JPH0727532A - Multi-beam laser sensor for monitoring semiconductor lead - Google Patents

Abstract

PURPOSE: To provide a method of measuring a component lead at a high speed with a high degree of accuracy, used for a machine for arranging component to be surface-mounted, and a multi-beam laser sensor therefor. CONSTITUTION: The sensor is used for detecting the respective positions and states of many leads of an integrated circuit with accuracy before arranging components such as integrated circuits. The leads 40 of each integrated circuit is passed through a focal point 800 of laser beams 700A to 700D from four laser diodes 60A to 60D, nominally, and accordingly, the positions of the leads can be determined when they blocks entirely or partly the laser beams. With the use of a processor, actual positions of the leads are computed, and then, difference between the actual positions and the nominal positions are computed. Further, the positions of the leads are sorted, and maximum deviations of the leads from optimum adaptable positions or a seating plane are determined. Next, in such a case that the deviation is out of an allowable range, a rejection signal or a rearrangement signal is delivered to the machine.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention accurately detects the position and condition of leads used in an integrated circuit by a surface mount pick and place machine before placing the integrated circuit on the circuit board. About the vision system to do. In particular, the present invention is a non-contact laser base capable of determining the lateral orientation and coplanarity of the leads of all integrated circuit components, including those with ultrafine pitch, at the highest resolution. Sensor system.

[0002]

BACKGROUND OF THE INVENTION As electronic devices are becoming smaller and more complex, the electronic components that are interconnected to achieve the electronic functions of such devices are also becoming smaller in size. Furthermore, various functions are being incorporated into a single unitary monolithic integrated circuit. These integrated circuit components, also referred to as semiconductor packages or chips, have numerous leads or elements that provide electrical connections. As a result of the reduction in size, the integrated circuit is made in a structure called a quad flat pack (QFP) or a quad pack, in which leads of close proximity and separation come out from each side of the quad pack. QFP is especially
It is configured to perform a surface mount placement by using a special lead called a gull wing. The QFP is precisely placed by a pick and place machine on the surface of the circuit board by gull wing leads to make proper contact with the intended circuit connections or pads on the circuit board or workpiece. Similarly, more and more manufacturers are making integrated circuits with finer leads constructed on tape used in a process called tape automated bonding or tab (TAB).

The distance between the centers of adjacent pairs of any pair of such components is called the pitch. Currently, the distance for commonly manufactured parts is 0.6
The pitch is 35 mm (25 mils) and the centers of each lead are 0.635 mm (25/1000 inch) apart. However, due to advances in parts manufacturing technology, 0.381
And 0.254 mm (15 and 10 mil) pitch integrated circuits have been manufactured, and TAB components with hundreds of leads separated by 0.1016 mm (4 mil) pitch have also been made.

In order to automatically manufacture electronic devices using these microelectronic components, the highest degree of precision is required in positioning and placing such components. Precise surface mount component placement machines have been developed to perform these delicate tasks.

There are two types of component placement machines commonly used today. One of them is a Cartesian coordinate system.
To move to the bin, pick up the component, properly orient the component, carry it to the circuit board or other work piece, and place the component in the correct position.
More than one vacuum quill has been used.
The proper position is the position where the leads make proper contact with the intended circuit connections on the circuit board or work piece.
Another type of placement system used is the carousel or turret placement system, which locates around a circular component-bearing mechanism as the components are picked from the bin and placed on the circuit board. It is sent in stages through the station. The parts are not aligned in their parts bin. Typically, gull wing style components can be misaligned by ± 1.27 mm (± 50 mils) and angularly offset by ± 5 degrees. Therefore, the orientation of the parts from the bin must be determined prior to placement. The present invention is also useful for both types of systems that require precise placement of components with the highest degree of speed and precision.

Surface mount circuit boards have a large number of individual small pads. Each lead from each electrical component must be properly placed on one circuit board to ensure proper electrical contact, which requires accurate angular and lateral positioning of the component. The dimensions of the parts to be placed are typically 0.508 mm (0.02 inch) and 50.8
Vary between mm (2.0 inches).

In surface mount component placement machines, a transport arm picks up a component from a component bin and transports it between the component bin and a circuit board using a vacuum quill as the primary tool to gently lift the component to be placed. . The transfer arm moves the part from the bin to the circuit board located on the working table. Most parts placement machines
It has a built-in position encoder that electromechanically determines the position of the placement head. A vacuum quill is attached to the placement head and moved by the placement head from the component bin to the circuit board. Therefore, the component placement machine knows the position of the quill based on the encoder reading. Encoders typically read with an accuracy of ± 4 microns. The accuracy of the encoder limits the ability of the component placement machine to place components accurately on the circuit board. During transport, the angular orientation of the part and the deviation of the part from the center of the quill are determined. Also,
The condition of the leads is also examined to determine if there is any bend or tilt. Then, any required corrections in placement are calculated and the placement head is adjusted to accommodate such corrections. The vacuum quill is then accurately lowered to fit the components onto the circuit board. In current part placement machines, the transfer arm and quill move at a rate of approximately 1 meter per second. However, the speed of movement can vary from 0 to 8 meters per second. At this speed, it is difficult for current systems to cope with the fine pitch and range of components to be placed and to achieve the precision required for alignment.

For high quality manufacturing, the component leads must be placed with at least 80% of the leads overlaid on corresponding pads on the circuit board. 0.508 mm
Devices with (20 mil) pitch are typically 0.2
It has 54 mm (10 mil) wide leads. 80%
By stacking, a lead width of at least 203.2 micrometers (8 mils) must be brought onto the pad with a lead width deviation of 5.08 micrometers (2 mils) or less from the pad. In general, the detection system used to align the components for placement should have a resolution 5-10 times better than required. Therefore, 50.8 specified for high quality manufacturing methods
5.08 to 10.16 micrometers (0.2 mils) to achieve a maximum placement error of 2 micrometers.
~ 0.4 mil) resolution is required.

The leads of the integrated circuit are made by various methods such as stamping and etching. The shape of leads made by different manufacturing methods varies. One method of forming leads is to apply photoresist to the surface of the metal,
To expose it to light. When exposed to light, the photoresist chemically etches through the metal to create leads. The etch rate may be such that the leads do not form generally vertical walls. Rather, the wall has a scalloped shape, and the lead has a substantially trapezoidal cross section. For example, if the leads are etched from top to bottom, their sides are jagged and the bottom or bottom surface is larger than the top. Due to such irregularities in the shape of the leads, the lead detection device needs to be specially configured to accommodate various shapes.

A quad flat pack (QFP) can be made with a plurality of bumpers extending outwardly at each corner of the component and beyond the leads of the component. Bumpers help protect the leads from bending and falling out of place. QFPs with bumpers are creating new problems for lead detection devices. Bumpers interfere with conventional vision detection systems to provide physical protection to keep the leads from bending. The bumper can extend 8 microns beyond the lead.

A conventional vision system used in conjunction with a component placement machine for lead determination is 512.
A solid state television camera with a resolution of x512 pixels or pixels is used. 50.8 mm
(2 inch) portion and corresponding 5 with 512 pixels
The 0.8 mm (2 inch) field of view is 0.1016 mm (4
It produces essentially a mil or four thousandths of an inch) resolution. This is not a sufficient resolution, and in fact, as pointed out earlier, it is necessary to achieve a resolution that is at least one order of magnitude higher. One solution that has been proposed is to use multiple lenses. Multiple fields of view are provided by using different lenses to achieve the required magnification. However, changing the lens system is time consuming and slows down the placement process, thus sacrificing speed for accuracy.

Another solution that has been proposed is to use super resolution. Back-illumination casts shadows onto the solid-state pixel components of the detector array and achieves super-resolution by applying a gray-scale image processing algorithm to the shadow edge intensity information projected onto the detector array. , Which is about four times the resolution achievable with binary processing.

Yet another proposed solution is 2000 ×
A two-dimensional detector array with 2000 pixels. Such products are available from Eastman Kodak of Rochester, NY
Manufactured by Rochester, New York). However, this solution is extremely expensive.

Finally, it is also possible to use a linear array with 1000 to 2000 detector elements. By using such a system, the movement of the part is synchronized with a series of sequentially read frames from the array to form a two-dimensional image. By using super-resolution, a resolution accuracy equal to that required for a 0.508 mm (20 mil) pitch device is achieved.

[0015]

What is needed to be solved by the invention and means for solving the problems are: 3.756,0.25
4 mm (15,10 mils) and even 0.1016 mm
It is an economical system that allows accurate placement of leads with even a (4 mil) pitch. These systems must have a resolution with an accuracy in the range of at least 0.508 micrometers (0.02 mils). What is needed is a system that can accurately position leads of various shapes and sizes. What is needed is a detection system that is unobstructed to prevent light rays from reaching the lead by a bumper that provides physical protection that prevents the lead from bending. The present invention is directed to satisfying these needs.

The present invention in its basic embodiment,
A multi-beam laser-based detection system capable of detecting lateral positions of integrated circuit (IC) array leads with a resolution of 508 micrometers (0.02 mils).
The preferred embodiment with multiple laser beams can also calculate lead height, colinearity, and coplanarity. The present invention optionally
It provides the ability for the bumper to inspect the leads of an integrated circuit with the bumper without blocking the necessary rays that detect all four corners of all leads.

The preferred embodiment incorporates four laser diodes that project four separate beams of laser light that focus to the same point in space. From that point, the laser beam is shunted onto the two detectors. The four lasers are energized sequentially to eliminate interference between the lasers, so that only one diode is active at any given time. Although only a single laser diode is energized at any given time, the multiplexing speed is such that with respect to the speed of the placement head, all four laser beams are present simultaneously for most calculations. Fast enough to think of as something to do.

The preferred embodiment calculates the lead height based on the lateral distance traveled along the scan axis during the interruption of a particular laser beam. The scan axis is the path of relative motion between the detection system and the part. The measured resolution of the lateral position along the scan axis directly affects the calculated height resolution. Pick and place
The resolution of the machine encoder can be increased by interpolation methods or phase locked loop rate multipliers.

The four lasers of the preferred embodiment are arranged in two pairs, each pair working together to detect a particular corner of the lead cross section. Since the four lasers are arranged in pairs so that they are directed to a particular corner of the lead being passed, the system can detect and measure the position of each of the four corners in the cross section of each lead. Thus, the system can accurately determine lateral position, co-linearity and co-planarity of integrated circuit leads.

One object of the present invention is to analyze the leads of an integrated circuit component to determine if any of the leads are bent. One object of the present invention is to detect the lateral position of each lead of an integrated circuit and determine if the lead position is offset from its normal position. One object of the present invention is to determine the co-linearity of the leads of integrated circuit components. One object of the present invention is to determine the lead coplanarity of integrated circuit components. One object of the present invention is to determine the position and height of the bottom corners of leads of integrated circuit components and their lateral position, co-linearity and co-planarity. One object of the present invention is to determine the position and height of the bottom corners of the leads of an integrated circuit component having bumpers and their lateral position, co-linearity and co-planarity. One object of the present invention is 0.254 micrometer (0.01 mil) for lateral position.
And 1.5 for co-linearity and co-planarity.
Inspecting lateral position of leads of integrated circuit components and co-linearity and co-planarity with an accuracy in the range of 24 micrometers (0.06 mils). One object of the present invention is to provide 0.254 micrometers (0.0
1 mil) and with a lateral accuracy of 1.524 micrometers (0.06 mils) for co-linearity and co-planarity with the lateral position of leads of integrated circuit components having bumpers. , Co-linearity and co-planarity. These and other objects and advantages of the invention will be apparent to those skilled in the art upon examination of the following description, accompanying drawings and claims.

[0021]

DETAILED DESCRIPTION OF THE INVENTION Current manufacturers of electronic products such as audio and video devices and computer systems desire to package such products in as small a package or area as possible using current technology. . It has also been necessary to automate the manufacturing process in order to achieve these objectives and economically manufacture such electronic devices. As a result, surface mount electronics have been developed. Using a pick and place machine, the electronic components are placed on a circuit board 28 that is specifically configured to receive the surface mount electronic components 24. For such automated manufacturing methods, it is essential for quality control and to make an actuatable device that each component 24 be accurately placed in proper position on a circuit board 28.

In FIG. 1, a placement head 30 is suspended and driven from a rail 32 to pick up an electronic component 24 to be placed by a vacuum quill 34 and carry the component 24 to a circuit board 28 at the fastest possible speed. , A typical component placement machine system for accurately placing electronic components 24 in the proper position by accurately placing leads 40 on corresponding circuit pads or lines on circuit board 28. "High-precision parts matching sensor system (H
high Precision Component A
ligence sensor system), filed February 22, 1991, co-pending U.S. patent application Ser. Vacuum quill 34 used to generate the correct angular orientation at and also carries the component 24 to the circuit board 28.
A non-contact laser-based alignment sensor 42 is disclosed that electronically determines any deviation of the center of the component 24 relative to the center of the. This disclosure is included herein for reference.

The high speed laser-based system 42 according to the prior invention described above is part 2 whose alignment is being detected.
A stripe of laser light (not shown) that is passed or blocked by 4 is utilized. The shadow projected by component 24 is detected by a linear array detector (not shown), and the detector data is analyzed to detect the leading and trailing edges of the shadow. This shadow edge detection information is analyzed to achieve angular and lateral alignment of component 24. Although angular orientation can be achieved with an accuracy of better than 0.03 degrees and lateral positioning can be achieved with an accuracy of better than 0.0254 mm (0.001 inch), the system described above uses each of the leads 40 of component 24. Is not configured to be placed exactly.

The conventional sensor system 42 is suitable for leadless components 24 and all other components 24 including components 24 that are detected and placed using the present invention 50. The sensor system 42 and the sensor system 50 of the present invention are suitable for handling the size of all electronic components 24 conventionally used for surface mount component pick and place machines. Generally, these electronic components 24 have a size of 0.508 mm (0.02 mm).
Between 5 inches and 50.8 mm (2.0 inches).
However, larger sizes can be accommodated.

It is usually the largest component 24 that is the most difficult to place accurately. Larger components 24 tend to be integrated circuits with many subtle leads 40. These are the most expensive parts 24, on average $ 100- $
Get to 500. Further, to ensure compactness, the size of leads 40 and the spacing between leads 40 are constantly reduced relative to monolithic integrated circuit 24. 0.6
35 mm (25 mil) pitch, or 0.635 mm
Current component 24, with a lead-to-lead distance of (0.025 inches), is common for quad flat pack components. However, from 0.1016 mm (0.004 inch) to 0.508 mm (0.020 inch) (4-2
Parts 24 with a spacing of 0 mils) are beginning to appear on the market. These components 24 have hundreds of leads 40, each of which needs to be precisely located and bonded to corresponding pads or circuit lines on the circuit board 28. For the smallest part 24,
The required spacing is approximately the same as the diameter of human hair. The resolution accuracy to achieve such placement must be on the order of greater than the maximum error allowed in placement of the component 24.

Problems that may occur or may occur when the leads 40 of the integrated circuit 24 or other semiconductor device 24 are accurately placed on the corresponding circuit board 28 using a pick and place machine. A number of embodiments 50 that provide solutions to are disclosed and claimed herein. In each of the above-described embodiments 50, the leads 40 of the semiconductor device 24 are accurately arranged, and the component 24 is arranged most accurately if it is within the design rule for the component 24, or the sensor 5 is used.
If 0 determines that there is a potential for a short circuit or an open circuit because one or more leads 40 are bent in the vertical or horizontal plane of the leads 40, then reject component 24. It is specifically configured to provide a signal to the placement machine.

The primary purpose of placing leads 40 and determining co-linearity and co-planarity is to determine where the bottom of leads 40 are located on circuit board 28. Co-linearity is the measurement of whether the calculated height of the lead 40 on one side of the integrated circuit component 24 is sufficient to fit the line. In the preferred embodiment, the line used is the best fit line. In this case, the best fit line (best fit)
line is the square deviation (squ) of the height of the lead 40.
It is a line where the sum of ares division) is the minimum. The positive or negative deviation of one lead 40 from the best fit line can be calculated. In most cases, the co-linearity measurement test rejects integrated circuit components 24 that also fail the co-planarity test.

There are two methods for performing coplanarity measurements: the Seating Plan.
n) ”and“ best fit plane (Best Fit Pla)
in) ”can be adopted. The industry standard test for coplanarity is the Joint Electronic Device Technology Council of Washington Dacy.
ce Engineering Counsel, Wa
Shinton, D.M. C. ) -Using the sheeting plane method set by JEDEC. JEDEC defines the seating plane of surface mount parts, 1
Coplanar surface mount semiconductor device dated November 991
Test ( Coplanarity Test for S
urface-Mount Semiconductor
r Devices ) standard (JESD22)
-B108) is issued. The JEDEC standard seating plane is defined as the plane set by the contact points of three or more leads 40 that support the device 24 when it is placed on a flat surface. The deviation from the coplanar plane is defined as the distance between the predetermined contact point of the lead 40 and the set sheeting plane. Any device with one or more leads 40 that exceed a certain deviation from the coplanar plane is rejected. On the other hand, the best fitting plane is the leads 40 from that plane.
It is defined so that the sum of the squares of the height differences is minimized (minimum square fit).

Referring again to FIG. 1, a station 50 is shown having one or more laser-based sensors constructed in accordance with the teachings of the present invention. As shown in FIG. 1, the component leads 40 are part of the present invention 50 and are capable of passing a laser beam 70 which analyzes the lateral position and / or coplanarity of each lead 40.
For a person skilled in the art, the focus 6 of the laser beam 70
The movement of the lead 40 through 6 is relevant and the component 24 is stopped at the sensor station 50,
There, it stabilizes and the sensor 50 moves to the translation table (tra
It is understood that it is moved by the lead 40 in an nslation table (not shown) or the like. It is also understood that both sensor 50 and component 24 move simultaneously during the measurement.

In FIG. 2, (a) shows a laser sensor 5.
0 shows the basic elements, and (b) and (c) are a side view and a top view of one physical configuration of the present invention 50. The arrangement includes a single light source 60 and a laser diode, which is focused through a condenser lens 62 and is of a laser beam 70 having a diameter of about 25.4 micrometers (about 1 mil). A focal point 66 is focused on the plane of the lead 40 that passes through the sensor system 50. The laser beam 70
If not blocked by the lead 40 being detected,
The laser beam 70 strikes a single detector 74.

A single laser beam 70 is shown in FIG.
However, in this figure and the subsequent figures, FIG.
It should be understood as showing a focused laser beam 70 as shown in FIG. 2 (b), which is about 25.4 in the nominal or nominal plane of the lead 40 being analyzed.
A micrometer (about 1 mil) focal point 66 is located and the depth of focus is about 0.508 mm (about 20 mil). Also, the detector 74 is shown in FIG. 2 (a) as a direct uninterrupted line through which the laser source is visible, but in practice the laser beam is bent using a mirror 76, lens or prism (not shown). Upward intervention of component 24 is minimized for the reasons described below. Since a single light source 60 is used with the control circuit 90 (shown in FIG. 4) to keep the laser power constant, a constant illumination is guaranteed. By using a single detector 74, the detector 7
Uniformity of response from 4 is guaranteed. There are no pixel-to-pixel variations that were common in the other proposed methods.
It is also possible to use a focused laser beam 70 as a distance sensor to detect the leads 40, but such an embodiment would have problems with the microstructure of the lead surface, nicks and scratches on the surface of the lead. Bring problems. Using the illustrated embodiment, the laser beam 70 is used as a backside illumination source to project a detectable shadow onto the detector 74, eliminating the possibility of surface microstructure problems. Finally, the laser beam 7 focused directly on the lead 40 itself
By using 0, it is possible to maximize the use of the data because all the data correspond to the read structure.
There is no need to review or reject the data surrounding the environment surrounding the chip 24 or the plastic packaging of the device 24 as is common in currently used solid state TV systems.

The multi-lead integrated circuit chip 24 is entirely releasable on a strip of tape or the like in the source tray 33, but is held in an adhesive state. Therefore, in general,
The leads 40 on the IC chip 24 have good coplanarity, that is, the bottom surfaces of the leads 40 are generally coplanar. However, during packaging, shipping, and loading of the source tray 33, all leads 40 are not always parallel for placement on the circuit board pads because the leads 40 are slightly bent or tilted. FIG. 3 (a) generally shows a component 24 having a lead 40, two of which 40, 401, 402 are bent and one 401 is biased in the negative direction, Subsequent to the position or nominal position, one 402 is offset in the positive direction. At least 80% of each lead 40, lead 4
It is a general design rule of quality assurance that it is necessary to accurately arrange the circuit pad 41 made to receive 0. This arrangement is shown in FIG. Lead 4
As the 0 shadow is projected onto the detector 74, data is stored that accurately represents the leading edge 404 and trailing edge 405 of each lead 40 of the IC chip 24.

Processor means (94 in FIG. 4) is then used to average the detected positions of the leading edge 404 and trailing edge 405 of each lead 40 to average each lead 4.
Find the center of 0. The difference between the measured center of lead 40 and the nominal position of lead 40 is then calculated. The position of each lead 40 is classified by the deviation of that lead 40 from its nominal position.

Finally, the most negative and most positive excursions are averaged to make the recommended adjustments to place the lead 40 on the receiving pad 41 on the circuit board 28. If no adjustment can be made within the 80% design rule described above, a reject signal can be generated and the component 24 is rejected and returned to the manufacturer. Moreover, the orientation and diagonal position of the component 24 can be adjusted.

The electric circuit used in this embodiment is schematically shown in FIG. The determination of the lead position 40 by the lead position probe measurement and detection instrument 50 is based on the IC lead 40 which intercepts the accurately guided and focused laser light source 60 beam. Each IC lead 40 passes through the focal point 66 of the guided and focused laser beam 70. The position of each lead 40 is determined when it blocks all or part of the beam of laser beam 70. The diameter of the laser beam 70 at the focal point 66 is typically 0.254.
The mm (0.001 inch) can very accurately indicate when the lead 40 enters and exits the beam 70.

The laser beam 70 is cut off by the photosensitive detector 7
4 output is monitored and compared with a reference voltage to determine. Edge portions 404, 40 of each IC lead 40
When 5 is detected, the position of the vacuum quill 34 is recorded. Next, the processor 94 uses the communication interface 112.
Via the leads 40 of the component 24 provide an output to the component placement machine to determine where to place it on the workpiece 28.

The electrical circuit of the lead position detecting instrument 50 includes a detector 74 and a comparator 81. Detector 74 provides a voltage output level corresponding to the amount of light rays striking detector 74 to comparator 81 via amplifier 106. Detector 74
Output voltage is therefore inversely related to the amount of interruption of the laser beam 70. Comparator 81 provides a digital output measurement of beam intercept by comparing the detector output voltage with a stepped reference voltage controlled by threshold adjustment logic 84 and digital-to-analog converter 85. The beam cutoff signal from the comparator 81 has two states: a blocked beam and an unblocked beam.

Threshold adjustment logic 84 drives a digital to analog (D / A) converter 85 to provide a stepped reference voltage to comparator 81, and threshold adjustment logic 84 is provided to comparator 81. And a calibration logic that automatically adjusts the reference voltage to compensate for blockage of the laser beam 70 due to dust in the optical path of the laser beam 70 and to provide long-term laser power drift compensation.
The output from the threshold adjustment logic 84 is provided via the D / A converter 85 and input to the comparator 81.

During the sequence of threshold adjustment logic 84, the output voltage of the detector 74 is determined when illuminated by the uninterrupted laser 70. The threshold voltage is then calculated using a selected percentage of the voltage output of the detector 74 of the uninterrupted laser, which is, for example, 50% in the preferred embodiment. The threshold value thus set is used to determine when a certain percentage of the laser beam 70 is blocked by the lead 40 during operation of the lead position probe 50. Thus, "beam interception" is the desired percentage of the amount of light rays 70 detected by detector 74 that impinge detector 74 when lead 40 is generally out of laser beam 70. When it is determined by the comparator 81.

The decision logic of the electrical circuit is based on the processor 94. Inputs to processor 94 include the beam intercept position received from beam intercept position recorder 98 regarding the instantaneous position of quill 34 from component placement machine encoder 99. The beam cutoff position information is recorded every time the beam cutoff signal changes state. Alternatively, a dynamic signal 9 that more accurately positions the component 24 and its leads 40.
8 can be generated using the embodiment 50 of the invention described in connection with FIG. When the component 24 is stationary and the laser is providing relative motion, the beam 70
It is necessary to record the position of. The processor 94, via the communication interface 112 to the parts placement machine,
The calculations described in connection with the various embodiments of the invention associated with FIGS. 2-13 may be performed. The processor 94 determines the position of the lead, the height of the lead, the width of the lead,
It can be programmed to calculate coplanarity, equations for rejection, equations for repositioning, statistics and various other relevant information. Also preferably included within the processor unit 94 is a library of components 24 in the component tray and sufficient information to calculate the displacement of the components 24 for placement of the components 24 on the circuit board 28. The output of the processor 94 is provided to the control module of the component placement machine via the communication interface 112 so that the component 24 can be accurately positioned. It will be apparent to those skilled in the art that there are many hardware and software possibilities for the processor 94 and the communication interface 112 to the component placement machine.

The laser control diode 90 is a photodiode 1 packaged with the laser diode 60.
60 is adopted. The photodiode 160 monitors the optical intensity of the laser diode 60 and holds a constant optical power output from the laser power control circuit 90.

The detector 74 has a transimpedance (transimpimp) that generates a voltage proportional to the amount of light rays incident on the photodiode 108 of the detector 74.
edance) amplifier 106 is included. The comparator 81 receives the output of the amplifier 106 of the detector 74 and compares the reference voltage provided by the output of the D / A converter 85 with the output of the detector. The D / A converter 85 receives the digital input from the threshold adjustment logic and compares the reference voltage with the comparator 8
Supply to 1.

Repeatability of the present invention
It) and accuracy are mainly influenced by three factors. First, the ability to accurately control the laser power output intensity, second, the detector's electronic signal to noise ratio, and third, the diameter of the focused laser beam 70. The ability to accurately detect the edge, or accurately when half of the beam 70 is blocked, is significantly affected by changing the three factors. Thus, when necessary, the laser power control circuit 90 may be modified to more accurately control and maintain a constant laser power output intensity. Also, the electronic signal to noise ratio of the detector can be improved by refinement of the detector circuit. To enhance the performance of the present invention, the diameter of the focused laser beam 70 can be reduced, but this involves a technical sacrifice that negatively impacts the depth of focus of the laser beam. As the diameter of the laser beam 70 decreases, the accuracy of the device improves, but a corresponding decrease in depth of focus negatively affects the measurement repeatability of the sensor. The lead 40, which has a shallow depth of focus, goes out of the best focus area of the laser beam 70. The preferred beam diameter and depth of focus are about 33.02 micrometers (about 1.3 mils) each.
And ± 0.635 mm (± 25 mil). With the preferred beam diameter and depth of focus, the measurement repeatability in the preferred embodiment is 0.508 micrometer (0.02 mil). This performance can be improved by changing the above three factors.

An embodiment of the invention was tested and found to have a resolution accuracy of 0.7 micron. With 25 microns per mil, or 25 microns per thousandth of an inch, the present invention provides about 25 times higher resolution than the commercial vision systems used today.

A laser-based lead location probe 50 in the lateral plane is shown in FIG. 2A with the detector 74 positioned directly in line with the location of the lead 40, but in an actual preferred embodiment. A condenser lens 62 and a laser diode 60 are used to bend the laser beam 70 to accommodate the physical structure of the component placement machine. For placement accuracy and to minimize vibration and deformation due to acceleration, it is desirable to make the vacuum quill 34 of the component placement machine as short as possible. Therefore, the detection arm 120 of the sensor 50 needs to be as thin or narrow as possible so that it fits between the component 24 and the placement head 30 and does not contact or interfere with the placement head 30. is there. As a result, a number of physical implementations are shown in FIGS. 2 (b) and 5 (a).
Through (c), which allow for an adequate focal length but minimize the physical dimensions of the detection arm 120. As shown in FIG. 2B, a single reflecting mirror 7 that bends the laser beam 70 towards a detector 74.
6 is shown. As shown in FIG. 5A, two reflection mirrors 76 and 761 are used, and the detector 74
Can be located in the lower housing of the sensor 50. As shown in FIG. 5B, the sensor cavity 124
Can be configured to a depth such that the detector 74 can be accurately positioned in the aperture pierced by the sensor 50 to receive the blocked laser beam 70 or the passed laser beam. Finally, as shown in FIG. 5C, the primitive laser beam and the reflected laser beam are flat horizontal mirrors 76 for reflecting the laser beam 70.
Can be placed in an appropriately opened hole in the sensor housing. In a precisely configured component placement machine, the mirror 76 shown in FIG. 5 (c) may be adhered to the bottom of the placement head 30, the alignment sensor 42, or otherwise secured to place the placement head 30 and the component 24.
The entry into the space between and can be minimized.

Coplanarity measurements are important to ensure rejection of electronically defective circuits, with all the bottoms of all leads 40 in the same horizontal plane. A common method of making leads 40 to connect to circuit board 28 involves using a stencil on the top of circuit board 28. Solder paste is applied to the pads 41 of the circuit board 28 via the stencil. After the integrated circuit 24 is properly placed on the solder paste, the circuit board 28 is baked in an oven. When fired, the solder melts and the component 24 becomes electrically and mechanically
It is attached to 8. Typically, the solder paste has a thickness when applied of about 0.2032 mm (about 8 in 1000). When the solder melts, the thickness is about 0.
1016 mm (about 4 mils or 1000 4 inches)
Is. One or more leads 40 are 0.1016 mm (1
If bent more than four thousandths of an inch) above or below the nominal plane of the other leads, they will not be connected to the circuit board 28 via solder paste, resulting in an open circuit.

The preferred embodiment of the invention shown in FIG. 6 is the lateral offset of any bent lead 40, the co-linearity of lead 40 on one side of component chip 24, the thickness of lead 40, and It can be used to test the co-planarity of the leads 40 of a given chip 24. FIG. 6 shows two pairs of laser diodes 60A-60D, both pairs of laser diodes 60A-60D being focused at a single point 800, each pair having its own photodetector 74A-74B. . Each pair of laser diodes is normal
From the first beam 700B, 70 at an angle 802 from β to
Second beams 700A, 700D are generated which are at an angle of 0 (C) and β (802) plus α (804) from the normal. Since the beams are generally located in the same vertical plane, the closest beam from two pairs of laser diodes is β
Form each degree of (802). In such a configuration, each pair of laser beams can detect one lower corner of the cross-sectional area of the lead and one upper corner of the cross-sectional area of the lead.

The preferred angular separation of the beam can be determined in consideration of two factors. Angular direction separation (α (8
04)) provides greater resolution and accuracy in height measurement, but the angular separation allows co-planarity errors of adjacent leads up to 0.127 mm (0.005 inch) in beam. It must be small enough to pass unimpeded through the finest pitch portion, including that which it has. The preferred angle β (802) is 20
The degree is in the range of 5 to 40 degrees. Angle β (80
2) Plus α (804) is optimal at 35 degrees, but can vary from about 25 degrees to 50 degrees. From the point of accuracy, it is preferable that the angle α (804) is as large as possible. However, the angle β (802) is limited by the angle formed by the sides of the trapezoidal lead with the bottom surface of the lead 40. The angle β (802) plus α (804) is limited by the laser beam intercepted by the adjacent lead 40.

The present embodiment is designed to cut off a specific pair of beams.
Calculate the lead height from the measured difference in lateral position along the scan axis. The measurement resolution at these locations directly affects the calculated height resolution. The geometry implemented in the lead height calculation provides a resolution multiplier of about 3. Therefore, the encoder 99 (in the placement machine)
Has a resolution of, for example, 5 microns, the resulting maximum resolution is about 15 μm. Since many QFPs have coplanarity tolerances of 100 μm or less, this 15 μm height resolution is often too coarse. Increase the resolution of the beam intercept position measurement (make it finer)
The interpolation method is used for this.

This embodiment calculates the seating plane and the heights of all leads 40 above the seating plane from the measured lead heights. Most consumers of such devices prefer this method of calculation over the best-fit plane method. In summary, the seating plane is defined as the plane on which the part (under measurement) sits when it is gently placed on a flat surface. A minimum of three leads 40 of component 24 make contact with this plane, and the center of gravity of the component is surrounded by the triangle formed by the three leads 40.

FIG. 6 illustrates each pair of laser beams 700A-7.
00B, 700C-700D photodetectors 74A, 74
B is shown. Although the preferred embodiment has one photodetector for each pair of beams, the present system all uses one photodetector or four detectors for every four beams, i.e. It can be configured with one photodetector for each beam. As the size of the photodetector increases, its response time decreases, which means that the sensor system 50
It has a negative effect on zero. To regulate the surface area of the photodetector and to maintain a reasonable response time,
There is a need to dedicate a single photodetector to the proper shape and size. This adds to the cost of the system over the two detector system. One and four separate photodetectors for each beam also add to the cost of the system.

When the sensor 50 is activated, it is important to be able to detect which laser beam is blocked by the shadow of the lead 40 and eliminate interference from adjacent beams. There are several ways in which the above can be achieved, including the use of different wavelengths of light. Four beams 700A-70
Although 0D exists simultaneously in FIG. 6, the preferred embodiment is to multiplex the beams so that there is only one beam at any one time. Beam 700A ~
Sequential multiplexing of 700D has proven to be the most economical solution.

7A to 7D show the sensor 4
4 shows the trajectory of the lead 40 through the book beams 700A-700D. FIG. 7 (a) shows the leading edge 404 of the beam 4 blocking the component. 7 (a) and 7 (b) show the operation of the leading edge detection system. 7C and 7D show a trailing edge detection system. Leading edge laser light source 60A, 60
B projects two laser beams 700A, 700B onto the leading edge detector 74B. FIG. 7B shows a lead 40 that starts blocking the beam 3 and continues to block the beam 4.
Indicates. At this time, photodetector 74B, which receives rays from beams 3 and 4, receives no laser beam rays.

FIGS. 7C and 7D show the operation of the trailing edge detection system. The trailing edge laser light sources 60C and 60D project two laser beams 700C and 700D onto the trailing edge detector 74A. In FIG. 7C, the beam 2 is avoided,
A trailing edge 405 of component 24 is shown that allows the beam to pass to photodetector 74A. In FIG. 7C, the beam 1
Remains blocked by the lead component 40. FIG. 7 (d) shows the lead 40 that does not block the beam 1 and allows the rays of the beam 1 to strike the photodetector 74A.
Indicates.

FIGS. 7A and 7B show the lower surface 414 of the leading edge 404 of the component detected by the lasers 3 and 4.
Is shown. By detecting the shadow generated by the leading edge 404 and blocking the lasers 3, 4, the sensor 50
Can determine the height of the lead 40. The sensor system 50 determines the height by the mathematical formula: height = d ·
cos (φ) cos (φ + α) / sin (φ) is used. Where d is the lateral distance traveled by the component 24 during the interruption of the beams 3 and 4, and the angle β (802)
And α (804) are set at the factory when the sensor 50 is configured. The preferred angles α (804) and β (8
02) is 15 degrees and 20 degrees, respectively. Thus, the relationship between height and lateral distance traveled is a constant based on the selected angle. By substituting the preferred angle into the above equation, the constant becomes 2.97 (h = 2.97d).

FIG. 8 shows a mode in which the four beams 700A to 700D can be multiplexed. The multiplexing cycle shown is 2.0 microseconds. In this case, each laser beam is energized for 0.5 microseconds. Beam 700A-700
D is sequentially multiplexed such that each beam is energized at the same time that the previous beam is deenergized. In the preferred embodiment, the beams 700A-700D are sequentially activated and de-energized such that the time interval between the same pair of beams is minimized. In other words, it is preferable that the beams 3 and 4 follow sequentially, and that the beams 1 and 2 follow sequentially. This minimizes the time that lead 40 can transition between state changes within a given pair of laser beams. By minimizing this time within a pair of beams, the system can treat the beams as if both were always on. Beam 70
The total cycle time for multiplexing 0A-700D is 2 microseconds. This cycle time has been found to be ideal, with microprocessor chips in the system up to 50 without compromising sensor accuracy.
Be able to exercise at 0 meters / second. The 2 microsecond cycle usually does not present a problem in electronic circuits.
The cycle is fast enough that four beams can always be considered on for most calculations.

As already mentioned, the resolution of the lateral position measurement along the scan axis directly influences the calculated height resolution. Although there are several ways to increase the resolution of the encoder 99, the time interpolation method is the preferred method. FIG. 9 shows a plot of encoder 99 position versus time.
In the preferred embodiment, encoder position and time values are recorded for each state change in one of the laser beams. For example, eight encoder values and eight time values are recorded for eight beam state changes. further,
If more data points are needed to perform the interpolation, the encoder position may occasionally be recorded exclusively to increase the number of data points. The accuracy of the time interpolation method is negatively affected by speed fluctuations. In particular,
FIG. 9 shows an attempt to interpolate the best fit line in the axis of time versus encoder position. In the preferred embodiment, temporal interpolation is used to increase the resolution of the encoder from 10 microns to 1 micron.

In the preferred embodiment, the sensor 50 is such that the positioning head of the component pick and place machine moves the chip through the sensor at a constant speed. The resolution of the encoder is typically about 10 microns. Since there is a ratio of about 3 to 1 between the lateral distance traveled and the height calculation, a 10 micron error in encoder position results in an error of about 30 micron in the height determination. If the encoder resolution is satisfactory (1 micron) then no interpolation is needed.

In the preferred embodiment, the interpolation method has the ability to check for errors. By checking for errors, the height calculated from the interpolated position is converted to the original (r
aw) Compared with the height calculated from the position, if the crossing between the heights exceeds a preset threshold value, the system indicates that the speed is not kept constant. If the tip 24 is not found to be moving at a constant speed, the LED lamp will illuminate, indicating a possible interpolation error.

Although in the preferred embodiment the tip 24 is moving at a constant velocity, the present invention works equally well with a non-constant velocity if the velocity profile of the tip is known. To obtain the maximum speed of the tip 24 through the sensor 50, the tip 24 must be accelerated until it reaches the center of the sensor 50 and then decelerated. After deceleration, the tip 24 can rotate to inspect the leads 50 on the other side of the tip 24.

The phase locked loop method can also be used to increase the resolution of the encoder. Phase locked loop (PLL) 114 can increase the resolution of the encoder by acting as a pulse rate multiplier. Phase locked loop multipliers take the pulse train at a given rate and output the pulse train at some multiple of the input rate. For example, Phase Locked Loop 114
Receives encoder pulses from encoder 99 and increases the resolution by producing ten times the number of these pulses in the same time. However, the phase locked loop method has certain limitations that make it less preferred compared to the temporal interpolation method.
As is well known, the phase locked loop 114
Potentially inconsistent velocity profile that does not work well over a wide range of frequencies. Therefore, the time interpolation method has been found to be the most accurate, cheaper, and least restrictive in increasing the resolution of the encoder.

In addition to LED error lighting for interpolation, the preferred embodiment has LED error lighting for software and hardware errors.

10 and 11 show the basic electronic circuitry of the preferred four laser beam embodiment. One multiplexing controller 200 assists in multiplexing the laser beams 700A-700D and produces synchronization signals from the photodetectors 74A, 74B.

When the beam energizing system is activated, signal A
Are sampled into the register as the laser power supply is switched from laser A to laser B. Similarly, laser B is sampled into a register as multiplexer 202 switches from laser beam B to laser C. This sampling process proceeds sequentially through laser D and then returns to laser A. The signal is most stable after a particular laser has been turned on for a period of time, so each laser is sampled at the moment before it is turned off, ie de-energized.

Although one electrical signal is received from each photodetector 74A, 74B, the system maintains a separate threshold voltage, or reference voltage, for each laser beam. Preferably, four comparators 220 correspond to each of the four reference voltages. Four reference voltages are provided to the comparator 220, and each reference voltage can be independently adjusted to compensate for dirty lenses or long-term laser drift. Each comparator compares the detector output voltage to a particular reference voltage to provide a digital output measurement for a particular beam break. Since there are four laser beams 700A-700D, there are four corresponding binary outputs.

Switch 204 is used to multiplex the four binary output values to utilize a single cable to carry the beam data from the amplification electronics to the data processor 230. The multiplexed signal is then demultiplexed by another switch 206 under the control of the multiplexing controller 200 and using the synchronization signal before processing the data. The second cable carries the synchronization signal to the data processor 230.

Data processing hardware and software 230 performs interpolation and height calculations. Further, the data is processed to determine if there are any errors. The data processor 230 can determine the width of the lead by taking the difference between the leading and trailing edges.
The thickness and twist of the lead 40 can be calculated from the lateral position and height of each corner. The data processing hardware and software 230 can calculate the lead thickness by taking the height difference between the top edge on either side and the bottom edge on either side. The data processing hardware and software 230 can also calculate the twist of the lead 40 based on the detected heights of the four corners.

Those skilled in the art will recognize that the system of the present invention allows for many alternative computations. The available memory associated with processor 230 enables a downloadable component library that defines the lead width and spacing of all components 24 to be analyzed and detected by sensor 50. It is possible to detect and analyze only the leading edge data and use the component library information to generate the centerline and proper position for each lead 40. For more accurate measurements, trailing edge data can be used for both lateral position and height analysis.

Although not shown in FIG. 6 or subsequent figures, it will be apparent to those skilled in the art that the beams of multiple beam systems need not physically intersect. It can be seen that the beams are offset from each other by a known amount of deviation and that the rays from laser beam A are not likely to diverge into detector B. Another way to make sure that there is no interference between the beams is to use multiple wavelengths of the beam and insert an appropriate optical bandpass filter in front of the detector so that only the beam from one source To be detected. Other means such as baffles (not shown) can be placed between the beams.

Instead of a single focus light beam 70, a stripe of laser light may be used. The stripes are generated from a single laser light source 60 using a source lens and stripe projection optics. The source lens and the stripe projection optical element can be combined to form a single optical element.

FIG. 12 shows a secondary laser sensor 50 'having a configuration generally similar to that shown and described in connection with FIGS. 2 and 3, which provides maximum accuracy and resolution. It may be advantageously used with the present invention 50 to more accurately position the placement head 30 of the component placement machine and to more accurately identify the position of the head 40.

A higher resolution is achieved in detection,
Some other problems may occur as the finer pitch components 24 are constructed. The component placement machine from the encoder 99, the rail frame 32 on which the placement head 30 is carried, the placement head 30 and the vacuum quill 36 have a definite rigidity. That is,
They are not infinitely stiff. The associated quill position encoder 99 includes a placement head 30 and a vacuum quill 32.
Head 30 by reading encoder 99 by relatively accurately monitoring the position of
Can be accurately positioned. In the dynamic case, further problems occur, especially while the placement head 30 is being accelerated, and become more pronounced during movement as the pitch becomes increasingly finer. Since the purpose of the sensor 50 of the present invention is to measure where the component 24 and its leads 40 are with respect to the quill 34, the placement head 3
Accurate positioning of 0 and the associated quill 34 requires a fine resolution of the order of 1 micron.

The accurate positioning of the placement head 30 is performed by the optical target 15 fixed to the bottom of the placement head 30.
This can be achieved by using a similar sensor 50 'as shown in FIG. 12 (a) configured to detect zero. As shown in FIGS. 12B and 12C, the optical target 150 alternates the transparent portions 152 and the opaque portions 154 so that the processing mechanism 50 ′ positions the placement head 30 with sub-micron accuracy. Can be detected,
This is the feature of the present invention. An alternative embodiment configured to not interfere with the movement of placement head 30 is shown in FIG. In order to construct the embodiment shown in FIG. 12 (c), the optical target 150 is composed of chrome 154 on the glass 152 so that the laser beam 7 from the reflective chrome surface 154.
The placement head 30 detects the 0'reflection by the detector 74 '.
And associated quill 34, component 24 and component lead 4
It is preferable to make a very accurate determination of 0.

By using any of the various embodiments of the present invention described above, the lateral offset position and height of various leads 40 on an IC or other semiconductor device 24 can be determined very accurately and accurately. it can. FIG.
Refer to the lead 40 in the component placement machine.
It is possible to use a single sensor 50 capable of inspecting and determining all four sets 1 through 4 in turn. Part 24 stopped, 4
Of course, this is the most time consuming process since it is necessary to pass the sensor 50 once. Similarly, component 24 is passed through sensor 50 to inspect the first set of leads, side 2 of FIG. 13 (b), and the opposite set of leads 40, FIG. 13 (a). 2) two sensors 50 for moving the side 1 by a predetermined distance (c) according to the parts library, along the trajectory 170 shown in FIG. Can be used. Next, the part 24 is rotated 90 degrees, and
Through the individual sensor system 50, the leads 40 on the sides 3 and 4 shown in FIG.
Similarly, the three sensor system 5 shown in FIG.
0 can be used, the first sensor 50 inspects the set of leads 40 on the side 1, the second sensor 50 inspects the set of leads 40 on the side 2, followed by the placement head 30 90. The second sensor 50 rotates the lead 40 of the side portion 4
Allows the third sensor 50 to inspect the lead 40 on the side 3. Obviously, it is also possible to use four independent sensors. For a merry-go-round placement system or a turret placement system, one or more stations can contain one or more sensor systems 50. For a fast Cartesian coordinate placement system, one or more laser sensors may be spaced in the same housing or enclosure at 101.6 to 152.4 mm (4 to 6 inches) to accommodate the largest size component 24. (Enclosure).

A basic laser sensor capable of detecting bent leads 40 of integrated circuit chip 24 at sub-micron resolution has been disclosed in connection with FIG. Although various alternative embodiments have been shown and described using the basic laser sensor 50 as a building block, these, when combined, allow the bottom surface of the leads 40 of a semiconductor device to be precisely 1 micron or less horizontally and vertically. With horizontal and vertical resolution, which allows the parts to be positioned within the accuracy and resolution of the part placement machine. Alternative embodiments have been described in which co-linearity, co-planarity, lead twist and lead width can be calculated. Although a number of alternative embodiments have been disclosed, it will be apparent to those skilled in the art that other modifications and variations are possible.
Each of the above modifications and variations within the scope of the teaching of the present invention are intended to fall within the intended scope of the present invention as set forth in the claims.

[Brief description of drawings]

FIG. 1 illustrates a component placement system using the present invention, where a component is picked up from a source tray by a placement head and associated vacuum quill, the component being located at the correct angular orientation on the vacuum quill. FIG. 5 shows how one or more sensors constructed according to claim 1 are passed through and then the component is accurately placed by a component placement machine on a circuit board adapted to receive the component.

2 (a) to (c) show a preferred embodiment of the present invention, FIG. 2 (a) is a diagram showing the basic elements of the present invention, and FIG.
And (c) are a side view and a top view of a mechanical structure according to an embodiment of the present invention.

Figure 3 (a) and (b), (a) shows a semiconductor package with the leads bent in the lateral plane, (b) of Figure 3 only partly on the corresponding circuit connection pad. The figure which shows the arrangement | positioning of the lead overlapped.

FIG. 4 is a schematic diagram of a sensor electronic circuit.

FIG. 5 (a)-(c), each configured to minimize physical interference of the components of the invention between the placement head of the component placement machine and the component whose lead is being analyzed. FIG. 9 illustrates an alternate physical embodiment of a keyed entry.

FIG. 6 shows a four-beam configuration that can be used to analyze lateral position, co-linearity and co-planarity at high resolution.

7 (a), (b), (c) and (d) show a four beam system for analyzing leads below the horizontal reference plane, (a) and (b) showing leading edge detection, FIG. (C) And (d) is a figure which shows a trailing edge detection.

FIG. 8 is a diagram showing multiplexing of a four beam system.

FIG. 9 is a graph showing interpolation of the best fit line of encoder position over time.

FIG. 10 is a diagram of the basic electronics for a four beam system, showing the electronics for the laser and photodetector.

FIG. 11 is a diagram of the basic electronic circuitry for a four beam system showing the interface to data processing hardware and software.

FIG. 12 (a)-(c) illustrates the use of the present invention to dynamically determine the position of the placement head of a component placement machine, using this embodiment of the invention for placement. The exact position of both the optical target of the head and the lead of the component is detected at the same time and the focused laser beam passes through the optical target as shown in (a), or
Alternatively, as shown in (c), it is possible to reflect from the optical target.

FIG. 13A is a diagram showing a semiconductor chip having four sides by, for example, a quad pack integrated circuit, and FIGS. 13B and 13C are diagrams when a plurality of sensors configured according to the present invention are used. The figure which shows the orbit taken by the component placement head.

[Explanation of symbols]

 24: Surface Mount Electronic Components 28: Circuit Board 30: Placement Head 33: Source Tray 34: Vacuum Quill 40: Lead 41: Circuit Pad 42: Non-contact Laser-based Matching Sensor 50: Sensor System 60: Laser Light Source 62: Collection Optical lens 66: Focus 70: Laser beam 74: Detector 76: Mirror

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor John P. Conicec Minnesota, USA 55409, Minneapolis, Stevens Avenue South 4632 (72) Inventor Steven Kay Case, Minnesota 55416, Sent Lewis Park, Inglewood Avenue 2829 (72) Inventor Timothy A. Skanes, Minnesota, USA 55421, Colombia Heights, East Upland Crest 4837.

Claims (51)

[Claims] 1. A multi-element device of the type in which each element is nominally arranged in a single horizontal plane for determining the height of each element of the multi-element device moving relative to a multi-beam sensor. In the beam sensor, at least one light ray providing means for providing a light ray, and condensing the light ray from the light ray providing means into a first beam of light rays having a focal point at an intersection of a horizontal plane and a vertical reference plane,
First forming an angle β of 5 to 25 degrees with respect to the vertical reference plane
And a light beam from the light beam providing unit, into a second beam of light beams having a focal point at an intersection of a horizontal plane and a vertical reference plane,
Second converging means forming an angle with respect to the vertical reference plane that is greater than the angle β and shadows that are located in the trajectories of each beam and that are generated by the interruption of the first beam by the edges of the elements of the multi-element device. A shadow detecting means for detecting a shadow generated by blocking the second beam by the edge of the element, and a plurality of elements as each shadow generated by the edge of the element of the multi-element device is detected by the shadow detecting means. Data receiving means for receiving data indicating an instantaneous lateral position of the device, connected to the shadow detecting means and the data receiving means,
By calculating the height of the edge of the element above the horizontal plane,
A multi-beam sensor, the height calculation being based on the instantaneous lateral position of the multi-element device as each shadow is generated by the edge of the element. 2. The multi-beam sensor according to claim 1, wherein the light beam providing means has two laser diodes. 3. The multi-beam sensor according to claim 1, wherein the shadow detecting means has two photodetectors. 4. The logic means determines a lateral distance traveled by an edge of the element between the first beam and the second beam,
The multi-beam sensor according to claim 1, wherein the height of the edge of the element from the horizontal reference plane is calculated by multiplying by a constant. 5. The logic means comprises interpolation means for interpolating between discrete recorded encoder positions and time values to obtain the instantaneous lateral position of the multi-element device. The multi-beam sensor according to claim 1. 6. The multi-beam sensor according to claim 5, wherein the interpolation means has a phase locked loop. 7. A multi-beam sensor according to claim 5, wherein the interpolating means comprises a timing device for performing temporal interpolation. 8. A best fit line is defined as one in which the sum of squared deviations of the lead height is minimized, and wherein said logic means deviates the height of the edge of the element from the best fit line. The multi-beam sensor according to claim 1, wherein 9. A seating plane is defined as a surface set by the contact points of three or more leads that support the component when the component is located on top of a flat surface, and the logic means is greater than the seating plane. The multi-beam sensor according to claim 1, wherein the height of the upper element edge is calculated. 10. The multi-beam sensor according to claim 1, wherein the light beam providing means further includes means for switching the beam so that the light beam is alternately turned on and off. 11. A multi-beam sensor for determining the height of each element of a multi-element device of the type wherein each element is nominally arranged in a single horizontal plane, said multi-element device comprising: A multi-beam sensor, wherein the leading edge detection device comprises at least one leading edge light source and a pair of leading rays of light from the leading edge light source. Focusing means for focusing the edge beam so that it intersects in a horizontal plane of the element, the focusing means wherein the beams of the pair of leading edges are blocked by the leading edges of the element; A leading edge detection device having at least one leading edge detection means for detecting a shadow of the leading edge generated by blocking the pair of leading edge beams by the following: a trailing edge detection device; A condensing means for converging a pair of trailing edge beams of light rays from the leading edge light source and the trailing edge light source so as to intersect in a horizontal plane of the element, wherein the pair of trailing edge beams are Focusing means blocked by a trailing edge of the element and at least one trailing edge detection for detecting a shadow of the trailing edge caused by blocking the pair of trailing edge beams by the trailing edge of the element. Means, and a trailing edge detecting device having, and connected to the leading edge detecting device and the trailing edge detecting device, the height of the leading edge and the trailing edge at the center height above the horizontal plane of the element Logic means for determining a center height which is an average of height and means for calculating a leading edge height of said element based on data received from said leading edge detecting device; and from said trailing edge detecting device. Based on the received data, the height of the trailing edge of the element Multibeam sensor, characterized in that it comprises a logic means and means for calculation. 12. How well the collinearity fits the best-fit line defined as the line where the height of the leads on one side of the integrated circuit is the smallest of the sum of the squared deviations of the leads. 12. The multi-beam sensor of claim 11, wherein said logical means further comprises means for determining element co-linearity based on center height. 13. The sheeting plane is a plane formed by contact points of three or more leads on a flat surface, and the logic means further comprises means for determining a sheeting plane of elements of a multi-element device. The multi-beam sensor according to claim 11. 14. A multi-beam sensor according to claim 11, wherein both said light sources comprise laser diodes and means for controlling the activation and deactivation of the leading edge beam and the trailing edge beam. . 15. The relative velocity of motion of the multi-element device is constant, and the time it takes for the multi-element device to move between the beams of light rays is multiplied by the velocity to obtain between the beams of light rays. 12. The multi-beam sensor of claim 11, further comprising means for determining the relative distance traveled by the multi-element device. 16. The method of claim 11, wherein the instantaneous lateral position is the position of the multi-element device at the time of detection, and further comprising receiving means for receiving the instantaneous lateral position of the multi-element device. Multi-beam sensor. 17. The receiving means receives lateral position information from an encoder, and the receiving means further comprises:
17. The multi-beam sensor of claim 16 including means for interpolating lateral position of a multi-element device to increase encoder resolution. 18. The multi-beam sensor according to claim 17, wherein the means for interpolating performs linear interpolation. 19. The multi-beam sensor of claim 16 wherein said receiving means further comprises a phase locked loop to increase encoder resolution. 20. A four-beam sensor for determining the height of each element of a multi-element device of the type where each element is nominally arranged in a single plane, wherein the multi-element device is a four-beam sensor. In the four-beam sensor that moves relative to the first beam of light, a first beam of light from the first beam of light, a second beam of light, a third beam of light from a fourth beam sensor,
First focusing means for focusing on the beam of
The focus of the beam is at the intersection of the horizontal plane and the vertical reference plane, and the first beam forms an angle β of 5 to 25 degrees with respect to the vertical reference plane; A ray of light from the means for providing the ray of light
Second condensing means for condensing into a beam of
Second beam converging means forming an angle larger than the angle β with respect to the vertical reference plane, the focal point of the beam of the second beam being at the intersection of the horizontal plane and the vertical reference plane, and the third light ray. The ray from the means for providing a third ray
Third converging means for condensing into a beam of
Is at the intersection of the horizontal plane and the vertical reference plane, and the third beam forms a minus β angle with the vertical reference plane so that the third beam forms an angle of 2β with the first beam. Focusing means and a fourth focusing means for focusing the light beam from the means for providing the fourth light beam into a fourth beam, the focus of the fourth beam being a horizontal plane and a vertical reference plane. A fourth concentrator located at the intersection of the first beam and the fourth beam forming an angle greater than 2β with the first beam; First detecting means for detecting an edge of a shadow generated by blocking the first beam by the edge of the element and a shadow generated by blocking the second beam by the edge of the element; Located in the orbit of the fourth beam and of the multi-element device Second detection means for detecting a shadow caused by the non-interruption of the third beam by the edge of the element and a shadow caused by the non-interruption of the fourth beam by the edge of the element; Data receiving means for receiving data indicative of an instantaneous lateral position of the multi-element device as edges of each shadow produced by the edges of the element are detected by the first and second detecting means; 1, the second detecting means and the data receiving means, and calculating the height of the edge of the element from a horizontal plane by calculating the height, the edge of each shadow being the edge of the element. And a logic means for performing based on the instantaneous lateral position of the multi-element device as generated by. 21. The four beam sensor of claim 20, including means for multiplexing light beams. 22. The four beam sensor of claim 20, including interpolating means for more accurately determining the lateral position of the multi-element device. 23. The four beam sensor of claim 20, wherein a multi-element device is in motion and the four beam sensor is stationary. 24. The four beam sensor of claim 20, wherein the motion profile of the multi-element device is known prior to calculation. 25. The four-beam sensor according to claim 20, wherein the first and second detecting means detect only the bottom edge portion of the lead. 26. The sheeting plane is a plane formed by contact points of three or more leads on a flat surface,
The four beam sensor of claim 20, further comprising means for determining a seating plane of the element. 27. The four beam sensor of claim 20, wherein the twist of the lead is a measure of the angular placement of the lead and the logic means further comprises means for calculating the twist of the lead. 28. The width of the lead is defined as the distance from one edge of the lead to the other edge, and the logic means further comprises means for calculating the width of the lead. The four-beam sensor described in. 29. The thickness of the lead is the distance from the top edge to the bottom edge of the lead and the logic means further comprises means for calculating the lead thickness.
4 beam sensor of 0. 30. How well the lead height on one side of the integrated circuit fits the best-fitting line where colinearity is defined as the line where the sum of the squared lead deviations is minimized. 21. The four beam sensor of claim 20, wherein the logic means further comprises means for determining colinearity of the set of leads. 31. Coplanarity is a measure of how well the height of an integrated circuit lead fits in the plane defined by the nominal position of the lead, said logic means further comprising: The four beam sensor of claim 20, including means for determining coplanarity of the leads. 32. The four-beam sensor according to claim 20, wherein the first and second detecting means detect only the upper edge portion of the lead. 33. A method of detecting the height of a bottom surface of a lead of a multi-element device from a horizontal reference plane and determining whether the height of the lead is within an allowed height deviation. A vertical reference plane is defined, a horizontal reference plane is defined, a first laser beam is generated, and the first laser beam is condensed at an intersection of the horizontal reference plane and the vertical reference plane to form a vertical reference plane. 5 to 3
An angle β of 0 degree is formed with respect to a vertical reference plane by forming a second laser beam and converging the second laser beam at the intersection of the horizontal reference plane and the vertical reference plane of the second laser beam. Is formed by adding an angle α between 7 and 23 degrees, and the multi-lead device is horizontally moved so that the lower surface of each lead passes through each of the two laser beams. Detected a change in the first state caused by the interruption of one of the laser beams of the two books, which was not included in the change of the first state, but was caused by the interruption of the laser beam by the edge of the lead. The change in the second state is detected, the lateral distance that the lead moves from the position where the lead causes the first state change to the position where the lead causes the second state change is measured, and the lateral distance moved And cos (β) and cos ( + Α)
Calculate the height of the edge of the lead above or below the horizontal reference plane by multiplying by and dividing the sum by sin (β) to determine the allowable height deviation and calculate the calculated lead The height of the edge of the lead is compared with an allowable height excursion, and if the lead edge height is greater than the allowable height excursion, each step of generating a rejection electrical signal is characterized by And how to. 34. The method further comprising calculating a center height of the lead by averaging the height calculated for the leading edge of the lead and the height calculated for the trailing edge of the lead. 34. The method according to claim 33. 35. The method of claim 33, wherein the step of measuring the lateral distance traveled comprises time interpolation. 36. The method of claim 33, further comprising the step of multiplexing the two laser beams. 37. Colinearity is a measure of how well the height of the leads on one side of the integrated circuit fits the best-fitting line, and the colinearity of a set of leads is calculated. 34. The method of claim 33, further comprising steps. 38. The method of claim 33, wherein the seating plane is the plane formed by the contact points of three or more leads on the flat surface and further comprising the step of positioning the seating plane. 39. The method of claim 33, further comprising receiving the instantaneous lateral position of the multi-lead device. 40. The method of claim 33, further comprising detecting a third state change and a fourth state change caused by a trailing edge that does not block the laser beam. 41. A method for determining the height of each element of a multi-element device of the type in which each element is nominally arranged in a single horizontal plane, wherein the multi-element device is in relative motion. A step of detecting a leading edge, the step of generating a leading edge ray, the pair of leading edges of rays from the leading edge ray, wherein each element has a leading edge and a trailing edge, Focusing the beam in the horizontal plane of the element, the pair of leading edge beams being intercepted by the leading edges of the element, and confirming the leading edge shadows caused by the interception of the pair of leading edge beams by the element leading edges. Detecting a leading edge, and detecting a trailing edge, wherein a trailing edge ray is generated and a pair of trailing edge beams of rays from a trailing edge ray are included in the element. Focus on the horizontal plane of A trailing edge beam is intercepted by the trailing edge of the element, the trailing edge comprising the steps of confirming the shadow of the trailing edge caused by the interception of the pair of trailing edge beams by the trailing edge of the element. Using the step of detecting, the data received by detecting the leading edge, and the data received by detecting the trailing edge,
Determining the center height which is the average of the leading edge height and trailing edge height at the center height of the element above a horizontal plane, the method being based on data received from detecting the leading edge. Determining a center height, which includes calculating a height of a leading edge of the element, and calculating a height of a trailing edge of the element based on data received from detecting a trailing edge. A method comprising :. 42. Colinearity is a measure of how well the height of the leads on one side of an integrated circuit fits the best fit line, and further based on the center height of the element. 42. The method of claim 41 including the step of calculating the co-linearity of the elements. 43. The seating plane is a plane formed by contact points of three or more leads on a flat surface,
42. The method of claim 41, further comprising calculating a seating plane for the element. 44. The method of claim 41, further comprising the step of multiplexing four beams. 45. Measuring the distance traveled by the multi-element device between the beams of light by multiplying the speed of the multi-element device moved at the time the multi-element device moved between the beams of light. 42. The method of claim 41, wherein: 46. The method of claim 41, further comprising receiving the instantaneous lateral position of the multi-element device. 47. The method of claim 46, wherein the instantaneous lateral position is received from the encoder and further comprising the step of interpolating the position of the multi-element device to increase the resolution of the encoder. 48. The method of claim 47, wherein the interpolation is linear interpolation. 49. The method of claim 46, further comprising the step of generating additional lateral position data using a phase locked loop. 50. The height of the leading edge is determined by the following mathematical expression: The height of the leading edge = Dcos (β) cos (β + α) / si
n (β) where D is the lateral distance traveled between the leading edge beams, the angle of β is equal to the angle between the leading edge beam and the vertical reference plane, and the angle of α is the angle between the leading edge beams. 42. The method of claim 41, which is equal to. 51. The method of claim 41, further comprising the step of determining a block of a beam of light rays generated by a bumper located in the multi-element device.