JP5813330B2 - Collet position detection method and apparatus - Google Patents

Description

  The present invention relates to a collet position detection method and apparatus for adsorbing and transferring a semiconductor chip.

  In the manufacturing process of a semiconductor device, when mounting a semiconductor chip (pellet) on a substrate such as a lead frame, a collet that sucks and transfers the semiconductor chip is used. The collet is usually supported by a bonding head provided with a drive mechanism, and transfers the semiconductor chip from a pickup stage in which the semiconductor chips are arranged in parallel to a bonding stage for bonding to the substrate.

  In order to make the semiconductor chip suction and mount position accurate by the collet, the collet is moved according to the set coordinates under image recognition by a CCD camera or the like, and the actual collet position is detected, and based on the detected coordinates. The method of correcting the set coordinates is employed.

  Conventionally, the collet position is detected by, for example, attaching a soft sheet such as a seal tape to one of the semiconductor chips on the pickup stage, moving the collet on the sheet according to the set coordinates, and then pressing the collet on the lower surface of the collet. This is done by forming indentations on the sheet and detecting the coordinates of the indentations on the image.

  However, in this method, the semiconductor chip is set at a predetermined set coordinate, a sheet for indentation formation is pasted, the collet is moved to a predetermined position on the sheet while recognizing the coordinates by the image, and then pressed against the sheet, It required a lot of labor to detect the coordinates of the indentation on the image. In addition, since the operator detects the coordinates while viewing the image, individual differences due to the proficiency level may affect the accuracy.

  Furthermore, in the position recognition on the image, the size of the pixel (pixel of the CCD camera) becomes the limit of resolution (for example, about 10 to 20 μm), and it is difficult to obtain higher accuracy. In particular, many large collets for large chips are made of rubber. In this case, a clear impression is not formed unless a large pressure is applied.

  On the other hand, for a small chip, a method using a pyramid collet has been proposed (Patent Document 1). This is because the needle base on which the needle is implanted is mounted on the stage, the collet is lowered from above the needle base, the needle is brought into contact with the conical recess of the collet, and the collet or stage is moved in the orthogonal X and Y directions. Then, the position of the collet center relative to the camera center is detected by recognizing the position at the time of contact with the concave portion on the image.

  However, in this method, it is necessary to add a needle base, and there is a drawback in that it takes time to carefully perform needle contact in each of the X and Y directions. In addition, as described above, in the position recognition on the image, the pixel size becomes the limit of resolution, and it is difficult to obtain higher accuracy.

Japanese Patent No. 3097314

  It is an object of the present invention to provide a collet position detection method and apparatus that can detect the position of the collet in a state in which these problems of the prior art are solved and the accuracy is not easily limited.

In order to achieve the above object, the present invention provides an optical sensor disposed in a collet movement path between a pickup stage and a bonding stage of a semiconductor chip in a reference coordinate system, and the upper side of the optical sensor is a collet or its mounting. This is a detection area for detecting the coordinates of the edge of the part, and while moving the collet in one direction along the horizontal plane according to the set coordinate value, the collet at the symmetrical position with respect to the center of the collet or the edge of its mounting part When the optical sensor passes through the detection area of the sensor , the optical sensor detects the coordinates of both edges, thereby detecting the coordinate value of the first central position, which is the center of both edges, and the one along the horizontal plane. While the collet is moved in the direction orthogonal to the direction according to the set coordinate value, the collet at the symmetrical position with respect to the collet center or the edge of the mounting portion thereof is When passing through the detection zone, by the light sensor detects the respective coordinates of the opposite edges, and detects the coordinate value of the second center position is the center of the opposite edges, said detected first A collet position detection method is provided, wherein a detected coordinate value of a collet center is obtained from a central position and a second central position, and a set coordinate value of the collet is corrected based on the detected coordinate value.

In order to achieve the above-mentioned object, the present invention also provides an optical sensor disposed in a collet movement path between a pick-up stage and a bonding stage of a semiconductor chip in one direction along the horizontal plane and the direction in the reference coordinate system. moves the collet according to the set coordinate value in a direction perpendicular, and a control unit for calculating a detection value obtained on the basis of this movement, above the light sensor, a collet or coordinates of the edge of the mounting portion When the collet moves in one direction, the collet or the edge of the collet mounting part passes through the detection area of the optical sensor. , the coordinate value signal of each of both edges and emitting said light sensor records the respective coordinate value as a first detection value, perpendicular to the one direction The movement of the collet to direction, the edge of the collet or the collet mounting portion located at symmetrical positions with respect to the collet center, when passing through the detection area of the light sensor, a coordinate value signal of each of both edge portions light sensor is issued, and records each of the coordinate values as the second detection value, wherein the first search of the detection value and the detection coordinate value of the collet from the center the second detection value, the detection collet set coordinates based on the coordinate values The value is corrected.

  In the collet position detection method according to the present invention, in the reference coordinate system, an optical sensor is arranged at a predetermined coordinate between a pickup stage and a bonding stage of a semiconductor chip, and the collet is moved by the optical sensor while moving the collet along a horizontal plane. Or the coordinate value of a collet mounting part is detected. The movement of the collet is performed in one direction along the horizontal plane according to the set coordinate value, and the optical sensor detects that the collet at the symmetric position with respect to the collet center or the edge of the collet mounting part passes through the detection area. The coordinate value of the first center position, which is the center of both edges, is detected. The movement of the collet is also performed in a direction perpendicular to the one direction along the horizontal plane according to the set coordinate value, and the optical sensor detects that the collet or the edge of the collet mounting portion that is symmetric with respect to the collet center passes through the detection area. By detecting at, the coordinate value of the second center position, which is the center of both edges, is detected. Then, a detected coordinate value of the collet center is obtained from the detected first center position and second center position, and the set coordinate value of the collet is corrected based on the detected coordinate value.

  Therefore, it is possible to accurately detect the collet position based on a simple operation of moving the collet according to the movement path between the pick-up stage and the bonding stage of the semiconductor chip.

The collet is usually provided with a suction hole having a circular or square cross section extending in the axial direction for adsorbing a semiconductor chip. Further, the outer peripheral edge portion of the collet in the horizontal direction is line symmetric with respect to each center line in a predetermined direction and a direction perpendicular thereto. The collet mounting part is a part where the collet is mounted on the collet arm, for example, like a collet holder, the center position of which coincides exactly with the coordinate position of the collet center, and the edge thereof is in a predetermined direction and perpendicular to this. It is line symmetric with respect to each center line in the direction to be. Therefore, by detecting each time these points pass through the detection area of the optical sensor, it is possible to detect the collet or the edge of the collet mounting portion that is symmetric with respect to the collet center. In the present invention, the edge is detected in one direction along the horizontal plane and in a direction perpendicular to the one direction, and the center of the edge of the collet or the collet mounting portion detected in each direction is set to the first central position and the first position. The detected coordinate value at the center of the collet is obtained from these two positions. The coordinates of the collet center are determined in the reference coordinate system as set values, but there may be a deviation due to an error in accuracy of the collet holder part, a tilt of the collet, or the like. Depending on the amount of deviation, the collet position accuracy at the time of pickup and bonding may not be sufficiently obtained.
On the other hand, as described above, the detection coordinate value of the collet center is obtained, and by correcting the set coordinate value of the collet based on the detection coordinate value, the coordinate of the collet can be made accurate. The deviation can be eliminated or reduced by matching the value.

  The control unit performs control for collet positioning. Since this control is performed with high accuracy (for example, 1 to several μm), if the detection accuracy of the optical sensor is increased, the detection value at the collet center is also at the same level. It can be obtained with high accuracy and is not easily limited by detection accuracy. In addition, since recognition does not require recognition work by an operator, high accuracy can be obtained without being influenced by individual differences.

  Moreover, in order to implement the said detection method, the collet position detection apparatus which concerns on this invention is provided with an optical sensor and a control part. When the photo sensor emits a coordinate value signal of the collet in the symmetric position with respect to the center of the collet or the edge of the collet mounting part by the movement of the collet in the one direction, the control unit calculates each coordinate value. The optical sensor emits a coordinate value signal at the edge of the collet or collet mounting portion that is symmetric with respect to the center of the collet by recording the first detection value and moving the collet in a direction orthogonal to the one direction. Sometimes, each coordinate value is recorded as a second detection value, the detection coordinate value of the collet center is obtained from the first detection value and the second detection value, and the set coordinate value of the collet is corrected based on the detection coordinate value To do. By this correction, the coordinates of the collet can be made accurate, and the set value and the detected value can be matched to eliminate or alleviate the deviation. Thereby, the effect similar to the said method of this invention can be acquired.

  As described above, according to the present invention, it is possible to provide a collet position detection method and apparatus that can detect the position of the collet easily and in a state in which it is difficult to be limited in accuracy.

It is a perspective view of the collet position detection apparatus concerning one embodiment of the present invention. It is a longitudinal cross-sectional view which shows the collet and optical sensor in the apparatus shown in FIG. It is explanatory drawing of the output of the optical sensor shown in FIG. It is explanatory drawing regarding the output of FIG. FIG. 10 is a perspective view of a collet position detecting device according to another embodiment of the present invention, which is a device for correcting eccentricity of collet rotation. It is a bottom view of the collet shown in FIG. 5, (a) before rotation, (b) has shown the state after rotation. It is explanatory drawing of eccentric correction of collet rotation by the apparatus of FIG.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. FIG. 1 is a perspective view of a collet position detection device according to an embodiment of the present invention.

  The collet position detection device 1 is provided in a bonding device 2. The bonding apparatus 2 includes a machine base 3, a pickup stage 4, a bonding stage 5, a transfer device 6, a bonding head 7, and the like. The pickup stage 4 and the bonding stage 5 are disposed on a substantially horizontal work surface of the machine base 3. The transport device 6 transports the substrate S along the rail 61 in the frontage direction (the left-right direction in the drawing) of the bonding device 2. The bonding head 7 extends in the front-rear direction of the machine base, and is driven at the rear end by a drive device (electric motor) 70 provided at the rear of the machine base so as to reciprocate in the front direction of the bonding apparatus. A bonding arm 8 extending in the vertical direction is supported on the front end portion of the bonding head 7.

  The bonding arm 8 includes a first intermediate member 81 supported up and down by the bonding head 7 and a second intermediate member supported up and down by the first intermediate member 81 for movement to a predetermined position. 82, a collet arm 83 extending downward from the second intermediate member 82, and a collet 80 supported by the lower end of the collet arm 83. For these driving operations, the bonding arm 8 is supported by the bonding head 7 and drives a first intermediate member 81 in the vertical direction (electric motor) 85, and is supported by the bonding head 7 and moves the second intermediate member 82 back and forth. And a driving device (electric motor) 86 for driving in the direction. A suction hole 801 extending in the axial direction is formed in the collet 80 and the collet arm 83, and a port 821 communicating with the suction hole 801 is provided in the second intermediate member 82, and a suction tube 822 is connected to the port. ing.

  A supply device and a discharge device for supplying and discharging the wafer ring holding the semiconductor chip to and from the pickup stage 4 are provided on the upstream side and the downstream side of the pickup stage 4. On the upstream side of the bonding stage 5, there are provided a loader for transferring the substrate S from the stocker to the transfer device 6, a preform part for applying an adhesive to the substrate, and the like on the downstream side of the bonding stage 5. An unloader for transferring the substrate S from the terminal end of the transfer device 6 to the outside of the machine base 3 is provided. Since these are used in a normal bonding apparatus, description and illustration are omitted here.

  The pick-up stage 4 includes a wafer table 41, holds the wafer ring supplied from the supply device, and enables the collet 80 to suck the semiconductor chips T one by one and peel them from the holding sheet. The wafer table 41 is provided with a drive mechanism for moving to the set position in the front and back direction of the machine base 3, that is, the X direction and the Y direction shown in FIG. A collet 80 that reciprocates between the semiconductor chips T allows the semiconductor chips T to be sequentially sucked and transferred. The movement of the collet 80 and the wafer table 41 in the X and Y directions is performed under the control of the control unit 31 provided at the lower part of the machine base 3, and the control part 31 uses the predetermined position of the machine base 3 as the origin. Position control is performed based on a reference coordinate system. In order to make the suction by the collet 80 smooth, the pickup stage 4 employs various known means such as a state in which the semiconductor chip T is pushed up by a push-up needle and partially peeled from the holding sheet. can do.

  The bonding stage 5 holds the holding plate H holding a large number of substrates S in a fixed state, and after the collet 80 transfers the semiconductor chips T to one row of substrates S, the holding plate H is sent to the next row. Is repeated, and the semiconductor chip T can be bonded to each substrate S on the holding plate H. This feeding operation in the bonding stage 5 is also performed by the control unit 31 based on the reference coordinate system.

  As shown in FIG. 2, the optical sensor 9 is provided on the upper wall 30 of the machine base 3. The optical sensor 9 is positioned between the pick-up stage 4 and the bonding stage 5 of the semiconductor chip, so that the moving distance of the collet 80 can be shortened. It is desirable to arrange the sensor 9. The optical sensor 9 includes, for example, a light emitting element and a light receiving element, and outputs whether or not reflected light of the light from the light emitting element reaches the light receiving element depending on the presence or absence of an object located in the detection area. It is possible to detect the presence or absence of an object by varying the height of the object. In this embodiment, the collet 80 includes an upper cylindrical portion 802 and a lower frustoconical portion 803, and has a circular cross section. The suction hole 801 is also formed in a circular cross section. Therefore, the output of the optical sensor 9 when the collet 80 moves in the X direction and the Y direction with respect to the optical sensor 9 is when the wall portion of the collet 80 approaches the optical sensor 9 as shown in FIG. The suction hole 801 is low when the suction hole 801 is positioned above the optical sensor 9, is high while the wall portion of the collet 80 is positioned above the optical sensor 9 again, and is low when the collet 80 passes through the optical sensor 9. In this embodiment, the collet 80 includes a truncated cone-shaped portion 803 in the lower portion, so that the reflected light at this portion reaches the optical sensor 9 as part of the irregularly reflected light. As a result, the input to the optical sensor 9 by the reflected light at the truncated cone portion 803 is smaller than that by the reflected light at the lower end surface of the collet, but a value clearly higher than the input from outside the suction hole and the collet is obtained. It is done.

  The control unit 31 records each coordinate value as the first detection value every time the edge of the suction hole 801 of the collet 80 passes over the optical sensor 9 by the movement of the collet 80 in the X direction. Similarly, each time the edge of the suction hole 801 of the collet 80 passes over the optical sensor 9 due to the movement of the collet 80 in the Y direction, the control unit 31 records each coordinate value as a second detection value.

  In order to obtain the coordinate value of the collet center based on the coordinate values thus recorded, the control unit 31 is provided with a calculation unit 310. The suction hole 801 is symmetrical with respect to a straight line passing through the center of the collet because the lower edge is circular. Based on this, the calculation unit 310 obtains the coordinate value of the central first central position from the two first detection values, obtains the coordinate value of the central second central position from the two second detection values, The coordinate value of the collet center is obtained from the one central position and the second central position, and this is set as the detected coordinate value of the collet center. Then, a predetermined set coordinate value of the collet center is corrected based on the detected coordinate value. This correction can be performed, for example, by moving the origin of the reference coordinates by the difference between the set coordinate value and the detected coordinate value. Along with this, the coordinate value of the pick-up position of the semiconductor chip by the collet is also corrected. The optical sensor 9 can be positioned using a jig or the like so as to be an accurate set position with respect to the reference coordinates. In this case, the origin position of the reference coordinates is also corrected by correcting the coordinate value of the collet center. With these corrections, the subsequent control position of the collet 80 by the control unit 31 can be optimized in a short time.

  When the collet end face detected by the optical sensor 9 has a shape that is axisymmetric with respect to the X axis and the Y axis like a circular suction hole 801, an accurate correction value even if the detection position is slightly deviated from the center of the collet. Can be obtained. That is, as shown in FIG. 4, the first detection value P1 detected when the detection position of the optical sensor deviates from the straight line X0, Y0 passing through the collet center O in the X-axis and Y-axis directions. , P2, the interval d is smaller than the diameter D of the suction hole 801, but the center in the Y-axis direction obtained by d / 2 is obtained as Y0, and the above deviation does not affect the detected value. This is the same when the detection position of the optical sensor is shifted in the Y-axis direction. Thereby, optimization of the control position can be reliably obtained.

  In particular, the control unit 31 uses a control mechanism for positioning the collet, and this control mechanism performs position control with a high level of accuracy of, for example, 1 to several μm. Therefore, by increasing the detection accuracy of the optical sensor to a level comparable to this, the detected coordinate value and the correction value can be obtained at the same high level as the control mechanism, and high-precision control optimization is possible. . Therefore, the accuracy of the collet position detection is not easily limited. In addition, since recognition does not require recognition work by an operator, high accuracy can be obtained without being influenced by individual differences.

  FIG. 5 shows an example in which the eccentric correction of the collet rotation is performed in the embodiment of the apparatus for carrying out the method of the present invention. In this example, a driving device 87 (electric motor) for rotating the collet is provided on the upper portion of the bonding head 7, and pulleys 88a and 88b mounted on the driving device 87 and the driving device 85 for vertical driving, The belt 88c wound around these constitutes the interlocking device 88. The pulley 88a is coupled to the collet arm 83, and the rotation of the driving device 87 rotates the collet 80a about the axis of the collet arm 83 (collet rotation axis) via the interlocking device 88 and the collet arm 83. The collet 80a is composed of an upper part 804 and a lower part 805 narrower than the upper part, and is formed in a rectangular cross section. Since other configurations are the same as those of the previous embodiment, description thereof is omitted.

  This apparatus is used as follows. First, in the same manner as described in the previous embodiment, the set coordinate value of the collet center is corrected based on the detected coordinate value. That is, the control unit 31 records each coordinate value as the first detection value every time the edge of the suction hole 801 of the collet 80a passes over the optical sensor by the movement of the collet 80a in the X direction, and the Y direction Each time the edge of the suction hole 801 passes over the optical sensor by the movement to, each coordinate value is recorded as the second detection value.

  Then, the calculation unit 310 obtains the coordinate value of the central first central position from the two first detection values, obtains the coordinate value of the central second central position from the two second detection values, The coordinate value of the collet center is obtained from the position and the second center position, and this is set as the first detected coordinate value (X1, Y1) of the collet center. Then, the predetermined set coordinate value of the collet center is corrected based on the first detected coordinate value.

  FIG. 6 shows the lower surface of the collet 80a, where (a) shows the state before rotation and (b) shows the state after rotation. Here, before the rotation, the collet center Oc and the rotation center R of the collet rotation axis are shifted by Xc and Yc in the X-axis and Y-axis directions, respectively. After the rotation, the collet center Oc ′ and the rotation center R are Assume a state where Xc ′ and Yc ′ are shifted. This difference in deviation is due to the eccentricity of the collet center Oc and the collet rotation center R.

  Furthermore, as shown in FIG. 7, the coordinate values (XOc, YOc) of the collet center with respect to the origin Os of the set coordinate system are detected as the first detected coordinate values (X1, Y1), and the correct origin Os ′ is the origin Os ′. Is assumed to be shifted by (XOc-X1, YOc-Y1). In this case, it is only necessary to correct the set coordinate value of the origin Os by (XOc-X1, YOc-Y1) based on the first detected coordinate value (X1, Y1) against the deviation.

  Next, the collet 80a is rotated 90 degrees around the collet rotation axis by the driving device 87 and the interlocking device 88. After this rotation, the third central position that is the center of the edge that is symmetrical with respect to the collet center in the same manner as the first central position in the previous embodiment, targeting the edge of the suction hole 801 of the collet 80a. The coordinate value of is detected. Similarly to the second center position, the coordinate value of the fourth center position, which is the center of the edge located at the symmetrical position with respect to the collet center, is detected. Then, the coordinate value of the collet center Oc ′ after rotation is obtained from the detected third center position and fourth center position, and is set as the second detected coordinate value (X2, Y2) (the reference of the coordinates is the corrected origin Os. 'Is). Then, the eccentric amount of the collet rotation is obtained based on the difference between the second detected coordinate value (X2, Y2) and the first detected coordinate value (X1, Y1). This is performed, for example, as follows.

  FIG. 7 shows a state in which the collet 80a is rotated by 90 degrees around the rotation center R by a one-dot chain line circle. In the calculation of the following eccentricity correction, the distances in the X-axis and Y-axis directions from the initial origin Os to the rotation center R are set as Xr and Yr, respectively. As described with reference to FIG. 6, the distances in the X-axis and Y-axis directions from the rotation center R to the collet center Oc ′ after rotation are Xc ′ and Yc ′, respectively.

In this state, the following relationship holds (see FIG. 7):
Xr, Yr, Xc ′, Yc ′ are unknown values, but since the rotation of the collet 80a is 90 degrees, two hatched squares have the same shape,
Xc '= Yc = Y1-Yr
Yc ′ = Xc = Xr−X1
-Correction amount from origin Os to Os' of set coordinate system:
XOc-X1, YOc-Y1 (1)
The amount of eccentricity of the rotation center R with respect to the collet center Oc before rotation:
Xc, Yc = (Xr−X1), (Y1−Yr) (2)
-Second detected coordinate value of collet center Oc 'after rotation: X2, Y2
That is the relationship.

Here, X2 and Y2 are based on the relationship shown in FIG.
X2 = Xr + Xc ′ = Xr + (Y1-Yr) (3)
Y2 = Yr + Yc ′ = Yr + (Xr−X1) (4)
From Equation (3) + Equation (4)
X2 + Y2 = 2Xr-X1 + Y1
Therefore, Xr = 1/2 (X1 + X2-Y1 + Y2) (5)
Equation (3)-Equation (4)
X2-Y2 = -2Yr + X1 + Y1
Therefore, Yr = 1/2 (X1-X2 + Y1 + Y2) (6)
Substituting the above formulas (5) and (6) into the formula (2) indicating the eccentric amounts Xc and Yc,
Xc = Xr-X1 = 1/2 (X1 + X2-Y1 + Y2) -X1
= 1/2 {(X2-X1) + (Y2-Y1)}
Yc = Y1-Yr = Y1-1 / 2 (X1-X2 + Y1 + Y2)
= 1/2 {(X2-X1)-(Y2-Y1)}
It becomes.

Therefore, the eccentric amount (X, Y coordinates) of the rotation center R with respect to the collet center Oc before the rotation is
Xc, Yc = 1/2 {(X2-X1) + (Y2-Y1)}, 1/2 {(X2-X1)-(Y2-Y1)}
Thus, the amount of eccentricity is obtained based on the first detected coordinate value (X1, Y1) and the second detected coordinate value (X2, Y2). Accordingly, it is only necessary to correct the eccentricity of the collet or the center of the collet mounting portion or the collet rotation axis in accordance with the amount of eccentricity.

In any of the above-described embodiments, when no problem in detection accuracy occurs, an optical sensor that outputs image data such as a CCD camera can be used to obtain detection coordinate values from an image. . Further, the detection by the optical sensor can be performed on the transmitted light instead of the reflected light. In this case, the light emitting element and the light receiving element are arranged to face the collet (or the collet mounting portion) on one side and the other side, and depending on whether or not the light emitted from the light emitting element is blocked, It is possible to detect passage of a collet (or collet mounting portion) in the detection area.

  The detection of the center of the collet and the correction of the set coordinate value based on the detection are performed immediately after the collet is mounted or replaced, so that the subsequent control position can be optimized. Further, by performing the detection and correction at the time of exchanging the wafer ring or replenishing the substrate, it is possible to eliminate or minimize the increase in the processing time of the bonding apparatus.

  In addition, the collet position detected by the optical sensor to detect the collet center may be a position that is line symmetric with respect to a straight line passing through the collet center in one direction along the horizontal plane and a direction perpendicular thereto. In addition to a circle, it can be a polygonal suction hole such as a square, and the outer peripheral edge of the collet or the collet mounting part is symmetrical with respect to a straight line passing through the center of the collet, such as a polygon such as a circle or a square. In this case, this can be set as a detection target. The collet mounting portion is a portion where the collet is mounted on the collet arm and the center position exactly matches the coordinate position of the collet center, and can be a collet holder in the collet arm, for example. In this case, an edge such as an outer peripheral portion of the collet holder is a detection target.

  When the collet suction hole is to be detected, detection is performed when the collet is not adsorbing the semiconductor chip. As described above, this is preferably performed immediately after wafer ring replacement, and can also be performed when the collet repeats the transfer of the semiconductor chip, when the collet moves backward, that is, when the collet moves from the bonding stage to the pickup stage. Alternatively, the detection with the suction hole as a detection target can be performed when the suction hole of the collet is opened and the suction flow rate through the suction hole is measured. Further, when the outer peripheral edge of the collet or the edge of the collet mounting portion is set as a detection target, if the semiconductor chip is smaller than these portions, the center of the collet can be detected even when the semiconductor chip is adsorbed. In any case, the detection accuracy can be further increased by detecting the collet center a plurality of times and obtaining the average value.

  The detection by the optical sensor is preferably performed in a state where the collet is close to the optical sensor. By doing so, high detection accuracy can be ensured even when there is an inevitable inclination that may occur slightly in the detection target such as a collet.

  Collets are sometimes used at high temperatures, such as when bonding is performed with solder. In this case, deviation from the set coordinate value is likely to occur due to thermal expansion. However, according to the detection and correction of the collet center, the detected coordinate value is obtained by avoiding or suppressing the influence of heat, and the set coordinate value is corrected. Can be performed.

  The present invention is not limited to the above embodiment, and various modifications can be made. For example, the arrangement of the optical sensor can shorten the movement distance and the detection time by setting the collet in the movement path between the pickup stage and the bonding stage in a normal processing operation. If there is a margin time, such as when the collet is out of processing operation, the detection time can be avoided or alleviated even if it is some distance from the moving route. .

1: Collet position detection device 2: Bonding device 4: Pickup stage 5: Bonding stage 6: Transport device 9: Optical sensor 31: Control unit 80: Collet 801: Suction hole O: Collet center S: Substrate P1, P2: First Detected value T: Semiconductor chip

Claims (5)

In the reference coordinate system, a photo sensor is arranged in a collet movement path between the pick-up stage and the bonding stage of the semiconductor chip, and a detection area in which the photo sensor is located above the collet or the edge of the mounting part. And
While moving the collet according to one direction setting coordinate values along the horizontal plane, when the edge of the collet or a mounting portion are symmetrical position, to pass through the detection zone of the optical sensor relative to the collet center, the optical sensor Detects the coordinate value of the first central position, which is the center of both edges, by detecting the coordinates of both edges,
While moving the collet according to the direction set coordinate value orthogonal to said one direction along the horizontal plane, the edge of the collet or a mounting portion located at symmetrical positions with respect to the collet center, when passing through the detection area of the optical sensor In addition, the optical sensor detects the respective coordinates of both edges, thereby detecting the coordinate value of the second central position that is the center of both edges,
A collet position detection method comprising: obtaining a detected coordinate value of a collet center from the detected first center position and second center position, and correcting a set coordinate value of the collet based on the detected coordinate value.   The collet position detection method according to claim 1, wherein an edge of a suction hole of the collet is detected by the optical sensor.   2. The collet position detection method according to claim 1, wherein an outer peripheral edge of the collet or an edge of the collet mounting part is detected by the optical sensor. After obtaining the detection coordinate value of the collet center from the first central position and the second central position, the collet is rotated 90 degrees around the collet rotation axis, and the rotated collet or the edge of the mounting portion is targeted, Similarly to the detection of the first center position, the coordinate value of the third center position, which is the center of the edge located symmetrically with respect to the collet center, is detected, and in the same manner as the second center position, The coordinate value of the fourth central position, which is the center of the edge at the symmetrical position, is detected,
A collet center coordinate value is obtained from the detected third center position and fourth center position, and the collet is determined based on a difference between the coordinate value and the detected coordinate value based on the first center position and the second center position. 4. The collet position detection method according to claim 1, wherein an eccentric amount of rotation is obtained, and eccentric correction of the collet, the collet mounting portion or the collet rotation shaft is performed in accordance with the eccentric amount. In a reference coordinate system, an optical sensor disposed in a collet movement path between a semiconductor chip pickup stage and a bonding stage;
The collet is moved according to the set coordinate value in one direction along the horizontal plane and in a direction perpendicular to the direction, and a control unit that calculates a detection value obtained based on the movement is provided.
Above the optical sensor is a detection area for detecting the coordinates of the edge of the collet or its mounting part,
When the edge of the collet or the collet mounting part that is in a symmetrical position with respect to the center of the collet passes through the detection area of the optical sensor by the movement of the collet in the one direction, respectively coordinate value signal and the optical sensor can occur, and records the respective coordinate value as a first detection value, by the movement of the collet in the direction orthogonal to the one direction, are symmetrical positions with respect to the collet center edge of the collet or the collet mounting portion recorded, when passing through the detection area of the light sensor, a coordinate value signal of each of both edges and emitting said optical sensor, each of the coordinate values as a second detection value A collet position detecting device characterized in that a detected coordinate value of the collet center is obtained from the first detected value and the second detected value, and the set coordinate value of the collet is corrected based on the detected coordinate value.

Images