KR100363707B1 - Semiconductor Element Inspection Method - Google Patents

Abstract

The present invention relates to a method for inspecting a semiconductor device, the method comprising: dividing an interior of a semiconductor device into a pad area, an inner lead area, an outward area, a tape area, and a volume area in order to distinguish characteristics of an object to be measured; Selecting a light source generated from a light source generator to optimally inspect each area, enlarging an object to precisely irradiate the selected light source to each area of the object, and according to the image of the enlarged object And changing the photographing range of the scanhead in order to photograph the area.

Description

Semiconductor element inspection method

The present invention relates to a method for inspecting a semiconductor device, and in particular, in a region having different characteristics for each functional element, not only for a lead frame, which is an essential material of a semiconductor manufacturing process, but also for devices with finer defects, such as wafers, LCDs, and TFTs. The present invention relates to a method of inspecting a semiconductor device that enables to perform a quick and precise inspection of the same.

A general area scan vision inspection system is basically a CCD camera (which has a digital or analog type), a light source that is combined with the CCD camera to provide illumination to the object, and an object captured using the CCD camera and the light source. It consists of a CPU that controls the image.

In general, the inspection area of the field of view (FOV) defined in the inspection of semiconductor devices is the resolution of the CCD camera, the required size, quantity, shape, and other inspection of the inspection object. Limited by purpose.

For example, in the case of a CCD camera having a resolution of 2,048 x 2,048 pixels of a digital type currently used in general (eg, manufactured by KODAK, Japan), an object having a specific defect is irrespective of an object to be inspected. Looking at the case of checking, we get the following result. In other words, when the size of the defective object was about 5 µm, the inspection could not be performed beyond the 10 mm x 10 mm area.

It is also known that digital cameras have a resolution of 2,048 × 2,048, while analog cameras have a resolution of 1,500 × 1,300. Therefore, when the analog camera is used as described above, if the size of the defective object is about 5 μm, the FOV becomes smaller. In this case, the appropriate inspection area is limited to 5mm × 5mm.

A method of inspecting a semiconductor device using a region scan vision inspection system that is generally used will be described with reference to FIGS. 1 and 2.

1 is an explanatory diagram for explaining a general inspection method of a semiconductor device, (a) of FIG. 2 shows the shape of an object group composed of a plurality of objects, and (b) is a diagram illustrating the audit procedure of each object. to be.

In FIG. 1, reference numeral 10 denotes an object to be measured, 11 denotes a table on which the object 10 can be placed, 50 denotes a light source for irradiating a constant light onto an object by combining a CCD camera, and 100 transfers inspection equipment. A handler for supplying an object to be inspected, 120 is a CCD camera as a photographing means of the object, a represents a working distance at an interval between the object 10 and the CCD camera 120, and b is a CCD camera ( 120 indicates a moving direction, and c indicates a moving direction of the object 10.

In FIG. 2A, each object A, B, C, D... Is arranged for inspection, and in FIG. 2B, a plurality of semiconductor packages are inspected for each unit package.

The inspection method of a semiconductor device generally used includes a moving camera method of moving a camera and a moving table method of moving a table on which an object is placed.

Referring to FIGS. 1 and 2, a method of inspecting a general semiconductor device will be described in more detail.

First, the moving camera method is a method in which the CCD camera 120 mounted vertically to the inspection position recognizes the object 10 while moving in the horizontal direction.

That is, when the inspection is performed by forming a group of objects in which a plurality of objects are arranged in sequence as shown in (a) of FIG. 2, as shown in (b) of FIG. 2, the CCD camera 120 is the first. After photographing the object A, the FOV of the camera is determined, and the object 10 is inspected. Then, the second object B is inspected in the horizontal direction. Likewise, when the inspection is completed after the FOV is determined for the second object B, the CCD camera 120 moves to the third object C to inspect the image. In this manner, the individual objects A, B, C, and D included in the object group 12 are sequentially inspected. At this time, in order to increase the illuminance of the object, illumination having a constant brightness is provided from the light source 50 combined with the CCD camera 120.

Next, the moving table method is a method in which an inspection of the entire area is performed by moving the table 11 on which the object 10 is placed in a fixed position, as opposed to the moving camera method. In other words, in the object group 12 as shown in Fig. 2A, the first object A is positioned within a predetermined FOV range of the camera, the inspection is performed, and then the table 11 is moved to move the object B. ), The object (C) and the object (D) in the order of inspection. Of course, the camera's field of view (FOV) when each object is moved must be adjusted separately. Similarly, illumination with constant brightness is provided from the light source 50 combined with the CCD camera 120 to increase the illuminance of the object.

The precision of the semiconductor inspection system depends on the FOV size, and when the FOV is small when the resolution of the camera has a constant value, the inspection can be performed as precisely. Therefore, in order to inspect the entire object group 12 composed of a plurality of objects at once, the working distance a between the CCD camera 120 and the object 10 should be increased. However, if the FOV is increased by increasing the working distance (a), a problem arises that the precision is lowered. In addition, since the movement of the CCD camera or the table is mechanically performed by a motor, the accuracy of the moving distance may be reduced, so it is difficult to use in a system requiring a precise level.

In this case, both of the above methods perform repeated inspections for each FOV of the mobile hardware assembly, which lowers the stability of the inspections and is sensitive to vibrations, so it takes a lot of time to scan the entire area. do.

In addition, a line scan camera may be used as an area scan vision inspection method. In general, when a line scan camera is used, a relatively inexpensive camera capable of obtaining 6,000 × 1 pixels and 4,000 × 1 pixels may be used. However, such a line scan camera has to move precisely in a very small unit of 1 pixel unit, and thus has a problem that a good image can be obtained only by scanning at a very slow speed and productivity is low.

As described above, in the general area scan vision inspection system, the FOV, which is an area that the camera can recognize at a time, is limited, and the FOV is an area of up to 10 mm x 10 mm.

This is a case where defect inspection is performed based on lead frame, which is an essential material of semiconductor manufacturing process, but when the inspection of other fields such as wafer, LCD, TFT, etc. with finer defect standards is performed, such FOV is smaller. Limited. Therefore, in order to perform precise inspection, a small FOV should be used to combine multiple camera assemblies into a multi-array.

In addition, when inspecting using only one camera assembly, the object to be inspected is inspected by step moving by the FOV given by the camera, or the camera itself is moved by the inspecting area step by step. Larger test objects will be examined. Therefore, by using the inspection method of a general semiconductor device having such a limitation, in the manufacturing process such as In Line Process, Back End Process, TFT, LCD, etc. Unnecessarily increased inspection time and very low productivity.

In other words, the output per hour is lowered, the manufacturing cycle is lengthened, and the manufacturing cost is increased. However, at present, the manufacturer continues to use this technology because it has not found a way to solve this problem.

Moreover, current semiconductor device manufacturers and general TFT, LCD and other semiconductor device manufacturers are constantly being demanded by end users for increased productivity, reduced manufacturing costs, increased manufacturing cycles, and faster supply of products.

Also, in the manufacture of semiconductor devices, there is a tendency to be compact, thin, multi-sized, highly integrated, and formed into one lead frame, substrate, film pattern, wafer, and the like, and the number of integrated devices per unit area is increased. As a result, the size of unit objects is growing.

For example, in the case of SOIC leadframes, the width is increased from 30mm to 75mm, and the length is largely reduced from 100mm to 270mm, and this trend is present in all processes. In addition, large caliber of all packages such as PLCC, TSOP, QFP, TQFP, BGA, FBGA, μBGA, TSSOP, SSOP, MLP, CSP is rapidly progressing.

However, since the FOV of the optical cameras of the conventional area scan vision inspection equipment has a maximum area of 10 mm × 10 mm in the case of a CCD camera having a resolution of 1500 × 1300 pixels, it has a time limit for rapid defect inspection.

The present invention is to solve the above-mentioned problems, the purpose of which is to precisely perform the defect inspection in a short time and shorten the manufacturing cycle, even if the number of semiconductor devices to be measured to reduce the manufacturing cost To provide a method for inspecting a semiconductor device that can be.

1 is an explanatory diagram illustrating a method of inspecting a general semiconductor device.

FIG. 2A shows the shape of an object group composed of a plurality of objects, and FIG. 2B is a view for explaining the inspection procedure of each object.

Fig. 3 is a block diagram showing the structure of an inspection system for executing the inspection method of a semiconductor device according to the present invention.

4 is a cross-sectional view showing an internal configuration of a semiconductor package for explaining a method for inspecting a semiconductor device according to the present invention.

5 is an explanatory diagram for explaining an outline of a data inspection function of a lead frame as a semiconductor device according to the present invention;

Fig. 6 is a schematic diagram illustrating an overview of functional sector inspection in the semiconductor device inspection method according to the present invention.

FIG. 7A is an explanatory diagram for explaining an inspection method using an area scan vision system according to the present invention, and FIG. 7B is an enlarged view of the object group 12 in FIG. 7A.

* Description of the symbols for the main parts of the drawings *

10: object 11: the table

12: object group 20: semiconductor package

21: pad area 23: inner lead area

25: Outstanding area 27: Tape area

29: volume area 30: lead frame

34: lead frame stream 40: recording mechanism

50: light source 100: handler

110: scan head 112: zoom lens

114: telescope 120: CCD camera

130: light source 132: light source generator

134: light source selection switch box 140: CPU

150: scan controller 160: servo driver

a: working distance b: direction of camera movement

c: direction of table movement

The inspection method of a semiconductor device according to the present invention,

Dividing the inside of the semiconductor device into a pad region, an inner lead region, an outer region, a tape region, and a volume region to distinguish characteristics of an object to be measured;

Selecting a light source of red (R), green (G), and blue (B) generated from a light source generator to optimally inspect each of the regions;

Enlarging an object to precisely irradiate the selected light source to each area of the object;

The method may be achieved by a method of inspecting a semiconductor device, comprising changing a photographing range of a scan head in order to photograph an area according to an image of the enlarged object.

As a result, defects and characteristics can be inspected with high speed and accuracy for the plurality of semiconductor devices to be measured.

The above objects and features and advantages of the present invention will become more apparent from the following detailed description taken in conjunction with the accompanying drawings.

(Example)

The inspection method by the inspection method of the semiconductor element by this invention is demonstrated, referring drawings.

3 is a cross-sectional view showing an internal configuration of a semiconductor package for explaining a method for inspecting a semiconductor device according to the present invention.

First, in FIG. 3, reference numeral 100 denotes a handler for transferring inspection equipment, supplying an object to be inspected, and exchanging signals with an external device such as a CPU, and 110, a scan head for moving and capturing an image according to an FOV, and moving at a very high speed. 112 is a zoom lens for adjusting a FOV in which a deviation occurs according to the size of an object, 114 is a telescope attached to the CCD camera to enlarge and observe an image of the object, 120 is a CCD camera for capturing an image, 130 is a light source unit that provides an auxiliary light for clearly distinguishing an object, and 140 is a built-in plurality of serially connected I / O ports and includes the handler 100, the light source unit 130, the CCD camera 120, and a scan controller. CPU to connect and control a controller, 150 is a scan controller to control the scan head 110, 160 is a servo drive to drive the scan head 110 Represent each live. Here, the scan head 110, the zoom lens 112, the telescope 114, and the CCD camera 120 form a unique photographing mechanism 40 of the present invention.

As shown in FIG. 3, in the semiconductor device inspection method, the object 10 to be measured is placed on the handler 100 and the inspection is performed.

When the object 10 to be measured is placed on the handler 100, the scan head 110 prepares to scan the object 10 by using a built-in actuator (not shown) such as a step motor or a galvanometer. Will be

Next, the light source 130 is controlled to more clearly identify the object 10. In this case, the light source unit 130 is provided with a plurality of light source generators 132 for generating different wavelengths. The light source generator 132 generates light having a wavelength for photographing the target object most clearly according to the shape or size of the object 10. For example, when an object has a relatively dark color and the thickness is thick, it is brightened to generate bright red color light for high-precision shooting, and when the object has a relatively light color and is thin In this case, blue light is generated to always photograph an object with high precision. At this time, the light generated from the light source generator 132 may be a color of the red (R), green (G), blue (B) system, but these colors have an appropriate color according to the shape, size, and state of the object. It can also be combined with light.

In addition, the light source 130 further includes a light source selection switch box 134 for selecting the light generated from the light source generator 132 and providing the light into the telescope 114. The light source selection switch box 134 selects a suitable light source according to the shape, size and state of the object 10.

The light source drawn out through the light source selection switch box 134 is irradiated into the telescope 114 coupled to the zoom lens 112 through a light guide to supply light to the FOV of the object 10. The FOV is enlarged to make precise adjustments. In this case, the light source unit 130 may receive an analog signal of a predetermined (approximately 0 to 5V) from the I / O port built in the CPU 140 and adjust the illumination according to the target object 10.

In this manner, when the FOV is determined according to the size of the image of the object 10, the object 10 is photographed using an imaging device such as a CCD camera 120, and the image data of the photographed object 10 is stored in the CPU ( It exchanges signals with other devices through the I / O port which is serially connected to the 140).

The scan controller 150 receives data about an area to be scanned from a main computer (shown in the figure) through a serial port and controls an actuator such as a galvanometer embedded in the scan head 110. At this time, the control amount detected through the scan controller 150 is provided to the servo drive 160 connected in parallel to each other to substantially drive the scan head 110.

That is, the scan data input through the serial port 142 built in the CPU 140 is uploaded to the scan controller 150 to sufficiently recognize the target area of the object 10. Galvanometer (not shown), which is a built-in actuator, is controlled, and the scan head 110 is driven according to the control amount through the servo drive 160.

Here, the recognition range varies depending on the actuator built into the scan head 110. When using the galvanometer with high precision as the actuator, the object 10 may be scanned while rotating up to ± 20 °. Can be. However, the angle range in which the object 10 can be scanned without distortion by maintaining the linearity of the scan area is approximately ± 10 °, and within this range, the error may be minutely scanned.

In other words, when the FOV is about ± 10 °, accurate measurement is possible from the range of 2.3 μm × 2.3 μm to 330 mm × 330 mm. This is a significant improvement over 10mm x 10mm, which is the maximum FOV of conventional general area scan vision inspection systems.

Therefore, according to the semiconductor device inspection method according to the present invention, a large number of objects can be inspected with high precision in a fast processing process at once.

Next, a method of inspecting a semiconductor device according to the present invention will be described in more detail with reference to FIG. 4.

4 is a cross-sectional view showing an internal configuration of a semiconductor package for explaining a method for inspecting a semiconductor device according to the present invention.

In FIG. 4, 20 represents a unit semiconductor package, 21 represents a pad area attaching an integrated circuit chip such as an IC, and 23 represents an inner lead area for wire bonding between a chip and a lead wire of an integrated circuit. lead area, 25 is an outer lead area for bonding the external lead wires, and 27 is a tape area for cutting and pressing the tape according to the standard and applying heat when pressing the cap to install the tape. Area, and 29 represents a volume area representing an outline of a package.

The pad region 21, the inner lead region 23, the outlier region 25, the tape region 27, and the volume region 29 constituting the semiconductor package 20 have different characteristics. That is, since the plating state, contamination, and the size of the lead wires in the respective parts are different from each other, the inspection method for these parts must also be different, so that accurate inspection can be performed.

However, according to the inspection method of a semiconductor device generally applied as described above, regardless of the specific characteristics of the object to be measured, either the mobile camera method or the moving table method, by providing a constant distance and a constant light source to inspect the product The defective rate of becomes high. However, in the method of inspecting a semiconductor device according to the present invention, the internal components of the semiconductor package are classified and inspected by the method as shown in FIG. 3.

That is, in the pad region 21, the chip attachment position and contamination of the integrated circuit, the damage state, the size and the conditions such as plating or distortion are determined, and a light source suitable for this is selected from the light source generator (see 132 of FIG. 3). From the zoom lens 120 and the telescope 114 is controlled to enable an optimal precision measurement.

In the inner lead region 23, the wire bonding state between the integrated circuit chip and the lead wire, whether the plating is sufficient, and whether the prescribed size of the lead wire is correct are examined in the same manner.

In this manner, the outward area 25, the tape area 27, and the volume area 29 also provide light sources with wavelengths suitable for the characteristics of each part, and precise control is possible by the control of the photographing mechanism 40. have.

In the above method for measuring and inspecting the respective parts, it is a feature that all internal components of the semiconductor device 20 can be inspected at once.

As described above with reference to FIG. 5, a method of inspecting a semiconductor device according to the present invention, which can inspect all the components of a semiconductor package at once, with high speed and accuracy, and by expanding the measurement range as compared with the related art. Shall be.

5 is an explanatory diagram for explaining an outline of a data inspection function of a lead frame as a semiconductor device according to the present invention.

In FIG. 5, the lead frames 30 are arranged in a singular or matrix type. At this time, the lead frame 30 has sizes of the pad region (see 21 in FIG. 4), the inner lead region, the outred region, and the package region. Taking this example as an example, the inspection method by the inspection method of the semiconductor device according to the present invention will be described.

First, in FIG. 3, a light source suitable for each area is generated from the light source generator 132 and irradiated to each area of the lead frame 30 to be measured. At this time, the zoom lens 112 and the telescope 114 to enlarge the respective areas through the CCD camera 120 to obtain the information of each area measured, and to scan according to the dimensions according to each area The controller 150 and the servo drive 160 are driven. By such an operation, the scanhead 110 searches for an area to be measured at high speed, thereby enabling highly accurate and stable inspection. In this case, the detailed operation of the semiconductor device inspection system according to the present invention is the same as described above and will be omitted. However, it can be seen that it is possible to obtain various region data of the semiconductor device such as a lead frame to be measured accurately and at high speed.

Next, a method of inspecting a plurality of semiconductor elements arranged in a plurality or matrix form will be described with reference to FIG. 6.

Fig. 6 is a schematic diagram illustrating a functional sector inspection overview in the semiconductor device inspection method according to the present invention.

6 to 34 show that the lead frames are arranged in a stream form.

For convenience of explanation, the description will be made by dividing the first step into the fifth step.

First, in the first step, as shown in FIG. 5, measurement is performed to obtain precise data for each region of the lead frame. At this time, the inspection method of the semiconductor device is the same as described above will be omitted. Through such a method of inspecting semiconductor devices, high-speed and precise unit devices can be inspected.

Next, the second step will be described. In practice this second step is carried out simultaneously with the inspection of the first step. That is, by controlling the scan head 110 by the inspection system of the semiconductor device according to the present invention shown in FIG. 4, the measurement range of the scan head 110 is extended and inspected. Of course, the measurement range (FOV) of the camera for the object to be applied is subdivided into each area to detect the same FOV as the measurement result in the first step.

Similarly, the third, fourth and fifth stages are inspected at the same time. Therefore, the inspection can be performed on the desired number of semiconductor devices at the same time as compared to the inspection by the semiconductor devices one by one by the conventional method, so that high-speed and precise inspection is possible.

That is, even if an object of any size is inputted by a CCD camera (not shown) or the like, the image of the object 10 inputted by controlling the mirrors (not shown) in the X and Y-axis directions built in the scan head 110 is controlled. The size is inspected precisely and printed out. The information of the object 10 to be input at this time may be two-dimensional information provided in the X, Y and Z directions, or may be processed as three-dimensional information provided from the X, Y and Z axes.

Generally, there are AC actuator, DC motor, step motor, and galvanometer for controlling actuators of mirror, and it can be used any one of them, but it is more precise considering the demand for high precision and high productivity. It is desirable to use a galvanometer that can measure data.

In practice, in order to obtain an image having a resolution of 5 μm using a general optical digital camera (2,048 × 2,048), the FOV should be 1 cm × 1 cm. However, the camera itself is expensive (approximately 3.6 times that of 1,024 × 1,024), and scanning such a small area as the camera or table moves is inefficient and cost-effective.

On the other hand, the inspection method of the semiconductor device according to the present invention is configured such that the FOV has an ultra-precision resolution of 2.3 mu m or less for an area of up to 300 mm x 300 mm.

Hereinafter, an inspection method for precisely measuring by using the area scan vision inspection system of the present invention will be described with reference to FIG. 7.

FIG. 7A is an explanatory diagram for explaining an inspection method using an area scan vision system according to the present invention, and FIG. 7B is an enlarged view of the object group 12 in FIG. 7A. to be.

As shown in (a) and (b) of FIG. 7, in the area scan vision system according to the present invention, both the object group 12 and the scan head 110 are installed at fixed positions to perform inspection. That is, the object group 12 covering the range θ according to the desired FOV is reflected inside the scan head 110 through the zoom lens 112 and the telescope 114 using a CCD camera (not shown). By enlarging the image of the object, the entire object group 12 can be taken simultaneously.

At this time, as shown in (b) of FIG. 7, the area scan vision system according to the present invention subdivides the FOV into several areas to accurately capture an image of individual objects A, B, C, and D.

The images taken by dividing into several pieces as described above are combined through a programming process in a main computer (not shown), and then formed and inspected as the original images. The finer the FOV, the higher the precision.

In addition, the movement of the FOV is made by a scanner without moving a camera or an object as in the prior art, and the moving distance of the FOV can be controlled to a precise range up to at least 2.3 μm. In addition, the FOV movement time can be controlled to at least 2.3㎛.

In addition, as described above, a zoom lens 112 is attached to the scan head 110 to more precisely adjust the FOV according to the size of the object. By attaching the zoom lens 112, inspection is performed by finely adjusting the FOV in an area where high precision is required, and inspection is performed by enlarging the FOV in an area not otherwise. In this case, the light source may be controlled according to the content of the object to improve the sharpness of the object.

As described above, according to the method of inspecting a semiconductor device according to the present invention, even when the quantity of semiconductor objects to be measured is large, accurate defect inspection can be performed in a short time, a manufacturing cycle can be shortened, and manufacturing cost can be greatly reduced.

Preferred embodiments of the present invention described above are disclosed for the purpose of illustration, and those skilled in the art may make various modifications, changes, substitutions, and additions through the spirit and scope of the present invention as set forth in the appended claims.

Claims (4)

As the inspection method of a semiconductor device, Dividing the inside of the semiconductor device into a pad region, an inner lead region, an outer region, a tape region, and a volume region to distinguish characteristics of an object to be measured; Selecting a light source of red (R), green (G), and blue (B) generated from a light source generator to optimally inspect each of the regions; Enlarging an object to precisely irradiate the selected light source to each area of the object; And changing a photographing range of a scan head in order to photograph an area according to the image of the enlarged object. The method of claim 1, And enlarging the object using a zoom lens and a telescope. As the inspection method of lead frame, Dividing the inside of the semiconductor device into a pad region, an inner lead region, an outer region, a tape region, and a volume region to distinguish characteristics of an object to be measured; Selecting a light source of red (R), green (G), and blue (B) generated from a light source generator to optimally inspect each of the regions; Enlarging an object to precisely irradiate the selected light source to each area of the object; And changing a photographing range of a scan head in order to photograph an area according to the image of the enlarged object. The method of claim 3, wherein And enlarging the object using a zoom lens and a telescope.