U.S. patent number RE34,615 [Application Number 07/829,727] was granted by the patent office on 1994-05-24 for video probe aligning of object to be acted upon.
This patent grant is currently assigned to MPM Corporation. Invention is credited to Gary T. Freeman.
United States Patent |
RE34,615 |
Freeman |
May 24, 1994 |
**Please see images for:
( Certificate of Correction ) ** |
Video probe aligning of object to be acted upon
Abstract
An aligning system includes an object to be acted upon at
predetermined locations, such as a circuit board to receive solder
paste. There is a device, such as a stencil, characterized by the
pattern for acting upon the object. A video probe is arranged to
look at both the device and the object for providing image signals
representative of both. A comparator compares the image signals to
provide an error signal representative of misalignment between the
device and object. A positioner responsive to the error signal
relatively positions the device and object to reduce the error. An
operator causes the device to operate upon the object at the
predetermined locations.
Inventors: |
Freeman; Gary T. (Beverly,
MA) |
Assignee: |
MPM Corporation (Franklin,
MA)
|
Family
ID: |
25672625 |
Appl.
No.: |
07/829,727 |
Filed: |
January 31, 1992 |
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
Reissue of: |
116490 |
Nov 2, 1987 |
04924304 |
May 8, 1990 |
|
|
Current U.S.
Class: |
348/95; 348/129;
348/87 |
Current CPC
Class: |
G03F
9/00 (20130101); H05K 13/0465 (20130101); H05K
3/1216 (20130101) |
Current International
Class: |
G03F
9/00 (20060101); H05K 13/04 (20060101); H05K
3/12 (20060101); H04N 007/18 () |
Field of
Search: |
;358/101,106,107,93 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Britton; Howard W.
Attorney, Agent or Firm: Fish & Richardson
Claims
What is claimed is:
1. Aligning apparatus comprising,
an object to be acted upon at predetermined locations defining a
pattern,
a device characterized by said pattern for acting upon said
object,
a video probe arranged to look at both said device and said object
for providing image signals representative of both,
means for comparing both image signals to obtain an error signal
representative of misalignment between the device and object,
means responsive to said error signal for .[.relatively
positioning.]. .Iadd.translating and rotating .Iaddend.said device
.[.and said object.]. to reduce said error,
and means for operating upon said object with said device upon said
predetermined locations.
2. Aligning apparatus is accordance with claim 1 wherein said video
probe comprises a mirror tube having a mirror at one end
intersecting the mirror tube axis at an angle of substantially
45.degree. and exposed through an opening in the wall of said
mirror tube,
a video camera located at the other end of said mirror tube for
receiving images reflected from said mirror,
said mirror tube being rotatable about the mirror axis between at
least a first position exposing said mirror to said device and a
second position exposing said mirror to said object.
3. Aligning apparatus .[.in accordance with claim 1 and further.].
comprising,
.Iadd.an object to be acted upon at predetermined locations
defining a pattern,
a device characterized by said pattern for acting upon said
object,
a video probe arranged to look at said device and said object for
providing image signals representative of both,
means for comparing both image signals to obtain an error signal
representative of misalignment between the device and object,
means responsive to said error signal for relatively positioning
said device and said object to reduce said error,
and means for operating upon said object with said device upon said
predetermined locations, .Iaddend.
a second of said video probes spaced from the first-mentioned video
probe arranged to look at both said device and said object for
providing image signals representative of both at areas thereof
different from the areas viewed by said first-mentioned video
probe,
and means for relatively displacing said video probes while in
fixed relative relation and one of said object and device
predetermined incremental distances in first and second orthogonal
directions and a rotational directional to provide a reference
image signal characterizing said pattern.
4. Aligning apparatus is accordance with claim 3 and further
comprising:
for each of said video probes video probe support means for
selectively moving the associated video probe between a first
position located between said device and said object and a second
position outside the region between said object and said
device,
and means for locating the associated video probe in said first
position before acting upon said object and then displacing said
video probe to said second position after relatively positioning
said device and said object to reduce said error and operating upon
said object with said device upon said predetermined locations with
the associated video probe in said second position.
5. Aligning apparatus in accordance with claim 4 wherein each of
said video probe support means comprises,
a movable probe support carrying said video probe,
a fixed base having vertical walls each formed with front and rear
slots with each slot having a horizontal leading portion and a
depending angled trailing portion,
said movable probe support having elements for riding in said slots
to allow said video probe to move between said first position with
said video probe extended and said second position with said video
probe retracted.
6. Aligning apparatus in accordance with claim .[.8.]. .Iadd.5
.Iaddend.wherein said movable probe support comprises a main clamp
supporting said video probe substantially at its center of gravity
and further comprising,
ball-and-cone pieces and reference balls,
said fixed base comprising a vertical bracket carrying one of said
reference balls and said ball-and-cone pieces,
said main clamp carrying the other of said reference balls and
ball-and-cone pieces arranged so that engagement of said
ball-and-cone pieces with mating reference balls defines a
reference position of said video probe.
7. Aligning apparatus in accordance with claim 1 and further
comprising,
video probe support means for selectively moving said video probe
between a first position located between said device and said
object and a second position outside the region between said object
and said device,
and means for locating said video probe in said first position
before acting upon said object and then displacing said video probe
to said second position after relatively positioning said device
and said object to reduce said error and operating upon said object
with said device upon said predetermined locations with said video
probe in said second position.
8. Aligning apparatus in accordance with claim 7 wherein said video
probe comprises a mirror tube having a mirror at one end
intersecting the mirror tube axis at an angle of substantially
45.degree. and exposed through an opening in the wall of said
mirror tube,
a video camera located at the other end of said mirror tube for
receiving images reflected from said mirror,
said mirror tube being rotatable about the mirror axis between at
least a first position exposing said mirror to said device and a
second position exposing said mirror to said object.
9. Aligning apparatus in accordance with claim 7 wherein each of
said video probes comprises a mirror tube having a mirror at one
end intersecting the mirror tube axis at an angle of substantially
45.degree. and exposed through an opening in the wall of said
mirror tube,
a video camera located at the other end of said mirror tube for
receiving images reflected from said mirror,
said mirror tube being rotatable about the mirror axis between at
least a first position exposing said mirror to said device and a
second position exposing said mirror to said object.
10. Aligning apparatus in accordance with claim 7 wherein said
video probe support means comprises,
a movable probe support carrying said video probe,
a fixed base having vertical walls each formed with front and rear
slots with each slot having a horizontal leading portion and a
depending angled trailing portion,
said movable probe support having elements for riding in said slots
to allow said video probe to move between said first position with
said video probe extended and said second position with said video
probe retracted.
11. Aligning apparatus in accordance with claim 2 wherein said
video probe comprises a mirror tube having a mirror at one end
intersecting the mirror tube axis at an angle of substantially
45.degree. and exposed through an opening in the wall of said
mirror tube,
a video camera located at the other end of said mirror tube for
receiving images reflected from said mirror,
said mirror tube being rotatable about the mirror axis between at
least a first position exposing said mirror to said device and a
second position exposing said mirror to said object.
12. A method of aligning an object to be acted upon at
predetermined locations defining a pattern with a device
characterized by said pattern for acting upon said object which
method includes the steps of,
positioning a video probe to look at first one of said device and
said object and then the other for providing image signals
representative of both,
comparing both said image signals to provide an error signal
representative of misalignment between said device and object,
.[.relatively positioning.]. .Iadd.translating and rotating
.Iaddend.said device .[.and said object.]. while sensing said error
signal to reduce said error,
and operating upon said object with said device upon said
predetermined locations with said object and said device then being
in alignment.
13. A method in accordance with claim 12 and further including the
steps of moving said video probe between an inside position located
between said device and said object,
then locating said video probe in an outside position outside the
region between said object and said device,
and operating upon said object with said device only with said
video probe in said outside position.
14. A method .[.in accordance with claim 13 and further including
the steps of.]. .Iadd.of aligning an object to be acted upon at
predetermined locations defining a pattern with a device
characterized by said pattern for acting upon said object which
method includes the steps of,
positioning a video probe to look at first one of said device and
said object and then the other for providing image signals
representative of both,
comparing both said image signals to provide an error signal
representative of misalignment between said device and object,
relatively positioning said device and said object while sensing
said error signal to reduce said error,
operating upon said object with said device upon said predetermined
locations with said object and said device then being in alignment,
.Iaddend.
looking first at one of said device and said object and then the
other with a second video probe spaced from the first-mentioned
video probe to provide a second set of image signals representative
of said device and said object different from the image signals
provided by said first-mentioned video and the areas representative
thereof,
and relatively displacing said video probes while in fixed relative
relation with respect to one of said object and device
predetermined incremental distances in first and second orthogonal
directions and a rotational direction to provide a reference image
signal characterizing said pattern.
Description
.Iadd.A computer program listing is set forth in a microfiche
appendix. .Iaddend.
The present invention relates in general to aligning and more
particularly concerns novel apparatus and techniques for aligning
an object to be acted upon at particular locations with a device
that performs the specific operations on the specified points. A
specific embodiment of the invention facilitates aligning circuit
boards for receiving surface mounted components with a stencil for
applying solder paste to a pattern on the circuit board while
facilitating rapid reorientation for applying solder paste to a
different pattern.
A typical screen printer is the commercially available ASP-24
automated screen printer available from MPM Corp., 71 West Street,
Medfield, Mass. 02052 incorporated herein by reference. This screen
printer includes a replaceable screen for printing solder paste on
a footprint of pads on a surface mount circuit board. This system
may also deposit other materials, such as epoxy, polymer, cermet
and most other screen printable materials. This system includes a
programmable controller for positioning each circuit board beneath
the screen or stencil above that deposits the screen printable
materials.
Prior art screen printers with vision require printing solder paste
on the circuit board, driving a table out beneath the cameras and
the cameras look at the solder paste. The solder paste in then
wiped off, the board enters beneath the stencil, and the board is
printed a second time, and reciprocates back out for examination by
the camera.
A prior art screen printer includes a camera that looks between the
board and the screen, but it does not look at an image on the
screen, only at the board. The camera on that device does not
reciprocate. It uses a fixed mounted camera that just looks down at
the board, and it requires that the board be outside of the stencil
which comes down upon it by a certain amount such that the camera
does not get in the way.
It is an important object of this invention to provide improved
apparatus and techniques for optically aligning an object to be
acted upon at one or more points with a device for repeatedly
acting upon like objects at like points.
According to the invention, at least one, and preferably two,
movable video probes are located between the acting device, such as
the screen or stencil, typically above, and the object to be acted
upon, such as a circuit board, typically below. The one or two
probes, first look at and locate a pattern, such as on the object
device, such as a circuit board, store information on features in
this pattern, such as x, y and .theta. information on key features
in the pattern. Associated data processing apparatus processes and
stores this information. The one or both video probes then rotate
downward while still in the region between the object to be acted
upon and the device to act, examine the other pattern such as on
the acting device, such as the screen or stencil, for a matching
pattern and relatively position the object and device so that the
two patterns are in alignment. The one or both video probes then
retract, and the device, such as the stencil or screen engages the
object, such as the circuit board, to deposit solder paste or other
screen printable material on selected points in registration with
corresponding points on the device, such as the screen or stencil.
Once a specific pattern is learned, the invention may repeatedly
relatively position by looking at only the object to be acted
upon.
According to a specific aspect of the invention, a positionable
base supports a video probe.[.,.]. .Iadd.. .Iaddend.This base is
formed with at least one and preferably two slots having a
generally horizontal leaning portion with a depending angled
trailing portion for guiding elements that carry the video probe as
it moves between the extended position at the leading end of the
slots, with the video probe between object and device, and the
retracted position at the trailing end of the slot, with the video
probe retracted from the region between object and device.
Numerous other features, objects and advantages of the invention
will become apparent from the following specification when read in
connection with the accompanying drawings in which:
FIG. 1 is a perspective view of a fully automated screen printer
with vision system according to the invention;
FIGS. 2 and 3 are perspective views of the video probe assembly in
retracted and extended positions, respectively;
FIG. 4 is an exploded view of the video probe and portions of the
supporting assembly;
FIG. 5 is an exploded view of the video probe support assembly;
and
FIG. 6 is a diagram illustrating the relationship among target
points and the center of rotation helpful in understanding
principles of the invention.
With reference now to the drawing and more particularly FIG. 1
thereof, there is shown a perspective view of an exemplary
embodiment of the invention comprising the commercially available
ASP-24 fully automated screen printer available from MPM Corp. with
movable video probes according to the invention. The apparatus
includes a base 11 for supporting a circuit board to be printed
upon, left and right video probe assemblies 12 and 13, a controller
14 and video monitor assembly 15. The system also includes a
positionable screen/stencil support assembly that may be moved to
align the screen or stencil with the circuit board to be printed
upon. A lower platform 16 supports various packages for containing
hardware used in the system.
The base components added to the ASP-24 automated screen printer
include the two video probe assemblies 12 and 13, a fiberoptic
light source for illuminating the screen and board during pattern
recognition, the trackball teaching module 14, and a vision
processor unit comprising a commercially available Cognex Type 2000
with suitable software, an example of which is set forth in
Appendix A.
The invention utilizes a user-friendly operator interface.
Menu-driven software, such as set forth in Appendix A, produces a
display on video monitor 15 that prompts the operator through each
phase of the four-step setup process. Help screens are accessible
at each step, outlining in detail the instructions for performing
that step. At the completion of each step, the system automatically
displays the next prompt on monitor 15. Setup errors may be quickly
corrected through the push-button editing feature with trackball
teaching module 14.
The operator uses the independent trackball teaching module 14 to
select a pattern on the PC board within each probe's field of view
of left and right probe assemblies 12 and 13. Teaching module 14
capture these two select patterns and then recognizes and locates
the matching patterns on the screen or stencil. Without further
operator involvement, the system implements the program by
calculating the pixel representation in x and y coordinates of
significant features and angular orientation in azimuth,
conveniently referred to a .theta. geometry. The apparatus then
determines an accurate "home" position for the screen/stencil in
relation to the circuit board then positioned on base
.Iadd.11.Iaddend.. The offset data corresponding to the difference
in x, y and .theta. positions of the observed circuit board
features and corresponding screen/stencil features is automatically
downloaded into the screen printer stepper motor programmable
indexes. These indexes activate the screen printer stepper motor
assembly to reposition the screen/stencil relative to each
stationary board for subsequent print cycles with screen/stencil in
horizontal alignment with each circuit board below. This automatic
positioning adjustment accommodates variations in board tolerances.
Between print cycles, the left and right video probes 12 and 13
automatically move in, look at the next board in place, move out
again, screen/stencil is positioned if necessary, and printing
occurs without operator intervention.
As an alternative feature the use may verify alignment with a
post-print inspection feature. As part of the setup menu, the
operator may choose how often to do post inspection. Then a choice
may be made from three tolerance threshold menu selections for
alignment error detection.
Printed circuit boards having patterns outside these parameters
produce an error message on video monitor 15. This error signal may
be used to activate a diverter to reroute that board for reworking.
The error signal may also be used to pause automatic operation so
that the operator can correct the problem before any more material
is wasted. To begin the setup process, the operator presses a
"video attention" button on teaching module 14. Actuating this
button produces the setup menu on video monitor 15. Then the
operator completes four steps, pressing a single button 14 on
teaching module 14 to signal completion of each step. First, the
operator clamps left and right video probe assemblies 12 and 13 in
place so that each probe when in the extended position may observe
two different patterns of significant features of the circuit board
and corresponding screen/stencil. The operator then teaches these
board patterns. Using trackball teaching module 14, the operator
moves joy stick 14A to define the windows within the .[.probes,.].
.Iadd.probes' .Iaddend.fields of view embracing significant
features, typically a pattern of pads for receiving solder paste,
and then pushes a single button on teaching module 14 to teach the
pattern to the apparatus which then captures the patterns. The
operator positions the probes of left and right probe assemblies 12
and 13 to view upward to learn the screen or stencil pattern above.
The associated data processing system automatically locates the
matching pattern on the screen or stencil, determines the pixel
representation of it in x, y and .theta. geometry; automatically
downloads offset data for x, y and .theta. for alignment purposes.
The operator then rotates the video probes downward toward the
circuit board, and the apparatus is now ready to print
automatically. As each board is positioned on base 16, the video
probes on left and right assemblies 12 and 13 extend above the
positioned circuit board, determine x, y and .theta. differences
between the positioned circuit board and the screen or stencil
above and operate stepper motors to reposition, if necessary, the
stencil or screen for alignment with the circuit board below.
Referring to FIGS. 2 and 3, there are shown perspective views of
the video probe assemblies in extended and retracted positions,
respectively. Each assembly includes a base 21 with vertical walls
22 and 23, each formed with front and rear slots 24 and 25,
respectively, having horizontal leading portions 24A and 25A,
respectively, and depending angled trailing portions 24B and 25B,
respectively. A vertical bracket 26 rests upon side walls 22 and
23. The movable probe support 27 is formed with elements, such as
31 for riding in slots 24 and 25. The movable base 27 supports
90.degree. mirror tube 32 that provides an image to camera 33 in
housing 34 and furnishes illuminating light to illuminate the
screen or stencil and the circuit board. FIG. 2 shows the video
probe in the extended position between circuit board and screen or
stencil. Arrow 35 illustrates the 180.degree. rotational range of
mirror tube 32 for looking above and below. FIG. 3 shows the video
probe in the retracted position outside the region between circuit
board and screen or stencil. Arrows 36 indicate the path traveled
by the probe up and forward to the extended position and rearward
and down to the retracted position.
Referring to FIG. 4, there is shown an exploded view of a number of
elements forming the video probe with some portions cut away. The
Volpie 90.degree. mirror tube 32 includes a mirror 32A mounted at
45.degree. angle relative to the axis of mirror tube 32 with a
Volpie intrascope approximately 12 inches long mounted inside
mirror tube 32. A rotate clamp mechanism is shown inside broken
line 42 with the parts exploded and functions to position mirror
tube 32 in one of two positions 180.degree. apart, looking up as
shown for viewing the bottom of the screen or stencil, or looking
down for viewing the top of the printed circuit board. Focus knob
43 allows for focusing the image of camera 33 by observing the
image on video monitor 15.
The main probe clamp 44 carries two shock absorbers 45 and 46 above
the left and right arms 47 and 48, respectively. Main clamp 44 also
carries three ball-and-cone pieces, two of which 51 and 52 are
visible in FIG. 4.
Referring to FIG. 5, there is shown an exploded view of a video
probe support assembly with the housing cut away and the video
probes omitted to better illustrate certain structural features.
Main clamp 44 supports the vision probe at the center of the camera
30 and Volpie intrascope unit 44. By supporting this assembly at
the center of gravity, when the ball-and-cone pieces 51, 52 and 61
engage reference balls, there is very low energy movement to
achieve precise reference position very quickly.
A linkage allows the center of effort to be aligned along the axis
of mirror tube 32 and Volpie intrascope unit 41. This linkage
includes a left link 62 and right link 63. Air cylinder 64 drives
the leading ends of these links with pin 65 seated in clevis 66
secured to the leading edge of air cylinder rod 64A. Main clamp 44
is formed with vertical recesses 44A and 44B that engage the left
and right arms 71 and 72, respectively. These arms accommodate four
rollers, a trailing pair in openings 71A and 72A, and a leading
pair in vertical slots 71B and 72B. Left and right side plates 22
are cammed as shown. Left and right angle brackets 73 and 74 are
attached to the top leading portions of left and right side plates
22 and 23, respectively, that rollers on the link system engage.
These angled brackets provided horizontal surfaces for insuring
that air cylinder 64 pulls the probe itself horizontally without a
vertical component of force. Rollers 75 and 76 ride on left and
right angled brackets 74 and 75, respectively. A pair of flow
controls 81 and 82 mount on the trailing portion of left side plate
22 and have a needle valve which allows re-exhaust in each
direction for aiding in providing smooth motion upon operating air
cylinder 64 for extension and retraction free of binding or violet
motion. Shock absorbers 45 and 46 also help allow smooth action
when the vision probe approaches a final location against the ball
and cone. Shock absorbers 45 and 46 provide damping that prevents
the CCD cameras 33 from being subjected to violet movements.
Main clamp 44 clamps the vision probe firmly to the tooling bed 11
of the machine, while allowing some relative adjustment on the
probe to position the mirror tube 32 between the stencil and
circuit board that may vary depending on the thickness of the
circuit board. Clamping screws may be loosened and two jacking set
screws 94A may be adjusted to control this position.
There is a right ball mount support 83 for supporting balls 83A and
83B and a left ball mount support 84 for supporting ball 84A. Balls
84A, 83A and 83B engage ball-and-cone pieces 51, 52, and 61,
respectively. Ball mounts 83C, 83D and 84B are independently
adjustable. When air cylinder 64 urges the probes to the extended
position, ball mounts 83C, 83D and 84B are tightened so that balls
83A, 83B and 84B are exactly seated in ball-and-cone pieces 52, 61
and 51, respectively. This position is then readily repeatable.
Left and right fiberoptic supports 85 and 86 are connected to left
and right links 62 and 63, respectively, and carry fibers optics
(not shown) in parallel to the Volpie intrascope. This arrangement
with fiber optics along and parallel to the Volpie intrascope
provides back lighting. That is to say, this structure provides a
very shallow light illuminating the object being viewed. The
shallow angle light reduces glare, and the image of the object
being viewed appears more clearly with back lighting.
Pins 85A and 86A pivotally support left and right fiber optic
supports 85 and 86 to slots 71B and 72B in left and right arms 71
and 72, respectively. This structural arrangement allows free
vertical rotation of these supports so that when air cylinder 64
extends and retracts the assembled slide unit, it prevents
significant free floating. Stated in other words, because the left
and right links are referenced on left and right angle brackets 73
and 74, respectively, at the extended end of the stroke for precise
horizontal actuation, the structural arrangement contains when not
actually contacting the angle brackets. Pins 85A and 86A in slots
71B and 72B provide this containment. Pin 65 pulls left and right
links 62 and 63 up and outward toward the extended end until
rollers 75 and 76 engage the underside of left and right brackets
73 and 74. On retraction pin 65 moves down and toward the retracted
position.
A focus hold bar 91 is secured to the top of main clamp 44 and
carries a piece of felt on the bottom for engagement by focus knob
43. This holding function helps prevent focus knob 43 from rotating
in the presence of repeated changes in position of the probe
assembly.
A proximity switch 92 is fastened to side plate 23 for providing a
signal indicating that the probe is in a safe retracted position to
allow the apparatus to print on the circuit board then below the
aligned screen or stencil.
There is a left clamp piece 93, left set screw adjust bar 94 and
left clamp nut plate 95 secured to the extended end of left side
plate 22. Similarly, there is a right clamp piece 96, right set
screw adjust bar 97 and right clamp nut plate 98 secured to the
extended end of right side plate 23. Set screws such as 94A and
97A, allow adjustment of set screw adjust bars 94 and 97.
Other components of the vision system include the vision processor,
a commercially available type.Iadd.. .Iaddend.
Cognex unit and monitor having Cognex commercially available
software for image searching for matching a stored image with an
observed image to permit alignment of the screen or stencil with a
circuit board below together with software for controlling the
menu-driven functions relative to the specific screen printer
application of the vision processor.
An operator communicates with the processor for aligning a
particular board through a trackball teaching module 14. This
teaching module has three buttons 14A, 14B and 14C and a ball 14D.
Rotating or pushing the ball with fingers allows the operator to
move the window within the field of view on monitor 15, and
actuating the buttons allows the operator to select a menu item or
move the cursor down through the menu. Operating button 14C selects
a help menu.
It may be advantageous to include a monitor mounted on an arm on
the side of the screen printer for observing what each vision probe
in assemblies 12 and 13 observes. Two vision probes are used to
detect rotational movement .theta. in addition to rectilinear
movement in x and y directions.
The light source is preferably a high intensity light coupled to
the fiber optics. These fiber optics carry light where needed on
each side of each vision probe, a total of four light sources.
There is also a source of illumination above the stencil for
backlighting the stencil to permit observation of features, such as
holes, in the stencil for identifying the image for capturing.
We turn now to a description of the process from the beginning when
an operator sets up the vision system to operate upon a specific
run of circuit boards to operation in an automatic cycle.
First, the operator adjusts the mechanisms on the screen printer to
handle the particular board size in conventional manner. That
involves adjusting the tracks and board stop, and inserting the
stencil needed to print on that board inside the screen printer.
These mechanical steps are the same as with the ASP-24 fully
automated screen printer without the vision features according to
the invention.
The operator then applies power to the apparatus to provide the
main menu on display. This main menu allows the operator to select
setup, edit setup, autoprint or address a help menu. First the
operator selects setup. After selecting setup, a menu prompt on the
screen instructs the operator to locate each vision probe over the
particular object on the circuit board which the operator feels is
unique and will be trained upon, such as a configuration of surface
mount circuit board footprint pads. The apparatus will then look
for this same unique pattern on subsequent boards in an automated
production run.
The operator manually positions a first vision probe, such as the
one in left assembly 12, over the board centered over the
particular unique object. The operator then clamps this vision
probe assembly in place. The operator then pushes button 14A on the
teaching track ball module to select the next prompt on the menu.
That prompt directs alignment of the second vision probe, such as
the one on right assembly 13. The operator aligns this second
vision probe in substantially the same manner as the first and
again presses button 14A to produce the next prompt. The board
patterns are taught and then the apparatus .[.loos.]. .Iadd.looks
.Iaddend.at the stencil. This next prompt instructs the operator to
rotate the vision probes .[.throu.]. .Iadd.through
.Iaddend.180.degree. to look at the stencil. The operator then
rotates mirror tubes 32 through the 180.degree. established by the
rotate clamps 42. The operator then drives the screen into position
above the board with joy stick and .theta. push button controls on
the ASP-24 machine. The operator positions the overhead structure
in x, y and .theta. into a position such that the vision probes
observe the pattern on the screen or stencil that matches the
pattern on the circuit board previously aligned centered within the
field of view of the probes. In response to each actuation of
button 14A the menu prompts the operator to teach the pattern that
is desirable after manually locating the stencil. The operator may
be required to make fine adjustments in the window by moving
trackball 14D and changing the window size until the object is
clearly defined and boxed in a square on monitor 15. The operator
then depresses push-button 14B and thereby teaches the system the
pattern it is looking for with the first probe. The operator
follows the same procedure for the second probe. Then the operator
pushes button 14C to indicate teaching is complete.
The screen printer then makes a number of automatic moves to learn
the geometry associated with this particular setup and this
particular type board.
The x stepper motor first moves the screen printer a predetermined
number of steps in the x direction. This movement defines the world
coordinate system for the cameras 33. Because of this feature the
cameras may be placed at any angle anywhere along the front of the
circuit board. The y stepper motors then move the screen printer a
predetermined number of steps in the orthogonal y direction to
confirm the world coordinate system. The system also recognizes the
number of steps per pixel during these x-y moves. These moves
enable recognition of how much the image moves for every step of
the stepper drive system. Then the stepper motors move the screen
printer through pure rotation a predetermined number of steps to
determine how the object translates in x and y coordinates during a
pure rotation. The printer repeats these translational and
rotational movements.
The system has thus recognized how the object moves in x and y and
how it translates in x and y during a rotational move. This
information on these moves allows the system to learn trigonometric
solutions of several triangles.
Referring to FIG. 6, there is shown a diagrammatic representation
of moves in x, y and .theta. helpful in understanding the
principles of the invention involved in learning the geometry of
the circuit boards and stencils. The pattern is characterized by a
center of rotation 111. The apparatus may learn from looking at the
stencil or screen driven by the stepping motors with reference to
two points, such as 112 and 113, that are some vector distance away
from the center of rotation 111. First performing a move in the x
direction facilitates learning the world coordinate system and
steps per pixel confirmed by a movement in the y direction. The
following rotational moves involves taking a picture of the
objects, such as 112 and 113, after an incremental move in one
direction, typically counterclockwise, followed by a move from the
initial position in the opposite direction by the same increment
from the initial position, typically clockwise. The moves in the x
and y direction basically define right triangles having a
hypotenuse of magnitude corresponding to the square root of the sum
of the squares of the incremental displacements in the x and y
directions. The angular displacements effectively create two
isosceles triangles from the shifts about points 112 and 113 with
the center of rotation 111 being the common vertex for both
isosceles triangles. The invention facilitates learning the
geometry of the board by looking at only two points and making
moves in x, y and .theta. directions for each of these points.
An operator may select any target point that appears unique on the
board such that it may be distinguished from other target points
around it, choose a second target point similarly distinguishable
from other surrounding targets, and teach the apparatus the
geometry of the board such that when any circuit board enters the
apparatus out of line with the stencil, the video probes looking at
the board automatically download the proper x, y and .theta. moves
for the stencil to bring stencil and board into alignment.
The operator then rotates both probes to look down at the board.
Looking now at the board, the system learns the pattern on the
board which correlates with a pattern on the screen. The operator
has then completed the automatic setup.
It may be desirable to modify this procedure slightly. For example,
it may be desirable to rotate the probes to look up again after the
system has learned the circuit board pattern to better correlate
circuit board and screen or stencil images.
With the probes then looking at the board, the operator may then
select auto print on the main menu, and the apparatus is then ready
for a production run. During a production run, both probe
assemblies index into position after the board has been brought in,
recognize the patterns that is learned on each of the probes and
automatically moves the screen relative to the board to align the
stencil very accurately with the board. The probes then moves into
the retracted position, printing occurs and the board just properly
printed exits the machine. A new board enters, set down against the
vacuum stop, the probes move to the extended position, the
apparatus recognizes the patterns, downloads to the stepper motors
the proper movements to align the stencil with the circuit board,
print, exits, and the process repeats.
As an alternative, the operator may edit a setup. If an operator
notices that the screen printer is printing consistently off the
pad in one direction or other, the operator may select edit setup
from the menu and modify where the screen printer is printing by
selecting a predetermined direction and distance of correction.
Thereafter, the apparatus will automatically print consistently in
the new location and continue to print in that location until
modified again.
A feature of the invention is the lighting arrangement for back
lighting the stencil. Stencils and screens are usually shiny or
have objects on the bottom which may be confused with object
features to be taught. By laying the translucent material on the
stencil and providing a light behind it, back lighting occurs which
prevents this problem. This arrangement disperses the light in a
manner that clearly defines each hole in the stencil relative to
any type of reflective background that might occur.
The invention has a number of features. The vision probes enter
between circuit board and stencil and perform the alignment as
distinguished from looking at the board outside the screen
printer.
There has been described novel apparatus and techniques for
aligning. It is evident that those skilled in the art may now make
numerous other uses and modifications of and departures from the
apparatus and techniques herein disclosed without departing from
the inventive concepts. Consequently, the invention is to be
construed as embracing each and every novel feature and novel
combination of features present in or possessed by the apparatus
and techniques herein disclosed and limited solely by the spirit
and scope of the appended claims. ##SPC1##
* * * * *