201005285 , 六、發明說明： [發明所屬之技術領域】 一般而言，本發明係涉及利用光學 的-個糸站“ ㈣用无学檢測物體表面瑕疵 们 幻糸統與方法，更具體而, -r， Μ S ，本發明係涉及本導艚 裝檢測的一個糸少次牛等體封 個系統與方法。本發明該系 測半導體封裝多重表面。 ’、Μ T同步檢 【先前技術】 ❹ ❹ 半導體封裝或設備之品質管制是半 極為重要且所費不實的一環。一 =製造過程 ，卞I 在銷售與輸出之前 *如四方扁平無接腳、球閘陣列與 裝技術必須要古钱故。 9圓級晶片尺寸等封 文釘义肩要有威格的品質管制與分 瑕疫偵測（以下稱之為品質檢 +導體封裝表面 銷售與輸出之前，有機會= 半導體封裝 體封裝之精確性斑σ皙恭 — '、此由於半導 體封F表面1 : I、 越來越受到重視，半導 裝表面m質檢驗也成為半導體封 導 為重要的步驟。 、盤製裎中一個極 —般都會針對半導體封裝檢測其表面 孔隙盘细餘播士今从^ 諸如内部 構成之精確性’此外也會對半導 核 二了解終端位置、間距、同平面與尖 Τ …進-步對半導體封裝實施檢測 。 裝表面之‘不或製造商資料是否漏印或拼錯的情开》。 :次疋出口之刚，必項對半導體封裝 而或疋側邊表面實施檢測。 3 201005285 目月1j有數種檢測半 裝檢測過程中，、 导體封裝的系統與方法。半導體封 以捕捉影像， 吏用堵如電何耦合元件之照相設備 用單一攝影機—檢測半導體封裝的典型系統與方法係運 表面實施檢洌Y =同步對半導體封裝底部、上端與側邊 過大，單—攝办’、、、而，由於底部、上端與側邊表面總面積 偵測瑕疲，此=機配置通常會導致檢測解析度下降或無法 用之鏡頭啟铲而在檢測過程中，通常與單-攝影機配置連 Ο ❹ 換通常會導勒^ y移置或更換，鏡頭與鏡面之移置或更 寸双杈準與調整回饋之遺失。 、f 一括士等人在其美國專利申請刊物US 2003/0086083 A1中詳 半導體封裝檢測之替代系統。_等人系統包括 夕重個別鏡面鱼鐘 ， "、兄頭該系統運用一個分光鏡，將反射自 "體封裝的光線沿著高、低放大路徑分隔之，以利高、 低放大攝影機分別捕捉半導體封裝之高、低放大影像。低 放大衫像用來識別所望區域，然後高放大攝影機會將該區 域放大，以利更細緻檢測作業之遂行。然而，Ebert等人 之檢測系統通常必須佔用相當大的空間，此外該系統必須 對多重鏡面與鏡頭進行個別放置，相當耗費資源，同時精 確性也會受到影響，再則該系統所捕捉之高、低放大影像 係屬同一所望區域的相同影像（雖然解析度不同）。 美國專利US 6, 952, 491 B2與美國專利申請刊物us 2004/026 3834 A1中，Alumot等人描述檢測半導體設備表面的一個方 法與裝置’該方法與裝置系運用兩個階段進行半導體設備 表面之檢測作業。第一階段中，在相對高速與極低的空間 4 201005285 - 解析度環境下，實施半導體表面之檢測。第二階段中 遥取之半導體表面位置則在相對向空間解析度環境下 檢測。接著利用一部外部處理器或控制器，對第一階 第二階段過程中所蒐集之資訊加以分析，並對第一階 第二階段所獲得影像實施比較，並參考相關影像，以 測半導體設備表面之瑕疵。 但是Alumot等人所公開之兩階段檢測方法非常浪 間與資源’此外該方法與裝置必須對所獲得影像實施 ❹ 並必須參考相關影像，是極不精確與耗費時間的一 業程序。再則Alumot等人所公開的方法與裝置僅能夠 半導體設備的一個表面，同時該裝置也包括多重組件 為笨重。Alumot等人所發明裝置必須具備較大空間才 裝，因此製造空間也必須足夠，最後導致製造成本的 四此枣領域專業人士應當了解，半導體封裝右 ’、方法有必要實施改良。經過改良之半導體封裝相 ’、方法應具備捕捉較高解析度影像的能力，同時 像以利瑕疵偵測與品質檢驗，出 改良的糸統必須節省空間。 發明内容】 ，所 進行 段與 段與 利偵 費時 比較 個作 檢測 ，極 能安 增加 系統 系統 夠捕 經過 測系統，該系統包 個觀點’此處揭露一個半 —個反光片總成，該反光 備檢 經過 5 201005285 - 型態模組，以接收反射自一個半導體設備一個第一表 光線，以及接收反射自該半導體設備至少一個第二表 光線。該第一表面沿著一個第一平面充分延伸，同時 少一個第二表面沿著至少一個第二平面充分延伸。該 一個第二平面與該第一平面形成一個角度，同時該反 總成進一步經過型態模組，以利沿著一個第一個方向 分引導反射自該第一表面與該至少一個第二表面的光 該半導體設備檢測系統進一步包括一個分光鏡，以利 著第一方向前進的光線分割成一個第一光束與一個第 〇 束，該第一光束沿著一個第一路徑前進，該第二光束 一個第二路徑前進。該第一路徑與一個第一影像捕捉 充分交會，該第二路徑與一個第二影像捕捉平面充分 ，該至少一個第一表面與該至少一個第二表面的影像 該第一影像捕捉平面與該第二影像捕捉平面獲得之， 隨後之檢測作業。 依據本發明第二個觀點，此處揭露一個半導體設 測方法，該方法包括捕捉一個第一影像，該第一影像 G 一個第一視界，該方法進一步包括捕捉一個第二影像 第二影像具備一個第二視界。該第一視界包括該半導 備一個第一表面一個影像，而該第二視界包括一個第 面其他影像，以及該半導體設備該第一表面與至少一 二表面之影像。該第一表面沿著一個第一平面充分延 同時該至少一個第二表面沿著至少一個第二平面充分 ，該至少一個第二平面與該第一平面形成一個角度。 法進一步包括處理至少該第一影像與該第二影像其中 ，以利檢測該半導體設備該第一表面與該至少一個第 面的 面的 該至 至少 光片 ，充 線。 將沿 二光 沿著 平面 交會 可由 以利 備檢 具備 ，該 體設 一表 個第 伸， 延伸 該方 之一 二表 6 201005285 - 面其中之一。第一影像捕捉與第二影像捕捉係同步或前 相繼執行二種方式其中之一，以利同步或相繼檢測該半 體設備該至少第一表面與該至少一個第二表面其中之一 依據本發明第三個觀點，此處揭露一個半導體設備 測方法’該方法包括捕捉一個第一影像，該第一影像之 色為至少一個第一視界與一個第一組特性，捕捉一個第 影像，該第二影像之特色為至少一個第二視界與一個第 組特性。該第一視界包括該半導體設備一個第一表面一 影像，而該第二視界包括一個第一表面其他影像，以及 〇 半導體設備該第一表面與至少一個第二表面之影像。該 一表面沿著一個第一平面充分延伸，同時該至少一個第 表面沿著至少一個第二平面充分延伸，該至少一個第二 面與該第一平面形成一個角度。第一組特性與第二組特 包括至少照明顏色其中之一、至少照明亮度其中之一與 少照明角度其中之一。該方法進一步包括處理該第一影 與該第二影像至少其中之一，以利檢測該半導體設備該 一表面與至少一個第二表面莫中之一。第一影像捕捉與 〇二影像捕捉係同步或前後相繼執行二種方式其中之一， 利同步或相繼檢測該半導體設備該至少第一表面與該至 一個第二表面其中之一。 【實施方式】 半導體封裝或設備之品質管制是半導體封裝製造過 極為重要且所費不貲的一環。目前有數種半導體封裝表 品質檢測與瑕疵偵測的系統與方法，現行系統與方法各 後 導 0 檢 特 個 該 第 平 性 至 像 第 第 以 少 程 面 種 7 201005285 缺失與問題已如前述 其缺點就是所捕捉的 動（例如鏡頭與鏡面 ’此外配備多重攝影 則必須彳占用極大的空 L如果採用配備單—攝景彡機的系 影像解析度不足。現行系統組件 )通常會導致校準或是調整回饋 機之現行半導體封裝品質檢測系 間，因而導致製造成本的増加。 ，本領域專業人士應當了解 有必要實施改良。 半導體封裝檢測 系統與 ❹201005285, VI. Description of the invention: [Technical field to which the invention pertains] In general, the present invention relates to the use of optical - "station" (4) to detect the surface of objects with no learning, and more specifically, - r, Μ S, the present invention relates to a system and method for reducing the number of times of cattle and the like. The invention measures multiple surfaces of a semiconductor package. ', Μ T synchronous detection [prior art] ❹ ❹ The quality control of semiconductor packaging or equipment is a semi-essential and unrealistic part. One = manufacturing process, 卞I before sales and output * such as quad flat no pin, ball grid array and loading technology must be ancient money 9 round wafer size and other sealed nails have Wigger quality control and plague detection (hereinafter referred to as quality inspection + conductor package surface sales and output, there is a chance = semiconductor package packaging accuracy Sex plaque — 皙 — - ', this is due to the semiconductor surface F: 1: I, more and more attention, semi-conducting surface surface quality inspection has become an important step in semiconductor sealing. As for the semiconductor package, the surface of the pore plate is tested for its accuracy. For example, the accuracy of the internal structure is also known. In addition, the terminal position, spacing, plane and sharpness of the semi-conducting core are also known. The package is tested and tested. The surface of the device is not “whether the manufacturer's information is missing or misspelled.”: The second exit is required to be tested on the semiconductor package or on the side surface. 3 201005285 A system and method for detecting a half-package inspection, a conductor package, a semiconductor package for capturing images, and a single camera for photographic equipment that plugs electrical and coupling elements - a typical system and method for detecting semiconductor packages Check Y = Synchronous to the bottom, upper end and side of the semiconductor package is too large, single-photo, ',,,, because the total area of the bottom, upper and side surfaces is detected to be fatigued, this machine configuration usually leads to detection and analysis The lens is lowered or can not be used to start the shovel. During the detection process, it is usually connected with the single-camera configuration. 换 Change usually leads to y y displacement or replacement, lens and The displacement of the surface or the loss of the adjustment and the loss of the adjustment feedback. F. et al., in its U.S. Patent Application Publication No. US 2003/0086083 A1, describes an alternative system for semiconductor package inspection. Mirror fish clock, ", brothers, the system uses a beam splitter to separate the light reflected from the "body package along the high and low amplification paths, to capture the high and low semiconductor package respectively. Enlarge the image. The low-magnification shirt image is used to identify the desired area, and then the high-magnification camera magnifies the area to facilitate more detailed inspections. However, the detection system of Ebert et al. usually has to occupy a considerable amount of space. The system has to place multiple mirrors and lenses separately, which is quite resource intensive and the accuracy is also affected. The high and low magnification images captured by the system are the same images in the same desired area (although the resolution is different). A method and apparatus for detecting the surface of a semiconductor device is described in U.S. Patent No. 6,952, 491, issued to U.S. Pat. Test the job. In the first stage, the detection of the semiconductor surface was carried out in a relatively high-speed and extremely low space 4 201005285 - resolution environment. The position of the semiconductor surface remotely taken in the second stage is detected in a relative spatial resolution environment. Then, an external processor or controller is used to analyze the information collected in the second stage of the first stage, and compare the images obtained in the second stage of the first stage, and refer to related images to measure the semiconductor device. The top of the surface. However, the two-stage detection method disclosed by Alumot et al. is very vain and resource. In addition, the method and device must implement the image obtained and must refer to the relevant image, which is a very inaccurate and time-consuming process. Furthermore, the method and apparatus disclosed by Alumot et al. can only be one surface of a semiconductor device, while the device also includes multiple components that are cumbersome. The device invented by Alumot et al. must have a large space for packaging, so the manufacturing space must also be sufficient, and finally the manufacturing costs of the professionals in the field should be understood, the semiconductor package right, the method needs to be improved. The improved semiconductor package phase, the method should have the ability to capture higher resolution images, and at the same time, like the detection and quality inspection, the improved system must save space. SUMMARY OF THE INVENTION The comparison between the segment and the segment and the profit detection time is very high, and the system system can be sufficiently captured to capture the system. The system includes a view that a half-reflective sheet assembly is disclosed here. The test is performed by 5 201005285 - a type module to receive a first meter of light reflected from a semiconductor device and to receive at least one second meter of light reflected from the semiconductor device. The first surface extends substantially along a first plane while at least one second surface extends substantially along the at least one second plane. The second plane forms an angle with the first plane, and the anti-assembly further passes through the type module to guide the reflection from the first surface and the at least one second surface along a first direction The semiconductor device detection system further includes a beam splitter for dividing the light traveling in the first direction into a first beam and a first beam, the first beam traveling along a first path, the second beam A second path advances. The first path is fully intersected with a first image capturing, the second path is sufficient for a second image capturing plane, and the image of the at least one first surface and the at least one second surface is the first image capturing plane and the first The image capture plane is obtained, and the operation is subsequently detected. According to a second aspect of the present invention, a semiconductor device is disclosed, the method comprising: capturing a first image, the first image G, a first field of view, the method further comprising capturing a second image, the second image having a Second horizon. The first field of view includes an image of the first surface of the semiconductor, and the second field of view includes a first image of the first surface and an image of the first surface and the at least one surface of the semiconductor device. The first surface is substantially extended along a first plane and the at least one second surface is substantially along at least one second plane, the at least one second plane forming an angle with the first plane. The method further includes processing at least the first image and the second image to facilitate detecting the at least light sheet of the first surface of the semiconductor device and the at least one first surface, charging. The meeting will be carried out along the plane along the two lights. It can be prepared by Eli. The body is set to extend one end and extends one of the two sides. 6 201005285 - One of the faces. The first image capture is synchronized with the second image capture system or performed in succession in advance to facilitate simultaneous or sequential detection of the half body device. The at least first surface and the at least one second surface are in accordance with the present invention. In a third aspect, a method for measuring a semiconductor device is disclosed herein. The method includes capturing a first image, the color of the first image being at least one first field of view and a first group characteristic, capturing a first image, the second The image is characterized by at least one second field of view and a first set of characteristics. The first field of view includes a first surface image of the semiconductor device, and the second field of view includes a first image of the first surface and an image of the first surface and the at least one second surface of the semiconductor device. The surface extends substantially along a first plane while the at least one first surface extends sufficiently along at least one second plane that forms an angle with the first plane. The first set of characteristics and the second set include one of at least one of the illumination colors, at least one of the illumination levels, and one of the illumination angles. The method further includes processing at least one of the first image and the second image to facilitate detecting one of the surface of the semiconductor device and the at least one second surface. The first image capture is performed synchronously with the second image capture system or sequentially, one of two ways, to synchronously or sequentially detect one of the at least first surface and the one to the second surface of the semiconductor device. [Embodiment] The quality control of semiconductor packages or devices is an extremely important and costly process for semiconductor package manufacturing. At present, there are several systems and methods for quality inspection and flaw detection of semiconductor package meters. The current system and method are each 0. The first level is the same as the first type. The number is missing. The problem is as described above. The disadvantage is that the captured motion (such as lens and mirror) is also required to take up a lot of space. If you use a single-camera camera, the image resolution is insufficient. The current system components usually lead to calibration or Adjusting the current semiconductor package quality inspection system of the feedback machine leads to an increase in manufacturing costs. Professionals in the field should understand the need to implement improvements. Semiconductor package inspection system and ❹
為提供簡單扼要之說明，以下有關本發明之 於具備多重放大與多重表面檢測功能之半導體封： 統與方法，但此種說明方式並無意將本發明各種應 例，自類似操作方式其他應用系統中排除之。在各 應用實施例中，本發明各種實施例所依據的基本原 仍有其普遍性，諸如操作、功能與性能等基本原理 半導體封裝檢測系統與方法實施例請參閱第^ 11圖，其中類似元件以類似參考數值標示之；其中 之L1係為LED設備；第3圖所指之[2為光路；第9In order to provide a brief and concise description, the following is a semiconductor package and method for the present invention with multiple amplification and multiple surface detection functions, but this description is not intended to be used in various applications of the present invention. Excluded from it. In various application embodiments, the basics on which the various embodiments of the present invention are based are still universal, such as operations, functions, and performance. The basic principles of semiconductor package inspection systems and methods are described in FIG. It is indicated by a similar reference value; among them, L1 is an LED device; Figure 3 refers to [2 is an optical path;
)圖係為第一攝影機24所捕捉之影像示意圖，且該圖 標示之L3係為半導體封裝12底部表面之影像，u為半 封裝12側邊表面之影像；而第9圖之（b)圖係為第二 機26所捕捉之影像示意圖，該圖中之L5係為半導體封 底部表面之影像D 依據本發明一個第一實施例，此處提供—個半導 裝（也稱為半導體基板）檢測系統1〇。半導體封裝12 但不限於以下事項：四方扁平無接腳、球閘陣列與晶 統， 之移 遺失 統， 因此 方法 僅限 測系 實施 各樣 原則 則。 至第 2圖 之（a 中所 導體 攝影 裝12 體封 包括 圓級 201005285 曰曰 片尺寸等封裝技術。可行的方式為該系統 明總幻4或照明配置、—個棱鏡總成16與數具攝景t機個照 可行的方式是每一半導體封裝12包括一個上沪 一個底部表面與四個側邊表面’言亥半導體封裝12二面、 面與底部表面也可分別稱之為該半導體封裝口 = 4可仃的方式是該半導體封裝12四個側邊表面的a 一矣 面與上端表面及底部表面形成肖度，本領域’- Ο 了解該半導體封|12側邊表面可以超過四自，例 體封裝口可以是八邊形或是十㈣，則該 ：該半導 邊表面分別4 8個表面或1〇個表面。 十装12側 可行的方式是該半導體封裝12之檢 士抓主心你 括上端表面 、底4表面與四個側邊表面每一表面之影 半導體封裝12之檢測應包括任何數量表面之影像捕：者該 可行的方式是該半導體封裝12檢測系統 $ 半導體封裝12底部影像，同時也能夠捕捉四個2捕捉該 一表面之影像。底部表面與側邊表面影像捕捉有表面每 瑕疵之偵測。 孩有助於表面 半：體封裝12表面瑕疵包括但不限 點、污染、型態錯誤與互連定位不精確等，表1痕、斑 有助於被偵測表面之瑕疵修 . 衣面瑕疵偵測 4百平導體封 疵偵測有助於處理有瑕疵 才裝12表面瑕 封裝…測有助於全面提升:::體:二半導體 。 子體封裝12之品質 9 201005285 ^ 在檢測與影像捕捉過程中’可行的方式是將半導體封 裝12置於一個預備位置18 ’可操作式照明總成14可提供或 引導光線指向預備位置18的該半導體封裝12。可行的方式 是該照明總成14包括一個角度照明總成20與一個同轴照明 總成22，可行的方式是該角度照明總成20與同軸照明總成 22包括數層的發光二極體設備’或者該角度照明總成20與 同軸照明總成22包括氙閃爍照明設備，或更進一步該角度 照明總成20與同軸照明總成22包括本領域專業人士所熟知 ❹ 的照明設備或裝置，或者該角度照明總成20與同轴照明總 成22分別包括不同的照明設備。 可行的方式是以一個 照明總成20與同軸照明總 一步一個可行方式是該角 之空間安排配置或型態模 人士應當了解在檢測過程 配置能充分固定，有利於 〇 提升作業效率。 預先律定的型態模組進行該角度 成22之空間安排配置，或是更進 度照明總成20與同軸照明總成22 組必須充分固定之。本領域專業 中’如果該照明總成14空間安排 該半導體封裝12之照明，同時可 遠角度照明總成2〇可提供弁 18該半導體封裝12的 、”或將光線指向預備位置 , " ，、四個侧邊表面的每一表面 〇仃、方式是該角度照明總成 層均以預先揸+ & Α 战20夕層LED設備中，每一 部表面，進—來 罝18該半導體封裝12的底 遇 步一個可行方式熹处糾 實施調整或修計，庶03 此夠針對該預先律定角度 訂，使，'、、月能夠依必I -V, 導體封裝。此要或所望角度指向該半 此外一個可行方式是 疋这夕層LED設備之空間配 10 位置18該半導體 的光線間產生相 201005285 置安排，應能热士 勹充分避免反射自預備 12底部表面盘v ^ m ，、四個側邊表面每一表 擾的情形。 @ J 4丁的大 Θ 1 J不式是該同軸照明The figure is a schematic image captured by the first camera 24, and the icon shows that L3 is an image of the bottom surface of the semiconductor package 12, u is an image of the side surface of the half package 12; and FIG. 9(b) The image is captured by the second machine 26, and the L5 in the figure is the image D of the bottom surface of the semiconductor package. According to a first embodiment of the present invention, a semi-conductive package (also referred to as a semiconductor substrate) is provided herein. Detection system 1〇. The semiconductor package 12 is not limited to the following: the quad flat no pins, the ball gate array and the crystal, and the missing system, so the method is limited to the implementation of various principles. To the second figure (a, the 12-body seal of the conductor photographic package includes a round-scale 201005285 cymbal size and other packaging techniques. The feasible way is the system illusion 4 or lighting configuration, a prism assembly 16 and several A possible way for the camera to be photographed is that each semiconductor package 12 includes a bottom surface and four side surfaces of the upper Shanghai. The two sides of the semiconductor package 12, the surface and the bottom surface may also be referred to as the semiconductor package, respectively. The mouth = 4 is 仃 仃 是 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体 半导体The body package port can be octagonal or ten (four), then: the semi-conductive side surface has 4 8 surfaces or 1 表面 surface respectively. The ten-mount 12 side feasible way is the inspection of the semiconductor package 12 The main point of view is that the surface of the upper surface, the bottom 4 surface and the four side surfaces are each. The detection of the semiconductor package 12 should include image capture of any number of surfaces. The possible way is that the semiconductor package 12 detection system $ semiconductor It is equipped with 12 bottom images, and can also capture four images capturing the surface of the bottom surface. The bottom surface and the side surface image capture the detection of each surface of the surface. The child helps the surface half: the body package 12 surface includes but not Limits, contamination, type errors, and inaccurate interconnection positioning, etc. Table 1 traces and spots help to repair the surface to be detected. The cover 瑕疵 detection 4 hundred flat conductor seal detection helps to deal with瑕疵 装 12 瑕 瑕 ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... : : : : : : : : : : : : : : : : : : : : : : : : : : : The ready position 18' operable illumination assembly 14 can provide or direct light to the semiconductor package 12 at the ready position 18. It is possible that the illumination assembly 14 includes an angular illumination assembly 20 and a coaxial illumination assembly 22 The feasible manner is that the angle illumination assembly 20 and the coaxial illumination assembly 22 comprise a plurality of layers of light emitting diode devices' or the angle illumination assembly 20 and the coaxial illumination assembly 22 comprise xenon flashing illumination devices, Still further, the angled illumination assembly 20 and coaxial illumination assembly 22 include illumination devices or devices well known to those skilled in the art, or the angle illumination assembly 20 and the coaxial illumination assembly 22 respectively include different illumination devices. The way to use a lighting assembly 20 and coaxial lighting is a feasible way. The space configuration of the corner or the type of mold should be understood that the configuration of the inspection process can be fully fixed, which is conducive to improving the efficiency of the work. The type module performs the spatial arrangement of the angle 22, or the more advanced lighting assembly 20 and the coaxial lighting assembly 22 must be sufficiently fixed. In the field of expertise, if the lighting assembly 14 is spatially arranged, the semiconductor package 12 illumination, at the same time, the remote angle illumination assembly 2 can provide the semiconductor package 12, or direct the light to the preparatory position, ", each surface of the four side surfaces, in the manner The angle illumination assembly layer is in the front 揸+ & 战 20 20 层 layer LED device, each surface, into the bottom of the semiconductor package 12 In a feasible way, you can adjust or repair, 庶03. This is enough for the pre-lawful angle, so that ',, month can be I-V, conductor package. Another possible way to point to the half or the angle of the angle is to set the space of the LED device to the position of the LED device. The position of the semiconductor light is between 201005285. The heat should be fully avoided. v ^ m , the case of each of the four side surfaces. @J 4丁的大Θ 1 J is not the coaxial lighting
光線指向預供A 頂備位置18該半導體 °亥半導體封裝12檢測系統 也稱之為反光片總&，該稜鏡 表面。可行的方式是表面數量 據相互之間關係加以型態模組 表面數量之間經過相互定位， 一表面均預先詳細定義其特性 其中之一包括一種塗層，該塗 、傳輸與吸收速率。此外複數表面至少其中之一是由 光學塗膠製程（Optical Cementing Process)建造而成。 可行的方式是提供給或指向預備位置18該半導體 12的光源，至少其中一部份光源是透過該半導體封裝 表面反射之，同時可行的方式是反射自該半導體封裝 面光線的一部份，在離開ji進入至少攝影機其中之— 進入該稜鏡總成16。可行的方式是攝影機數量為2部 就是一部第一攝影機24與一部第二攝影機26(也稱為 第一影像捕捉設備與一個第二影像捕捉設備）。可行 式是該第一攝影機24與第二攝影機26係屬高解析度攝 ’該第一攝影機24與第二攝影機26解析度相當，或者 一攝影機24與該第二攝影機26具備不同解析度。 封裝 互干 或將 16， 反射 是依 式是 。每 至少 反射 一種 封裝 12的 12表 前， ，也 —個 的方 影機 該第 總成22主要在提供光線 封.裳12的底部表面。 10進一步包括稜鏡總成 總成16包括數個表面或 經過特殊安排配置，或 ’進一步一個可行的方 以律定其間的光線通道 或特色，例如表面數量 層具備預先詳細定義之 〇 11 第—攝影機24與第 二攝影機26能夠分 行影像捕捉作業 影機26能夠捕捉高解析度影像’ @此，具備高解析度 201005285 可行的方式是因為光線進入該 影機26 ’使得該第一攝影機24與第 具備尚解析度的第一攝影機24與第 -攝影機24與第二攝影機26使系統1〇更易執行表面瑕 測’同時也提升其精確性。 可行的方式是透過電腦導向系統與一個可編程控 (未顯示）其中之一或是其組合型態，控制第一攝影 ® 、第二攝影機26與照明總成14之作業。可行的方式是 編程控制器至少與角度照明總成2〇與同軸照明總成22 之一連結，或是與其中之一構成訊號通聯，該可編程 器控制並決定至少該角度照明總成2〇與同軸照明總成 中之一所提供光源之特性或特色，該角度照明總成2〇 軸照明總成22提供光源給預備位置18的該半導體封裝 該可編程控制器所控制光源特性包括但不限於亮度、 、顏色與照明肖度’例如該可編程控制器可依據預備 18該半導體封裝12底部表面所定義的參考平面，控制 LED設備每一層的角度，因此就能控制多層LED設備 預備位置18的該半導體封裝12的照明角度。 可行的方式是反射自該半導體設備12底部表面的 充分的與參考平面垂直，進一步一個可行方式是該半 設備12四個側邊表面每—表面都與參考平面垂直，反 該半導體設備12四個側邊表面每一表面的光線充分的 考平面平行。 —攝 別執 二攝 的第 疵偵 制器 機24 該可 其中 控制 22其 與同 12 〇 強度 位置 多層 指向 光線 導體 射自 與參 12 201005285 - 可控制角度照明總成20與同軸照明總成22，使該角度 …、月""成20與同軸照明總成22能夠同步提供該半導體封裝 12之昭明，式玉 '、、、 ％考可控制角度照明總成20與同軸照明總成22 -依據疋時間間隔輪流提供該半導體封裝12角度照明或 同軸，.、、明可編程控制器對照明總成14所提供光線各個不 同特（·生具備個別控制之能力，此外一個可行的方式是該控 制此夠决义並控制該照明總成14所提供光線的一組特性 ¥ 光是由基本 會呈現波與粒子 二元性現象。一 波，每一光束或 率，當光束撞擊 性、吸收性與反 正如前述， φ 發生該半導體封 導體封裝12表面 影機24或第二攝 可行的方式 前，反射自該半 16。可行的方式 循穿透該稜鏡總 方式是當光線最 鏡總成16的一個 粒 子 ( 也 就 是 所 謂 的 的 特 性 > 也 就 是 通 常 般 而 言 通 常 我 們 了 光 波 有 其 波 長 光 束 一 個 表 面 時 > 光 束 的 射 性 的 一 個 因 素 0 當 半 導 體 封 裝 12 位 於 裝 12 表 面 檢 測 或 影 像 的 光 線 ( 可 互 相 稱 之 影 機 26 時 > 影 像 即 被 是 在 進 入 第 攝 影 機 導體 封 裝 12 表 面 的 光 是 當 光 線 最 終 進 入 第 成 16 的 一 個 第 一 光 路 終 進 入 第 二 攝 影 機 26 第 JIL 光 路 〇 光子）所組成的，光 所謂的粒子與波動的 解光是一束光束或光 之波長決定光束的頻 頻率是決定光束傳輸 預備位置18時，即會 捕捉。當反射自該半 為光束）進入第一攝 捕捉。 24或第二攝影機26之 線應先穿過稜鏡總成 一攝影機24時，應遵 。進一步一個可行的 時’應遵循穿透該稜 13 201005285 ，該第 導體封 由一個 第一入 撞擊該 步 反射表 ❹層，可 分別為 輸與吸 實際需 反 擊某一 之直線 面反射 反 塗層反 第二反 步 過稜鏡 光線垂 射表面 學塗膠 個第—光路100流程實 一光路100步驟U0中 裝12底部表面與四個側 第一入口表面50進入棱 口表面50塗布一層抗反 第一入口表面50光線之 驟120中，進入該稜鏡 面52反射出來，該第二 行的方式是該特殊塗層 50% 、50% 與 0%，進一 收比率是可變動的，其 要可對反射、傳輸與吸 射定律說明當某一入射 介面時，該角度係測量 間的角度，且該入射光 出去，換句話說光束之 射自該第二反射表面52 射、傳輸與吸收比率而 射表面52光線之反射角 驟130中，將反射自該 總成16内部的一個第三 直撞擊該第三反射表面 54反射出去’可行的方 製程製造之，另外一個 所示 讀半 ，緩 是該 降低 第二 殊塗 率應 、傳 依據 之。 )揸 垂直 該介 特殊 自該 導通 式是 三反 以光 膠製 施 例 如 第 4 圏與 第 5 圖 反 射 白 預 備位 置 18 的 邊 表 面 每 -—* 表面 的 先 綠 鏡 總 成 16 0 可行 的 方 式 射 塗 層 以 防止 或 充 分 反 射 0 總 成 16 的 光 線會 由 個 反 射 表 面 52 塗佈 層 特 其 反 射 、 傳 輸與 吸 收 th 步 — 個 可 行 方式 是 反 射 容 忍 誤 差 達 +/- 5°/〇 » 或 收 比 率 實 施 調整 或 修 訂 光 束 以 j 角 度（ 入 射 角 該 介 面 與 某 一與 該 介 面 束 也 會 以 相 同/ 角 度自 入 射 角 等 於 其反 射 角 〇 光 束 數 量 的 多募 端 視 該 定 可 行 的 方式 是 反 射 等 於 其 入 射 角。 第 二 反 射 表 面的 光 線 引 反 射 表 面 54 ，可 行 的 方 54 ， 以 防 止 光線 自 該 第 式 是 該 第 三 反射 表 面 54 可 行 的 方 式 是由 光 學 塗 14 201005285 -程所製造之表面兩側的屈光指數相等或相告 牙過第二反射表面54的光線，則會在 稜鏡總成16内部一個第四反射表面郎反射出驟140 式是光線應以某—角度（例如入射角）導入來可 面56 ’該第四反射表面56應塗布—層與第_ I第四 同之特殊塗層，更進一步一個可行 反射表 面56特殊塗層之反射、傳輸與吸收比率應與X第四 52之特殊塗層相當，或者該第四反射 反 ^ ^ ^ 町表面沾特殊塗 射、傳輸與吸收比率與第二反射表面52之 φ 叫之特殊塗層 第二反射表面52與第四反射表面56均可傲盔、 j j做為分光器 是說導向該第二反射表面52與該第四反射表面邛的 能夠以數道光束自該表面反射出去，而每道光束均 同的路徑或方向。 可行的方式是反射自該第四反射表面56的光線 射角應等於入射角。反射自該第四反射表面56光束 數量的多募端視該特殊塗層反射、傳輸與吸收比率 可行的方式是將反射自該第四反射表面56的光 一個第五反射表面58，該第五反射表面58應具備一 ，其反射率大於92% ’或者可行的方式是該鏡面具 不同的反射率數值’如果增加該第五反射表面58鏡 射率，則會增加撞擊該處光線被反射出去的百分率 第一光路100步驟150中，導向該第五反射表 光線會反射出去’可行的方式是光線自該第五反射 反射出去的角度應等於其入射角。 步驟160中，自該第五反射表面58反射出去的 中，自 行的方 反射表 面52相 反射表 射表面 層之反 不同。 ，也就 光線， 導向不 ，其反 或光線 而定。 線導向 個鏡面 備一個 面之反 〇 面58的 表面58 光線會 15 201005285 經過一個第一出口表面60離開稜鏡總成16。可行的方式是 該第一出口表面60其特性應與第一表面50相同或相當’也 就是說該第一出口表面60應具備抗反射塗層’以防止或降 低光線的反射，或者該第一出口表面60的特性與該第一表 面50的特性不同。 在最終步驟170中，進入第一攝影機24之前，經該第 一出〇表面60離開稜鏡總成16的光線會穿過一個第一攝影 機鏡頭62，可行的方式是該第一攝影機24定義一個第一影 像 捕 捉 平 面 9 該 第 與 該 第 一 攝 影 機 24 式 是 該 第 攝 影 機 充 分 固 定 之 可 行 件 距 離 配 備 一 個 長 預 先 律 定 的 焦 長 > 素 ， 可 行 的 方 式 是 術 > 對 該 第 一 攝 影 可 行 的 方 式 是 大 率 J 可 讓 反 射 白 與 四 個 側 邊 表 面 每 方 式 疋 該 第 一 攝 影 就 是 具 備 捕 捉 預 備 側 邊 表 面 每 一 表 面 可 個 第 二 光 路 行 的 方 式 第 一 0 應 分 別 與 第 — 光 相 似 〇 因 此 在 步 驟The light is directed to the pre-supplied A-mount position 18 of the semiconductor. The semiconductor package 12 detection system is also referred to as the retroreflective sheet total & A possible way is to compare the surface quantities according to the relationship between the types of modules. The number of surfaces is mutually positioned, and a surface is defined in advance in detail. One of them includes a coating, the rate of coating, transmission and absorption. In addition, at least one of the plurality of surfaces is constructed by an optical cementing process. A possible way is to provide or point the light source of the semiconductor 12 to the preliminary position 18, at least a part of the light source is reflected through the surface of the semiconductor package, and a feasible way is to reflect a part of the light from the semiconductor package surface. Leave ji to enter at least the camera - enter the 稜鏡 assembly 16. A feasible method is that the number of cameras is two, that is, a first camera 24 and a second camera 26 (also referred to as a first image capturing device and a second image capturing device). It is possible that the first camera 24 and the second camera 26 are high-resolution cameras. The first camera 24 and the second camera 26 have the same resolution, or the camera 24 and the second camera 26 have different resolutions. The package is either coherent or the reflection of 16, is based on . Each of the 12 front and back of each of the 12 packages of the package 12 is reflected. The first assembly 22 is primarily provided with a light seal to the bottom surface of the skirt 12. 10 further includes a 稜鏡 assembly assembly 16 comprising a plurality of surfaces or specially arranged configurations, or 'further a feasible side to define a ray channel or feature therebetween, such as a surface number layer having a pre-detailed definition 第 11 - The camera 24 and the second camera 26 can branch the image capturing operation. The camera 26 can capture the high resolution image. This is possible because the high resolution 201005285 is feasible because the light enters the camera 26' such that the first camera 24 and the first camera 24 The first camera 24 and the first camera 24 and the second camera 26, which are still in resolution, make the system 1 〇 easier to perform surface surveys' while also improving their accuracy. A feasible way is to control the operation of the first camera ® , the second camera 26 and the illumination assembly 14 through one of the computer-guided system and a programmable control (not shown) or a combination thereof. The feasible way is that the programming controller is connected to at least one of the angle illumination assembly 2 and the coaxial illumination assembly 22, or is connected to one of the signals, and the programmable device controls and determines at least the angle illumination assembly. The characteristics or characteristics of the light source provided by one of the coaxial illumination assemblies, the angle illumination assembly 2 x-axis illumination assembly 22 provides the light source to the semiconductor package of the preparatory position 18, the programmable light source controlled by the programmable controller includes but not Limited to brightness, color, and illumination, 'For example, the programmable controller can control the angle of each layer of the LED device according to the reference plane defined by the bottom surface of the semiconductor package 12, thereby controlling the multi-layer LED device preparation position 18 The illumination angle of the semiconductor package 12. A possible way is to reflect from the bottom surface of the semiconductor device 12 substantially perpendicular to the reference plane. Further, it is possible that the four side surfaces of the half device 12 are perpendicular to the reference plane, and the semiconductor device 12 is opposite. The light on each surface of the side surface is sufficiently parallel to the test plane. - Photographing the second detector of the second detector, which can control 22 and its 12-inch intensity position, multi-layer pointing light conductors from the reference 12 201005285 - controllable angle illumination assembly 20 and coaxial illumination assembly 22, The angle ..., the month "" into 20 and the coaxial illumination assembly 22 can simultaneously provide the semiconductor package 12, the type of jade ',,, the % control angle illumination assembly 20 and the coaxial illumination assembly 22 - The semiconductor package 12 is provided with angular illumination or coaxial according to the time interval of the turn, and the programmable controller has the ability to provide individual control of the light provided by the illumination assembly 14 (in addition, a feasible way is Controlling this set of characteristics that are sufficient to determine and control the light provided by the illumination assembly 14 is essentially a phenomenon in which the wave and the particle are dualistic. One wave, each beam or rate, when the beam is impinging, absorptive And contrary to the foregoing, φ is reflected from the half 16 before the semiconductor package conductor 12 surface camera 24 or the second camera is feasible. The way is when the light is the most mirrored component of a particle 16 (also known as the characteristic > is usually the usual light wave has a surface of its wavelength beam > a factor of the beam's radiance 0 when the semiconductor The package 12 is located on the surface of the 12-detection or image (when it can be called the camera 26). The image is the light that enters the surface of the camera conductor package 12 when the light finally enters the first optical path of the first 16 Entering the second camera 26, the JIL optical path photon), the so-called particle and wave of the light is a beam or the wavelength of the light determines the frequency of the beam is determined when the beam transfer preparation position 18 is captured. When the reflection from the half is the beam) enters the first camera capture. 24 or the line of the second camera 26 should first pass through the 稜鏡 assembly a camera 24, should be followed. Further feasible when the 'should follow the rib 13 201005285, the first conductor seal A first intrusion strikes the reflective surface layer of the step, which can respectively be used for the transmission and suction, and the actual need to counterattack a certain linear surface reflection, the reverse coating, the second reverse step, the light, the vertical surface, the coating, the first light path 100 The process of the optical path 100, the step U0 of the bottom surface 12 and the four side first inlet surfaces 50 enter the edge surface 50 and apply a layer of light 120 against the first inlet surface 50, and enter the surface 52 to be reflected. The second line is 50%, 50% and 0% of the special coating. The ratio of the first to the second is variable. It is necessary to explain the reflection, transmission and the law of the absorption when the incident interface is measured. Between the angles, and the incident light exits, in other words, the beam is incident from the second reflective surface 52, the transmission and absorption ratio is reflected, and the reflection surface angle 130 of the surface 52 is reflected from the interior of the assembly 16. A third direct impact on the third reflective surface 54 is reflected out of the 'feasible square process manufacturing, and the other one is read half, which is the basis for reducing the second special coating rate.揸 该 该 该 特殊 特殊 特殊 特殊 特殊 特殊 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸 揸Shooting the coating to prevent or adequately reflect the light of the 0 assembly 16 will be reflected, transmitted and absorbed by the reflective layer 52 coating layer. A possible way is to have a reflection tolerance of +/- 5°/〇» or The ratio is adjusted to adjust or revise the beam at j angle (incident angle of the interface with a certain angle with the interface beam also from the angle of incidence equal to the angle of reflection of the beam) Equal to its angle of incidence. The light-reflecting surface 54 of the second reflecting surface, the feasible side 54 to prevent light from being self-contained is the third reflecting surface 54 is feasible by the surface of the optical coating 14 201005285 If the side's refractive index is equal or the light passing through the second reflecting surface 54 is reflected, a fourth reflecting surface inside the cymbal assembly 16 is reflected. The light is supposed to be at an angle (for example, the angle of incidence). The surface of the fourth reflecting surface 56 should be coated with a layer and the special coating of the fourth layer, and the reflection, transmission and absorption ratio of the special coating of the viable reflective surface 56 should be the same as X The special coating of the fourth 52 is equivalent, or the fourth reflective surface is coated with a special coating, transmission and absorption ratio and the second reflecting surface 52 is called a special coating second reflecting surface 52 and fourth reflection The surface 56 can be proud of the helmet, jj as a beam splitter, which means that the second reflective surface 52 and the fourth reflective surface can be reflected from the surface by a plurality of beams, and each beam has the same path or direction. A feasible way is that the angle of incidence of the light reflected from the fourth reflecting surface 56 should be equal to the angle of incidence. The multi-reporting of the number of beams reflected from the fourth reflecting surface 56 depends on the manner in which the reflection, transmission and absorption ratio of the special coating is feasible. Is going to The light from the fourth reflective surface 56 is a fifth reflective surface 58 which should have a reflectivity greater than 92% 'or a feasible way to have different reflectance values for the mirror mask' if increased The fifth reflecting surface 58 mirroring rate increases the percentage of the first light path 100 in which the light is reflected. The first light path 100 in the step 150 is performed, and the light directed to the fifth reflecting surface is reflected. The feasible way is that the light is from the fifth. The angle at which the reflection is reflected should be equal to its angle of incidence. In step 160, the reflection from the fifth reflective surface 58 is different from the reflection of the reflective surface layer of the self-reflecting surface 52. , that is, the light, the guidance is not, its inverse or the light depends. The line guides the mirror surface. The surface 58 of the surface 58 is illuminating. 15 201005285 A first exit surface 60 exits the cymbal assembly 16. It is possible that the first outlet surface 60 should have the same or equivalent characteristics as the first surface 50. That is, the first outlet surface 60 should be provided with an anti-reflective coating to prevent or reduce the reflection of light, or the first The characteristics of the exit surface 60 are different from the characteristics of the first surface 50. In a final step 170, prior to entering the first camera 24, light exiting the cymbal assembly 16 via the first exit surface 60 will pass through a first camera lens 62, possibly by defining the first camera 24 The first image capturing plane 9 is first and the first camera 24 is a feasible distance of the camera that is sufficiently fixed to be equipped with a long pre-determined focal length > prime, and the feasible way is to perform the first photography The feasible way is that the large rate J can be reflected white and the four side surfaces each way. The first photography is to have a second light path for each surface of the preparatory side surface. The first 0 should be respectively and the first Light is similar, so in the steps
一攝影機鏡頭62應在 之間進行空間配置， 24與該第一攝影機鏡 的方式是該第一攝影 鏡頭，該第一攝影機 以達到一個預先律定 必要時可運用本領域 機鏡頭62以及放大率 受該第一攝影機鏡頭 預備位置18的該半導 一表面的光線進入攝 機24具備捕捉多重表 位置18該半導體封裝 影像的能力。 200的實施例如第6 光路200步驟210 、 路100步驟110 、步 210中’反射自預備 該第一出口表面60 進一步一個可行方 頭62之空間配置應 機鏡頭62應依據物 鏡頭62應具備一個 的放大率或放大因 專業人士熟知的技 進行調整修訂之。 62控制的焦長與放 體封裝12底部表面 影機。因此可行的 面影像的能力，也 12底部表面與四個 圖與第7圖所示。 步驟220與步驟23 驟120與步驟130 位置18該半導體封 16 201005285 裝 12 底 部 表 〇 然 後 在 步 二 反 射 表 面 二 反 射 表 面 在 第 —— 過 第 — 反 射 行 的 方 式 是 第 四 反 射 表 可 行 的 別 為 50% ·> 9 其 中 50% 在 步 驟 會 經 過 一 個 是 該 第 二 出 或 相 當 j 也 層 > 以 防 止 64 的 特 性 與 第 二 光 離 開 稜 鏡 總 可 行 的 方 式 > 該 第 二 攝 二 攝 影 機 26 與 該 第 二 攝 是 該 第 _ 攝 此 外 該 第 二 mA camera lens 62 should be spatially arranged between, and the first camera lens is the first photographic lens. The first camera can use the machine lens 62 and the magnification when a predetermined law is reached. Light entering the camera 24 by the semi-conductive surface of the first camera lens ready position 18 has the ability to capture the multi-table position 18 of the semiconductor package image. For example, the sixth lens 200 step 210, the road 100 step 110, the step 210, the reflection from the preparation of the first exit surface 60, the space of the further feasible square 62, the lens 61 should be provided according to the object lens 62. The magnification or magnification is adjusted and revised by techniques well known to the professional. 62 control of the focal length and the bottom surface of the package 12 is the camera. The ability to make a face image is therefore also 12 bottom surface with four figures and Figure 7. Step 220 and Step 23 Step 120 and Step 130 Position 18 The semiconductor seal 16 201005285 is equipped with 12 bottom surface 〇 then in step 2, the reflective surface is reflected on the surface, and the second reflection table is feasible. 50% ·> 9 of which 50% will pass in the step is the second out or equivalent j also layer> to prevent the characteristics of 64 from leaving the second light 稜鏡 always feasible> Second photo two The camera 26 and the second camera are the first camera and the second camera
面的光線， 驟220中， 52反射出去 52的光線會 光路200步 表面54的光 傳輸通過第 面56特殊塗 方式是該特 50% 與 ， 會傳輸通過 25〇中，傳 第二出口表 口表面64的 就*是說該第 或充分降低 該第一出口 路2〇〇步驟 成1丨6的光線 是镑第二攝 I機鏡頭66 之間。進一 影機輓頭66 影機鏡頭66 #影機鏡頭 表面50進入 於入射角的 驟230中， 三反射表面 在稜鏡總成 至第四反射 56的光線百 傳輸與吸收 射、傳輸與 第四反射表 表面56。 四反射表面 鏡總成16。 出口表面60 64應塗布一 ’或者該第 性不同。 進入第二攝 個第二攝影 一個第二影 第二出口表 方式是該第 分固定之， 距離配備一 個預先律定 會經過第一 光線會以等 。接著在步 前進經過第 驟240中， 線會被傳輸 四反射表面 層的反射、 殊塗層的反 因此撞擊該 該第四反射 輸經過該第 面64離開稜 特性與第— 一出口表面 光線的反射 表面60的特 260中，在 會先通過― 影機26定義 應配置在該 步一個可行 之配置應充 應依據物件 66應具備一 稜鏡總成16 角度，自第 反射自該第 54。 16内部，通 表面56 ，可 分率端視該 比率而定。 吸收比率分 面56的光束 56的光線， 可行的方式 的特性相同 層抗反射塗 一出口表面 影機之前， 機鏡頭66。 像捕捉平面 面64與該第 二攝影機26 可行的方式 個長鏡頭， 的焦、長，以 201005285 - 達到一個預先律定的放大 般’可行的方式是必要時 術’對該第二攝影機鏡碩 .之。 可行的方式是該第二 影機鏡頭62的放大率不同 鏡頭66控制的焦長與放大 導體封裝12底部表面的光 第二攝影機26具備捕捉預 面影像的能力，更具體的 可行的方式是該第二攝影 機鏡頭62的放大率，因此 像而言’第二攝影機26具 由於第一攝影機鏡頭 大率不同’使得系統1〇能 。本領域專業人士應當了 的 數 量 ， 而 且 每 — 鏡 頭 焦 總 成 16 表 面 之 數 量 則 系 不 同 放 大 率 與 視 界 的 影 像 可 行 的 方 式 是 系 統 10 分 的 鄰 接 配 置 並 分 別 與 64 對 齊 0 該 第 一 攝 影 機 24 空 間 決 定 該 棱 鏡 總 成 16 第 之 間 所 需 的 距 離 與 空 間 〇 與 現 行 半 導 體 封 裝 12 的 系 統 10 之 安 裝 較 為 節 省 。正如同第一攝影機鏡頭62 — 運用本領域專業人士熟知的技 之焦長與放大率進行調整修訂 影機鏡頭66的放大率與第—攝 可行的方式是受該第二攝影機 ’僅讓反射自預備位置18該半 進入該苐一攝影機26。因此該 位置18該半導體封裝12單一表 是捕捉底部表面影像的能力。 鏡頭66的放大率大於第一攝影 該半導體封裝12底部表面的影 更尚的影像放大率。 與第二攝影機鏡頭66焦長與放 捕捉多重放大率與視界的影像 ’如果增加系統10攝影機鏡頭 不同’同時相對應地增加稜鏡 10就能夠捕捉為數可觀的各種 一攝影機24與第二攝影機26充 一出口表面60與第二出口表面 第二攝影機安排配置所需實體 反射表面56與第五反射表面58 測系統相較，採用稜鏡總成16 間。進一步一個可行的方式是 18 201005285 - 該稜鏡總成16的第一表面50、第二反射表面52、第三反射 表面54、第四反射表面56與第五反射表面58之空間配置應 牛固不動，s玄稜鏡總成16上述各表面的穩定性有助於消除 或充分降低校準或調整回饋之遺失，現行採用移動式鏡面 與鏡頭系統常見的問題就是校準或調整回饋之遺失。 本發明第二個實施例半導體封裝12檢測的方法如第η 圖所示’可行的方式是該方法300使用系統10。 .步驟310中’半導體封裝12置於預備位置18。可行的 方式疋§亥預備位置18應配置在系統1〇側視稜鏡結構68之中 0 ’該側視稜鏡結構68包括數個表面或鏡面，更具體而言就 是四個表面。可行的方式是該預備位置18位於該側視稜鏡 結構68四個表面所界定的範圍之内。可行的方式是該侧視 稜鏡結構68四個表面在接收反射自該半導體封裝12四個側 邊表面的光線。 步驟320中，照明總成會提供一個第一閃光或照明， 該第—閃光或照明指向該半導體封裝12，可行的方式是該 第一閃光會提供預備位置18該半導體封裝12底部表面以及 ^ 四個側邊表面之照明。 可仃的方式是該第-閃光具備一乡且預先定義之特性， 該組預先定義之特性由控制器決定或控制。正如前述，該 組特性包括-個特定的亮度或強度、顏色與照明角度。例 Τ :玄 '、且特性包括一個亮纟5〇%、紅色、照明角度5。度等。 可行的方式是必要時，&夠對該組特性内 施修訂之。 特疋篩選，以強調該半 本領域專業人士應當了 可行的方式是該組特性應經過 導體封裝12特定型態之表面瑕疵， 19The light of the surface, in step 220, 52 reflects the light out of 52. The light transmission on the surface 54 of the optical path 200 passes through the first surface. The special coating method is 50% of the special coating, which is transmitted through the 25 ,, and the second exit surface is transmitted. The surface 64 is said to be the first or the lowering of the first exit path 2, and the light of the step 1 is 6 is between the second camera lens 66. Into a camera head 66 camera lens 66 #影机 lens surface 50 into the incident angle of the step 230, the three reflective surface in the 稜鏡 assembly to the fourth reflection 56 of the light transmission and absorption, transmission and fourth Reflecting surface 56. Four reflective surface mirror assembly 16. The exit surface 60 64 should be coated with a ' or the same. Entering the second camera, the second camera, the second film, the second exit table, the method is that the first point is fixed, and the distance is equipped with a pre-law to pass the first light to wait. Then, as the step advances through the second step 240, the line is transmitted by the reflection of the four reflective surface layer, and the opposite of the coating, thereby striking the fourth reflection through the first surface 64 and leaving the edge characteristic and the first exit surface ray. The special surface 260 of the reflective surface 60, which is first defined by the "mechanical machine 26", should be configured in this step. A feasible configuration should be based on the object 66 should have a 稜鏡 assembly 16 angle, since the first reflection from the 54th. 16 internal, through surface 56, the rate can be determined depending on the ratio. The absorption ratio of the light beam 56 of the facet 56 is the same as that of the viable manner. The layer is anti-reflective coated with an exit surface before the camera lens 66. Like the long plane of the capture plane 64 and the second camera 26 is feasible, the focal length, with the 201005285 - reaching a pre-determined magnification, the 'feasible way is necessary when the surgery' . The feasible manner is that the magnification of the second camera lens 62 is different from the focal length controlled by the lens 66 and the light of the bottom surface of the amplifying conductor package 12. The second camera 26 has the ability to capture the pre-image, and a more specific feasible manner is The magnification of the second camera lens 62, so that the 'second camera 26 has a different rate of the first camera lens', causes the system 1 to be disabled. The number of professionals in the field should be, and the number of surfaces per lens focal length 16 is different from the magnification and view of the field of view. The way the system is 10 points adjacent to each other and aligned with 64 respectively. The first camera 24 The space determines the distance and space required between the prism assembly 16 and the installation of the system 10 of the current semiconductor package 12 is relatively economical. As with the first camera lens 62, the magnification and magnification of the camera lens 66 are adjusted by using the focal length and magnification that are well known to those skilled in the art, and that the second camera is only allowed to reflect from the second camera. The preparatory position 18 enters the first camera 26 halfway. Thus, the location 18 of the semiconductor package 12 is a single watch that captures the image of the bottom surface. The magnification of the lens 66 is greater than the first image of the bottom surface of the semiconductor package 12, and the image magnification is more. With the second camera lens 66 focal length and the image capturing the multiple magnification and the field of view 'if the system 10 camera lens is increased', the corresponding increase of 稜鏡10 can capture a considerable number of cameras 24 and 2. The charge-out surface 60 and the second exit surface are arranged in a second camera arrangement with a desired solid reflective surface 56 and a fifth reflective surface 58. A further feasible way is 18 201005285 - the spatial arrangement of the first surface 50, the second reflective surface 52, the third reflective surface 54, the fourth reflective surface 56 and the fifth reflective surface 58 of the cymbal assembly 16 Immobility, s Xuanzi assembly 16 The stability of the above surfaces helps to eliminate or substantially reduce the loss of calibration or adjustment feedback. The current common problem with moving mirrors and lens systems is the loss of calibration or adjustment feedback. The method of detecting the semiconductor package 12 of the second embodiment of the present invention is as shown in the figure η. A possible way is that the method 300 uses the system 10. The semiconductor package 12 is placed in the preliminary position 18 in step 310. A viable manner 疋 预备 预备 预备 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 It is possible that the preliminary position 18 is within the range defined by the four surfaces of the side view structure 68. It is possible that the four surfaces of the side view germanium structure 68 receive light reflected from the four side surfaces of the semiconductor package 12. In step 320, the illumination assembly provides a first flash or illumination that is directed toward the semiconductor package 12. It is possible that the first flash will provide the preliminary position 18 of the bottom surface of the semiconductor package 12 and Illumination of the side surfaces. The embarrassing way is that the first-flash has a predefined and pre-defined characteristic, and the set of predefined characteristics is determined or controlled by the controller. As mentioned earlier, this set of characteristics includes - a specific brightness or intensity, color and illumination angle. Example Τ : Xuan ', and features include a bright 纟 5〇%, red, lighting angle 5. Degrees, etc. The feasible way is to & it is necessary to revise the characteristics of the group. Special screening to emphasize that the half of the field should be feasible in a way that the set of characteristics should pass through the surface of the conductor package 12, 19
201005285 解運用特定特性之光線’即可很輕易或精確地偵測出 體封裝12特定之表面瑕疵。 指向預備位置18該半導體封裝12的第一閃光，可 半導體封裝12底部表面與四個側邊表面每一表面反射 。可行的方式是自該半導體封裝12底部表面反射出來 線，應與反射自該半導體封裝12四個侧邊表面每一表 光線成垂直狀態。 步驟330中’第一攝影機24捕捉到該半導體封裝 一個第一影像。可行的方式是該第一攝影機24曝光時 先律s丁’使光線此夠進入第一攝影機24，以利該第_ 機24捕捉影像。可行的方式是光線經由第一光路ι〇〇 第一攝影機24。 可行的方式是該第一攝影機24所捕捉的影像來自 位置18該半導體封裝12多重表面的影像，也就是說該 攝影機24所捕捉的影像包括來自預備位置18該半導體 12底部表面與四個側邊表面每一表面的影像。此外可 方式是所捕捉的第一影像由預先定義的第一閃光特性 其特色。 步驟340中，照明總成提供一個第二閃光，該第 光指向預備位置18該半導體封裝12。可行的方式是該 閃光具備與第一閃光相同的預先定義的特性組，或者 二閃光具備與第一閃光不同的預先定義的特性組。 ^步驟350中，第二攝影機26捕捉到一個第二影像 灯的方式是該第二攝影機26曝光時間預先律訂， 夠進入第二攝影機26，正如同第一攝影機％ 半導 自該 出去 的光 面的 12的 間預 攝影 進入 預備 第一 封裝 行的 突顯 二閃 第二 該第 線能 利該 20 201005285 - 第二攝影 200進入 可行 置18該半 機26所捕 表面的一 可行 定義的第 像分享共 © 可行 其特性， 色與照明 性。 正如 组影像。 捕捉不同 ❹時，照明 。正如前 ，其目的 可行 26所捕捉 編程控制 式是可編 構成訊號 至可編程 機 26 捕 捉 影 像 0 第 二 攝 影 機 26 0 的 方 式 是 該 第 二 導 體 封 裝 12 單 一 捉 的 影 像 僅 屬 於 個 影 像 0 的 方 式 0 疋 所 捕 捉 二 閃 光 特 性 突 顯 同 的 一 組 預 先 定 的 方 式 是 每 組 影 例 如 每 組 影 像 内 角 度 > 或 者 每 組 上 述 y 透 過 執 行 可 行 的 方 式 是 必 組 別 的 影 像 〇 可 總 成 所 提 供 的 光 述 ， 提 供 給 該 半 乃 在 提 升 不 同 表 的 方 式 是 下 載 或 的 第 一 影 像 與 第 器 或 控 制 器 9 以 程 控 制 器 應 與 該 通 聯 ， 可 行 的 方 控 制 器 0 可 行 的 可行的方式是光 攝影機26所捕捉 表面的影像，也 預備位置18該半 的第二影像以及 其特色，因此第 義特性，以下稱 像内的影像都由 之個別影像分享 影像内之個別影 步驟320〜350以捕 要時可反覆執行 行的方式是每次 線之特性組具備 導體封裝12照明 面瑕疲彳貞測之效 傳輸第一攝影機 二影像（或是第 利在步驟360分 第一攝影機24與 式是以影像訊號 方式是可編程控 線經由第二光路 &景彡像為預備位 就是說第二攝影 導體封裝12底部 胃—影像由預先 二影像與第一影 之為第一組影像 相同特性組突顯 相同的亮度、顏 像分享不同的特 捉或獲得該第一 步驟320〜350，以 執*行步驟320〜350 可變性或選擇性 特性之變動修訂 率 〇 24或第二攝影機 一組影像）至可 析之。可行的方 該第二攝影機26 形式將影像下載 制器經過特定編 21 201005285 ❹ 程’以利 人士所熟 進行處理 像二值化 之表面瑕 半導體封 能力。 該半 攝影機26 行的方式 捉一系列 第一攝影 影像交互 26交互捕 二攝影機 一般 提尚攝影 輸時間對 效率有極 方法300 .因此 捕捉影像 導體封裝 雖然 本領域專 攝影機24 像。可行的方式是運用本領域專業 算，對下載至可編程控制器的影像 像差異法、負片法（negativing)、影 測以利鑑識或偵測該半導體封襄12 個可行的方式是運用系統10檢測該 程控制器具備自動偵測表面瑕寂的 分析所 知的方 作業， ，或是 疲。進 裝12， 導體封 具備捕 是該第 連續影 機24與 捕捉作 捉影像 26影像 而言， 機解析 第一攝 大的影 之全盤 ，相當 之措施 12檢消ij 本發明 業人士 與第二 下載影 法與演 例如影 邊緣偵 一步_ 該可編 裝12檢 捉—系 一攝影 像，可 第二攝 業之遂 之運用 捕捉時 影像捕 度，則 影機24 響，進 效率。 明顯的 ’可降 處理速 實施例 應當了 攝影機 測方法 列連續 機24與 行的方 影機26 行，該 ，有助 資訊傳 捉過程 資料傳 與第二 而影響 300讓 影像的 第二攝 式是該 於降低 輸延遲 之資訊 輸的時 攝影機 半導體 第一攝影機24與第 能力，進 影機26以 可編程控 交互曝光之同步 第一攝影機24與 該第一攝 之衝擊。 傳輸需要 間就會增 26捕捉影 封裝12檢 一步一個可 交互方式捕 制器控制該 作業，以利 第二攝影機 影機24與第 時間，如果 長，資料傳 像之速度與 測系統10與 第一攝影機24與第二攝影機26交 低資訊傳輸延遲之衝擊，有助於 度，進而提升產能。 運用時間協同交互影像捕捉方法 解運用本發明所提供系統1〇，第 26也具備同步捕捉影像的能力。 22 201005285 . 上述實施例所揭示者係藉以具體說明本發明，且 雖透過特定的術語進行說明，當不能以此限定本發明 利範圍；熟悉此項技術領域之人士當可在瞭解本發明 神與原則後對其進行變更與修改而達到等效之目的， 等變更與修改，皆應涵蓋於如后所述之申請專利範圍 定範疇中。 〇 文中 之專 之精 而此 所界 23 201005285 - 【圖式簡單說明】 第1圖：本發明之實施例一，係顯示一半導體封裝檢測系 統之空間系統型態模組。 第2圖：承第1圖所示之系統角度照明總成與側邊稜鏡結 構之正剖面視圖。 （、 第3圖：承第2圖所示之側邊棱鏡結構部份放大正剖面視 圖與光線的重新定向示意圖。 第4圖：光皞沿著第1圖所示系統中該半導體封裝與一第 一攝影機之間一實施例之第一光路前進流程圖。 〇 第5圖：係光線沿著第4圖所示實施例第一光路前進之路 線示意圖。 第6圖：係光線沿著第1圖所示系統中該半導體封裝與一 個第二攝影機之間一個實施例第二光路前進的流 程圖。 第7圖：係光線沿著第6圖所示該實施例第二光路前進之 路線。 第8圖：係光線沿著第4圖所示該實施例第一光路前進與 Ο 第6圖所示該實施例第一光路前進路線之組合圖 示。 第9圖：係顯示第1圖所示系統第一攝影機與第二攝影機 所捕捉之影像。 第10圖：係光線穿過該系統稜鏡總成前進之路線，以利檢 測第1圖所示半導體封裝。 第11圖：係依據本發明一個第二實施例，顯示一個半導體 封裝檢測方法之流程圖。 【主要元件符號說明】 24 201005285201005285 The use of light with specific characteristics can easily or accurately detect the specific surface flaws of the package 12. The first flash of the semiconductor package 12 is directed to the ready position 18, and the bottom surface of the semiconductor package 12 and each of the four side surfaces are reflective. A possible way is to reflect the line from the bottom surface of the semiconductor package 12 and to be perpendicular to each of the four sides of the semiconductor package 12. In step 330, the first camera 24 captures a first image of the semiconductor package. It is feasible that the first camera 24 is exposed to allow the light to enter the first camera 24, so that the camera 24 captures the image. A possible way is to pass the light through the first optical path to the first camera 24. It is possible that the image captured by the first camera 24 is from the position 18 of the multiple surface of the semiconductor package 12, that is, the image captured by the camera 24 includes the bottom surface and four sides of the semiconductor 12 from the preliminary position 18. An image of each surface of the surface. In addition, the first image captured is characterized by a predefined first flash characteristic. In step 340, the illumination assembly provides a second flash that is directed toward the semiconductor package 12 at the ready position 18. It is possible that the flash has the same predefined set of characteristics as the first flash, or the two flashes have a predefined set of characteristics different from the first flash. In step 350, the second camera 26 captures a second image light in such a manner that the second camera 26 is pre-adjusted by the exposure time to enter the second camera 26, just as the first camera % is semi-guided from the outgoing light. The pre-photography of the face 12 enters the preliminary first package line and highlights the second flash. The second line can benefit the 20 201005285 - the second picture 200 enters a feasible definition of the surface of the half-machine 26 Share a total of possible characteristics, color and lighting. Just like group images. Capture different ❹ when lighting. As before, the purpose of the 26 programmable programming control is to encode the signal to the programmable machine 26 to capture the image 0. The second camera 26 0 is the second conductor package 12. The single captured image belongs to only one image 0. 0 疋 The captured two flash characteristics highlight the same set of predetermined methods: each set of images, for example, the angle of each set of images> or each set of y above is implemented by a feasible method that is required by the image set. The way of providing the half is to improve the different tables in the way that the first image is downloaded or the controller or controller 9 should be connected to the controller. The feasible controller 0 is feasible. The image captured by the light camera 26 also prepares the second image of the half of the position 18 and its characteristics. Therefore, the first feature, the image in the image below, is shared by the individual image in the individual image steps 320 to 35. The way to repeat the execution of the line is to transmit the first camera image of the conductor package 12 illumination surface after each characteristic line of the line (or the first camera 24 in step 360). The mode is that the image signal is the programmable control line via the second optical path & the image is the preparatory position, that is, the bottom of the second imaging conductor package 12 is the first image of the first image and the first image. The same characteristic group highlights the same brightness, the image sharing different special capture or obtain the first step 320~350, to perform the step 320~350 variability or the change characteristic of the selective characteristic 修订24 or the second camera one Group image) to the analysis. Feasible party The second camera 26 forms the image downloader through a specific process of processing, such as binarization of the surface 半导体 semiconductor package capability. The semi-camera 26-line approach captures a series of first photographic image interactions. 26 Interactive capture of two cameras. General photographic still-transfer time-efficient method 300. Therefore, the captured image conductor package is similar to the camera in the field. The feasible way is to use the professional calculation in the field, the video image difference method, the negative method (negativing), the shadow test downloaded to the programmable controller to identify or detect the semiconductor package. The test controller has the ability to automatically detect surface deadness, or to fatigue. Into the package 12, the conductor seal is equipped with the capture of the continuous camera 24 and capture the image of the capture image 26, the machine analyzes the first shot of the whole film, the equivalent of 12 measures to eliminate ij the invention industry and the second Download the shadow method and perform the shadow edge detection step _ This can be edited and captured 12 - a photographic image, which can be used in the second shooting industry to capture the image capture, then the camera 24 rings, the efficiency. The obvious 'degradable speed processing example should have the camera measuring method of the continuous machine 24 and the line of the square machine 26 lines, which helps the information transmission process data transmission and the second influence 300 to make the image second shot The first camera 24 and the first capability of the camera semiconductor are used to reduce the transmission delay. The first camera 24 and the first camera are synchronized by the programmable exposure of the camera 26. Between the transmission needs will increase 26 capture shadow package 12 check one step interactive mode capture device to control the operation, in order to benefit the second camera camera 24 and the first time, if long, data transmission speed and measurement system 10 and A camera 24 and the second camera 26 have a low impact on the information transmission delay, which contributes to the degree of productivity. The use of time-coordinated interactive image capture method utilizes the system provided by the present invention. The 26th also has the ability to simultaneously capture images. 22 201005285. The above embodiments are disclosed to illustrate the present invention, and although the specific terms are used to describe the present invention, the scope of the present invention is not limited thereto; those skilled in the art can understand the present invention. After the principle is changed and modified to achieve the equivalent purpose, such changes and modifications shall be included in the scope of the patent application scope as described later. 〇 之 之 专 23 23 23 23 2010 2010 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 23 Figure 2: A front cross-sectional view of the system's angular illumination assembly and side 稜鏡 structure shown in Figure 1. (Fig. 3: an enlarged front view of the side prism structure shown in Figure 2 and a schematic diagram of the reorientation of the light. Figure 4: The semiconductor package and the aperture in the system shown in Figure 1 A first optical path advancement flowchart of an embodiment between the first cameras. Fig. 5 is a schematic diagram of a path of light traveling along the first optical path of the embodiment shown in Fig. 4. Fig. 6: ray along the first A flow chart of the second optical path advancement of an embodiment between the semiconductor package and a second camera in the system shown in the drawing. Fig. 7 is a line of light traveling along the second optical path of the embodiment shown in Fig. 6. Fig. 8 is a combination diagram showing the first light path of the embodiment shown in Fig. 4 and the first light path advancing path of the embodiment shown in Fig. 6. Fig. 9 is a view showing Fig. 1 The image captured by the first camera and the second camera of the system. Fig. 10 is a route through which the light passes through the system , assembly to facilitate the detection of the semiconductor package shown in Fig. 1. Fig. 11: According to the present invention A second embodiment showing a semiconductor A flowchart of a method of detecting apparatus. The main reference numerals DESCRIPTION 24,201,005,285
1 0 系 統 1 2 半 導 體 封 裝 1 4 昭 i 明 總 成 1 6 稜 鏡 總 成 1 8 預 備 位 置 2 0 角 度 照 明 總 成 2 2 同 轴 照 明 總 成 2 4 第 一 攝 影 機 2 6 第 二 攝 影 機 5 0 第 一 入 σ 表 面 5 2 第 二 反 射 表 面 5 4 第 二 反 射 表 面 5 6 第 四 反 射 表 面 5 8 第 五 反 射 表 面 6 0 第 一 出 D 表 面 6 2 第 一 攝 影 機 鏡 頭 6 4 第 二 出 σ 表 面 6 6 第 二 攝 影 機 鏡 頭 6 8 側 視 稜 鏡 結 構 1 0 0 第 一 光 路 1 1 0〜1 7 0 步 驟 2 0 0 第 二 光 路 2 1 0〜2 6 0 步 驟 3 0 0 方 法 3 1 0 - -3 6 0 步 驟 251 0 System 1 2 Semiconductor package 1 4 Zhao i Ming assembly 1 6 稜鏡 assembly 1 8 Preparation position 2 0 Angle illumination assembly 2 2 Coaxial illumination assembly 2 4 First camera 2 6 Second camera 5 0 One into σ surface 5 2 second reflecting surface 5 4 second reflecting surface 5 6 fourth reflecting surface 5 8 fifth reflecting surface 6 0 first out D surface 6 2 first camera lens 6 4 second out σ surface 6 6 Second camera lens 6 8 side view structure 1 0 0 first light path 1 1 0~1 7 0 step 2 0 0 second light path 2 1 0~2 6 0 step 3 0 0 method 3 1 0 - -3 6 0 step 25