200929425 九、發明說明： 【發明所屬之技術領域】 本發明係關於一種如申請專利範圍第丨項之前序中所述 類別之方法’用以拾取設在晶圓台上之半導體晶片。本發 明更關於一種用以配置經取去之半導體晶片於一基板上之 方法。 【先前技術】 用以裝配半導體晶片之裝配機械裝置於黏晶機之領域中 爲已知的。該裝配機械裝置係用以將互相毗鄰設置於一晶 片載具上之許多同型之晶圓晶片一個接著一個地裝配於一 基板上，例如金屬導線’架。該黏晶機包含一晶圓台，於該 晶圓台上設有該晶片載具；一運送系統，用以提供該基板 及一取置系統，以自該晶片載具取去該等半導體晶片，並 將其放在該基板上。該取置系統包含一具有晶片夾具之接 合頭，其中該晶片夾具係藉由一驅動系統來回移動。該晶 片夾具係以一垂直軸旋轉，故若必要時可改變該等半導體 晶片之旋轉位置。該晶片夾具包含一可交換的夾取部件， 其爲一種於其上實施真空之吸取部件，其爲所屬技術領域 中習知之”拾取工具”或”晶粒夾套”。 對於此種裝配機械裝置有非常高的需求。對於該等已裝 配之晶片之進一步處理，該等晶片需被放在該基板上的一 精確位置上。於該黏晶機上設置二台攝影機，以確保該等 半導體晶片可在微米範圍內正確地被置於該基板上。該第 一攝影機測量將由該晶片夾具所拾取之該半導體晶片之位 200929425 置’以及提供關於一第一座標系統之位置資料。該第二攝 影機測量該等半導體晶片需要被放置於其上之該基板所在 地之位置，以及提供關於一第二座標系統之位置資料。依 據該等攝影機所提供的資訊，該取置系統以下列方式控制 該接合頭：該晶片夾具可自該晶圓台取去該半導體晶片， ' 以及可以一精確位置方式，將該半導體晶片放在該基板所 在地之正確位置處。該取置系統之位置與一第三座標系統 ^ 有關，其中該第三座標系統係與該攝影機之座標系統無關。 在該黏晶機之操作期間，會發生該等三個座標系統的相 對位置會因不同情況而改變之問題。該黏晶機在不同位置 處的溫度時常有意或無意地改變。此多半造成由該第一攝 影機之座標系統中或該第二攝影機之座標系統中所判定之 目標座標轉換爲該取置系統之運動座標，不再如所要求之 準確的後果。 【發明內容】 © 本發明係基於提供一種用以拾取及裝配半導體晶片之方 法爲目的，其中可確保該等半導體晶片放置時之高準確 度，而不管外界環境與變化。此目的係依據本發明申請專 利範圍第1項之特徵而達成。 本發明係關於一種用以拾取及可選擇地裝配半導體晶片 於一基板上之方法，其中： -將該等半導體晶片供應在一晶圓台上； -一半導體晶片接續在另一個之後被供應在一基板台上； 200929425 -一第一攝影機檢測提供在該晶圓台上並被裝配爲下一 個之該半導體晶片之位置及定位； -一第二攝影機檢測之將被裝配於基板上之該半導體晶 片，其在基板所在地之位置及定位；以及 -一晶片夾具拾取提供在該晶圓台上之該半導體晶片，以 ' 及裝配該半導體晶片於該基板上，其中該晶片夾具係被固 定在一接合頭上以及一最好具有二個線性驅動裝置之取置 0 系統係於該晶圓台與該基板之間來回運送具有該晶片夾具 之接合頭。 依據本發明，藉由該第一攝影機所檢測之下一個將被裝 配之該半導體晶片之位置係以關於一第一座標系統KSI2 位置資料的形式被提供，將裝配該半導體晶片於基板上之 該基板所在地之位置係以關於一第二座標系統KS2之位置 資料之形式被提供，以及該接合頭之位置係關於一第三座 標系統KS3。 〇 本發明提出一種提供在該接合頭上之標記，其中該標記 之位置可藉由該等攝影機測得而知。由於該標記因爲結構 上埋由無法配置在該等攝影機之焦平面內，故本發明在一 較佳具體實施例更提出於該標記上方連接一透鏡，該透鏡 可確保該標記也以清晰方式成像。 本發明更提出使用一第一固定映射函數F與一第一可變 校正向量，用以將該第一座標系統KSii座標轉換爲該 取置系統之該第三座標系統KS3，以及一第二固定映射函 200929425 數G與一第二可變校正向量Κ2,用以將該第二座標系統KS2 之座標轉換爲該取置系統之該第三座標系統KS3。當首次 設定該黏晶機或在該黏晶機之一般新設定的情況下，一方 面決定該等映射函數F與G以及其反函數，以及另一方面 將該等二個校正向量1與K2設爲0。該等校正向量1^與 ' Κ2係依一預定事件的發生作重新調整，然而並沒有改變該 等映射函數F與G直到該黏晶機之下個一般新設定。將了 ^ 解到一預定事件係爲具有高或然率之可被預期的事件，其 中該等三個座標系統KS!、KS2與KS3之相對位置相對於彼 此而改變至一降低配置精確度的範圍。 【實施方式】 第1圖係槪要顯示用以裝配半導體晶片之裝配機械裝置 (其爲所謂的黏晶機）之俯視圖，在其爲用以了解本發明所 必要範圍者。第2圖係以側視圖顯示部分該裝配機械裝 置。該黏晶機包含一晶圓台1，於其上提供有將被裝配之 Ο 該等半導體晶片2; —基板台3，藉由一運送設備（沒有顯 示）而將被裝配之基板4於其上提供；以及一取置系統5’ 其自該晶圓台1拾取該等半導體晶片2並將其放置在該基 板4上，以及二台攝影機6與7。該取置系統5包含一具有 可交換晶片夾具9之接合頭8(第2圖）以及二個線性位置控 制驅動器，用以在二個指定爲X方向與y方向之正交方向 來移動該接合頭8。第三驅動裝置（沒有顯示）係用以在z方 向舉起或降下該接合頭8或該晶片夾具9 ’其中該z方向係 200929425 延伸與該圖面垂直。該第一攝影機6係用以判定下個將被 取去半導體晶片2之位置。該第二攝影機7係用以判定於 基板4上將被放置該半導體晶片2在該基板之所在地之位 置。該第一攝影機6 —通常以一固定方式配置。該第二攝 ' 影機7亦以一固定方式配置或可以獨立的驅動裝置’在至 ' 少一個或二個平行於該基板4之表面的方向延伸移動。該 取置系統5例如可爲習知第 TW 1 25280、TW23 1 5 6 1、 A TW237297以及TW287841號中之系統。 ❹ 一標記1 〇(第2圖）係以下列方式側向連接至該接合頭8 : 當該接合頭8設在該第一攝影機6之視野中時’其可在該 第一攝影機6所供應之影像中看到，以及當該接合頭8設 在該第二攝影機7之視野中時，其可在該第二攝影機7所 供應之影像中看到。 第2圖顯示該第一攝影機6、該接合頭8以及該晶圓台1 之側視圖。其藉由面向該晶圓台1之線6 a來限定圖式中之 〇 視野，使得在由其所供應之影像中，下個將被取去之半導 體晶片2係以清晰界定方式成像。該第一攝影機6之焦平 面位在由將被取去之該半導體晶片2所定義之平面中。該 第二攝影機7之焦平面（第1圖）位在由將被裝配之該基板4 之表面所定義之平面中。將該標記10連接於該接合頭8 上，使得其可藉由二台攝影機6與7以一清晰界定方式來 成像但不調整該焦平面是不可行的。爲了仍可確保該標記 10以清晰界定方式來成像，連接一透鏡10於該標記10上 200929425 方之該接合頭8是有助益的。該透鏡11係設在該標記10 與該各個攝影機6及7之間，並可確保該標記1〇於該各個 攝影機6與7之影像中以足夠的清晰界定方式來成像。爲 了可確保該標記10以清晰界定方式來成像，其也可預先調 整該等攝影機之焦平面以取代該透鏡11之設置。由於該透 鏡11，結果該等攝影機6與7之透鏡系統較小調整範圍， 故以該透鏡11之解決方式是較簡單、快速且更經濟的。 0 該第一攝影機6將其影像資料提供給一第一影像處理單 元’其從該影像資料判定接著將被裝配之該半導體晶片2 之位置與定位’以及將其以關於一第一座標系統KS,之位 置資料的形式來提供。這些位置資料係由三個數（p，q，φ) 所組成，其中ρ與q二個數指定該半導體晶片2之參考點 的位置’以及數値Φ判定沿著該半導體晶片2對其設定點 位置轉動的角度。 該第二攝影機7將其影像資料提供給—第二影像處理單 〇 元’其從該影像資料判定將被裝配於基板上之該半導體晶 片2 ’其在該基板所在地之位置與定位，以及將其以關於 —第二座標系統KS2之位置資料的形式來提供。這些位置 資料係由三個數（u，ν’ ψ)所組成，其中數値u與v指定該 基板所在地之參考點的位置，以及數値ψ爲沿著該基板所 在地對其設定點位置轉動的角度。 該取置系統之第一線性驅動裝置提供〜數値^以該取置 系統之第二線性驅動裝置提供一數値yN1，它們係相對於該 10- 200929425 第三座標系KS3而一起形成代表該標記10之位置（XM，yM) 之該位置資料。 該晶片夾具9係可以繞一旋轉軸12轉動（第2圖）。該晶 片夾具9之吸入口定義該晶片夾具9之夾具軸13(第2圖） 之位置。該夾具軸13於該第三座標系統KS3之位置（Xc，Xc) ' 係由下式給定：BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of picking up a semiconductor wafer disposed on a wafer stage as described in the preamble of the patent application. The invention further relates to a method for configuring a removed semiconductor wafer on a substrate. [Prior Art] The assembly mechanism for assembling a semiconductor wafer is known in the field of die bonding machines. The assembly mechanism is used to mount a plurality of wafer wafers of the same type disposed adjacent to each other on a wafer carrier one after another on a substrate, such as a metal wire. The die bonder includes a wafer stage on which the wafer carrier is disposed; a transport system for providing the substrate and a pick-up system for removing the semiconductor wafers from the wafer carrier And place it on the substrate. The pick-up system includes a mating head having a wafer holder that is moved back and forth by a drive system. The wafer holder is rotated by a vertical axis so that the rotational position of the semiconductor wafers can be changed if necessary. The wafer holder includes an exchangeable gripping member which is a suction member on which a vacuum is applied, which is a "pickup tool" or "die jacket" as is known in the art. There is a very high demand for such assembly mechanisms. For further processing of the assembled wafers, the wafers need to be placed at a precise location on the substrate. Two cameras are placed on the die bonder to ensure that the semiconductor wafers are properly placed on the substrate in the micrometer range. The first camera measures the position of the semiconductor wafer to be picked up by the wafer holder 200929425 and provides location information about a first coordinate system. The second camera measures the location of the substrate on which the semiconductor wafers need to be placed and provides location information about a second coordinate system. According to the information provided by the cameras, the pick-up system controls the bond head in such a manner that the wafer holder can take the semiconductor wafer from the wafer stage, and the semiconductor wafer can be placed in a precise position. The correct location of the substrate location. The location of the retrieval system is associated with a third coordinate system ^, wherein the third coordinate system is independent of the coordinate system of the camera. During the operation of the die bonder, the relative position of the three coordinate systems may change due to different conditions. The temperature of the die bonder at different locations is often intentionally or unintentionally changed. This mostly results in the conversion of the target coordinates determined by the coordinate system of the first camera or the coordinate system of the second camera to the motion coordinates of the access system, without the exact consequences as required. SUMMARY OF THE INVENTION The present invention is based on the object of providing a method for picking up and assembling a semiconductor wafer in which high accuracy in the placement of the semiconductor wafers is ensured regardless of the external environment and variations. This object is achieved in accordance with the features of the first item of the patent application of the present invention. The present invention relates to a method for picking up and optionally assembling a semiconductor wafer on a substrate, wherein: - the semiconductor wafers are supplied on a wafer stage; - a semiconductor wafer is successively supplied after the other a substrate stage; 200929425 - a first camera detects the position and positioning of the semiconductor wafer provided on the wafer stage and assembled as the next; - a second camera detects the semiconductor to be mounted on the substrate a wafer at a position and location of the substrate; and - a wafer holder picking up the semiconductor wafer provided on the wafer table to ' and assembling the semiconductor wafer on the substrate, wherein the wafer holder is fixed to the substrate A pick-up system on the bond head and preferably having two linear actuators is used to transport the bond head having the wafer clamp back and forth between the wafer table and the substrate. According to the present invention, the position of the semiconductor wafer to be assembled by the first camera is provided in the form of positional information about a first coordinate system KSI2, and the semiconductor wafer is mounted on the substrate. The position of the substrate is provided in the form of positional information about a second coordinate system KS2, and the position of the joint head is related to a third coordinate system KS3. The present invention provides a marker provided on the bond head, wherein the position of the marker is known by the cameras. Since the marking is not configurable in the focal plane of the camera because of the structural burying, the present invention further proposes to attach a lens above the marking in a preferred embodiment, the lens ensures that the marking is also imaged in a clear manner. . The present invention further proposes to use a first fixed mapping function F and a first variable correction vector for converting the coordinate of the first coordinate system KSii into the third coordinate system KS3 of the access system, and a second fixed The mapping function 200929425 and the second variable correction vector Κ2 are used to convert the coordinates of the second coordinate system KS2 into the third coordinate system KS3 of the accommodating system. When the die bonder is first set or in the case of a general new setting of the die bonder, on the one hand the mapping functions F and G and their inverse functions are determined, and on the other hand the two correction vectors 1 and K2 are Set to 0. The correction vectors 1^ and 'Κ2 are re-adjusted according to the occurrence of a predetermined event, but the mapping functions F and G are not changed until the next new setting of the die bonder. The solution to a predetermined event is an event that can be expected with a high probability, wherein the relative positions of the three coordinate systems KS!, KS2, and KS3 are changed relative to each other to a range that reduces the accuracy of the configuration. [Embodiment] Fig. 1 is a plan view showing an assembly mechanism for mounting a semiconductor wafer, which is a so-called die bonder, which is necessary for understanding the scope of the present invention. Figure 2 shows a portion of the assembly mechanism in a side view. The die bonder includes a wafer table 1 on which the semiconductor wafer 2 to be mounted is provided; a substrate table 3 on which the substrate 4 to be assembled is mounted by a transport device (not shown) Provided thereon; and a pickup system 5' which picks up and places the semiconductor wafers 2 from the wafer table 1 on the substrate 4, and two cameras 6 and 7. The pick-up system 5 includes a bond head 8 (Fig. 2) having an exchangeable wafer holder 9 and two linear position control drivers for moving the joint in two orthogonal directions designated as the X and y directions. Head 8. A third driving means (not shown) is used to lift or lower the bonding head 8 or the wafer holder 9' in the z direction, wherein the z-direction system 200929425 extends perpendicular to the drawing. The first camera 6 is used to determine the position at which the next semiconductor wafer 2 will be removed. The second camera 7 is for determining where the semiconductor wafer 2 is to be placed on the substrate 4 at the location of the substrate. The first camera 6 is typically configured in a fixed manner. The second camera 7 is also arranged in a fixed manner or can be independently driven to extend in a direction that is less than one or two parallel to the surface of the substrate 4. The accommodating system 5 can be, for example, a system in the conventional TW 1 25280, TW23 1 5 6 1 , A TW237297, and TW287841. ❹ a mark 1 〇 (Fig. 2) is laterally connected to the joint head 8 in the following manner: when the joint head 8 is disposed in the field of view of the first camera 6, it can be supplied at the first camera 6. As seen in the image, and when the bond head 8 is disposed in the field of view of the second camera 7, it can be seen in the image supplied by the second camera 7. Fig. 2 shows a side view of the first camera 6, the bonding head 8, and the wafer stage 1. The field of view in the drawing is defined by the line 6a facing the wafer table 1 such that the next semiconductor wafer 2 to be removed is imaged in a clearly defined manner among the images supplied therefrom. The focal plane of the first camera 6 is in the plane defined by the semiconductor wafer 2 to be removed. The focal plane (Fig. 1) of the second camera 7 is located in a plane defined by the surface of the substrate 4 to be assembled. Attaching the marker 10 to the bond head 8 makes it possible to image in a clearly defined manner by the two cameras 6 and 7 without adjusting the focal plane. In order to still ensure that the indicia 10 is imaged in a clearly defined manner, it is helpful to connect a lens 10 to the indicia 10 on the indicia 10 200929425. The lens 11 is disposed between the indicia 10 and the respective cameras 6 and 7, and ensures that the indicia 1 is imaged in a sufficiently sharply defined manner in the images of the respective cameras 6 and 7. In order to ensure that the indicia 10 are imaged in a clearly defined manner, it is also possible to pre-adjust the focal planes of the cameras in place of the arrangement of the lenses 11. As a result of the lens 11, as a result, the lens systems of the cameras 6 and 7 have a small adjustment range, so that the solution of the lens 11 is simpler, faster and more economical. 0 The first camera 6 supplies its image data to a first image processing unit 'which determines the position and position of the semiconductor wafer 2 to be assembled from the image data' and relates it to a first coordinate system KS , the location of the location information to provide. The position data is composed of three numbers (p, q, φ), wherein the two numbers ρ and q specify the position of the reference point of the semiconductor wafer 2 and the number 値 Φ determination is set along the semiconductor wafer 2 The angle at which the point position is rotated. The second camera 7 supplies its image data to a second image processing unit that determines from the image data that the semiconductor wafer 2 to be mounted on the substrate is positioned and positioned at the substrate location, and It is provided in the form of positional information about the second coordinate system KS2. The position data consists of three numbers (u, ν' ψ), where the numbers u and v specify the position of the reference point of the substrate location, and the number of turns is the position of the set point along the substrate location. Angle. The first linear driving device of the accommodating system provides a number 値 NN1, which is formed by the second linear driving device of the accommodating system, which are formed together with the third coordinate system KS3 The location data of the location (XM, yM) of the marker 10. The wafer holder 9 is rotatable about a rotating shaft 12 (Fig. 2). The suction port of the wafer holder 9 defines the position of the jig axis 13 (Fig. 2) of the wafer holder 9. The position (Xc, Xc) of the clamp shaft 13 at the third coordinate system KS3 is given by:
(xg，υ〇) = (χμ ’ yμ) + D + E q 其中向量D說明相關於該標記10之位置（xm，yM)的旋轉 軸12位置，以及向量E爲相關於該旋轉軸12之位置的該 夾具軸13位置。向量D爲一即刻被決定的固定向量。向量 E爲一與該晶片夾具9共同旋轉之向量：其長度具有一固 定量，但當該晶片夾具9繞著該旋轉軸12轉動時，其方向 則會改變。在理想情況下，該旋轉軸12與該夾具軸1 3總 是一致，亦即E = 0，不管該晶片夾具9之旋轉位置。 第3圖係說明該等三個座標系統KS!、KS2與KS3之間的 © 相關性。爲了確保該等半導體晶片2可被以一正確地放置 方式而放在該基板4上，必須可計算在該第一座標系統KS· 與該第二座標系統KS2二者內之該晶片夾具9之夾具軸13 之目前位置。因此，於首次設定或該裝配機械裝置之一般 新設定時，判定一第一映射函數F，其中該函數F將該第 一座標系統1^51映射至該第三座標系統KS3。此映射藉助於 該標記10而發生：該取置系統5之二個線性驅動裝置在該 第一攝影機6之視野內，將該具有該標記10之接合頭8 — Γ Γ1 a. w 2 -11· 200929425 同移至k個不同位置（χη，yn)，其中n=l到k ’以及該第一 影像處理單元將該第一攝影機6所供應之影像判定爲標記 10之相關位置（Ρ»，qn)。自所獲得之資料記錄中計算該第一 映射函數F。接著應用下列式子： (X ， y) = F(p ， q) 接著計算該映射函數F之反函數F〃，使得 (p » q) = F '(x > y) ❹ 此外，將一第一校正向量L設爲Κπ〇。 同理，判定將該第二座標系統KS 2映射至該第三座標系 統KS3的一第二映射函數G及其反函數G1。接著應用下式： > (X ， y) = G(u ， ν) 並且反之亦然 (u ' v) = G '(x > y) 此外’將一第二校正向量K2設爲K2 = 0。 使用該第一攝影機6與該第一座標系統KS,以判定關於 © 該第一座標系統KS,的目標座標，其中該取置系統5必須 將該接合頭8移至該第一座標系統KS,以致使該晶片夾具9 可拾起提供在該晶圓台1上之該半導體晶片2。使用該第 二攝影機7與該第二座標系統KS2以判定關於該第二座標 系統KS 2的目標座標，其中該取置系統5必須將該接合頭8 移至該第二座標系統KS2以致使該晶片夾具9可以正確之 放置方式來放置該半導體晶片2。所有計算於這二個座標 系統KS i與KS2中執行。僅在所有計算完成後該經判定之 -12- 200929425 目標座標，藉由各自的映射函數F與G而被轉換爲該第三 座標系統KS3之運動座標。因而向量d與E可均被判定爲 關於該第一座標系統KS，之向量〇1與Ei，以及關於該第二 座標系統KS2之向量Eh與E2。因此，該第三座標系統KS3 被僅用來移動接合頭8而於此第三座標系統尺33中不作任 • 何計算。該第三座標系統KS3係藉由該取置系統5之機械 裝置指定’亦即’座標X與y爲由該等二個線性驅動裝置 q 之編碼器所供應的位置値，以及因此其不是一準確的正交 座標系統。 一旦已判定該等映射函數F與G、其反函數F·1與G'1以 及該等向量Di、E，、D2與E2，則於該生產階段時，一半導 體晶片2接在另一個之後而被裝配，其中 -由該第一攝影機6拍攝下個半導體晶片2之影像，以及 自該影像計算關於該半導體晶片2之第一座標系統KS!之 位置資料（pw，qw，<pw)，其中當該半導體晶片2沒有對其設 Ο 定點位置旋轉時，（pw = 〇; -關於該第三座標系統KS3之位置（xw，yw)，其中該第三 座標系統需要藉由該標記10來拍攝，使得該晶片夾具9之 夾具軸13通過該半導體晶片2之參考點，其以下式計算： (xw ' yw) = F[(pw，qw) —Di-Ει + Κι] -該經計算之位置（xw，yw)爲近似値並且該半導體晶片2 係藉由該晶片夾具9拾起； -由該第二攝影機7拍攝於基板上將被裝配之該半導體晶 -13- 200929425 片2’其在該基板所在地的影像，以及自該影像計算關於 該第二座標系統KS2之該基板所在地之位置資料（Us，vs , Ψ3) ’其中當該基板所在地沒有對其設定點位置旋轉時， Ψ s 二 0 ; -關於該第三座標系統KS3之位置（xs’ ys)，其中該第三座 標系統需要藉由該標記10來拍攝，使得該晶片夾具9之夾 具軸13通過該基板所在地之參考點，其以下式計算： ❹ (xs > ys) = G[(us ' vs)-D2-E2 + K2] -該經計算之位置（xs’ ys)爲近似値，並且該晶片夾具9 係以ψ5-φ5之角度作可選擇地轉動，並且該半導體晶片2 係放在該基板所在地上。 爲了於全部生產期間以同樣高的水準保持該黏晶機之放 置精確度，於一預定事件發生期間執行該第一校正向量Kl 與該第二校正向量K2之重新調整。使用設在該接合頭8上 之該標記1 0，該標記係被帶至該第一攝影機6之視野中以 © 重新調整該第一校正向量K,，以及被帶至該第二攝影機7 之視野中以重新調整該第二校正向量K2。該第一校正向量 1之重新調整係藉由以下而發生： -將該接合頭8移至一設定點位置R = (xr，yR)，其中該標 記10係以關於該第三座標系統KS3之座標（xR，yR)而位在 該第一攝影機6之視野中； -計算相對於該第一座標系統KSi之該標記1〇設定點位 置（PR，QR)當作（PR，qR) = F-1(XR，yR); -14- 200929425 -以該第一攝影機6拍攝該標記10之影 攝影機6之影像以判定相對於該第一座標 記10之真實位置（pM，qM);以及 -計算該第一校正向量L作爲該近似設 測量之真實位置之間的差異：(xg, υ〇) = (χμ ' yμ) + D + E q where vector D indicates the position of the axis of rotation 12 associated with the position (xm, yM) of the mark 10, and the vector E is associated with the axis of rotation 12 The position of the clamp shaft 13 of the position. Vector D is a fixed vector that is immediately determined. The vector E is a vector co-rotating with the wafer holder 9: its length has a fixed amount, but its orientation changes as the wafer holder 9 rotates about the rotary shaft 12. Ideally, the axis of rotation 12 is always coincident with the axis of the clamp 13, i.e., E = 0, regardless of the rotational position of the wafer holder 9. Figure 3 illustrates the © correlation between the three coordinate systems KS!, KS2 and KS3. In order to ensure that the semiconductor wafers 2 can be placed on the substrate 4 in a correctly placed manner, it is necessary to calculate the wafer holder 9 in both the first coordinate system KS· and the second coordinate system KS2. The current position of the clamp shaft 13. Therefore, a first mapping function F is determined at the first setting or a general setting of the assembly mechanism, wherein the function F maps the first landmark system 1^51 to the third coordinate system KS3. This mapping takes place by means of the marking 10: the two linear drives of the handling system 5, within the field of view of the first camera 6, the bonding head 8 with the marking 10 - Γ Γ 1 a. w 2 -11 · 200929425 moves to k different positions (χη, yn), where n=l to k ' and the first image processing unit determines the image supplied by the first camera 6 as the relevant position of the mark 10 (Ρ», Qn). The first mapping function F is calculated from the obtained data record. Then apply the following formula: (X , y) = F(p , q) Then calculate the inverse function F〃 of the mapping function F such that (p » q) = F '(x > y) ❹ In addition, one will The first correction vector L is set to Κπ〇. Similarly, it is determined that the second coordinate system KS 2 is mapped to a second mapping function G of the third coordinate system KS3 and its inverse function G1. Then apply the following formula: > (X , y) = G(u , ν) and vice versa (u ' v) = G '(x > y) In addition, 'set a second correction vector K2 to K2 = 0. Using the first camera 6 and the first coordinate system KS to determine a target coordinate with respect to © the first coordinate system KS, wherein the pick-up system 5 must move the joint head 8 to the first coordinate system KS, The wafer holder 9 can pick up the semiconductor wafer 2 provided on the wafer stage 1. Using the second camera 7 and the second coordinate system KS2 to determine a target coordinate with respect to the second coordinate system KS 2, wherein the picking system 5 must move the joint head 8 to the second coordinate system KS2 to cause the The wafer holder 9 can place the semiconductor wafer 2 in a proper placement. All calculations are performed in these two coordinate systems KS i and KS2. The determined -12-200929425 target coordinates are converted to the motion coordinates of the third coordinate system KS3 by the respective mapping functions F and G only after all calculations have been completed. Thus, vectors d and E can both be determined as vectors 〇1 and Ei for the first coordinate system KS, and vectors Eh and E2 for the second coordinate system KS2. Therefore, the third coordinate system KS3 is used only to move the joint head 8 and the third coordinate system ruler 33 does not perform any calculation. The third coordinate system KS3 specifies, by means of the mechanism of the pick-up system 5, that the coordinates X and y are the positions 供应 supplied by the encoders of the two linear drives q, and therefore it is not a Accurate orthogonal coordinate system. Once the mapping functions F and G, their inverse functions F·1 and G'1, and the vectors Di, E, D2 and E2 have been determined, then at the production stage, one semiconductor wafer 2 is connected to the other. And assembled, wherein - the image of the next semiconductor wafer 2 is taken by the first camera 6, and the position data (pw, qw, <pw) of the first coordinate system KS! of the semiconductor wafer 2 is calculated from the image. , wherein when the semiconductor wafer 2 is not rotated at a fixed point position, (pw = 〇; - about the position (xw, yw) of the third coordinate system KS3, wherein the third coordinate system needs to be marked by the mark To take a picture so that the jig axis 13 of the wafer holder 9 passes through the reference point of the semiconductor wafer 2, which is calculated by the following formula: (xw ' yw) = F[(pw, qw) - Di-Ει + Κι] - the calculation The position (xw, yw) is approximately 値 and the semiconductor wafer 2 is picked up by the wafer holder 9; the semiconductor crystal 13-200929425 piece 2' to be assembled on the substrate by the second camera 7 An image of the location of the substrate, and calculation of the second coordinate system from the image Location data of the substrate location of the KS2 (Us, vs, Ψ 3) 'When the substrate location does not rotate at its set point position, Ψ s 2 0; - the position of the third coordinate system KS3 (xs' ys) Wherein the third coordinate system needs to be photographed by the mark 10 such that the jig axis 13 of the wafer holder 9 passes through a reference point of the substrate location, and the following formula is calculated: ❹ (xs > ys) = G[(us 'VS)-D2-E2 + K2] - the calculated position (xs' ys) is approximately 値, and the wafer holder 9 is selectively rotated at an angle of ψ5-φ5, and the semiconductor wafer 2 is placed At the location of the substrate, in order to maintain the placement accuracy of the die bonder at the same high level during all production periods, the readjustment of the first correction vector K1 and the second correction vector K2 is performed during a predetermined event occurrence. Using the mark 10 provided on the bond head 8, the mark is brought into the field of view of the first camera 6 to re-adjust the first correction vector K, and is brought to the second camera 7. Re-adjusting the second correction vector K2 in the field of view The re-adjustment of the first correction vector 1 occurs by: - moving the bond head 8 to a set point position R = (xr, yR), wherein the mark 10 is associated with the third coordinate system KS3 The coordinates (xR, yR) are located in the field of view of the first camera 6; - calculating the mark 1 〇 set point position (PR, QR) relative to the first coordinate system KSi as (PR, qR) = F -1 (XR, yR); -14- 200929425 - taking the image of the camera 6 of the mark 10 with the first camera 6 to determine the true position (pM, qM) with respect to the first seat mark 10; Calculating the difference between the first correction vector L as the true position of the approximate set measurement:
Ki=(pR ’ qR)-(pM ’ qM)。 可明顯得知，該第一校正向量L與該第 _ 相關聯。 〇 該第二校正向量K2之重新調整係藉由相 -將該接合頭8移至一設定點位置Τ = (χτ 記10係以關於該第三座標系統KS3之座框 該第二攝影機7之視野中； -計算相對於該第二座標系統KS2之該標 置（Ut，Vt)成（UT，Vt) = G」（Xt，yT); -以該第二攝影機7拍攝該標記10之影 〇 攝影機7之影像以判定相對於該第二座標 記10之真實位置（UM，VM)，以及 -計算該第二校正向量K2作爲該近似設 測量之真實位置之間的差異： K2=(UT ， Vt)-(UM ， Vm)， 可明顯得知，該第二校正向量K2與該第 相關聯。 有許多不同事件可觸發該等校正向量Κ 像，使用該第一 系統KSi之該標 定點位置與該經 一座標系統KSi 似方式而發生： ，y〇，其中該標 冥（X T，y T)而位在 記1 0設定點位 像，使用該第二 系統K S 2之該標 定點位置與該經 二座標系統K； S 2 I與K 2之重斩調 -15- 200929425 整，特別是下列四個事件： -由於最後校正，故已裝配預定數量的半導體晶片2; -由於該最後校正，故於該取置系統5之預定位置處所測 量的溫度已被一比預定値還大的値所改變； " -生產被停止； * -於裝配後由該第二攝影機7檢測與計算之該裝配的半導 體晶片之實際位置係以大於一預定量偏離該設定點位置。 _ 在完成該等校正向量10與K2之重新調整後，可依照如 ❹ 上所述之步驟繼續該等半導體晶片2之裝配，但現在更新 之校正向量1與Κ2可不同於0。 Λ 本發明可被用在習知之取置系統中，其中晶圓台1與該 基板4之平台3係以平行平面配置，以及在ΕΡ1 480507中 所述之取置系統中，其中該晶圓台1與該基板之平台3係 以彼此關聯傾斜的方式配置，以及其中該接合頭8除了於 X方向與y方向中之運動外，其繞著一水平軸執行一樞轉 ❹ 運動。 上述之具體實施例爲一較佳具體實施例，其中爲了調整 與重新調整而將該接合頭分別移至該第一設定點位置R以 及移至該第二設定點位置T，以及儲存關於該第三座標系 統KS 3之該第一設定點位置R與該第二設定點位置τ之座 標，並用以重新調整該等二個校正向量1^與K2。在此例示 中，該標記10之各個設定點位置係分別藉由該反函數F-1 與cr1來計算。接下來說明另一具體實施例，其中當該接合 -16- 200929425 頭8位在該第一或第二設定點位置中時，則附加儲存關於 該第一座標系統KS12該標記10之座標（或該接合頭8上之 任何其它隨機參考點）或者關於該第二座標系統KS2之該標 記10之座標（或該接合頭8上之任何其它隨機參考點），接 著用於該等二個校正向量1^與K2之重新調整。 ' 該取置系統之一部分包含該拾取系統，用以自該晶圓台 拾取該等半導體晶片。該第三座標系統KS3爲該拾取系統 φ 或該取置系統之內在座標系統，以及因此以下將稱爲KS 座標系統。爲了確保該重新調整可被執行，首先於一設定 階段執行一調整，其中該接合頭8係移至一第一設定點位 置（其係位在第一攝影機6之視野內），以及判定與儲存關 於該座標系統KS之第一設定點位置的座標（XSPI，ysP1)與關 於該第一攝影機6之座標系統KSii第一設定點位置之座 標（pSPI，qspi)。該重新調整係以下列方式而發生在該生產 階段中：該接合頭8係移至該第一設定點位置之座標（XSP1， Ο ρΡ1)，並再判定關於該第一攝影機6之座標系統ks1之設 疋點位置之座標（PSPl’’ qspi’）。（PSFM’，qSPl’）-(PSP 丨，qspi)之 向量差含有關於該第一座標系統KSi相對於該座標系統KS 之位移資訊，其中該座標系統KS自在該設定階段中設定以 後就已發生。在該接合頭8上之任何隨機參考點可被用以 定義相對於該第一座標系統KS,之該接合頭8之第一設定 點位置。如上所述’該標記1〇最好用作定義該參考點。 同理’檢測及校正該第二攝影機7之該第二座標系統KS2 -17- 200929425 相對於該接合頭8之座標系統KS之位置，使得於該設定階 段時執行進一步之調整’其中該接合頭8係被移動至一位 在該第二攝影機7之視野內之第二設定點位置，以及 判定與儲存關於該座標系統KS之第二設定點位置的座 標（xs^’yspO與關於該第二攝影機7之座標系統KS2之第二 設定點位置之座標（USP2，VSP2)。在該生產階段中之該重新 調整係以下列方式發生：該接合頭8係移至該第二設定點 φ 位置之座標（xsP2，ySP2)，並再判定關於該第二攝影機7之 座標系統KS2之第二設定點位置之座標（USP2’，VsP2’p (uSP2’’ vsP2’）-(uSP2’ VSP2)之向量差含有關於該第二座標系統 KS2相對於該座標系統KS之位置資訊，其中該座標系統KS 自該在該設定階段中設定以後就已發生。同樣，在此情況 下，在該接合頭8之任何隨機參考點可被用以定義相對於 該第二座標系統KS2之該接合頭8之第二設定點位置。如 上所述，該標記10最好用於定義該參考點。 ❹ 判定關於該第一座標系統KS,與該第二座標系統KS2i 該等參考點之座標包含各自以攝影機6與7拍攝影像，以 及藉由傳統影像評估判定該參考點之座標。 接著，該等半導體晶片之裝配較佳地以下列方式發生： -由該第一攝影機6所檢測而接著將被裝配之該半導體晶 片2的位置係以關於該第一座標系統位置上的資料 的形式來提供； -由該第二攝影機7所檢測，將被裝配之於基板上該半導 -18- 200929425 ， 體晶片2其在該基板所在地的位置，係以關於該第二座檁 系統KS:之位置上的資料的形式來提供； -在設定階段時’ 一第一映射函數將該第一座標系統KS, 映射至該座標系統KS並判定其反函數，以及設定一第一校 正向量爲0値，一第二映射函數將該第二座標系統KS2映 射至該座標系統KS並判定其反函數，以及設定一第二校正 向量爲0値； 0 -在生產階段時，一半導體晶片2而接在另一個被裝配， 使得 -以該攝影機6拍攝接著將被裝配之該半導體晶片2之影 像，並且判定該半導體晶片2相對於該第一座標系統KS, 之位置，以及藉由該第一映射函數及考慮該第一校正向 量，從其中計算該取置系統5需要移動該接合頭8以拾取 該半導體晶片2相關於該KS座標系統的位置； -以該第二攝影機7拍攝將被裝配於基板上之半導體晶片 © 2，其在該基板所在地的影像，以及判定相對於該第二座標 系統KS2之該基板所在地之位置，以及藉由該第二映射函 數及考慮該第二校正向量，從其中計算該取置系統5需要 移動該接合頭8以裝配該半導體晶片2於該基板所在地相 關於該KS座標系統的位置； 以及於該生產階段時之重新調整包括以下列步驟重新調 整該第一校正向量及該第二校正向量： -將該接合頭8移動至該第一設定點位置； -19- 200929425 -以該第一攝影機6拍攝該標記10之影 一攝影機6之影像判定該標記10相對於 的實際位置，以及 -計算該第一校正向量Kl作爲該儲存設 定之實際位置之間的差異； " -將該接合頭8移動至該第二設定點位置 •以該第二攝影機7拍攝該標記10之影 0 二攝影機7之影像判定該標記10相對於 KS2的實際位置，以及 -計算該第二校正向量K2作爲該儲存設 Ί 定之實際位置之間的差異。 雖然已顯示及說明本發明之具體實施例 技術領域中具有此揭露權益之熟悉技藝 是，在不脫離本發明之槪念下，遠超過上 可行的。因此，除了附隨之申請專利範圍舆 © 本發明將不被限制。 隨附圖式係說明本發明之一個或多個具 詳細之發明說明一同用以解釋本發明之原 倂入並構成本說明書之—部分。該等圖式 來顯示。 【圖式簡單說明】 第1圖係顯示用以裝配半導體晶片之 圖； 像，以及從該第 該第一座標系統 定點位置與所判 像，以及從該第 該第二座標系統 定點位置與所判 及應用，但所屬 者可顯而易知的 述範圍之修正爲 I其等效範圍外， 體實施例，並與 則與實施，其係 並非以真實比例 裝配機械之俯視 -20- 200929425 第2圖係顯示一攝影機、一接合頭與一晶圓台之側視 圖，以及 第3圖係顯示該接合頭與三個不同的座標系統之俯視 圖。 【主要元件符號說明】Ki = (pR ' qR) - (pM ' qM). It can be clearly seen that the first correction vector L is associated with the first _. The re-adjustment of the second correction vector K2 is performed by phase-moving the bonding head 8 to a set point position Τ = (χτ 10 is for the second camera 7 of the frame of the third coordinate system KS3) In the field of view; - calculating the index (Ut, Vt) relative to the second coordinate system KS2 to (UT, Vt) = G" (Xt, yT); - taking the shadow of the mark 10 with the second camera 7. The image of the camera 7 is determined to determine the true position (UM, VM) relative to the second block mark 10, and - the second correction vector K2 is calculated as the difference between the true positions of the approximation settings: K2 = (UT , Vt)-(UM , Vm), it is apparent that the second correction vector K2 is associated with the first. There are a number of different events that can trigger the correction vector image, using the calibration point of the first system KSi The position occurs in a manner similar to the KSi of the landmark system: , y〇, where the standard (XT, y T) is located at the setpoint image, and the calibration point position of the second system KS 2 is used. With the two-coordinate system K; S 2 I and K 2 heavy adjustment -15- 200929425, especially the following four events - a predetermined number of semiconductor wafers 2 have been assembled due to the final correction; - due to this final correction, the temperature measured at the predetermined position of the handling system 5 has been changed by a larger than a predetermined threshold; " - production is stopped; * - the actual position of the assembled semiconductor wafer detected and calculated by the second camera 7 after assembly is deviated from the set point position by more than a predetermined amount. _ After completing the correction vectors 10 and K2 After re-adjustment, the assembly of the semiconductor wafers 2 can be continued in accordance with the steps described above, but the updated correction vectors 1 and Κ2 can now be different from 0. Λ The present invention can be used in conventional pick-and-place systems. Wherein the wafer table 1 and the platform 3 of the substrate 4 are arranged in a parallel plane, and in the pick-up system described in ΕΡ 1 480 507, wherein the wafer table 1 and the platform 3 of the substrate are tilted in association with each other The configuration, and wherein the bonding head 8 performs a pivoting motion about a horizontal axis in addition to the movement in the X and y directions. The above specific embodiment is a preferred embodiment, wherein Adjusting and re-adjusting to move the joint head to the first set point position R and to the second set point position T, respectively, and storing the first set point position R with respect to the third coordinate system KS 3 The coordinates of the second set point position τ are used to readjust the two correction vectors 1^ and K2. In this illustration, the respective set point positions of the mark 10 are respectively by the inverse functions F-1 and cr1 Next, another embodiment will be described in which when the first 8 bits of the joint-16-200929425 are in the first or second set point position, the mark 10 is additionally stored with respect to the first coordinate system KS12. The coordinates (or any other random reference point on the bond head 8) or the coordinates of the mark 10 (or any other random reference point on the bond head 8) of the second coordinate system KS2 are then used for the second Correction of correction vector 1^ and K2. A portion of the pick-up system includes the pick-up system for picking up the semiconductor wafers from the wafer station. The third coordinate system KS3 is the pick-up system φ or the intrinsic coordinate system of the pick-up system, and thus will hereinafter be referred to as the KS coordinate system. In order to ensure that the readjustment can be performed, an adjustment is first performed in a set phase, wherein the bond head 8 is moved to a first set point position (which is within the field of view of the first camera 6), and determined and stored. The coordinates (XSPI, ysP1) of the first set point position of the coordinate system KS and the coordinates (pSPI, qspi) of the first set point position of the coordinate system KSii of the first camera 6. The re-adjustment occurs in the production phase in that the joint head 8 is moved to the coordinates of the first set point position (XSP1, ΟρΡ1), and the coordinate system ks1 regarding the first camera 6 is again determined. Set the coordinates of the point (PSPl'' qspi'). The vector difference of (PSFM', qSPl')-(PSP 丨, qspi) contains information about the displacement of the first coordinate system KSi relative to the coordinate system KS, wherein the coordinate system KS has occurred since the setting phase was set . Any random reference point on the bond head 8 can be used to define a first set point position of the bond head 8 relative to the first coordinate system KS. As described above, the mark 1 is preferably used to define the reference point. Similarly, detecting and correcting the position of the second coordinate system KS2 -17- 200929425 of the second camera 7 relative to the coordinate system KS of the bonding head 8 enables further adjustment during the setting phase. The 8 series is moved to a second set point position within the field of view of the second camera 7, and the coordinates (xs^'yspO and the second are determined and stored with respect to the second set point position of the coordinate system KS) The coordinate of the second set point position of the coordinate system KS2 of the camera 7 (USP2, VSP2). This re-adjustment in this production phase occurs in the following manner: the joint head 8 is moved to the second set point φ position Coordinates (xsP2, ySP2), and then determine the coordinates of the second set point position (USP2', VsP2'p (uSP2'' vsP2')-(uSP2' VSP2) of the coordinate system KS2 of the second camera 7. The difference contains information about the position of the second coordinate system KS2 relative to the coordinate system KS, wherein the coordinate system KS has occurred since the setting in the setting phase. Also, in this case, at the bonding head 8 any A random reference point can be used to define a second setpoint position of the bond head 8 relative to the second coordinate system KS2. As noted above, the marker 10 is preferably used to define the reference point. The coordinates of the coordinate system KS and the second coordinate system KS2i include the respective images taken by the cameras 6 and 7, and the coordinates of the reference point are determined by the conventional image evaluation. Then, the assembly of the semiconductor wafers is better. Ground occurs in the following manner: - the position of the semiconductor wafer 2 to be assembled by the first camera 6 and then the position of the semiconductor wafer 2 is provided in the form of information on the position of the first coordinate system; - by the second camera 7 detected, will be mounted on the substrate on the semiconductor -18- 200929425, the position of the bulk wafer 2 at the location of the substrate, in the form of information on the location of the second coordinate system KS: - in the set phase 'a first mapping function maps the first coordinate system KS to the coordinate system KS and determines its inverse function, and sets a first correction vector to 0 値, a second mapping The function maps the second coordinate system KS2 to the coordinate system KS and determines its inverse function, and sets a second correction vector to 0 値; 0 - in the production phase, one semiconductor wafer 2 is connected to the other, Having imaged the semiconductor wafer 2 to be assembled with the camera 6 and determining the position of the semiconductor wafer 2 relative to the first coordinate system KS, and by considering the first mapping function and considering the first a correction vector from which it is calculated that the pick-up system 5 needs to move the bond head 8 to pick up the position of the semiconductor wafer 2 in relation to the KS coordinate system; - to photograph the semiconductor wafer to be mounted on the substrate with the second camera 7 2, its image at the location of the substrate, and determining the position of the substrate relative to the second coordinate system KS2, and calculating the access system therefrom by the second mapping function and considering the second correction vector 5 that the bonding head 8 needs to be moved to mount the semiconductor wafer 2 at a position of the substrate relative to the KS coordinate system; and re-adjust at the production stage The whole step comprises: re-adjusting the first correction vector and the second correction vector by: - moving the bonding head 8 to the first set point position; -19- 200929425 - taking the mark 10 with the first camera 6 The image of the camera 6 determines the actual position of the marker 10, and - calculates the difference between the first correction vector K1 as the actual position of the storage setting; " - moves the bonding head 8 to the second Setting the position of the mark • Taking the image of the mark 10 by the second camera 7 The image of the second camera 7 determines the actual position of the mark 10 with respect to the KS 2, and - calculating the second corrected vector K2 as the actual position of the storage setting difference between. Having shown and described the specific embodiments of the present invention in the technical field, it is far from being practicable without departing from the scope of the invention. Therefore, the scope of the appended patent application 舆 © the present invention is not limited. BRIEF DESCRIPTION OF THE DRAWINGS One or more detailed descriptions of the invention are intended to be illustrative of the invention and are intended to be These patterns are shown. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a view showing a semiconductor wafer to be mounted; an image, and a fixed position and a judgment from the first coordinate system, and a fixed position and a position from the second coordinate system. The application is subject to change, but the scope of the description that can be readily understood by the owner is I, the equivalent of the scope, the embodiment, and the implementation, which is not the true proportion of the assembly machine -20- 200929425 2 shows a side view of a camera, a bonding head and a wafer table, and Figure 3 shows a top view of the bonding head and three different coordinate systems. [Main component symbol description]
1 晶 圓 台 2 半 導 體 晶 片 3 基 板 台 4 基 板 5 取 置 系 統 6 第 一 攝 影 機 7 第 二 攝 影 徴 8 接 合 頭 9 晶 片 夾 具 10 標 記 11 透 鏡 12 旋 轉 軸 13 夾 具 軸 KSi 第 一 攝 影 機 之 座 標 系 統 KS2 第 二 攝 影 機 之 座 標 系 統 KS3 接 合 頭 之 座 標 系 統 F、G 映 射 函 數 F-1、G'1 映 射 函 數 之 反 函 數 D、E 向 量 -21-1 Wafer table 2 Semiconductor wafer 3 Substrate table 4 Substrate 5 Pick-up system 6 First camera 7 Second image 徴 8 Bond head 9 Chip holder 10 Mark 11 Lens 12 Rotary shaft 13 Fixture axis KSi Coordinate system KS2 of the first camera Coordinate system of the second camera KS3 coordinate system of the joint head F, G mapping function F-1, G'1 inverse function of the mapping function D, E vector-21-