JPH05102235A - Wirebonding method - Google Patents

Abstract

PURPOSE:To cut down the lead detecting time for enhancing the productivity of wirebonding work. CONSTITUTION:Before starting the wire bonding work, the numbers of detection ranges processible in an image plane are determined per detection point of a camera while teaching the coordinates of the detection ranges e1. e2... of individual leads and the ideal bonding points P1. P2... as well as considering the detection ranges e1, e2 of respective leads to the effective detection range E (DELTAX, DELTAY). At this time, within the actual bonding step, the leads the number of which has determined of an image to be taken-in from the camera are processed and image processed so as to process the positional difference from the regular position. Finally, the actual bonding point coordinates are processed conforming to the positional difference and the position of the ideal bonding point so as to perform the bonding step conforming to the processed bonding point coordinate.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a wire bonding method.

[0002]

2. Description of the Related Art Usually, when wire-bonding between an electrode of a semiconductor pellet and a lead of a lead frame,
The amount of deviation of the bonding point of the semiconductor pellet or the lead positioned at the bonding position from the ideal position is detected.

Conventionally, this detection has been performed using a camera as follows.

First, before the bonding work, images of a semiconductor pellet serving as a reference and a plurality of leads located in the periphery thereof are sequentially captured by a camera, and the captured images are binarized or multivalued by an image processing device. The characteristic patterns of the two areas on the semiconductor pellet and the ideal bonding point coordinates in each area, and the characteristic patterns in each lead and the ideal bonding point coordinates in each characteristic pattern are stored in the storage unit (teaching). That).

Therefore, in an actual bonding operation, first, when the semiconductor pellet is positioned at the bonding position which is the detection position or the position before it, a plurality of semiconductor pellets located in the two regions on the semiconductor pellet and in the periphery of the semiconductor pellet are detected by the camera. The images of the leads of the book are sequentially captured individually, and the captured images are binarized or multivalued by the image processing device. Next, each image is compared with one corresponding feature pattern stored in the storage unit of the image processing apparatus in advance by the teaching operation, and the mutual positional deviation amount is detected. Then, the actual bonding point coordinates are calculated from this deviation amount and the ideal bondin point coordinates. Wire bonding is performed on the calculated bonding point coordinates.

[0006]

However, according to the above-mentioned technique, in the actual bonding operation, the semiconductor pellets are detected in units of one or each of the leads. Therefore, particularly when the number of leads is large. However, it takes a lot of time for the detection work, and as a result, the bonding time becomes long, which hinders the improvement of productivity.

An object of the present invention is to provide a wire bonding method which can shorten the lead detection time and improve the productivity of wire bonding work.

[0008]

In order to achieve the above object, the present invention according to claim 1 sets a detection range for each lead before wire bonding work and takes in an image from a camera for image processing. It is characterized in that the number of detection ranges that can be processed in the screen is determined in advance, and in detecting the position of the lead, the determined number of the leads are processed from the one screen.

According to the second aspect of the present invention, before the wire bonding work, the detection range for each semiconductor pellet is set, and the detection range that can be processed in one screen which is taken in from the camera and image-processed is set. It is characterized in that the number of the semiconductor pellets and the leads are processed in the one screen in order to detect the positions of the semiconductor pellets and the leads around the semiconductor pellets.

[0010]

According to the first aspect of the present invention, a plurality of leads located around the semiconductor pellet are detected.

According to the second aspect of the present invention, the semiconductor pellet and the leads located around the semiconductor pellet are detected in units of a plurality of semiconductor pellets.

[0012]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of a wire bonding apparatus used for carrying out the present invention, FIG. 2 is a plan view of a lead frame, FIG. 3 is a view showing a lead detection range, and FIG. 4 is an effective detection. FIG. 5 to FIG. 8 are diagrams showing the range, and FIG. 5 to FIG. 8 are diagrams showing the lead detection method.

In the figure, the semiconductor pellets 2 are mounted on the lead frame 1 by the pellet mounting device installed in the previous step, and are positioned at the bonding position by a carrier device (not shown). A bonding head 10 is provided at the bonding position. The bonding head 10 is mounted on an XY table 11, and a capillary 13 is attached to the tip of a bonding arm 12 which is held so as to be vertically movable.

A camera 14 capable of picking up an image of a predetermined position on the lead frame 1 is movably arranged in the XY direction above the lead frame 1 positioned at the bonding position, and a control device is provided via an image processing device 15. It is connected to 16. This camera 14 has a bonding head 10
It may be fixed to the main body. Control device 16
Controls the movement of the bonding head 10 and the XY table 11, and has a storage unit 17, a calculation unit 18, and a comparison unit 19. In the figure, reference numeral 20 is a monitor on which an image captured by the camera 14 is displayed.

Next, the operation will be described.

Before the bonding work, while watching the monitor 20, the teaching work is performed as follows.

First, the images of the leads 1a are sequentially captured by the camera 14, and the characteristic pattern of each lead and its detection range ei (Δxi, Δyi) are set in the order of wire bonding as shown in FIG. It is stored in the storage unit 17. At this time, the coordinates (xi, yi) of the ideal bonding point Pi are also set and stored. The detection range ei may be automatically set from the lead width.

Next, as shown in FIG. 4, an effective detection range E of the camera 14 (ΔX, ΔY: a detection range in which the detection accuracy of the camera is not adversely affected by lens distortion etc.) is set and stored in the storage unit 17. Remember. This effective detection range is
It is set as all or part of the range projected on the monitor 20.

When the above setting is completed, the processing unit 18 of the control unit 16 can detect a processable detection range in one screen which is captured from the camera 14 and binarized or multivalued by the image processing unit 15. Determine the number of.

This determination method will be described with reference to FIGS. 5 to 8.

First, the lead L to be bonded first
The detection range e1 of 1 (hereinafter referred to as the first lead L1) and the effective detection range E of the camera 14 are compared. This is the lateral length Δ in the detection range e1 and the effective detection range E.
x1 and ΔX, and vertical lengths Δy1 and ΔY are compared with each other to determine the magnitude relationship between them. Δx1 ≦ ΔX,
Since Δy1 ≦ ΔY is established, it is determined that the detection range e1 is included in the effective detection range E (see FIG. 5). Therefore, next, a comparison is made between the minimum detection rectangle S2 (Δm2, Δn2) including both the detection range e1 and the detection range e2 of the second lead L2 and the effective detection range E. This is to compare the horizontal lengths Δm2 and ΔX and the vertical lengths Δn2 and ΔY in the quadrangle S2 and the effective detection range E, respectively, and determine the magnitude relationship between them, where Δm2 ≦ ΔX and Δn2 ≦ ΔY are satisfied. Since it is satisfied, it is determined that the quadrangle S2 is included in the effective detection range E (see FIG. 6). In the same manner as described above, the determination is performed in order, such as the third lead L3 and the fourth lead L4. In this example, as shown in FIG. 7, the detection ranges e1 and e1 from the first lead L1 to the fourth lead L4
The smallest quadrangle S4 (Δm containing all 2, e3, e4)
4, Δn4) is compared with the effective detection range E, Δm4 ≧ ΔX and Δn4 ≦ ΔY are satisfied for the first time, and thus it is determined that the quadrangle S4 is not included in the effective detection range E. Therefore, in one effective detection range E, the first lead L1
If it is determined that the third to the third leads L3 can be processed, these are set as the first block, and the arithmetic unit 18 sets each detection range e.
Center position Pa of the smallest quadrangle S3 including 1, e2, e3
Calculate the coordinates of. Then, with the center position Pa as a base point, the relative relationship (relative positional relationship in the X direction and the Y direction, that is, vector information) with the ideal bonding points P1 to P3 in each lead L1 to L3 is calculated, and the storage unit 17 is calculated.
(See FIG. 8). After that, the same processing is performed for the fourth and subsequent reads.

Next, a method of detecting the position of the lead actually positioned at the bonding position will be described. First, the camera 14 moves to the position where the center of the field of view is the first detection point, that is, the center position Pa, and captures the lead images corresponding to the three leads L1 to L3 as the first block. The captured image is stored in the storage unit 17 after being processed by the image processing device 15. Then, the detection areas e1, e2, and e3 are cut out from this image, and the comparison unit 19 performs matching processing with the corresponding feature patterns stored in the storage unit 17, respectively. Then, the comparison unit 19 detects the positional deviation amount of the lead, and finally the arithmetic operation unit 1 based on the deviation amount, the coordinates of the center position Pa, and the vector information of the ideal bonding points P1 to P3.
At 8, the actual bonding point coordinates are calculated.

The camera 14 uses three leads L1 to L3.
When the lead corresponding to is read and stored in the storage unit 17, the camera 14 moves to the next detection point, and the same processing is sequentially performed on the leads after the fourth lead for each block. It

Regarding the positional deviation of the semiconductor pellet 2, the positional deviation amounts at two points on the semiconductor pellet are detected as in the conventional case, and all the bonding points on the semiconductor pellet are corrected and calculated based on this.

When all the bonding point coordinates are calculated in this way, the controller 16 operates the XY table 11 and the bonding head 10 based on the bonding point coordinates to perform bonding.

According to the above embodiment, the detection range ei is set for each lead, and the number of processable detection ranges in one screen which is captured from the camera 14 and image-processed while considering the detection range ei. In addition to the above determination, in the wire bonding operation, the determined number of leads are processed in one screen, so that the detection time can be shortened and the productivity of the bonding work can be improved.

In the above-described embodiment, the determination is made after the teaching of the detection range of each lead and the ideal bonding point coordinates is finished, but the invention is not limited to this. May be performed.

Further, in the above-described embodiment, the read processing for a plurality of lines is carried out in one screen captured by the camera 14, but as shown in FIG. When there are a plurality of images of the semiconductor pellets 2, the detection range ea including the leads located in the periphery of each of the semiconductor pellets is set in advance, and the determination is performed in the same manner as in the above-described embodiment. It is possible to process a plurality of semiconductor pellets and the corresponding leads in one screen which is taken in and image-processed, which can significantly improve the productivity of the bonding work.

[0030]

According to the present invention, the lead detection time can be shortened and the productivity of the bonding work can be improved.

[Brief description of drawings]

FIG. 1 shows a configuration diagram of an embodiment of a wire bonding apparatus used for implementing the present invention.

FIG. 2 shows a plan view of a lead frame.

FIG. 3 is a diagram showing a lead detection range.

FIG. 4 is a diagram showing an effective detection range.

FIG. 5 is a diagram showing a lead detection method.

FIG. 6 is a diagram showing a lead detection method.

FIG. 7 is a diagram showing a lead detection method.

FIG. 8 is a diagram showing a lead detection method.

FIG. 9 is a schematic diagram showing a second embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 lead frame 2 semiconductor pellet 10 bonding head 14 camera 15 image processing device 16 control device 17 storage unit 18 arithmetic unit 19 comparison unit 20 monitor ei detection range E effective detection range

Claims (2)

[Claims] 1. The position of each lead is detected by positioning a semiconductor pellet mounted on a lead frame at a detection position, capturing images of a plurality of leads located around the semiconductor pellet with a camera, and performing image processing. Then, the amount of deviation from the regular position is obtained, and in the wire bonding method in which the preset bonding point coordinates are corrected based on the obtained amount of deviation, before the wire bonding work, for each of the leads. In addition to setting the detection range, the number of processable detection ranges in one screen captured from the camera and subjected to image processing is determined in advance.
The wire bonding method is characterized in that the determined number of the leads are processed from the screen. 2. A plurality of semiconductor pellets mounted on a lead frame are sequentially positioned at detection positions, and a camera is used to capture images of the semiconductor pellets and a plurality of leads located around the semiconductor pellets for image processing. In the wire bonding method, the position of each semiconductor pellet and each lead is detected to obtain the amount of deviation from the normal position, and the preset bonding point coordinates are corrected based on the obtained amount of deviation. Before the bonding work, the detection range per semiconductor pellet is set, and the number of processable detection ranges in one screen which is captured from the camera and image-processed is determined in advance. When detecting the position of the lead, a few minutes before the determined number from the one screen A wire bonding method, wherein the semiconductor pellet and the lead are treated.