this way, the algorithm using the variance value is preferably applied when the inspection features in the inspection area are very small and mixed with the size of the inspection background.

Among the above algorithms, the analysis process for the cross-sectional and longitudinal cross-sectional images of the inspection area is described by applying an algorithm using an area in a binarized image.

First, referring to FIG. 22, a process of analyzing a cross-sectional image of an inspection region is binarized based on a preset image brightness boundary value, after which the cross-sectional image of the inspection region restored in S810 is binarized. Read the video values of the bridge window around the lead ball shape in S820 to calculate the presence of the bridge, calculate the amount of lead including lead balls by calculating the area of the black ball-shaped part in S830, and in S840 The void size in the lead ball is calculated by calculating the area of the part defined in white.

Subsequently, the soldering state of the lead ball being inspected is determined by comparing the results calculated in S820, S830, and S840 with the inspection reference values of the preset defective item in S850. That is, it is determined whether the soldering state of the lead ball under test is normal or whether it is a defective item such as insufficient lead, excess lead, lead balloid, bridge, and no lead ball.

At this time, when the area of the normal cross section is defined as 100 and the area of the insufficient lead cross section is 90 or less in the inspection standard, as shown in FIG. . In addition, when the area of the normal cross section is defined as 100 and the lead and multi-area area is 120 or more in the inspection standard, if the area of the black portion of the binarized cross-sectional image is 120 or more, it is determined that the lead is excessive.

In addition, when the area of the lead balloid defect is defined as 10 or more in the inspection standard, even if the area of the entire black portion of the binarized cross-sectional image is normal, if the area of the white portion within the black portion is 10 or more, it is determined as a lead balloid defect. do.

In addition, if the image average value of the bridge window is read between the lead ball and the lead ball among the binarized cross-sectional images, it is determined that the bridge is bad.

In addition, when the inspection standard defines the normal cross-sectional area as 100 and the lead ball-free section as 50 or less, if the area of the black portion of the binarized cross-sectional image is 50 or less, it is determined as a lead-free ball defect.

Next, referring to FIG. 24, the analysis process of the longitudinal section image of the inspection region is binarized based on a preset image brightness boundary value after binarizing the longitudinal section image of the inspection region reconstructed in S910. In S920, the lead area including lead balls is calculated by calculating the area of the lead ball shape defined in black, and in S930, the area of the part defined in white inside the lead ball shape is calculated to calculate the void size in the lead ball, and in S940, the PCB is calculated. Calculate the degree of lead ball moxibustion by measuring the distance between the land and lead ball shape.

Subsequently, the solder state of the lead ball being inspected is determined by comparing the results calculated in S920, S930, and S940 described above with the inspection reference values of the defective item. In other words, it is determined whether the soldering state of the lead ball under test is normal or whether it is a defective item such as insufficient lead, excessive lead, lead ball void, lead ball moxibustion, and no lead ball.

At this time, when the area of the normal longitudinal section is defined as 100 and the area of the insufficient lead section is 90 or less in the inspection criterion as shown in FIG. 25, if the area of the black portion of the binarized longitudinal section image is 90 or less, it is determined that the defect is insufficient. . In addition, when the area of the normal longitudinal section is defined as 100 and the lead and multi-area area is 120 or more in the inspection standard, if the area of the black portion of the binarized longitudinal section image is 120 or more, it is determined that the lead excess is bad.

In addition, when the area of lead balloid defects is defined as 10 or more in the inspection standard, even if the area of the entire black portion of the binarized longitudinal section image is normal, if the area of the white portion within the black portion is 10 or more, it is determined as a lead balloid defect. do. In addition, when the distance between the lead ball shape and the land shape of the PCB in the binarized longitudinal section image is larger than 0, it is determined that the lead ball is bad, and when there is no lead ball shape, it is determined that the lead ball is bad. At this time, the poor lead moxibustion may be determined to be a lack of lead, and the lack of lead ball may be determined as a lack of lead and a poor lead decay, respectively.

Thereafter, the total number of inspection areas according to the automatic inspection plan of the inspection target PCB 10 is compared with the number of inspection areas already inspected in S440. As a result of the step S440, if the number of inspection areas to be inspected is smaller than the total number of inspection areas, that is, if the inspection area to be inspected remains, the positioning unit 20 operates in S450 to move the inspection target PCB 10 to the next inspection area. Move to position

The above process is repeatedly performed until there is no inspection area to be inspected on the inspection target PCB 10.

In the tomographic inspection method of the PCB using the X-ray according to the present invention constructed and operated as described above, by the part so that the number of pixels required for restoring the cross-sectional image of the inspection area during the inspection area inspection of the inspection target PCB 10 By applying the variable mesh consisting of the pixel size according to the importance of inspection, the time required for retrieving the cross-sectional image after obtaining the sloping perspective image of the inspection area is greatly shortened, thereby increasing the accuracy of the inspection and enabling high-speed inspection. There is an advantage.

As described above, the present invention maintains the accuracy of the inspection, which is an advantage of the CT method by introducing a variable mesh into the existing CT method, and improves the speed of the inspection so that the longitudinal or cross-sectional image of the inspection area can be quickly restored. do. Therefore, there is an advantage that can be inlined on the production line can be useful for industrial use.

In addition, the single-layer inspection method of the PCB using the X-ray according to the present invention, since the upper and lower surfaces of the inspection target PCB 10 can be selectively inspected by acquiring an image once, the inspection efficiency is maximized and the high speed inspection is performed. In addition, since the inspection is performed by applying the algorithm that is most suitable for each inspection region, various mounting state defects of the components can be detected, and the release parts having various electrodes can be effectively inspected.

In addition, in the single-layer inspection method of the PCB using the X-ray according to the present invention, it is possible to selectively inspect the longitudinal section or the cross-sectional image of the inspection area as necessary, thereby maximizing the efficiency of the inspection as well as the inspection target PCB (10) According to the type of electronic components mounted there is an advantage that can perform the most suitable inspection to maximize the accuracy of the inspection results.

Claims (3)

After soldering an electronic component to a printed circuit board (hereinafter referred to as a PCB), the size of the pixel is changed according to the inspection importance of each part of the inspection area in the inspection area of the PCB to automatically check the mounting state and the soldering state of the electronic part. PCB tomographic inspection method using X-ray to determine the component mounting status and soldering status after retrieval and analysis of cross section or longitudinal section image of the inspection area of PCB by applying the variable mesh To The first process of acquiring the sloping perspective image of the inspection area of the PCB, and after restoring the cross-sectional image of the upper surface of the PCB by using the image obtained in the first process is a component failure item for the components mounted on the upper surface of the PCB In the second process of inspecting no part, movement, and polarity reversal, and if the result of the second process is a good product, the parts mounted on the upper surface of PCB are not soldered, unpaid, overpaid, bridged, voided. The third step of performing the inspection and the cross-sectional image of the lower surface of the PCB by restoring the image obtained in the first process and then the parts missing, moving, polarity reversal If the inspection result of the fourth process and the result of the fourth process is a good product, the inspection can be performed on the solderless items such as lead-free, unpaid, over-paid, bridge, and void for the components mounted on the lower surface of the PCB. If the inspection result of the fifth process and the third process and the fifth process is good and the component mounted in the inspection region is a component having a ball-type electrode, the PCB inspection region may be 6. A method of inspecting a single layer of a PCB using an X-ray after restoring a longitudinal cross-sectional image, wherein a sixth process of inspecting ball solder and voids, which is a poor soldering item, is performed on components mounted on upper and lower surfaces of the PCB. The method of claim 1, Before the second, third, fourth, fifth, and sixth processes, the number of inspection areas, the position and size of the inspection areas, an algorithm applied to each inspection area, for each inspection item, PCB tomography inspection method using the X-rays, characterized in that it further includes the process of adding, modifying, deleting the inspection plan including the parameters required for each algorithm. The method of claim 2, The algorithm applied to each of the inspection areas includes an algorithm for detecting defective items by calculating an average value of the image brightness in the inspection area and comparing the average value of the image brightness with a preset reference value, or a component in the binarized image by binarizing the inspection area. Alternatively, the area of the soldered area is calculated by comparing the area with a preset reference value and detecting the defective item by binarizing the area, or the ratio of the area of the component or soldered area to the area of the entire inspection area in the binarized image. ) And then compare this area ratio with a preset reference value to detect the defective item or binarize the inspection area to find the maximum length in the horizontal or vertical direction of the component or solder area in the binarized image. Egg to detect defective items by comparing with preset reference value By setting the threshold edge intensity for edge intensity, which is the brightness difference between neighboring pixels in the algorithm or inspection area, calculate the number of pixels with edge intensity larger than this threshold edge intensity, and compare the number of pixels with the preset reference value. Algorithm for detecting defective items or an algorithm for detecting defective items by calculating a variance value in the inspection area and comparing the variance value with a preset reference value. method of inspection.