JP3580407B2 - Optical spectrum analyzer - Google Patents

Description

ただし、ｒはアレイ素子のピッチ
Ｐは一方の３つの素子の出力から求めたパワーレベル
Ｐ’は他方の３つの素子の出力から求めたパワーレベル
Δｘは一方の３つの素子におけるずれ量
Δｘ’は他方の３つの素子におけるずれ量である。
【００１４】
このように、ビーム中心とアレイ素子の中心とのずれ量に応じて、一方の３素子の出力で求めたパワーレベルを採用するか、または両方の３素子出力により求めたパワーレベルの加重平均を採用することにより、容易に測定パワーレベルの変動を軽減することができる。
【００１５】
【発明の実施の形態】
以下図面を用いて本発明を詳しく説明する。本発明は図４に示す構成と同様であるが、演算装置３０の機能のみ以下に述べるように異なっている。
【００１６】
ここではガウス分布で近似する場合を例に採って説明する。通常、アレイ素子１５における出力が最大の素子と、その両隣の素子の３つの素子からの出力Ｖ_Ａ，Ｖ_Ｂ，Ｖ_Ｃから、元のスペクトルをガウス分布としてスポットサイズωを求め、トータルパワーを類推する。
【００１７】
スペクトルが理想的な曲線からずれているときは、ωの誤差が大きく、測定レベルは変動する。特にスペクトルのピーク位置が図１（ａ）に示すようにアレイ素子の中間に位置するとき、上記ずれは顕著になる。
【００１８】
特にスペクトルが左右非対称のときには、隣接する第４のアレイ素子出力Ｖ_Ｄを用いてＶ_Ｂ，Ｖ_Ｃ，Ｖ_Ｄの３点で近似した方が元のスペクトルをより忠実に反映したものになることもある。
【００１９】
したがって、下記のような方法でパワーレベルを演算すると、相対的なパワーレベルの変動を軽減することができる。
（ａ）図１（ｃ）に示すように、それぞれの３点（Ｖ_Ａ，Ｖ_Ｂ，Ｖ_Ｃ），（Ｖ_Ｂ，Ｖ_Ｃ，Ｖ_Ｄ）で演算したスポットサイズとパワーレベル（ω，Ｐ），（ω’，Ｐ’）のうちスポットサイズが小さい方のパワーレベルを採用する（以下この方式を本発明の（ａ）の方式と呼ぶ）。
なお、スポットサイズの小さい方がパワーの集中度が高くなるため、スポットサイズが小さくなる方の系列から求めたパワーの方がより真値に近くなる。したがって、上記のように、２組のデータのうちスポットサイズが小さい方のパワーレベルを採用する。
（ｂ）スペクトルの中心と素子の中心とのずれ量Δｘに応じて、ＰとＰ’との加重平均を採用する（以下この方式を本発明の（ｂ）の方式と呼ぶ）。
【００２０】
なお、３点（例えば（Ｖ_Ａ，Ｖ_Ｂ，Ｖ_Ｃ）の３点）からスポットサイズとパワーレベルを求める演算のアルゴリズムは、特開平０８−２５４４６５号に記載されたアルゴリズムに同じである。
【００２１】
以下そのアルゴリズムについて説明する。アレイ素子は、図２に示すように、短冊状の受光部（素子または受光素子という）が配列されたもので、いま素子の数をｎ_０、素子の長さをＬ、素子の幅をｄ、素子のピッチをｒとする。このような構成のアレイ素子の中で出力が最大となるｎ番目の素子の中心と、出射ビーム（強度分布ｆ（ｘ，ｙ））の中心（ｘ＝０）とがΔｘだけずれている場合、本発明では各素子の出力Ｐ_ｎ−２ ，Ｐ_ｎ−１ ，Ｐ_ｎ（ｍａｘ） ，Ｐ_ｎ＋１ ，・・・からΔｘを求め、波長を補間してトータルパワーを推定する。
【００２２】
ただし、前提として
（１） 被測定光のスペクトル線幅は、十分小さく（ここではレーザ光を想定している）、出射ビームの広がり（あるいは形状）は主としてレンズによる回折および入射口のパワー分布によって決まる。
（２） アレイ素子の番号と波長とは予め対応が付けられており、アレイ素子間もΔｘによって補間できるものとする。素子番号１，・・・，ｎ−１ ，ｎ ，ｎ＋１ ，・・・，ｎ_０に対し、波長λ_１，・・・λ_ｎ−１，λ_ｎ，λ_ｎ＋１ ，・・・，λ_０が対応し、各素子間の波長差は
λ_ｋ−λ_ｋ−１≒Δλ
ただし、ｋ：２〜ｎ_０
とする。
【００２３】
以上の事柄を前提とすれば、図５のフローチャートに示すように、実測値
Ｐ_１，・・・，Ｐ_ｎ−１，Ｐ_ｎ，Ｐ_ｎ＋１，・・・，Ｐ_ｎ０を用い、Δｘを推定し、中心波長
λ_０、トータルパワーＰ_{ｔｏｔａｌ}を求めることができる。以下中心波長λ０ 、トータルパワーＰｔｏｔａｌを求める場合の動作について説明する。
【００２４】
シングルモードファイバ入射のように、入射口での光強度分布ｇ（ξ，η）がガウス型の場合、トータルパワーを１とすれば、
ｇ（ξ，η）＝（２／πω_０ ^２）ｅｘｐ｛−２（ξ^２＋η^２）／ω_０ ^２｝……（１）
ただし、ω_０はスポットサイズ
となる。
【００２５】
レンズ系によってビームが変換あるいは多少の回折の効果を受けても、アレイ素子上での光強度分布ｆ（ｘ，ｙ）はやはりガウス型となり、損失がないものとすれば、
ｆ（ｘ，ｙ）＝（２／πω^２）ｅｘｐ｛−２（ｘ^２＋ｙ^２）／ω^２｝……（２）
ただし、ωはスポットサイズ
となる。
【００２６】
したがって、ｋ番目の素子に照射される光パワーｐ_ｋは、
【数１】

ただし、ｘ_１ ＝（ｎ−ｋ）ｒ−ｄ／２＋Δｘ
ｘ_２ ＝（ｎ−ｋ）ｒ＋ｄ／２＋Δｘ
で表わされる。
【００２７】
ここで、Ｌ／２≫ωとし、積分範囲（ｙ軸）を
【数２】

と置き換えることができれば、
【数３】

より、
【００２８】
【数４】

さらに、変数変換２^１／２×（ｘ／ω）＝ｔを行えば、
【数５】

ただし、ｔ_１＝２^１／２（１／ω）｛（ｎ−ｋ）ｒ−ｄ／２＋Δｘ｝
ｔ_２＝２^１／２（１／ω）｛（ｎ−ｋ）ｒ＋ｄ／２＋Δｘ｝
となる。
【００２９】
この積分は数値積分も容易であるが、補誤差関数
【数６】

を用いれば、
ｐ_ｋ＝（１／２）｛Ｅ_ｒｆ_ｃ（ｔ_２）−Ｅ_ｒｆ_ｃ（ｔ_１）｝
となる。
【００３０】
さて、ｐ_ｎ＋１／ｐ_ｎ−１の対数値すなわちｌｎ（ｐ_ｎ＋１／ｐ_ｎ−１）とΔｘとはほぼ直線的な関係にあるから、実測値Ｐｎ−１ ，Ｐｎ ，Ｐｎ＋１ が得られれば、Δｘは容易に求められ、中心波長λ_０およびトータルパワーＰ_{ｔｏｔａｌ}はそれぞれ次式で求められる。
λ_０＝λ_ｎ−Δｘ・Δλ／（ｒ／２）
Ｐ_{ｔｏｔａｌ}＝Ｐ_ｎ／ｐ_ｎ
【００３１】
このようにして、隣接する３点（Ｖ_Ａ，Ｖ_Ｂ，Ｖ_Ｃ），（Ｖ_Ｂ，Ｖ_Ｃ，Ｖ_Ｄ）のアレイ素子出力によりそれぞれスポットサイズとパワーレベルを求め、スポットサイズの小さい方のパワーレベルを採用する（本発明の（ａ）の方式）。
【００３２】
また、スペクトルの中心と素子の中心とのずれ量Δｘに応じて次のようにパワーレベルを求めるようにすることもできる（本発明の（ｂ）の方式）。
すなわち、
（１） |Δｘ|≦α (αは適宜の定数) の場合
パワーレベル＝Ｐ
（２） |Δｘ|＞α （αは適宜の定数）の場合
パワーレベル＝Ｐ×（ｒ−｜（Δｘ＋Δｘ’）／２｜）／ｒ ＋Ｐ’×（｜（Δｘ＋Δｘ’）／２｜）／ｒ
ただし、ｒはアレイ素子のピッチ
Ｐは一方の３つの素子の出力から求めたパワーレベル
Ｐ’は他方の３つの素子の出力から求めたパワーレベル
Δｘは一方の３つの素子におけるずれ量
Δｘ’は他方の３つの素子におけるずれ量である。
【００３３】
図３に同一パワーのスペクトルの波長を変えて（つまりスペクトル中心と素子中心を変えて）測定した結果を示す。横軸が波長、縦軸がレベルである。実線Ａは従来方式、鎖線Ｂは本発明の（ａ）の方式、鎖線Ｃは本発明の（ｂ）の方式による場合を示す。
【００３４】
従来方式でパワーレベルが急激に変動しているところは、ビームが素子と素子の間（真中に近い）のときである。特にビームの左右非対称性が大きいときは上記変動も大となる。
しかし、本発明の（ａ），（ｂ）の方式では、上記のような不連続的変動はなくなる。更に、（ｂ）の方式では変動幅も小さくなっている（ただし、α＝１０である）。
【００３５】
なお、以上の説明は、本発明の説明および例示を目的として特定の好適な実施例を示したに過ぎない。したがって本発明は、上記実施例に限定されることなく、その本質から逸脱しない範囲で更に多くの変更、変形をも含むものである。
【００３６】
【発明の効果】
以上説明したように本発明によれば次のような効果がある。
請求項１に記載の発明によれば、３点のアレイ素子出力から、ガウス分布を仮定して、ビーム径、パワーレベル、スペクトルの中心とアレイ素子の中心とのずれ量を求めるという従来の技法がそのまま利用でき、簡単な演算により測定パワー変動を容易に低減できる。要するに、分光器は従来の構成のままで演算装置のみの変更で、相対パワー変動（フラットネス）を改善することができる。
【００３７】
また、請求項２のように、スペクトルの中心と素子の中心とのずれ量Δｘに応じてパワーレベルを求めるようにしても、測定パワー変動を容易に低減することができる。
【図面の簡単な説明】
【図１】スペクトル形状とアレイ素子出力の関係を説明する図である。
【図２】アレイ素子と出射ビームの関係を示す図である。
【図３】パワーレベルの変動についての従来方式と本発明の方式との比較図である。
【図４】光スペクトラムアナライザの一実施例を示す構成図である。
【図５】中心波長およびトータルパワーを求める動作フローである。
【符号の説明】
１０ 分光装置
１１ スリット
１２ コリメーティングミラー
１３ 分散素子
１４ フォーカシングミラー
１５ アレイ素子
２０ 駆動装置
３０ 演算装置
４０ 表示装置[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to improvement of wavelength accuracy and power level accuracy in an optical spectrum analyzer using an array detector, and more particularly to improvement of relative level accuracy in quasi-monochromatic light spectrum measurement.
[0002]
[Prior art]
The conventional optical spectrum analyzer indicates an average value of optical power existing in a certain finite slit width, that is, a spectral width.
Further, even in a spectroscope using an array element, the output of each element has to be displayed as it is. This is because the spectrum shape of the measured light is assumed to be arbitrary, and this spectrum shape and the output of the array element do not always correspond to one body.
Therefore, it has been difficult to estimate the true spectral shape from the output of the array element.
[0003]
To solve this problem, there is JP-A-08-254465 filed by the applicant of the present invention, entitled "Optical Spectrum Analyzer". In this method, the central wavelength and the total power of the corresponding monochromatic light can be obtained by a simple operation based on the outputs of the adjacent array elements.
[0004]
FIG. 4 shows a configuration example of the optical spectrum analyzer. The light to be measured incident through the entrance of the slit 11 becomes parallel light by the collimating mirror 12 and enters the dispersion element 13. The light emitted from the dispersion element 13 is focused on the array element 15 by the focusing mirror 14. In this case, the dispersion element 13 is fixed, and the position of the light spot hitting the array element 15 moves according to the wavelength of the light to be measured.
[0005]
The array device is driven by the driving device 20, the output signal of the array device hit by the light spot is read, and the arithmetic device 30 performs arithmetic processing as shown in FIG.
Note that the array elements are formed by arranging strip-shaped light receiving portions. Here, the number of elements is n ₀ , the element length is L, the element width is d, and the element pitch is r.
[0006]
When the center of the n-th element having the maximum output in the array element having such a configuration is shifted by Δx from the center of the output beam, Δx is obtained from the output of each element to obtain the center wavelength. At the same time, the total power is estimated by interpolating the wavelength.
[0007]
In other words, when measuring the power level of quasi-monochromatic light with a spectrometer using a photodiode array element as described above, the spectrum shape is approximated to a Gaussian distribution or the like from the outputs of three elements including the element that becomes the peak output. The beam diameter ω and the total power (power level) are inferred from the output pattern of each array element.
[0008]
[Problems to be solved by the invention]
However, such an optical spectrum analyzer has a problem that when the spectrum shape deviates from an ideal curve (for example, Gaussian distribution), an error such as the beam diameter ω is large, and the measurement power level fluctuates.
[0009]
In other words, there is a limit in estimating the shape of the spectrum only from the outputs of three elements including the element having the peak output. In particular, when the spectrum is near between the elements, there has been a problem that the above-described deviation increases.
[0010]
An object of the present invention is to solve the above-mentioned problem, and to carry out a simple calculation while inheriting a conventional method for obtaining a beam diameter, a power level, a beam position, and the like assuming a Gaussian distribution from three element outputs. An object of the present invention is to provide an optical spectrum analyzer capable of reducing fluctuations in measured power and improving fluctuations in relative power (flatness).
[0011]
[Means for Solving the Problems]
In order to achieve such an object, in the present invention of claim 1,
A spectroscopic device configured to enter the light to be measured from the entrance into the dispersive element, to converge the light emitted from the dispersive element, and to irradiate the array element including a plurality of light receiving elements;
Among the four adjacent elements including the element having the largest output among the array elements irradiated with the light, the spot size and the power level are obtained by calculation from the outputs of three consecutive elements, and An arithmetic unit adapted to employ power levels obtained from outputs of the three smaller elements of the spot size as power levels.
[0012]
When the spectrum shape deviates from an ideal curve, an error such as a beam diameter is large, and the measured power level also varies. Obtain a spot size and a power level based on the output of each of three consecutive elements from among four consecutive elements, compare the two spot sizes, and adopt the power level of the smaller three elements. Accordingly, it is possible to reduce an error such as a beam diameter and to reduce fluctuations in the measured power level.
[0013]
In the invention of claim 2,
A spectroscopic device configured to enter the light to be measured from the entrance into the dispersive element, to converge the light emitted from the dispersive element, and to irradiate an array element including a plurality of light receiving elements; Among the four adjacent elements including the element having the largest output among the array elements obtained, the power level and the shift amount between the beam center and the center of the array element are obtained from the outputs of three consecutive elements, respectively. According to another feature of the present invention, there is provided an arithmetic unit configured to calculate a power level by the following arithmetic expression according to the shift amount.
(1) When | Δx | ≦ α (α is a constant), power level = P
(2) When | Δx |> α (α is a constant)

Here, r is the pitch P of the array elements, the power level P ′ obtained from the output of one of the three elements is the power level Δx obtained from the output of the other three elements, and the deviation Δx ′ of the one of the three elements is This is the shift amount of the other three elements.
[0014]
As described above, according to the amount of deviation between the beam center and the center of the array element, the power level obtained from the output of one of the three elements is employed, or the weighted average of the power levels obtained from the outputs of the three elements is used By adopting, the fluctuation of the measured power level can be easily reduced.
[0015]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, the present invention will be described in detail with reference to the drawings. The present invention has the same configuration as that shown in FIG. 4, but differs only in the function of the arithmetic unit 30 as described below.
[0016]
Here, a case of approximation by a Gaussian distribution will be described as an example. Normally, the output is the largest element in the array element 15, the output V _A of the three elements of the device of both sides _thereof, V _B, from V _C, determine the spot size ω the original spectrum as a Gaussian distribution, the total power By analogy.
[0017]
When the spectrum deviates from the ideal curve, the error of ω is large and the measurement level fluctuates. In particular, when the peak position of the spectrum is located in the middle of the array element as shown in FIG.
[0018]
Especially when spectra of asymmetrical is that V _B using a fourth array element output V _{D adjacent,} V _C, is better approximated by three points V _D will reflect the original spectrum more faithfully There is also.
[0019]
Therefore, when the power level is calculated by the following method, the relative fluctuation of the power level can be reduced.
(A) As shown in FIG. 1 (c), each of the three points _{(V A, V B, V} C), (V B, V C, V D) calculated in spot size and power level (omega, P ), (Ω ′, P ′), the power level with the smaller spot size is adopted (hereinafter, this method is referred to as the method (a) of the present invention).
Since the smaller the spot size, the higher the power concentration, the power obtained from the series with the smaller spot size is closer to the true value. Therefore, as described above, the power level with the smaller spot size of the two sets of data is adopted.
(B) A weighted average of P and P 'is employed according to the amount of shift Δx between the center of the spectrum and the center of the element (this method is hereinafter referred to as the method (b) of the present invention).
[0020]
Note that three points (e.g. _{(V A,} V B, three points of _{V C))} algorithm calculation for obtaining a spot size and power level from the same to the algorithm described in JP-A-08-254465.
[0021]
Hereinafter, the algorithm will be described. As shown in FIG. 2, the array element is an array of strip-shaped light receiving portions (referred to as elements or light receiving elements). The number of elements is n ₀ , the element length is L, and the element width is d. , And the element pitch is r. When the center of the n-th element having the maximum output among the array elements having such a configuration and the center (x = 0) of the output beam (intensity distribution f (x, y)) are shifted by Δx. In the present invention, Δx is obtained from the outputs P _n−2 , P _n−1 , P _n (max), P _{n + 1} ,... Of each element, and the total power is estimated by interpolating the wavelength.
[0022]
However, it is assumed that (1) the spectral line width of the light to be measured is sufficiently small (laser light is assumed here), and the spread (or shape) of the output beam is mainly due to diffraction by the lens and power distribution at the entrance. Decided.
(2) The numbers of the array elements and the wavelengths are previously associated with each other, and the inter-array elements can be interpolated by Δx. Element number 1, ···, n-1, n, n + 1, ···, to _{n 0,} the wavelength _{λ 1, ··· λ n-1} , λ n, λ n + 1, ···, is lambda ₀ Correspondingly, the wavelength difference between each element is λ _k -λ _k-1 ≒ Δλ
However, k: 2 to n ₀
And
[0023]
Assuming the above, as shown in the flowchart of FIG. 5, Δx is estimated using the actually measured values P ₁ ,..., P _n−1 , P _n , P _{n + 1 ,.} Then, the center wavelength λ ₀ and the total power P _total can be obtained. The operation for obtaining the center wavelength λ0 and the total power Ptotal will be described below.
[0024]
When the light intensity distribution g (ξ, η) at the entrance is Gaussian, as in the case of single mode fiber incidence, if the total power is 1,
g (ξ, η) = (2 / πω ₀ ² ) exp {−2 (ξ ² + η ² ) / ω ₀ ² } (1)
However, ω ₀ is the spot size.
[0025]
Even if the beam is converted or slightly diffracted by the lens system, the light intensity distribution f (x, y) on the array element is still Gaussian and if there is no loss,
f (x, y) = (2 / πω ² ) exp {−2 (x ² + y ² ) / ω ² } (2)
Here, ω is the spot size.
[0026]
Thus, the optical power p _k to be irradiated on the k-th element,
(Equation 1)

Where x ₁ = (nk) rd−2 + Δx
x ₂ = (n−k) r + d / 2 + Δx
Is represented by
[0027]
Here, L / 2≫ω, and the integration range (y-axis) is given by

If you can replace
(Equation 3)

Than,
[0028]
(Equation 4)

Further, if the variable conversion 2 ^1/2 × (x / ω) = t is performed,
(Equation 5)

Here, t ₁ = 2 ^1/2 (1 / ω) {(nk) r−d / 2 + Δx}
t ₂ = 2 ^1/2 (1 / ω) {(nk) r + d / 2 + Δx}
It becomes.
[0029]
This integration can be easily performed by numerical integration, but the complementary error function

If you use
_{p k = (1/2) {E} r f c (t 2) -E r f c (t 1)}
It becomes.
[0030]
Now, since the logarithmic value of _{pn + 1} / _pn-1 or ln ( _{pn + 1} / _pn-1 ) and .DELTA.x are almost linear, if the measured values Pn-1, Pn, Pn + 1 are obtained, Δx is easily obtained, and the center wavelength λ ₀ and the total power P _total are respectively obtained by the following equations.
λ ₀ = λ _n -Δx · Δλ / (r / 2)
P _total = P _n / p _n
[0031]
In this way, the adjacent three points _{(V A, V B, V} C), (V B, V C, V D) respectively determine the spot size and power level by the array element output, a spot size smaller A power level is adopted (the method (a) of the present invention).
[0032]
Further, the power level can be obtained as follows according to the shift amount Δx between the center of the spectrum and the center of the element ( the method (b) of the present invention).
That is,
(1) | Δx | ≦ α (α is an appropriate constant) Power level = P
(2) | Δx |> α (α is an appropriate constant) Power level = P × (r− | (Δx + Δx ′) / 2 |) / r + P ′ × (| (Δx + Δx ′) / 2 |) / r
Where r is the pitch of the array element
P is the power level obtained from the output of one of the three elements
P 'is the power level obtained from the outputs of the other three elements
Δx is the amount of displacement in one of the three elements
Δx ′ is a shift amount of the other three elements.
[0033]
FIG. 3 shows the result of measurement by changing the wavelength of the spectrum having the same power (that is, changing the center of the spectrum and the center of the element). The horizontal axis is wavelength and the vertical axis is level. The solid line A shows the case using the conventional method, the chain line B shows the case using the method (a) of the present invention, and the chain line C shows the case using the method (b) of the present invention.
[0034]
In the conventional method, the power level fluctuates rapidly when the beam is between the elements (close to the center). In particular, when the left-right asymmetry of the beam is large, the above fluctuation also becomes large.
However, in the methods (a) and (b) of the present invention, the discontinuous fluctuation as described above is eliminated. Further, in the method (b) , the fluctuation width is small (however, α = 10).
[0035]
It should be noted that the foregoing description has been directed to specific preferred embodiments for the purpose of explanation and illustration of the present invention. Therefore, the present invention is not limited to the above-described embodiment, but includes many more changes and modifications without departing from the spirit thereof.
[0036]
【The invention's effect】
As described above, the present invention has the following effects.
According to the first aspect of the present invention, a conventional technique for obtaining a beam diameter, a power level, and a shift amount between the center of a spectrum and the center of an array element, assuming a Gaussian distribution, from outputs of three array elements. Can be used as it is, and measurement power fluctuation can be easily reduced by a simple calculation. In short, the relative power fluctuation (flatness) can be improved by changing only the arithmetic unit while keeping the spectroscope in the conventional configuration.
[0037]
Further, even when the power level is determined according to the shift amount Δx between the center of the spectrum and the center of the element, the fluctuation of the measured power can be easily reduced.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a relationship between a spectrum shape and an array element output.
FIG. 2 is a diagram illustrating a relationship between an array element and an output beam.
FIG. 3 is a comparison diagram of a conventional method and a method of the present invention regarding power level fluctuation.
FIG. 4 is a configuration diagram illustrating an embodiment of an optical spectrum analyzer.
FIG. 5 is an operation flow for obtaining a center wavelength and a total power.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 10 Spectroscopy apparatus 11 Slit 12 Collimating mirror 13 Dispersion element 14 Focusing mirror 15 Array element 20 Drive device 30 Computing device 40 Display device

Claims (2)

A spectroscopic device configured to enter the light to be measured from the entrance into the dispersive element, to converge the light emitted from the dispersive element, and to irradiate the array element including a plurality of light receiving elements;
Among the four adjacent elements including the element having the maximum output among the array elements irradiated with the light, the spot size and the power level are obtained by calculation from the outputs of three consecutive elements, and An optical spectrum analyzer comprising an arithmetic unit configured to employ power levels determined from outputs of the three smaller spot sizes as power levels. A spectroscopic device configured to enter the light to be measured from the entrance into the dispersive element, to converge the light emitted from the dispersive element, and to irradiate the array element including a plurality of light receiving elements;
Among the four adjacent elements including the element having the maximum output among the array elements irradiated with light, the power level and the deviation between the beam center and the center of the array element from the output of every three consecutive elements. An optical spectrum analyzer comprising an arithmetic unit configured to calculate the respective amounts and calculate the power level according to the following arithmetic expression according to the shift amount.
(1) When | Δx | ≦ α (α is a constant), power level = P
(2) When | Δx |> α (α is a constant)

Here, r is the pitch P of the array elements, the power level P ′ obtained from the output of one of the three elements is the power level Δx obtained from the output of the other three elements, and the shift amount Δx ′ of the one of the three elements is This is the shift amount of the other three elements.

Images