JP3848007B2 - Solder bump measurement method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a method for measuring solder bumps formed on a semiconductor module such as an LSI or TAB (Tape Automated Bonding).
[0002]
[Prior art]
FIG. 1 is an explanatory view showing the structure of a semiconductor module 1 having a large number of solder bumps.
A large number of solder bumps 10 are formed on the back surface of the semiconductor module 1, and each bump connects to a wiring board on which the semiconductor module 1 is mounted. The semiconductor module 1 has a square shape of, for example, a 10 mm square, and 23 × 23 solder bumps 10 are formed on the surface 8 at a pitch interval of 450 μm.
Each solder bump 10 has a substantially spherical shape and shape as shown in an enlarged view.
Conventionally, there is no apparatus for automatically measuring the height of such a large number of spherical bumps, and there has been only a visual measurement method using a depth of focus microscope or the like.
[0003]
[Problems to be solved by the invention]
A work (object to be measured) such as a semiconductor module on which a large number of bumps are formed is positioned on the wiring board in the next step, and the solder bumps are heated and melted in a heating furnace to be connected.
In order to ensure this connection, the height of the top of each bump needs to be accurately formed.
In addition, it is necessary to prevent a short circuit with an adjacent bump or wiring by controlling the size of each bump to a constant value.
However, it is very difficult to accurately measure the height dimensions of many bumps in a short time.
The present invention has developed a technique for measuring the height of a workpiece using the light beam disclosed in Japanese Patent Laid-Open No. 2-80905 previously proposed by the present applicant, and measures bumps on the workpiece at high speed and accurately. It provides a way to do this.
[0004]
[Means for Solving the Problems]
The measuring method of the present invention is a method for measuring a solder bump formed on a measured object, the step of attaching the measured object on a table, and a predetermined solder bump among a plurality of solder bumps on the measured object Irradiating a light beam to the predetermined solder bump and detecting the position of the predetermined solder bump by irradiating the reflected light from the predetermined solder bump; and irradiating the solder bump row including the predetermined solder bump with the light beam. Receiving the reflected light from the solder bump array, determining the vertex position of each solder bump in the solder bump array, calculating the amount of deviation of the posture of the object to be measured from the determined result, and the table Adjusting the posture of the object to be measured and correcting the deviation of the position of the object to be measured, irradiating each solder bump on the object to be measured with the light beam, and receiving the reflected light from each solder bump. And a step of measuring the height of the apex of the solder bumps.
[0005]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 2 is an explanatory view showing the appearance of the apparatus used in the method for measuring solder bumps of the present invention.
A measuring apparatus denoted as a whole by reference numeral 100 monitors an operation stage 110 and a control unit 120, a control panel 130, a printer 140 for outputting the result, and a measurement unit of the operation stage 110 for placing a workpiece and performing measurement. A monitor TV 150 and the like are provided.
FIG. 3 is a block diagram showing the configuration of the control system of the second apparatus 100. The table side on which the workpiece 1 is mounted is disposed on the vibration isolation table 210, and is around the vertical axis (Z axis) of the workpiece. A θ stage 212 for controlling the angle, an X stage for controlling movement in one guide axis (X axis) direction in a plane orthogonal to the vertical axis, and movement in the guide axis (Y axis) direction orthogonal to the X axis. A Y stage to be controlled, a β tilt stage 218 and an α tilt stage 220 for controlling the tilt of the surface of the workpiece positioning mechanism, and a workpiece positioning mechanism (work table) 230 for holding the workpiece 1 are provided.
[0006]
Each control shaft is controlled by an output from the shaft control drive 240.
The optical microsensor 250 is attached to the Z-axis drive mechanism 270 and is controlled in the Z-axis direction with respect to the work table 230. The optical microsensor 250 is configured to be retracted from a measurement position by a retracting mechanism 272 in order to facilitate attachment and removal of the work 1 with respect to the work table 230. The amount of movement in the Z-axis direction is managed by the digital micro 274.
An optical camera 252 is mounted adjacent to the optical microsensor 250, and the status of the measurement unit can be visually recognized by the CRT 150 for monitoring.
[0007]
The optical microsensor 250 is controlled by the controller 260, and the measurement data is A / D converted and recorded in the personal computer 266 via the digital input / output interface 264.
The measurement result is sent to the digital input / output interface 280 side of the master personal computer 110 and displayed on the screen of the personal computer 110 and also outputted to the printer 140.
The determination result of the solder bump measurement is operated by the operation switch 284 using the coordinate data of the position of each bump and the correction amount data of each control axis.
[0008]
FIG. 4 shows the structure of the optical microhead 250, which includes a semiconductor laser 252 and a light receiving element 255. The laser output from the semiconductor laser 252 is irradiated onto the bump 10 as a beam LB through the lens 253, and the beam reflected on the surface of the bump 10 is received by the light receiving element 255 through the lens 254. The light receiving element 255 measures the height position of the surface of the bump based on the principle of triangulation, based on the light receiving position of the reflected light. At the same time, the amount of reflected light is also detected.
The X axis of the table is set in the optical axis direction of the laser beam LB.
[0009]
FIG. 5 shows a path in which the bump 10 is scanned by the laser beam LB, and the vicinity including the apex of the bump 10 is measured by three passes.
[0010]
FIG. 6 shows an example of measurement result data, where the horizontal axis represents the X-axis coordinate and the vertical axis represents the detection level. The first curve C1 indicates the level of reflected light, and the second curve C2 indicates the displacement of the bump height.
The reference value TL of the reflected light C1 is determined in advance, the X coordinate position where the received light exceeds the reference value TL and the X coordinate position where the received light falls below the reference value are detected, and the displacement signal C2 at the intermediate coordinate position X10 is detected. The value is adopted as the height position of the bump apex.
[0011]
FIG. 7 shows the relationship between the surface 8 of the substrate 2 and the height positions of the bumps 10.
The laser beam LB scans the surface 8 of the substrate 2 and the bumps 10 to detect the height position of the surface 8 and the height position of the bumps 10.
The substrate 2 does not necessarily form an absolute plane and generates swells and the like. Therefore, as shown in FIG. 8, a regression plane P1 formed by the regression plane P1 of the surface 8 of the substrate 2 and the top of each bump is calculated.
Then, the absolute height H1 of the bump 10 from the regression plane P1 and the relative height H2 from the regression plane P2 are obtained.
The shape and size of the substrate of the workpiece to be measured and the position of the solder bump formed on the substrate do not always have a fixed relationship. Therefore, after the work is mounted on the table, the mounting posture of the work is adjusted (alignment) before the measurement is started.
[0012]
FIG. 9 shows a means for positioning the LSI carrier 1 as a work on a work table 230 that moves in the X and Y directions.
The work 1 has a square shape in plan view, and a right-angle block 232 corresponding to the work 1 is mounted on the table 230. A stopper 233 serving as a reference protrudes from the block 232, and the workpiece 1 is pressed against the stopper 233 by a pressing pin 234 that moves forward toward the stopper 233 to position the workpiece 1.
The table 230 is provided with a suction device 235 using negative pressure, and holds the back surface of the work 1 by suction.
The mounted workpiece 1 has an error between the outer shape of the LSI substrate and the bump position due to cutting error of the outer shape of the substrate, contraction of the substrate material, and the like.
[0013]
Therefore, a step of detecting the position of the first bump 10-1 of the workpiece 1 shown in FIG.
Since the controller recognizes the design position of the first bump 10-1 with respect to the substrate of the work 1, first, the X-axis coordinate of the table is positioned at the design coordinate of the first bump 10-1. The measurement is performed by the first scan S-1 while moving the table in the Y direction.
[0014]
By this scanning S-1, a curve of change in the amount of reflected light received by the first bump 10-1 as shown by a curve C1 in FIG. 6 is obtained. Therefore, the center position of the light quantity obtained from this curve is obtained and assumed to be the center point (origin) of the first bump 10-1. When the amount of received light does not reach the reference value, the X-axis coordinate is shifted by the distance D1, the second scan S-2 is executed, the temporary origin is detected, and the Y-axis coordinate is determined.
Next, this Y-axis coordinate is fixed, scanning X-3 in the X-axis direction is performed, and the position of the first bump 10-1 is detected from the center of the amount of received light. When the amount of received light on the X-axis coordinates does not reach the reference value, the Y-axis is shifted by the distance D2 and scanning S-4 is performed, and the same processing is performed.
[0015]
The coordinate position obtained as described above is set as a temporary origin K1.
The first row of bumps including the temporary origin K1 is scanned (S-10), and the vertex position is obtained for the bumps whose received light amount is greater than or equal to the reference. An average value of deviation amounts between the X-axis coordinate position of the vertex position and the designed X-axis coordinate position is calculated (FIG. 11).
Next, the same scanning (S-11) is performed on the bump row closest to the opposite end of the substrate, and the average value of the deviation amounts of the X-axis coordinates is calculated.
With this amount of deviation, the X-axis coordinate of the origin is corrected as a correction value in the X-axis direction from the substrate end surface of the workpiece to the first bump.
Similar scanning is performed in the Y-axis direction, the amount of deviation of the Y-axis coordinates is calculated, and the Y-axis coordinates of the origin are corrected.
[0016]
If the workpiece is mounted around the central axis, the peak value of the amount of received light that appears on the scanning line changes in a certain direction when the scanning interval is moved to the designed value and scanning is performed. There is a difference between the position and the design position (FIG. 12).
With this change, the correction amount of the rotation angle θ of the workpiece is calculated, and the θ stage is rotated to correct it.
[0017]
At the time of this θ correction scanning in the X-axis direction, as shown in FIG. 13, the height positions of the positions 8-1, 8-2, 8-3,. Based on the change in the height position, the regression line L-1 is calculated, the correction value α of the tilt angle from the reference line (horizontal line) in the X-axis direction is calculated, and the α tilt stage 218 is corrected.
Similarly, a tilt correction value β in the Y-axis direction is calculated, and the β tilt stage 216 is corrected.
After completing the above alignment process, all bumps are scanned to measure the apex height.
[0018]
FIG. 14 shows a regression plane PL-1 formed by the vertex positions of the bumps 10 based on the measurement results. Then, the deviation of the apex position of each bump from the regression plane PL-1 is calculated, and the determination result is displayed on the display 112.
[0019]
FIG. 15 is a graph of measurement results displayed on the screen.
The horizontal axis represents the difference in height of each bump apex from the regression plane, and the vertical axis represents the number of bumps.
In the case of this work, the number of bumps is 5200, the standard deviation is 0.65 μm, and 3δ is ± 1.95 μm.
For example, if the target specification is set to ± 2 μm, the determination based on the standard deviation of the workpiece becomes a non-defective product and is sent to the next process.
[0020]
FIG. 16 shows examples of various screens displayed on the screen 112 of the measuring apparatus.
The operation screen 300 includes a display area 302 such as a current operation mode, an area 304 that displays a relative height frequency distribution graph, an area 306 that displays a determination result, and an area 308 that displays a detailed result for each LSI 1 that is a work. .
Area 302 is
a. Current operation mode (alignment, vertex measurement, base measurement, judgment calculation, individual remeasurement)
b. Display current inspection position (row, column) numerically and graphically c. After 1 LSI inspection, the coordinates of the defective bump are displayed as numerical values and graphics.
The position of the defective bump is displayed as an image of the position of the defective bump 10 -N of the work 1 on the area 302.
[0021]
The relative height frequency distribution graph 304 is shown in FIG. The determination result is displayed in area 306 as OK or NG. The detailed results for each LSI displayed in the area 308 indicate the alignment results as numerical values of X, Y, Z, θ, α, and β, the average height of bumps, the standard deviation of relative height, and the average light reception. The amount and the base shrinkage in the X and Y directions are displayed.
[0022]
The positional deviation graph screen 310 is a graph showing the distribution of the deviation amount of the vertex position from the design value for each bump.
The height solid distribution graph screen 320 shows a 3D graph of relative height for each bump and a planar distribution graph by color coding.
The defective LSI information screen 330 shows a defective LSI determination mode after completion of full automatic operation (in magazine units).
The inspection condition setting screen 340 is a screen for setting inspection parameters such as a determination threshold.
[0023]
【The invention's effect】
As described above, the present invention provides a method for determining the quality of a work in a short time by measuring the height of all the bumps of an LSI or the like having a large number of solder bumps. It is also possible to cope with bumps to be changed.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram showing an LSI that is a measurement object (work) according to the present invention.
FIG. 2 is a perspective view of a measuring apparatus according to the present invention.
FIG. 3 is a block diagram showing a configuration of a measuring apparatus according to the present invention.
FIG. 4 is a perspective view showing a configuration of an optical microhead.
FIG. 5 is a plan view showing scanning of an optical microhead.
FIG. 6 is a graph showing an example of measurement results.
FIG. 7 is an explanatory view showing a method for measuring a substrate surface and a solder bump apex.
FIG. 8 is an explanatory diagram showing the influence of substrate waviness.
FIG. 9 is a plan view showing attachment of a work to a work table.
FIG. 10 is a plan view showing a method for detecting a vertex position of a solder bump.
FIG. 11 is an explanatory view of a measurement result caused by a mounting error of a substrate.
FIG. 12 is an explanatory diagram of a measurement result caused by a mounting error of a substrate.
FIG. 13 is an explanatory diagram of a measurement result caused by a mounting error of a substrate.
FIG. 14 is an explanatory diagram of a regression plane formed by the vertices of solder bumps.
FIG. 15 is a graph of standard deviation of measurement results.
FIG. 16 is an explanatory diagram showing a display screen of the monitor device.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Semiconductor module 8 The surface 10 of the semiconductor module 1 which has a square planar shape Solder bump 100 Measuring apparatus 110 Operation stage (master personal computer)
112 Display 120 Control unit 130 Operation panel 140 Output printer 150 Monitor TV (CRT)
210 Anti-vibration table 212 θ stage 218 for controlling the angle around the vertical axis (Z axis) of the workpiece β tilt stage for controlling the tilt of the surface of the workpiece positioning mechanism (β tilt stage 216)
220 α tilt stage (α tilt stage 218)
230 Work positioning mechanism (work table)
232 Right angle block 233 Stopper 234 Pressing pin 235 Suction device 240 Axis control drive device 250 Optical microsensor 252 Optical camera (semiconductor laser)
253, 254 Lens 255 Light receiving element 260 Controller 264 Digital input / output interface 266 Personal computer 270 Z-axis drive mechanism 272 Retraction mechanism 274 Digital micro 280 Digital input / output interface 284 of master personal computer 110 Operation switch

Claims (3)

  In the method for measuring solder bumps formed on the object to be measured, the step of attaching the object to be measured on a table, and irradiating a predetermined solder bump among a plurality of solder bumps on the object to be measured with a light beam, Receiving reflected light from the predetermined solder bump and detecting the position of the predetermined solder bump; irradiating the solder bump row including the predetermined solder bump with the light beam; Receiving the reflected light to determine the apex position of each solder bump in the solder bump row, calculating the amount of deviation of the posture of the object to be measured from the obtained result, adjusting the posture of the table and Correcting the deviation of the posture of the object to be measured; irradiating each solder bump on the object to be measured with the light beam; receiving the reflected light from each solder bump; Measuring method of the solder bumps, characterized by a step of measuring the height.   The measurement method according to claim 1, wherein after measuring the heights of the vertices of the plurality of solder bumps, calculating a regression plane formed by the vertices of the solder bumps, And a step of determining whether or not a relative vertex height of each solder bump is within a predetermined range.   3. The measuring method according to claim 2, further comprising a step of calculating a standard deviation of a relative vertex height of each solder bump, and a step of determining whether the standard deviation is good or bad.

Images