JP2005159110A - Component-mounting method and device thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a component-mounting method and a device capable of improving mounting accuracy, by obtaining an optimum offset value which corresponds to the size of a board. <P>SOLUTION: Reference marks 201, arranged at each prescribed interval on a glass substrate 200, are recognized, and the offset values for respective areas, corresponding to a substrate size, are decided as numerical values for correction from the recognized result. By respectively reflecting the pertinent offset values of the respective moving positions of a component-mounting head 136 as the numerical values for the correction at the the mounting position correction, at the mark recognition correction or at the mounting of position offset value measurement, high accuracy mounting is carried out. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a component mounting method and apparatus for mounting components on a substrate with high accuracy.

  The component suction head is moved in the XY direction by driving the XY robot, and component mounting such as component suction by the head nozzle, recognition of the suction component by the camera, and mounting on the board is performed. However, high mounting accuracy could not be achieved due to distortion of the component mounting apparatus itself. This distortion of the component mounting apparatus itself is caused by poor machining accuracy of the XY robot of the component mounting apparatus or poor assembly accuracy.

  By analyzing more specifically that it is impossible to mount components with high accuracy when mounting on a substrate due to distortion of the XY robot due to such processing accuracy, yawing of the guide member of the XY robot (on the XY robot) Position in the XY direction by rolling (rolling in a direction orthogonal to the moving direction of the moving head), pitching (poor linearity in the moving path of the head), rolling (pitching in a direction different from the above roll by 90 degrees), etc. Misalignment will occur.

  Therefore, conventionally, camera calibration is performed, and the reference mark on the reference board is viewed with the board recognition camera fixed to the XY robot, and the positional deviation between the target position where the reference mark should originally exist and the actual position of the reference mark is changed. The amount is calculated, and the calculated positional deviation amount is added to each position as a mounting position offset value and correction is performed, so that component mounting can be performed with high accuracy (see, for example, Patent Document 1).

  Here, the camera calibration in the board recognition camera is a position coordinate calculated based on the recognition result by causing the board recognition camera to recognize a jig whose position coordinates are known in advance in order to detect mounting errors of the board recognition camera. In other words, the mounting error of the board recognition camera is calculated from the difference from the position coordinates known in advance, and the position correction is performed. At the time of the camera calibration, not only the position correction of the board recognition camera but also the position correction of the component recognition camera and the nozzle are performed together.

JP-A-6-126671

  However, in the method of correcting each of the above positions, for example, there is a possibility that the position of the reference substrate may be displaced by about 1 mm between the first positioning of the reference substrate and the second positioning of the reference substrate. Is very expensive because it requires very high accuracy. From the standpoint of preventing breakage, it is necessary to stop and position the reference substrate at a general position in the X direction without using a substrate stopper. Since there is a gap of less than 1 mm in the Y direction for conveyance, there is no reproducibility of positioning in the substrate holding portion of the reference substrate in the component mounting apparatus, which causes a reduction in mounting accuracy.

  After positioning the reference board at the approximate position in this way, the relative displacement between each position of the robot is obtained by recognizing the reference mark of the reference board, and when mounting the displacement, the mounting board Since this is reflected in the position data to be mounted, the mounting accuracy is a factor.

  On the other hand, in the method of recognizing and correcting a glass reference substrate provided with a grid in a matrix, the reference substrate grid is measured on the assumption that the reference substrate is accurately positioned, and the measured data is obtained. It is conceivable to use the correction value as it is.

  However, as described above, it is very difficult to accurately hold the reference substrate on the substrate holding unit in units of microns, and a special positioning device is required to accurately hold the reference substrate on the substrate holding unit of the component mounting apparatus. After all, if the measured data is directly used as a correction value, accurate correction of the XY robot cannot be performed unless the reference substrate is accurately positioned with good reproducibility.

  By the way, when considering the entire component mounting area of the component mounting apparatus, the distortion of the head operation due to the distortion of the XY robot changes depending on the position of positioning, so that conventional camera calibration and mounting Since only the position offset value is insufficiently corrected, there is a problem that the mounting accuracy cannot be secured.

  This is because even if the reference substrate itself on which a large number of reference marks are arranged in a lattice at equal intervals is precisely manufactured, the XY robot and the reference substrate cannot be absolutely parallel, and the XY robot As a result, the absolute perpendicularity is not guaranteed, so that there is no reference, and the head having a board recognition camera that recognizes the reference board arranged in the component mounting area of the component mounting apparatus is supported. Since the robot is distorted, the position obtained from the reference board cannot be used as a reference, and the mounting accuracy is increased (for example, the robot accuracy is increased to about ± 2 μm, or the total accuracy as a mounting machine is ± 20 μm). Could not be raised to the extent).

  Accordingly, an object of the present invention is to provide a component mounting method and apparatus capable of improving the mounting accuracy by obtaining an optimal offset value corresponding to the size of a board, in order to solve the above-described problem. It is in.

  In order to achieve the above object, the present invention is configured as follows.

According to the first aspect of the present invention, component mounting for mounting a component held by a component holding head movable with respect to the substrate holding device at a component mounting position of the component mounting substrate held by the substrate holding device. In the method
Recognizes the position coordinates of the reference marks arranged at predetermined intervals of the reference substrate held by the substrate holding device in a state where the reference substrate for reference mark recognition is held by the substrate holding device and positioned in the component mounting area. Then, the position coordinates of each recognized reference mark are obtained,
The difference between the NC coordinates of the respective reference marks and the position coordinates is obtained as a correction value, respectively.
Obtain the NC coordinates of the position coordinates of at least two board reference position calculation marks on the component mounting board,
From the recognized reference marks, reference marks that are close to the two substrate reference position calculation marks are respectively extracted.
While converting the position coordinates of the extracted reference marks so that the correction values of the extracted reference marks become zero or substantially zero, the offset values at the respective reference marks are obtained.
In place of the reference mark recognizing reference substrate, the component mounting substrate is held by the substrate holding device and positioned in the component mounting region, and at least two of the component mounting substrates held by the substrate holding device are used. Recognizing each of the two substrate reference position calculation marks, and obtaining the position coordinates of the two recognized substrate reference position calculation marks,
Based on the obtained position coordinates of the two substrate reference position calculation marks, the NC coordinates of the two substrate reference position calculation marks are respectively corrected,
When the component held by the component holding head is positioned above each component mounting position of the component mounting board, the offset value of the reference mark closest to the recognition device provided in the component holding head is used. In addition, after correcting the position coordinates of the component mounting position, the component is mounted at the component mounting position based on the corrected position coordinates of the component mounting position. Provide an implementation method.

According to the second aspect of the present invention, the extracted reference marks are adjusted such that the correction values of the extracted reference marks close to the two substrate reference position calculation marks are zero or substantially zero. When converting the position coordinates, and obtaining the offset value at each reference mark,
Coordinate conversion is performed by rotating and moving the graph connecting the extracted reference marks so that the correction value of the extracted reference marks close to the two substrate reference position calculation marks is zero or substantially zero. Accordingly, the component mounting method according to the first aspect is provided, in which the position coordinates of the extracted reference marks are respectively transformed to obtain the offset values at the respective reference marks.

According to the third aspect of the present invention, the extracted reference marks are adjusted such that the correction values of the extracted reference marks close to the two substrate reference position calculation marks are zero or substantially zero. When converting the position coordinates, and obtaining the offset value at each reference mark,
From the extracted reference mark, a correction value in at least one of the X direction of the substrate holding device and the Y direction orthogonal to the X direction is calculated, the inclination of the reference substrate is obtained, and the correction value is calculated. According to the first or second aspect, the position coordinates of the extracted reference marks are respectively transformed so that the offset value at each reference mark is obtained so that is zero or substantially zero. Provide a component mounting method.

According to the fourth aspect of the present invention, the component mounting for mounting the component held by the component holding head movable with respect to the substrate holding device at the component mounting position of the component mounting substrate held by the substrate holding device. In the device
At each predetermined interval of the reference substrate held by the substrate holding device, provided in the component holding head, with the reference mark recognition reference substrate held by the substrate holding device and positioned in the component mounting area. A recognition device for recognizing the position coordinates of the arranged reference marks;
The position mark of the reference mark is obtained from the recognition result of the reference mark recognized by the recognition device, and the difference between the NC coordinate of the reference mark and the position coordinate is obtained as a correction value. Based on the NC coordinates of the position coordinates of at least two substrate reference position calculation marks, the reference marks close to the two substrate reference position calculation marks are extracted from the recognized reference marks, respectively. The coordinate values of the extracted reference marks are converted so that the correction value of the extracted reference mark becomes zero or substantially zero, and the offset value at each reference mark is obtained. In a state where the component mounting board is held in the board holding device instead of the recognition reference board and is positioned in the component mounting area, Recognizing each of the at least two board reference position calculation marks of the component mounting board held by the board holding device, and obtaining the position coordinates of the two recognized board reference position calculation marks, respectively. A calculation unit for correcting the NC coordinates of the two substrate reference position calculation marks based on the position coordinates of the two substrate reference position calculation marks,
When the component held by the component holding head is positioned above each component mounting position of the component mounting board, the offset value of the reference mark closest to the recognition device provided in the component holding head is calculated. And a controller for mounting the component to the component mounting position based on the corrected position coordinate of the component mounting position after correcting the position coordinate of the component mounting position. A component mounting apparatus is provided.

  According to the fifth aspect of the present invention, the calculation unit is configured to extract the correction values of the extracted reference marks that are close to the two substrate reference position calculation marks, respectively, so that the correction value is zero or substantially zero. When the coordinate values of the obtained reference marks are converted and the offset values at the respective reference marks are obtained, the correction values of the extracted reference marks that are close to the two substrate reference position calculation marks are zero or By rotating and moving the graph connecting the extracted reference marks so as to be substantially zero, the coordinates of the positions of the extracted reference marks are converted, and the respective reference marks are converted. A component mounting apparatus according to a fourth aspect is provided for obtaining an offset value at.

  According to the sixth aspect of the present invention, the calculation unit is configured so that the correction value of the extracted reference mark close to the two substrate reference position calculation marks is zero or substantially zero. When the coordinate position of each of the reference marks thus obtained is converted and the offset value at each reference mark is obtained, the X direction of the substrate holding device and the Y direction perpendicular to the X direction are obtained from the extracted reference marks. And calculating the correction value in at least one direction, obtaining the inclination of the reference substrate, and coordinate the position coordinates of the extracted reference marks so that the correction value is zero or substantially zero. The component mounting apparatus according to the fourth or fifth aspect is provided in which the offset value at each reference mark is obtained by conversion.

  According to the seventh aspect of the present invention, the two Y-axis robots installed parallel to each other along the Y-axis direction, and the two Y-axis movably along the X-axis direction orthogonal to the Y-axis direction. An XY robot arranged on the robot and having one X-axis robot movably supporting the component holding head along the X-axis direction, and the component in the XY-axis direction with respect to the substrate holding device A component mounting apparatus according to any one of the fourth to sixth aspects, in which a holding head is movable between the two Y-axis robots and one X-axis robot.

  According to an eighth aspect of the present invention, the component holding head includes a plurality of component suction nozzles that can hold each component by suction and are arranged along the X-axis direction, and the plurality of components. The component according to the seventh aspect, wherein the substrate recognition camera is arranged on the component holding head so that the imaging center of the substrate recognition camera is positioned coaxially with a straight line passing through the center of the suction nozzle. A mounting apparatus is provided.

  According to the present invention, the reference marks are arranged at predetermined intervals of the reference substrate held by the substrate holding device in a state where the reference substrate for reference mark recognition is held by the substrate holding device and positioned in the component mounting area. Recognizing the position coordinates of the marks, obtaining the position coordinates of the respective recognized reference marks, obtaining the difference between the NC coordinates of the respective reference marks and the position coordinates as correction values, and the component mounting board; NC coordinates of the position coordinates of at least two substrate reference position calculation marks are respectively obtained, and reference marks close to the two substrate reference position calculation marks are respectively extracted from the recognized reference marks, The position coordinates of the extracted reference marks are respectively coordinate-transformed so that the correction values of the extracted reference marks are zero or substantially zero. , And to obtain the offset value for each of the reference marks. Thereafter, in place of the reference mark recognizing reference substrate, the component mounting substrate held by the substrate holding device and positioned in the component mounting region, the component mounting substrate held by the substrate holding device Recognize at least two substrate reference position calculation marks, respectively, determine the position coordinates of the two recognized substrate reference position calculation marks, and use the obtained position coordinates of the two substrate reference position calculation marks as the position coordinates. The NC coordinates of the two board reference position calculation marks are respectively corrected based on the above, and when the component held by the component holding head is positioned above each component mounting position of the component mounting board, After correcting the position coordinates of the component mounting position based on the offset value of the reference mark closest to the recognition device provided in the component holding head, And to perform the attachment to the component mounting position of the component based on the corrected position coordinates of the component placing position. As a result, the reference marks arranged at predetermined intervals on the reference substrate for reference mark recognition are recognized, and the numerical value for correcting the position coordinates for each area according to the substrate size is determined as an offset value from the recognition result. In the mounting position correction, the mark recognition correction, and the mounting position offset value measuring operation, or any of those operations, the corresponding offset value of each moving position of the component mounting head is used. As a result, the cause of deviation due to distortion of the XY robot operation is absorbed, and an optimum offset value corresponding to the size of the board is obtained, so that high-precision mounting (for example, positioning at a level of ± 0.005 mm during mounting) Mounting under precision conditions).

  Also, when the reference mark is recognized, the offset value corresponding to each movement position of the component mounting head is reflected as a numerical value for correction, thereby absorbing the shift factor due to distortion of the XY robot operation and the size of the board. By obtaining an optimum offset value corresponding to the position, it is possible to perform mounting with higher accuracy.

  Embodiments according to the present invention will be described below in detail with reference to the drawings.

  As shown in FIGS. 1 to 4, a component mounting apparatus 100 capable of performing a component mounting method according to an embodiment of the present invention includes a gantry 110, an XY robot 120, and a board recognition camera 140 as basic components. And a component recognition camera 150 and a control device 170, and further a component supply device 180 and a substrate transfer device 190.

  The pedestal 110 is a platform for installing the XY robot 120, the component recognition camera 150, the control device 170, the component supply device 180, and the substrate transfer device 190, and includes a rectangular parallelepiped base 111 and The base portion 111 and the Y-axis robot leg portion 112, that is, the gantry 110, are integrally formed by casting. The Y-axis robot leg portion 112 protrudes from the base portion 111 at both ends of the base portion 111 in the X-axis direction 51 and extends along the Y-axis direction 52 orthogonal to the X-axis direction 51. . Each Y-axis robot leg 112 is provided with a linear guide 123 or the like in a Y-axis robot 121, which will be described in detail later, constituting the XY robot 120. Each linear guide 123 as a guide support member of the nut portion 126 of FIG. 4 is for a Y-axis robot along a linear guide installation surface 123a formed along the Y-axis direction 52 on each Y-axis robot leg portion 112. As described above, each Y-axis robot leg 112 is constructed integrally with the base 111 by casting.

  The XY robot 120 includes two Y-axis robots 121 installed in parallel to each other along the Y-axis direction 52 on each of the Y-axis robot legs 112, that is, a gantry 110 formed in an integral structure by casting. And one X-axis robot 131 disposed along the X-axis direction 51 perpendicular to the Y-axis direction 52 on the two Y-axis robots 121.

  Each Y-axis robot 121 includes a Y-axis ball screw structure 122 and the linear guide 123. The Y-axis ball screw structure 122 has one end 122 a as a fixed end and the other end 122 b as a support end, and linearly expands and contracts only in the Y-axis direction 52 by heat, and moves the X-axis robot 131 in the Y-axis direction 52. More specifically, as shown in FIGS. 1 and 4, the one end 122 a of the Y-axis ball screw structure 122 is fixed to the Y-axis robot leg 112, and a motor 124 is provided as a drive source of the ball screw 125. Are coupled to the ball screw 125. The other end 122b is attached to the Y-axis robot leg 112 while supporting the ball screw 125 so as to be rotatable in the circumferential direction and extendable in the axial direction, that is, the Y-axis direction 52.

  When the Y-axis robot 121 configured as described above is continuously operated, the portions that generate heat are the ball screw 125 and the motor 124, and the other end 122b allows expansion and contraction of the ball screw 125 in the Y-axis direction 52 due to heat. To do. Further, since the motor 124 is fixed to the gantry 110 having an integral structure as described above, the expansion / contraction of each Y-axis robot 121 due to heat, that is, thermal expansion / contraction, can be linear only in the Y-axis direction 52. . Since the operations of the two Y-axis robots 121 are the same, the amount of thermal expansion and contraction in the Y-axis direction 52 in each Y-axis robot 121 is equal.

  Further, as shown in FIG. 4, a nut portion 126 is attached to the ball screw 125 of each Y-axis robot 121, and the nut portion 126 moves in the Y-axis direction 52 by the rotation of each ball screw 125. An X-axis robot 131 constituting the XY robot 120 is installed between the nut portions 126 along the X-axis direction 51. As described above, the amount of expansion / contraction in each Y-axis robot 121 in the Y-axis direction 52 is the same, so the X-axis robot 131 installed between the nuts 126 moves in the Y-axis direction 52 in a state parallel to the X-axis. Can move.

  FIG. 4 is a diagram conceptually showing the structure of the gantry 110 and the XY robot 120, and a component mounting head described later is not shown. 2 to 4, the illustration of the component supply device 180 is omitted.

  The X axis robot 131 includes an X axis frame 132 and an X axis ball screw structure 133. The X-axis frame 132 is fixed at both ends to the nut portion 126 of the ball screw structure 122 in each Y-axis robot 121 as described above, and extends in the X-axis direction 51. The X-axis ball screw structure 133 is formed on the X-axis frame 132 and linearly expands and contracts only in the X-axis direction 51 by heat with one end 133a as a fixed end and the other end 133b as a support end. The component mounting head 136 is attached, and the component mounting head 136 is moved in the X-axis direction 51.

  The X-axis frame 132 is a member made of substantially prismatic aluminum, and both ends thereof are fixed to the nut portion 126 as described above. The one end 133a of the X-axis ball screw structure 133 formed on the side surface of the X-axis frame 132 is provided with a motor 135 that is fixed to the X-axis frame 132 and serves as a drive source for the ball screw 134, as shown in FIG. And coupled to the ball screw 134. The other end 133b is attached to the X-axis frame 132 while supporting the ball screw 134 so as to be rotatable in the circumferential direction and extendable in the axial direction, that is, the X-axis direction 51. When the X-axis robot 131 is continuously operated, the portions that generate heat are the ball screw 134 and the motor 135, and the other end 133b allows the ball screw 134 to expand and contract in the X-axis direction 51 due to heat.

  Further, as shown in FIG. 1, a nut portion 134a for attaching the component mounting head 136 is attached to the ball screw 134, and the nut portion 134a, that is, the component mounting head 136 is rotated by the rotation of the ball screw 134. Move in the axial direction 51.

  The component mounting head 136 confirms the positional deviation between the component suction nozzle 1361 serving as an example that functions as a component holding member that holds the electronic component 62 and the circuit board 61 that is loaded and installed in this embodiment. Therefore, the circuit board 61 is used to image the board reference position calculation marks 202-1 and 202-2, and to image the reference marks 201 arranged at predetermined intervals on the reference mark recognition reference board 200 described later. And a substrate recognition camera 140. As shown in detail in FIG. 5, the component suction nozzle 1361 is provided with eight component suction nozzles 1361 on a straight line along the X-axis direction 51 in the present embodiment. The electronic component 62 is a small component such as a chip component or a large component such as QFP. Therefore, the component suction nozzle 1361 is also installed with an optimal size and shape corresponding to various components to be sucked. As described above, the board recognition camera 140 is arranged so that the imaging center of the board recognition camera 140 is positioned coaxially with a straight line passing through the center of each component suction nozzle 1361 arranged along the X-axis direction 51. Yes. The component mounting head 136 is also provided with a rotation motor 1363 for rotating each component suction nozzle 1361 in the direction around its axis.

  Each component suction nozzle 1361 is configured to suck the electronic component 62 from the component supply device 180 and mount the sucked electronic component 62 on the circuit board 61 as an example of a component mounting board. It is necessary to move along the direction, that is, the Z-axis direction 53. In the present embodiment, the component mounting head 136 is used for moving a component suction nozzle 1361 as an example of a component holding member, and each component suction nozzle 1361 is an example that functions as a drive source for moving the component holding member. A motor 1362 is provided. Therefore, compared to the case where all of the plurality of component suction nozzles are conventionally driven by one large output motor, a low output motor can be used, and the amount of heat generated from the motor can be suppressed. As an example, the output of the moving motor 1362 is 20 W, and there is almost no heat generation from the moving motor 1362. Further, conventionally, when one large output motor having a large heat generation amount is provided, a temperature gradient is generated according to the distance from the large output motor in the conventional component mounting head, and the distance between the component suction nozzles in the arrangement direction. Differ due to differences in thermal expansion and contraction. On the other hand, in this embodiment, since each component suction nozzle 1361 is provided with a moving motor 1362, there is almost no heat from each moving motor 1362, and even if there is heat, the component mounting head. In 136, a temperature gradient that affects the component mounting accuracy does not occur. Therefore, even if the component mounting head 136 is continuously operated, the distance between the component suction nozzles 1361 in the X-axis direction 51 can be kept equal or almost equal. Note that the substantially equal state means that the component mounting accuracy is not affected.

  In addition, since there is no temperature gradient that affects the component mounting accuracy in the component mounting head 136 as described above, the relative position between each component suction nozzle 1361 and the board recognition camera 140, that is, each component suction nozzle 1361 The distance from the substrate recognition camera 140 can be fixed. Here, the immobility means that the distance between each component suction nozzle 1361 and the board recognition camera 140 does not expand or contract to the extent that the component mounting accuracy is affected by heat.

  In the component mounting apparatus 100 of the present embodiment, the component supply apparatus 180 is a so-called cassette type component supply apparatus having a plurality of reels wound with tapes that house electronic components 62, and the front side of the component mounting apparatus 100. Two sets are provided for each of 100a and rear side 100b.

  The board transfer device 190 is a device that carries in, holds and holds the circuit board 61 with respect to the mounting position of the circuit board 61 in the component mounting area of the component mounting apparatus 100, as shown in FIG. In addition, the component mounting apparatus 100 is disposed along the X-axis direction 51 at a substantially central portion. The substrate transfer device 190 has a transfer table 165 as an example of a substrate holding device at the mounting position, and can suck and hold the circuit board 61 that has been carried in. Can be carried out.

  As shown in FIG. 6, the control device 170 is connected to the XY robot 120, the substrate recognition camera 140, the component recognition camera 150, the component supply device 180, and the substrate transfer device 190, which are the components described above. Operation control is performed, and the mounting operation of the electronic component 62 on the circuit board 61 is controlled. The control device 170 has a program and mounting data necessary for the mounting operation and the like (for example, each moving position coordinate data of the component mounting head 136 during the mounting operation, mounting position coordinate data of each component, and component mounting) Data such as relation information between each movement position of the head 136 and each component mounting position, reference mark recognition reference board size and reference mark position coordinate data, board size to be mounted and board reference position calculation Mounting position information such as mark position coordinate data, respective component data, nozzle size data, component supply data of the component supply device 180, etc., recognition information by the board recognition camera 140, and a calculation unit 171 described later. For example, a board recognition camera 140 that performs various calculations. Recognition information (for example, recognition information of the reference marks 201A and 201B by the substrate recognition camera 140, recognition information of the reference mark 201 by the substrate recognition camera 140 and substrate reference position calculation marks 202-1 and 202-2 by the substrate recognition camera 140). Based on the recognition information and the data of each mounting position among the mounting information stored in the storage unit 173, and the like. And an operation unit 171 for calculating and calculating the error. The control device 170 causes a component mounting operation to be performed based on data and information stored in the storage unit 173. The component mounting operation, particularly the correction operation, of the control device 170 configured as described above will be described in detail below.

  The operation of the component mounting apparatus 100 configured as described above, that is, the component mounting method executed by the component mounting apparatus 100 will be described in more detail. In addition, about the operation | movement from the component adsorption | suction by the XY robot 120 including the component mounting head 136 to the component mounting to the circuit board 61 by the XY robot 120 including the component mounting head 136 by the circuit board conveying apparatus 190. Since these operations are basically similar to the operations performed in the conventional component mounting apparatus, these operations will be briefly described below.

  That is, the component mounting head 136 is moved to the component supply device 180 by the XY robot 120. Next, one or more electronic components 62 are sucked and held by one or more nozzles 1361 of the component mounting head 136 from the component supply device 180. Next, after the component mounting head 136 passes above the component recognition camera 150 by the XY robot 120 and the orientation of the electronic component 62 sucked and held by the nozzle 1361 is recognized by the component recognition camera 150, the mounting of the circuit board 61 is performed. Head to the position. After the electronic component 62 sucked and held by one nozzle 1361 of the component mounting head 136 is positioned above the corresponding mounting position by the XY robot 120, the nozzle 1361 is lowered to bring the electronic component 62 to the mounting position. Mounting. At this time, the nozzle 1361 is rotated around its axis based on the component orientation recognition result of the component recognition camera 150, and the position of the component mounting head 136 is corrected in consideration of an offset value described later, and then the mounting operation described above. To perform the mounting operation. The series of mounting operations is performed for all the components 62 to be mounted on the substrate 61.

  The component mounting method according to the present embodiment is characterized in the position correction operation of the component mounting head 136 during the mounting operation in consideration of the offset value, and will be described in detail below with reference to FIG.

  That is, the component mounting method according to the present embodiment recognizes the reference marks 201 arranged at predetermined intervals on the glass substrate 200 as an example of the reference mark recognition reference board, and recognizes each of the recognized reference marks. Position coordinates (coordinates composed of the X coordinate value in the X direction in the plane of the glass substrate 200 for indicating the position of the reference mark and the Y coordinate value in the Y direction orthogonal to the X direction) are obtained, The difference between the NC coordinates of the mark (designed numerical position coordinates of a predetermined reference mark) and the position coordinates is obtained as a correction value, and at least two board reference position calculation marks on the component mounting board are obtained. NC coordinates of the respective position coordinates are obtained, and reference marks close to the two substrate reference position calculation marks are extracted from the recognized reference marks. Then, the position coordinates of the extracted reference marks are respectively transformed to obtain offset values at the respective reference marks so that the correction values of the extracted reference marks become zero or substantially zero. . Then, in place of the reference mark recognizing reference substrate, the component mounting substrate held by the substrate holding device and positioned in the component mounting region, the component mounting substrate held by the substrate holding device Recognize at least two substrate reference position calculation marks, respectively, determine the position coordinates of the two recognized substrate reference position calculation marks, and use the obtained position coordinates of the two substrate reference position calculation marks as the position coordinates. Based on the NC coordinates of the two substrate reference position calculation marks, respectively, at the time of mounting position correction, at the time of mark recognition correction, and at the time of mounting position offset measurement operation, or any of those operations, The reference mark closest to the recognition device provided in the component holding head when positioned at each movement position of the component mounting head 136 Based on the offset value, by correcting the position coordinates of the mobile position, in which highly accurate mounting is to allow.

  Here, as described later, the offset value is zero or substantially zero as the correction value of the reference mark extracted as a reference mark close to each of the two board reference position calculation marks of the component mounting board. As described above, it means a numerical value for correcting the position coordinates of the reference mark obtained by coordinate conversion of the extracted position coordinates of the reference mark.

  The correction value means a difference between each NC coordinate of the reference marks arranged at predetermined intervals of the reference substrate and each recognized position coordinate.

  First, an outline of how to obtain the offset value will be described.

  The positioning accuracy of the component mounting head 136 is greatly affected by the distortion of the XY robot 120 (see FIGS. 7 and 8), and a positioning error occurs. For example, FIG. 7 is a diagram showing the relationship between the distortion of the X-axis robot and the component mounting head 136, and FIG. 8 is a diagram showing the relationship between the strain of the Y-axis robot and the component mounting head 136. This positioning error varies depending on the position where the component mounting head 136 moves, and affects the mounting accuracy. Therefore, as shown in FIG. 9, the NC value is used as a correction numerical value for removing errors such as positioning of the XY robot 120 that occurs when the XY robot 120 moves the head 136 to an arbitrary NC coordinate position. The offset value of the reference mark position closest to the coordinate position (in other words, the correction offset value of the area where the NC coordinate position exists) is used. That is, an offset value used as a correction numerical value for correcting an error such as positioning is set to a maximum component mounting area (a board to be produced, for example, a board of XL size: 510 mm × 460 mm, M size: 330 mm × It is determined using a reference substrate for recognizing a reference mark within a region including a 250 mm substrate.

  Specifically, first, in step S1 of FIG. 11, a glass substrate 200 as an example of a reference substrate for recognizing reference marks is held on a transfer table 165 as an example of a substrate holding device and positioned in a component mounting area.

  Next, in step S <b> 2 of FIG. 11, the position recognition coordinates of all the reference marks 201 arranged at predetermined intervals of the glass substrate 200 held on the transfer table 165 are recognized by the substrate recognition camera 140 of the component mounting head 136. To do. Here, more specific recognition of the reference mark for measuring the correction value is performed as follows. In this correction value measurement, as an example of the reference mark recognition reference substrate which is the measurement substrate, a reference mark (a circle having a diameter of 1 mm) is placed on a glass substrate 200 of XL size: 510 mm × 460 mm (M size: 330 mm × 250 mm). ) 201 is a dedicated glass substrate (hereinafter referred to as a glass substrate) formed by printing or the like in a grid shape (lattice shape). That is, as an example of the glass substrate 200, as shown in FIG. 10, for a XL size, a circular reference mark (Y direction: 44 rows and X direction: 49 columns) on a 510 mm × 460 mm glass plate with a pitch of 10 mm. The one printed with 201 (diameter 1 mm) is used. Therefore, the number of reference marks used in the measurement is 2156 points. For M size, a circular fiducial mark (diameter 1 mm) 201 with a pitch of 10 mm is used on a glass plate of 410 mm × 240 mm, and the fiducial mark 201 in the Y direction: 22 rows and the X direction: 39 columns is used for measurement. Therefore, the number of reference marks used in the measurement is 858 points.

  In principle, the size of the reference board for reference mark recognition may be any size as long as it is equal to or larger than the maximum component mounting area of the component mounting apparatus, but is smaller than the maximum component mounting area as will be described later. In some cases, the size may be larger than the virtually maximum component mounting area using a synthesis method. If the interval between the reference marks is made fine, the accuracy increases, but the data acquisition time becomes longer and the data storage amount increases. Therefore, about ¼ to ５ of the ball screw lead of the ball screw structure of the XY robot is economically sufficient. As a specific example, the reference mark pitch can be 10 mm with respect to the lead of 40 mm.

  Next, in step S <b> 3 of FIG. 11, the position coordinate of each recognized reference mark 201 is obtained by the calculation unit 171 based on the recognition result and stored in the storage unit 173. That is, for example, as shown in FIG. 13, all the reference marks 201 are arranged in parallel with the substrate transfer direction of the substrate transfer device 190 from the leftmost reference mark 201 in the same row in order to reduce positional deviation. The substrate recognition camera 140 of the head 136 is moved to the reference mark 201, and all the reference marks 201 in the row are sequentially recognized, and the position coordinates are obtained by the calculation unit 171 based on the recognition result and stored in the storage unit 173. . Next, after moving backward diagonally to the left, the substrate recognition camera 140 of the head 136 is moved from the reference mark 201 at the left end of the row immediately above the bottom row to the reference mark 201 at the right end of the same row, and that row All the fiducial marks 201 are recognized in order, and the position coordinates are obtained by the calculation unit 171 based on the recognition result and stored in the storage unit 173. Next, after moving diagonally to the left, the substrate recognition camera 140 of the head 136 is moved from the reference mark 201 at the left end of the two rows above the bottom row to the reference mark 201 at the right end of the same row. All the fiducial marks 201 are recognized in order, and the position coordinates are obtained by the calculation unit 171 based on the recognition result and stored in the storage unit 173. According to such an order, all the reference marks 201 in all the rows are recognized, and the position coordinates are obtained by the calculation unit 171 based on the recognition result and stored in the storage unit 173. The lower side of the glass substrate 200 in FIG. 13 corresponds to the front side of the component mounting apparatus, that is, the front side of the operator.

  In order to improve the recognition accuracy of each reference mark 201, the recognition process of each reference mark 201 may be repeated a plurality of times. In this case, the average value of the position coordinates obtained from the recognition results for the number of times is calculated by the calculation unit 171 and stored in the storage unit 173 as the position coordinates of each reference mark 201. It is preferable that the number of times can be arbitrarily changed from the operation screen of the component mounting apparatus.

  In this way, the position coordinates of all the reference marks 201 are stored in the storage unit 173.

  Next, in step S4 in FIG. 11, the difference between the NC coordinate of each reference mark 201 and the position coordinate is obtained as a correction value by the calculation unit 171 and stored in the storage unit 173. This correction value is a numerical value for correcting the holding deviation, recognition deviation, positioning error of the XY robot, and the like when the glass substrate 200 is sucked and held by the transfer table 165.

  Next, in step S5 of FIG. 11, the NC coordinates of the position coordinates of at least two board reference position calculation marks 202-1 and 202-2 of the component mounting board 61 are acquired by the calculation unit 171.

  Next, in step S6 of FIG. 11, from among the recognized reference marks 201 of the glass substrate 200 based on the NC coordinates of the position coordinates of the two substrate reference position calculation marks 202-1 and 202-2. The calculation unit 171 extracts the reference marks 201 close to the two board reference position calculation marks 202-1 and 202-2 of the component mounting board 61, respectively. Specifically, in FIG. 12, two reference marks 201A, for example, on the upper right and lower left diagonals on the glass substrate 200, which are close to the two substrate reference position calculation marks 202-1 and 202-2, respectively. The substrate recognition camera 140 recognizes 201B while moving the head 136 with the XY robot 120. That is, it is difficult for the glass substrate 200 to be held on the transfer table 165 so as to be completely parallel to the substrate transfer direction of the substrate transfer device 190, and a positional deviation occurs. In order to correct the positional deviation at the time of holding the glass substrate, first, the reference marks 201 at the lower left corner and the upper right corner of the glass substrate 200 are recognized as the reference marks 201A and 201B.

  Next, in step S7 in FIG. 11, the position coordinates of the extracted reference marks 201A and 201B are respectively coordinate-transformed so that the correction values of the extracted reference marks 201A and 201B become zero or substantially zero. (Coordinate conversion is performed in consideration of parallel shift, inclination, and expansion / contraction rate), and offset values at the respective reference marks 201A and 201B are obtained. That is, the parallel displacement and inclination of the glass substrate 200 are obtained by the calculation unit 171 from the position coordinates of the recognition results of the two reference marks 201A and 201B obtained in step S3 of FIG. Formulas for calculating parallel shift and inclination will be described later. The parallel shift means a positional shift in the X direction and / or the Y direction. The tilt means a rotation shift caused by rotating the substrate with respect to the X direction and the Y direction which is an orthogonal direction when the substrate is stopped at the mounting position of the transfer table 165 by the substrate stopper. At this time, in the case of normal substrate reference position calculation mark correction, since it is necessary to consider expansion / contraction of the substrate due to heat, the expansion / contraction ratio is also obtained, but it is not necessary to consider expansion / contraction of the substrate due to heat In the case of the idea of using as a reference, the expansion / contraction rate of the glass substrate 200 is 1. Here, the expansion / contraction rate means the ratio of expansion / contraction due to heat of the substrate itself.

  Next, based on the correction values (parallel shift and inclination) obtained by the calculation unit 171, the correction values of the two reference marks 201A and 201B become zero (in other words, the two reference marks 201A). , 201B to rotate and move the graph connecting the two reference marks 201A and 201B so as to be substantially zero (so as to be within a range of ± 5 μm, for example) Then, the coordinates are converted, and the offset values at the position coordinates of all the reference marks 201 are obtained and stored in the storage unit 173. As a result, each area corresponding to the size of the reference substrate for recognizing the reference mark {a rectangular area obtained by dividing the reference substrate into unit areas based on the reference mark (for example, surrounded by four reference marks)} The offset value for each area can be determined using the offset value for each area as a numerical value for correcting the movement position of the component mounting head in each area. It is possible to improve the mounting accuracy by correcting the position by using it at the time of mounting and mounting the component to the board to be mounted.

  An error such as a unique positioning of the XY robot 120 can be grasped as a relative displacement between the mounting positions based on the offset values obtained in steps S1 to S7 in FIG. In addition, the offset value obtained in this way is used for correction in the reference mark recognition operation, the component mounting operation, and the mounting offset value measuring operation or when calculating the head positioning position in each of these operations. By using it as a numerical value for correcting the position coordinates, it is possible to absorb a shift factor due to distortion of the XY robot operation and improve the mounting accuracy.

  Here, the reason why the correction based on the deviation of the reference board for reference mark recognition is added to the position coordinates of all the reference marks 201 is that the positioning error of the XY robot 120 is included in the reference mark recognition when measuring the correction value. This is because it has been lost. In the first place, the positioning operations of all the XY robots 120 include errors, and even if the glass substrate 200 can be manufactured with a desired high accuracy, it cannot be accurately positioned at the mounting position of the component mounting apparatus. Since there is no reference, it is impossible to accurately measure the positioning error of the XY robot 120.

Here, FIG. 14 showing that the reference marks 201A and 201B are recognized at positions shifted from the visual field center positions O ₁ and O ₂ of the substrate recognition camera 140 is shown in FIG. 14 as the reference marks 201A and 201B at the time of reference mark recognition. When the recognition result, the position coordinates deviation obtained from the recognition result of the first point of the reference mark _{201A (ΔX 1, ΔY 1)} , the recognition result from the obtained position coordinates displacement of the second point of the reference mark 201B ([Delta] X ₂ , ΔY ₂ ) is obtained as the positional coordinate deviation obtained from the reference mark recognition result.

Ideally, the deviation component included in the positional coordinate deviation obtained from each recognition result is ideally only the parallel deviation amount when the glass substrate 200 is held on the transport table 165, but in practice, The error of recognition processing and the positioning error of the XY robot 120 are included. Therefore, the positional coordinate deviation obtained from the recognition results of the reference marks 201A and 201B is
(Position coordinate deviation of recognition result) = (Substrate holding deviation) + (Recognition deviation) + (XY robot positioning error)
The amounts of parallel displacement of the reference marks 201A and 201B are (X _pcb1 , Y _pcb1 ) and (X _pcb2 , Y _pcb2 ), and the recognition errors of the reference marks 201A and 201B are (X _rec1 , Y _rec1 ) and (X _rec2 ). , Y _rec2 ), and the positioning error amount of the XY robot 120 at the reference marks 201A and 201B is (X _e1 , Y _e1 ) and (X _e2 , Y _e2 ), the position coordinate deviation (ΔX ₁ , ΔY ₁ ), (ΔX ₂ , ΔY ₂ )
[Equation 1]
ΔX ₁ = X _pcb1 + X _rec1 + X _e1
ΔY ₁ = Y _pcb1 + Y _rec1 + Y _e1
ΔX ₂ = X _pcb2 + X _rec2 + X _e2
ΔY ₂ = Y _pcb2 + Y _rec2 + Y _e2
It becomes.

  That is, using the recognition result, the position coordinates of the reference mark obtained by correcting the positional coordinate deviation of the glass substrate 200 with respect to the position coordinates of each reference mark 201 are not actually the coordinates where the reference mark 201 exists. . This is because the corrected reference mark position coordinates include a deviation due to the positioning error of the XY robot 120.

If the recognition errors (X _rec1 , Y _rec1 ) and (X _rec2 , Y _rec2 ) of the reference marks 201A and 201B are set to zero, the position coordinates (X _m , Y _m ) of the reference mark obtained by correction are If the NC coordinate of the reference mark 201 is (X _mnc , Y _mnc ), and the NC coordinate of each reference mark 201A, 201B is (X _nc1 , Y _nc1 ), (X _nc2 , Y _nc2 ),
[Equation 2]
X _m = (X _mnc −X _nc1 ) cos Δθ− (Y _mnc −Y _nc1 ) sin Δθ + ΔX ₁
= (X _mnc −X _nc1 ) cos Δθ− (Y _mnc −Y _nc1 ) sin Δθ + X _pcb1 + X _e1 ... [1]
[Equation 3]
Y _m = (X _mnc −X _nc1 ) sin Δθ + (Y _mnc −Y _nc1 ) cos θ + ΔY ₁
= (X _mnc −X _nc1 ) sin Δθ + (Y _mnc −Y _nc1 ) cos θ + Y _pcb1 + Y _e1 ... [2]
It becomes.

On the other hand, if the position coordinates where the actual reference mark 201 exists are (X _t , Y _t ),
[Equation 4]
X _t = (X _mnc −X _nc1 ) cos Δθ− (Y _mnc −Y _nc1 ) sin Δθ + X _pcb1 ... [1] ′
Y _t = (X _mnc −X _nc1 ) sin Δθ + (Y _mnc −Y _nc1 ) cos θ + Y _pcb1 ... [2] ′
It becomes.

Here, originally, the corrected NC coordinates must coincide with the actual reference mark position coordinates ([1] = [1] ′, [2] = [2] ′). However, comparing the above equations,
[Equation 5]
X _m −X _t = X _e1 ≠ 0
Y _m −Y _t = Y _e1 ≠ 0
Thus, the corrected NC coordinates do not match the actual reference mark position coordinates. The fact that the head 136 cannot be positioned at the actual position coordinates of the reference mark means that the position coordinate deviation obtained from the recognition result obtained there results in a correction value including a positioning error. I can not use it.

  As described above, the XY robot operation of the component mounting apparatus always includes a positioning error, and even if the correction value is measured with reference to the glass substrate 200, it does not become a true value and is absolute. There is no standard.

  Therefore, in order to make this error as zero as possible (in other words, to make the position coordinate data of the reference mark 201 coincide with the NC coordinate data), the following processing is performed on the correction value obtained above. .

  In the actual component mounting operation in the component mounting apparatus, the component mounting apparatus recognizes all the reference marks as described above in order to correct the holding displacement of the production board (board to be mounted) on the transfer table 165. As a result, each mounting position is corrected. The result of recognizing the two substrate reference position calculation marks 202-1 and 202-2 at this time is as shown in FIG. Here, in addition to the amount of holding displacement, the positional deviation of the position obtained from the recognition result of the two substrate reference position calculation marks 202-1 and 202-2 includes two substrate reference position calculation marks 202-1 and 202. Positioning error at position -2 is included.

  When the component 62 is actually mounted on the mounting position 205 of the substrate 61 to be mounted, the parallel displacement, inclination, and expansion / contraction rate are obtained from the substrate reference position calculation mark recognition result, and each mounting position 205 is corrected. And use it. Specifically, the deviation amount (holding deviation + positioning error) at the position of the reference mark close to the two substrate reference position calculation marks 202-1 and 202-2 is zero (in other words, the two This is done by rearranging all mounting positions 205 (so that the position coordinate data of the substrate reference position calculation marks 202-1 and 202-2 matches the NC coordinate data).

  Specifically, as shown in FIG. 16, the position of the reference mark which is the original data of the correction value is from the original position (the center position of the rectangular field of view in FIG. 17) as shown in FIG. Since it is misaligned in the X and Y directions, it is not zero. In FIG. 16, the vertical axis indicates the amount of positional deviation, the horizontal axis indicates the position in the X direction, the upper graph indicates ΔX, that is, the positional deviation in the X direction, and the lower graph indicates ΔY, that is, the positional deviation in the Y direction. Show.

  Accordingly, as shown in FIGS. 18 and 19, correction values of the reference marks 201a and 201b in the vicinity of the two substrate reference position calculation marks 202-1 and 202-2 of the relatively small substrate 61S to be mounted are shown. The graph connecting the two reference marks 201a and 201b is rotated and moved to change the coordinates so that it becomes zero or substantially zero (for example, within a range of ± 5 μm). The position is rearranged. In the graph of FIG. 18, the reference marks 202-1 and 202-1 (on the diagonal line) are plotted on the same graph, but the data itself has a constant Y coordinate and an X coordinate at 10 mm intervals. It is measured. Accordingly, in the data displayed as “202-2” on the graph, the Y coordinate data of the reference mark 202-1 is the same, and the reference mark 202-2 and the X coordinate data are the same reference mark data. Yes. The same applies to FIG.

  As shown in FIGS. 20 and 21, the correction value of the reference mark 201 near the two substrate reference position calculation marks 202-1 and 202-2 of the relatively large substrate 61L to be mounted is as follows. All the mounting positions are rearranged by rotating and moving the graph so as to be zero or substantially zero (for example, within a range of ± 5 μm) and converting the coordinates. Thus, the actual use data of the correction value varies greatly depending on the board to be mounted.

  Since there is no absolute reference in the process of obtaining the XY robot positioning error, the measured XY robot positioning error amount of each area matches the board 61 to be mounted at the time of production. Only the positions of the substrate reference position calculation marks 202-1 and 202-2 are shown. Therefore, using the correction values of the reference marks close to the positions of the two substrate reference position calculation marks 202-1 and 202-2 on the production substrate 61, the correction values at the two points are zero or substantially zero. And rearrange the coordinates (for example, within a range of ± 5 μm). As the processing at this time, as in the correction processing of the two substrate reference position calculation marks 202-1 and 202-2, parallel shift, inclination, and expansion / contraction rate are obtained, and all mounting positions 205 are re-established based on the results. Deploy.

  In FIG. 22, all the reference mark positions are calculated based on the XY robot positioning error amounts of the reference marks 201a and 201b on the glass substrate 200 closest to the substrate reference position calculation marks 201-1 and 202-2 of the production substrate 61. The XY robot positioning error amount is subjected to coordinate conversion (coordinate conversion in consideration of parallel displacement, inclination, and expansion / contraction ratio) by the calculation unit 171 and stored in the storage unit 173.

  The above coordinate conversion is performed at the time of board type selection, and the offset value obtained by the conversion is set to each moving position as a numerical value for correction in each of the mark recognition operation, the component mounting operation, and the mounting offset measurement operation. Each is added by the control device 170. By using the offset value in this way, an error unique to the robot can be grasped as a relative displacement between the positions.

  Next, the following steps, that is, steps S8 to S12 in FIG. 11, are steps for correcting the position, inclination, and shrinkage of the component mounting board 61 when mounting. That is, when mounting, the following steps are performed to correct the position, inclination, and contraction of the component mounting board 61.

  Specifically, in step S8 in FIG. 11, the component mounting board 61 is held on the transfer table 165 and positioned in the component mounting area.

  Next, in step S9 of FIG. 11, the at least two board reference position calculation marks 202-1 and 202-2 of the component mounting board 61 held on the transfer table 165 are recognized and recognized. The position coordinates of the two substrate reference position calculation marks 202-1 and 202-2 are obtained.

  Next, in step S10 of FIG. 11, based on the obtained position coordinates of the two substrate reference position calculation marks 202-1 and 202-2, the two substrate reference position calculation marks 202-1 and 202-2. Each of the NC coordinates is corrected. That is, based on the difference between the position coordinates of the two substrate reference position calculation marks 202-1 and 202-2 and the NC coordinates of the two substrate reference position calculation marks 202-1 and 202-2, the 2 The NC coordinates of the two substrate reference position calculation marks 202-1 and 202-2 are corrected to the position coordinates of the two substrate reference position calculation marks 202-1 and 202-2.

  Next, in step S11 in FIG. 11, when the component 62 held by the component holding head 136 is positioned above each component mounting position 205 of the component mounting board 61, the component holding head 136 is provided. The component mounting based on the offset value of the reference mark 201 closest to the board recognition camera 140 as an example of the recognition apparatus (in other words, the offset value of the area including the reference mark 201 closest to the board recognition camera 140). The position 205 is corrected. Specifically, a reference nozzle among a plurality of nozzles 1361 of the head 136 (for example, at the left end of FIG. 5) is set to the NC coordinate of each reference mark 201 on the glass substrate 200 as an example of a reference substrate for recognizing reference marks. Nozzle) 1361 is positioned, and the substrate recognition camera 140 fixed to the head 136 reads the offset value of the reference mark 201 closest to the camera 140 from the storage unit 173, and based on this, the component mounting position 205 Make corrections.

  Next, in step S12 in FIG. 11, the component 62 is mounted on the corrected component mounting position 205.

  In the above description, the offset value is used in step S11. However, in step S9, the substrate recognition camera is moved by adding the offset value to the NC coordinate data of the substrate reference position calculation mark, so The position may be obtained.

  The above is the outline of the measurement of the correction value for obtaining the offset value of the area and the mounting position correction operation based on the measurement result.

  Hereinafter, a more specific example of the component mounting method according to the present embodiment will be described with reference to FIGS.

  (1) First, for example, a reference mark recognition operation is performed before shipping a component mounting apparatus from a component mounting apparatus manufacturing factory to a user. Note that the following reference mark recognition operation is performed in the same manner even when an overhaul is made after handing over to the user.

  That is, as shown in FIG. 24, as step S13A in FIG. 24, the operation of the component mounting apparatus is performed so as to select the reference board recognition reference board type program for measuring the correction value for obtaining the offset value of each area. Prompt the operator on the screen. In this reference mark recognition reference substrate type program, the type and size of the glass substrate 200 as an example of the reference mark recognition reference substrate and NC coordinate data of each position of the reference mark 201 on the glass substrate 200 are stored. By selecting a substrate type, the glass substrate 200 is specified, and NC coordinate data of each position of the reference mark 201 on the glass substrate 200 is sent from the storage unit 173 to the control device 170.

  As one more specific example, on a 410 mm × 240 mm glass substrate, when 858 reference marks of 22 rows × 39 rows are arranged at intervals of 10 mm vertically and horizontally, the coordinates of the first reference mark are ( 10, 10), the coordinates of the second fiducial mark are (20, 10),. . . . . The coordinates of the 880th reference mark are (390, 220). As another specific example, on a 510 mm × 460 mm glass substrate, when 2156 reference marks of 44 rows × 49 columns are arranged at intervals of 10 mm vertically and horizontally, the coordinates of the first reference mark are (10, 10), the coordinates of the second fiducial mark are (20, 10),. . . . . The coordinates of the 2156th reference mark are (490, 440). These are examples of the NC coordinate data.

  Next, during or after the NC coordinate data is sent from the storage unit 173 to the control device 170, as shown in FIG. 24, the reference marks 201 are arranged at regular intervals in a grid pattern as shown in FIG. The placed glass substrate 200 is positioned in the component mounting area by the transfer table 165 of the substrate transfer device 190 (see step S1 in FIG. 11).

  Next, after the glass substrate 200 is positioned in the component mounting area, the XY robot 120 is driven based on the NC coordinate data of each position of the reference mark 201 sent from the storage unit 173 as step S13C in FIG. The head 136 is moved to move the substrate recognition camera 140 to each position of the reference mark 201 to recognize all the reference marks 201 on the glass substrate 200 (see step S2 in FIG. 11). The position coordinate deviation (ΔX, ΔY) obtained from each recognition result or the position coordinates (X + ΔX, Y + ΔY) including the deviation is stored in the storage unit 173 (see step S3 in FIG. 11). At this time, the position coordinates of each reference mark 201 may be recognized more than once to obtain the coordinates of the position of each reference mark 201 with higher accuracy.

  The position of each reference mark 201 is stored and managed in the storage unit 173 as the movement position of each component mounting head 136. Accordingly, a reference mark recognition operation, component mounting operation, and mounting offset value measurement operation (particularly, mounting offset value measurement operation when mounting a chip component or a QFP component) in component mounting production, or a component mounting head of any of those operations Based on the positioning position 136, the control device 170 determines which area's offset value is reflected. Specifically, for example, an area surrounded by four reference marks 201 is assigned as one area, and the above four points are set as area offset values for the mounting position of the component 62 mounted in the area. The offset value of the position of any one of the reference marks 201 is adopted, and this offset value is added to the position coordinates of the mounting position as an area offset value in the area to perform correction.

  In the 410 mm × 240 mm glass substrate of the specific example, the positional coordinate deviation (−0.132, −0.051) obtained from the recognition result of the first reference mark or the positional coordinates including the deviation ( 10-0.132, 10-0.051) is stored in the storage unit 173. Further, the position coordinate deviation (−0.132, −0.051) obtained from the recognition result of the second reference mark or the position coordinates (20−0.132, 10−0.051) including the deviation is stored. Stored in the unit 173. Further, the positional coordinate deviation (−0.139, −0.050) obtained from the recognition result of the third reference mark or the positional coordinates (20−0.139, 20−0.050) including the deviation is stored. Stored in the unit 173. Further, a positional coordinate deviation (−0.139, −0.049) obtained from the recognition result of the fourth reference mark or a positional coordinate (10−0.139, 20−0.050) including the deviation is stored. Stored in the unit 173. As the area offset value, a positional coordinate shift (−0.132, −0.051) of the first reference mark is adopted. As another example, the position coordinate deviation (-0.132, -0.051) obtained from the recognition result of the fifty-first reference mark or the position coordinates including the deviation (210-0.132, 93-0). .051) is stored in the storage unit 173. Further, the position coordinate deviation (−0.130, −0.067) obtained from the recognition result of the 52nd fiducial mark or the position coordinates (220−0.130, 93-0.067) including the deviation is stored. Stored in the unit 173. Further, the positional coordinate deviation (−0.139, −0.050) obtained from the recognition result of the 53rd fiducial mark or the positional coordinates (220−0.139, 103−0.050) including the deviation is stored. Stored in the unit 173. Further, the positional coordinate deviation (−0.139, −0.049) obtained from the recognition result of the 54th reference mark or the positional coordinates (210−0.139, 103−0.050) including the deviation is stored. Stored in the unit 173. As the area offset value, a positional coordinate shift (−0.132, −0.051) of the 51st reference mark is adopted. This is similarly performed for other reference marks.

  (2) Next, the production board type is selected.

  First, as shown in FIG. 25, in step S21, the board type selection program is transferred from the storage unit 173 to the control device 170, and the board type selection of the board 61 to be produced (to be mounted) is selected by the component mounting apparatus. Prompt the operator on the operation screen. When the substrate type is selected by the operator, the size of the selected substrate and the NC coordinate data of the position coordinate of the reference mark 201 are read from the storage unit 173 by the control device 170.

  Next, in step S22, two substrate reference position calculation marks 202-1 for the substrate 61 of the selected substrate type from the NC coordinate data read out according to the selected substrate type by the control device 170, Each of the position coordinates 202-2 is extracted.

  In the 410 mm × 240 mm glass substrate of the above specific example, (15, 18) and (215, 111) are extracted as the position coordinates of the substrate reference position calculation marks 202-1 and 202-2.

  Next, in step S23, on the glass substrate 200 that is closest to the positions of the two substrate reference position calculation marks 202-1 and 202-2 by calculation by the calculation unit 171 based on the data stored in the storage unit 173. One reference mark 201 is extracted. For example, in FIG. 22, the lower left first reference mark 201a is extracted from the first substrate reference position calculation mark 202-1 and the lower left 52nd reference mark 201b is extracted from the second substrate reference position calculation mark 202-2. Extract.

  In the 410 mm × 240 mm glass substrate of the above specific example, the position coordinates (10, 10) of the first reference mark 201a at the lower left are the position coordinates (15, 18) of the first substrate reference position calculation mark 202-1. ) And the position coordinates (210, 110) of the lower left 52nd reference mark 201b are extracted from the position coordinates (215, 111) of the second substrate reference position calculation mark 202-2.

  Next, in step S24, the parallel displacement, the inclination, and the expansion / contraction rate are obtained by calculation by the calculation unit 171 from the respective recognition results of the two extracted first reference marks 201a and 52nd reference marks 201b.

  Specifically, of the two first reference marks 201a and the 52nd reference mark 201b, the parallel shift is considered based on the first reference mark 201a.

Therefore, assuming that the offset value of the first reference mark 201a is (ΔX _a , ΔY _a ), the parallel shift amounts (ΔX _ab , ΔY _ab ) can be described by the following equations.
[Equation 6]
ΔX _ab = ΔX _a
ΔY _ab = ΔY _a
In the 410 mm × 240 mm glass substrate of the above specific example, when the area offset value of the first reference mark 201 a is (−0.132, −0.051), the parallel shift amount is calculated from the above equation (6). (−0.132, −0.050).

  On the other hand, the inclination of the glass substrate 200 is obtained by adding respective offset values to the straight line connecting the NC coordinates of the first reference mark 201a and the 52nd reference mark 201b and the NC coordinates of the first reference mark 201a and the 52nd reference mark 201b. This is the angle formed by the straight line connecting the coordinates.

The NC coordinates of the first reference mark 201a and the 52nd reference mark 201b are (X _a , Y _a ) and (X _b , Y _b ), and the offset values of the first reference mark 201a and the 52nd reference mark 201b are respectively Assuming (ΔX _a , ΔY _a ) and (ΔX _b , ΔY _b ), the inclination Δθ _ab between the first reference mark 201a and the 52nd reference mark 201b can be described by the following equation.
[Equation 7]
Δθ _ab = tan ⁻¹ {(Y _b −Y _a ) / (X _b −X _a )} − tan ⁻¹ [{(Y _b + ΔY _b ) − (Y _a + ΔY _a )} / {(X _b + ΔX _b) )-(X _a + ΔX _a )}]
In the 410 mm × 240 mm glass substrate of the specific example, the NC coordinates of the first reference mark 201a and the 52nd reference mark 201b are (10, 10) and (210, 110), and the first reference mark 201a and If the offset values of the 52nd fiducial mark 201b are (−0.132, −0.051) and (−0.130, −0.067), respectively, the first fiducial mark 201a and the 52nd fiducial mark 201b The slope Δθ _ab of
[Equation 8]
Δθ _ab = tan ⁻¹ {(110−10) / (210−10)} − tan ⁻¹ [{(110−0.067) − (10−0.051)} / {(210−0.130) − (10−0.132)}]
= -0.004125 °
It becomes.

  In addition, the expansion / contraction rate E of the glass substrate 200 is set to 1 when taking the idea of using the glass substrate 200 as a reference.

Next, in step S25, the position coordinates of the positions of all the reference marks 201 stored in step S3 of FIG. 11 and corresponding to the area of the substrate 61 to be mounted are calculated by the calculation unit 171 using the parallel shift and inclination (and expansion / contraction ratio). The position coordinate of the corrected reference mark 201 is stored in the storage unit 173. Specifically, the correction value of each reference mark 201 is corrected in consideration of the parallel shift, inclination, and expansion / contraction rate of the first reference mark 201a and the 52nd reference mark 201b, and then stored in the storage unit 173 as an offset value. Will do. Here, the parallel deviation is (ΔX _ab , ΔY _ab ), the inclination is Δθ _ab , the expansion / contraction rate is E, and the NC coordinates of the first reference mark 201a are (X _a , Y _a ), and any reference mark to be corrected If the NC coordinates of 201 are (X _nc , Y _nc ) and the offset values are (ΔX _R , ΔY _R ), the corrected offset values (ΔX _off , ΔY _off ) of each reference mark 201 are described by the following equations. it can.
[Equation 9]
_{X off = E {((X} nc + ΔX R) -X a)} cosΔθ ab - ((Y nc + ΔY R) -Y a) sinΔθ ab} - (X nc -X a) + ΔX ab
Y _off = E {((X _nc + ΔX _R ) −X _a )} sin Δθ _ab + ((Y _nc + ΔY _R ) −Y _a ) cos Δθ _ab } − (Y _nc −Y _a ) + ΔY _ab
It becomes.

In the 410 mm × 240 mm glass substrate of the above specific example, the parallel displacement is (−0.132, −0.050), the inclination is Δθ _ab = 0.004125 °, and the expansion / contraction rate is E = 1.000226. When the NC coordinates of the first reference mark 201a are (10, 10) and the offset values are (−0.132, −0.050), the corrected offset values (ΔX _off , ΔY of the first reference mark 201). _off ) becomes (0, 0). Similarly, if the NC coordinate of the reference mark 201 of 15 rows and 8 columns to be corrected is (150, 80) and the offset value is (−0.132, −0.060), the offset after correction of the reference mark 201 is corrected. The values (ΔX _off , ΔY _off ) are (−0.001, −0.015).

  (3) Next, reference mark recognition and component mounting operations are performed.

  First, as shown in FIG. 26, in step S31, the controller 170 determines the movement position where the head 136 should move for the reference mark recognition operation, the component mounting operation, or the mounting offset value measurement operation. To obtain a recognition position or a mounting position.

  At this time, for example, at the time of component mounting operation, the head 136 is moved by the XY robot 120 and stopped at a certain moving position, and the nozzle 1361 with the head 136 is placed on the mounting position after correcting the component 62 with the substrate 61. When the sucked and held component 62 is positioned and ready for mounting, the reference mark 201 closest to the center of the visual field of the board recognition camera 140 of the head 136 at that time is considered as the reference mark 201 for the component 62.

  Similarly, at the time of the reference mark recognition operation, the head 136 is moved by the XY robot 120 and stopped at a certain movement position, and the head 136 is moved over the position of the reference mark 201 after correction of the reference board for reference mark recognition 200. When a certain nozzle 1361 is positioned, the reference mark 201 closest to the center of the visual field of the substrate recognition camera 140 of the head 136 at that time is considered as the reference mark 201 for the certain reference mark 201.

  Similarly, during the mounting offset value measuring operation, the head 136 is moved by the XY robot 120 and stopped at a certain moving position, and a certain reference position calculation mark 202-after correction of the reference mark recognition reference substrate 200 is performed. When the nozzle 1361 with the head 136 is positioned on the position 1 or 202-2, the reference mark 201 closest to the visual field center of the substrate recognition camera 140 of the head 136 at that time is used for calculating the substrate reference position. Consider the reference mark 201 for the mark 202-1 or 202-2.

  Next, in step S <b> 32, the calculation unit 171 adds the area offset value corresponding to the movement position of the head 136 in step S <b> 31 to the position coordinates of the movement position of the head 136. Specifically, as shown in FIG. 23, when there are M rows of reference marks 201 in the vertical direction and N columns in the horizontal direction of the substrate 61 to be mounted (therefore, a total of M × N reference marks 201). An area surrounded by four reference marks 201 (area indicated by P in FIG. 23) is assigned as one area. Any one of the four reference marks 201 as an area offset value with respect to the position coordinates of the mounting position of the component 62 to be mounted in the area (or the position coordinates of the individual mark as a guide for the mounting position). For example, the offset value of the position of the lower left reference mark 201c is adopted, and this offset value is added as the area offset value to the position coordinates of the mounting position (or the position coordinates of the individual marks that serve as a guide for the mounting position). Make corrections.

  Next, by moving the head 136 to the corrected position coordinates, positioning with high accuracy can be ensured, and reference mark recognition operation, component mounting operation, or mounting offset value measurement operation with high accuracy can be performed. In particular, in component mounting operations, individual components such as IC components (BGA components, etc.) that require high mounting accuracy (for example, XY robot positioning accuracy of about ± 2 μm and total accuracy as a mounting machine of about ± 20 μm) are required. The area offset value can be used as a numerical value for correcting individual marks.

  In addition, when the position coordinate (position coordinate) of the recognized reference mark 201 is stored in the storage unit 173 in step S3 of FIG. 11, the following correction may be further added. That is, as shown in FIG. 12, the position coordinates of each reference mark 201 recognize the two reference marks 201A and 201B at the lower left and upper right of the glass substrate 200, and the parallel displacement of the glass substrate 200 with respect to the transfer table 165. Then, the calculation unit 171 calculates the recognition positions of all the reference marks 201 to be measured in consideration of the correction values.

  Regarding the parallel displacement of the glass substrate 200, the reference mark 201A is considered as a reference among the two points of the reference marks 201A and 201B. Further, when recognizing the reference marks 201A and 201B, the center of the substrate recognition camera 140 is moved to the position of the reference mark 201 in the NC coordinates, so that the parallel shift amount (ΔX, ΔY) This is a positional coordinate shift (a shift amount from the center of the recognition visual field of the substrate recognition camera 140) obtained from the recognition result.

Therefore, assuming that the positional coordinate shift obtained from the recognition result of the reference mark 201A is (ΔX _A , ΔY _A ) (see FIG. 34), the parallel shift amount (ΔX _g , ΔY _g ) of the glass substrate 200 is expressed by the following equation. It can be described by.
[Equation 10]
ΔX _g = ΔX _A
ΔY _g = ΔY _A
Note that the coordinates are converted from the position coordinate system to the NC coordinate system.

  The inclination of the glass substrate 200 is an angle Δθ formed by a straight line connecting the reference mark 201A and the reference mark 201B on the NC coordinate and a straight line connecting the recognized reference mark 201A ′ and the reference mark 201B ′.

That is, the NC coordinates of the reference marks 201A and 201B are set to (X _A , Y _A ) and (X _B , Y _B ), and the position coordinate deviation (from the center of the field of view) obtained from the recognition result when the reference marks 201A and 201B are recognized. (ΔX _A , ΔY _A ), (ΔX _B , ΔY _B ), the substrate tilt Δθ _g can be described by the following equation.
[Equation 11]
Δθ _g = tan ⁻¹ {(Y _B −Y _A ) / (X _B −X _A )} − tan ⁻¹ [{(Y _B + (− ΔY _B )) − (Y _A + (− ΔY _A ))] } / {(X _B + ΔX _B ) − (X _A + ΔX _A )}]
= Tan ⁻¹ {(Y _B −Y _A ) / (X _B −X _A )} − tan ⁻¹ [{(Y _B −ΔY _B ) − (Y _A −ΔY _A )} / {(X _B + ΔX _B )-(X _A + ΔX _A )}]
Note that the coordinates are converted from the position coordinate system to the NC coordinate system.

Therefore, the position coordinate of each recognized reference mark 201 is calculated by the calculation unit 171 in consideration of the parallel shift and inclination of the glass substrate 200 described above. Here, the parallel shift is (ΔX _g , ΔY _g ), the inclination is Δθ _g , the NC coordinate of the reference mark 201A is (X _A , Y _A ), and the NC coordinate of the reference mark N at an arbitrary position on the glass substrate 200 Where (X _N , Y _N ) is the recognition position (X _RN , Y _RN ) of the reference mark N at an arbitrary position,
[Equation 12]
X _RN = (X _n −X _A ) cos θ− (Y _m −Y _A ) sin θ + ΔX _g
Y _RN = (X _n −X _A ) sin θ + (Y _m −Y _A ) cos θ + ΔY _g
It becomes.

  Accordingly, the recognition position of the reference mark N obtained in this way may be stored in the storage unit 173 as the position coordinate (position coordinate) of the recognized reference mark 201 in step S3 of FIG. Good.

  According to the above-described embodiment, the reference marks 201 arranged at predetermined intervals on the glass substrate 200 as an example of the reference substrate for recognizing reference marks are recognized, and from the recognition result, each area corresponding to the substrate size is determined. The offset value is determined as an area offset value, and the corresponding movement position of the component mounting head 136 corresponds to either the mounting position correction, the mark recognition correction, the mounting position offset value measurement operation, or any of those operations. By reflecting each area offset value as a numerical value for correction, the cause of deviation due to distortion of the XY robot operation is absorbed, and an optimum offset value corresponding to the size of the board is obtained, so that high-precision mounting is possible. Yes.

  Also, when the reference mark is recognized, the area offset value corresponding to each movement position of the component mounting head 136 is reflected as a numerical value for correction, thereby absorbing a shift factor due to distortion of the XY robot operation, and By obtaining an optimum offset value corresponding to the size, mounting with higher accuracy can be performed.

  In addition, this invention is not limited to the said embodiment, It can implement with another various aspect.

  For example, the two first and 52nd fiducial marks 201a, 201b or 201A, 201B or 202-1 and 202-2 are positioned at different diagonal positions of either the fiducial mark recognition reference board or the board to be mounted, Alternatively, different positions along any one of the XY directions, in other words, any two different points other than the same point may be used.

  When the reference mark recognition reference board 200 is smaller than the board 61 to be mounted, the reference mark recognition reference board 200 is positioned at either end of the component mounting area of the board 61 to be mounted. After recognizing and acquiring the position coordinates of the mark 201, the reference mark recognition reference board 200 is moved to one of the other ends of the component mounting area of the board 61 to be mounted, and the position coordinates of the reference mark 201 are recognized and acquired again. Then, the data may be handled as if the position coordinates of the reference mark 201 were recognized and acquired by a single large virtual reference mark recognition reference substrate 200 by overlapping the common portions. For example, as shown in FIG. 27, specifically, the position coordinates data [1] of the reference mark 201 measured at the normal position of the substrate and the position coordinates of the reference mark 201 measured at the position moved 350 mm to the left. The data [2] is synthesized. Data [1] and data [2] are only subjected to rotation and movement correction so that the common parts coincide. When the expansion / contraction ratio is added, the common parts will not match, so do not multiply.

  The actual example about the change of the deviation | shift amount between the case where the offset value of each area concerning the said embodiment is not applied, and the case where it applies and the change of component mounting precision is shown.

  The offset value of each area was measured using the reference mark 201 of the substrate of 428 mm × 250 mm shown in FIG.

  In FIG. 27, when the reference mark 201 is recognized, the head 136 is arranged in such a manner that the center of the visual field of the substrate recognition camera 140 is 60 mm away from the center of the rightmost nozzle 1361 (that is, rightward in FIG. 27). Therefore, in order to be able to position all the nozzles 1361 from the leftmost nozzle 1361 to the rightmost nozzle 1361 in all regions on the substrate 61, the substrate recognition camera 140 abuts on the left end of the substrate 61. Thus, it is necessary to move 720.5 mm (XL = substrate width 510 mm + 60 mm + 150.5 mm between the nozzles at both ends) from the position of the substrate stopper that positions the substrate 61 at the mounting position of the transfer table 165 (that is, rightward in FIG. 27). There is.

  However, if the reference mark recognition reference substrate used for recognizing the reference mark 201 is only 410 mm in the X direction from the position of the substrate stopper, the reference mark recognition reference substrate is shifted in the X direction, By recognizing the reference mark 201 twice, the entire region (0 mm to 720.5 mm) of the substrate 61 can be covered.

  The graphs shown in FIGS. 28 and 29 are plots of the output data of the positional coordinate deviation obtained from the recognition result when using the offset value of each area. The two graphs in FIG. 28 show the relationship between the position in the X direction and the amount of deviation in the X direction when the head 136 moves at a pitch of 10 mm in the X direction, and the graph [1] shows the offset value of each area. Before use, graph [2] is after using the offset value of each area. The two graphs in FIG. 29 show the relationship between the position in the Y direction and the amount of deviation in the Y direction when the head 136 moves at a pitch of 10 mm in the Y direction, and graph [1] shows the offset value of each area. Before use, graph [2] is after using the offset value of each area.

  The graph [1] before using the offset value of each area in FIG. 28 shows an upward error of 20 μm at the position moved by 200 mm from the substrate stopper before using the offset value of each area in the X direction. Has a convex shape. On the other hand, the corrected graph [2] transitions around near zero.

  From the graph of FIG. 29, the graph [1] before using the offset value of each area in the Y direction transitions with a slight inclination, but the graph [2] after using the offset value of each area is X Similar to the direction, it is changing near zero.

  In graph [2] after using the offset value of each area in FIGS. 28 and 29, the error is within ± 5 μm in both the X direction and the Y direction.

  Next, regarding the change in the component mounting accuracy, when the ceramic capacitor which is a chip component of 1.6 points × 0.8 mm of 400 points is mounted on the substrate with respect to the substrate of 428 mm × 250 mm, the above embodiment FIG. 30 shows the mounting accuracy when the offset value of each area is not used, and FIG. 31 shows the mounting accuracy when the offset value of each area according to the embodiment is applied. In addition, when a large number of QFPs are mounted on a substrate, the mounting accuracy when the offset values of each area according to the above embodiment are not applied is shown in FIG. 32, and when the offset values of each area according to the above embodiment are applied. The mounting accuracy is shown in FIG. The dimension values in each figure are on the order of mm.

  From the above results, as shown in FIGS. 31 and 33, there is an improvement trend in the mounting accuracy in the X direction and the Y direction. That is, it can be seen that the amount of deviation between the corrected apparatus position data and the true mounting position data is numerically smaller than when the offset value of each area according to the above embodiment is not applied. .

  As a specific numerical value as an example, the correction value is about 10 μm to 30 μm. When coordinate conversion is performed on a 400 mm × 250 mm substrate as an example of a small substrate, the expansion / contraction rate is 1.0002005. As an example of a large substrate, when coordinate conversion is performed on a 600 mm × 250 mm substrate, the expansion / contraction ratio is about 1.00005. In addition, a small substrate such as 100 × 100 mm is also effective.

  The present invention can be applied to mounting almost all electronic components. For example, small components such as a square chip capacitor, a square chip resistor, and a transistor, or an IC targeted for fine pitch mounting such as QFP or BGA. It is applicable to.

  Instead of measuring the reference mark recognition reference board with a camera, it can also be achieved by measuring the movement position of the board camera unit with a laser length measuring instrument (in this case, the reference mark recognition reference board is It becomes unnecessary.)

  In addition to the correction by the area offset value described above, mark recognition operation (substrate mark recognition, individual mark recognition corresponding to IC components, pattern mark recognition displayed on each substrate of multi-chip board, each component group (Recognized group mark recognition, bad mark recognition indicating failure display), component mounting operation, mounting offset value measurement operation, “substrate camera offset value” used for head movement position calculation at each operation of reference mark recognition, and By reflecting the “substrate camera offset value” at the time of camera calibration and the area offset value at the measurement position of “nozzle pitch” at the time of camera calibration, it is possible to improve the accuracy.

  In the camera calibration described above, the offset value of the substrate camera 140 and the pitch between nozzles (distance between nozzles of a plurality of nozzles) are obtained. In the process of obtaining the offset value, each area for correcting distortion of the XY robot is obtained. The correction value is not reflected. Therefore, by reflecting the offset value and inter-nozzle pitch of the substrate camera 140 used when calculating the head movement position during mark recognition, component mounting operation, and / or mounting offset value measurement operation, higher accuracy can be achieved. Can be installed. The offset value and inter-nozzle pitch of the substrate camera 140 are given by the distance from the first nozzle 1361-1. Therefore, when reflecting in the offset value of the substrate camera 140 and the pitch between nozzles used when calculating the head movement position during the mark recognition, the component mounting operation, or the mounting offset value measuring operation, the substrate camera offset value or the nozzle pitch measurement is performed. The difference between the area offset value at the time and the area offset value at the time of measuring the position of the first nozzle 1361-1 is reflected during each operation.

  Hereinafter, the positional relationship among the nozzle, the component recognition camera 150, and the board recognition camera during measurement will be described with reference to FIG.

  As shown in FIG. 37A, when measuring the position of the first nozzle (referred to as a reference nozzle) 1361-1, the first nozzle 1361-1 is positioned on the component recognition camera 150, and the first nozzle 1361-1 is measured. Measure the position. The value of the position of the first nozzle 1361-1 obtained by the measurement in this state is defined as an area offset value (X1, Y1).

  Subsequently, when the inter-nozzle pitch of the n-th nozzle 1361-n is measured as shown in FIG. 37B, the n-th nozzle 1361-n is positioned on the component recognition camera 150, and the n-th nozzle 1361 is measured. -N position measurement. The value of the position of the nth nozzle 1361-n obtained by the measurement in this state is set as an area offset value (Xn, Yn). In the case of the head shown in FIG. 37, since the total number of nozzles is 8, n is sequentially measured from 2 to 8, and is set as the area offset value of each first nozzle 1361-1.

  Subsequently, as shown in FIG. 37C, when the board camera 140 is measured, the board camera 140 is positioned on the component recognition camera 150, and the position of the board camera 140 is measured. The value of the position of the substrate camera 140 obtained by the measurement in this state is defined as an area offset value (Xp, Yp).

  As shown in FIG. 38, the offset value of the substrate camera and the pitch between nozzles are given by the distance from the first nozzle 1361-1. Therefore, when the area offset value is reflected, the difference between the substrate camera offset value or the area offset value at the time of measuring the pitch between nozzles and the area offset value at the time of measuring the position of the first nozzle 1361-1 is calculated during each operation. To reflect.

  For example, referring to FIG. 38, the area offset value at the time of position measurement of the first nozzle 1361-1 at the time of camera calibration is (X1, Y1), and the nozzle of the nth nozzle 1361-n at the time of camera calibration. Assuming that the area offset value at the time of the inter-pitch measurement is (Xn, Yn) and the area offset value at the time of measuring the substrate camera offset value at the time of camera calibration is (Xp, Yp), "(Xp-X1, Yp-Y1)". Further, the area offset value reflected in the “inter-nozzle pitch” of the n-th nozzle 1361-n in the component mounting operation is (Xn−X1, Yn−Y1).

  As shown in the flowchart of FIG. 35, during the reference mark recognition operation, an area offset value corresponding to the position measurement position of the first nozzle 1361-1 during camera calibration is obtained in step S51.

  In step S52, an area offset value corresponding to the board camera offset value measurement position at the time of camera calibration is obtained.

  Next, when the area offset value is reflected in the substrate camera offset value in step S53, the movement position of the head 136 is obtained, and in step S22 (FIG. 25), the area offset value corresponding to the movement position of the head 136 is obtained. Further, in step S23 (FIG. 25), an area offset value corresponding to the position where the first nozzle (nozzle serving as a reference position of the inter-nozzle pitch and the substrate camera offset value) 1361-1 is located on the recognition camera is obtained, and step S24 is performed. In FIG. 25, an area offset value corresponding to the position where the substrate camera 140 is on the recognition camera is obtained. In step S25, the area offset value obtained in step S22 during the reference mark recognition operation is reflected, and in step S54, the difference between the area offset value obtained in step S23 and the area offset value obtained in step S24 (in step S24). The obtained area offset value—the area offset value obtained in step S23) is reflected. Specifically, in step S54, the difference between the area offset values obtained in step S52 and step S53 (area offset value in step S53−area offset value in step S52) is added to the substrate camera offset value. Next, in step S55, the substrate mark recognition movement position is obtained using the substrate camera offset value in step S54. Next, in step S56, an area offset value corresponding to the movement position obtained in step S55 is obtained. Next, in step S57, an area offset value corresponding to the movement position obtained in step S56 is added. Next, in step S58, the substrate camera is moved to the movement position obtained in step S57.

  With this configuration, it is possible to reflect the area offset value due to the distortion of the XY robot operation included in the inter-nozzle pitch and the substrate camera offset value, and it is possible to perform mounting with higher accuracy.

  The flowchart of FIG. 36 shows the procedure for performing the component mounting operation by reflecting the area offset value at the measurement position of the inter-nozzle pitch.

  First, as described above in steps S62 and S63, the area offset values of the first nozzle and the nth nozzle at the time of camera calibration are obtained. That is, in step S62, an area offset value of the area corresponding to the position measurement position of the first nozzle at the time of camera calibration is obtained. Next, in step S63, an area offset value corresponding to the area of the pitch measurement position between the nth nozzles during camera calibration is obtained.

  Next, in step S64, the difference between the area offset values obtained in steps S62 and S63 (area offset value in step S63−area offset value in step S62) is added to the nth nozzle pitch.

  Next, in step S65, the component mounting position is obtained using the inter-nozzle pitch in step S64.

  Next, in step S66, an area offset value corresponding to the movement position obtained in step S65 is obtained.

  Next, in step S67, the area offset value of the area corresponding to the movement position obtained in step S66 is added.

  Next, in step S68, the nozzle is moved to the movement position obtained in step S67.

  It is to be noted that, by appropriately combining arbitrary embodiments of the various embodiments described above, the effects possessed by them can be produced.

  The component mounting method and apparatus according to the present invention recognizes the reference marks 201 arranged at predetermined intervals on the glass substrate 200 and corrects the offset value for each area according to the substrate size from the recognition result. The value is determined as a numerical value, and the corresponding offset value of each movement position of the component mounting head 136 is reflected as a correction numerical value at the time of mounting position correction, mark recognition correction, or mounting position offset value measurement, respectively. Thus, the mounting accuracy can be improved, which is useful.

It is a top view of the component mounting apparatus which can enforce the component mounting method concerning one Embodiment of this invention. It is a front view of the said component mounting apparatus shown in FIG. It is a right view of the said component mounting apparatus shown in FIG. It is a conceptual diagram of the mount frame and XY robot with which the said component mounting apparatus shown in FIG. 1 is equipped. It is a front view of the component mounting head of the X-axis robot with which the said component mounting apparatus shown in FIG. 1 is equipped. It is a block diagram which shows the relationship between each component of the said component mounting apparatus shown in FIG. 1, and a control apparatus. It is explanatory drawing which shows the relationship between the distortion | strain of an X-axis robot and a component mounting head for demonstrating that the positioning accuracy of a component mounting head is influenced greatly by the distortion | strain of an XY robot. It is explanatory drawing which shows the relationship between the distortion | strain of a Y-axis robot, and a component mounting head for demonstrating that the positioning accuracy of a component mounting head is influenced greatly by the distortion | strain of an XY robot. It is explanatory drawing for demonstrating the view of the offset value of the said component mounting method concerning the said embodiment of this invention. It is a top view which shows the specific example of the glass substrate used in the said component mounting method concerning the said embodiment of this invention. It is a flowchart which shows the procedure which calculates | requires and uses the offset value of the said component mounting method concerning the said embodiment of this invention. It is a top view which shows the reference mark of the glass substrate used in the said component mounting method concerning the said embodiment of this invention. It is explanatory drawing for demonstrating the recognition method of the reference mark of the glass substrate used in the said component mounting method concerning the said embodiment of the present invention. In the component mounting method according to the embodiment of the present invention, it is an explanatory view showing that reference mark is recognized to the displacement position from the field center position O _1, O ₂ substrate recognition camera. It is explanatory drawing which shows the result at the time of the recognition of two board | substrate reference position calculation marks in the said component mounting method concerning the said embodiment of this invention. The vertical axis indicates the amount of positional deviation, the horizontal axis indicates the position in the X direction, the upper line graph indicates ΔX, that is, the positional deviation in the X direction, and the lower line graph indicates ΔY, that is, the positional deviation in the Y direction. . It is explanatory drawing which shows the state which has shifted | deviated to the X direction and the Y direction from the center position of the rectangular visual field area | region whose reference mark position is an original position. The coordinate conversion is performed by rotating and moving the graph so that the correction value of the reference mark in the vicinity of the two substrate reference position calculation marks on the board to be mounted is zero or substantially zero. It is a graph which shows the state which rearranges a mounting position. FIG. 19 is a plan view showing two board reference position calculation marks of a relatively small board to be mounted in FIG. 18. By rotating and moving the graph so that the correction value of the reference mark near the two board reference position calculation marks of the relatively large substrate to be mounted is zero or substantially zero, the coordinates are converted, It is a graph which shows the state which rearranges a mounting position. FIG. 21 is a plan view showing two board reference position calculation marks of a relatively large board to be mounted in FIG. 20. It is explanatory drawing which shows the reference mark on the glass substrate nearest to the board | substrate reference position calculation mark of a production substrate. It is explanatory drawing which shows the state which allocates the area | region P enclosed by the four reference marks as one area, when there are M rows of reference marks in the vertical direction of the substrate to be mounted and N columns of horizontal marks. It is a flowchart of the reference | standard mark recognition operation | movement in the more specific example of the component mounting method concerning the said embodiment. It is a flowchart of the kind selection operation | movement in the more specific example of the component mounting method concerning the said embodiment. 6 is a flowchart of a reference mark recognition operation and a component mounting operation in a more specific example of the component mounting method according to the embodiment. It is explanatory drawing at the time of combining the data [1] of the position coordinates of the reference mark measured at the normal position of the substrate and the data [2] of the position coordinates of the reference mark measured at the position moved 350 mm to the left. 28 is a graph showing the relationship between the position in the X direction and the amount of deviation in the X direction when the head is moving at a pitch of 10 mm in the X direction on the substrate of FIG. 28 is a graph showing the relationship between the position in the Y direction and the amount of deviation in the Y direction when the head is moving at a pitch of 10 mm in the Y direction on the substrate of FIG. Mounting accuracy when the offset value according to the above embodiment is not applied when a ceramic capacitor which is a chip component of 1.6 mm × 0.8 mm of 400 points is mounted on a substrate having a size of 428 mm × 250 mm. FIG. 6 is a graph showing the amount of wearing displacement in the Y direction on the vertical axis and the amount of wearing deviation in the X direction on the horizontal axis. Mounting accuracy when applying the offset value according to the above embodiment when a ceramic capacitor which is a chip component of 1.6 mm × 0.8 mm of 400 points is mounted on a substrate having a size of 428 mm × 250 mm. FIG. 6 is a graph showing the amount of wearing displacement in the Y direction on the vertical axis and the amount of wearing deviation in the X direction on the horizontal axis. When a large number of QFPs are mounted on a substrate having a size of 428 mm × 250 mm, the graph shows the mounting accuracy when the offset value according to the above embodiment is not applied, and the mounting shift amount in the Y direction Is a graph in which the vertical axis indicates the amount of misalignment in the X direction and the horizontal axis indicates the amount of displacement. A graph showing the mounting accuracy when applying the offset value according to the above embodiment when a large number of QFPs are mounted on a substrate having a size of 428 mm × 250 mm, and the amount of mounting misalignment in the Y direction Is a graph in which the vertical axis indicates the amount of misalignment in the X direction and the horizontal axis indicates the amount of misalignment. It is explanatory drawing which shows the positional offset amount to the X direction and Y direction of the reference mark from the visual field center of a board | substrate recognition camera. As an application example of the above-described embodiment, it is a flowchart showing an operation of reflecting an area offset value due to distortion of an XY robot operation included in an inter-nozzle pitch and a substrate camera offset value. It is a flowchart which shows the procedure which performs component mounting operation | movement reflecting an area offset value in the measurement position of the pitch between nozzles. (A), (b), (c) is a figure which shows the positional relationship of the nozzle at the time of a measurement, a component recognition camera, and a board | substrate recognition camera. It is a figure for demonstrating the offset value of a board | substrate camera, and the pitch between nozzles.

Explanation of symbols

DESCRIPTION OF SYMBOLS 51 ... X-axis direction, 52 ... Y-axis direction, 53 ... Z-axis direction, 61 ... Circuit board, 62 ... Electronic component, 100 ... Component mounting apparatus, 110 ... Mount, 120 ... XY robot, 121 ... Y-axis robot, 122 ... Y-axis ball screw structure, 122a ... one end, 122b ... other end, 131 ... X-axis robot, 132 ... X-axis frame, 133 ... X-axis ball screw structure, 133a ... one end, 133b ... other end, 136 ... component mounting head, 140 DESCRIPTION OF SYMBOLS ... Board | substrate recognition camera, 150 ... Component recognition camera, 160 ... Reference | standard mark, 165 ... Conveyance table, 170 ... Control apparatus, 171 ... Calculation part, 173 ... Memory | storage part, 180 ... Component supply apparatus, 190 ... Substrate conveyance apparatus, 200 ... Glass substrate, 201, 201a, 201b, 201A, 201B ... reference mark, 202, 202-1 and 202-2 ... substrate reference position calculation mark , 1361 ... suction nozzle, 1362 ... motor.

Claims (8)

The component (62) held by the component holding head (136) movable with respect to the substrate holding device is mounted at the component mounting position of the component mounting substrate (61) held by the substrate holding device (165). In the component mounting method,
Reference marks (201) arranged at predetermined intervals of the reference substrate held by the substrate holding device in a state where the reference substrate for reference mark recognition (200) is held by the substrate holding device and positioned in the component mounting area. ) To determine the position coordinates of each of the recognized reference marks,
The difference between the NC coordinates of the respective reference marks and the position coordinates is obtained as a correction value, respectively.
NC coordinates of the position coordinates of at least two board reference position calculation marks (201A, 201B) of the component mounting board are acquired,
From the recognized reference marks, reference marks that are close to the two substrate reference position calculation marks are respectively extracted.
While converting the position coordinates of the extracted reference marks so that the correction values of the extracted reference marks become zero or substantially zero, the offset values at the respective reference marks are obtained.
In place of the reference mark recognizing reference substrate, the component mounting substrate is held by the substrate holding device and positioned in the component mounting region, and at least two of the component mounting substrates held by the substrate holding device are used. Recognizing each of the two substrate reference position calculation marks, and obtaining the position coordinates of the two recognized substrate reference position calculation marks,
Based on the obtained position coordinates of the two substrate reference position calculation marks, the NC coordinates of the two substrate reference position calculation marks are respectively corrected,
When the component held by the component holding head is positioned above each component mounting position of the component mounting board, the offset value of the reference mark closest to the recognition device provided in the component holding head is used. In addition, after correcting the position coordinates of the component mounting position, the component is mounted at the component mounting position based on the corrected position coordinates of the component mounting position. Implementation method. The position coordinates of the extracted reference marks are respectively coordinate-transformed so that the correction values of the extracted reference marks close to the two substrate reference position calculation marks are zero or substantially zero, When calculating the offset value at the reference mark of
Coordinate conversion is performed by rotating and moving the graph connecting the extracted reference marks so that the correction value of the extracted reference marks close to the two substrate reference position calculation marks is zero or substantially zero. The component mounting method according to claim 1, further comprising: converting the position coordinates of the extracted reference marks to obtain an offset value at each reference mark. The position coordinates of the extracted reference marks are respectively coordinate-transformed so that the correction values of the extracted reference marks close to the two substrate reference position calculation marks are zero or substantially zero, When calculating the offset value at the reference mark of
From the extracted reference mark, a correction value in at least one of the X direction of the substrate holding device and the Y direction orthogonal to the X direction is calculated, and the inclination of the reference substrate is obtained to obtain the extracted value. 2. An offset value at each reference mark is obtained by converting the coordinates of the position of the extracted reference mark so that the correction value of the reference mark becomes zero or substantially zero. Or the component mounting method of 2. The component (62) held by the component holding head (136) movable with respect to the substrate holding device is mounted at the component mounting position of the component mounting substrate (61) held by the substrate holding device (165). In component mounting equipment,
A reference board provided in the component holding head and held in the substrate holding device in a state where the reference mark recognition reference substrate (200) is held in the substrate holding device and positioned in the component mounting area. A recognition device (140) for recognizing the position coordinates of the reference marks (201) arranged at intervals;
The position mark of the reference mark is obtained from the recognition result of the reference mark recognized by the recognition device, and the difference between the NC coordinate of the reference mark and the position coordinate is obtained as a correction value. Based on the NC coordinates of the position coordinates of at least two substrate reference position calculation marks, the reference marks close to the two substrate reference position calculation marks are extracted from the recognized reference marks, respectively. The coordinate values of the extracted reference marks are converted so that the correction value of the extracted reference mark becomes zero or substantially zero, and the offset value at each reference mark is obtained. In a state where the component mounting board is held in the board holding device instead of the recognition reference board and is positioned in the component mounting area, Recognizing each of the at least two board reference position calculation marks of the component mounting board held by the board holding device, and obtaining the position coordinates of the two recognized board reference position calculation marks, respectively. A calculation unit (171) for correcting the NC coordinates of the two substrate reference position calculation marks based on the position coordinates of the two substrate reference position calculation marks,
When the component held by the component holding head is positioned above each component mounting position of the component mounting board, the offset value of the reference mark closest to the recognition device provided in the component holding head is used. And a control device (170) that, after correcting the position coordinates of the component mounting position, mounts the component at the component mounting position based on the corrected position coordinates of the component mounting position. A component mounting apparatus characterized by that.   The calculation unit coordinates the position coordinates of the extracted reference marks so that the correction values of the extracted reference marks that are close to the two substrate reference position calculation marks are zero or substantially zero, respectively. When converting and obtaining offset values at the respective reference marks, the correction values of the extracted reference marks that are close to the two substrate reference position calculation marks are zero or substantially zero. The coordinates connecting the extracted reference marks are rotated and moved to perform coordinate conversion, thereby converting the position coordinates of the extracted reference marks and obtaining offset values at the respective reference marks. Item 5. The component mounting apparatus according to Item 4.   The arithmetic unit coordinates the position coordinates of the extracted reference marks so that the correction values of the extracted reference marks close to the two substrate reference position calculation marks are zero or substantially zero, respectively. When converting and obtaining an offset value at each reference mark, a correction value in at least one of the X direction of the substrate holding device and the Y direction orthogonal to the X direction is extracted from the extracted reference mark. And calculating the inclination of the reference substrate, and converting the position coordinates of the extracted reference marks so that the correction value is zero or substantially zero. The component mounting apparatus according to claim 4 or 5, wherein an offset value is obtained.   Two Y-axis robots (121) installed parallel to each other along the Y-axis direction (52), and the two Y-axis robots movable along the X-axis direction (51) perpendicular to the Y-axis direction And an XY robot (120) having one X-axis robot (131) that movably supports the component holding head (136) along the X-axis direction. The component mounting apparatus according to any one of claims 4 to 6, wherein the component holding head can be moved in the XY axis direction by the two Y axis robots and one X axis robot. The component holding head (136) has a plurality of component suction nozzles (1361) capable of sucking and holding the components and arranged along the X-axis direction, and the center of the plurality of component suction nozzles. The board recognition camera is arranged on the component holding head (136) so that the imaging center of the board recognition camera (140) is positioned coaxially with a straight line passing through the component holding head (136). Component mounting equipment.

Images