592008 ⑴ 玖、發明說明 【發明所屬之技術領域】 本發明是關於觀測多層印刷配線板的導引記號後，對 該應對之指定的基準孔進行鑽孔的鑽孔機，特別是關於在 對各內層的偏移爲無規則性時等形成有用基準孔的鑽孔機 【先前技術】 _ 最近，伴隨著I c晶片、電阻、電容器等表面安裝用 電子零件的小型化，要安裝該等的印刷配線板也被要求爲 需高密度化，因此印刷配線板大多數都被多層化。即使爲 民生用也是使用4層、6層等的多層印刷配線板，於產業 用上則有使用更多層之高多層印刷配線板的趨勢。 多層印刷配線板是以露出在表背2層外部的導體層， 及數層不露出的內層導體層所構成，於各導體層之間插入 有絕緣性的基板，由該基板使導體層爲結合構造。 β 做爲多層印刷配線板的導體層，例如是使用厚度爲 1 8 程度的銅箔。 做爲基板材料，是以使用熱凝性的玻璃•環氧樹脂爲 1 主流，至於對高多層配線板也使用玻璃•聚醯亞氨樹脂、 玻璃· B T樹脂等耐熱樹脂。 由於多層印刷配線板爲6層以上的配線板單純只是內 層的導體數爲多，因此做爲多層印刷配線板的製造方法如 以下，於此參考第8圖及第9圖，對6層之多層印刷配線 -7- (2) (2)592008 板的製造方法進行簡單說明。 第8 ( a )圖爲表示6層印刷配線板的構成模式透視 · 圖，第8 ( b )圖爲表示內層之兩面配線板的導體部份上 所形成的印刷圖案模式平面圖，第8 ( c )圖爲表示於下 述之結層時所使用之結層用工具板側面圖。第9圖爲6層 配線板的剖面圖，（a )表示其在熱壓工程前的狀態，（ b )表示在熱壓工程基板使其熱凝後，結合導體層成爲1 片多層印刷配線板的狀態。 ® 如第8 ( a )圖所示，6層之多層印刷配線板6 0是 在2層露出的導體層6 2、6 2和爲內層之2片的兩面印 刷配線板6 1 、6 1之間夾著聚酯膠片6 4、6 4 a、 6 4而形成。 於構成內層的兩面印刷配線板6 1 、6 1上，通常是 藉由鈾刻在其表背銅箔面形成有做爲最終成品（於圖中爲 6個）的單一配線板圖案6 1 a.......6 1 a等。 於事先，在上述之兩面配線板6 1 、6 1上至少鑽孔 # 有2個基準孔65、65，由於是以該基準孔65、65 爲基準形成表背兩面的圖案6 1 a .......6 1 a等，因 此於平面上，兩面配線板6 1、6 1的表背兩面的圖案是 ‘ 能夠互相確保其位置。如此般2片兩面配線板6 1 、6 1 、 是使用相同座標位置的基準孔6 5、6 5形成圖案6 1 a .......6 1 a 等。 於成爲內層板的兩面配線板6 1上所形成的圖案中， 除了形成著單一配線板的圖案6 1 a .......6 1 a以外 -8- (3) (3)592008 ，通常還準備有於後工程要使用之4個以上（最少3個） 基準孔用的導引記號6 6、6 6 ......，將這些導引記號 ^ 也在蝕刻工程形成。 把複數片之蝕刻完成的內層用兩面印刷配線板重疊， * 使分別形成在各配線板之導體部上的圖案位置關係得以對 正整齊之步驟稱爲結層（Lay up)。 首先，準備好設有與在兩面印刷配線板6 1上形成圖 案61a .......61a等時所使用之基準孔65、65 φ 的中心距離爲相等中心距離之結層釘6 8 a、6 8 a的結 層用工具板6 8。把1片蝕刻完成的兩面印刷配線板6 1 放在結層用工具板6 8上使結層用工具板6 8的結層釘 68a、68a得以插穿其基準孔65、65。在其上面 鑽孔有基準孔6 5、6 5，使加熱前的基板材料（稱聚酯 膠片）6 4 a得以重疊在其上方。又把另一片兩面印刷配 線板6 1重疊放在結層用工具板6 8上使結層用工具板 68的結層釘68a、68a得以插穿其基準孔65、 · 6 5。於該階段時2片的兩面印刷配線板6 1 、6 1和它 們之間的聚酯膠片6 4 a的外周是被暫時固定而完成結層 〇 如此，上述基準孔6 5、6 5也是做爲結層時的基準 、 孔來使用，故可稱其爲結層用基準孔。以下爲避免產生混 亂時也會將基準孔6 5、6 5稱爲結層用基準孔6 5、 6 5 〇 如第9圖剖面圖所示，當於被結層之2片的兩面印刷 -9 - (4) (4)592〇〇8 配線板6 1、6 1的兩側放置聚酯膠片6 4、6 4和導體 材料的銅箔6 2、6 2再以熱壓機加壓加熱時，插入在銅 箱6 2或兩面配線板6 1之間的聚酯膠片6 4、6 4 a、 6 4會熱凝（絕緣）變化成基板6 3，各導體間的結合也 完成，而形成1片的多層印刷配線板6 0。 然後，應對於多層配線板之內層圖案的新基準孔被開 孔，以該新基準孔爲基準進行最外層之導體配線圖案的触 刻、通孔等的鑽孔加工等。接著又施以電鍍工程、防銹處 理工程等，以機械加工分割成單一配線板，切出成所要的 外形形狀完成多層印刷配線板。 如所說明般，於要形成爲內層板之兩面配線板的導體 層中，除了單一配線板的圖案6 1 a .......6 1 a以外 ’還準備3個以上之基準孔形成用的導引記號6 6、6 6 、6 6等，該導引記號也會在蝕刻工程被進行蝕刻。這些 的導引記號的座標，由於是被定位成與單一配線板的圖案 6 1 a .......6 1 a確保有某種位置關係，因此就變成 要對這些導引記號的位置進行測定，來推定構成電路之圖 案的配置，然後再決定於後工程上要使用的基準孔位置。 爲了要鑽有新的基準孔來應對於形成在上述多層配線 板60之內層的導引記號66、66 .......，通常是使 用X光（基準孔）鑽孔機。 如以上所述，以熱壓機所加壓加熱的多層配線板的表 背兩外面市以無垢的導體層包覆著，用肉眼使用可視光線 是不可能明瞭透視形成在內層的導引記號。 -10- (5) 於現今一般的方式雖然是以微弱的X光透視多層配線 板，來測定形成在內層板上的導引記號位置，但對使用超 音波等其他的測定方式也有種種的硏究。無論是任何方式 ，都被包括在使用可視光線以外的測定方式內來測定形成 在內層板上的導引記號位置。 基準孔鑽孔機是具有用可視光線以外的測定方式來觀 測導引記號的觀測裝置，及基準孔鑽孔用的鑽孔裝置，其 是由可載置固定住做爲加工物之多層配線板（表背是具有 無垢的導體層）的工作台及對該等進行支承的機架等所構 成。 於一般，其具有固定在機架上被做爲以設有X、γ、 Z的3軸來敘述直角座標的機械座標系統，上述之觀測裝 置、鑽孔裝置及工作台是沿著X軸和Y軸移動在X Y平面 上。然後，觀測裝置和鑽孔裝置是構成爲可相對性移動在 工作台上所載置的多層配線板的任意位置上，進行導引記 號的觀測和基準孔的鑽孔加工。 導引記號普通是形成爲要在基準孔鑽孔機進行觀測時 所適合的形狀，而配線板上之導引記號的配置位置普通是 形成在配線板的4角落等外周部以反映配線板整體之變形 程度。 通常，在單一配線板之圖案6 1 a .......6 1 a的 配置時，是先設定設計座標系統，然後使用該設計座標系 統的座標値來敘述構成圖案之焊墊的位置、孔位等。 也使用同一設計座標系統來敘述導引記號6 6的座標 -11 - (6) (6)592008 ，並且又是由該設計座標系來敘述在結層後的工程上所要 使用之基準孔的座標。 爲工作件之結層後的多層印刷配線板的外周形狀和內 層圖案之間嚴格來說並未保有位置性關係。 因此當工作件被載置在鑽孔機的工作台上時，上述設 計座標系統的原點座標、座標軸方向是不明。 因此，基準孔鑽孔機是把爲了要對被載置在工作台上 多層印刷配線板之內層的導體層中所形成的導引記號進行 觀測後，從被做爲機械座標系統的座標値來敘述之導引記 號的觀測値要推定出多層配線板之內層導體圖案所屬的設 計座標系統的原點座標、X Y軸方向時所需要的各種計算 手段內藏有在其控制裝置中。 當在觀測裝置對被載置在工作台上之多層配線板的內 層中所形成的導引記號進行觀測時，其座標値是顯示在被 固定於機架上的機械座標系統上。最近以每1導體層至少 形成有3個以上（通常爲4個）的導引記號來進行觀測的 狀況較多。 在熱壓工程中因受到熱和壓力，使要形成上述1層之 導體層的1片銅箔內產生不規則的移動，及，甚至在各層 間各導引記號的位置也會產生不規則的移動，因此需要從 這般具有參差不齊（誤差）的觀測値中，推定出各個觀測 値和設計上之導引記號的座標之間誤差爲最小的設計座標 系統的原點座標、X Y軸方向。 要從這般含有誤差之數値來推定某種分佈（函數）時 -12- (7) (7)592008 ，於一般是使用最小二乘法（method of least squares)， 求算其誤差値二次方之和做爲最小的函數。 以具體性的程序而言，首先，是對基準孔鑽孔機的導 引記號進行觀測來測定出機械座標系統的座標値。這之後 想要求算的數値是導引記號被形成時所使用的設計座標系 統的原點位置和座標傾斜度的2個數値，此爲顯現上述之 機械座標系統上的數値。若得知該2數値時，就能夠把設 計座標系統上已知的基準孔座標値換算成在機械座標上所 敘述的座標値（使用習知的座標變換式來換算）。 因此以被設定在基準孔鑽孔機上沿著機械座標系統的 X軸和Y軸進行移動的鑽孔裝置，是容易對換算成機械座 標系統之座標値的基準孔座標位置進行鑽孔。 另，在上述例使用最小二乘法時，由於把設計座標系 統的原點位置固定在某數値進行計算時推論會較容易，因 此該原點就稱爲不動點。 於大多數的狀況，在以熱壓機對多層配線板進行加熱 加壓使其成一體性的熱壓工程中所發生之多層配線板的變 形，大槪是以多層配線板中心附近的1點爲不動點往外側 位移的變形，各點的位移量可認爲是與離該不動點的距離 成比例。 該不動點嚴格來講是會因內層的導體層而有所不同， 雖然熱壓工程的施力狀況也會使其有微妙的變化，但因導 引記號被設在圖案外周部的狀況較多，所以把各導引記號 全部視爲（無具質量m大小的點）質點，來計算導引記號 •13- (8) (8)592008 的重心時，其重心的座標値，幾乎是形成在多層配線板的 中心附近。爲了方便起見把導引記號的重心視爲不動點， 大多也不會有很大的不妥。 因此，於熱壓工程中之多層配線板變形的中心（即不 動點）假定爲是導引記號的重心使用最小二乘法時，就可 求出設計座標系統的原點位置和座標軸的傾斜度。 本申請人是於「日本特願2000 - 1 30764 ( 2 000年4月28日提出申請）」中，提出以下計算方 法做爲專利申請··用變形的中心假定是爲導引記號的重心 方式對複數個導引記號位置進行測定，然後計算出內層圖 案座標的設計座標系統的原點位置和座標軸傾斜度。 如此，多層印刷板的變形中心（不動點）假定是爲導 引記號的重心當進行基準孔鑽孔熱壓以後的加工時，大多 數的狀況是在實用上能夠充分做爲偏差少之基準孔來使用 〇 但是，偶而也會有基準孔位置不妥的狀況。例如有內 層特定的導體層是只有1層時之變形的狀況等這是因爲把 不動點假定爲重心時使1層的圖案會有很大的變形。 尤其是在貫穿多層印刷配線板的全導體層之通孔或穿 孔（接電孔......於零件之安裝時爲不使用之層間的接電 孔）或1層之焊墊的位移較大，使其成爲瑕疵品。 第1 0 ( a ) 、（ b )圖爲表示設置在多層印刷配線 板上之通孔的模式平面圖及板厚方向的剖面投影圖，在全 導體層（於圖中爲6層）的同一座標上設有直徑爲2 R的 -14- (9) (9)592008 圓形焊墊1 0 0，在這些焊墊1 0 0的中心貫穿有直徑爲 2 r的通孔1 0 1 ，在孔內面施以金屬電鍍使各焊墊接電 〇 於各導體層無位移時，如第1 0 ( a )圖所示各焊墊 1 0 0會重疊成1個圓。通孔1 0 1是能夠完整地開在全 部的焊墊1 0 0中心。 若導體層產生位移時焊墊位置會動搖不整，變成例如 其平面圖爲第10 (c)圖所示，側面圖爲第10 (d) 所示的樣子。 當各焊墊的配置爲參差不齊時，假定通孔1 0 1還是 在全部焊墊1 0 0的內部時，各焊墊的位置的偏差容許爲 未滿最大極限（R - r )値。 如第10(c) 、（d)圖所示，當形成在導體層之 第2層上的焊墊1 0 0 ( 2 )的位移爲特別大時，以重心 爲不動點所定位的通孔1 〇 1和焊墊1 〇 〇 ( 2 )的座標 偏差就會超過（R - r )値，此時若把通孔1 〇 1開在中 心102 ，在第10 (c)的圖號103所示位置上，通 孔1 0 1就會超出焊墊1 0 〇 ( 2 )的外緣，使通孔成爲 瑕疵品。 另’若要把如該圖所示般成爲重疊的各焊墊的外形線 分離進行觀測，使用平常的X光是無法透視，但爲方便說 明’平面圖（a ) 、（ C )爲表現各導體層的外周可透視 在多層基板厚度方向的想像圖。 然而’桌10 (e) 、 （f)圖是和第10 (c)、 -15- (10) (10)592008 (d )圖所表示的爲完全相同之焊墊的偏差，但只要把之 前在點1 0 2爲中心的通孔移成以點1 0 2 a爲中心時， 就可使通孔1 0 1 a是在全部之焊墊1 〇 0的內部，使多 層印刷配線板可做爲良品使用。 於現實上雖爲不可能，但假設作業員能夠對如第1 ◦ (c ) 、（ e )圖般需要透視的全部孔同時進行透視時， 就能夠總體性修正鑽孔位置，和採用以不動點是在導引記 號的重心去進行基準孔鑽孔相較之下，使作業員能把握更 佳的基準孔位置。 【發明內容】 〔發明所欲解決之課題〕 如在上述之通孔例所見，可知當在設定將成爲熱壓以 後之加工規範的基準孔時，也有比以不動點是在導引記號 的重心來決定基準孔位置，還更佳的基準孔位置存在。 若能夠得知如此之基準孔位置時，就能把經常要視導 引記號爲座標原點的問題，放在如何提昇多層印刷配線板 的精度上，因此現狀基準孔鑽孔機被要求的一大問題是要 能夠提昇製品的精度。 雖然就機率而言尙屬少，但以鑽孔機所開孔的基準孔 是使用在製造工程的後半，特別是製造工程過半以後的瑕 疵產生其累計投下的成本也大，即使只提昇些許的精度也 能產生較大的經濟效果，因此從使用者方面要求改善的期 望也®。 -16 - (11) (11)592008 〔用以解決課題之手段〕 本發明爲解決上述問題點，提供一種基準孔鑽孔機， 是在具備有：已設定有具垂直之Xm軸和Ym軸的機械座 標系統之被固定在基準孔鑽孔機外殼的機架；對形成在多 層印刷配線板之構成要素的印刷配線板導體層上並且以被 固定在上述印刷配線板之導體層上的設計座標系統來敘述 其座標値的導引記號，至少進行3點觀測之導引記號的觀 測裝置；對上述印刷配線板進行基準孔開孔的鑽孔裝置，· 具與上述X m軸和Y m軸爲平行之輸送方向，對上述印刷 配線板和上述觀測裝置及上述鑽孔裝置之相對位置進行變 換的輸送裝置；及，從上述觀測裝置所觀測的上述導引記 號座標値算出上述設計座標系統的配置，把在上述設計座 標系統上所敘述的基準孔座標値換算成上述機械座標系統 的座標値，對上述輸送裝置進行控制的控制裝置之基準孔 鑽孔機中，上述控制裝置，是具備有：把加權函數設定成 指定値來算出上述印刷配線板之上述設計座標系統的原點 做爲敘述在上述機械座標系統上之座標値的原點座標値算 出手段；算出上述設計座標系統的座標軸和上述機械座標 系統的座標軸所形成之角度的座標軸角度算出手段；及從 變換成上述機械座標値之上述導引記號的座標値和導引記 號的觀測値算出個別誤差，對該個別誤差進行比較，執行 複數上述設計座標之篩選的誤差比較手段。 本發明之基準孔鑽孔機，是在上述導引記號之座標値 觀測後，把全部之導引記號的加權函數當做1使其在算出 -17- (12) (12)592008 第1上述設計座標系統的上述原點座標値，以及上述座標 軸角度之後，改變加權函數的値再算出上述第1設計座標 系統的上述原點座標値，以及上述座標軸角度，根據上述 導引記號之個別誤差的比較來決定採用第1或第2上述設 計座標系統中任一設計座標系統。 此外，本發明之基準孔鑽孔機，是只有把根據第1上 述設計座標系統的上述原點座標値及上述座標軸角度所算 出之具有上述導引記號的個別誤差最大値及最小値的上述 導引記號的上述加權函數當做1 ，來算出第2上述設計座 標系統的上述原點座標値，以及上述座標軸角度，對第1 以及弟2上述導引記號的上述個別誤差進fj比較來決定採 用第1或第2上述設計座標系統中誤差範圍爲小的任一設 計座標系統。 再者，本發明之基準孔鑽孔機又可構成爲：是把全部 之導引記號的加權函數當做1在算出第1上述設計座標系 統的上述原點座標値，及，上述座標軸角度之後，改變加 權函數的値，複數次重覆進行上述設計座標系統的上述原 點座標値，及，上述座標軸角度的算出，在每次算出的同 時，都進行上述導引記號之上述個別誤差的比較，在上述 個別誤差是否會增加時，或者是，在事先所設定的最小誤 差是否會往下降時’或者是在算出次數達到事先所設定的 重覆次數時就結束算出作業，採用其中之一的設計座標系 統。 -18- (13) (13)592008 【實施方式】 〔發明之實施形態〕 爲了方便說明起見’分成以下4個項目來進行說明。 1.鑽孔機的構造說明。 2從導引記號的觀測値，推定出最小二乘法所適用 之內層導體圖案的設計座標系統之說明。 3.計算式之說明。 4 .本發明之實施形態的說明。 1 .鑽孔機的構造說明。 參考第1圖〜第3圖對本發明之實施形態的多點分配 方式基準孔鑽孔機（以下簡稱鑽孔機）進行說明。帛i n 爲本發明之鑽孔機1的外觀透視圖，以透視外殻2來表現 〇 桌2圖、弟3圖均爲表不鑽孔機的投影圖，第2 ( a )圖爲鑽孔機的正面圖，第2(b)圖爲其側面圖，第3 (a) 、（b)圖爲改變鑽孔機1之可動工作台12之位 置時的平面圖，第2圖、第3圖均以透視外殼2來表現。 此外，標示在各圖中的機械座標系統（X m、Y m、 Z m、原點〇m )爲被固定在鑽孔機1的不動部份（例如 外殻1或機架3 )的座標系統，輸送裝置的各種機械部份 的移動方向是平行於該座標軸。以X光攝影機觀測多層配 線板的導引記號後所得的座標値，或基準孔的鑽孔座標於 基本上也是使用該座標系統來算出。 -19- (14) (14)592008 另，第1圖中的白色箭頭1 7表示作業員的立定位置 ，作業員是朝箭頭方向（Ym軸的正値方向）站立，放入 要鑽基準孔的多層印刷配線板，於鑽孔結束後從鑽孔機將 其取出。 於此，是以於首先放入其熱壓完成後的導體層仍舊是 爲一面銅箔的（6層）多層印刷配線板，接著分別觀測其 內層導體的導引記號，然後再進行基準孔鑽孔的作業來做 爲說明。 第4圖爲表示被設在多層印刷配線板6 0之4個角落 上的導引記號6 6周邊部之模式圖。該範例的形狀，大致 上是接近實際的導引記號形狀。 如第4 (a)圖所示，內層導體62a是具有（於圖 中爲6個的）單一配線板的圖案6 1 a 、6 1 a ，其周邊 部是留有鑲邊形的銅箔，在該鑲邊形的銅箔的四角落形成 有角窗7 1 a〜7 1 d，於形成在角窗附近得圓窗7 3的 內部，留有比基準孔孔徑還大做爲基準孔記號7 2的圓形 銅箱。 另，第4 (a)圖是從同一方向透視繪製各內層導體 62a (2) .......62a (5)。 在參考第4 ( c )圖之多層印刷配線板6 0的剖面圖 時，除去爲無垢銅箔之表層的導體62、62後，角窗 7 1 a〜7 1 d是被配置在4層之內層導體6 2 a ..... • · 6 2 a之能夠透視的位置上。內層導體6 2 a ....... 6 2 a的導引記號6 6是被配置成在角窗7 1內透視時不 -20- (15) (15)592008 會重疊。因此，當寬廣地透視被配置在左上部的角窗7 1 3時，各導體層的角窗71 a附近就會和把第4 (d)〜 (g )圖所示者合成後爲相同結果，如第4 ( b )圖所示 4個導引記號66 (2)〜66 (5)會出現在角窗71 ' a內。 此外，於圓窗7 3的內部中，基準孔記號7 2是重疊 且出現有環狀間隙。因此在多層印刷配線板6 0的四角落 出現有相同記號。 φ 於上述範例中，鑽孔機是對多層配線板角落的每1處 各進行導引記號4個觀測，記憶其座標値，在4個角落合 計爲觀測1 6個導引記號。於全部導引記號觀測結束後對 換算成機械座標系統的基準孔位置進行基準孔鑽孔。 要在6層的多層印刷配線板上進行基準孔鑽孔，是需 要有該程度的觀測次數，但在以下的動作說明中是把觀測 省略成1處1次來進行說明。此外，導引記號和基準孔同 時爲4個，於要應對的導引記號附近假定有基準孔。 ® 如第1圖〜第3圖所示，在鑽孔機1之外殻2的內部 ，固定著機架3。左右1對的X方向移動架1 0、1 0， 大致是形成爲槽鋼狀，左右形狀是爲鏡射關係。該X方向 ’ 移動架1 0、1 〇是由配置在機架3之上端的直線導件 . 1 〇 a、1 〇 a支承著。藉由滾珠螺桿1 〇b和與其卡合 之被安裝在X方向移動架1 〇下面的滾珠螺帽（未圖示） ’根據要鑽基準孔之配線板的大小，於事先，使其待機在 移動成平行於X m軸時可觀測導引記號的位置上。 -21 - (16) (16)592008 另，爲要對X方向移動架1 0、1 0進行個別驅動， 每個X方向移動架1 0各配置有滾珠螺桿1 0 b。 在X方向移動架1 0、1 〇的上部固定著X光產生裝 置4、4，在下部安裝著直線導件1 1 a、1 1 a。然後 ，Y方向移動架1 1 、1 1 ，是以該直線導件1 1 a支承 著。藉由滾珠螺桿1 0 b和與其卡合之被安裝在X方向移 動架1 0下面的滾珠螺帽（未圖示），使Y方向移動架 1 1、1 1可移動成平行於Y m軸。 φ Y方向移動架1 1 、1 1是形成爲槽鋼狀，在上部配 置有X光防護管5，如第2圖所示，設有與X光防護管5 排列的夾緊片9和使夾緊片9上下移動的氣缸9 a。在下 部固定著心軸7和X光攝影機6。 與Y m軸配置成平行藉由固定在外殼的中央部份的直 線導件1 2 a和滾珠螺桿1 2 b使其支承、驅動，可搭載 多層印刷配線板的可動工作台1 2是活動成平行於Y m軸 • 可動工作台1 2是在1 2A的位置上，載置爲工作件 之要鑽基準孔的多層印刷配線板，接著沿著γ m軸移動來 測定導引記號，然後將其拉入在基準孔鑽孔位置上。 · 另，對滾珠螺桿1 〇 b、1 1 b、1 2 b進行驅動， . 來控制X方向移動架1 〇、1 〇和γ方向移動架1 1 、 1 1 ’及可動工作台1 2之移動的控制裝置並未圖示於圖 中〇 於此’做爲鑽孔機的主要構成要素，分別是以X光產 -22- (17) (17)592008 生裝置4和X光防護管5及X光攝影機6形成導引記號的 觀測裝置，以心軸7和夾緊片9來成鑽孔裝置，以X方向 移動架1 0和Y方向移動架1 1及可動工作台1 2以及對 該等進行支承、驅動的直線導件1 0 a、1 1 a、1 2 a 及滾珠螺桿l〇b、1 lb、12b等形成驅動裝置。 此外，並未圖示的控制裝置，是依照一連貫的鑽孔作 業次序，進行上述各種裝置的控制。又從觀測裝置所觀測 到的導引記號X光圖算出座標値，也扮演著從該座標値和 事先輸入的基準孔設計座標値中計算基準孔鑽孔位置的角 色。 可動工作台1 2通常是以金屬製形成爲平坦的板狀， 左右開孔有導引記號透視用、基準孔鑽孔用的開口（ 1、 2 孔用）1 3 、1 3 及開口（ 3、4 孔用）1 3 a 、1 3 a ° 多層印刷配線板6 0，是依照其外形大小，使配線板 的前端部份是爲整齊地放入在可動工作台1 2上。基準孔 之1 、2是使用開口 1 3 ，基準孔之3、4是使用開口 1 3 a進行鑽孔。 2.作業次序 使用上述鑽孔機，以4孔基準孔鑽孔爲範例對其作業 次序的說明如下。 第4圖爲表示被設在多層印刷配線板6 0之4個角落 上的導引記號6 6周邊部之模式圖，該範例的形狀，大致 -23- (18) (18)592008 上是接近實際的導引記號形狀。 如第4 (a)圖所示，內層導體62a是具有6個的 單一配線板的圖案6 1 a、··· 6 1 a ，其周邊部是留 有鑲邊形的銅箔，在該鑲邊形的銅箔的四角落形成有角窗 7 1 a〜7 1 d ，於形成在角窗附近得圓窗73的內部， 留有比基準孔孔徑還大做爲基準孔記號7 2的圓形銅箔。 另，第4 (a)圖是從同一方向透視繪製各內層導體 6 2a ( 2 ) .......62a(5)。 · 如第4 ( c )圖所示，在參考之多層印刷配線板6 0 的剖面圖時，除去爲無垢銅箔之表層的導體6 2、6 2後 ，角窗7 1 a〜7 1 d是被配置在4層之內層導體6 2 a .......6 2 a的大置相同位置上。形成在內層導體62 a .......6 2 a上的導引記號6 6是被配置成在角窗 7 1內透視時不會重疊。當把圖示著被配置在左上部之角 窗71 a周邊的第4 (d)〜（g)圖合成時，如第4 ( b)圖所不4個導引記號66 (2)〜66 (5)會出現 春 在角窗7 1 a內。 此外，於圓窗7 3的內部中，基準孔記號7 2是重疊 且出現有環狀間隙。因此分別在多層印刷配線板6 0的四 | 角落，將出現有角窗71a〜71d及基準孔記號72a . 〜7 2 d做爲和上述相同之記號。 另’基準孔記號雖然是可不必形成在導體層上，但大 多數是將其設置做爲基準孔加工之確認及之後的作業記號 -24- (19) (19)592008 鑽孔機是對每1處各進行導引記號4個觀測，記憶其 座標値，合計爲觀測1 6個導引記號。於全部導引記號觀 . 測結束後對換算成機械座標系統的基準孔位置進行基準孔 鑽孔。 ^ 要在6層的多層印刷配線板上進行基準孔鑽孔是需要 有該程度的觀測次數，但在以下的動作說明中是把觀測省 略成1處1次來進行說明。此外，導引記號和基準孔同時 爲4個，導引記號和基準孔是假定成鄰近者。 鲁 在鑽孔機側攝影機位置是被編排成在觀測導引記號時 ，導引記號是1個1個的進入攝影機的視野。 因此，做爲槪略的次序，是需要以下3個步驟：（1 )對導引記號6 2進行觀測；（2 )從導引記號6 2的觀 測値推定出設計座標系統得座標原點和座標軸傾斜度；（ 3 )把在設計座標系統上所敘述的基準孔位置換算成機械 座標系統。 另，於以下的次序說明中爲避免繁雜，因此於1個角 · 窗中有1個導引記號，在各導引記號的附近設有基準孔， 配置在同一角落的導引記號和基準孔是爲鄰近者。 熱壓結束後的多層印刷配線板的內層導體6 2 a的圖 ’ 案，和配線板周邊部的形狀嚴格上來講是無座標性關係， . 但大多數的狀況是能以周邊部的外形導引使導引記號被收 入在攝影機的視野內。因此，於一般上，放入在鑽孔機中 的多層印刷配線板，是對其外形進行引導使導引記號置於 攝影機的視野內。 -25- (20) (20)592008 首先，從要鑽基準孔之工作件的多層印刷配線板（例 如第4圖或第8圖、第9圖所示）6 0的外形尺寸和（設 計上的）基準孔座標來決定沿著X方向移動架1 0、1 〇 之Xm軸上的位置，於事先，使X方向移動架10、10 是移動至該位置上待機著。 可動工作台1 2是在（第1圖的）1 2 A的位置，作 業員是把多層印刷配線板6 0載置在可動工作台1 2上之 指定的位置。配線板6 0是暫時固定在可動工作台1 2上 。從作業員來看，可動工作台1 2是移動到使配線板6〇 前端側的2個導引記號（6 6、6 6 )位於X光產生裝置 4所內藏之X光產生管4 a的下面。 以X光透視前端側的導引記號（6 6、6 6 )在X光 攝影機6、6進行觀測，測定其座標値。座標値是被記憶 在未圖示之控制裝置的記憶體中。 可動工作台1 2往Y m方向移動的距離，僅僅是使靠 近跟前的導引記號（66、66)移至X光產生管4a下 的距離。接著照射X光，在X光攝影機6、6觀測2個導 引記號，記憶其座標値。 於此’採用將於下述計算方法，從導引記號4點的座 標求出設計座標的原點座標和座標軸傾斜度，從該値計算 出基準孔4個的（機械座標系統的）座標値，首先，心軸 7、7是移動至靠近跟前的2個基準孔的座標位置上，進 行基準孔鑽孔。 接著，可動工作台1 2會移動使前端側的導引記號（ -26- (21) (21)592008 6 6、6 6 )回到X光透視的位置附近，心軸7、7是移 動至前端側的基準孔的座標位置上，進行2個基準孔鑽孔 〇 然後，可動工作台12會移動至放入位置12a，當 作業員取出鑽孔完畢的配線板6 0時就結束基準孔加工工 程。 另，如第4圖所示，即使是在導引記號的個數於實際 上是各導引記號分別爲4個時，測定次數也只不過是變成 4倍，其次序的執行是和上述內容幾乎沒有改變。 於此所鑽開的基準孔，是使用在外層之2面導體層的 圖案形成或通孔、穿孔等之圖案內部的開孔等。該等孔都 是在結層用工具板上所設的定位用結層釘上插穿多層基板 的基準孔來決定該等位置。 於實際上，爲了要維持加工精度，大多數都會1次使 用4支以上的結層釘，惟恐若插穿於結層用工具板的結層 釘好幾次將造成孔徑擴大而不再度使用基準孔，因此也會 有於每1工程使用不同基準孔的狀況。該狀況時，是在後 工程對要使用之複數的結層用工具板的結層釘數或其座標 所配合的複數組的基準孔進行鑽孔。， 這些複數組之基準孔的座標，由於也都是被敘述在同 一的設計座標系統上，因此只要決定該設計座標系統的原 點座標和座標軸傾斜度，接著只是增加各個的座標計算和 鑽孔工次而已，於原理上是可重覆相同事宜。 -27- (22) (22) 592008 3 .計算式的說明 於一般’由於基準孔是在後工程插穿於結層用工具板 的結層釘使用在工作件的定位上，因此1組的基準孔是具 有與要使用之結層用工具板的結層釘的座標位置爲相同的 配置，各基準孔相互的距離是要在指定値的範圍內，與導 體圖案等爲一定之誤差內開孔。由於這是藉由把誤差分配 在各基準孔而達成要求，因此通常稱爲分配式。要觀測之 導引記號的數量爲3個以上時，有時也稱爲多點分配式。 如以上所述，形成在導體層上的導引記號是和導體圖 案同時形成。在多層印刷配線板的內層上所形成的導體圖 案和基準孔，及導引記號等之全部導體層所屬要素的座標 ，是全部被敘述在1個設計座標系統上，其關係是已知。 因此，需要以下程序：（1 )觀測鑽孔機工作台上之 多層印刷配線板上所形成的導引記號，讀取固定在鑽孔機 上的機械座標系統所表示之各導引記號的座標値；（2 ) 從該値求出設計座標系統的原點座標和座標軸傾斜度；（ 3 )若要以搭載在鑽孔機中的鑽孔裝置鑽孔，需要把在設 計座標系統上所敘述之基準孔的座標値變換成機械座標系 統上所表示的座標値之程序。 該一連貫的程序中，（1 )單純是導引記號的觀測， (3 )是只要採用眾所周知的三角函數所使用的座標變換 式就可容易計算。 (2 )之從所觀測的各導引記號之機械座標系統的座 標値求出設計座標系統的主要的原點座標和座標軸傾斜度 -28- (23) (23)592008 之程序爲多點分配式之計算式的主要部。 由於該計算式已在上述「日本特願」的說明書中有詳 細說明，因此只對其要點進行敘述如下。 參考第5圖，對導體圖案設計時之設計座標系統和鑽 孔機上所設定的機械座標系統之間的關係進行說明。 第5 ( a )圖爲設計座標系統之說明用平面模式圖， 設計座標系統是以0 D爲座標原點，具有互爲垂直之相當 於X軸的U D軸、相當於Y軸的V D軸。放置著多層印刷 配線板6 0，形成在其內層導體層上（於圖中爲6個）的 單一配線板的圖案6 1 a A.......6 1 a A及導引記號 P D 1 .......PD4以及基準孔HI .......H4等的 位置，是以設計座標系統的座標値來敘述。使用該設計座 標系統，同樣也敘述著單一配線板之圖案6 1 a A內部的 形狀或孔位置爲該系統的座標値。 如第5 ( b )圖所示，是著眼於形成在多層印刷配線 板6 0上的導引記號P D 1 .......P D 4 ，把導引記號 分別視爲質點時之重心做爲旋轉中心，設定（新）設計座 標系統：以◦爲座標原點，具有平行於U D軸、V D軸之 U軸、V軸。 第5 ( C )圖，是把在X光攝影機所個別觀測之導引 記號位置P 1〜P 4 (根據其之觀測座標値）繪製在鑽孔 機上所設定之機械座標系統（原點◦，以Xm、Ym爲互 成垂直之座標軸）上的模式圖。然後還是把導引記號P 1 〜P 4視爲質點求出其重心〇g，設定（新）機械座標系 -29 - (24) (24)592008 統：以〇g爲座標原點，具有平行於X m軸、Y m軸之χ 軸、Y軸。 另’設計座標系統、機械座標系統雖然都是把原點移 至導引記號的重心位置上，而形成有（新）設計座標系統 、（新）機械座標系統，但因兩者之新舊的座標軸均互爲 平行’所以只要單純對新原點之座標値進行加減計算就能 執行座標變換。 於此，要在第5 (c)圖的上面，描繪出第5 (b) 圖所示之多層印刷配線板6 0上之導引記號P D 1〜P D 4。首先，想成多層印刷配線板6 0是被放置成（新）設 計座標系統的原點〇是和（新）機械座標系統的原點〇g 爲一致，U軸和X軸（X m軸均是）爲平行，V軸和Y軸 (Y m軸均是）爲平行。從之前的（新）設計座標系統和 (新）機械座標系統設定的過程得知，X m軸和U D軸、 Y m軸和V D軸均互成平行。 於第5 ( d )圖中以1點虛線所示之長方形S各頂點 的黑點是相當於（設計上的）導引記號P D 1〜P D 4。 (設計上的）導引記號P D 1〜P D 4和（所觀測的）導 引記號P 1〜P 4雖是被放在一張紙的上面，但於該階段 只有兩者的重心是爲一致。 把（設計上的）導引記號P D 1與（所觀測的）導引 記號P 1之間的距離L 1乘以二次方，把P D 2與P 2的 距離L 2乘以二次方，同樣地把距離L 3、L 4各自乘以 二次方的合計是等於最小之α ° (例如U軸和X軸所形成 -30- (25) (25)592008 的角度）是可透過使用最小二乘法來計算求其結果。 在第5 ( d )圖上，把上述多層印刷配線板6 0繞著 (新）設計座標系統的原點◦旋轉成α ° ，例如使以黑點 表示的（設計上的）導引記號（P D 1 )旋轉至以白點表 示的P D 1。此時（所觀測的）導引記號Ρ 1與（設計上 的）導引記號PD1的距離是以L1表示，PD2與Ρ2 的距離以L 2表示，同樣地其餘距離以L 3、L 4表示。 在第5 ( d )圖上繪出L 1〜L 4之二次方的和等於最小 的値之位置上。 在上述之說明書中已詳細說明有求出α的式子和其之 算出法，在α的絕對値爲較小時，當把（新）機械座標系 統所表示之導引記號P i的座標爲X i 、y i時，把（新 )設計座標系統和所表示之導引記號P D i的座標爲U i 、V i時，就可用下述〔方程式1〕求出^的正切値。 Σ (y V,) sin a i _ l____ tan a =---- ' ~ C〇S a Σ (χ,υ, + y i -1 ·· ··[計算式1 ] 反之，雖是根據上述（新）設計座標系統的原點座標 (重心位置），及座標軸的角度a ° ’把（設計上的）導 引記號P D 1〜P D 4記入在（新）機械座標系統及機械 座標系統上，但也可說是以第5 ( d )圖來表示著。 如第5 ( d )圖所示，（設計上的）導引記號P D 1 〜P D 4與（所觀測的）導引記號Ρ 1〜P 4的距離L 1 -31 - (26) (26)592008 〜L 4之二次方的和爲最小値之事宜，也可於實際上的驗 算得知 。 如此般，使用最小二乘法，就可求出（新）機械座標 系統所表示之（新）設計座標系統的原點座標（重心位置 )’及座標軸的角度α ° 。藉由該（新）設計座標系統的 原點座標（重心位置），及座標軸的角度使在設計座標系 統上所敘述的座標能夠全部換算成機械座標系統上的座標 〇 因此，使鑽孔機能夠把在設計座標系統上所敘述的基 準孔變換成機械座標系統上的座標値進行鑽孔。 本發明之實施形態的說明。 最小二乘法於一般上是被說明成如下所述。即，測定 對象而獲得η個的觀測値〔f ( X i ) ...... i = 1〜^ ......(式1 )〕時，適當選出對η個的觀測値進行加權 的加權函數〔ω ( X ) : weighting function〕，並且假定 成較簡單之函數形g n ( X )求出與加權函數之差値二次 方的積値，再求出其之和Ω ( i = 1〜η)爲最小時之 gn(x)。因此，Ω是以下述式（2)來表示。 { Ω = Σ ω ( X i )x [ f ( X i ) — g n ( x i )]2......(2)} 於此，Σ是表示把1〜n代入ω (xi)x[f(xi)-gn(xi)]2 的i所得之合計。 於本例中，是觀測η個的導引記號來獲得其之（所觀 測的）η個導引記號座標値（X i 、y i )。從該觀側値 中欲得知的數値爲機械座標系統上所表示之設計座標系統 -32- (27) (27)592008 的原點座標和座標軸的傾斜度。 最小二乘法使用的次序，是假定適當的加權函數以決 定設計座標系統的原點座標，來求出在該原點位置上誤差 爲最小之座標軸的角度。 另，雖是重覆敘述，已知於多層配線板的導電層中， 導體圖案或其所附帶的孔種類、導引記號、基準孔等諸要 素的座標値等是全部敘述在該設計座標系統上。因此，得 知設計座標系統的原點座標和座標軸的傾斜度，就表示能 使導電層所屬之要素的全部配置（座標）換算成機械座標 系統。 加權函數對於（所觀測的）η個導引記號全都當做1 ，把導引記號的重心做爲設計座標系統的不動點來求出設 計座標系統之座標軸的傾斜度之求出方法已說明在上述「 曰本特願」的說明書中，於該說明書中示意著亦可採用其 他的不動點。 雖然先前所說明之加權函數的定義嚴格上來講也有不 合的點，但對於（所觀測的）η個導引記號，爲要決定其 設計座標系統的原點座標値，把假定適當加權函數ω ( X )之事宜，於以下稱爲加權函數ω ( χ )的使用。 以上述的稱謂方式，是把所觀測的導引記號Ρ 1〜 Ρ 4視爲質點，使其重心成爲不動點則表示加權函數全部 是1 ，即，非要(X) =1不可。 例如··當導引記號總數爲4點時，以X i爲其X座標 ’要求出重心之X座標G ( X)時，以m爲其質量’其算 -33- (28) (28)592008 出式爲 [G(x) = (mxxH-mxx2 + mxx3+mxx4/4m = I(xi)/4......(3)] ’相當於各導引記號的加權函數全部是當做1。 當改變加權函數的選擇方式時，就可對使用最小二乘 法的結果做大改變。 第6圖爲表示設置在1片多層印刷配線板之4個角落 上的（設計上的）導引記號P D 1〜P D 4及（所觀測的 )導引記號P 1〜P 4的模式圖，（所觀測的）導引記號 P 1〜P 4如第4圖所示是構成爲各角落4個。 多層印刷配線板之各角落的4個導引記號是總括各設 計座標値以畫有斜線的1個圓來表示，做爲（設計上的） 導引記號P D 1〜P D 4。相當於（設計上的）導引記號 P D 1 ，（所觀測的）導引記號P 1的4個是以對其中心 只移動誤差量之圓的外周線來表示。其他的3個角落也以 同樣方式表示，（設計上的）導引記號P D 1〜P D 4的 中心是位於長方形S的各頂點。 該等之關係是和第1 0圖之通孔與焊墊之間的關係爲 相同·，於槪念上當表示（設計上的）導引記號P D 1〜 P D 4的斜線圓超出表示（所觀測的）導引記號P 1〜 P 4的圓之外周線時超過誤差的界限就表示可認定爲瑕疵 品。 第6 (a)圖是表示用（設計上的）導引記號PD1 〜P D 4和（所觀測的）導引記號P 1〜P 4的各自重心 是爲一致位置（做爲不動點）的習知方法決定導引記號 -34- (29) (29)592008 P D 1〜P D 4位置時所得到的結果，（設計上的）導引 記號P D 4的圓會和第2層之（所觀測的）導引記號P 4 (2 )交差使其成爲瑕疵品。 針對於此，本發明之實施形態的要點，是在於以重心 點爲不動點使用最小二乘法後，改變加權函數的選擇方式 來變更不動點位置，再度使用最小二乘法，將兩者之結果 進行比較而採用誤差較少者。 不動點位置變更後之結果如第6 ( b )圖所示，設計 座標的座標原點將從0 g移至0 2，當以新的原點座標使 用上述〔計算式1〕時，座標軸角度也將從α °變成 α 2 ° 。其結果，使（設計上的）導引記號P D 1〜P D 4是仍舊保持著長方形S的關係進行移動及旋轉，從第6 (a )圖得位置移到第6 ( b )圖的位置上。 其結果，使（設計上的）導引記號P D 4和（所觀測 的）導引記號P 4 ( 2 )不交差，判斷出基準孔是被開孔 在被認定爲良品的位置上。 做爲本發明之實施形態的1範例，設計座標系統推定 用之計算的執行、判斷是以基準孔鑽孔機的控制裝置來進 行。 即，在對上述基準孔鑽孔機的觀測裝置、鑽孔裝置及 輸送裝置執行控制之控制裝置內，內藏有原點座標値算出 手段、座標軸角度算出手段及誤差比較手段的計算程式， 根據來自控制裝置內之C P u的指令，對各計算進行分擔 處理。 -35- (30) (30)592008 參考第7圖之方塊圖對該3手段要處理之計算內容和 實際的處理次序流程的應對進行說明。第7 ( a )圖爲表 示控制裝置內之局部性構成，第7 ( b )圖爲表示沿著箭 頭進行處理程序的流程。第7 ( a )圖之各手段，是把記 載在其下方之第7 ( b )圖的各處理隨著箭頭之指示進行 處理。 計算程序是進行如下：以導引記號的觀測値爲根據， (1 )藉由原點座標値算出手段計算出座標原點（重心位 · 置），以重心爲不動點使用最小二乘法，計算出（新）機 械座標系統所表示之（新）設計座標系統的座標原點，以 座標軸角度算出手段，進行〔計算式1〕的計算求出座標 軸的角度（α ° )，而得到第第1座標系統。然後把（設 計上的）導引記號P D i的座標値換算成機械座標系統的 座標値。以前是使用該設計座標系統求出基準孔之機械座 標系統的座標値，立即將其形成爲基準孔。 延續第1座標系統的計算，進行以下的計算。 · (2 )求出與（設計上的）導引記號P D i應對之各 (所觀測的）導引記號P i的誤差（個別誤差）。 個別誤差是以座標軸爲區別求出在機械座標系統上。 即，把各（設計上的）導引記號P D i的座標値爲 (P D i X、P D i y )把各（所觀測的）導引記號 P i ( j )的座標値爲〔P i ( j ) X、P i ( j ) y〕 ，把以座標軸爲區別的個別誤差爲△ P i ( j ) x、 △ Pi ( j ) y時，x軸、y軸的計算如下。 -36- (31) (31)592008 X 軸之計算爲〔APi(j)x = Pi(j)x-PDix......(4)〕 y 軸之計算爲〔ΔΡΚΌγζΡΚηγ-ΡΟίγ......(5)〕 於此，i是相當於（設計上的）導引記號P D i (基 板外周各角落之編號）’ j是相當於導體層的編號〔例如 參考第4 (c)圖〕。 (3 )選擇具有最大最小之誤差的（所觀測的）導引 記號P i ( j )。 做爲一般的程序’是選擇其誤差是以昇順序或降順序 排列之誤差系列兩端的導引記號。即，把以座標軸爲區別 的個別誤差包括符號排列成從小往大的順序，或者排列成 從大往小的順序，選擇位於其兩端的導引記號P i ( j ) 。例如：對於X軸是選擇P 2 ( 3 )和P 4 ( 1 )，對於 γ軸是選擇P1 (4)和P3 (2)。 另，當出現有相同誤差之複數個的導引記號時，只要 選擇其中之一即可。 (4 )求出各導引記號之機械座標系統所表示之座標 値的平均計算値，把該値做爲第二的座標原點。例如爲上 例時， 〔[Ρ2(3)χ + Ρ4(1)χ]/2 =新不動點的 X 座標......（6)〕 〔[Ρ 1 (4)χ + Ρ3(2)χ]/2 =新不動點的 y 座標......（7)〕 上述的計算，是相當於以通常的重心計算式把誤差系 列兩端之導引記號的加權函數爲1 ，其他爲〇時之狀況。 把用上述座標値的平均計算値分別做爲x、y座標的 點當做第2原點0 g 2製成第2機械座標系統。第2機械 -37- (32) (32)592008 座標系統的X 2軸、Y 2軸是與原來的X m軸、Y m軸平 行，這可應對於第6 ( b )圖。只是，因與U V軸非常接 近故未記入X 2軸、Y 2軸。（所觀測的）導引記號p 土 (j )的座標値是藉由座標原點從◦ g平行移往〇 g 2， 而得以更新成新的値。 把（新）設s十座標系統的原點座標移到0 g 2。因在 (新）設計座標系統的內部中於（設計上的）導引記號 P D 1〜P D 4的重心已有座標原點，所以也就是說（新 )設計座標系統整體（也包括在其座標上的點）是從〇g 平行移往〇g 2 。 於座標軸角度算出手段中，是以第2原點〇g 2爲第 2不動點，使用最小二乘法求出設計座標系統之座標軸和 機械座標系統之座標軸所形成的角度（α 2。），其計算 式採用〔計算式1〕即可。 藉由第2設計座標系統的原點座標〇2 (等於第2原 點〇g 2 )和座標軸的角度（α 2 ° )，使（設計上的） 導引記號P D i得以敘述在機械座標系統上。 被表示在該等第2座標系統上的（設計上的）導引記 號P D i的座標値是被輸送到誤差比較手段；（5 )計算 從第2設計座標系統換算成機械座標系統之（設計上的） 導引記號P D i和（所觀測的）導引記號p丨（j )之座 標値以座標軸爲區別的個別誤差。計算方法是和上述之第 1座標系統的個別誤差的計算相同。 (6 )比較第2座標系統的個別誤差（5 )和習知手 -38- (33) (33)592008 法之第1座標系統的個別誤差（2 )後進行判定。於一般 ，是採用誤差範圍（最大誤差-最小誤差）較少者即可。 (7 )採用有利之設計座標系統的配置。 另，若有某種特定要求時，例如：當某導體層爲特別 重要時等，是選擇其層之誤差爲較少者等，可適宜決定是 否採用的判定條件。 如此般，是在對基準孔鑽孔機的觀測裝置、鑽孔裝置 及輸送裝置執行控制的控制裝置內內藏有計算程式，以原 點座標値算出手段、座標軸角度算出手段及誤差比較手段 推定出設計座標系統的位置，藉由設定在鑽孔機上的機械 座標系統，來決定多點分配式之基準孔的座標値。 特別是，因新增設有：算出導引記號之個別誤差，選 出具特定誤差的導引記號，判定何者爲有利之座標系統的 誤差比較手段，故得以從複數的設計座標系統中求出最適 當者；藉由所選之最適當的設計座標系統，來獲得精度更 高的基準孔位置之事宜，對要降低工作件瑕疵率而言可以 算是一大貢獻。 通常是沿著上述程序，把第1次之以（所觀測的）導 引記號P i ( j )的重心爲不動點使用最小二乘法所得的 結果，和第2次之以具有最大最小之誤差的導引記號座標 値的平均計算値所求得之第2不動點使用在最小二乘法時 所得的結果進行比較就可得到良好的結果。 因應需求，也可複數次重覆使用由平均計算値所求得 之第2次數以後的最小二乘法。於該狀況時，是採用當初 -39- (34) (34)592008 次所得的誤差比初次所指定的最大誤差還小時就停止重覆 的方法，或採用當2次數以後的計算已進行了指定次數時 就停止等之方法，也可想成有2個以上的座標系統然後選 擇最棒的座標系統。 第2次數以後的計算中即使是改變所使用之加權函數 的加權方式也會有不同的結果。例如：在事先對全部的導 體層分配1 ，當把有重要之要素的導體層的加權函數爲2 時’是可使形成在該導體層的圖案比形成在其他之導體層 的圖案來得其誤差要少。 〔發明效果〕 如以上說明，本發明如申請專利範圍第2項所述，是 在用導引記號的重心爲不動點之習知手法來決定多層印刷 配線板的第1配置之後，再度變更不動點位置以算出第2 多層基板的配置，因是採用其中之誤差較少者所以增加成 爲良品的機會因此對減少瑕疵率有貢獻。如先前所述，熱 壓以後的工程因其成形度是持續進行，所以即使是降低些 許的瑕疵率也可成爲大廢棄原價的削減，對生產成本的幫 助是較大。 此外，如申請專利範圍第3項所述，當多層配線板之 導體層之中有1層的偏差爲較大時等選擇第1之誤差較大 的導引記號來設定第2不動點時，對多層配線基板之導體 層之中有1層的偏差爲較大時等能特別發揮效果。由於有 1層的偏差爲較大之狀況是表示瑕疵整體之占有率相當大 -40- (35) (35)592008 ，因此就能防止其發生以降低瑕疵率。如先前所述，熱壓 以後的工程因其成形度是持續進行，所以即使是降低些許 . 的瑕疵率也可成爲大廢棄原價的削減，對生產成本的幫助 是很大。 又，於本發明中如申請專利範圍第4項所述，藉由重 覆求出多層印刷配線板的配置，能夠獲得達成滿意的結果 。特別是對層數多的高多層配線板之瑕疵率的降低有較大 的效果。 _ 再者，如申請專利範圍第1項所述，若要實施本發明 是幾乎不用變更市場上所供應的分配式基準孔鑽孔機的機 械構造部份，以只改善內藏在控制裝置中的程式就能應對 。程式的變更即使是在交貨後也是比較容易進行，對於已 市販的基準孔鑽孔機的性能提昇也具有能夠處理的效果。 【圖式簡單說明】 第1圖爲表示本發明之實施形態之基準孔鑽孔機的構 · 成模式透視圖。 第2圖爲表示本發明之實施形態之基準孔鑽孔機的構 造模式正面圖及側面圖。 ^ 第3圖爲表示本發明之實施形態之基準孔鑽孔機的構 _ 造模式平面圖。 第4圖爲表示實際使用之多層配線板上所設置之導引 記號的配置和形成在導引記號附近的基準孔（記號）模式 圖。 -41 - (36) (36)592008 第5圖爲多數孔基準之分配開孔原理說明用的模式圖 〇 第6圖爲表示多層配線板之導體層上所設置之導引記 號與所觀測的導引記號之間的關係模式圖。 第7圖爲基準孔鑽孔機之控制裝置的局部構成及比對 設計座標系統位置之算出程序的方塊圖。 第8圖爲多層印刷配線板之構成說明用透視圖和內層 之導體層的平面圖及結層用工具板的側面圖。 ® 第9圖爲多層印刷配線板之構成說明用之板厚方向的 剖面圖。 第1 0圖爲表示多層印刷配線板之圖案內所設廠之通 孔的平面圖及板厚方向的剖面圖。 〔圖號說明〕 1 :鑽孔機 2 :外殼 @ 3 :機架 4 : X光產生裝置 4 a : X光產生管 ’ 5 : X光防護管 · 5 a ·孔 6 : X光攝影機 7 :心軸 7 a :夾頭 -42- (37) (37)592008 7 b :孔鑽 7 c :氣缸 8 :心軸架 9 :夾緊片 9 a :氣缸 9b:夾緊片支撐具 9 c :夾緊螺桿 10 :X方向移動架 1 1 : Y方向移動架 1 2 :可動工作台 1 0 a、1 1 a、1 2 a :直線導件（l Μ導件） l〇b、lib、12b:滾珠螺桿 1 6 :鑽孔位置（1、2孔） 1 6 a :鑽孔位置（3、4孔） 1 7 :作業員位置（白色箭頭） 5 0 :機械座標系統（Xm、Ym、Zm、機械原點〇m ) PI、P2、P3、P4 :(所觀測的）導引記號（敘述在機械 座標系統上） PD1、PD2、PD3、PD4 :(設計上的）導引記號（敘述在 設計座標系統上） H1、H2、H3、H4:基準孔 α :旋轉角度（U軸和X軸所形成的角度） 6 0 :多層印刷配線板 6 1 ' 6 1 A :兩面配線板 -43- (38)592008 6 1 a 6 1 a • 單 — 配 線 板 的 圖 案 6 2 : 導 體 6 3 : ( 絕 緣 ) 基 板 6 4 : 聚 酯 膠 片 6 4 a 、 6 4 a A ； 聚 酯 膠 片 ( 附 有 基 準 孔） 6 5 6 5 A % 6 5 B ： ( 結 層 用 ) 基 準 孔 6 6 • 導 引 記 號 6 7 : 基 準 孔 6 8 6 8 A ； 結 層 用 工 具 板 6 8 a 6 8 a A 6 8 a B ·· 定 位 用 結 層釘 7 1 ( b C d e ) 二 方: 形 窗 7 2 ( b % C d e ) ； 圓: 形 記 號 7 3 ； 圓 形 窗 1 0 0 1 0 2 ( 2 ) ·· 焊 墊 1 0 1 1 0 1 a : 通 孔 1 0 2 1 0 2 a ； 通 孔 中 心592008 玖 发明 Description of the invention [Technical field to which the invention belongs] The present invention relates to a drilling machine that observes a guide mark of a multilayer printed wiring board and drills a corresponding reference hole, and particularly relates to Drilling machines for forming useful reference holes when the inner layer is deviated [prior art] _ Recently, with the miniaturization of surface mount electronic parts such as IC chips, resistors, capacitors, etc. Since printed wiring boards are also required to have high density, most printed wiring boards are multilayered. Even for the people's livelihood, multilayer printed wiring boards with 4 or 6 layers are used. For industrial use, there is a tendency to use more multilayer printed wiring boards with more layers. The multilayer printed wiring board is composed of a conductor layer exposed on the outer two layers of the front and back, and several inner conductor layers that are not exposed. An insulating substrate is inserted between the conductor layers, and the conductor layer is formed by the substrate. Combine construction. β is used as a conductive layer of a multilayer printed wiring board, for example, a copper foil having a thickness of about 18 is used. As the substrate material, thermosetting glass and epoxy resins are used as the mainstream. For high-layer wiring boards, heat-resistant resins such as glass, polyimide resin, glass, and B T resin are also used. Since the multilayer printed wiring board is a wiring board with more than 6 layers, it is simply the number of conductors in the inner layer. Therefore, the manufacturing method of the multilayer printed wiring board is as follows. Refer to Figures 8 and 9 here. Multilayer Printed Wiring-7- (2) (2) 592008 A method for manufacturing a board will be described briefly. Fig. 8 (a) is a perspective view showing a configuration pattern of a 6-layer printed wiring board, and Fig. 8 (b) is a plan view showing a pattern of a printed pattern formed on the conductor portion of the inner-side two-sided wiring board. c) The figure is a side view showing a bonding tool plate used in the following bonding. Figure 9 is a cross-sectional view of a 6-layer wiring board. (A) shows the state before the hot-pressing process, and (b) shows the bonding conductor layer becomes a multilayer printed wiring board after the hot-pressed engineering substrate is allowed to heat set. status. ® As shown in Fig. 8 (a), the 6-layer multilayer printed wiring board 60 is a two-sided printed wiring board 6 1 and 6 1 which are the conductor layers 6 2 and 6 2 exposed on the 2 layers. Polyester films 6 4, 6 4 a, and 6 4 are formed therebetween. On the double-sided printed wiring boards 6 1 and 6 1 constituting the inner layer, a single wiring board pattern 6 1 as a final product (6 in the figure) is usually formed on the front and back copper foil surfaces by uranium engraving. a. . . . . . . 6 1 a and so on. Beforehand, drill at least two wiring boards 6 1, 6 1 above. # There are 2 reference holes 65, 65, because the reference holes 65, 65 are used to form the pattern 6 1 a on the front and back sides. . . . . . . 6 1 a, etc. Therefore, on a flat surface, the patterns on the front and back sides of the two-sided wiring boards 6 1 and 61 are ‘able to secure their positions with each other. In this way, the two double-sided wiring boards 6 1 and 6 1 use the reference holes 6 5 and 6 5 at the same coordinate position to form a pattern 6 1 a. . . . . . . 6 1 a etc. In the pattern formed on the double-sided wiring board 61 which becomes the inner layer board, except for the pattern 6 1 a where a single wiring board is formed. . . . . . . In addition to 6 1 a -8- (3) (3) 592008, usually more than 4 (minimum 3) guide marks 6 6 and 6 6 for reference holes are prepared. . . . . . These guide marks ^ are also formed in the etching process. The step of overlapping a plurality of etched inner layers with printed wiring boards on both sides so that the positional relationship of the patterns formed on the conductor portions of each wiring board can be aligned properly is called a lay-up. First, prepare and provide a pattern 61a on the double-sided printed wiring board 61. . . . . . . The center distances of the reference holes 65 and 65 φ used at the time of 61a are equal to the centering distance between the centering nails 6 8 a and 6 8 a. One etched double-sided printed wiring board 6 1 was placed on the bonding tool board 68 so that the bonding pins 68a and 68a of the bonding tool board 68 could pass through the reference holes 65 and 65 thereof. Reference holes 6 5 and 6 5 are drilled on the upper surface, so that the substrate material (called polyester film) 6 4 a before heating can be superimposed on it. The other two-sided printed wiring board 61 was placed on the bonding tool board 68 to overlap the bonding pins 68a and 68a of the bonding tool board 68 through the reference holes 65 and 65. At this stage, the outer periphery of the two double-sided printed wiring boards 6 1 and 6 1 and the polyester film 6 4 a between them are temporarily fixed to complete the lamination. Thus, the above reference holes 6 5 and 65 are also made. It is used as the reference hole for the layering, so it can be called the reference hole for the layering. In the following, to avoid confusion, the reference holes 6 5 and 65 are also referred to as reference holes 6 5 and 6 5 for the junction layer. As shown in the cross-sectional view of FIG. 9, when printed on both sides of the two layers to be layered- 9-(4) (4) 592.08 Wiring boards 6 1 and 6 1 are placed on both sides of polyester film 6 4, 6 4 and copper foil of conductive material 6 2, 6 2 and then heated by hot press At this time, the polyester film 6 4, 6, 4 a, 6 4 inserted between the copper box 6 2 or the wiring board 61 on both sides will be thermally coagulated (insulated) and changed to the substrate 6 3, and the bonding between the conductors is also completed, and One multilayer printed wiring board 60 is formed. Then, a new reference hole of the inner layer pattern of the multilayer wiring board should be opened, and the new reference hole should be used as a reference for the outermost conductor wiring pattern etching, through-hole drilling, etc. Then, electroplating process and rust-prevention treatment process are applied, and the machine is divided into single wiring boards by machining, and cut into a desired shape to complete a multilayer printed wiring board. As explained, in the conductor layer of the two-sided wiring board to be formed as the inner layer board, except for the pattern 6 1 a of the single wiring board. . . . . . . In addition to 6 1 a ′, three or more guide marks 6 6, 6 6, 66 for preparing reference holes are prepared, and the guide marks are also etched in the etching process. The coordinates of these guide marks are positioned as a pattern with a single wiring board 6 1 a. . . . . . . 6 1 a ensures a certain positional relationship, so it becomes necessary to measure the positions of these guide marks, to estimate the layout of the pattern that constitutes the circuit, and then to determine the position of the reference hole to be used in later engineering. In order to drill new reference holes, guide marks 66, 66 for the inner layer of the above-mentioned multilayer wiring board 60 should be used. . . . . . . , Usually using X-ray (reference hole) drilling machine. As described above, the front and back surfaces of the multilayer wiring board heated under pressure by a hot press are covered with a scale-free conductive layer. It is impossible to see the guide marks formed on the inner layer using visible light with the naked eye. . -10- (5) Although the general method used today is to measure the position of the guide mark formed on the inner layer board with a weak X-ray through the multilayer wiring board, there are various other measurement methods such as the use of ultrasound research. Either way, it is included in the measurement method other than visible light to measure the position of the guide mark formed on the inner plate. The reference hole drilling machine has an observation device for observing the guide mark by a measurement method other than visible light, and a drilling device for reference hole drilling. It is a multilayer wiring board that can be mounted and fixed as a processed object. (The front and back are provided with a non-scale conductive layer.) It consists of a table and a frame that supports them. In general, it has a mechanical coordinate system that is fixed on a frame and is described by three axes provided with X, γ, and Z. The above-mentioned observation device, drilling device and table are along the X axis and The Y axis moves on the XY plane. Then, the observation device and the drilling device are configured to be relatively movable to an arbitrary position of the multilayer wiring board placed on the table to perform observation of the guide mark and drilling of the reference hole. The guide marks are generally formed into a shape suitable for observation in a reference hole drilling machine, and the placement positions of the guide marks on the wiring board are generally formed on the outer periphery of the four corners of the wiring board to reflect the entire wiring board. Degree of deformation. Usually, the pattern 6 1 a on a single wiring board. . . . . . . In the configuration of 6 1 a, the design coordinate system is set first, and then the position and hole position of the pads constituting the pattern are described using the coordinate system 値 of the design coordinate system. The same design coordinate system is also used to describe the coordinates of the guide mark 6 6-11-(6) (6) 592008, and the design coordinate system is used to describe the coordinates of the reference hole to be used in the project after layering. . Strictly speaking, there is no positional relationship between the outer peripheral shape and the inner layer pattern of the multilayer printed wiring board after the work piece is laminated. Therefore, when the work piece is placed on the worktable of the drilling machine, the origin coordinate and the direction of the coordinate axis of the above-mentioned design coordinate system are unknown. Therefore, the reference hole drilling machine is used to observe the guide marks formed in the conductor layer of the inner layer of the multilayer printed wiring board placed on the table, and then use them as the coordinates of the mechanical coordinate system. For the observation of the guide marks for description, it is necessary to estimate the origin coordinate of the design coordinate system to which the inner layer conductor pattern of the multilayer wiring board belongs, and various calculation methods required for the XY axis direction are embedded in its control device. When an observation device observes a guide mark formed in the inner layer of a multilayer wiring board placed on a workbench, its coordinates 値 are displayed on a mechanical coordinate system fixed on a rack. Recently, there have been many observations with at least three (usually four) guide marks formed per conductor layer. Due to heat and pressure in the hot-pressing process, irregular movement occurs in one piece of copper foil to form the above-mentioned one-layer conductor layer, and even irregular positions occur in the positions of the guide marks between the layers. It is necessary to estimate the origin coordinate and XY axis direction of the design coordinate system with the smallest error between the observations and the coordinates of the guide mark on the design. . In order to estimate a certain distribution (function) from such numbers containing errors -12- (7) (7) 592008, the method of least squares is generally used to find the error 値 quadratic The sum of squares is used as the smallest function. In terms of specific procedures, first, the reference mark of the reference hole drilling machine is observed to determine the coordinate 値 of the mechanical coordinate system. After that, the number to be calculated is the two numbers of the origin position and the inclination of the design coordinate system used when the guide mark is formed. This is the number on the mechanical coordinate system described above. If the number 2 is known, the reference hole coordinate 値 known in the design coordinate system can be converted into the coordinate 値 described in the mechanical coordinate (converted using a conventional coordinate transformation formula). Therefore, it is easy to drill the reference hole coordinate position converted to the coordinate 値 of the mechanical coordinate system with the drilling device set on the reference hole drilling machine to move along the X and Y axes of the mechanical coordinate system. In addition, when the least square method is used in the above example, since the origin position of the design coordinate system is fixed at a certain number and it is easier to infer when calculating, the origin is called a fixed point. In most cases, the deformation of the multilayer wiring board occurs in a hot pressing process in which the multilayer wiring board is heated and pressurized by a hot press to form an integrated body. It is a point near the center of the multilayer wiring board. To deform the fixed point toward the outside, the amount of displacement of each point can be considered to be proportional to the distance from the fixed point. Strictly speaking, this fixed point will be different due to the inner conductor layer. Although the force applied by the hot pressing process will also cause subtle changes, the situation in which the guide mark is placed on the outer periphery of the pattern is slightly different. There are many guide marks, so all of them are regarded as (no points with mass m) mass points. When calculating the center of gravity of the guide mark • 13- (8) (8) 592008, the coordinates of the center of gravity 値 are almost formed. Near the center of the multilayer wiring board. For the sake of convenience, the center of gravity of the guide mark is regarded as a fixed point, and most of them will not be greatly wrong. Therefore, when the center of deformation (ie, the fixed point) of the multilayer wiring board in the hot pressing project is assumed to be the center of gravity of the guide mark and the least square method is used, the origin position of the design coordinate system and the inclination of the coordinate axis can be obtained. The applicant proposed the following calculation method as a patent application in "Japanese Patent Application No. 2000-1 30764 (filed on April 28, 2000)." The center of deformation is assumed to be the center of gravity of the guide mark. The positions of a plurality of guide marks are measured, and then the origin position of the design coordinate system of the inner pattern coordinates and the inclination of the coordinate axes are calculated. In this way, the deformation center (fixed point) of the multi-layer printed board is assumed to be the center of gravity of the guide mark. When the reference hole is drilled and hot-pressed, most of the conditions are practically sufficient as a reference hole with less deviation.来 用 〇 However, occasionally the position of the reference hole may be incorrect. For example, when there is only one layer of the inner layer-specific conductor layer, the deformation is caused when the fixed point is assumed to be the center of gravity. Especially in the through-holes or through-holes (electrical connection holes) of the full conductor layer penetrating the multilayer printed wiring board. . . . . . When the parts are installed, the contact holes between the layers are not used) or the displacement of the pads on the first layer is large, making it a defective product. Figures 10 (a) and (b) are a schematic plan view of a through hole provided on a multilayer printed wiring board and a cross-section projection view in the thickness direction. The same coordinates are on the all-conductor layer (6 layers in the figure). There are -14- (9) (9) 592008 circular solder pads 1 0 0 with a diameter of 2 R. A through hole 1 0 1 with a diameter of 2 r is penetrated in the center of these solder pads 1 0 0. The inner surface is plated with metal to make each pad electrically connected. When there is no displacement of each conductor layer, as shown in FIG. 10 (a), each pad 100 will overlap into a circle. The through hole 101 is the center of the entire pad 100 which can be completely opened. When the conductor layer is displaced, the position of the pad will be shaky, for example, the plan view is shown in Fig. 10 (c) and the side view is shown in Fig. 10 (d). When the arrangement of each pad is uneven, assuming that the through hole 1 01 is still inside all the pads 100, the position deviation of each pad is allowed to be less than the maximum limit (R-r). As shown in Figures 10 (c) and (d), when the displacement of the solder pad 1 0 0 (2) formed on the second layer of the conductor layer is particularly large, the through hole is positioned with the center of gravity as the fixed point. The deviation between the coordinates of 1 〇1 and the pad 1 〇〇 (2) will exceed (R-r) 値. At this time, if the through-hole 1 〇1 is opened at the center 102, it is shown in the drawing number 103 of (c) At the indicated position, the through hole 100 will exceed the outer edge of the pad 100 (2), making the through hole a defective product. In addition, 'If you want to separate and observe the outlines of the pads that are overlapping as shown in this figure, you cannot use ordinary X-rays to see through them. However, for convenience, the plan views (a) and (C) show the conductors. The perimeter of the layer can be seen through the imaginary image in the thickness direction of the multilayer substrate. However, the table 10 (e) and (f) diagrams are exactly the same as the pad deviations shown in Figs. 10 (c), -15- (10) (10) 592008 (d), but as long as the previous When the through-hole centered at point 102 is moved to center at point 102a, the through-hole 10a can be made inside all the pads 100, so that the multilayer printed wiring board can be made. Used for good quality. Although it is impossible in reality, if the operator can see all the holes that need to be seen at the same time as shown in Figure 1 ((c), (e)), he can correct the drilling position and adopt The point is to make the reference hole drilling at the center of gravity of the guide mark, so that the operator can grasp the better reference hole position. [Summary of the Invention] [Problems to be Solved by the Invention] As can be seen in the example of the through hole described above, it can be seen that when setting a reference hole that will become a processing standard after hot pressing, there is also a point of gravity that is more important than the fixed point in the guide mark. To determine the reference hole position, there are also better reference hole positions. If the position of such a reference hole can be known, the problem of often considering the guide mark as the origin of the coordinate can be placed on how to improve the accuracy of the multilayer printed wiring board. Therefore, the current standard hole drilling machine is required The big problem is to be able to improve the accuracy of the product. Although it is rare in terms of probability, the reference hole opened by the drilling machine is used in the second half of the manufacturing process, especially the defects after the half of the manufacturing process cause the cumulative investment cost is large, even if only a small increase Accuracy can also produce a large economic effect, so the expectation of improvement from the user side is also ®. -16-(11) (11) 592008 [Means to solve the problem] In order to solve the above problems, the present invention provides a reference hole drilling machine, which is provided with: a vertical Xm axis and a Ym axis have been set Of the mechanical coordinate system of the present invention is fixed to the frame of the reference hole drilling machine housing; the design of the printed wiring board conductor layer formed on the constituent elements of the multilayer printed wiring board and fixed on the conductive layer of the printed wiring board Coordinate system to describe the guide mark of its coordinate 値, the observation device of the guide mark for at least 3 points observation; the drilling device for the reference hole opening of the printed wiring board, with the above X m axis and Y m A conveying device that converts the relative positions of the printed wiring board, the observation device, and the drilling device, and the axis is a parallel conveying direction; and calculates the design coordinate system from the guide mark coordinates 値 observed by the observation device The configuration of the reference hole coordinate 値 described in the above-mentioned design coordinate system is converted into the coordinate 値 of the above-mentioned mechanical coordinate system, and the above-mentioned conveying device is controlled. In the reference hole drilling machine of the manufacturing equipment, the control device is provided with: setting a weighting function to designated 値 to calculate an origin of the design coordinate system of the printed wiring board as a coordinate described on the mechanical coordinate system Means for calculating the origin coordinate 値 of 的; means for calculating the coordinate axis angle of the angle formed by the coordinate axis of the above-mentioned design coordinate system and the coordinate axis of the above-mentioned mechanical coordinate system; and the coordinates 上述 and The observation of the guide mark calculates the individual error, compares the individual error, and executes the error comparison means of filtering the above-mentioned design coordinates. According to the reference hole drilling machine of the present invention, after the coordinates 値 of the above-mentioned guide marks are observed, the weighting function of all the guide marks is taken as 1 to calculate -17- (12) (12) 592008. The first design described above After comparing the origin coordinate 値 of the coordinate system and the angle of the coordinate axis, change the weight function 値 to calculate the origin coordinate 値 of the first design coordinate system and the angle of the coordinate axis according to the individual errors of the guide marks. It is decided to adopt any one of the first or second design coordinate system mentioned above. In addition, the reference hole drilling machine of the present invention is only the guide with the maximum error 値 and minimum 个别 of the individual error with the guide mark calculated from the origin coordinate 値 and the angle of the coordinate axis according to the first design coordinate system. The weighting function of the quotation mark is taken as 1 to calculate the origin coordinate 値 of the second design coordinate system and the angle of the coordinate axis. The individual errors of the first and second guidance marks are compared with each other to determine whether to use the first Any of the design coordinate systems in which the error range is small in the design coordinate system of 1 or 2 above. Furthermore, the reference hole drilling machine of the present invention may be configured as follows: the weighting function of all the guide marks is taken as 1, and after calculating the origin coordinate 値 of the first design coordinate system, and after the angle of the coordinate axis, Change the weighting function 値, repeat the origin coordinate 値 of the above-mentioned design coordinate system multiple times, and calculate the angle of the coordinate axis, and at the same time each calculation, perform the comparison of the individual errors of the guide marks, When the above-mentioned individual errors increase, or when the minimum error set in advance decreases, or when the number of calculations reaches the preset number of repetitions, the calculation operation is terminated, and one of the designs is adopted. Coordinate system. -18- (13) (13) 592008 [Embodiments] [Embodiments of the Invention] For convenience of explanation, the explanation is divided into the following four items. 1. Construction instructions for the drilling machine. 2 From the observation of the guide mark, estimate the design coordinate system of the inner conductor pattern to which the least square method is applicable. 3. Explanation of calculation formula. 4. Description of the embodiment of the present invention. 1 . Construction instructions for the drilling machine. A reference hole drilling machine (hereinafter abbreviated as a drilling machine) of a multi-point distribution method according to an embodiment of the present invention will be described with reference to FIGS. 1 to 3.帛 in is a perspective view of the appearance of the drilling machine 1 of the present invention, which is represented by the perspective of the casing 2. The table 2 and 3 are the projection views of the surface drilling machine, and the second (a) is the drilling The front view of the machine, Figure 2 (b) is its side view, Figures 3 (a) and (b) are plan views when the position of the movable table 12 of the drilling machine 1 is changed, and Figures 2 and 3 Both are represented by see-through casing 2. In addition, the mechanical coordinate system (X m, Y m, Z m, origin 0 m) shown in each figure is the coordinate fixed to the fixed part (such as the housing 1 or the frame 3) of the drilling machine 1. The movement direction of various mechanical parts of the system and the conveying device is parallel to the coordinate axis. The coordinates 値 obtained by observing the guide marks of the multilayer wiring board with an X-ray camera, or the drilling coordinates of the reference hole are basically calculated using this coordinate system. -19- (14) (14) 592008 In addition, the white arrow 17 in the first figure indicates the standing position of the operator. The operator is standing in the direction of the arrow (the direction of the Ym axis), and inserts the reference hole to be drilled. After the drilling is finished, the multilayer printed wiring board is taken out from the drilling machine. Here, so that the conductor layer after the hot pressing is still placed is a copper foil (6 layers) multilayer printed wiring board, and then the guide marks of the inner conductors are observed separately, and then the reference hole is made. The drilling operation is explained. Fig. 4 is a schematic diagram showing peripheral portions of a guide mark 66 provided at four corners of the multilayer printed wiring board 60. The shape of this example is approximately the shape of an actual guide mark. As shown in FIG. 4 (a), the inner layer conductor 62a has a pattern 6 1 a and 6 1 a (six in the figure) of a single wiring board, and a peripheral-shaped copper foil is left there. Corner windows 7 1 a to 7 1 d are formed at the four corners of the edge-shaped copper foil. Inside the round window 7 3 formed near the corner window, a larger diameter than the reference hole is left as the reference hole. Round copper box with symbol 7 2. In addition, Figure 4 (a) is a perspective drawing of each inner layer conductor 62a (2) from the same direction. . . . . . . 62a (5). When referring to the cross-sectional view of the multilayer printed wiring board 60 of FIG. 4 (c), after removing the conductors 62 and 62 which are the surface layer of the non-scale copper foil, the corner windows 7 1 a to 7 1 d are arranged on the 4th layer. Inner conductor 6 2 a. . . . . • The position where 6 2 a can see through. Inner conductor 6 2 a. . . . . . . 6 2 a's guide mark 6 6 is configured so that it does not overlap when seen through the angle window 7 1 -20- (15) (15) 592008. Therefore, when the corner window 7 1 3 arranged at the upper left is widely seen, the corner window 71 a of each conductor layer will have the same result as the one shown in Figures 4 (d) to (g). As shown in Fig. 4 (b), four guide marks 66 (2) to 66 (5) will appear in the corner window 71'a. In addition, in the inside of the round window 7 3, the reference hole mark 72 is overlapped and a ring gap is formed. Therefore, the same marks appear at the four corners of the multilayer printed wiring board 60. φ In the above example, the drilling machine makes four observation marks on each corner of the multilayer wiring board, memorizes the coordinates 値, and it makes 16 observation marks on the four corners. After the observation of all the guide marks, the reference hole position converted into the mechanical coordinate system is drilled with the reference hole. In order to drill a reference hole in a 6-layer multilayer printed wiring board, the number of observations is required. However, in the following description of the operation, the observation will be omitted and described once. In addition, there are four guide marks and reference holes at the same time, and it is assumed that there are reference holes near the guide mark to be handled. ® As shown in Figures 1 to 3, the frame 3 is fixed inside the casing 2 of the drilling machine 1. The left and right pair of X-direction moving frames 10 and 10 are formed roughly in a channel shape, and the left and right shapes are in a mirror relationship. The X-direction ′ moving frames 10 and 10 are guided by linear guides arranged on the upper end of the frame 3. 10a and 10a are supported. The ball screw 1 〇b and the ball nut (not shown) mounted on the lower side of the X-direction moving frame 1 〇 engaged with the ball screw 10 ′ According to the size of the wiring board to drill the reference hole, make it stand by in advance. When moving parallel to the X m axis, the position of the guide mark can be observed. -21-(16) (16) 592008 In addition, in order to individually drive the X-direction moving frames 10 and 10, each X-direction moving frame 10 is provided with a ball screw 10b. X-ray generating devices 4, 4 are fixed to the upper portions of the X-direction moving frames 10, 10, and linear guides 1 1 a, 1 1 a are installed at the lower portion. Next, the Y-direction moving frames 1 1 and 1 1 are supported by the linear guide 1 1 a. The ball screw 1 0 b and the ball nut (not shown) mounted below the X-direction moving frame 10 are engaged with it, so that the Y-direction moving frame 1 1, 1 1 can be moved parallel to the Y m axis. . The φ Y-direction moving frames 1 1 and 11 are formed in a channel shape, and an X-ray protective tube 5 is arranged on the upper part. As shown in FIG. 2, a clamping piece 9 arranged with the X-ray protective tube 5 and Cylinder 9 a with clamp plate 9 moving up and down. A mandrel 7 and an X-ray camera 6 are fixed to the lower portion. Arranged in parallel to the Y m-axis, the linear guide 1 2 a and the ball screw 1 2 b fixed to the center of the housing are used to support and drive the movable table 1 2 which can mount a multilayer printed wiring board. Parallel to the Y m axis • The movable table 12 is placed at a position of 12 A, and a multilayer printed wiring board to be used as a reference hole of the work piece is drilled, and then the guide mark is measured by moving along the γ m axis. It is pulled into the reference hole drilling position. · Drive ball screws 10b, 1b and 1b, The control devices for controlling the movement of the X-direction moving frames 1 0, 10 and the γ-direction moving frames 1 1, 1 1 ′ and the movable table 12 are not shown in the figure. The main components are observation devices that use X-ray production -22- (17) (17) 592008 to form guide marks on the X-ray production device 4 and X-ray protective tube 5 and X-ray camera 6, respectively. 9 to form a drilling device, which moves the carriage 10 in the X direction, the carriage 11 and the movable table 12 in the Y direction, and the linear guides 1 0 a, 1 1 a, 1 2 a that support and drive the same. And ball screws 10b, 1 lb, 12b, etc. form a driving device. In addition, the control device (not shown) controls the above-mentioned various devices in accordance with a continuous drilling operation sequence. Coordinate 値 is calculated from the guide X-ray image observed by the observation device, and it also plays the role of calculating the reference hole drilling position from this coordinate 値 and the reference hole design coordinate 输入 entered in advance. The movable table 1 2 is usually made of metal and formed into a flat plate shape. The left and right openings have openings for guide marks for perspective, reference hole drilling (1,2 holes), 1 3, 1 3, and openings (3 For 4 holes) 1 3 a, 1 3 a ° The multilayer printed wiring board 60 is placed on the movable table 12 in order to make the front end of the wiring board neatly according to its external dimensions. Reference holes 1 and 2 use openings 1 3, and reference holes 3 and 4 use openings 1 3 a for drilling. 2. Working sequence The following describes the working sequence using the above-mentioned drilling machine, taking a 4-hole reference hole as an example. Fig. 4 is a schematic diagram showing a peripheral portion of a guide mark 66 located at four corners of the multilayer printed wiring board 60. The shape of this example is approximately -23- (18) (18) 592008. Actual guide mark shape. As shown in FIG. 4 (a), the inner-layer conductor 62a is a pattern 6 1 a, 6 1 a with 6 single wiring boards, and the peripheral portion is a copper foil with a rim shape. Corner windows 7 1 a to 7 1 d are formed at the four corners of the rim-shaped copper foil. Inside the round window 73 formed near the corner window, a larger diameter than the reference hole is left as the reference hole mark 7 2 Round copper foil. In addition, Figure 4 (a) is a perspective drawing of each inner layer conductor 6 2a (2) from the same direction. . . . . . . 62a (5). · As shown in Figure 4 (c), when referring to the cross-sectional view of the multilayer printed wiring board 60, the conductors 6 2, 6 2 which are the surface layer of the copper foil are removed, and the corner windows 7 1 a to 7 1 d It is an inner layer conductor 6 2 a arranged in 4 layers. . . . . . . 6 2 a large set in the same position. Formed as inner conductor 62 a. . . . . . . The guide mark 6 6 on 6 2 a is configured so as not to overlap when seen through the corner window 7 1. When the 4th (d) to (g) diagrams illustrating the periphery of the corner window 71 a arranged at the upper left are combined, the four guide marks 66 (2) to 66 are shown in the 4th (b) diagram. (5) Spring will appear in the corner window 7 1 a. In addition, in the inside of the round window 7 3, the reference hole mark 72 is overlapped and a ring gap is formed. Therefore, at the four | corners of the multilayer printed wiring board 60, corner windows 71a to 71d and reference hole marks 72a will appear. ~ 7 2 d is the same symbol as above. In addition, although the reference hole mark does not need to be formed on the conductor layer, most of them are set as confirmation of reference hole processing and subsequent work marks-24- (19) (19) 592008 Four observation marks were taken at each place, and the coordinates 値 were memorized, and a total of 16 observation marks were observed. In All Guiding Signs. After the measurement, the reference hole position converted into the reference position of the mechanical coordinate system is drilled. ^ In order to drill a reference hole on a 6-layer multilayer printed wiring board, the number of observations is required. However, in the following description of the operation, the observation will be omitted at one point for explanation. In addition, there are four guide marks and reference holes at the same time, and the guide marks and reference holes are assumed to be adjacent. Lu The camera position on the side of the drilling machine is arranged so that when observing the guide marks, the guide marks are one by one and enter the field of view of the camera. Therefore, as an order of strategy, the following three steps are needed: (1) Observing the guide mark 6 2; (2) Deducing the coordinate origin and the coordinate origin of the design coordinate system from the observation 导引 of the guide mark 6 2. Coordinate axis inclination; (3) Convert the reference hole position described in the design coordinate system into a mechanical coordinate system. In order to avoid complication in the following sequence description, there is a guide mark in one corner and window, and a reference hole is provided near each guide mark. The guide mark and reference hole are arranged in the same corner. Is for neighbors. The diagram of the inner layer conductor 6 2 a of the multilayer printed wiring board after the hot pressing is strictly a non-coordinate relationship with the shape of the peripheral portion of the wiring board. However, in most cases, it is possible to guide the outline of the peripheral portion so that the guide mark is incorporated in the field of view of the camera. Therefore, in general, a multilayer printed wiring board placed in a drilling machine guides its outer shape so that the guide mark is placed in the field of view of the camera. -25- (20) (20) 592008 First, from the multilayer printed wiring board of the work piece to be drilled as the reference hole (for example, shown in Figure 4 or Figure 8 and Figure 9), the external dimensions of 0 and the (design The reference hole coordinates are used to determine the position on the Xm axis of the moving frame 10 and 10 along the X direction. In advance, the X moving frame 10 and 10 are moved to this position and waited. The movable table 12 is at the position 12 A (of FIG. 1), and the worker places the multilayer printed wiring board 60 at the designated position on the movable table 12. The wiring board 60 is temporarily fixed on the movable table 12. From the point of view of the operator, the movable table 12 is moved so that the two guide marks (6, 6 6) on the front end side of the wiring board 60 are located on the X-ray generating tube 4 a built in the X-ray generating device 4. Below. The guide marks (6, 6 and 6) on the front side of the X-ray perspective are observed on the X-ray cameras 6, 6 and the coordinates 値 are measured. The coordinates 値 are stored in the memory of a control device (not shown). The distance that the movable table 12 moves in the direction of Y m is only the distance to move the guide mark (66, 66) in front of it closer to the X-ray generating tube 4a. Next, X-rays are irradiated, and two guide marks are observed on the X-ray cameras 6, 6 and the coordinates 値 are memorized. Here ', the following calculation method will be used to calculate the origin coordinate of the design coordinate and the inclination of the coordinate axis from the coordinates of the four points of the guide mark, and calculate the four (mechanical coordinate system) coordinates of the reference hole from this 値First, the mandrels 7 and 7 are moved to the coordinates of the two reference holes in front of them, and the reference hole is drilled. Next, the movable table 12 is moved to return the guide mark on the front side (-26- (21) (21) 592008 6 6, 6 6) to the position near the X-ray perspective, and the spindles 7 and 7 are moved to At the coordinate position of the reference hole on the front side, two reference holes are drilled. Then, the movable table 12 is moved to the insertion position 12a. When the operator removes the drilled wiring board 60, the reference hole processing ends. engineering. In addition, as shown in FIG. 4, even when the number of the guide marks is actually 4 for each guide mark, the number of measurement times is only 4 times, and the execution of the sequence is the same as the above. Hardly changed. The reference holes drilled here are the openings in the pattern formation of the two-sided conductor layer on the outer layer or the inside of the pattern such as through holes and perforations. These holes are reference holes that are inserted through the multi-layer substrate in the positioning junction nails provided on the junction tool board to determine the positions. In fact, in order to maintain the processing accuracy, most of them use 4 or more tie nails at a time. However, if the tie nails inserted in the tie tool board are used several times, the hole diameter will be enlarged and the reference hole will no longer be used. Therefore, there may be a situation where different reference holes are used in each process. In this case, the reference hole of the complex array in which the number of layer nails of the plurality of layering tool plates to be used or the coordinates are used is to be drilled in a later process. Since the coordinates of the reference holes of these complex arrays are also described on the same design coordinate system, as long as the origin coordinates of the design coordinate system and the inclination of the coordinate axes are determined, then only the respective coordinate calculation and drilling are added. Just the number of trips can, in principle, repeat the same thing. -27- (22) (22) 592008 3. The description of the calculation formula is general. 'Because the reference holes are post-processed piercing nails that are inserted into the tooling board for bonding, which is used for positioning the work piece, the reference holes of one group have the same tools as those used for bonding. The coordinate positions of the layered nails of the board are the same. The distance between the reference holes should be within the specified range, and the holes should be opened within a certain error from the conductor pattern. Since this is achieved by allocating the error to each reference hole, it is usually called a distribution type. When the number of guidance marks to be observed is three or more, it is sometimes referred to as a multi-point distribution type. As described above, the guide mark formed on the conductor layer is formed at the same time as the conductor pattern. The coordinates of the conductor patterns and reference holes formed on the inner layer of the multilayer printed wiring board, as well as the coordinates of all conductor layer elements such as guide marks, are all described on a design coordinate system, and the relationship is known. Therefore, the following procedures are needed: (1) Observe the guide marks formed on the multilayer printed wiring board on the drilling machine table, and read the coordinates of each guide mark indicated by the mechanical coordinate system fixed on the drill machine.値; (2) Obtain the origin coordinate of the design coordinate system and the inclination of the coordinate axis from this 値; (3) If you want to drill with a drilling device mounted in a drilling machine, you need to describe the design coordinate system The procedure of transforming the coordinate 値 of the reference hole into the coordinate 表示 represented on the mechanical coordinate system. In this coherent program, (1) is simply observation of the guide mark, and (3) is easy to calculate as long as the coordinate transformation formula used for the well-known trigonometric function is used. (2) The main origin coordinates of the designed coordinate system and the inclination of the coordinate axis are obtained from the observed coordinates of the mechanical coordinate system of each guide mark. -28- (23) (23) 592008 The procedure is multi-point allocation The main part of the formula. Since this calculation formula has been explained in detail in the above-mentioned "Japanese Patent Application", only the main points are described below. Referring to Fig. 5, the relationship between the design coordinate system when designing the conductor pattern and the mechanical coordinate system set on the drilling machine will be described. Fig. 5 (a) is a plan pattern diagram for explaining the design coordinate system. The design coordinate system is based on 0 D as the origin, and has a U D axis corresponding to the X axis and a V D axis corresponding to the Y axis which are perpendicular to each other. A multilayer printed wiring board 60 is placed, and a single wiring board pattern 6 1 a A is formed on the inner conductor layer (6 in the figure). . . . . . . 6 1 a A and guide mark P D 1. . . . . . . PD4 and reference hole HI. . . . . . . The positions of H4 and so on are described by the coordinate 値 of the design coordinate system. Using this design coordinate system, the shape or hole position inside the pattern 6 1 a A of the single wiring board is also described as the coordinate 値 of the system. As shown in Fig. 5 (b), the focus is on the guide mark P D 1 formed on the multilayer printed wiring board 60. . . . . . . P D 4, the center of gravity when the guide mark is regarded as the mass point as the center of rotation, set (new) design coordinate system: with ◦ as the origin, with U axis, V axis parallel to U D axis, V D axis. Figure 5 (C) shows the mechanical coordinate system (origin point) set on the drilling machine by drawing the guide mark positions P1 ~ P4 (according to their observation coordinates 値) individually observed by the X-ray camera. , With Xm and Ym as vertical coordinate axes). Then take the guide marks P 1 to P 4 as the mass points and find the center of gravity of 0g, and set the (new) mechanical coordinate system-29-(24) (24) 592008 system: with 0g as the origin of the coordinates, with parallel On the X m axis and Y m axis. In addition, although the design coordinate system and the mechanical coordinate system both move the origin to the position of the center of gravity of the guide mark, a (new) design coordinate system and (new) mechanical coordinate system are formed. The coordinate axes are all parallel to each other, so the coordinate transformation can be performed by simply adding and subtracting the coordinates 新 of the new origin. Here, the guide marks P D 1 to P D 4 on the multilayer printed wiring board 60 shown in Fig. 5 (b) are drawn on the top of Fig. 5 (c). First, think of the multilayer printed wiring board 60 as the origin of the (new) design coordinate system. 0 is consistent with the origin of the (new) mechanical coordinate system. 0g is the same as the origin of the (new) mechanical coordinate system. (Yes) are parallel, and the V-axis and Y-axis (both Y and m-axis are) are parallel. From the previous (new) design coordinate system and (new) mechanical coordinate system setting process, we know that the X m axis and U D axis, Y m axis and V D axis are parallel to each other. In FIG. 5 (d), the black points at the vertices of the rectangle S indicated by a dotted line in one point are equivalent to (designed) the guide marks P D 1 to P D 4. (Design) Guidance marks PD 1 ~ PD 4 and (Observed) Guidance marks P 1 ~ P 4 are placed on a piece of paper, but at this stage only the centers of gravity of the two are consistent. . Multiply the distance L 1 between the (designed) pilot mark PD 1 and the (observed) pilot mark P 1 by the quadratic power, and multiply the distance L 2 between PD 2 and P 2 by the quadratic power. Similarly, the total of the distances L 3 and L 4 multiplied by the quadratic power is equal to the minimum α ° (for example, the angle formed by the U axis and the X axis is -30- (25) (25) 592008). The result is calculated by the double method. In Figure 5 (d), the above-mentioned multilayer printed wiring board 60 is rotated around the origin of the (new) design coordinate system to α °, for example, a (design) guide mark (designed) indicated by a black dot ( PD 1) Rotate to PD 1 represented by a white dot. At this time, the distance between the (observed) guidance mark P1 and the (designed) guidance mark PD1 is represented by L1, and the distance between PD2 and P2 is represented by L2. Similarly, the remaining distances are represented by L3, L4. . On the 5th (d) graph, the position where the sum of the squares of L 1 to L 4 is equal to the minimum 値 is plotted. In the above description, the formula for calculating α and its calculation method have been described in detail. When the absolute value of α is small, when the coordinates of the guidance symbol P i indicated by the (new) mechanical coordinate system are For X i and yi, when the coordinates of the (new) design coordinate system and the indicated guide mark PD i are U i and Vi, the tangent ^ of ^ can be obtained by the following [Equation 1]. Σ (y V,) sin ai _ l____ tan a = ---- '~ C〇S a Σ (χ, υ, + yi -1 ···· [calculation formula 1] Conversely, although it is based on the above (new ) Design the origin coordinate (position of the center of gravity) of the design coordinate system and the angle of the coordinate axis a ° 'Enter the (designed) guide marks PD 1 to PD 4 in the (new) mechanical coordinate system and mechanical coordinate system, but also It can be said that it is represented by Fig. 5 (d). As shown in Fig. 5 (d), the guide marks (designed) PD 1 to PD 4 and the (observed) guide marks P 1 to P The distance between the distance of 4 L 1 -31-(26) (26) 592008 ~ The sum of the least squares of L 4 can also be learned from actual verification. In this way, using the least square method, Find the origin of the (new) design coordinate system (center of gravity position) 'and the angle α of the coordinate axis represented by the (new) mechanical coordinate system. With the origin of the (new) design coordinate system (center of gravity position) , And the angle of the coordinate axis, so that the coordinates described in the design coordinate system can be converted into the coordinates on the mechanical coordinate system. Therefore, the drilling machine The reference hole described in the design coordinate system can be transformed into a coordinate 値 on a mechanical coordinate system for drilling. Explanation of the embodiment of the present invention. The least square method is generally described as follows. That is, measurement Object to obtain n observations 値 (f (X i). . . . . . i = 1 ~ ^. . . . . . (Expression 1)], appropriately select a weighting function [ω (X): weighting function] that weights n observations 値, and assume a simpler function form gn (X) to find the difference from the weighting function 値The product of the quadratic power is calculated, and the gn (x) when the sum Ω (i = 1 to η) is the smallest is obtained. Therefore, Ω is expressed by the following formula (2). {Ω = Σ ω (X i) x [f (X i) — g n (x i)] 2. . . . . . (2)} Here, Σ represents the total obtained by substituting 1 ~ n into i of ω (xi) x [f (xi) -gn (xi)] 2. In this example, n pilot symbols are observed to obtain (observed) n pilot symbol coordinates 値 (X i, y i). The number you want to know from this perspective is the origin coordinate of the design coordinate system -32- (27) (27) 592008 indicated on the mechanical coordinate system and the tilt of the coordinate axis. The order used in the least square method is to assume an appropriate weighting function to determine the origin coordinate of the design coordinate system, and to find the angle of the coordinate axis with the smallest error at the origin position. In addition, although it is repeated, it is known that in the conductive layer of a multilayer wiring board, the coordinates of the conductor pattern or the elements such as the types of holes, guide marks, reference holes, etc. are all described in the design coordinate system. on. Therefore, knowing the origin coordinate of the designed coordinate system and the inclination of the coordinate axis indicates that all the configurations (coordinates) of the elements to which the conductive layer belongs can be converted into a mechanical coordinate system. The weighting function treats all (observed) n guidance marks as 1, and uses the center of gravity of the guidance marks as the fixed point of the design coordinate system to obtain the inclination of the coordinate axis of the design coordinate system. In the manual of "Yoshimoto", it is indicated in the manual that other fixed points can also be adopted. Although the definition of the weighting function explained earlier has strict points, for n observation marks (observed), in order to determine the origin coordinate 値 of its design coordinate system, the appropriate weighting function ω ( X) is referred to as the use of the weighting function ω (χ). In the above-mentioned appellation method, the observed guide marks P 1 to P 4 are regarded as particles, and the center of gravity becomes a fixed point, which means that all weighting functions are 1, that is, (X) = 1 is not necessary. For example, when the total number of guidance marks is 4, use X i as its X coordinate 'to request the center of gravity of the X coordinate G (X), use m as its mass' whose calculation is -33- (28) (28) The 592008 formula is (G (x) = (mxxH-mxx2 + mxx3 + mxx4 / 4m = I (xi) / 4. . . . . . (3)] 'All weighting functions corresponding to each navigation mark are regarded as 1. When the choice of the weighting function is changed, the result of using the least square method can be greatly changed. FIG. 6 is a schematic diagram showing (designed) guide marks PD 1 to PD 4 and (observed) guide marks P 1 to P 4 provided at four corners of a multilayer printed wiring board. The (observed) guide marks P 1 to P 4 are formed in four corners as shown in FIG. 4. The four guide marks in each corner of the multilayer printed wiring board are collective design coordinates. They are represented by a circle with diagonal lines as the (design) guide marks P D 1 to P D 4. Corresponding to the (designed) guide mark P D 1, the (observed) guide mark P 1 is represented by the outer perimeter of a circle whose center is moved only by an error amount. The other three corners are also shown in the same way. The centers of the guide marks P D 1 to P D 4 (designed) are located at the vertices of the rectangle S. The relationship between these is the same as the relationship between the through hole and the pad in Fig. 10, and the oblique circle of the guide mark PD 1 to PD 4 (designed) is shown on the display (observed) When the outer circumference of the circle of the guide marks P 1 to P 4 exceeds the margin of error, it can be regarded as a defective product. Fig. 6 (a) shows the practice of using the (designed) guide marks PD1 to PD 4 and the (observed) guide marks P 1 to P 4 to be at the same position (as a fixed point). (29) (29) 592008 The results obtained when PD 1 ~ PD 4 positions are determined by the known method. The circle of the (design) guide mark PD 4 and the (observed) ) The guide mark P 4 (2) crosses to make it a defective product. In view of this, the main point of the embodiment of the present invention is to use the least square method with the center of gravity as the fixed point, and then change the selection method of the weighting function to change the position of the fixed point. Compare with less error. The result of changing the position of the fixed point is shown in Fig. 6 (b). The origin of the coordinate of the design coordinate will be moved from 0 g to 0 2. When the above [calculation formula 1] is used as the new origin coordinate, the angle of the coordinate axis will also be Will change from α ° to α 2 °. As a result, the (designed) guide marks PD 1 to PD 4 are moved and rotated while maintaining the relationship of the rectangle S, and moved from the position shown in FIG. 6 (a) to the position shown in FIG. 6 (b). . As a result, the guide mark (designed) P D 4 and the (observed) guide mark P 4 (2) do not intersect, and it is determined that the reference hole is opened at a position recognized as a good product. As an example of the embodiment of the present invention, the calculation and execution of the design coordinate system estimation are performed by the control device of the reference hole drilling machine. That is, the control device that controls the observation device, drilling device, and conveying device of the reference hole drilling machine includes a calculation program for calculating the origin coordinate, the angle calculation method for the coordinate axis, and the error comparison method. The command from CP u in the control device performs the sharing processing for each calculation. -35- (30) (30) 592008 With reference to the block diagram of Fig. 7, the description of the calculation content to be processed by these 3 methods and the actual processing sequence flow will be explained. Fig. 7 (a) shows the local structure in the control device, and Fig. 7 (b) shows the flow of the processing procedure along the arrow. Each means in Fig. 7 (a) is to process each process in Fig. 7 (b) recorded below it as indicated by the arrow. The calculation procedure is as follows: Based on the observation mark of the guide mark, (1) Calculate the coordinate origin (center of gravity position and position) by the calculation method of origin coordinate 値, and use the least square method to calculate the center of gravity as the fixed point. The (origin) coordinate origin of the (new) design coordinate system represented by the (new) mechanical coordinate system is calculated by the calculation method of the coordinate axis angle, and the angle (α °) of the coordinate axis is calculated to obtain the first Coordinate system. Then, the coordinate 値 of the guide mark P D i (on the design) is converted into the coordinate 値 of the mechanical coordinate system. In the past, the coordinate system 座 of the mechanical coordinate system of the reference hole was obtained using this design coordinate system, and it was immediately formed into a reference hole. Continuing the calculation of the first coordinate system, the following calculations are performed. · (2) Find the error (individual error) of each (observed) guide mark P i (designed) with the guide mark P D i. The individual error is calculated on the mechanical coordinate system based on the difference of the coordinate axis. That is, the coordinate 値 of each (designed) guide mark PD i is (PD i X, PD iy), and the coordinate 各 of each (observed) guide mark P i (j) is [P i (j ) X, P i (j) y], when the individual errors distinguished by the coordinate axes are Δ P i (j) x, Δ Pi (j) y, the calculation of the x and y axes is as follows. -36- (31) (31) 592008 The calculation of the X axis is (APi (j) x = Pi (j) x-PDix. . . . . . (4) The calculation of the y-axis is (ΔΡΚΌγζΡΚηγ-ΡΟίγ. . . . . . (5)] Here, i is equivalent to (designed) the guide mark P D i (numbers at the corners of the outer periphery of the substrate) 'j is a number corresponding to the conductor layer [for example, refer to FIG. 4 (c)]. (3) Select the (observed) guidance symbol P i (j) with the largest and smallest errors. As a general procedure, it is to select guide marks at both ends of an error series whose errors are arranged in ascending or descending order. That is, the individual errors that are distinguished by the coordinate axis include the symbols arranged in order from small to large, or arranged in order from large to small, and the pilot marks P i (j) located at both ends thereof are selected. For example, for the X axis, select P 2 (3) and P 4 (1), and for the γ axis, select P1 (4) and P3 (2). In addition, when there are a plurality of guidance marks with the same error, only one of them needs to be selected. (4) Find the average calculation 値 of the coordinate 値 indicated by the mechanical coordinate system of each guide mark, and use this 値 as the second coordinate origin. For example, in the above example, [[Ρ2 (3) χ + Ρ4 (1) χ] / 2 = the X coordinate of the new fixed point. . . . . . (6)] [[ρ 1 (4) χ + ρ3 (2) χ] / 2 = the y-coordinate of the new fixed point. . . . . . (7)] The above calculation is equivalent to the case where the weighting function of the guide marks at both ends of the error series is set to 1 by the usual calculation formula of the center of gravity, and the other is 0. A point where the average calculation 値 of the above coordinates 値 is used as the x and y coordinates as the second origin 0 g 2 is used to form a second mechanical coordinate system. Machine 2 -37- (32) (32) 592008 The X 2 axis and Y 2 axis of the coordinate system are parallel to the original X m axis and Y m axis. This can be applied to Fig. 6 (b). However, because it is very close to the U V axis, it is not included in the X 2 axis and Y 2 axis. The (observed) coordinate 値 of the guide mark p soil (j) is updated to a new 値 by moving the coordinate origin from ◦ g to 0 g 2 in parallel. Move the (new) origin coordinate of the s-ten coordinate system to 0 g 2. Since the center of gravity of the (design) guide marks PD 1 to PD 4 already has the origin of the coordinates in the (new) design coordinate system, that is to say, the (new) design coordinate system as a whole (including its coordinates) The upper point) is a parallel shift from 0 g to 0 g 2. In the calculation method of the coordinate axis angle, the second fixed point 0 g 2 is used as the second fixed point, and the angle formed by the coordinate axis of the design coordinate system and the coordinate axis of the mechanical coordinate system (α 2) is obtained by the least square method. The calculation formula can be [Calculation Formula 1]. With the origin coordinate 2 of the second design coordinate system (equivalent to the second origin 0g 2) and the angle of the coordinate axis (α 2 °), the (designed) guide mark PD i can be described in the mechanical coordinate system on. The coordinates (designed) of the guide mark PD i indicated on the second coordinate system are transmitted to the error comparison means; (5) The calculation (conversion) from the second design coordinate system to the mechanical coordinate system (design Coordinates of the pilot mark PD i and the (observed) pilot mark p 丨 (j) are individual errors distinguished by the coordinate axis. The calculation method is the same as the individual error calculation of the first coordinate system described above. (6) The individual error (2) of the second coordinate system is compared with the individual error (2) of the first coordinate system of the conventional hand -38- (33) (33) 592008 method, and the judgment is made. In general, it is sufficient to use the smaller error range (maximum error-minimum error). (7) Adopt a favorable design coordinate system configuration. In addition, if there are certain specific requirements, such as when a certain conductor layer is particularly important, or if the error of its layer is selected to be small, etc., the determination conditions for whether or not to use it may be appropriately determined. In this way, a calculation program is built in the control device that controls the observation device, drilling device, and conveying device of the reference hole drilling machine, and estimates using the origin coordinate calculation method, coordinate axis angle calculation method, and error comparison method. The position of the designed coordinate system is determined, and the coordinate system of the multi-point distributed reference hole is determined by the mechanical coordinate system set on the drilling machine. In particular, it is newly added that: the calculation of the individual errors of the guide mark, the selection of a guide mark with a specific error, and the determination of which is a favorable coordinate system error comparison means, it is possible to obtain the most from a plurality of design coordinate systems. Appropriate; by choosing the most appropriate design coordinate system to obtain a more accurate reference hole position, it can be considered a major contribution to reducing the defect rate of the work piece. Usually, following the above procedure, the result obtained by using the least square method with the center of gravity of the (observed) guide mark P i (j) as the fixed point, and the second time with the largest and smallest error The second fixed point obtained by the average calculation of the guide mark coordinate 値 of is compared with the result obtained in the least square method, and a good result can be obtained. In accordance with the requirements, the least square method after the second number obtained by averaging calculation can be used repeatedly. In this case, the method adopted is to stop repeating when the error obtained at -39- (34) (34) 592008 is smaller than the maximum error specified for the first time, or when the calculation has been specified after 2 times. The method to stop when the number of times, can also be thought of as having more than two coordinate systems and then choose the best coordinate system. In the calculation after the second time, even if the weighting method of the weighting function used is changed, there will be different results. For example: 1 is assigned to all conductor layers in advance, and when the weighting function of conductor layers with important elements is 2, the error of the pattern formed on the conductor layer is greater than the pattern formed on other conductor layers. Less. [Effects of the Invention] As described above, as described in item 2 of the scope of patent application, the present invention determines the first arrangement of the multilayer printed wiring board by the conventional method of using the center of gravity of the guide mark as a fixed point, and then changes it again. The position of the point is used to calculate the arrangement of the second multilayer substrate. Since the error among them is less, the chance of becoming a good product is increased, and thus it contributes to reducing the defect rate. As mentioned earlier, since the hot-pressing process is continuously performed, even a small reduction in the defect rate can be a reduction in the original cost of the large waste, which greatly helps the production cost. In addition, as described in item 3 of the scope of the patent application, when the deviation of one of the conductor layers of the multilayer wiring board is large, such as when the first guide mark with a larger error is selected to set the second fixed point, This is particularly effective when the deviation of one of the conductor layers of the multilayer wiring board is large. The fact that there is a large deviation in one layer indicates that the overall occupancy rate of the defect is quite large -40- (35) (35) 592008, so it can be prevented from occurring to reduce the defect rate. As mentioned earlier, since the hot-pressing process is continuously performed, even if it is slightly reduced. The defect rate can also be a reduction in the original price of the large waste, which greatly helps the production cost. In addition, in the present invention, as described in item 4 of the scope of patent application, by repeatedly determining the arrangement of the multilayer printed wiring board, a satisfactory result can be obtained. In particular, it has a great effect on reducing the defect rate of high multilayer wiring boards with many layers. _ Furthermore, as described in item 1 of the scope of patent application, to implement the present invention, it is almost unnecessary to change the mechanical structure part of the distributed reference hole drilling machine supplied on the market to improve only the built-in control device Program can cope. It is easy to change the program even after delivery, and it has the effect of improving the performance of commercially available standard hole drilling machines. [Brief Description of the Drawings] Fig. 1 is a perspective view showing the construction pattern of a reference hole drilling machine according to an embodiment of the present invention. Fig. 2 is a front view and a side view showing a construction pattern of a reference hole drilling machine according to an embodiment of the present invention. ^ Fig. 3 is a plan view showing a construction pattern of a reference hole drilling machine according to an embodiment of the present invention. Fig. 4 is a schematic diagram showing the arrangement of guide marks provided on a multilayer wiring board actually used and a reference hole (mark) formed near the guide marks. -41-(36) (36) 592008 Figure 5 is a pattern diagram for the explanation of the principle of the distribution and opening of the majority of holes. Figure 6 is a guide mark and observations set on the conductor layer of the multilayer wiring board. Diagram of the relationship between the guide marks. Fig. 7 is a block diagram of a part of a control device of a reference hole drilling machine and a calculation procedure for designing and comparing a coordinate system position. Fig. 8 is a perspective view for explaining the structure of a multilayer printed wiring board, a plan view of a conductor layer of an inner layer, and a side view of a tool board for a junction layer. ® Figure 9 is a cross-sectional view in the thickness direction of the structure of the multilayer printed wiring board. Fig. 10 is a plan view of a through hole provided in a pattern of a multilayer printed wiring board and a cross-sectional view in the direction of board thickness. [Illustration of drawing number] 1: Drilling machine 2: Housing @ 3: Frame 4: X-ray generating device 4a: X-ray generating tube 5: X-ray protective tube 5a · Hole 6: X-ray camera 7: Mandrel 7 a: Collet-42- (37) (37) 592008 7 b: Hole drill 7 c: Cylinder 8: Mandrel holder 9: Clamping piece 9 a: Cylinder 9b: Clamping piece support 9 c: Clamping screw 10: X-direction moving frame 1 1: Y-direction moving frame 1 2: Movable table 1 0 a, 1 1 a, 1 2 a: linear guide (1 Μ guide) 10b, lib, 12b : Ball screw 1 6: Drilling position (1, 2 holes) 1 6 a: Drilling position (3, 4 holes) 1 7: Operator position (white arrow) 5 0: Mechanical coordinate system (Xm, Ym, Zm , Mechanical origin 0m) PI, P2, P3, P4: (observed) guidance marks (described on the mechanical coordinate system) PD1, PD2, PD3, PD4: (designed) guidance marks (described on Design coordinate system) H1, H2, H3, H4: reference hole α: rotation angle (angle formed by U axis and X axis) 6 0: multilayer printed wiring board 6 1 '6 1 A: double-sided wiring board -43- (38) 592008 6 1 a 6 1 a • Single—pattern 6 2: conductor 6 3: (Insulation) Substrate 6 4: Polyester film 6 4 a, 6 4 a A; Polyester film (with reference hole attached) 6 5 6 5 A% 6 5 B: (For layering) Reference hole 6 6 • Guidance mark 6 7: Reference hole 6 8 6 8 A; Tooling board for bonding 6 8 a 6 8 a A 6 8 a B ·· Positioning bonding pin 7 1 (b C de) 7 2 (b% C de); circle: shape symbol 7 3; round window 1 0 0 1 0 2 (2) ·· pad 1 0 1 1 0 1 a: through hole 1 0 2 1 0 2 a; Through hole center
-44--44-