00191913-01.pdf - 第223页
User Manual SIPLACE 80S-20/F4 6 Vision functions Software version SR.406.xx 02/2000 Issue US 6.3 Component vision systems 221 6.3 Component visi on systems The PCB vi sion syste m recor ds the pre cise positi on of a com…
6 Vision functions User Manual SIPLACE 80S-20/F4
6.2 PCB vision system Software version SR.406.xx 02/2000 Issue US
220
– Number of fiducials
When using ceramic substrates and small boards, tow fiducials will usually be sufficient. How-
ever, with larger boards it is recommended that three fiducials be defined. The individual fidu-
cials can vary in structure. You can simplify the recognition procedure if you use the same
structure for each fiducial.
– Correction with two fiducials X-position
Y-position
PCB skew
– Correction with three fiducials: ideally, the straight lines which each pass through the cen-
ters of two fiducial will be parallel with the x and y axes
X-position
Y-position
PCB skew
Shear
Warpage of PCB in the X direction
Warpage of PCB in the Y direction
NOTE 6
You should never position 3 fiducials so that they are located on a straight line. 6
– Spacing between the fiducials
You may locate the fiducials at any point on the board. However, it is a good idea to space the
fiducials as far apart as possible on the two axes. The further apart the fiducials are from one
another, the more accurate optical position and angle recognition will be.
User Manual SIPLACE 80S-20/F4 6 Vision functions
Software version SR.406.xx 02/2000 Issue US 6.3 Component vision systems
221
6.3 Component vision systems
The PCB vision system records the precise position of a component by determining the offset be-
tween the component center and the nozzle’s axis of symmetry, and by calculating the rotational
offset from the relative rotational position of the nozzle. It is also possible to analyze the state of
the lead configuration in the X and Y directions. The component vision system consists of the op-
tical system for recognizing the component position, and the vision evaluation unit. 6
6.3.1 Component vision system on the 12-segment revolver head
6.3.1.1 System description
The 12-segment revolver head has its own component position recognition system in star station
7 (see Fig. 6.1 - 3, page 206). The vision evaluation unit for PCB and component position recog-
nition is located in the control unit (see Fig. 6.1 - 6, page 211). 6
The optical component position recognition system consists of a CCD camera with deflector, im-
age lens, and LED illumination system. The usable field of view of the CCD camera (SONY XC75
camera) measures 24 mm x 24 mm. For position recognition, and for the lead test, the component
is evenly illuminated by the rows of LEDs using the front lighting method, and is projected onto the
CCD chip in sharp focus by the lens. The position, rotational angle, and lead condition parameters
are determined using the h
igh-accuracy lead extraction (HALE) method of digital image process-
ing. 6
The vision evaluation unit (MVS) was described in section 6.1.4, page 210, since it performs both
PCB and component evaluation functions. 6
6.3.1.2 Technical data
Camera type: SONY XC75
Number of pixels: 484 x 484
Field of view: 24mm x 24mm
Illumination method: Front lighting (infrared light), 3 LED levels
Image processing: HALE gray scale method
(H
igh Accuracy Lead Extraction)
Screen: RGB monitor (VGA mode), 640 x 484 pixels
Component sizes: 0.5mm x 0.5mm ... 18.7mm x 18.7mm
Range of recognizable components: TSOP, LCC, PLCC, QFP, SO series up to SO28
essentially all components with J and
gull wing leads,
µ
BGAs
Minimum lead pitch: 0.3mm for the camera
0.5mm for the machine
Minimum ball diameter for
µ BGAs
:250µm
6 Vision functions User Manual SIPLACE 80S-20/F4
6.3 Component vision systems Software version SR.406.xx 02/2000 Issue US
222
6.3.1.3 Functional description
A segment of the 12-segment revolver head picks up a component at star station 1. As the star
advances, further components are picked up. Once a component reaches star station 7, where
the component vision system is located, three offset rows of LEDs evenly illuminate the compo-
nent with infrared light. The lens maps components up to 5mm high in sharp focus onto the cam-
era’s CCD chip. 6
The digital image of the component generated by the camera is then transferred to the vision eval-
uation unit, where the HALE digital image processing method is used to compare the component
image with a synthetic model previously created in the GF (package form) editor. The parameters
thus obtained provide information on deviations in position, the rotational angle, lead condition,
and component reidentification. The HALE method has proved to be highly resistant to disruptive
factors such as unwanted reflection, different reflection behavior of leads, diffused light, etc, and
it is faster and more accurate than the matching method.Once the measurement is complete, the
segment advances to star station 9, and rotates the component into the correct orientation for
placement. Finally, the component is placed in the correct position on the PCB at star station 1.6
6.3.2 Component camera module for the Pick&Place head
6.3.2.1 System description
Component vision module for the Pick&Place head: fine pitch vision module 6
All the optical components of the system 6
– CCD camera (SONY XC77 camera)
–lens
– optical strip filter for suppressing unwanted reflections
are located in a dust-tight housing. The field of view of the CCD camera measures 38 mm x
38 mm. For position recognition, and for the lead test, the component is illuminated by three LED
levels using the front lighting method, and is projected onto the CCD chip in sharp focus by the
lens. The position, rotational angle, and lead condition parameters are determined for fine pitch
components and BGAs (Ball Grid Arrays) using digital image processing methods. 6
Component vision module for the Pick&Place head: flip-chip vision module 6
All the optical components of the system, such as the 6
– CCD camera (SONY XC75C camera)
–lens