200928327 九、發明說明 【發明所屬之技術領域】 本發明，係有關於一種，對於構成 LED ( Light Emitting Diode)或 LD ( laser diode)等之發光元件晶片 ，就算是在每一晶片中具有相異之配光特性，亦能夠將其 發光特性迅速且高精確度地作特定之測光裝置。 D 【先前技術】 在LED等之製造廠商中，作爲最終檢查的其中一環 ，係有必要對各發光元件晶片之光量（quantity of light) 作計測。作爲光量値，例如，係要求有光度（Luminous intensity )。於此，光度Ιχ之燭光（candela )値，係爲 相對於光通量（Luminous flux) Φ γ之流明（lumen)値 ，其每一單位立體角（1Ω之比例，而爲Ir = <1φ 7/οΙΩ。 另一方面，光通量Φ r ，係爲將放射通量<x>e( ❹ radiant flux )，以 CIE ( International Commission on Illumination )所規定之標準分光視感效率（Spectral luminous efficiency) V ( 7 )作加權而積分者，亦考慮最 大視感效果度（Maximum luminous efficacy) Km，而如同 下式一般地被定義。另外，所謂放射通量Φ e ( radiant flux )，係爲光源所放射至空間之功率〔W〕，亦即是， 爲每一單位時間之放射能量〔J/s〕。 200928327 〔數式1〕 Φτ^ΚιηίΓΦε") ·ν(λ)(1(λ) 爲了計測此種光量，在受訊發光元件晶片之放射光的 位置處，配置光二極體。光二極體，係經由視感度濾鏡而 接收受訊光。根據從光二極體而來之起電流値，而求取出 ^ 光通量（Luminous flux)値或光度（Luminous Intensity )値。於此，視感度濾鏡，係爲對應於人類之視感度，而 將CIE檫準分光視感效率之特性V ( λ )作再現的濾鏡。 光二極體，由於係接收有此種通過了視感度濾鏡後之 光，因此，根據該光二極體之起電流値，能夠求取出光通 量之流明値，或是光度ir之燭光値。亦即是，係成 爲只要對光二極體之輸出，將光量以既知之其他的主光源 來校正，或是對同一發光源而以其他之標準測定器來作測 Ο 定，並根據該測定値來對光二極體之輸出作修正即可。 然而，藉由上述方法所得到之檢測値，係極大的依存 於視感度濾鏡之特性。特別是，由於視感度濾鏡之特性’ 在藍綠波長域亦多會有所偏差，因此，會有在檢測値之精 確度上缺乏信賴性的缺點。又，從光二極體所產生之電流 ，雖會對應於受訊光之波長而變化，但是，在上述之方法 中，係亦存在著完全沒有考慮到對於此受訊光之波長的感 度之變化的缺點。 又，更進而’對於具備有相對於放射方向而放射強度 -6- 200928327 爲不均勻之配光特性的發光元件晶片，係有無法將其之光 量作特定之致命的問題點。此係因爲，發光元件晶片之配 光特性，就算是從相同之半導體經原所切出之晶片，亦會 在每一個晶片中存在有差異，因此，就算是再怎麼樣的將 _ 測定對象之晶片作精密的位置決定，亦無法得到正確的計 '測値之故。 於此，雖然可以將光二極體之受光面積增大，但是， Φ 在大面積之受光面中，由於係無法對從點光源而來之放射 光作最適當的受光（受光面與射入光不會直交），因此， 測定誤差亦會隨之增加。 又，在先前技術之裝置中，就算是能夠對朝向上方之 放射光作檢查，針對朝向下方之放射光，亦有完全無法檢 測出其之強度的問題。例如，在先前技術之裝置中，對於 預定作爲覆晶封裝之發光元件，係完全無法對應。 本發明，係有鑑於上述之問題點而進行者，其目的， 〇 係在於提供一種：不需要使用視感度濾鏡，又，就算是對 於具有不均勻之配光特性的發光元件，亦可將其之光量迅 速且高精確度的作測定之測光裝置。又，係以提供一種亦 可檢測出全方向之配光特性的測光裝置爲目的。 【發明內容】 爲了達成上述目的，本發明，係爲一種對具備有對於 放射方向而放射強度爲不均勻之配光特性的發光元件之光 量作自動計測的測光裝置，其特徵爲，具備有：複數之r 200928327 個的受光部，係受訊前述發光元件之發光，並輸 所受訊之受訊光的受訊強度之檢測資料11〜In ; 或是複數個的分光分析部，係受訊前述發光元件 並針對所受訊之受訊光，而輸出將受訊強度之相 ' 定之波長間隔來作特定的頻譜分佈資料Ρ(λ) ' 部，係將把對於在前述受光部中之受訊光的波長 以前述特定之波長間隔而作特定的分光感度資料 0 )〜PDn (λ)，對應於η個的受光部而各別作 控制部，係從前述受光部、前述分光分析部、以 憶部，而接收所需要之資料，並實行演算動作， 部，係具備有以下之處理：第1處理，係根據η 部所輸出之檢測資料Π〜In、和前述記憶部所記1 的分光感度資料PD1 ( λ )〜PDn ( λ )、和前 析部所輸出之頻譜分佈資料Ρ(λ) ’而對η個 之每一者，分別計算出前述發光元件之放射能量 〇 佈EG1(A)〜EGn(A):和第2處理，係根 1處理所計算出之頻譜分佈EG1(A)〜EDn ( 計算出在η個的受訊部處之放射通量EG1〜EGn 處理，係根據藉由第1處理所計算出之頻譜分佈 )〜EGn ( λ )、和代表相對於光之波長λ的人 度的分光視感效率ν( λ )，而計算出在η個的 之光通量Φ i。 在本發明中之受訊部的個數η，雖係因應於 之配光特性的不均勻度而被決定’但是’不論是 出對應於 和1個又 之發光， 對値以特 ;和記憶 之感度， PD1 ( λ 記憶；和 及前述記 前述控制 個的受光 隱之η個 述分光分 的受光部 的頻譜分 據藉由第 λ )，而 :和第3 EG1 ( λ 間之視感 受光部處 發光元件 在何種情 -8- 200928327 況下，均係爲複數個。故而，能夠將受訊部之受光面積設 定爲最適當之大小，藉由使受光面與射入光直交，而提昇 測定之精確度。又，η個的受訊部，較理想，係爲在從發 光元件而離開有相同之距離的位置處被規則地配置爲理想 ' 。在最適當的情況，η個的受訊部，係以將發光元件藉由 球面又或是半球面而包覆的方式來配置。 在本發明之第1處理中，較理想，係作爲第i個的受 Q 光部處之頻譜分佈EGi (λ)，而實行式（A)之演算。 EGi ( λ ) =ΙίχΡ ( λ ) /〔 Σ (Ρ (入）xPDi ( λ ))〕……式（A) 於此，輸出頻譜分佈資料P ( λ )之分光分析部，係 以近接於η個的受光部而設置複數個爲理想。但是，由於 就算是在發光源爲具備有不均勻之配光特性的情況時，亦 多有在頻譜分佈資料Ρ ( λ )的分佈形狀中不產生有變化 〇 的情形，因此，在此種情況時，分光分析部係亦可爲單一 個。. 在本發明之第3處理中，較理想，係作爲第i個的受 光部處之光通量Φί’而實行式（B)之演算。 Φ i= Kmx [ Σ ( V ( λ ) xEGi ( λ ))〕......式（Β ) 如此這般，針對所算出之η個的光通量値Φ1〜Φη’ 若是將此些作綜合之評價’則就算是針對具備有非均勻之 -9- 200928327 配光特性的發光元件，亦能夠更正確地對光度作 型來說，係採用積分處理或總和處理。 另外，檢查對象之發光元件，雖然亦可在將 搬送線的狀態下而直接作檢查，但是，較理想， ' 配置在專用之檢查台而作檢查。此時，係以將檢 ' 發光元件，以半球面又或是球面狀來作包圍的方 置η個的受光部爲理想。 Q 又，若是將檢查台以透光性之材料來構成， 於全方向（360度）而對光量作檢查。又，若是 以半透光性之材料來構成，則由於在通過檢查台 ，從點光源而來之放射光係被變換爲散亂光，因 是大面積之受光面，亦能夠對光量正確地作把握 若藉由以上所說明之本發明，則不需要使用 鏡，且，就算是對於具有不均勻之配光特性的發 亦可將其之光量迅速且高精確度的作測定。又， Q 透光性之檢查台，便可檢測出全方向之配光特性 【實施方式】 以下，根據實施例，對本發明作更進一步之 。圖1(a)，係爲展示實施例之測光裝置EQU 成之電路區塊圖。在此測光裝置EQU中，檢查 光元件晶片CH，係被載置於半透明玻璃製之檢1 並被點燈驅動。而，測光裝置EQU，係自動計算 性並輸出。 特定。典 其配置於 係應將其 查對象之 式，而配 則能夠對 將檢查台 之過程中 此，就算 〇 視感度濾 光元件， 僅需使用 詳細說明 的槪略構 對象之發 台ΕΧ上 其發光特 -10- 200928327 發光元件晶片，例如，係爲構成發光二極體之半導體 晶片。而，將半導體晶圓切斷所切出之多數的發光元件晶 片CH…CH，係被保持於黏著薄片SE上並作待機（圖3) 。而，機械手，係將待機狀態之發光元件晶片CH —個一 個的取出，並將其載置於檢查台EX上。 - 發光元件晶片，其放射光之放射強度相對於放射方向 係具備有不均勻之配光特性（參考圖4)，且，其配光特 ❹ 性，在各個發光元件晶片中係並不爲相同。 但是，在此測光裝置EQU中，係藉由設置複數之η 個的計測點，並將η個的計測値作總合評價，而對於配光 特性爲相異之各個的發光元件晶片CH，將其發光特性正 確地作特定。具體而言，此測光裝置EQU，係針對各個的 發光元件晶片CH，將放射通量（Radiant flux) 〔W:瓦 特〕、和被作了視感度修正之光通量（Luminous flux)〔 lm:流明〕、和光度（Luminous intensity) 〔Cd:燭光 φ 〕自動作計算並輸出。 圖2，係爲將測光裝置EQU之計測點作圖示者。如圖 2(a)之立體圖與圖2(b)之平面圖所示一般，在此實施 例中，於檢查台ΕΧ之上方，8個的光二極體PD1〜PD8， ' 係以將發光元件晶片CH以半球面狀來覆蓋的方式而被配 置。具體而言，8個的光二極體PD1〜PD8，係在離開發 光元件晶片CH相等距離R之位置而被配置。4個的光二 極體PD1〜PD4，係近接於通過發光元件晶片CH之鉛直 線（Ζ軸），而以90°之間隔而被配置爲同心圓狀。在此 -11 - 200928327 光二極體PD1〜PD4之外側，4個的光二極體PD5〜PD8， 係以90°之間隔而被配置爲同心圓狀。 另外，在此實施例中，發光元件晶片CH之配光特性 的不均勻度，由於係較爲輕微（參考圖4)，因此係使用 有8個的光二極體，但是，不用說，因應於配光特性之不 ' 均勻度，會使配置個數η增加。 如圖2(c)之正面圖所示，在檢查台ΕΧ之鉛直下方 0 ，係被配置有受光面大的光二極體PD9。在本實施例中， 由於檢查台ΕΧ係藉由半透明玻璃（毛玻璃）所構成，因 此，從發光元件CH而來之放射光，係在檢查台ΕΧ處成 爲散亂光，並到達光二極體PD9。故而，光二極體PD9， 係成爲接收直交於受光面之散亂光，而使得受光面積較大 一事並不會成爲問題。另外，放射光之通過檢查台ΕΧ時 所致的衰減量，由於係可預先藉由實驗而作把握，因此係 經由計測動作時之修正演算而被修正。 〇 接下來，對測光裝置EQU之電路構成作說明。如圖1 (a )所示一般，測光裝置EQU，係以對發光元件晶片CH 作點燈控制之發光驅動部DR、和9個的光二極體PD 1〜 PD9、和輸出比例於光二極體PD1〜PD9之起電流的電壓 — 之9個的PD放大器1…1、和將從發光元件晶片CH而來 之放射光的分光頻譜正確地作求取之分光計2、和接收從 PD放大器1以及分光計2而來之輸出’並計算出各種的 測光値之演算部3、和對裝置各部之動作適宜作控制的機 器控制部4爲中心而構成。 200928327 演算部3，典型來說，係由個人電腦所構成，並作爲 測光裝置EQU之中央控制部而起作用。而，藉由演算部3 所計算出之測光値，係依序被記憶在內藏又或是連接於演 算部4之記憶部中，同時，被輸出至顯示輸出部DISP。 於實施例之情況中，光二極體PD1〜PD8，係全部具 備有直徑l〇mm左右之受光面。但是，由於殼體等之限制 ，其實效面積，作爲半徑7 = 4之圓，，係成爲S=r2X 0 π = 50.27〔mm2〕。另一方面，發光元件晶片CH與光二 極體PD之距離R，係被設定爲42mm左右。因此，對於 各光二極體PD1〜PD8之立體角，係成爲約72xtt/(RxR )=0.02 8。另外’光二極體PD9之受光面積，係被設定 爲其他之光二極體PD1〜PD8的受光面積之10倍〜20倍 左右。 如圖1(b)之原理圖所示一般，9個的PD放大器1 ，係分別由OP放大器6、和η個的回饋電阻Ri ( R1〜Rn Φ )、和負載電阻RL、和η個的開關元件Si ( S 1〜Sn )所 構成。於此’開關元件Si，其中之任一者係經由機器控制 部4而被設爲ON狀態，其餘的則係成爲OFF狀態。〇p 放大器6之反轉輸入端子與非反轉輸入端子係成爲假想短 路之狀態’ OP放大器6之輸入阻抗係成爲幾乎無限大。 故而’光二極體PD之起電流，係成爲全部流動至回饋電 阻Ri°PD放大器1之輸出電壓Vo，係比例於光二極體 PD之起電流，而成爲Vo=-RixI。另外，回饋電阻Ri( R1〜Rn)之電阻値’由於係因應於在pD放大器1處所要 -13- ❹200928327 IX. Description of the Invention [Technical Field] The present invention relates to a light-emitting element wafer constituting an LED (Light Emitting Diode) or an LD (laser diode), even if it is different in each wafer The light distribution characteristics can also be used as a specific photometric device with rapid and high-accuracy illumination characteristics. D [Prior Art] In the manufacturer of LEDs and the like, as one of the final inspections, it is necessary to measure the amount of light of each of the light-emitting element wafers. As the amount of light 値, for example, Luminous intensity is required. Here, the candela 値 is a lumen 相对 with respect to Luminous flux Φ γ, which is a solid angle per unit (a ratio of 1 Ω, and Ir = < 1 φ 7/ οΙΩ. On the other hand, the luminous flux Φ r is the standard spectral spectroscopic efficiency (Spectral luminous efficiency) V (the radiant flux), which is defined by CIE (International Commission on Illumination). 7) For weighting and integrating, the maximum luminous efficacy Km is also considered, and is generally defined as the following formula. In addition, the so-called radiant flux Φ e (radiant flux) is emitted by the light source. The power of space [W], that is, the radiated energy per unit time [J/s]. 200928327 [Expression 1] Φτ^ΚιηίΓΦε") ·ν(λ)(1(λ) In order to measure this The amount of light is such that a photodiode is disposed at a position where the light emitted from the light-emitting element chip is received. The photodiode receives the received light through the viscous filter. The current is obtained from the photodiode. Remove ^ Luminous flux (Lum Inos flux) Luminous Intensity 値 Here, the viscous filter is a filter that reproduces the characteristic V ( λ ) of the CIE 檫 spectroscopic visual efficiency corresponding to the human visual sensitivity. Since the light diode receives such light that has passed through the opacity filter, it is possible to extract the lumen of the luminous flux or the illuminance of the illuminance of the candle according to the current 値 of the photodiode. Yes, as long as the output of the photodiode is corrected, the amount of light is corrected by another known main light source, or the same light source is measured by another standard measuring device, and the measurement is performed according to the measurement. The output of the photodiode can be corrected. However, the detection enthalpy obtained by the above method is greatly dependent on the characteristics of the viscous filter. In particular, due to the characteristics of the viscous filter 'in the blue-green wavelength region There are also some deviations, so there is a disadvantage of lack of reliability in detecting the accuracy of the flaw. Moreover, the current generated from the photodiode varies depending on the wavelength of the received light, but Above In the middle, there is also a disadvantage that the sensitivity of the wavelength of the received light is not considered at all. Further, the light distribution characteristic of the radiation intensity -6-200928327 is relatively uniform with respect to the radiation direction. The light-emitting element wafer has a problem that it is impossible to make the amount of light of the light-emitting element specific. This is because the light distribution characteristics of the light-emitting element wafer, even if the wafer is cut out from the same semiconductor, will be There is a difference in one wafer. Therefore, even if the wafer of the measurement target is determined to be a precise position, the correct measurement cannot be obtained. In this case, although the light receiving area of the photodiode can be increased, Φ can not receive the most appropriate light (light receiving surface and incident light) from the point light source in the large-area light receiving surface. It will not be straightforward), so the measurement error will also increase. Further, in the prior art device, even if it is possible to inspect the emitted light toward the upper side, there is a problem that the intensity of the emitted light toward the lower side cannot be detected at all. For example, in the prior art device, it is completely impossible to correspond to a light-emitting element that is intended to be a flip-chip package. The present invention has been made in view of the above problems, and an object thereof is to provide a method in which a luminosity filter is not required, and even a light-emitting element having uneven light distribution characteristics can be used. A photometric device for measuring light with rapid and high precision. Further, it is an object of the invention to provide a photometric device that can detect omnidirectional light distribution characteristics. In order to achieve the above object, the present invention is a photometric device that automatically measures the amount of light of a light-emitting element having a light distribution characteristic in which the radiation intensity is uneven in the radial direction, and is characterized in that: The light-receiving part of 200928327 is the light-emitting part of the light-emitting element, and the detection data of the received light of the received light is 11~In; or a plurality of spectroscopic analysis parts are received. The light-emitting element outputs a specific spectral distribution data λ(λ)′ portion for the received signal light, and outputs a wavelength interval of the received signal intensity, which is to be received by the light-receiving portion. The wavelength of the light is made to have a specific spectral sensitivity data 0) to PDn (λ) at a predetermined wavelength interval, and is used as a control unit for each of the n light receiving portions, and is configured by the light receiving unit and the spectroscopic analysis unit. In order to receive the required data and perform the calculation operation, the department has the following processing: the first processing is based on the detection data output from the η unit In~In, and the memory unit is recorded. The spectral sensitivity data PD1 ( λ ) to PDn ( λ ) and the spectral distribution data Ρ (λ) ' output by the pre-analysis unit, and for each of η, calculate the radiation energy of the light-emitting element EG1 (A) to EGn (A): and the second processing, the spectral distribution EG1 (A) to EDN calculated by the root 1 processing (calculated the processing of the radiation fluxes EG1 to EGn at the n receiving portions, The luminous flux at n is calculated based on the spectral distribution calculated by the first processing, EGn ( λ ), and the spectral sensibility ν( λ ) representing the humanity with respect to the wavelength λ of the light. Φ i. In the present invention, the number η of the signal receiving portion is determined in accordance with the unevenness of the light distribution characteristics, but 'there is a corresponding light and a light, and the memory is Sensitivity, PD1 (λ memory; and the spectrum of the light-receiving portion of the η-segment light-splitting portion of the aforementioned control group by the λ), and the third EG1 (the SPECT between the λ1) In the case of the illuminating elements in the case of -8-200928327, the number of light-receiving elements can be set to the most appropriate size, by making the light-receiving surface and the incident light orthogonal. The accuracy of the measurement is improved. Further, the n receiving portions are preferably arranged to be ideally arranged at a position separated from the light emitting element by the same distance. In the most appropriate case, n The signal receiving unit is disposed such that the light emitting element is covered by a spherical surface or a hemispherical surface. In the first processing of the present invention, it is preferably used as the spectrum of the i-th Q-ray-receiving portion. Spread EGi (λ) and perform the calculus of equation (A). EGi ( λ ) =Ιί Ρ ( λ ) / [ Σ (Ρ (in) xPDi ( λ ))] (A) Here, the spectroscopic analysis unit that outputs the spectral distribution data P ( λ ) is connected to the n light receiving portions. It is desirable to set a plurality of numbers. However, even in the case where the light source has uneven light distribution characteristics, there is a case where there is no change in the distribution shape of the spectrum distribution data (λ). Therefore, in this case, the spectroscopic analysis unit may be a single one. In the third processing of the present invention, it is preferable to carry out the equation (B) as the luminous flux Φί' at the i-th light-receiving portion. Φ i = Kmx [ Σ ( V ( λ ) xEGi ( λ ))] (Formula (Β) In this way, for the calculated η luminous fluxes 値 Φ1 Φ Φ η ' For the comprehensive evaluation, even for the illuminating elements with non-uniform -9-200928327 grading characteristics, it is possible to use the integral processing or the sum processing for the luminosity modeling more correctly. The light-emitting element can be directly inspected while the transport line is being used, but It is preferable to perform the inspection on a dedicated inspection table. In this case, it is preferable to detect the light-emitting elements, which are surrounded by a hemispherical surface or a spherical shape. If the inspection table is made of a light transmissive material, the amount of light is inspected in all directions (360 degrees). If it is made of a semi-translucent material, it passes through the inspection table from the point source. The emitted light system is converted into scattered light, and because of the large-area light-receiving surface, the amount of light can be accurately grasped. According to the invention as described above, it is not necessary to use a mirror, and even if it has The uneven light distribution characteristics can also be measured quickly and with high precision. Further, the Q light transmissive inspection table can detect the omnidirectional light distribution characteristics. [Embodiment] Hereinafter, the present invention will be further developed based on the embodiments. Fig. 1(a) is a circuit block diagram showing the photometric device EQU of the embodiment. In this photometric device EQU, the optical element wafer CH is inspected and placed on a test made of translucent glass and driven by lighting. However, the photometric device EQU is automatically calculated and output. specific. The code should be configured in the system, and the matching can be used in the process of checking the table. Even if the sensation filter element is used, only the detailed description of the object should be used. Luminous Special-10-200928327 A light-emitting element wafer is, for example, a semiconductor wafer constituting a light-emitting diode. On the other hand, the plurality of light-emitting element chips CH...CH cut by the semiconductor wafer are held on the adhesive sheet SE and standby (Fig. 3). On the other hand, the robot takes out the light-emitting element chips CH in the standby state one by one and mounts them on the inspection table EX. - A light-emitting element wafer in which the radiation intensity of the emitted light has uneven light distribution characteristics with respect to the radiation direction (refer to FIG. 4), and the light distribution characteristics are not the same in each of the light-emitting element wafers. . However, in the photometric device EQU, by setting a plurality of n measurement points and n-measurement measurement as a total evaluation, the light-emitting element wafer CH having the different light distribution characteristics will be Its luminescent properties are correctly specified. Specifically, the photometric device EQU uses a radiation flux (W: watt) and a luminosity corrected Luminous flux (lm: lumen) for each of the light-emitting element wafers CH. And Luminous intensity [Cd: Candle φ] is automatically calculated and output. Fig. 2 is a diagram showing the measurement points of the photometric device EQU. As shown in the perspective view of Fig. 2(a) and the plan view of Fig. 2(b), in this embodiment, above the inspection table, eight photodiodes PD1 to PD8, 'to be used for the light-emitting element wafer The CH is arranged in such a manner as to cover the hemispherical shape. Specifically, the eight photodiodes PD1 to PD8 are arranged at a position equal to the distance R from the developed optical element wafer CH. The four photodiodes PD1 to PD4 are arranged close to the vertical line (the x-axis) passing through the light-emitting element wafer CH, and are arranged concentrically at intervals of 90°. Here, on the outside of the photodiodes PD1 to PD4, the four photodiodes PD5 to PD8 are arranged concentrically at intervals of 90°. In addition, in this embodiment, the unevenness of the light distribution characteristics of the light-emitting element wafer CH is relatively slight (refer to FIG. 4), so eight photodiodes are used, but needless to say, The non-uniformity of the light distribution characteristics increases the number of configurations η. As shown in the front view of Fig. 2(c), the photodiode PD9 having a large light receiving surface is disposed under the vertical line 0 of the inspection table. In the present embodiment, since the inspection table is made of translucent glass (ground glass), the emitted light from the light-emitting element CH becomes scattered light at the inspection table and reaches the light diode. PD9. Therefore, the photodiode PD9 is intended to receive scattered light that is orthogonal to the light receiving surface, so that the light receiving area is large and does not become a problem. Further, since the amount of attenuation caused by the passage of the emitted light through the inspection of the table is grasped by the experiment in advance, it is corrected by the correction calculation at the time of the measurement operation. 〇 Next, the circuit configuration of the photometric device EQU will be described. As shown in FIG. 1(a), the photometric device EQU is an illumination driving unit DR for controlling the light-emitting element wafer CH, and nine photodiodes PD1 to PD9, and an output ratio to the photodiode. Voltages of the starting currents of PD1 to PD9 - 9 PD amplifiers 1 and 1 and the spectrometer 2 of the spectral spectrum of the emitted light from the light-emitting element wafer CH are accurately obtained, and the receiving slave PD amplifier 1 The calculation unit 3 that calculates the output of the spectrometer 2 and calculates various kinds of photometers is mainly composed of the machine control unit 4 that appropriately controls the operation of each unit of the apparatus. 200928327 The calculation unit 3 is typically composed of a personal computer and functions as a central control unit of the photometric device EQU. On the other hand, the photometric unit calculated by the calculation unit 3 is sequentially stored in the memory unit connected to the calculation unit 4, and is output to the display output unit DISP. In the case of the embodiment, the photodiodes PD1 to PD8 each have a light receiving surface having a diameter of about 10 mm. However, due to the limitation of the casing and the like, the effective area, as a circle with a radius of 7 = 4, is S = r2X 0 π = 50.27 [mm2]. On the other hand, the distance R between the light-emitting element wafer CH and the photodiode PD is set to about 42 mm. Therefore, the solid angle of each of the photodiodes PD1 to PD8 is about 72xtt/(RxR) = 0.02. Further, the light receiving area of the photodiode PD9 is set to be about 10 to 20 times the light receiving area of the other photodiodes PD1 to PD8. As shown in the schematic diagram of FIG. 1(b), generally, nine PD amplifiers 1 are respectively composed of an OP amplifier 6, and n feedback resistors Ri (R1 RRn Φ ), and load resistors RL, and n. The switching element Si (S 1 to Sn ) is formed. Here, the switching element Si is set to the ON state via the machine control unit 4, and the rest is in the OFF state. The inverting input terminal and the non-inverting input terminal of the 〇p amplifier 6 are in a virtual short circuit state. The input impedance of the OP amplifier 6 is almost infinite. Therefore, the current of the photodiode PD is the output voltage Vo of all the current flowing to the feedback resistor Ri°PD amplifier 1, which is proportional to the current of the photodiode PD, and becomes Vo=-RixI. In addition, the resistance 値' of the feedback resistor Ri (R1 to Rn) is due to the fact that it is required at the pD amplifier 1 -13- ❹
200928327 求之增益而被設定，因此，例如，若是爲被要求 106、107、108之4種類的增益的情況，則4個 阻R1〜R4之電阻値，係成爲1〇5〜1〇8〔 Ω〕。 9個的PD放大器1…1之輸出電壓Vo，係 給至所對應之A/D變換器7…7。另外，A/D變換 解析度，例如，係爲16位元又或是16位元。被供 變換器7處之基準電壓等的A/D變換器7之動作狀 經由機器控制部7，而因應於PD放大器1之增益 宜作變更。 A/D變換器7…7所輸出之各數位資料（計沏 ，係全部被供給至第1通訊部8。而，第1通訊咅I 將9個的計測資料依序送訊至演算部3。在此實施 由於係經由平行通訊而傳送計測資料，因此，演算 係可在短時間內取得多數之計測資料。 作爲分光計2，在此實施例中，係僅使用有1 用繞射光柵（grating )而將放射光分光之光柵分 grating spectrometer)。如圖 2 ( c)所示一般，在 例中，分光計2之光纖纜線Fi，係被配置在發光元 CH之正上方。另外，當發光元件晶片CH之分光特 有方向性的情況時，係將複數之分光計2相分離而 參考圖2 ( c )之虛線部分）。 不論何者，分光計2，均係將放射光之頻譜（ 佈資料），作爲代表光能量強度之相對値的計數値 。此計數値，係透過第2通訊部9，而被傳送至演 I 105 ' 1回饋電 .別被供 器7之 .給至/D :態，係 :來被適 資料） 5 8，係 例中， 部3， 個的利 光計（ 此實施 件晶片 性爲具 配置（ 分光分 而輸出 算部3 -14- 200928327 。第2通訊部9，在此實施例中，爲了以8位元寬幅又或 是16位元寬幅而高速的傳送資料，係採用根據SCSI ( Small Computer System Interface)規格之平行通訊方式。 在演算部3中，針對9個的光二極體PD1〜PD9，係 ' 設置有如圖5所示一般之記憶分光感度資料PD1 ( λ )〜 Ρϋ9(λ)的分光感度表格TBL1(參考圖8(b))。如 圖5所示一般，一般而言，光二極體，雖係會對應於波長 0 而提昇感度，但是，在此實施例中，對於可視域之波長範 圍（λ = 300〜800〔 nm〕），係記憶有1〔 nm〕間隔之分 光感度資料〔A/W〕。 故而，若是參考分光感度表格TBL1，對於單位能量 •單位時間照射（1〔 # J/S〕= 1〔 V W〕），係成爲能夠 將各光二極體PD1〜PD9之起電流値〔/i A〕，以單位波 長（1〔 nm〕）之間隔來作特定。另外，爲了建構分光感 度表格TBL1，雖然亦可對每一光二極體而測定分光感度 Q 資料，但是，係以使用由元件製造商所提供之資料爲較簡 易。 不論如何，在分光感度表格TBL1中，由於係對於各 個光二極體，而記憶有分光感度表格’因此，係成爲不受 ' 到光二極體之特性的拘束，而可作正確之測定。 又，在演算部3中，係亦設置有將人類的視感度特性 V ( Λ )，以單位波長（1〔 nm〕）之間隔而作記億的視 感度表格TBL2(參考圖6、圖8(b))。作爲視感度特 性，係使用有CIE所致之標準分光視感度效率（Spectral -15- 200928327 1 u m i η 〇 u s e f f i c i e n c y )。此視感度特性，係將對於波長入 =5 5 5〔 nm〕之單色放射所感到的明亮度正規化爲1，而 表示在其他波長所感到之相同放射強度的明亮度之比。如 圖6所示一般，例如’波長；I = 470〔 nm〕之光，就算在 ' 物理上係爲相同之放射強度，亦係成爲僅能感受到波長λ • = 5 5 5〔 nm〕之光的約十分之一的明亮度。 接下來，將圖1之測光裝置EQU的動作內容（測光 Q 機制），根據圖8 ( a )之流程圖來作說明。如圖示一般， 最初，機器控制部4係根據從演算部（中央控制部）3而 來之指示而動作，並設定裝置各部之動作條件（ST1)。 例如，對測光時間間隔 '或是發光元件晶片CH之點燈時 間、或是PD放大器1…1之增益、或是分光計2之曝光時 間等，因應於所需要之測定條件來作設定。 若是結束此種之初期處理，則機器控制部4係根據從 演算部3而來之指示而對發光驅動部DR作控制，而對發 φ 光元件晶片CH供給驅動訊號（ST2)。又，對應於發光 元件晶片CΗ之點燈，而開始分光計2之動作（ST3 )。 接下來，演算部3，係透過第1通訊部8，而取得比 例於光二極體PD 1〜PD9之起電流的電壓値（計測資料） (ST4 )。而，演算部3，係根據A/D變換器7之解析度 或是PD放大器1…1之增益，而求取出光二極體PD1〜 PD9之起電流値〔# A〕並作記憶。另外，於此，假設光 二極體PD〜PD9之起電流値，係爲π〜I9〔 /ζΑ〕。 接下來’演算部3，係透過第2通訊部9，而取得從 -16- 200928327 發光元件晶片CH而來之放射光的分光分佈資料p (又） ，並儲存在記憶表格TBL3中（ST6)。 分光分佈資料Ρ(λ)，係爲將放射光之頻譜分佈以 相對値來表示者，將光能量之相對強度，作爲以1〔 nm〕 間隔而計測的計數値來展示（參考圖7、圖8 ( b ))。 _ 若是如以上一般而取得了計測結果，則演算部3，係 根據所取得之計測資料，而實行適當的演算處理，來計算 0 出受光面之放射通量〇e〔//W〕、和受光面之光通量φγ 〔// lm〕、和光度I 7 〔 " cd〕，並作記憶與顯示（ST7 ) 。以下，同樣的，針對下一個檢查對象（發光元件晶片） ，亦反覆進行步驟ST1〜ST7之處理。 圖9，係爲展示資料演算處理（ST6)之具體內容的 流程圖。最初，係根據第1光二極體之輸出11，和分光計 之輸出Ρ(λ)，而特定出發光元件晶片CH之放射能量 的頻譜分佈EG1(A)與光通量（D1(ST61)。具體而言 Q ，係由 EGi(A) = IlxP ( λ ) / Σ ( Ρ ( λ ) xPDl ( λ ) 〕......式（1 )與 Φ i = Kmx〔 Σ ( V ( λ ) xEGi ( λ )) 〕……式（2)之演算所致，以下，對式（1)之意義依序 作說明。 首先，經由式（1)之分母所示的ρ(λ) xPDl (入 )的演算，身爲由分光計2所得之計數値的分光分佈資料 Ρ ( λ )，係配合於第1光二極體之感度特性ΡΕΠ(λ) 而被修正。故而，係可作爲具備有修正後之分光分佈資料 Ρ ( λ ) X P D1 ( A )的放射光被照射於具有平坦之感度特 -17- 200928327 性的光二極體PD 1後之結果，而擬似地得到了起電流II 〇 接下來’經由式（1)之分母所示的Σ 〔P(又）xPDl (λ )〕的演算，而計算出將修正後之分光分佈資料P ( λ ) xPDl ( λ )對於全波長而積分後之總計數値SUM = • Σ 〔 P ( λ ) xPDl ( λ )〕。故而，ρ ( λ ) XPD1 ( λ ) /SUM之値，係成爲代表針對各波長之與總和値（=SUM 0 )的分配率。 故而，經由Ι1χΡ(又）xPDl ( λ ) /SUM之演算，第 1光二極體PD1之起電流値II，係成爲被分解爲每一波長 λ成分。換言之，相當於從第1光二極體PD1所檢測出之 總物理量的起電流値11，係擬似地被分解爲每一個其之構 成要素（波長成分）。 於此’接下來，若是將起電流値II之構成要素IlxP (入）xPDl ( λ ) /SUM，以第1光二極體PD1之分光感 φ 度資料PD 1 ( λ )來作除算，則從發光元件晶片Ch而來 之放射光的功率，係成爲被分解爲每一單位波長λ 〔= 1 nm〕成分。 若是對以上所說明之所有演算作總結，則係成爲11 xp (λ ) /SUM = 11 xP ( Λ ) / Σ〔P( λ ) xPD 1 ( λ )〕， 其結果’係與式（1 )之演算一致。此式（1 )演算之結果 ，係將僅代表（將放射光之頻譜分佈正確地作特定者的） 相對値之分光分佈資料P ( λ )，經由相當爲比例於放射 光之絕對値的物理量之起電流値II而作校正。換言之， -18- 200928327 僅係代表相對値之分光分佈資料P ( λ )，係成爲Μ變換 爲放射能量之頻譜分佈EG 1 ( λ )。另外，若是將頻譜分 佈EG 1 ( λ )對於全波長而求取其總和，則係求取出光二 極體PD1之受光面處的視感功率量〔^W〕，亦即是求取 ' 出放射通量EG I。 ' 而，光通量Φ〔 Lm〕，係爲將放射通量，根據分光視 感效率（Spectral luminous efficiency) V ( γ )與最大視 ❹ 感效果度（Maximum luminous efficacy ) Km而作變換者 ，並經由式（2)而被計算。另外，分光視感效率ν( λ )，係預先被記憶在視感度表格TBL2中。 Φ 1= Kmx ( Σ ( V ( λ ) xEGl ( λ ))] 式（2) 以上，雖對步驟S Τ 6 1之處理作了說明，但是，步驟 ST62〜ST68之處理，亦幾乎爲相同。亦即是，根據第2 ❹ 光二極體PD2〜第8光二極體PD8之起電流値12〜18，和 分光計2之輸出Ρ( λ)，而特定出發光元件晶片CH之 放射能量的頻譜分佈EG2 ( λ )〜EG8 ( λ )與光通量φ 2 〜Φ 8 ( ST62 〜〜ST68 )。 例如，針對第i光二極體PDi，係使用式（A )與式 (B )。 EGi ( λ ) = ΠχΡ (λ)/〔ΣΡ(λ) xPDi ( λ )〕......式（A) Φ i= Kmx〔 Σ V ( Λ ) xEGi (入）〕......式（B ) -19- 200928327 另一方面，第9光二極體PD 9，由於係經由半透明玻 璃製之檢查台EX而接收放射光，因此’考慮在檢查台EX 處之衰減量，而使用下述之式（3)與式（4)。另外’修 正參數Χ(λ)，係預先藉由實驗而決定。 EG9 ( λ ) =Χ ( λ ) χΙ9χΡ ( λ ) /Σ [ Ρ ( λ ) xPD9 ( Λ )]......式（3) φ9 = Κηιχ〔 Σ (V ( λ ) xEG9 ( λ ))〕……式（4) ❹ 經由以上之處理，在9處的計測點之放射能量的頻譜 分佈EG1(A)〜EG9(A)，與光通量Φ1〜Φ9，係被 特定。又，藉由將在各計測點處之放射能量的頻譜分佈 EGi ( λ )，設爲對於全波長之總和演算Σ EGi (又），而 亦特定出放射通量EG1〜EG9。 於此，對在9處之計測點的放射通量E G 1〜E G 9作總 合性的評價，而決定放射通量Φε(3Τ70)。對於總合評 φ 價之手法，雖並未特別限定，但是，係可採用因應於各光 二極體PDi之配置位麗或受光面積而附加上權重的總和演 算。 最簡易者’係只要將在各光二極體PDi處之放射通量 EGi作加算’則便能夠特定出關於各光二極體之受光面積 的總和面積之放射通量SEGi。此算出値2EGi，係成爲對 應於發光元件晶片之全放射通量Φε的相關性高之値。 又，如圖2(b)所示一般’近接於光二極體pd3之 被描繪有顏色較濃之網格的部分（表面積S1)，係可擬 -20- 200928327 似爲與在光二極體PD3處之放射通量EG3爲同一値。又 ，近接於光二極體PD7之被插i繪有'顏色較淡之網格的部分 (表面積S2)，係可擬似爲與在光二極體PD7處之放射 通量EG7爲同一値。 在此種擬似的情況中’經由〔Slx(EG1+EG2+EG3 + E G 4 ) + S2x ( EG5 + EG6 + EG7+ EG8) 〕/ S 的演算’ 可以特定出被放射至上方之放射光的放射通量之全量°另 ^ 外，面積S，係爲光二極體PD1〜PD8之受光面積。在此 實施例中’由於計測點係有8處’因此’雖然仍些許缺乏 正確性，但是，相較於如先前技術一般之根據1處的計測 値來對發光元件晶片之發光特性作評價的情況，係大幅地 提昇精確度。而且，若藉由本發明’則係具備有：只要使 計測點增加，便可無限制地將精確度提昇的優點。 又，在本實施例中，由於係將受光面積大之光二極體 PD7配置在發光元件晶片CH之背面側，因此，例如，係 〇 成爲亦可對被預定作爲覆晶封裝之發光元件晶片CH作計 測。另外，所謂覆晶封裝，係指在晶粒（die )與連接端 子間之連接處，並不使用導線，而經由被直接形成在晶粒 表面上之導電性的銲墊而連接者。 以上’雖針對放射通量而作了說明，但是，針對光通 量φ r，亦係對在9處之計測點的光通量値φ 1〜Φ 9作總 合評價並作決定（ST7 1 )。總合評價之手法，雖並未特別 限定’但是’例如係將光通量値φ 1〜φ 8平均化而作爲光 通量値Φ r。如同最初所說明一般，在本實施例中，由於 -21 - 200928327 係並未在光二極體處裝著視感度濾鏡，因此，係有能夠同 時取得放射通量與光通量之優點。 但是，光度ir 〔燭光〕，係爲每單位立體角（1球 面度，steradian )之光通量値〔流明〕。在本實施例 中，由於發光元件晶片CH與光二極體PD1〜PD8之相隔 距離係爲R，各光二極體PD1〜PD8之受光面積S係爲相 同，因此，經由xRxR/S之計算，而計算出光度 φ I r ( ST72 )。另外，在此實施例中，R与42mm，S与r2x π = 50.27 [ mm2) 〇 如以上所說明一般，在本實施例中，係根據從分光計 2所得之頻譜資料Ρ(λ)，和各光二極體之分光感度資 料PD1 ( λ )〜PD9 ( λ )，而將光二極體之檢測値Π〜 19，變換爲對每一波長之放射通量的實値。於此，由於將 代表相對値之頻譜資料Ρ ( λ )的精確度提昇一事係爲較 容易，因此，所計算初之對每一波長的放射通量實値的精 〇 確度亦爲高。故而，從放射通量實値所計算出並被作顯示 輸出的各測光値，亦成爲高精確度者。另外，作爲本實施 例之測光値，雖係計算出受光總量之放射通量〔// w〕、 將波長放射通量作視感度修正並換算後之光通量〔/xLm〕 ' 、身爲1球面度之光通量的光度値〔/zCd〕，但是，不用 說，作顯示輸出之測光値，係可適宜地作選擇。 以上，雖係針對本發明而作了具體說明，但是，作爲 實施例所例示之記載內容，係並非爲對本發明作限定者。 例如，係並不限定爲可視域之計測，針對紫外域或紅外域 -22- 200928327 等之任意的區域，亦可進行計測。 【圖式簡單說明】 圖1，係爲展示實施例之測光裝置的槪略構成之區塊 圖。 圖2，係爲例示光二極體之配置位置的圖面。 圖3，係爲展示待機狀態之發光元件晶片的平面圖。 Q 圖4，係爲展示發光元件晶片之配光特性的圖面。 圖5，係爲例示光二極體之分光感度特性的特性圖。 圖6，係爲展示人類之視感度效率的特性圖。 圖7，係爲例示從發光元件晶片而來之放射光的頻譜 分佈之特性圖。 圖8，係爲展示圖1之測光裝置的動4 F 5 C之流程圖 (a)、和展示演算部之內部構成的圖面（b)。 圖9，係爲用以說明圖1之測光裝置的演算處理之流 G 程圖。 【主要元件符號說明】 1 : PD放大器 2 :分光計 3 =演算部 4 :機械控制部 6 : OP放大器 7 : A/D變換器 -23- 200928327 8:第1通訊部 9 :第2通訊部 CH :發光元件晶片 EX :檢查台 EQU :測光裝置 ' S E :黏著薄片 PD1 :光二極體 φ PD2 :光二極體 PD3 :光二極體 PD4 :光二極體 PD5 :光二極體 PD6 :光二極體 PD7 :光二極體 PD8 :光二極體 PD9 :光二極體 〇 Ri :回饋電阻 RL :負荷電阻 Si :開關元件 -24-200928327 is set to gain. Therefore, for example, if four types of gains of 106, 107, and 108 are required, the resistance of four resistors R1 to R4 is 1〇5~1〇8. Ω]. The output voltages Vo of the nine PD amplifiers 1...1 are supplied to the corresponding A/D converters 7...7. In addition, the A/D conversion resolution is, for example, 16 bits or 16 bits. The operation of the A/D converter 7 such as the reference voltage supplied to the inverter 7 is controlled by the device control unit 7, and the gain of the PD amplifier 1 is preferably changed. Each of the digital data output by the A/D converters 7...7 is supplied to the first communication unit 8. The first communication unit I sequentially transmits the nine measurement data to the calculation unit 3. In this implementation, since the measurement data is transmitted via parallel communication, the calculation system can obtain a majority of the measurement data in a short time. As the spectrometer 2, in this embodiment, only one type of diffraction grating is used ( The grating is divided into grating spectrometers. As shown in Fig. 2(c), in general, in the example, the optical fiber cable Fi of the spectrometer 2 is disposed directly above the illuminating element CH. Further, when the splitting of the light-emitting element wafer CH is characteristic, the plurality of spectrometers 2 are separated and referred to the broken line portion of Fig. 2(c). In either case, the spectrometer 2 uses the spectrum of the emitted light (the cloth data) as the count 値 representing the relative enthalpy of the intensity of the light energy. This count is transmitted to the I 105 '1 feedback through the second communication unit 9. It is not supplied to the /D: state by the supplier 7. The system is: In the middle, the third part of the light meter (the implementation of the wafer is configured (divided into light and the output of the calculation unit 3-14-200928327. The second communication part 9, in this embodiment, in order to have an 8-bit width In addition, it is a parallel communication method based on the SCSI (Small Computer System Interface) specification. In the calculation unit 3, the nine optical diodes PD1 to PD9 are set. There is a spectral sensitivity table TBL1 (refer to FIG. 8(b)) of the general memory spectral sensitivity data PD1 (λ) to Ρϋ9 (λ) as shown in FIG. 5. As shown in FIG. 5, generally, the photodiode is The sensitivity is increased corresponding to the wavelength 0, but in this embodiment, for the wavelength range of the visible field (λ = 300 to 800 [nm]), the spectral sensitivity data with 1 [nm] interval is stored [A/ W]. Therefore, if it is the reference spectral sensitivity table TBL1, for unit energy • unit time exposure (1 [ #J/S] = 1 [VW]), the current 値[/i A] of each of the photodiodes PD1 to PD9 can be specified at intervals of a unit wavelength (1 [nm]). In order to construct the spectral sensitivity table TBL1, although the spectral sensitivity Q data can be measured for each photodiode, it is simpler to use the information provided by the component manufacturer. In any case, in the spectral sensitivity table TBL1, Since the spectroscopic sensitivity table is memorized for each photodiode, it is determined that it is not restricted by the characteristics of the photodiode, and can be accurately measured. Further, in the calculation unit 3, the system 3 is also provided. The human visual sensitivity characteristic V ( Λ ) is expressed in the illuminance table TBL2 (see FIG. 6 and FIG. 8(b)) at intervals of a unit wavelength (1 [nm]). There is a standard spectroscopic visual sensitivity efficiency (Spectral -15- 200928327 1 umi η 〇usefficiency ) caused by CIE. This visual sensitivity characteristic is that the brightness of the monochromatic radiation with a wavelength of = 5 5 5 [nm] is normal. Turned into 1, and expressed in it The ratio of the brightness of the same radiation intensity that he feels at the wavelength. As shown in Fig. 6, for example, 'wavelength; I = 470 [nm] light, even if it is physically the same radiation intensity, it is only The brightness of about one tenth of the light of the wavelength λ • = 5 5 5 [nm] can be felt. Next, the action content (photometry Q mechanism) of the photometric device EQU of Fig. 1 is based on Fig. 8 (a) The flow chart is used for explanation. As shown in the figure, first, the machine control unit 4 operates in accordance with an instruction from the calculation unit (central control unit) 3, and sets the operating conditions of each unit of the apparatus (ST1). For example, the metering time interval 'or the lighting time of the light-emitting element chip CH, the gain of the PD amplifier 1...1, or the exposure time of the spectrometer 2, etc., is set in accordance with the required measurement conditions. When the initial processing is completed, the device control unit 4 controls the light-emission drive unit DR based on the instruction from the calculation unit 3, and supplies a drive signal to the φ-optical element wafer CH (ST2). Further, the operation of the spectrometer 2 is started in response to the lighting of the light-emitting element wafer C (ST3). Then, the calculation unit 3 obtains a voltage 计 (measurement data) (ST4) of the currents of the photodiodes PD 1 to PD9 through the first communication unit 8. On the other hand, the calculation unit 3 obtains the current 値[# A] of the photodiodes PD1 to PD9 and stores them based on the resolution of the A/D converter 7 or the gain of the PD amplifiers 1...1. Further, here, it is assumed that the currents of the photodiodes PD to PD9 are π to I9 [ / ζΑ]. Next, the calculation unit 3 acquires the spectral distribution data p (also) of the emitted light from the light-emitting element wafer CH of the-16-200928327 through the second communication unit 9, and stores it in the memory table TBL3 (ST6). . The spectral distribution data λ(λ) is expressed by comparing the spectral distribution of the emitted light with respect to 値, and the relative intensity of the optical energy is displayed as a count 计 measured at intervals of 1 [nm] (refer to FIG. 7 and FIG. 8 (b)). _ If the measurement result is obtained as described above, the calculation unit 3 performs an appropriate calculation process based on the acquired measurement data to calculate the radiance 〇e[//W] of the light-receiving surface. The luminous flux φγ [// lm] and the illuminance I 7 [ " cd] of the light receiving surface are recorded and displayed (ST7). In the same manner, the processing of steps ST1 to ST7 is repeated for the next inspection target (light-emitting element wafer). Fig. 9 is a flow chart showing the details of the data calculation processing (ST6). Initially, based on the output 11 of the first photodiode and the output Ρ(λ) of the spectrometer, the spectral distribution EG1 (A) and the luminous flux (D1 (ST61)) of the emission energy of the light-emitting element wafer CH are specified. Q, by EGI(A) = IlxP ( λ ) / Σ ( Ρ ( λ ) xPDl ( λ ) 〕 Equation (1 ) and Φ i = Kmx [ Σ ( V ( λ ) xEGi ( λ )) 〕 Equation (2) is calculated by the calculation. Hereinafter, the meaning of the formula (1) will be described in order. First, ρ(λ) xPDl (in) shown by the denominator of the formula (1) In the calculation, the spectral distribution data (λ) of the count 値 obtained by the spectrometer 2 is corrected in accordance with the sensitivity characteristic ΡΕΠ(λ) of the first photodiode. Therefore, it can be corrected as The spectral distribution data ( λ ) XP D1 ( A ) is irradiated to the photodiode PD 1 having a flat sensitivity of -17-200928327, and the current is obtained by pseudo-II 〇 Calculating the corrected spectral distribution data P ( λ ) xPDl ( λ ) for the full wavelength by the calculation of Σ [P(又)xPDl (λ )] shown by the denominator of equation (1) The total count after 値SUM = • Σ [ P ( λ ) xPDl ( λ )]. Therefore, ρ ( λ ) XPD1 ( λ ) / SUM is the sum of the sums for each wavelength = (= SUM 0 ) Therefore, the current 値II of the first photodiode PD1 is decomposed into the λ component of each wavelength by the calculation of Ι1χΡ(also)xPDl(λ)/SUM. In other words, it is equivalent to the first one. The starting current 値11 of the total physical quantity detected by the photodiode PD1 is pseudo-decomposed into each of its constituent elements (wavelength components). Here, if the current element III is formed, IlxP ( When xPD1 ( λ ) /SUM is subtracted from the spectral φ degree data PD 1 ( λ ) of the first photodiode PD1, the power of the emitted light from the light-emitting element wafer Ch is decomposed into Each unit wavelength λ [= 1 nm] component. If all the calculations described above are summarized, it is 11 xp (λ ) /SUM = 11 xP ( Λ ) / Σ [P( λ ) xPD 1 ( λ )], the result 'is consistent with the calculation of formula (1). The result of this formula (1) calculus will only represent (will radiate The spectral distribution of the light is correctly corrected to the specific spectral distribution data P ( λ ), which is corrected by the current 値 II which is proportional to the physical quantity of the absolute enthalpy of the emitted light. In other words, -18- 200928327 is only the spectral distribution data P ( λ ) representing the relative enthalpy, which is the spectral distribution EG 1 ( λ ) converted into radiant energy. Further, if the spectral distribution EG 1 ( λ ) is obtained for the total wavelength, the visual power amount (^W) at the light receiving surface of the photodiode PD1 is taken out, that is, the result is obtained. Radiation flux EG I. The luminous flux Φ [ Lm ] is a change in the radiation flux based on the Spectral luminous efficiency V ( γ ) and the Maximum luminous efficacy Km. It is calculated by the formula (2). Further, the spectral visual efficiency ν(λ) is previously stored in the visual sensitivity table TBL2. Φ 1 = Kmx ( Σ ( V ( λ ) xEGl ( λ ))) (2) Although the processing of the step S Τ 6 1 has been described above, the processing of steps ST62 to ST68 is almost the same. That is, the spectrum of the radiant energy of the light-emitting element wafer CH is specified based on the currents 値12 to 18 of the second to eighth photodiodes PD2 to PD8 and the output Ρ(λ) of the spectrometer 2. Distributions EG2 (λ) to EG8 (λ) and luminous fluxes φ 2 to Φ 8 (ST62 to ST68). For example, for the ith photodiode PDi, the equations (A) and (B) are used. EGi (λ) = ΠχΡ (λ) / [ΣΡ(λ) xPDi ( λ )] (A) Φ i = Kmx [ Σ V ( Λ ) xEGi (in)] (B) -19- 200928327 On the other hand, the ninth photodiode PD 9 receives the emitted light through the inspection table EX made of translucent glass. Therefore, considering the amount of attenuation at the inspection table EX, the following is used. Equations (3) and (4). In addition, the 'correction parameter Χ(λ) is determined in advance by experiment. EG9 ( λ ) = Χ ( λ ) χΙ 9 χΡ ( λ ) / Σ [ Ρ ( λ ) xPD9 ( Λ )]...Formula (3) φ9 = Κηιχ[ Σ (V ( λ ) xE G9 ( λ ))] (4) 经由 Through the above processing, the spectral distributions EG1(A) to EG9(A) of the radiation energy at the measurement points at nine points, and the luminous fluxes Φ1 to Φ9 are specified. Further, by arranging the spectral distribution EGi ( λ ) of the radiation energy at each measurement point as the sum of the total wavelengths Σ EGi (also), the radiation fluxes EG1 to EG9 are also specified. The radiation fluxes EG 1 to EG 9 at the measurement points at 9 points are evaluated for the totality, and the radiation flux Φ ε (3 Τ 70) is determined. The method for the total evaluation of the φ price is not particularly limited, but A summation calculation is performed by adding a weight to the arrangement of the light-emitting or light-receiving areas of the photodiodes PDi. The simplest one is to add the radiation flux EGi at each photodiode PDi to be specified. The radiation flux SEGi of the total area of the light-receiving areas of the respective photodiodes is calculated as 値2EGi, which has a high correlation with the total radiation flux Φε of the light-emitting element wafer. Further, as shown in Fig. 2(b) Generally shown as being close to the light diode pd3 is depicted as a network with a stronger color Portion (surface area S1), the proposed system can -20-200928327 seems to be the same and the radiation flux EG3 PD3 Zhi in the light diode. Further, the portion of the photodiode PD7 which is closely connected to the light-colored grid (surface area S2) is similar to the radiation flux EG7 at the photodiode PD7. In such a pseudo-like case, 'the calculation of the radiation emitted to the upper side by the calculation of [Slx(EG1+EG2+EG3 + EG 4 ) + S2x ( EG5 + EG6 + EG7+ EG8)] / S The total amount of the amount is also the area of the light receiving area of the photodiodes PD1 to PD8. In this embodiment, 'because there are 8 places in the measurement point, 'there is still a lack of correctness, but the luminescence characteristics of the light-emitting element wafer are evaluated in comparison with the measurement 値 according to 1 in the prior art. The situation is to greatly improve the accuracy. Further, according to the present invention, there is an advantage that the accuracy can be improved without limitation as long as the measurement point is increased. Further, in the present embodiment, the photodiode PD7 having a large light receiving area is disposed on the back side of the light-emitting element wafer CH. Therefore, for example, the light-emitting element wafer CH which is intended to be a flip-chip package can be used. Take measurements. Further, the flip chip package refers to a connection between a die and a connection terminal without using a wire, and is connected via a conductive pad directly formed on the surface of the die. Although the above description has been made with respect to the radiation flux, the luminous flux φ r is also evaluated for the total of the luminous fluxes 値 φ 1 to Φ 9 at the measurement points at nine points (ST7 1 ). The method of the total evaluation is not particularly limited 'but', for example, the luminous flux 値 φ 1 to φ 8 is averaged as the luminous flux 値 Φ r . As described in the first place, in the present embodiment, since -21 - 200928327 does not have a luminosity filter at the photodiode, there is an advantage that the radiant flux and the luminous flux can be simultaneously obtained. However, the luminosity ir [candle] is the luminous flux [lumen] per unit solid angle (1 steradian). In the present embodiment, since the distance between the light-emitting element wafer CH and the photodiodes PD1 to PD8 is R, and the light-receiving areas S of the photodiodes PD1 to PD8 are the same, the calculation is performed by xRxR/S. The luminosity φ I r (ST72 ) is calculated. Further, in this embodiment, R and 42 mm, S and r2x π = 50.27 [mm2), as described above, in the present embodiment, based on the spectrum data λ(λ) obtained from the spectrometer 2, and The spectral sensitivity data PD1 ( λ ) to PD9 ( λ ) of each photodiode are converted into 値 19 of the photodiode, which is converted into the actual radiance of each wavelength. Here, since it is easier to increase the accuracy of the spectral data Ρ ( λ ) representing the relative enthalpy, the accuracy of the calculation of the radiant flux for each wavelength is also high. Therefore, each of the photometers calculated from the radiation flux and outputted as a display is also highly accurate. Further, as the photometer of the present embodiment, the radiation flux [//w] of the total amount of received light is calculated, and the luminous flux [/xLm] of the wavelength radiance is corrected and converted. The luminosity of the spherical luminous flux 値 [/zCd], however, needless to say, the photometric 作 for display output can be suitably selected. The above description of the present invention has been specifically described, but the description of the examples is not intended to limit the invention. For example, it is not limited to the measurement of the visible field, and measurement can be performed for any region such as the ultraviolet region or the infrared region -22-200928327. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a block diagram showing a schematic configuration of a photometric device of an embodiment. Fig. 2 is a view showing the arrangement position of the photodiode. Fig. 3 is a plan view showing a light-emitting element wafer in a standby state. Q Fig. 4 is a view showing the light distribution characteristics of the light-emitting element wafer. Fig. 5 is a characteristic diagram illustrating the spectral sensitivity characteristics of the photodiode. Figure 6 is a characteristic diagram showing the efficiency of human visual sensitivity. Fig. 7 is a characteristic diagram illustrating a spectral distribution of emitted light from a light-emitting element wafer. Fig. 8 is a flow chart (a) showing the movement 4 F 5 C of the photometric apparatus of Fig. 1, and a drawing (b) showing the internal structure of the calculation unit. Fig. 9 is a flow chart for explaining the calculation processing of the photometric device of Fig. 1. [Description of main component symbols] 1 : PD amplifier 2 : Spectrometer 3 = Calculation unit 4 : Mechanical control unit 6 : OP amplifier 7 : A/D converter -23 - 200928327 8: First communication unit 9 : Second communication unit CH: light-emitting element wafer EX: inspection table EQU: photometric device 'SE: adhesive sheet PD1: photodiode φ PD2: photodiode PD3: photodiode PD4: photodiode PD5: photodiode PD6: photodiode PD7 : Photodiode PD8: Photodiode PD9: Photodiode 〇Ri: Feedback resistor RL: Load resistance Si: Switching element-24-