Data processing: generic control systems or specific application – Specific application – apparatus or process – Product assembly or manufacturing
Reexamination Certificate
1999-07-27
2003-07-01
Follansbee, John (Department: 2121)
Data processing: generic control systems or specific application
Specific application, apparatus or process
Product assembly or manufacturing
C029S740000
Reexamination Certificate
active
06587743
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention generally relates to automatic configuration of semiconductor device handling equipment to accommodate multiple applications and random variations among machines and devices.
In the semiconductor industry, a considerable number of electronic devices are provided by vendors in programmable form with blank memories or unspecified connections between arrays of logic. Users can then custom configure or program the electronic devices to perform their intended function by programming them, transferring or “burning in” a sequence of operating codes into the memory, or by specifying a particular arrangement of gating logic connections.
Numerous manufacturers have developed automated machinery for handling and programming such devices. Such machinery moves blank devices from a source medium (e.g., trays, tubes, etc.) to one or more programming sites, carries out the programming operation on each device, and moves programmed devices from the programming sites to an output medium (e.g., trays, tubes, etc.).
Both to allow flexible handling of a wide variety of automated programming operations (different types of input or output media, different device package types, etc.) and to account for inevitable manufacturing variations from machine to machine, it is necessary for the equipment operator to configure (or “teach”) the automated programming machinery the precise locations from which to pick up devices and to which to place devices. This includes all input and output media locations, the locations of all programming site sockets, and any other such locations within the system.
Accurate teaching is critical to the robust operation of automated programming systems. While older, larger programmable devices are relatively insensitive to placement accuracy, modem fine pitch devices have very delicate leads and suffer damage unless placement operations are highly accurate (for instance, correct to within 0.001″).
Not all automated equipment can achieve such accurate placement. In order to do so, high-end equipment uses a technique known as vision centering. The system picks up each device with a pick and place nozzle and holds the device in the path of a series of parallel laser beams. The device is then rotated in the path of the laser beams. A bank of sensors monitors which of the beams is interrupted during the rotation of the device. This information can be processed numerically to identify the precise angle and position of the device on the pick and place nozzle. Linear encoder technology allows the system to place the nozzle at any desired location. The results of the vision centering operation allow a correction move in the lateral horizontal (i.e., X, Y) and angle coordinates to be performed so that the part can be precisely placed at the desired location.
Traditional vision centering does not assist the user in determining the correct pick and place locations. Rather, it merely enables the system to place with extreme accuracy once those locations are specified.
While the ease with which the operator can carry out the teaching operation does not directly affect the physical treatment of devices, it does affect the efficiency of system operation. Typical users of automated programming equipment are highly sensitive to system throughput (measured in correctly programmed devices per hour) and yield (defined as the percentage of devices which are correctly programmed). Furthermore, more difficult teaching techniques require more highly trained personnel that draw higher wages. As a result of all of these considerations, automated programming equipment users prefer fast, easy teaching techniques.
A variety of teaching techniques has been implemented to date. These include none, file-controlled, trial-and-error, single point downward vision teaching, and double point downward vision teaching. Each of these techniques is characterized by advantages and disadvantages when considered in terms of accuracy and ease of use.
Some equipment requires no teaching whatsoever. This equipment is simple to set up and use, but is limited to devices that are rugged and can tolerate the relatively imprecise component placement that results from inevitable manufacturing variations from machine to machine. Such equipment is normally also limited to a specific device type or a small range of types and offers little flexibility to handle new devices.
Some equipment requires no active teaching on the part of the user, but offers improved flexibility by utilizing CAD data files to determine the location of device pick and place points. Systems of this type offer more flexible handling of a variety of jobs, but are still limited to rugged components that can tolerate relatively poor placement accuracy, because this teaching technique doesn't account for random manufacturing variations.
A wide variety of equipment is available that provides operator control over the various pick and place locations but offers no systematic technique to help the operator determine the proper settings. The operator of such equipment must pursue a trial-and-error approach until correct settings are determined. Once accurately configured, such equipment can operate very reliably, but the trial-and-error process can be very time consuming and can result in many damaged (and hence unusable) parts. Equipment of this type still normally handles only rugged devices.
Some automated device programming systems are equipped with downward vision cameras. Such a camera is mounted to the movable portion of the system and can be positioned over any point in the system workspace. The camera can “look down” on the components or component locations. The operator can observe the camera field of view on a monitor which is normally equipped with crosshairs for precise positioning. Downward vision cameras can be used in “single point” or “double point” teaching mode.
In a system that employs single point downward vision teaching, the operator positions the camera crosshairs over the estimated center of the component location to be taught and indicates via a keystroke, mouse click, or some other event that the proper position has been identified. The system then stores this position and returns to it when necessary. This approach provides better accuracy than all previously described techniques, and requires only a single camera positioning operation by the user. However, the approach requires that the operator visually estimate the proper crosshair location. This can be inaccurate unless the location being taught exhibits some kind of distinct “landmark” at the center point, which is not always the case. When such a landmark is available, systems of this type can properly handle delicate fine pitch parts, while some device damage can result in the absence of such landmarks.
Double point downward vision teaching improves upon the single point technique by allowing the user to teach two points and taking as the teach point the arithmetic average of those two points. In most systems it's much easier to find two symmetrically located “landmarks” than to find a single landmark at the precise target location. A disadvantage of this approach, however, is that the operator must position the crosshairs twice, doubling the labor involved in the teaching process. Historically, double point downward vision teaching has provided the most reliable results and has proven most successful in handling fine pitch devices.
None of the approaches described above fully exploit the fine positioning capability of modem motion control equipment. Even the most accurate of the above methods, double point downward vision teaching, requires that the operator visually align two references (the camera crosshair and the image landmark) with one another. Human vision is not capable of performing this feat to the full accuracy of motion control hardware. The approaches that do work well require extensive operator involvement and thus admit the possibility of human error.
A much more serious limitation of all existing appr
Ingram Samuel K.
White William H.
B P Microsystems, Inc.
Gain, Jr. Edward F.
Gray Gerald T.
LandOfFree
Pick and place teaching method and apparatus for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Pick and place teaching method and apparatus for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Pick and place teaching method and apparatus for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3042174