JPH039621B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is a face-down method in which bumps of a chip are bonded by directly overlapping the pads or leads of a substrate (the case of leads will be described below) to manufacture semiconductor devices. The present invention relates to an alignment method for chip bonding.

[Prior Art] Conventionally, in such chip bonding, the alignment between the leads of the board and the bumps of the chip was performed in the fifth step.
This is done by the method shown in FIG.

In the method shown in FIG. 5, a chip 2 is placed above a substrate 1 which is positioned and held by a bonding station (not shown), and is held by a bonding tool 3, and a camera 4 disposed below the substrate 1 is used to detect the substrate 1. Copy chip 2 on the same screen, and
The leads 1a of the chip 2 and the bumps 2a of the chip 2 are aligned so that they overlap. In the figure, 5 is a half mirror, 6
indicates a lamp.

In the method shown in FIG. 6, a camera 10 and a half mirror 11 are disposed above the substrate 1, and the light reflected from the substrate 1 by the lamp 12 and the light reflected by the chip 2 by the lamp 13 are transferred to the substrate 1 using the half mirror 11. The chip 2 is photographed on the same screen of the camera 10 and aligned so that the leads 1a of the substrate 1 and the bumps 2a of the chip 2 overlap.

[Problems to be Solved by the Invention] Since the method shown in FIG. 5 projects the chip 2 through the substrate 1, it cannot be applied unless the substrate 1 is transparent.

Since the method shown in FIG. 6 uses reflected light, it can be detected even if the substrate 1 is opaque. However, in order to minimize the number of half mirrors 11 used and to simplify the structure, this method requires that the center of the substrate 1 and the center of the chip 2 be on a vertical line, and that the center of the substrate 1, the chip 2, and the half mirror 11 must be horizontal, and all angles of incidence and reflection must be at the same angle θ ₁ . For this reason, a table on which the substrate 1 is placed, a bonding tool 3, a camera 10,
The arrangement of the half mirror 11 may be restricted, making installation difficult. Furthermore, there is a drawback that if even one of the angles θ ₁ deviates, a positional deviation occurs.

An object of the present invention is to provide a chip bonding alignment method that is applicable regardless of whether the substrate is transparent or opaque, and allows for more accurate alignment.

[Means for solving the problem] The above purpose is to place a board on a board positioning stand,
The chip is adsorbed by a bonding tool, the board on the board positioning table is photographed by a board detection camera placed above the board positioning table, and the chip detection camera is placed at a location away from the board positioning table. The chip adsorbed to the bonding tool was photographed, and the two cameras were used to take out the chip.
This is achieved by combining and displaying two images on one monitor using a combining circuit, and aligning the board and chip.

[Operation] The board and chip are captured as two images using two cameras installed in separate locations, and the composite circuit is used to combine and display the images on one monitor. Align at a remote location. Therefore, it can be applied regardless of whether the substrate is transparent or opaque, and alignment can be performed easily and accurately. Furthermore, the camera can be installed at any position, and the angle of the camera can be easily set, making it difficult for positional deviation to occur.

[Example] An example of the present invention will be described below with reference to FIGS. 1 to 4. As shown in FIG. 1, a substrate detection camera 21 that detects the substrate 1 placed on the substrate positioning table 20 and a chip detection camera 22 that detects the chip 2 held by the bonding tool 23 are arranged. It is set up. Board detection camera 2
1 is fixed at point B of the substrate positioning table 20, which is directly above the reference point A (for example, the bonding position).
The chip detection camera 22 is rotatable and movable in the X and Y directions, and is disposed at a point C that is distant from a reference point A. The bonding tool 23 has upper and lower (Z)
It is installed so that it can move and rotate in the XY directions.
It can be located at point D, which is directly above point C, and point E, which is directly above point A. Further, a substrate feeder 24 for feeding and feeding the substrate 1 to and from the substrate positioning table 20 is provided before and after the substrate positioning table 20.

The two cameras 21 and 22 are connected to a deflection circuit 31 by a timing circuit 30, as shown in FIG.
The cameras 21 and 22 convert the video beams into signals, and the synthesis circuit 32 converts the two cameras 2
The signals from 1 and 22 are combined and amplified, and the monitor 33
It is now possible to input a composite image into the image.

Next, the method of the present invention will be explained. First, as shown in FIG. 3, a jig 43 stores two transparent plates 41 and 42 with a mark 40 or an alternative thereon on the point A of the substrate positioning table 20 with the marks 40 aligned. to be fixed. This is photographed by the board detection camera 21. Naturally, in this state, 2
The marks 40 on the two transparent plates 41 and 42 are overlapped and transferred to the monitor 33 (see FIG. 2).

Next, the bonding tool 23 is moved in the X and Y directions using an electric manipulator to position it at point E directly above point A. Next, bonding tool 2
3 is lowered to adsorb the uppermost transparent plate 41, and is again raised to point E, or moved in the XY direction while rising, and positioned at point D. The distances X ₁ and Y ₁ of the bonding tool 23 in the X and Y directions at this time are stored in a computer or the like. Bonding tool 23
When the bonding tool 23 is located at point D, the chip detection camera 22 photographs the transparent plate 41 that is attracted to the bonding tool 23. Therefore, the two images of the remaining transparent plate 42 captured by the board detection camera 21 and the transparent plate 41 captured by the chip detection camera 22 are used as a second image.
A synthesized image is created through the synthesis circuit 32 shown in the figure, and 1
The position and magnification of the chip detection camera 22 are adjusted so that the alignment marks 40 on the transparent plates 41 and 42 coincide with each other. The position and magnification of the board detection camera 21 are determined by the bonding tool 2.
The setting is made before adsorbing the transparent plate 41 in step 3. This completes the adjustment of the magnification and position of the two cameras. Therefore, the transparent plates 41, 42 and the jig 4
Remove 3.

Next, as shown in FIG. 1, the substrate 1, which is an actual sample, is positioned and placed on the substrate positioning table 20 at point A, the chip 2 is attracted to the bonding tool 3, and the substrate detection camera 21 is placed at point B. , the bonding tool 23 is positioned at point D, respectively. Then, the board 1 is detected by the board detection camera 21, and the chip 2 is
When each image is captured by the chip detection camera 22, a synthesized image is displayed on the monitor 33 through the synthesis circuit 32 shown in FIG. 4 as shown in FIG. 4a. Next, the tip 2, that is, the bonding tool 23
by rotating and moving in the XY directions, or rotating and moving the board positioning table 20 in the XY directions.
The leads 1a of the substrate 1 and the bumps 2a of the chip 2 are aligned as shown in FIG. Various methods can be considered for this positioning method, which will be described later.

Next, the bonding tool 23 is moved the distance X ₁ and Y ₁ determined in the explanation of FIG. When pressed, the bumps 2a of the chip 2 are pressed onto the leads 1a of the board 1.
are superimposed and bonded accurately. Thereafter, the operations explained in FIG. 1 are performed manually or automatically, and bonding is performed automatically in sequence.

Next, two methods for aligning the substrate 1 and the chip 2 as shown in FIG. 4C from the state shown in FIG. 4A will be explained.

In Fig. 4a, x and y shown by solid lines are the reference cross lines, x' and y' shown by dashed dotted lines are the reference lines of board 1, and x'' and y'' shown by dotted lines are the reference lines of chip 2, respectively. show. Also, θ' is the slope of the reference lines x', y' from the reference lines x, y, that is, the substrate 1
The inclination angle, θ'' indicates the inclination of the reference lines x'', y'' from the reference lines x, y, that is, the inclination angle of the chip 2, respectively. Then, θ', θ'' are determined, and the chip 2 or the substrate 1 is When rotated by ＝θ″−θ′, the same figure b
As shown in the figure (the figure shows a state in which the substrate 1 is rotated), the substrate 1 and the chip 2 are parallel to each other. Next, X the chip 2
When moving △x in the direction and △y in the Y direction (parallel movement), point a of chip 2 aligns with point b of lead 1a,
As shown in figure c, the lead 1a and the bump 2a overlap. In reality, the θ direction and the XY direction are moved simultaneously, and the state shown in FIG. 4 a is changed to the state shown in FIG. 4 c at one time.

As _another method, as _shown in FIG.
b ₁ and b ₂ are determined in advance. First, the inclination angle α of the chip 2 is determined. Predetermined reference lines for a ₁ and a ₂
If the deviation in the x and y directions is △x ₁ and △y ₁ , then α=tan ^-1
△y ₁ / △x ₁ . Similarly, when determining the tilt angle β of the substrate 1, it becomes β=tan ⁻¹ Δy ₂ /Δx ₂ . Therefore, the deviation angle θ between the substrate 1 and the chip 2 is determined by θ=α−β. Here, θ>0 means clockwise rotation when viewed from the figure, and θ<0 means counterclockwise rotation when viewed from the figure. Therefore, when the substrate 1 is rotated by .theta., the substrate 1 and the chip 2 become parallel to each other as shown in FIG. Next, 1 point from the predetermined chip 2
The deviation △x, △y between a ₁ and one point b ₁ on the board 1 is detected, and the chip 2 is moved △x in the x-axis direction and △y in the y-axis direction.
When moved, the lead 1a and the bump 2a overlap as shown in FIG. In this way, if two arbitrary points are determined in advance on each of the substrate 1 and the chip 2, and it is determined which points correspond to which points, automatic detection and automatic bonding can be performed.

Note that the two cameras 2 explained in FIG.
The magnification and position adjustments of 1 and 22 only need to be performed once at the beginning, and are not required during subsequent position alignment and bonding. Further, when adjusting the magnification, as shown in FIG. 3, cross marks 40 are attached to the transparent plates 41 and 42, but the present invention is not limited to this.
Triangular, square, or round marks may be attached, or the mark 40 may not be attached, and the transparent plates 41 and 42 may have the same size and match in outer shape. In addition, in the embodiment, the position in the XY direction was calculated by moving the bonding tool 23 from the position of the substrate positioning table 20 to above the chip detection camera 22, but the position of the chip detection camera 22 was already calculated from the position of the substrate positioning table 20. It goes without saying that when the distance to the top is known in advance, there is no need to calculate this distance anew.

[Effects of the Invention] As is clear from the above explanation, according to the chip bonding alignment method of the present invention, two cameras installed correspondingly to the substrate and the chip can be used at separate locations. The two images are taken out, combined and displayed on a single monitor using a compositing circuit, and the board and chip are aligned at a separate location, so it can be applied regardless of whether the board is transparent or opaque, and alignment is easy. and be able to do it accurately. Furthermore, the camera can be installed at any position, and the angle of the camera can be easily set, making it difficult for positional deviations to occur as in the prior art.

[Brief explanation of the drawing]

FIG. 1 shows an embodiment of the present invention, in which a is a front explanatory view, b is a perspective explanatory view, FIG. 2 is a circuit diagram, and FIG. 3 is a camera magnification adjustment method, a is a front explanatory view, b is a perspective explanatory diagram; FIGS. 4a, b, and c are explanatory diagrams showing the position adjustment of the substrate 1 and the chip 2 projected on the monitor; and FIG. 5 is an explanatory diagram showing the conventional method in the case of a transparent substrate. 6 are explanatory diagrams showing a conventional method in the case of an opaque substrate. 1: Board, 2: Chip, 20: Board positioning stand, 21: Board detection camera, 22: Chip detection camera, 23: Bonding tool, 32: Synthesis circuit, 33: Monitor, 40: Mark, 41,4
2: Transparent plate.

Claims (1)

[Claims] 1. Place the substrate on a substrate positioning table, adsorb the chip with a bonding tool, and photograph the substrate on the substrate positioning table with a substrate detection camera disposed above the substrate positioning table. A chip detection camera placed at a location away from the substrate positioning table photographs the chip adsorbed to the bonding tool, and the two images taken by the two cameras are displayed on one monitor using a compositing circuit. A chip bonding alignment method characterized by combining and projecting and aligning a substrate and a chip. 2. Place one of two transparent plates with marks or of the same size on a substrate positioning table, adsorb the other on a bonding tool, and position the substrate using a substrate detection camera disposed above the substrate positioning table. Photographing the transparent plate on the table, and photographing the transparent plate adsorbed to the bonding tool with a chip detection camera disposed at a location away from the substrate positioning table;
The two images taken by the two cameras are combined and displayed on one monitor using a composition circuit, the magnification and position of the two cameras are adjusted, and then the board is placed on a board positioning table. , the chip is adsorbed to the bonding tool, the board and the chip are taken out as two images by the two cameras, the composite circuit is used to combine and display the images on the one monitor, and the board and chip are aligned. A chip bonding alignment method characterized by the following.