201024717 六、發明說明： 【發明所屬之技術領域】 ' 本發明是關於一種同時地可測定形於 TAB (Tape201024717 VI. Description of the invention: [Technical field to which the invention pertains] ' The present invention relates to a simultaneously measurable shape to TAB (Tape)
Automated Bonding)卷帶等的透光性的基板的配線圖案的 上部線寬與下部線寬的圖案檢查裝置。 【先前技術】 φ 一般，藉由蝕刻，在基板形成配線圖案（以下也稱爲 圖案），則其斷面是有成爲下部寬度比上部寬度還要寬的 梯形狀的趨勢。所以，當蝕刻不足，則即使上部線寬爲良 品的範圍內，而在下部產生稱爲「留下根」的配線變粗， 產生與相鄰的配線引起短路的可能性。所以，在配線圖案 的檢查，測定圖案的下部寬度很重要。 眾知作爲測定圖案的下部寬度的檢查方法，有將照明 光透射形成有圖案的基板來進行（依透射照明的檢查（例如 φ 參照專利文獻1或專利文獻2)。 在專利文獻1 ’揭示利用透射照明來測定配.線圖案的 下部線寬’檢測出是否產生短路等。在專利文獻2的第5 圖，也揭示利用透射照明，可測定出配線圖案的下部線 寬。 專利文獻1 ··日本特開2003-303862號公報 專利文獻2:日本特開2000-113191號公報 【發明內容】 -5- 201024717 使用透射照明，則圖案的下部線寬度是可測定。但 是’無法測定上部的線寬度。但是，最近不但圖案的下部 線寬度’而且成爲也盼望須測定上部的線寬度。其理由表 示於以下。 在第9圖表示形成於基板上的配線圖案的斷面形狀。 如上述地，利用蝕刻所形成的圖案斷面，是成爲下部的寬 度比上部還要寬廣的梯形狀。亦即，如第9(a)圖所示地， 配線圖案是下部寬度b對於上部寬度a是成爲a<b，而將 高度作爲h，則斷面積S是成爲S = (a + b)xh/2。 但是，最近，利用配線圖案的微細化，下部線寬是良 品範圍內，惟上部線寬極窄小，如第9(b)圖所示地，配線 圖案的斷面形狀有成爲三角形狀的情形。 如第9(b)圖所示地，斷面形狀成爲三角形，將配線圖 案的下部寬度作爲b，將高度作爲h，則斷面積S’是成爲 S’ = bxh/2，即使下部寬度相同，與斷面爲梯形狀的情形相 比較，斷面積變小。 亦即與第9(a)圖相比較，第9(b)圖的斷面積是面積僅 小 a X h / 2 〇 配線圖案的斷面積是由流在其圖案的電流値所設計。 因此，在圖案的檢查裝置中，斷面積比容許範圍還要小的 圖案，是必須作爲不良品。 但是，僅測定配線圖案的下部寬度，無法知道圖案的 斷面積的大小。爲了求出斷面積，必須測定圖案的上部與 下部的配線寬度。 -6- 201024717 然而，眾知欲檢測出配線圖案的上部寬度，可利用反 射照明光。 因此，對於欲檢查的圖案，照射反射照明光來測定圖 案的上部線寬，之後，照射透射照明光來測定圖案的下部 線寬，則由其雙方的測定結果就可求出斷面積。 但是，在其方法中，對於一個檢查圖案，成爲進行依 反射照明光的測定與依透射照明光的測定的兩次測定，使 Φ 得檢查時間變久。所以，期盼同時地可進行測定圖案的上 部線寬，與測定下部線寬的檢查裝置。 本發明是鑑於上述事項而發明者，本發明的目的，是 在於在圖案的檢查裝置，同時地可進行檢測出圖案的上部 形狀，及檢測出下部形狀。 在本發明中，如下地解決上述課題。 依據在形成於透光性的基板上的配線圖案照射照明光 所攝影的畫像，進行判定上述圖案的良否的配線圖案檢査 φ 裝置，對於形成在透光性的基板上的圖案，設置從3個不 同的方向進行照明的照明手段，而從3方向同時地照明。 第1照明手段是從形成有基板的圖案的一側，對於檢 查領域斜斜地入射照明光的方式進行照射。第2照明手段 是從與形成有基板的圖案的一側的相反側。對於檢査領域 斜斜地入射照明光的方式進行照射。第3照明手段是從與 形成有基板的圖案的一側的相反側，對於檢查領域大約正 交地入射照明光的方式進行照射。 藉由如此地進行照明，同時地可檢測出圖案的上部形 201024717 狀與下部形狀，例如，可檢測出配線圖案的下部形狀是正 常，惟在上部的一部分產生缺口等的缺陷的情形。 又，視需要，依照所檢測的圖案的上部形狀（線寬）與 下部形狀（線寬），也可計算出圖案的斷面積。比較所計算 的圖案的斷面積是否在容許範圍內，若爲容許範圍外，則 將其圖案作爲不良者。 在本發明中，對於檢査圖案，以1次測定，就可同時 地檢測出圖案的上部形狀與下部形狀。因此，成爲檢查時 間不會變久，就可檢測出配線圖案的缺陷等。 【實施方式】 第1圖是本發明的實施例的配線圖案檢査裝置的方塊 圖。又，以下的實施例是針對於基板爲所謂TAB卷帶或 COF(Chip On Film)的薄膜狀工件的情形加以說明，惟本發 明是也可適用於具透光性的其他的基板的圖案檢查。 如同圖所示地，本實施例的圖案檢査裝置是具備：搬 運TAB卷帶5的送出捲軸21或捲取捲軸22等所構成卷 帶搬運機構20，在從送出捲軸21所送出的TAB卷帶5照 射照明光並進行攝影檢查圖案6的檢查部1，將檢查部1 在TAB卷帶的檢査圖案6上進行掃描的掃描手段2，在不 良圖案標示標誌的標誌部3。 在標誌部3，對於被判定爲不良的圖案，施以用衝孔 機的穿孔，或其部分爲不良品以目視就可立即確認的方 式，施以塗上顔色等的標誌。 -8 - 201024717 又，圖案檢查裝置是具備控制部4。控制部4是控制 檢查部1，標誌部3，及卷帶搬運機構20的動作，而依據 所檢測出的圖案的上部與下部形狀進行檢測圖案的缺陷。 又，依據上部與下部的線寬進行計算斷面積，也可判定圖 案之良否。 檢查部1是具備：對於TAB卷帶5，從形成有圖案的 一側斜斜地照射照明光的第1照明手段1 a，從與形成有圖 φ 案的一側的相反側斜斜地照射照明光的第2照明手段1 b， 從與形成有圖案的一側的相反側對於檢查領域（大約）正交 地入射的方式照射照明光的第3照明手段1 c，與第1照明 手段la相同側，設於檢查領域的正上方方向的攝影手段 1 1 » 第1、第2、第3照明手段la、lb、lc的光源，在本 實施例爲使用LED，惟使用鹵素燈也可以。將鹵素燈使用 作爲光源時，則藉由導光光纖引導來自燈的光，而成爲所 Φ 盼望的角度的方式分別設定著從光纖所出射的光對於檢查 領域的入射角。 攝影手段11是對上述照明光的波長具有受光靈敏度 的例如CCD線感測器或面感測器。 又，在攝影手段的光入射側，設有放大進行TAB卷 帶5的檢查的領域並予以投影的透鏡（未圖示）。又，此透 鏡是組合複數透鏡而被收納於鏡筒者。 控制部4是控制第1、第2、第3照明手段1 a、1 b、 的照明光的點燈與熄燈，及攝影手段11的攝影’依掃 -9 - 201024717 描手段2的檢查部1的移動，TAB卷帶5的搬運。 又，控制部4是依據所檢測的圖案的上部與下部形狀 來判定圖案的良否。又，依據上部與下部的線寬來計算圖 案的斷面積，而從該斷面積來判定圖案的良否也可以。 因此，在控制部必須事先輸入圖案的上部線寬，下部 線寬的容許範圍，配線圖案的斷面積的容許範圍等。 在第2圖及第3圖，表示檢查部1的擴大圖。第2圖 是檢査部的立體圖，第3圖是沿著TAB卷帶5的長度方 向的斷面圖。 如第2圖所示地，第1照明手段1 a與第2照明手段 1 b，是在檢查領域全領域全面地，從全方向入射相同入射 角度的光，或是從各方向所入射的光強度成爲相同的方式 可進行照明，而圓環狀地配置複數LED的構成，例如第3 圖所示地，在LEDlOa的光出射側設置稜鏡片l〇b的構 成。又，在其光出射側，安裝有擴散板10c。 第3照明手段lc是沿著直線狀的檢查領域配置 LEDlOa，而在其光出射側安裝擴散板l〇c者。 在第4圖表示上述圓環狀地形成的第1照明手段la 與第2照明手段lb的具體性的構成例。第4(a)圖是表示 扇型式地切剖的稜鏡片，第4(b)圖是表示使用圖示於第 4(a)圖的稜鏡片以構成照明手段時的從第4(a)圖的A方向 所觀看時的光的出射方向，而第4(c)圖表示使用圖示於第 4(a)圖的稜鏡片以構成照明手段時的從第4(a)圖的B方向 所觀看時的光的出射方向。 -10- 201024717 稜鏡片1 〇b是在透明的片的單面側排列斷面呈三角形 的多數稜鏡成爲平行者，如第4(a)圖所示地，稜鏡的長軸 方向朝著扇型的圓弧的切線方向的方式扇狀切剖該稜鏡 片。又，將該稜鏡片予以圓環狀地排列，並配置於安裝在 支撐構件12上的LED 10a的光出射側。在其上面又安裝擴 散板10c，以構成第1照明手段la，第2照明手段lb。 從LED 10a所出射的主光線平行的光，是入射於稜鏡 φ 片l〇b，從第4(a)圖的A方向所觀看時，在稜鏡折射而如 第4(b)圖所示地仍然保持平行狀態，以一定角度入射於攝 影領域R。又，從第4(a)圖的B方向所觀看時，如第4(c) 圖所示地，在稜鏡不會折射而朝下方向照射。 回到第1圖，攝影手段11是攝影藉由第1、第2、第 3照明手段la、lb、lc同時地被照明的檢查領域。又，被 使用於攝影手段11的CCD是線感測器，而攝影領域是沿 著CCD線感測器的細長領域。攝影手段11是與第1、第 φ 2、第3照明手段la、lb、lc 一體而朝TAB卷帶5的寬度 方向移動，進行攝影檢査領域整體。 以下，表示爲了作成同時地進行圖案的上部線寬與下 部線寬的測定，調査進行那一種照明就可以的實驗結果。 第5圖是表示進行檢查的樣品的圖案者，第5(a)圖是 從上面觀看圖案的圖式，第5(b)圖是第5(a)圖的A-A斷面 圖。又，第5圖是模式地表示以雷射顯微鏡觀察實際的圖 案的結果者。 如第5(a)、（b)圖所示地，樣品圖案是下部寬爲約 -11 - 201024717 20μπι，上部寬度14μχη，斜面的寬度爲3μιη，而圖案的高 度爲7〜8μηι左右，如同圖所示地，與良品比較’配線的 上部缺83 % (下部也有缺口，惟上部的缺口較大）。因此， 以圖案檢查裝置進行此圖案時，可檢測出下部的線寬之同 時，可檢測出在上部有「83%的缺口」較理想。 又，該缺口的比率，是在攝影的圖案畫像，由缺口的 部分的像素一個一個的亮度計算所求出。 在實驗中，使用（a)對於TAB卷帶，從形成有圖案的 —側斜斜地照射照明地的第1照明手段1 a，（b)從與形成 有圖案的一側的相反側斜斜地照射照明光的第2照明手段 lb，（c)從與形成有圖案的一側的相反側對於檢查領域（大 約）正交地入射方式照射照明光的第3照明手段lc，將上 述（a)，（b)，（c)的3種類的照明，分別以單獨地照明的情 形，及經組合進行照明的情形，利用攝影元件1 1進行攝 影樣品圖案的畫像而加以比較。 (1) 未使用第3照明手段lc(正交透射）的情形。 針對於僅第1照明手段la(斜斜地反射）的照明，僅第 2照明手段lb(斜斜地透射）的照明，第1照明手段la與第 2照明手段lb的同時照明（斜斜地反射+斜斜地透射）的各 個照明，以攝影手段進行上述樣品的圖案的畫像。 任何情形，畫像黑暗、對比不好，不僅圖案的下部線 寬’上部線寬也很難確認，可知很難精度優異地測定圖案 線寬。 (2) 僅以第3照明手段lc(正交透射）進行照明樣品圖案 201024717 的情形。 第6(a)圖是模式地表示僅以第3照明手段ic(正交透 射）進行照明樣品圖案時的畫像的圖式。如此地，當使用 第3照明手段lc，則沒有配線圖案的基板的部分透射照明 光’可得對比良好的畫像。又，可檢測出圖案的下部線寬 並可加以測定。 但是，僅在第3照明手段lc，無法測定圖案的上部線 ❹ 寬。在第6(a)圖所檢測出的圖案的上部缺口的比率是 47%。 如上述地，實際的缺口比率是8 3 %，所檢測出的缺口 大小是與實際者相比較小到只有一半左右。此乃爲未能正 確地檢測出圖案的上部線寬。 (3)在第3照明手段lc(正交透射）加上第丨照明手段 la(斜斜地反射）或是第2照明手段lb(斜斜地透射）進行照 明的情形。 Φ 第6(b)圖是在第3照明手段lc(正交透射）加上第1照 明手段la(斜斜地反射）進行照明的情形，第6(c)圖在第3 照明手段lc(正交透射）加上第2照明手段lb(斜斜地透射） 的情形。 如第6(b)、（c)圖所示地，在利用第3照明手段lc(正 交透射）的照明上，加上利用第2照明手段lb或第3照明 手段的照明，則可以檢測出下部線寬與上部線寬者。 但是，第6(b)圖的情形是圖案的上部缺口的大小被檢 測出爲64%，而6(c)圖的情形是圖案的上部缺口的大小被 201024717 檢測出爲5 5 %。任何情形，都比僅第6 (a)圖的第3照明手 段lc(正交透射）的情形，缺口大小接近於正的數値 (83%)，不能說是充分者。 (4)在第3照明手段lc(正交透射）加上第丨照明手段 1 a(斜斜地反射）及第2照明手段1 b(斜斜地透射）進行照明 的情形。 第6(d)圖是同時地照明第3照明手段lc(正交透射）， 第1照明手段la(斜斜地反射），第2照明手段ib(斜斜地 透射）的全部的情形。被檢測出的缺口大小是73%，以最 接近於實際的缺口比率的數値。又，下部及上部線寬，也 與上述（3)的情形相比較，更清楚地可檢測出。 又，如後述地，僅第3照明手段〗c(正交透射）的照明 的情形’則圖案的側面與下面部黑黑（黑暗）地被顯示，惟 在此加上來自第1照明手段la(斜斜地反射）的照明光，或 是來自第2照明手段1 b (斜斜地透射）的照明光，則圖案的 側面成稍明亮。藉此，可區別圖案的上部與圖案的側面， 成爲可檢測出上部的線寬。 第6(e)圖是模式地表示擴大第6(d)圖的缺口部分的畫 像者’在此’表7K配線圖案的下部寬20 μιη，上部寬 16μιη，側面寬2μιη的情形，如同圖所示地，圖案的側面 部分是比圖案上部稍明亮，而由此就可求出圖案下部的寬 度。 上述第6(a)圖至第6(e)圖是模式地表示畫像者，惟將 在攝影手段1所攝影的畫，在控制部4進行畫像處理，依 201024717 據1像素別的亮度，來計算缺口大小，就可計測缺口的大 小。 如以上述地，利用同時地進行第1、第2、第3照 明，以1次測定，就可較清楚地檢測出圖案的上部形狀與 下部形狀，藉此，成爲可檢測出圖案的缺陷。 又，同時地可檢測出圖案的下部線寬與上部線寬，也 可求出斷面積。 φ 第6(d)圖的情形，依據攝影的畫像的1像素別的亮度 進行線寬，則下部線寬是約20μιη，發生缺口的部分的上 部線寬是大約4μιη。 因此，發生缺口的部分的斷面積，是成爲（20μιη + 4μη〇 X圖案高度xl/2。又，圖案高度是無法從第6圖的畫像求 出之故，因而代入設計値。 在裝置的控制部4，事先輸入流在圖案的電流値的斷 面積的下限値。控制部4是將在上述計算所得到的圖案的 缺口部分的斷面積與此斷面積的下限値比較，若比下限僅 小，則將其圖案作爲不良。 利用同時地進行第1、第2、第3照明，就可同時地 檢測出圖案的下部線寬與上部線寬的理由，是有如下理 由。 針對對此情形，使用第7圖加以說明。 利用來自第3照明手段lc(正交透射）的照明光（3)，檢 測出圖案的下部線寬。 還有，入射來自第1照明手段1 a(斜斜地反射）的照明 -15- 201024717 光（1)，及入射來自第2照明手段lb(斜斜地透射）的照明 光（2)。又，來自第1照明手段la的照明光（1)，及來自第 2照明手段lb的照明光（2)，是在第7圖中，表示成從圖 中左右入射，惟實際上，對於圖案從360。全方向入射有 光。 實際的圖案的側面並不是平滑的面，而發生多數的細 凹凸。因此’來自第1照明手段la的照明光（1)，及來自 第2照明手段lb的照明光（2)，是在此圖案的側面漫反 射，使得其一部分入射於攝影手段11。藉此，圖案的側面 是比圖案的上面稍明亮地被攝影。 亦即，僅第3照明手段1 c(正交透射）的照明的情形， 則圖案的側面與上面都顯示黑黑（黑暗），惟在此加上來自 第1照明手段1 a(斜斜地反射）的照明光（i)與來自第2照明 手段lb(斜斜地透射）的照明光（Π)，則圖案的側面成爲稍 明亮，而可區別圖案的上部與圖案的側面。因此，圖案上 部的形狀成爲明確，可檢測出其線寬。 藉由以上，同時地可檢出圖案的下部形狀與上部形 狀。 第8圖是表示用以求出第1照明手段la與第2照明 手段lb的照明光的最適當角度的實驗結果的圖式。 將照明TAB卷帶5的照明手段30,從與形成有圖案 的一側的相反側，移動至形成有圖案的一側，藉由攝影樣 品圖案的側面的明亮的變化的畫像進行測定。 如第8(a)圖所示地，照明手段3〇是照明光爲將從與 201024717 成有圖案的一側的相反側正交所入射的位置（亦即第3照 明手段的位置）作爲0°，並朝著形成有圖案的一側移動至 160°的位置進行測定。 又，在本實驗中，很難變更從圓環狀照明手段所出射 的照明光的入射角度之故，因而將LED放在檢査領域的 長度方向兩側，如第8圖所示地，藉由移動其LED，來變 更照明光對於TAB卷帶的入射角度。在此，作爲照明手 φ 段30使用將晶片排成1列的LED，流著70mA的電流。 在第8(b)圖表示其結果。橫軸是照明手段的角度〇, 縱軸是圖案的側面亮度（任意的單位）。如同圖所示地，以 約30°〜60°範圍進行照明的情形，與以120°以上進行照明 的情形，使得圖案的側面成爲明亮。圖案的側面愈明亮， 會使與圖案上部之境界愈成爲明確之故，因而適用作爲照 明手段的位置。 因此，第1照明手段la是照明光對於檢查領域的入 φ 射角度設定成爲12 0°〜16 0°的範圍。又’第2照明手段lb 是照明光對於檢査領域的入射角度設定成爲30°〜60°的範 圍。 【圖式簡單說明】 第1圖是本發明的實施例的配線圖案檢查裝置的方塊 圖。 第2圖是擴大第1圖的檢査部的立體圖。 第3圖是沿著TAB卷帶的長度方向切剖第1圖的檢 -17- 201024717 查部的斷面圖。 第4(a)圖至第4(c)圖是表示圓環狀地形成的第1照明 手段la與第2照明手段lb的具體性構成的圖式。 · 第5(a)圖及第5(b)圖是模式地表示進行檢查的樣品的 圖案的圖式。 第6(a)圖至第6(d)圖是模式地表示變更照明手段的組 合而攝影第5圖的樣品時所得到的畫像的圖式。 第7圖是說明利用同時地進行第1、第2、第3照 _ 明，而可同時地檢測出圖案的下部線寬與上部線寬的理由 的圖式。 第8(a)圖及第8(b)圖是表示用以求出第1照明手段la 與第2照明手段lb的照明光的最適當角度的實驗結果的 圖式。 第9(a)圖及第9(b)圖是表示形成於基板上的配線圖案 的斷面形狀的圖式。 【主要元件符號說明】 1 :檢査部 1 a :第1照明手段 1 b :第2照明手段 1 c :第3照明手段 2 :掃描手段 3 :標誌部 4 :控制部 18- 201024717 5 : TAB卷帶 6 :檢查圖案 10a ： LED l〇b :稜鏡片 l〇c :擴散板 1 1 :攝影手段 1 2 :支撐構件 φ 20 :卷帶搬運機構 21 :送出捲軸 22 :捲取捲軸 3 〇 :照明手段 -19-Automated Bonding) A pattern inspection device for an upper line width and a lower line width of a wiring pattern of a light-transmissive substrate such as a tape. [Prior Art] φ Generally, when a wiring pattern (hereinafter also referred to as a pattern) is formed on a substrate by etching, the cross section has a tendency to have a trapezoidal shape in which the lower width is wider than the upper width. Therefore, when the etching is insufficient, even if the upper line width is within the range of the good product, the wiring called "leaving the root" is thickened in the lower portion, which may cause a short circuit with the adjacent wiring. Therefore, in the inspection of the wiring pattern, it is important to measure the lower width of the pattern. It is known that the inspection method of the lower width of the measurement pattern is performed by transmitting the illumination light to the substrate on which the pattern is formed (in accordance with the inspection of the transmission illumination (for example, φ refer to Patent Document 1 or Patent Document 2). In the transmission illumination, the lower line width of the line pattern is measured to detect whether or not a short circuit has occurred. In the fifth drawing of Patent Document 2, it is also disclosed that the lower line width of the wiring pattern can be measured by the transmission illumination. Patent Document 1 JP-A-2003-303862, JP-A-2000-113191, SUMMARY OF THE INVENTION - 5 - 201024717 When transmission illumination is used, the lower line width of the pattern is measurable. However, the upper line width cannot be measured. However, recently, not only the width of the lower line of the pattern but also the line width of the upper part is expected to be measured. The reason is shown below. Fig. 9 shows the cross-sectional shape of the wiring pattern formed on the substrate. The cross-section of the pattern formed by the etching is a trapezoidal shape in which the width of the lower portion is wider than the upper portion. That is, as shown in Fig. 9(a), The line pattern is such that the lower width b is a < b for the upper width a and the height S is the same, and the sectional area S is S = (a + b) xh/2. However, recently, by miniaturization of the wiring pattern, The lower line width is within the range of the good product, but the upper line width is extremely narrow. As shown in Fig. 9(b), the cross-sectional shape of the wiring pattern has a triangular shape. As shown in Fig. 9(b) The cross-sectional shape is a triangle, the lower width of the wiring pattern is b, and the height is h, the sectional area S' is S' = bxh/2, and even if the lower width is the same, the cross-sectional shape is the same as the trapezoidal shape. The cross-sectional area becomes smaller. That is, compared with the 9th (a) figure, the sectional area of the 9th (b) figure is only a small area of a x h / 2 〇 the cross-sectional area of the wiring pattern is the current flowing in its pattern Therefore, in the pattern inspection apparatus, a pattern having a smaller sectional area than the allowable range is required as a defective product. However, only the lower width of the wiring pattern is measured, and the size of the sectional area of the pattern cannot be known. To determine the sectional area, it is necessary to measure the wiring of the upper and lower parts of the pattern. -6- 201024717 However, it is known to detect the upper width of the wiring pattern, and it is possible to use the reflected illumination light. Therefore, for the pattern to be inspected, the reflected illumination light is irradiated to measure the upper line width of the pattern, and then the transmission illumination is irradiated. When the lower line width of the pattern is measured by light, the cross-sectional area can be obtained from the measurement results of both of them. However, in the method, the measurement of the reflected illumination light and the measurement of the transmitted illumination light are performed for one inspection pattern. In the two-time measurement, the inspection time of Φ is made longer. Therefore, it is expected that the upper line width of the measurement pattern can be simultaneously measured and the inspection apparatus for measuring the lower line width. The present invention has been made in view of the above, and the present invention The purpose is to detect the upper shape of the pattern and the lower shape at the same time in the pattern inspection device. In the present invention, the above problems are solved as follows. The wiring pattern inspection φ device for determining whether the pattern is good or not is formed by irradiating the image captured by the illumination light on the wiring pattern formed on the light-transmitting substrate, and the pattern formed on the light-transmitting substrate is set to three. Lighting means for illuminating in different directions, while illuminating simultaneously from 3 directions. The first illumination means irradiates the illumination light obliquely from the side where the pattern of the substrate is formed, obliquely to the inspection area. The second illumination means is from the side opposite to the side on which the pattern of the substrate is formed. Irradiation is performed in such a manner that the inspection field obliquely enters the illumination light. The third illumination means irradiates the illumination light in a manner orthogonal to the inspection area from the side opposite to the side on which the pattern of the substrate is formed. By illuminating in this manner, the shape of the upper portion of the pattern and the shape of the lower portion can be detected at the same time. For example, it is possible to detect that the shape of the lower portion of the wiring pattern is normal, but a defect such as a notch is formed in a part of the upper portion. Further, the sectional area of the pattern can be calculated in accordance with the upper shape (line width) and the lower shape (line width) of the detected pattern as needed. It is compared whether the calculated sectional area of the pattern is within the allowable range, and if it is outside the allowable range, the pattern is regarded as a bad one. In the present invention, for the inspection pattern, the upper shape and the lower shape of the pattern can be simultaneously detected by one measurement. Therefore, defects such as wiring patterns can be detected without increasing the inspection time. [Embodiment] Fig. 1 is a block diagram of a wiring pattern inspection device according to an embodiment of the present invention. Further, the following embodiments are described in the case where the substrate is a film-like workpiece of a so-called TAB tape or COF (Chip On Film), but the present invention is also applicable to pattern inspection of other substrates having light transmissivity. . As shown in the figure, the pattern inspection device of the present embodiment includes a tape transport mechanism 20 including a delivery reel 21 or a take-up reel 22 that transports the TAB tape 5, and the TAB tape is fed from the delivery reel 21. (5) The inspection unit 1 that irradiates the illumination light and performs the photographing inspection pattern 6, and the scanning means 2 that scans the inspection pattern 6 of the TAB tape by the inspection unit 1 is marked on the indicator portion 3 of the defective pattern. In the indicator portion 3, a punch which is determined to be defective is applied with a punch for punching, or a portion thereof is a defective product which can be immediately confirmed by visual inspection, and a mark such as a color is applied. -8 - 201024717 Further, the pattern inspection device includes the control unit 4. The control unit 4 controls the operation of the inspection unit 1, the indicator unit 3, and the tape transport mechanism 20, and detects defects in the pattern in accordance with the upper and lower shapes of the detected pattern. Further, by calculating the sectional area based on the line widths of the upper and lower portions, it is also possible to determine whether the pattern is good or not. The inspection unit 1 is provided with a first illumination means 1a for obliquely illuminating the illumination light from the side on which the pattern is formed on the TAB tape 5, and obliquely irradiated from the side opposite to the side on which the pattern φ is formed. The second illumination means 1b for illuminating light, the third illumination means 1c that illuminates the illumination light so as to be orthogonally incident on the inspection area (about) on the side opposite to the side on which the pattern is formed, and the first illumination means 1a On the same side, the photographing means 1 1 » the light source of the first, second, and third illumination means 1a, 1b, and 1c are provided in the direction of the inspection area. In the present embodiment, the LED is used, but a halogen lamp may be used. When a halogen lamp is used as the light source, the light from the lamp is guided by the light guiding fiber, and the angle of incidence of the light emitted from the optical fiber with respect to the inspection field is set so as to be an angle of Φ. The photographing means 11 is, for example, a CCD line sensor or a surface sensor having a light receiving sensitivity to the wavelength of the illumination light. Further, a lens (not shown) that enlarges and projects the area where the inspection of the TAB tape 5 is performed is provided on the light incident side of the photographing means. Further, this lens is a combination of a plurality of lenses and is housed in a lens barrel. The control unit 4 controls the lighting and the light-off of the illumination light of the first, second, and third illumination means 1a, 1b, and the imaging of the imaging means 11 by the inspection unit 1 of the scanning means 2 - 201024717 The movement, TAB tape 5 handling. Further, the control unit 4 determines whether or not the pattern is good or bad depending on the upper and lower shapes of the detected pattern. Further, the sectional area of the pattern is calculated based on the line widths of the upper portion and the lower portion, and the quality of the pattern may be determined from the sectional area. Therefore, the control unit must input the upper line width of the pattern, the allowable range of the lower line width, the allowable range of the sectional area of the wiring pattern, and the like. In the second and third figures, an enlarged view of the inspection unit 1 is shown. Fig. 2 is a perspective view of the inspection unit, and Fig. 3 is a cross-sectional view along the length direction of the TAB tape 5. As shown in Fig. 2, the first illumination means 1a and the second illumination means 1b are all light that enters the same incident angle from all directions in the entire field of inspection, or light incident from each direction. In the same manner, the illumination can be performed in the same manner, and the configuration of the plurality of LEDs is arranged in a ring shape. For example, as shown in FIG. 3, the configuration of the cymbal 10b is provided on the light exit side of the LED 10a. Further, a diffusion plate 10c is attached to the light exit side. The third illumination means lc is such that the LED 10a is disposed along the linear inspection area, and the diffusion plate 10c is attached to the light emission side. Fig. 4 shows an example of the specific configuration of the first illumination means 1a and the second illumination means 1b which are formed in the above-described annular shape. Fig. 4(a) is a fragment showing a fan-shaped cut, and Fig. 4(b) is a view showing a fourth (a) when the cymbal of the fourth (a) diagram is used to constitute an illumination means. The direction of light emission when viewed in the A direction of the figure, and Fig. 4(c) shows the B direction from the 4th (a) figure when the cymbal shown in Fig. 4(a) is used to constitute the illumination means. The direction in which the light is emitted when viewed. -10- 201024717 稜鏡 1 是 是 是 是 是 是 是 是 是 是 是 是 是 是 是 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡 稜鏡The sipe is cut in a fan shape in a tangential direction of the circular arc of the fan. Further, the cymbal pieces are arranged in a ring shape and arranged on the light outgoing side of the LED 10a attached to the support member 12. Further, a diffusion plate 10c is attached to the upper surface to constitute a first illumination means 1a and a second illumination means 1b. The light parallel to the chief ray emitted from the LED 10a is incident on the 稜鏡φ sheet l〇b, and is viewed in the 方向 direction when viewed from the direction A of the fourth (a) diagram, as shown in the fourth (b) diagram. The display remains parallel and is incident on the field of photography R at a certain angle. Further, when viewed in the B direction of the fourth (a) diagram, as shown in Fig. 4(c), the crucible is not refracted and is irradiated downward. Returning to Fig. 1, the photographing means 11 is an inspection field in which the first, second, and third illumination means la, lb, and lc are simultaneously illuminated. Further, the CCD used in the photographing means 11 is a line sensor, and the field of photographing is an elongated field along the CCD line sensor. The photographing means 11 is integrated with the first, second, and third illumination means 1a, 1b, and 1c, and moves in the width direction of the TAB web 5 to perform the entire photographing inspection field. Hereinafter, in order to measure the upper line width and the lower line width of the pattern, the experimental results of the illumination can be investigated. Fig. 5 is a view showing a pattern of a sample to be inspected, Fig. 5(a) is a view in which a pattern is viewed from above, and Fig. 5(b) is a cross-sectional view taken along line A-A in Fig. 5(a). Further, Fig. 5 is a view schematically showing the result of observing the actual pattern by a laser microscope. As shown in Fig. 5(a) and (b), the sample pattern has a lower width of about -11 - 201024717 20μπι, an upper width of 14μχη, a bevel width of 3μηη, and a pattern height of about 7~8μηι, as shown in the figure. As shown in the figure, compared with the good product, 'the upper part of the wiring is lacking 83% (the lower part also has a notch, but the upper part has a large gap). Therefore, when the pattern is inspected by the pattern inspection device, the line width of the lower portion can be detected, and it is possible to detect that there is "83% of the notch" in the upper portion. Further, the ratio of the notch is obtained by calculating the luminance of the pixel in the notch portion by the pattern image of the photograph. In the experiment, (a) for the TAB tape, the first illumination means 1 a obliquely illuminating the illumination from the side where the pattern is formed, (b) is inclined from the side opposite to the side on which the pattern is formed. The second illumination means 1b that illuminates the illumination light, (c) the third illumination means lc that illuminates the illumination light orthogonally to the inspection area (about) on the side opposite to the side on which the pattern is formed, and the above (a) In the case of the three types of illumination of (b) and (c), respectively, in the case of separately illuminating, and in the case of combining illumination, the image of the sample pattern is imaged by the photographic element 1 1 and compared. (1) The case where the third illumination means lc (orthogonal transmission) is not used. For the illumination of only the first illumination means 1a (obliquely reflected), only the illumination of the second illumination means 1b (transversely transmitted), the first illumination means 1a and the second illumination means 1b are simultaneously illuminated (slantly Each of the illuminations of the reflection + oblique transmission is used to perform an image of the pattern of the sample by means of photographing. In any case, the image is dark and the contrast is not good, and not only the lower line width of the pattern, but also the upper line width, is difficult to confirm, and it is found that it is difficult to measure the line width of the pattern with excellent precision. (2) The case where the sample pattern 201024717 is illuminated only by the third illumination means lc (orthogonal transmission). Fig. 6(a) is a view schematically showing an image when the sample pattern is illuminated by the third illumination means ic (orthogonal transmission). As described above, when the third illumination means lc is used, the portion of the substrate having no wiring pattern transmits the illumination light', and a relatively good image can be obtained. Further, the lower line width of the pattern can be detected and measured. However, the upper line width of the pattern cannot be measured only in the third illumination means lc. The ratio of the upper notch of the pattern detected in Fig. 6(a) was 47%. As described above, the actual gap ratio is 83%, and the detected gap size is as small as about half of the actual one. This is the upper line width of the pattern that was not correctly detected. (3) A case where the third illumination means lc (orthogonal transmission) is irradiated with the second illumination means la (obliquely reflected) or the second illumination means 1b (transversely transmitted). Φ Fig. 6(b) shows a case where the third illumination means lc (orthogonal transmission) is irradiated with the first illumination means 1a (obliquely reflected), and the sixth (c) figure is shown by the third illumination means lc ( Orthogonal transmission) plus the second illumination means lb (transversely transmitted). As shown in FIGS. 6(b) and 6(c), the illumination by the third illumination means lc (orthogonal transmission) can be detected by adding illumination by the second illumination means 1b or the third illumination means. The lower line width and the upper line width are the ones. However, in the case of Fig. 6(b), the size of the upper notch of the pattern is detected to be 64%, and in the case of the 6(c) diagram, the size of the upper notch of the pattern is detected by 201024717 as 55 %. In any case, the gap size is close to a positive number 83 (83%) than in the case of the third illumination hand lc (orthogonal transmission) of the sixth (a) diagram, and it cannot be said that it is sufficient. (4) A case where the third illumination means lc (orthogonal transmission) is illuminated by the second illumination means 1a (obliquely reflected) and the second illumination means 1b (transversely transmitted). Fig. 6(d) is a view showing all of the third illumination means lc (orthogonal transmission), the first illumination means 1a (reflected obliquely), and the second illumination means ib (transversely transmitted). The detected gap size is 73%, which is the closest to the actual gap ratio. Further, the lower and upper line widths are more clearly detectable than in the case of the above (3). Further, as will be described later, only the case of illumination of the third illumination means 〖c (orthogonal transmission) is displayed in the black side (dark) of the side surface and the lower surface of the pattern, but the first illumination means la is added here. The illumination light (reflected obliquely) or the illumination light from the second illumination means 1b (transversely transmitted) is slightly brighter on the side surface of the pattern. Thereby, the upper portion of the pattern and the side surface of the pattern can be distinguished, and the line width of the upper portion can be detected. Fig. 6(e) is a view schematically showing a case where the image of the notch portion of the sixth figure (d) is enlarged. Here, the lower portion of the wiring pattern of the table 7K is 20 μm wide, the upper portion is 16 μm wide, and the side width is 2 μm. In the illustration, the side portion of the pattern is slightly brighter than the upper portion of the pattern, and thus the width of the lower portion of the pattern can be found. The sixth (a)th to the sixth (e)th diagrams are diagrams schematically showing the portrait, but the image photographed by the photographing means 1 is subjected to image processing by the control unit 4, according to the brightness of one pixel of 201024717. Calculate the size of the gap and measure the size of the gap. As described above, the first, second, and third illuminations are simultaneously performed, and the upper and lower shapes of the pattern can be clearly detected by one measurement, whereby the pattern can be detected. Further, the lower line width of the pattern and the upper line width can be detected simultaneously, and the sectional area can also be obtained. In the case of Fig. 6(d), the line width is set according to the brightness of one pixel of the photographed image, and the lower line width is about 20 μm, and the upper line width of the notched portion is about 4 μm. Therefore, the sectional area of the portion where the notch is formed is (20 μm η + 4 μη〇X pattern height xl/2. Further, the pattern height cannot be obtained from the image of Fig. 6, and is substituted into the design 値. In the portion 4, the lower limit 断 of the cross-sectional area of the current 流 flowing in the pattern is input in advance. The control unit 4 compares the cross-sectional area of the notch portion of the pattern obtained by the above calculation with the lower limit 此 of the cross-sectional area, and is smaller than the lower limit. The reason for this is that the pattern is defective. The reason why the lower line width and the upper line width of the pattern can be simultaneously detected by simultaneously performing the first, second, and third illuminations is as follows. The lower line width of the pattern is detected by the illumination light (3) from the third illumination means lc (orthogonal transmission). Further, the incidence is from the first illumination means 1a (obliquely reflected). Illumination -15-201024717 Light (1), and illumination light (2) incident from the second illumination means lb (obliquely transmitted). Further, the illumination light (1) from the first illumination means la, and from Illumination light (2) of the second illumination means lb, In Fig. 7, it is shown that the light is incident from the left and right in the figure, but actually, the pattern is incident on the light from 360 in all directions. The side surface of the actual pattern is not a smooth surface, but a large number of fine irregularities occur. The illumination light (1) from the first illumination means 1a and the illumination light (2) from the second illumination means 1b are diffused and reflected on the side surface of the pattern so that a part thereof is incident on the imaging means 11. Thereby, the pattern The side surface is photographed slightly brighter than the upper side of the pattern. That is, in the case of only the illumination of the third illumination means 1 c (orthogonal transmission), both the side and the top of the pattern are black (dark), but only The illumination light (i) from the first illumination means 1a (obliquely reflected) and the illumination light (Π) from the second illumination means 1b (transversely transmitted) are slightly brighter, and the side surface of the pattern is slightly brighter. The upper portion of the pattern is distinguished from the side surface of the pattern. Therefore, the shape of the upper portion of the pattern is made clear, and the line width can be detected. By the above, the lower shape and the upper shape of the pattern can be simultaneously detected. Fig. 8 is a view showing The first lighting means la and the first 2 Schematic of the experimental result of the most appropriate angle of the illumination light of the illumination means 1b. The illumination means 30 for illuminating the TAB tape 5 is moved from the side opposite to the side on which the pattern is formed to the side on which the pattern is formed, The measurement is performed by a brightly changing image of the side surface of the photographic sample pattern. As shown in Fig. 8(a), the illumination means 3 〇 is that the illumination light is orthogonal to the side opposite to the side patterned with 201024717. The incident position (that is, the position of the third illumination means) is 0°, and is measured at a position shifted to 160° toward the side on which the pattern is formed. Further, in this experiment, it is difficult to change from the ring shape. The incident angle of the illumination light emitted by the illumination means is such that the LEDs are placed on both sides in the longitudinal direction of the inspection field, and as shown in Fig. 8, the illumination light is incident on the TAB tape by moving the LEDs thereof. angle. Here, as the illumination hand φ segment 30, an LED in which the wafers are arranged in one row is used, and a current of 70 mA flows. The result is shown in the figure 8(b). The horizontal axis is the angle 〇 of the illumination means, and the vertical axis is the side luminance of the pattern (arbitrary unit). As shown in the figure, when the illumination is performed in the range of about 30 to 60, and the illumination is performed in the range of 120 or more, the side surface of the pattern is made bright. The brighter the side of the pattern, the more clearly the boundary with the upper part of the pattern is, and thus the position as a means of illumination is applied. Therefore, the first illumination means 1a is a range in which the illuminating angle of the illumination light to the inspection area is set to be 120 to 160 degrees. Further, the second illumination means lb is a range in which the incident angle of the illumination light in the inspection area is set to 30 to 60. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a block diagram of a wiring pattern inspection device according to an embodiment of the present invention. Fig. 2 is a perspective view showing an enlarged inspection unit of Fig. 1. Fig. 3 is a cross-sectional view of the inspection section taken along the longitudinal direction of the TAB tape, taken along the line -17-201024717 of Fig. 1. 4(a) to 4(c) are diagrams showing specific configurations of the first illumination means 1a and the 2nd illumination means 1b which are formed in an annular shape. - Fig. 5(a) and Fig. 5(b) are diagrams schematically showing the pattern of the sample to be inspected. Fig. 6(a) to Fig. 6(d) are diagrams schematically showing an image obtained when the sample of Fig. 5 is photographed by changing the combination of the illumination means. Fig. 7 is a view for explaining the reason why the lower line width and the upper line width of the pattern can be simultaneously detected by simultaneously performing the first, second, and third illuminations. Figs. 8(a) and 8(b) are diagrams showing experimental results of the optimum angles for obtaining the illumination light of the first illumination means 1a and the second illumination means 1b. Figs. 9(a) and 9(b) are diagrams showing the cross-sectional shape of the wiring pattern formed on the substrate. [Description of main component symbols] 1 : Inspection unit 1 a : First illumination means 1 b : Second illumination means 1 c : Third illumination means 2 : Scanning means 3 : Marking part 4 : Control part 18 - 201024717 5 : TAB volume Belt 6 : Inspection pattern 10a : LED l〇b : cymbal sheet l 〇 c : diffusion plate 1 1 : photographic means 1 2 : support member φ 20 : tape transport mechanism 21 : delivery reel 22 : take-up reel 3 〇: illumination Means-19-