583462 玖'發明說明 【相關之專利申請案】 本案相關於美國專利申請案編號[RDI-123-3]，標題爲 「具有寬視角之大型、穩健、單塊和類似單塊的AMCLCD 顯示器的建構」，在此同時申請並於此納入參考。 【發明所屬之技術領域】 本發明針對透光平板電光顯示器，其使用低剖面、大 面積準直背光，尤其針對具反射光源的顯示器，諸如於準 直背光中加入螢光燈和/或LED的大型、無接縫、拼接之 主動矩陣液晶顯示器。 【先前技術】 透光平板電光顯示器，諸如在可攜式電腦上的主動矩 陣液晶顯示器（AMLCD)，需要背光。這樣的平板顯示器包 含但不限定在拼接的FPD、單塊和類似單塊的FPD。這樣 的顯示器的標準配置是由一個裝在顯示器面板後面的一薄 背光所構成。背光的主要光學組件一般爲在顯示器側之白 色矩形反射腔穴開口、裝在腔穴中或以一光學波導引連接 至腔穴的螢光型式之燈泡、和光調節結構，諸如裝在顯示 器下之腔穴開口側的散光器和/或隨空間變化之中性密度 濾光片。這樣的背光所產生的光分佈大約爲Lambertian。 透光平板顯示器的一些型式需要比前述標準的背光所 產生的更準直光線。通常在這些情況下，額外的特別目的 7 583462 光調節結構在散光器上加至背光，以產生較高的增益（也就 是較小的Lambertian和較準直之光分佈）。這在顯示器之橫 幅或尺寸增加時尤其真確。 顯示器需要較準直之背光之例子爲授讓給本申請人之 美國專利No.5661531中所描述之拼接之AMLCD。在此例中 ，需要準直光線以可忽略之映像點間的放大且不阻斷拼塊 之間的接縫，把在映像點平面的影像投射至位在短矩離外 的一螢幕。本發明的主要動機爲提供光學上更有效率之背 光給這樣的顯示器。 一準直背光之第二應用例的動機在於克服眾所皆知的 事實，即液晶型式透光顯示器通常會依觀看角度變化而有 不佳的對比。這是由於液晶（LC)材料本身與生倶來之非等 向性。一種準Lambertian背光的替代方案，以其在漸增之 非常態視角下與生倶來不佳的對比而導向穿過液晶顯示器 面板，則爲一更準直的背光，其透過液晶螢幕面板投射至 直接在顯示器面板頂端的顯示螢幕。通過液晶的準直光線 有比準Lambertian光更高的平均對比。螢幕重新分配此更 高的對比影像給在不同觀看角度之觀看者。 一準直背光的另一應用例爲顯示器製造商選擇這樣的 背光是出於美學、功能性和/或效率的理由。在這樣的配 置中，顯示器從正常接近法線的入射角觀看，會相對較明 亮且有更佳的對比。這樣的應用可以包含飛機、軍事、汽 車的顯示器、自動提款機顯示器以及可攜式電腦顯示器。 .爲這樣的顯示器設計和製造具有特殊光線準直結構的 8 583462 背光組件是一件格外有挑戰性的工作。前述應用之理想的 光調節結構應該是製造上的花費不高、厚度薄、重量輕， 而且長寬大約如同在其下的背光一般。從附近之Lambertian 背光光源產生大面積、均勻、大致準直輸出光束之在一側 或兩側的簡易光學結構，位在光線準直結構附近。使用此 光調節結構的理想背光會使用高效能、便宜、半標準、白 色、發散的螢光燈，或是有輸出功率之LED，其在燈數和 尺寸上都可放大。結合的背光組件會有相當接近背光來源 光束功率之輸出光束功率（高效能）。較不理想的準直背光會 犧牲準直的程度、光學功率效能和製造成本，並且使用較 多的特殊燈。這樣的背光系統之光學設計可以從成像和非 成影光學兩種互補原則切入。這兩者以其特殊的光線準直 結構，而與準直背光有關聯。 在成像光學領域中，眾所皆知一個小的物體來源放置 在一簡易的薄凸透鏡的焦距上，會產生準直光束。然而， 對於物體爲背光腔穴而言，這樣配置的透鏡尺寸和距離是 不合實際的，而需要其它的方法。一個能更有效率地從發 散來源獲取光學能量的透鏡傾向於比效率差的透鏡來得大 。透鏡可以設計且製造爲能對與透鏡大約相等大小或比稍 微大一點的來源之影像加以聚焦，但是代價是光學功率效 能降低。微透鏡陣列無法解決這些問題。 在非成像光學領域中，光學設計師使用非影像保留光 學原則，諸如完全內部反射（TIR)、多重反射和折射、以及 在組件之間的光線重循環，以產生導引及重塑來源照明的 9 583462 光學功能，而非嚴格地將物體或來源加以成像。必然要在 光學表現、效率和成本之間的作一困難的抉擇。 基於非成像光學之準直背光組件的一個例子爲描述於 美國專利 No.5739931 之 Allied Signal SpectraVueTM 準直片 。此準直片有薄且長的燈管，有效率地藉由密封的反射鏡 連接至一薄的平面波導的一個或兩個平行邊上。Fresnel透 鏡陣列層合在波導頂部表面上。波導的長寬大約與要照明 的顯示器相同。適當角度的光線藉由Fresnel透鏡陣列準直 化，那些沒有適當角度的則完全在內部反射，直到它們攔 截燈泡和/或反射鏡爲止，在此它們重新進入波導以做另 一次通行。這樣的配置產生了一個非常有效率的區域準直 背光，但是不適用於大的或是多重的發散來源，而且當準 直片面積增加時（亦即輸入功率隨一邊的尺寸線性增加時) 但是需要的功率爲了穩定的輸出功率而隨一邊的平方而增 加，故其擴充性也不好。 非成像光學的另一例，應用於準直背光組件的特殊光 準直結構爲3M光學系統（St. Paul，Minn)的亮度增強膜（BEF) 。這些光準直結構加至傳統的發散器頂端上的背光，以產 生更高的增益。它們是透明的塑膠膜，其在顯示側的表面 有分光槽之線性陣列。顯示側有會反射的表面，且背面側 的表面要就是完全反射，不然就是有些黯淡無光澤。這些 濾鏡的準直動作是基於從特定極角和方位角之入射光束的 小百分比的空氣一槽介面之折射。大約50%的進入光束會 藉由包含雙重完全內部反射和多重反射和折射之反射機制 10 583462 的組合，而重新導回至背光。BEF膜厚度薄、重量輕，且 大量生產時價格比較便宜。它們可以製成大的尺寸。然而 ，爲求系統效率，它們也強烈地依靠背光腔穴重覆循環的 特性。大約它們光線輸出的一半是在相對於薄膜法線之30 度角錐之外。在嚴格需要準直化或大致準直化光線的應用 中，此特性是無法接受的。使用BEF薄膜於此應用的系統 設計者會面臨兩個困難的抉擇：允許更大的角度光線透過 顯示器而損失了光學的效能，或是藉由機械孔洞或其它方 法移除大角度光線而造成無效率的系統。這樣的應用仍需 要大的、薄的、便宜的、高準直且有效率的光線準直方法 〇 另外一個準直背光組件的非成像光學爲美國專利 No.5598281 Zimmerman等人的技術。此背光使用一孔洞陣 列至逐漸縮小的光學元件，後者使用TIR產生部份的準直 而進入一微透鏡陣列，以作更進一步的準直化。這些逐漸 縮減的光學元件有平面的光線輸入及輸出表面。孔洞表面 面積與整體表面面積的比例最好是在20%至30%之間。 【發明內容】 本發明的一個目標是提供一個大的、薄的、具擴充性 的、光學上有效率的、和高度準直的背光。 本發明是基於三個槪念的組合：前兩個槪念以高的背 反射比產生一個新穎的、強力準直化、可透光的光線調節 結構。第三個槪念以發明的光線調節結構，提供作爲一有 11 583462 效率背光組件的製造。 第一個槪念是一高度準直化的光線調節結構是從一個 經過修正的基本折射光發散器（諸如描述於美國專利 No.2378252)所實現。然而本發散器是以反向來操作。也就 是說，進入和出去表面是顛倒的。這個槪念也應用在基本 折射光發散器的變體上，例如於美國專利No.5781344中所 述。這些光調節結構的種類是透明的珠設於大約一珠半徑 厚的不透明的、吸收性的、黑色基底的複合物。此珠與基 底複合物通常是在透明的基質上；這些珠要就是接觸不然 就是快接觸到基板。這些光發散器通常用來作爲投射螢幕 〇 一個來自投射器之最小發散的光束入射在半球形的珠 表面。這些珠以折射方式把投射器光束聚焦透過它們的近 焦點，在焦點之後該光束分散爲大的視角，觀看螢幕因而 有作用。黑色基底作用爲珠的離開瞳孔，也能改善光線準 直結構螢幕的週圍對比。在本修正後之相反的應用，腔穴 提供發散光給接近透鏡元件之近焦點的開孔，隨後離開透 鏡變爲更準直的光束。隨著珠折射率或半球透鏡元件橫跨 塑膠和光學玻璃的範圍，接近緊密配置之珠的有效珠開孔 率大約爲16%±10%。 從相反之基本折射發散器製造的光線準直結構爲一優 良的準直器，然而它本身卻是無效率的，理由有二：第一 ，基底材質對有好的週圍對比的觀看顯示器而言，很明顯 地爲高吸收性且低反射性，多數光線很容易被吸收。第二 12 583462 個理由是光線進入珠之開孔區域大約爲整體面積的5%。因 此，入射光只有一小部份會經由此結構穿過。 本發明依據的第二個槪念爲解決前述的第一個問題， 此藉由用高反射、低吸收率、白色材質或結構，取代上述 折射光準直結構之高吸收率、低反射性基底，以致於無法 交會珠的進入開孔的光會被反射回到背光腔穴，而非被吸 收或消失。 然而，基本折射光準直結構螢幕之不透明、吸收性、 黑色基底的材料特性不同於本發明的光準直結構之高反射 、低吸收性、白色基底的特性。在實務應用上，珠化螢幕 光準直結構黑色基底在厚度數十微米時會有所需的特性， 且較薄切面相當地透明，然而白色基底只在厚度大約爲三 分之一毫米至幾毫米時才會有所需的特性。這些材料特性 的限制意謂本發明之珠或半球形透鏡的半徑，會大於或等 於腔穴出口板上之反射結構的厚度。一額外的設計元件爲 開孔的側壁爲非高吸收性。大多數阻斷於側壁的光束都沒 被吸收，而是被反射或是散射。 本發明所依據的第三個槪念解決了第二個前述提及的 無效率，其藉由使用本發明的光線準直結構，以內含之高 度反射光源，作爲一個實質上不漏光、低吸收、白色、高 反射腔穴的腔穴出口板。本發明使用一開孔陣列，其並非 顯著地依賴TIR或平面輸入或輸出表面。較佳的開孔率大 約是16%±10%。本發明的背光組件包含一實質上不漏光、 高度反射、低吸收腔穴，其包含高度反射的光源。一個包 13 583462 含光源的低損失腔穴對一有效率、準直化、背光組件而言 是相當重要的。從來源產生的光束有許多機會交會於一珠 的進入瞳孔。那些交會的便被有效率地準直化，那些沒有 交會的則被重覆循環直到交會爲止。一由這樣的腔穴和出 口板光準直結構所構成的背光不僅僅是一優良的準直器， 也是一個高效率準直背光。 【實施方式】 本發明爲一具有高反射、實質上密封、薄的矩形光腔 穴的大型、高效能、高功率之準直背光組件，腔穴包含高 反射表面和一個或多個光源。腔穴的大型表面其中之一爲 一光出口板，其包含一可透光的光準直結構。準直結構的 腔穴側包含一高反射白色的平面，其包含具有最小側壁吸 收的圓形開孔陣列。開孔是位在光學軸心的中心，其位置 接近於位在準直結構外表面之緊密配置之半球形或球形透 鏡陣列之焦距。光束在腔穴的高反射表面、光源和開孔壁 之間捕獲，直到它們進入透鏡爲止，其輸出進入一準直光 束之大部份光束。特定的參數使背光效率和準直化所需的 角度最佳化：具光源之腔穴的有效反射率和幾何、開孔面 積和透鏡元件剖面積的比例、開孔幾何和吸收率、以及透 鏡元件之折射率和幾何。背光組件用於透光型電光顯示器 上是有利的，尤其是那些效能是藉由有效率之準直光線所 增強的，諸姐無接縫拼接之液晶顯示器。 參考圖la，所顯示的依據本發明的一背光組件，圖lb 14 583462 所顯不的爲側視圖。背光組件1由三個主要的次組件所構 成：具有透鏡陣列3的腔穴出口板光準直結構2、光源4、 和背光腔穴5(圖lb)。 同光組件1的另一實施例顯示在圖1 c的上視圖中’以 及圖1 d中的側視圖。在此配置中，多個小光源4排列在背 光組件1的背面。 參考圖2a，較佳實施例（圖ia和lb)之背光腔穴5的內 部腔穴表面形成一高反射、具有大面積之實質上不漏光之 腔穴。它是薄的且實質上爲矩形。背光腔穴5之內部表面 的主要光學組件爲側面8和9、上部6、底部7和背部10。 同時也顯示光源4的非照射端用的孔11。 參考圖2b，另一實施例（圖lc和Id)的背光腔穴5的內 部腔穴表面也會形成有大面積之高反射、實質上不漏光之 腔穴。它也是薄的且實質上爲矩形。背光腔穴5的內部表 面的主要光學組件爲側邊8和9、上部6、底部7和背部 10。一倂顯示的爲用於光源4之機械安裝或電供應通過的 孔U。 多種材料可以用於內部腔穴表面，以達90%-99%的有 效全腔穴反射率（亦即發散加上反射）。這造成光波長的可見 光譜中（即白光）實質上相同的或全反射。這樣的反射材料的 一般例子包含SpectralonTM材質和披覆Teflon™的鋁、多層 聚合物薄膜、和在鋁上披覆特殊的磷光劑混合物。更高的 材質反射性和不漏光的腔穴會產生更有效率的背光。 腔穴5的幾何性質也有助於背光效率。在其它條件相 15 583462 同的情況下，更薄的背光腔穴會產生更有效率的背光。在 光源4周圍的腔穴5的形狀也是重要的。在燈和腔穴的非 出口表面區域之間捕獲光的形狀，會比從這樣的區域由大 而小照入光線至腔穴出口表面的形狀，產生較無效率的背 光。上部、底部和側面的形狀不需要如圖2所示的爲矩形 或平坦面，因爲斜的或曲面形狀可以提供有效率的優勢。 一般而言，腔穴的表面積應該最小化。這意謂以曲面將側 面、上部、底部和背部融合在一起，而非垂直的接點。垂 直的接點也容易在接點處捕獲光線，而非導引光線通過腔 穴出口板光準直結構2。也因爲同樣的原因，垂直接點並非 最有效率的。 高反射、高效率的光源4包含線性管狀的螢光燈管， 其有諸如T5的直徑。這樣的光源顯示於圖la和圖lb。這 樣的燈管的全反射性在95%至97%之間。燈管4裝在腔穴 5，並以非照明端區域在腔穴5外，而通過孔11突出經過 腔穴5的側面8和9，以及燈管4和孔11周圍之間有不漏 光的接合。 、在另一實施例中（圖lc、Id和2b)，光源4可爲多重小 光源，諸如僅以安裝孔或是提供電力給光源的穿孔而裝在 腔穴5的LED。 再次參考圖la和lb，腔穴5的前側爲出口板光準直結 構2。圖3顯示具有透鏡陣歹[J 3、透鏡直徑12、開孔陣列 13、開孔直徑14、開孔深度20、開孔側壁21、腔穴出口板 16、高反射腔穴側結構15、接著劑17、光學輸入側18、光 16 583462 學輸出側19的一個腔穴出口板光準直結構2的較佳實施例 之一剖面的側視圖。就是這整個結構產生本發明有效率的 準直作用。 參考圖3，腔穴出口板光準直結構2有四個光學上重要 的機能性組件：高反射腔穴側結構15(例如 質和披覆Teflon™的鋁、多層聚合物薄膜、以及在鋁上披 覆特殊的磷光劑混合物）、開孔陣列13、透鏡陣列3、以及 黑色、光吸收之輸出表面16。前三者是所有較佳實施例都 包含的必要光學元件。第四個爲一增強作用，用來減少來 自輸出側上透鏡一空氣介面的反射及空氣一顯示器反射所 造成之大角度背景光線，從10%的等級減少到1%的等級。 透鏡陣列3包含實質上爲球形透鏡或半球形（平凸）透鏡 或不規則形狀的透鏡之緊密配置的陣列，以構成開孔的功 能，其配置在腔穴出口板光準直結構2的光學輸出側19上 。開孔陣列13的圓形開孔位在光學軸心的中央，且接近開 孔出口板光準直結構2之光學輸入側18上透鏡之焦矩。必 須了解的是開孔13未必要圓形，也未必有開孔的實體結構 ，如底下會討論到的。透鏡直徑12實際上比最小的開孔直 徑14大。開孔陣列13的內部表面，在高反射腔穴側結構 15和透鏡陣列3的進入瞳孔之間，不會吸收每一開孔上光 入射的實質百分比。這可能是因爲開孔深度20是小的，且 開孔側壁21沒有阻斷入射於開孔入口之光的實質百分比。 另外可選擇的，開孔側壁21實質上是非吸收性的。所有較 佳實施例都包含這些必要光學元件間的幾何及結構關係。 17 583462 基於上述針對腔穴出口板光準直結構2的原則，可以 想像更多的較佳實施例。 在圖3的較佳實施例中，高反射腔穴側結構15顯示爲 提供透鏡陣列之結構支撐的不透明腔穴出口板16上的反射 媒介。透鏡藉由接著劑17固定在正確的位置。腔穴出口板 光準直結構2之高反射腔穴側結構15可以是與用在腔穴其 它側相同的高反射材料，或是爲了厚度、製造成本或其它 理由，而是不同的高反射材料。可以藉由像在腔穴出口板 16上塗料或陽極電鍍表面的黑色塗層，而在透鏡之間和之 下的光學輸出側19上製造黑色的吸收表面，而腔穴出口板 16本身是由像黑色塑膠的黑色吸收材料所製造，或是替代 地，接著劑17也可以是黑色的且吸收光線的。 在圖4的替代性實施例中，腔穴出口板16本身是由高 反射材料所製成，其提供放置透鏡的結構支撐，舉例而言 ，材料爲Spectralon™。開孔陣列13直接作於板16中，形 成腔穴出口開孔板。一黑色吸收表面可以藉由像在腔穴出 口板16上塗料的黑色塗層，而製作在透鏡之間和之下的光 學輸出側19上，或是替代地，接著劑Π也可以是黑色的 且吸收光線的。 參考圖5，在另一實施例中，腔穴出口板16可以是透 明的平面基板，諸如玻璃或塑膠，在腔穴出口板光準直結 構2的光學輸入側上沒有開孔。高反射腔穴側結構是由在 反射箔片23上的反射材質22的組合物所達成。開孔是在 反射組合物中。一黑色吸收表面可以藉由像在反射箔片23 18 583462 上塗料的黑色塗層，製作在透鏡之間和之下的光學輸出側 19上，或是替代地，接著劑Π也可以是黑色的且吸收光線 的。 更進一步的實施例顯示在圖6中。腔穴出口板16再一 次的爲透明的平面基板，諸如玻璃或塑膠，在腔穴出口板 光準直結構2的光學輸入側上沒有實體開孔。一薄層的光 學接著劑Π固定透鏡在腔穴出口板16上。一環形物在每 一透鏡3的最外周形成，其提供必須的光傳輸特性，其與 顯示於前述實施例（圖3至圖5)中的相同。高反射腔穴側結 構15是由光學接著劑17之光學輸出側19上的一層反射媒 體22所達成。有效的開孔直徑14有兩個組件：在光學接 著劑至透鏡的圓形淸楚的開孔、以及來自反射媒體中光線 的平均穿透深度的環形物。一黑色吸收表面可以藉由像在 反射媒體22上塗料的黑色塗層，.製作在透鏡之間和之下的 光學輸出側19上。 必須注意的是，在圖3、4、5所顯示的實施例中，光 學接著劑17可以出現在開孔中，而不顯著影響腔穴出口板 光學準直結構2的效能。在圖16、17、18和19中顯示的 不同的結構配置，舉例而言，接著劑不需要用來固定透鏡 至腔穴出口開孔板。一鉗夾板29(圖16)或鉗夾板29與襯墊 30的組合（圖17)可以用來實體上固定透鏡24在板16上。 此外，鉗夾板29也可以置於透鏡24的末端周圍（圖18) ’ 實體上限制透鏡24於所需的位置。在另一實施例中（圖19) ，擠製物31可以形成作爲透鏡24的一部份，並藉由鉗夾 19 583462 板29固定位置。一黑色吸收表面可以藉由像在腔穴出口板 16和鉗夾板29上塗料的合適黑色塗層，製作在透鏡之間 和之下的光學輸出側19上。 開孔陣列13的開孔可以是簡單的直孔，或是有複雜的 幾何，例如：a)有大側在光學輸入側18之由大變小的孔， 並且小端形成透鏡進入瞳孔；b)—個從腔穴出口板16之光 學輸出側19到珠進入瞳孔的球切面形狀；或c)光學輸入側 18和一透鏡進入瞳孔之間具最窄直徑的一不規則的沙漏形 狀。開孔可以簡單地如同在珠下的一薄的出口腔穴塗層， 乃至於相較於非常高反射性所需厚度的出口腔穴板之進入 表面，如描繪於圖6。開孔可以是簡單的直孔。透鏡陣列3 的透鏡的光學輸出側19可以有額外的光學媒體薄層，以修 正準直動作，或採用腔穴出口板光準直結構2的反射特性 〇 參考圖7，所顯示的是結構2的替代性實施例，在其中 描繪一混合鏡頭24’，其爲透鏡陣列（沒有顯示出來）的其中 一員，而接近開孔13，其也爲開孔陣列（沒有顯示出來）的 其中一員。透鏡24’的整體厚度或直徑描述於圖中爲尺寸d 。透鏡24’的輸入側爲一凸透鏡，其半徑爲R2，並位在接 近開孔13處。透鏡24’的輸出側爲一半球形，其半徑爲R1 。一黑色吸收表面可以藉由像在腔穴出口板16或透鏡陣列 24之非透鏡表面上塗料的合適黑色塗層，製作在透鏡之間 和之下的光學輸出側19上。 圖8顯示兩個發散的輸入光束25入射在腔穴出口板光 20 583462 準直結構2上。一光束是入射在開孔24上，另一則入射在 高反射腔穴側結構15上。圖9顯示高反射腔穴側結構15 將入射發散的輸入光束25轉換成輸出發散之反射光束26 ° 圖1〇顯示透鏡24將入射在開孔24之發散之輸入光束25 轉換至準直的輸出光束27。 多種折射率可以用在透鏡上，此取決於開孔的幾何特 性和所需的準直度。準直度是由增益對極角Θ 28(圖8)所指 定。增益是在極角爲㊀時開孔上穿透光對入射光的比例。 圖11顯示對一代表性開孔幾何，其增益對三個可能的透鏡 折射率。 圖12顯示光入射於開孔上而以角度小於或等於極角Θ 的之輸出比例。相同於圖11的透鏡折射率和開孔幾何用於 圖12。 在描述過背光組件和元件之次組件的結構和功能以及 它們較佳的實施例後，腔穴、準直動作以及背光效率的最 佳化將描述於下。 腔穴的效率可以視爲傳送至開孔的輸入光功率的比例 。爲了說明起見，假設所有傳送至開孔的光束皆消失至腔 穴，並由透鏡所輸出，不然就是消失。使用上述之腔穴和 開孔，腔穴的效率被建立成一個腔穴出口板光準直結構的 開孔率X的函數f。稱此函數爲f(x)。開孔率X爲開孔陣列 13的開孔之整體斷面積與整體透鏡陣列3所橫跨之矩形區 域之整體斷面積之比例。 函數f(x)爲流至開孔的光流量除以輸入至腔穴的光功 21 583462 率。f(x)的一個簡單有用的模型爲f(x)=x/((C/P)+ α *(1· χ)+χ)，其中C爲一參數，取決於腔穴之吸收率和幾何特性 和光源之吸收率和幾何特性；Ρ爲腔穴出口板光準直結構2 的透鏡陣列3所橫跨之整體面積；而α爲高反射腔穴側結 構15的吸收率。 函數f(x)以0起始：f(0)=0，單調地增加至某漸近線。f 之準確函數形式對本發明並非關鍵，但是腔穴效率是隨開 孔率而實質單調地增加。選擇包含不同之C、P、α之參數 會產生不同的曲線組。最佳化的腔穴爲一在材料、組件、 成本和幾何限制上具有最高效率者。圖13爲參數C、Ρ、 α之不同代表性値的f(x)函數圖。 背光最佳化流程的第二個部份有關於從在腔穴出口板 光準直結構2之開孔入口上的發散光束，至透鏡元件之輸 出側上一更準直光束的傳輸函數。設h(x)爲來自透鏡而在 所需之準直錐中的光輸出除以進入開孔的光功率。再一次 ，開孔率爲X，函數h(x)爲準直效率。設計參數包含折射率 和詳細的表面特性和開孔的結構的幾何特性，以及所有來 自開孔入口的表面，直到光束離開透鏡最後的光學表面。 函數h(x)爲一 S或乙形的曲線，其h(0)接近百分之百，而 當X增大時單調地減至一漸近線。函數h(x)可以藉由此技 術專精的專家所熟知的光追蹤方法，爲特定的設計所決定 並最佳化。h(x)之準確函數形式對本發明並非關鍵，但是準 直效率是隨開孔率單調地減少。最佳化的開孔/透鏡系統爲 一在材料、組件、成本和幾何限制上具有最高效率者。 22 583462 圖14顯示代表性開孔幾何和透鏡折射率的三個可能的 選擇之準直效率h(x)與開孔率X之關係。 背光最佳化過程的第三個和最後一部份爲結合腔穴效 率f(x)和準直效率h(x)，成爲背光準直效率g(x)。設g(x)爲 f(x)*h(x)的乘積。函數g(x)爲進入目標準直錐之光輸出除以 腔穴輸入光功率。由於f(x)的單調增加的特性和h(x)的單調 減少的特性，g(x)在開孔率X的某些値有最大値。假設此X 値爲x_opt，並假設其代表腔穴和準直效率兩者之獨立最佳 化，如同應用條件所限制的。 圖15顯示圖11、12和13之相同代表性開孔幾何和透 鏡折射率的背光準直效率g(x)與開孔率x的關係。 至此所描述的實施例的透鏡和開孔都爲圓形橫斷面。 很顯然地，所描述的觀念、結構和方法對其它幾何形狀也 是適用的，諸如正方形、棒狀、狹縫、金字塔狀…等。在 特定的應用中，這樣的幾何形狀可能是更好的，而包含於 本發明。只要大多數或全部的入射光是透過開孔陣列傳送 而造成準直，則開孔和透鏡的幾何形狀是無關緊要的。類 似地，本發明不會被在此描述之有或無接著劑之特定的安 裝配置所限制。 由於其它修正和改變以符合特定之操作需求和環境， 對那些習知此技術的人是相當顯而易見的，本發明不應限 於被選擇用作揭示的例子，而應涵蓋所有不悖離本發明真 正的精神和範圍的改變和修正。 在如此地描述本發明後，需要受到專利證書保護的將 23 583462 呈現在所附的申請專利範圍中。 【圖式簡單說明】 (一）圖式部分 要完全了解本發明可以藉由參考附圖並加上上述的詳 細說明而獲得，其中： 圖la爲本發明之背光組件的上視圖，其顯示具有透鏡 陣列及光源之非準直斷面的腔穴出口板光準直結構； 圖lb爲圖la之背光組件的側視圖，其顯示具有透鏡 陣列、光源之非準直斷面、及背光腔穴的腔穴出口板光準 直結構； 圖lc爲一用其它替代光源之背光組件的上視圖； 圖Id爲圖lc之背光組件的側視圖，其顯示具有透鏡 陣列、光源之非準直斷面、和背光腔穴的腔穴出口板光準 直結構； 圖2a爲在圖la中所描繪之背光腔穴之內部表面的上 透視圖，其中不包含出口板光準直結構和光源； 圖2b爲在圖lc中所描繪之背光腔穴之內部表面的上 透視圖，其中不包含出口板光準直結構和光源； 圖3、4、5、6爲腔穴出口板光準直結構的多個實施例 之斷面的各別側視圖； 圖7爲具有近似開孔陣列之半球/凸透混合透鏡陣列之 腔穴出口板光準直結構之另一實施例之側視圖； 圖8爲腔穴出口板光準直結構的一剖面之側視圖，其 24 583462 顯示光入射在濾鏡的兩個不同區域上； 圖9爲腔穴出口板光準直結構之一剖面的側視圖，其 顯示從光準直結構的表面之光入射和反射； 圖10爲腔穴出口板光準直結構之一剖面的側視圖，其 顯示光入射在開孔上以及光準直結構的透鏡之準直動作； 圖11爲極角與增益之座標圖； 圖12極角與效能之座標圖； 圖13爲開孔率與腔穴效能的座標圖； 圖14爲開孔率與準直效能之座標圖； 圖15爲開孔率與背光效能之座標圖； 圖16、17、18、19爲具有將透鏡附加在一腔穴出口開 孔板上之機械固定方法的腔穴出口板光準直結構之多個實 施例的剖面的各別側視圖。 爲了簡潔淸楚起見，在所有的圖形中，類似的元件和 組件將會使用相同的名稱和編號。 (二）元件代表符號 1. 背光組件 2. 腔穴出口板光準直結構 3. 透鏡陣列 4. 光源 5. 背光腔穴 6. 上部 7. 底部 8.側邊 25 583462 9. 側邊 10. 背部 11. 孔 12. 透鏡直徑 13. 開孔陣列 14. 開孔直徑 15. 高反射腔穴側結構 16. 腔穴出口板 17. 接著劑 18. 光學輸入側 19. 光學輸出側 20. 開孔深度 21. 開孔側壁 22. 反射材質 23. 反射箔片 24. 透鏡 2 4 混合透 25. 發散的輸入光束 26. 發散的反射光束 27. 準直的輸出光束 28. 極角 29. 鉗夾板 30. 襯墊 31. 擠製物 26583462 发明 'Invention Description [Related Patent Application] This case is related to US patent application number [RDI-123-3], entitled "Construction of Large, Robust, Monolithic and Monolithic AMCLCD Displays with Wide Viewing Angle ", Apply here at the same time and incorporate it here for reference. [Technical field to which the invention belongs] The present invention is directed to a translucent flat-panel electro-optic display, which uses a low-profile, large-area collimated backlight, especially for a display with a reflective light source, such as a fluorescent lamp and / or LED added to the collimated backlight. Large, seamless, spliced active matrix LCD display. [Prior art] Light-transmissive flat-panel electro-optic displays, such as active matrix liquid crystal displays (AMLCDs) on portable computers, require a backlight. Such flat panel displays include, but are not limited to, spliced FPDs, monoliths, and similar monolithic FPDs. The standard configuration of such a display consists of a thin backlight mounted behind the display panel. The main optical components of the backlight are generally white rectangular reflective cavity openings on the display side, fluorescent-type bulbs installed in the cavity or connected to the cavity with an optical wave guide, and light adjustment structures, such as installed under the display A diffuser on the open side of the cavity and / or a spatially varying neutral density filter. The light distribution produced by such a backlight is approximately Lambertian. Some versions of translucent flat panel displays require more collimated light than the aforementioned standard backlight. Often in these cases, an additional special purpose 7 583462 light adjustment structure is added to the backlight on the diffuser to produce a higher gain (ie, a smaller Lambertian and a more collimated light distribution). This is especially true when the width or size of the display increases. An example of a display that requires a more collimated backlight is the spliced AMLCD described in U.S. Patent No. 5,615,531, assigned to the applicant. In this example, it is necessary to collimate the light with negligible magnification between the image points without blocking the joints between the tiles, and project the image in the plane of the image point to a screen located at a short moment away. The main motivation of the present invention is to provide such a display with an optically more efficient backlight. The motivation for the second application example of a collimated backlight is to overcome the well-known fact that liquid crystal type transmissive displays usually have poor contrast depending on the viewing angle change. This is due to the anisotropy of the liquid crystal (LC) material itself and the nature. An alternative to the quasi-Lambertian backlight, which is guided through the LCD panel with its poor contrast with the increasing abnormal viewing angle, is a more collimated backlight that is projected through the LCD panel to Display screen directly at the top of the display panel. Collimated light passing through the liquid crystal has a higher average contrast than quasi-Lambertian light. The screen redistributes this higher contrast image to viewers at different viewing angles. Another example of a collimated backlight is for display manufacturers to choose such a backlight for aesthetic, functional, and / or efficiency reasons. In such a configuration, the display is relatively brighter and has better contrast when viewed from an angle of incidence normally close to the normal. Such applications can include aircraft, military, automotive displays, cash dispenser displays, and portable computer displays. Designing and manufacturing 8 583462 backlight assemblies with special light collimation structures for such displays is a particularly challenging task. The ideal light adjustment structure for the aforementioned application should be low in manufacturing cost, thin in thickness, light in weight, and approximately as long as the backlight underneath it. A simple, optical structure that produces a large area, uniform, and approximately collimated output beam from one nearby Lambertian backlight source is located near the light collimation structure. The ideal backlight using this light adjustment structure will use high-efficiency, cheap, semi-standard, white, divergent fluorescent lamps, or LEDs with output power, which can be amplified in both the number of lamps and the size. The combined backlight assembly will have output beam power (high efficiency) that is quite close to the beam power of the backlight source. Less desirable collimation backlights sacrifice the degree of collimation, optical power efficiency and manufacturing costs, and use more special lamps. The optical design of such a backlight system can be cut from two complementary principles of imaging and non-image forming optics. These two are related to the collimated backlight with their special light collimation structure. In the field of imaging optics, it is well known that a small object source placed at the focal length of a simple thin convex lens will produce a collimated beam. However, for objects with backlight cavities, the lens size and distance configured in this way are impractical, and other methods are required. A lens that can more efficiently obtain optical energy from a divergent source tends to be larger than a less efficient lens. Lenses can be designed and manufactured to focus an image from a source about the same size or slightly larger than the lens, but at the cost of reduced optical power efficiency. Microlens arrays cannot solve these problems. In the field of non-imaging optics, optical designers use non-image-preserving optical principles such as total internal reflection (TIR), multiple reflections and refraction, and the recirculation of light between components to produce a guide that reshapes the source of illumination. 9 583462 Optical function rather than strictly imaging an object or source. A difficult choice must be made between optical performance, efficiency, and cost. An example of a non-imaging optical-based collimated backlight assembly is the Allied Signal SpectraVueTM collimator described in US Patent No. 5,739,931. This collimator has a thin and long tube, which is efficiently connected to one or two parallel sides of a thin planar waveguide by a sealed mirror. Fresnel lens arrays are laminated on the top surface of the waveguide. The length and width of the waveguide are approximately the same as the display to be illuminated. Light at the proper angle is collimated by the Fresnel lens array, and those without the proper angle are completely reflected internally until they intercept the bulb and / or mirror, where they re-enter the waveguide for another pass. This configuration produces a very efficient area-collimated backlight, but is not suitable for large or multiple sources of divergence, and when the area of the collimator increases (that is, when the input power increases linearly with the size of one side), but The required power increases with the square of one side for stable output power, so its expandability is also not good. Another example of non-imaging optics is a special light used for collimating backlight components. The collimating structure is a 3M optical system (St. Paul, Minn) brightness enhancement film (BEF). These light collimation structures are added to the backlight on top of a traditional diffuser to produce higher gain. They are transparent plastic films with linear arrays of spectroscopic grooves on the surface of the display side. There is a reflective surface on the display side, and if the surface on the back side is completely reflective, otherwise it is a bit dull. The collimation action of these filters is based on the refraction of a small percentage of the air-slot interface from incident light beams at specific polar and azimuth angles. Approximately 50% of the incoming light beam is redirected back to the backlight by a combination of reflection mechanisms 10 583462 including double total internal reflection and multiple reflections and refraction. BEF film is thin and light, and it is cheaper in mass production. They can be made in large sizes. However, for system efficiency, they also strongly rely on the nature of the backlight cavity's repeated cycles. About half of their light output is outside a 30-degree pyramid with respect to the film's normal. This feature is unacceptable in applications where strict or substantially collimated light is required. System designers using BEF films for this application will face two difficult choices: allowing greater angles of light to pass through the display and losing optical performance, or removing large angles of light by mechanical apertures or other methods that cause Efficient system. Such applications still require large, thin, inexpensive, highly collimated and efficient light collimation methods. Another non-imaging optic for collimating backlight components is the technology of US Patent No. 5,559,281 Zimmerman et al. This backlight uses a hole array to shrink optical elements, which uses TIR-generated collimation to enter a microlens array for further collimation. These tapered optics have flat light input and output surfaces. The ratio of the hole surface area to the total surface area is preferably between 20% and 30%. SUMMARY OF THE INVENTION An object of the present invention is to provide a large, thin, expandable, optically efficient, and highly collimated backlight. The present invention is based on a combination of three intents: the first two intents produce a novel, powerful collimating, light-transmitting light-regulating structure with a high back reflection ratio. The third idea is the invention of a light-adjusting structure that provides the manufacturing of a backlight assembly with 11 583462 efficiency. The first idea is that a highly collimated light-regulating structure is implemented from a modified basic refracting light diffuser (such as described in U.S. Patent No. 2378252). However, this diffuser operates in the reverse direction. In other words, entering and exiting are reversed. This idea is also applied to variations of the basic refracting light diffuser, such as described in U.S. Patent No. 5,781,344. The type of these light-regulating structures is a composite of transparent beads on an opaque, absorptive, black substrate with a thick bead radius. This bead and substrate complex is usually on a transparent substrate; if these beads are in contact, they will be in rapid contact with the substrate. These light diffusers are usually used as projection screens. A minimal divergent light beam from the projector is incident on the surface of a hemispherical bead. These beads refractively focus the beam of the projector through their near focus. After the focus, the beam disperses into a large viewing angle, which is useful for viewing the screen. The black base acts as the bead's exit from the pupil, and can also improve the contrast of the surroundings of the light collimation screen. In the reverse application after this correction, the cavity provides divergent light to an opening near the focal point of the lens element, and then leaves the lens to become a more collimated beam. As the refractive index of a bead or hemispherical lens element spans the range of plastic and optical glass, the effective bead openness of a close-packed bead is approximately 16% ± 10%. The light collimation structure made from the opposite basic refracting diffuser is an excellent collimator, but it is inefficient for two reasons. First, the base material is good for viewing displays with good surrounding contrast. Obviously, it is highly absorbing and low reflecting, and most light is easily absorbed. The second 12 583462 reason is that the area where the light enters the opening of the bead is about 5% of the total area. Therefore, only a small part of the incident light passes through this structure. The second idea on which the present invention is based is to solve the aforementioned first problem, by replacing the high-absorptance and low-reflective substrate of the refracted light collimation structure with a high reflection, low absorption, white material or structure , So that the light entering the opening that cannot meet the beads will be reflected back to the backlight cavity instead of being absorbed or disappeared. However, the opaque, absorptive, and black-based material characteristics of the screen of the basic refracted light collimation structure are different from those of the high-reflective, low-absorptive, and white-based properties of the light collimation structure of the present invention. In practical applications, the black substrate of the pearlized screen light collimation structure will have the required characteristics when the thickness is tens of microns, and the thin section is quite transparent, but the white substrate only has a thickness of about one third to several millimeters The required characteristics are only available in millimeters. These material property limitations mean that the radius of the bead or hemispherical lens of the present invention will be greater than or equal to the thickness of the reflective structure on the cavity exit plate. An additional design element is that the side walls of the openings are non-absorbent. Most of the light beams blocked by the side walls are not absorbed, but are reflected or scattered. The third idea on which the present invention is based solves the second inefficiency mentioned above. By using the light collimating structure of the present invention, the light source with a high degree of reflection is used as a light leak-proof, low Cavity exit plate for absorbing, white, highly reflective cavity. The present invention uses an array of apertures, which does not rely significantly on TIR or planar input or output surfaces. The preferred opening ratio is about 16% ± 10%. The backlight assembly of the present invention includes a substantially leak-free, highly reflective, low-absorption cavity that includes a highly reflective light source. A package 13 583462 low-loss cavity with a light source is important for an efficient, collimated, backlight assembly. The beam from the source has many opportunities to intersect with a bead entering the pupil. Those who meet will be efficiently collimated, while those who do not meet will be re-circulated until they meet. A backlight composed of such a cavity and exit plate light collimation structure is not only an excellent collimator, but also a high-efficiency collimation backlight. [Embodiment] The present invention is a large, high-performance, high-power collimated backlight assembly with a highly reflective, substantially sealed, thin rectangular light cavity. The cavity includes a highly reflective surface and one or more light sources. One of the large surfaces of the cavity is a light exit plate that includes a light-transmissive light collimation structure. The cavity side of the collimation structure includes a highly reflective white plane that includes a circular array of apertures with minimal sidewall absorption. The opening is located at the center of the optical axis, and its position is close to the focal length of a closely arranged hemispherical or spherical lens array located on the outer surface of the collimating structure. The beam is captured between the highly reflective surface of the cavity, the light source, and the opening wall until they enter the lens, and its output enters most of the beam of a collimated beam. Specific parameters optimize backlight efficiency and the angle required for collimation: effective reflectivity and geometry of the cavity with the light source, the ratio of the aperture area to the cross-sectional area of the lens element, the aperture geometry and the absorptivity, and the lens Refractive index and geometry of the device. The backlight assembly is advantageous for use in transmissive electro-optic displays, especially those LCD displays in which the efficiency is enhanced by efficient collimated light and the joints are spliced seamlessly. Referring to FIG. 1a, a backlight assembly according to the present invention is shown. FIG. Lb 14 583462 shows a side view. The backlight assembly 1 is composed of three main sub-assemblies: a cavity exit plate light collimation structure 2 with a lens array 3, a light source 4, and a backlight cavity 5 (Figure lb). Another embodiment of the same light assembly 1 is shown in a top view 'in FIG. 1c and a side view in FIG. 1d. In this configuration, a plurality of small light sources 4 are arranged on the back of the backlight module 1. Referring to Fig. 2a, the inner cavity surface of the backlight cavity 5 of the preferred embodiment (Figs. Ia and lb) forms a highly reflective cavity having a large area and substantially light-tight. It is thin and essentially rectangular. The main optical components of the inner surface of the backlight cavity 5 are the sides 8 and 9, the upper part 6, the bottom part 7, and the back part 10. A hole 11 for the non-irradiated end of the light source 4 is also shown. Referring to FIG. 2b, the cavity surface of the backlight cavity 5 of another embodiment (FIGs. Lc and Id) also forms a cavity with a large area, high reflection, and substantially no light leakage. It is also thin and essentially rectangular. The main optical components of the internal surface of the backlight cavity 5 are the sides 8 and 9, the upper part 6, the bottom part 7 and the back part 10. A hole U is shown for the mechanical installation or electrical supply of the light source 4. A variety of materials can be used on the interior cavity surface to achieve effective full-cavity reflectivity (ie, divergence plus reflection) of 90% -99%. This causes the visible wavelength of the light spectrum (ie, white light) to be substantially the same or totally reflected. Common examples of such reflective materials include SpectralonTM materials and Teflon ™ -coated aluminum, multilayer polymer films, and special phosphor blends on aluminum. Higher material reflectivity and non-leakage cavities result in more efficient backlighting. The geometry of the cavity 5 also contributes to backlight efficiency. Under the same conditions as other 583462, a thinner backlight cavity will produce a more efficient backlight. The shape of the cavity 5 around the light source 4 is also important. The shape of the light captured between the lamp and the non-exit surface area of the cavity will produce a less efficient back light than the shape from which light is radiated from the area to the cavity exit surface. The shape of the top, bottom, and sides does not need to be rectangular or flat as shown in Figure 2, because oblique or curved shapes can provide efficient advantages. In general, the surface area of the cavity should be minimized. This means that the sides, upper, bottom, and back are fused together with curved surfaces, rather than vertical joints. Vertical contacts are also easier to capture light at the contacts, rather than directing light through the cavity exit plate light collimation structure 2. For the same reason, vertical contacts are not the most efficient. The highly reflective, highly efficient light source 4 includes a linear tubular fluorescent tube having a diameter such as T5. Such a light source is shown in FIGS. 1a and 1b. The total reflectivity of such a lamp is between 95% and 97%. The lamp tube 4 is installed in the cavity 5 and is outside the cavity 5 with a non-illuminated end area, and the sides 8 and 9 protruding through the cavity 5 through the hole 11 and there is no light leakage between the lamp tube 4 and the periphery of the hole 11 Join. 2. In another embodiment (Figures lc, Id, and 2b), the light source 4 may be multiple small light sources, such as an LED installed in the cavity 5 only with a mounting hole or a perforation that provides power to the light source. Referring again to Figures la and lb, the front side of cavity 5 is the exit plate light collimation structure 2. Figure 3 shows a lens array [J 3, lens diameter 12, aperture array 13, aperture diameter 14, aperture depth 20, aperture sidewall 21, cavity exit plate 16, highly reflective cavity side structure 15, and A side view of a cross section of one of the preferred embodiments of a cavity exit plate light collimation structure 2 of the agent 17, the optical input side 18, the light 16 583462 and the academic output side 19. It is this entire structure that produces the efficient collimation effect of the present invention. Referring to Figure 3, the cavity exit plate light collimation structure 2 has four optically important functional components: a highly reflective cavity-side structure 15 (such as Teflon ™ -coated aluminum, multilayer polymer film, and aluminum Covered with a special phosphor mixture), aperture array 13, lens array 3, and black, light-absorbing output surface 16. The first three are necessary optical components included in all preferred embodiments. The fourth is an enhancement effect, which is used to reduce the large angle background light caused by the reflection from the lens-air interface on the output side and the reflection from the air-monitor from 10% to 1%. The lens array 3 includes a compactly arranged array of substantially spherical lenses or hemispherical (plano-convex) lenses or irregularly shaped lenses to form a function of openings, which are arranged in the optics of the light exiting structure 2 of the cavity exit plate On the output side 19. The circular apertures of the aperture array 13 are located at the center of the optical axis, and are close to the focal moment of the lens on the optical input side 18 of the light collimation structure 2 of the aperture exit plate. It must be understood that the openings 13 do not have to be circular, nor do they necessarily have a solid structure of openings, as will be discussed below. The lens diameter 12 is actually larger than the smallest aperture diameter 14. The internal surface of the aperture array 13 between the highly reflective cavity-side structure 15 and the entrance pupil of the lens array 3 will not absorb a substantial percentage of the light incident on each aperture. This may be because the opening depth 20 is small and the opening sidewall 21 does not block a substantial percentage of the light incident on the entrance of the opening. Alternatively, the opening sidewall 21 is substantially non-absorbent. All preferred embodiments include geometric and structural relationships between these necessary optical components. 17 583462 Based on the above-mentioned principle for the light collimation structure 2 of the cavity exit plate, more preferred embodiments can be imagined. In the preferred embodiment of Fig. 3, the highly reflective cavity-side structure 15 is shown as a reflective medium on the opaque cavity exit plate 16 that provides structural support for the lens array. The lens is fixed in the correct position by the adhesive 17. Cavity exit plate light collimation structure 2 The high-reflection cavity-side structure 15 may be the same highly reflective material used on the other side of the cavity, or it may be a different highly reflective material for thickness, manufacturing cost, or other reasons . A black absorbing surface can be made on the optical output side 19 between and under the lens by a black coating such as paint or anodized surface on the cavity exit plate 16, and the cavity exit plate 16 itself is made of The adhesive 17 may be made of a black absorbing material like black plastic, or alternatively, the adhesive 17 may be black and absorb light. In the alternative embodiment of FIG. 4, the cavity exit plate 16 itself is made of a highly reflective material, which provides structural support for placing the lens, for example, the material is Spectralon ™. The perforated array 13 is formed directly in the plate 16 to form a cavity exit perforated plate. A black absorbing surface can be made on the optical output side 19 between and under the lens by a black coating like paint on the cavity exit plate 16, or alternatively, the adhesive Π can also be black And absorb light. Referring to FIG. 5, in another embodiment, the cavity exit plate 16 may be a transparent planar substrate such as glass or plastic, and there is no opening on the optical input side of the cavity exit plate light collimation structure 2. The highly reflective cavity side structure is achieved by a composition of a reflective material 22 on a reflective foil 23. The openings are in a reflective composition. A black absorbing surface can be made on the optical output side 19 between and under the lens by a black coating like coating on reflective foil 23 18 583462, or alternatively, the adhesive Π can also be black And absorb light. A further embodiment is shown in FIG. 6. The cavity exit plate 16 is again a transparent flat substrate, such as glass or plastic, and there is no physical opening on the optical input side of the light collimation structure 2 of the cavity exit plate. A thin layer of optical adhesive II holds the lens on the cavity exit plate 16. A ring is formed on the outermost periphery of each lens 3, which provides the necessary light transmission characteristics, which are the same as those shown in the foregoing embodiments (FIGS. 3 to 5). The highly reflective cavity-side structure 15 is achieved by a layer of reflective medium 22 on the optical output side 19 of the optical adhesive 17. The effective aperture diameter 14 has two components: a circularly shaped aperture in the optical contact to the lens, and a ring from the average penetration depth of light from the reflective medium. A black absorbing surface can be formed on the optical output side 19 between and under the lens by a black coating like a coating on the reflective medium 22. It must be noted that, in the embodiment shown in Figs. 3, 4, and 5, the optical adhesive 17 may appear in the opening without significantly affecting the efficiency of the optical collimation structure 2 of the cavity exit plate. The different structural configurations shown in Figs. 16, 17, 18, and 19, for example, do not require an adhesive to secure the lens to the cavity exit aperture plate. A jaw plate 29 (Fig. 16) or a combination of the jaw plate 29 and the spacer 30 (Fig. 17) can be used to physically fix the lens 24 to the plate 16. In addition, the jaw plate 29 can also be placed around the end of the lens 24 (Fig. 18) 'to physically limit the lens 24 to a desired position. In another embodiment (FIG. 19), the extruded article 31 may be formed as a part of the lens 24 and fixed in position by a clamp 19 583462 plate 29. A black absorbing surface can be made on the optical output side 19 between and under the lens by a suitable black coating like coating on the cavity exit plate 16 and jaw plate 29. The openings of the opening array 13 may be simple straight holes or have complicated geometries, for example: a) there are holes with large sides on the optical input side 18 that change from large to small, and the small end forming a lens into the pupil; b ) A spherical cut shape from the optical output side 19 of the cavity exit plate 16 to the bead entering the pupil; or c) an irregular hourglass shape with the narrowest diameter between the optical input side 18 and a lens entering the pupil. The opening can be as simple as a thin exit cavity coating under the bead, or even the entrance surface of the exit cavity plate with a thickness required for very high reflectivity, as depicted in Figure 6. The opening can be a simple straight hole. The optical output side 19 of the lens of the lens array 3 may have an additional thin layer of optical media to correct the collimation action, or the reflection characteristics of the light collimation structure 2 using the cavity exit plate. Referring to FIG. 7, the structure 2 is shown An alternative embodiment of the invention depicts a hybrid lens 24 ', which is one of the lens arrays (not shown), and close to the opening 13, which is also one of the opening arrays (not shown). The overall thickness or diameter of the lens 24 'is depicted in the figure as dimension d. The input side of the lens 24 'is a convex lens with a radius of R2 and located near the opening 13. The output side of the lens 24 'is hemispherical, and its radius is R1. A black absorbing surface can be made on the optical output side 19 between and under the lens by a suitable black coating like coating on the non-lens surface of the cavity exit plate 16 or lens array 24. FIG. 8 shows that two divergent input light beams 25 are incident on the cavity exit plate light 20 583462 collimation structure 2. One beam is incident on the opening 24 and the other is incident on the highly reflective cavity-side structure 15. Figure 9 shows a highly reflective cavity-side structure 15 that converts an incident divergent input beam 25 into an output divergent reflected beam 26 °. Figure 10 shows a lens 24 that converts a divergent input beam 25 incident on the opening 24 to a collimated output. Light beam 27. Various refractive indices can be used on the lens, depending on the geometry of the opening and the required collimation. Collimation is specified by the gain-to-polar angle Θ 28 (Figure 8). Gain is the ratio of penetrating light to incident light at the aperture when the polar angle is ㊀. Figure 11 shows the gain versus the refractive index of three possible lenses for a representative aperture geometry. FIG. 12 shows the output ratio of light incident on the opening at an angle less than or equal to the polar angle Θ. The same lens refractive index and aperture geometry as in FIG. 11 are used in FIG. After describing the structure and function of the backlight assembly and its sub-assemblies and their preferred embodiments, the optimization of the cavity, collimation action, and backlight efficiency will be described below. The efficiency of the cavity can be viewed as the ratio of the input optical power delivered to the opening. For the sake of explanation, it is assumed that all the light beams transmitted to the opening disappear into the cavity and are output by the lens, otherwise they disappear. Using the cavity and opening described above, the efficiency of the cavity is established as a function f of the aperture ratio X of the light collimation structure of the cavity exit plate. Call this function f (x). The aperture ratio X is the ratio of the entire cross-sectional area of the openings in the aperture array 13 to the entire cross-sectional area of the rectangular area spanned by the overall lens array 3. The function f (x) is the light flow rate to the opening divided by the light work rate input to the cavity 21 583462. A simple and useful model of f (x) is f (x) = x / ((C / P) + α * (1 · χ) + χ), where C is a parameter that depends on the absorption rate of the cavity and Geometrical characteristics and absorptivity and geometrical characteristics of the light source; P is the entire area spanned by the lens array 3 of the light collimation structure 2 of the cavity exit plate; and α is the absorptivity of the highly reflective cavity-side structure 15. The function f (x) starts at 0: f (0) = 0 and monotonically increases to some asymptote. The exact function form of f is not critical to the invention, but the cavity efficiency increases substantially monotonically with the porosity. Selecting different parameters including C, P, and α will result in different curve groups. The optimized cavity is the one with the highest efficiency in terms of materials, components, cost, and geometry. FIG. 13 is a graph of f (x) functions of different representative values of parameters C, P, and α. The second part of the backlight optimization process concerns the transfer function from the divergent beam at the entrance of the aperture of the light collimation structure 2 of the cavity exit plate to a more collimated beam on the output side of the lens element. Let h (x) be the light output from the lens in the required collimation cone divided by the optical power entering the aperture. Once again, the aperture ratio is X and the function h (x) is the collimation efficiency. The design parameters include the refractive index and detailed surface characteristics and the geometrical characteristics of the opening structure, as well as all the surfaces from the entrance of the opening until the beam leaves the last optical surface of the lens. The function h (x) is an S or B-shaped curve, h (0) is close to 100%, and decreases monotonically to an asymptote as X increases. The function h (x) can be determined and optimized for a particular design using light tracing methods that are well known to experts in this technology. The exact functional form of h (x) is not critical to the invention, but the alignment efficiency decreases monotonically with the porosity. The optimized aperture / lens system is the one with the highest efficiency in terms of materials, components, cost, and geometry. 22 583462 Figure 14 shows the relationship between the collimation efficiency h (x) and the aperture ratio X for three possible choices of representative aperture geometry and lens refractive index. The third and last part of the backlight optimization process is to combine the cavity efficiency f (x) and the collimation efficiency h (x) to become the backlight collimation efficiency g (x). Let g (x) be the product of f (x) * h (x). The function g (x) is the light output entering the standard straight cone divided by the cavity input optical power. Due to the characteristic of monotonically increasing f (x) and the characteristic of monotonically decreasing h (x), g (x) has the largest 値 at some 値 of the aperture ratio X. Assume that X 値 is x_opt and assume that it represents independent optimization of both cavity and collimation efficiency, as limited by application conditions. Fig. 15 shows the relationship between the backlight collimation efficiency g (x) and the aperture ratio x for the same representative aperture geometry and lens refractive index of Figs. 11, 12, and 13. The lenses and apertures of the embodiments described so far have a circular cross-section. Obviously, the described concepts, structures, and methods are also applicable to other geometric shapes, such as squares, rods, slits, pyramids, etc. In certain applications, such geometries may be better and are included in the present invention. As long as most or all of the incident light is transmitted through the array of apertures to cause collimation, the geometry of the apertures and the lens is irrelevant. Similarly, the present invention is not limited by the particular installation configuration described herein with or without an adhesive. Due to other modifications and changes to meet specific operating needs and circumstances, it will be quite obvious to those skilled in the art that the present invention should not be limited to the examples chosen for disclosure, but should cover all without departing from the true nature of the invention Changes and amendments to the spirit and scope. After describing the invention in this way, 23 583462, which needs to be protected by a patent certificate, is presented in the scope of the attached patent application. [Brief description of the drawings] (1) The schematic part of the invention can be fully understood by referring to the drawings and adding the above detailed description, in which: Figure la is a top view of a backlight assembly of the present invention, which shows Lens array and light collimation structure of cavity exit plate of non-collimated section of light source; Figure lb is a side view of the backlight assembly of Figure la, showing a lens array, non-collimated section of light source, and backlight cavity Light collimation structure of the cavity exit plate; Figure lc is a top view of a backlight assembly using other alternative light sources; Figure Id is a side view of the backlight assembly of Figure lc, which shows a non-collimated section with a lens array and a light source And the cavity exit plate light collimation structure of the backlight cavity; Figure 2a is an upper perspective view of the internal surface of the backlight cavity depicted in Figure la, which does not include the exit plate light collimation structure and light source; Figure 2b It is a top perspective view of the internal surface of the backlight cavity depicted in Figure lc, which does not include the light collimation structure and light source of the exit plate; Figures 3, 4, 5, and 6 show the light collimation structure of the exit plate Sectional differences Side view; Figure 7 is a side view of another embodiment of a light collimation structure of a cavity exit plate with a hemispherical / convex hybrid lens array having an approximately aperture array; Figure 8 is a light collimation structure of a cavity exit plate A side view of the section, 24 583462, showing light incident on two different areas of the filter; Figure 9 is a side view of a section of a light collimating structure of a cavity exit plate, showing light from the surface of the light collimating structure Incident and reflection; Figure 10 is a side view of a cross section of a light collimation structure of a cavity exit plate, showing the collimation action of light incident on an opening and the lens of the light collimation structure; Figure 11 is the polar angle and gain Coordinates; Figure 12: Polar angle and efficiency; Figure 13: Coordinates of aperture ratio and cavity performance; Figure 14: Coordinates of aperture ratio and collimation performance; Figure 15: Porosity and backlight efficiency Coordinate diagrams; Figures 16, 17, 18, 19 are respective sides of the cross-section of a plurality of embodiments of a cavity exit plate light collimation structure having a mechanical fixing method for attaching a lens to a cavity exit opening plate view. For the sake of brevity, in all figures, similar elements and components will use the same name and number. (II) Symbols for components 1. Backlight assembly 2. Cavity exit plate light collimation structure 3. Lens array 4. Light source 5. Backlight cavity 6. Top 7. Bottom 8. Side 25 583462 9. Side 10. Back 11. Hole 12. Lens diameter 13. Opening hole array 14. Opening hole diameter 15. High reflection cavity side structure 16. Cavity exit plate 17. Adhesive 18. Optical input side 19. Optical output side 20. Opening hole Depth 21. Opening sidewall 22. Reflective material 23. Reflective foil 24. Lens 2 4 Mixed through 25. Divergent input beam 26. Divergent reflected beam 27. Collimated output beam 28. Polar angle 29. Clamp plate 30 Gasket 31. extruded material 26