JPH0240516A - Spectrophotometer - Google Patents

Abstract

PURPOSE:To obtain a spectrophotometer characterized by a simple controlling method, excellent linearity and a broad dynamic range by constituting the apparatus with a spectroscopic filter, a plurality of solid-state image sensing element and an output selecting circuit. CONSTITUTION:A spectroscopic filter 1 is formed so that a spectrum- transmitting main wavelength is continuously changed in the direction X and the wavelength is uniform in the direction Y. For the filter 1, a CCD solid-state image sensing element 7 having high sensitivity picture elements and a CCD solid-state image sensing element 8 having low sensitivity picture elements are arranged. An output selecting circuit 19 selects the output of the CCD solid- state image sensing element 7 which is imparted through a line 17 and the output of the CCD solid-state image sensor 8 which is imparted through a line 18 and outputs the signal. When the amount of the incident light is small, the output of the image sensing element 7 having the high sensitivity is selected. When the amount of the incident light is large, the image sensing element 8 having the low sensitivity is selected. The signal can be taken out in this way. Therefore, the dynamic range of this apparatus is expanded.

Description

[Detailed Description of the Invention] The present invention relates to a spectroscopic measurement device, and more specifically, it spatially disperses light of each wavelength of incident light, and
The present invention relates to a spectroscopic measurement device in which a solid-state image sensor receives the dispersed light of each wavelength.

Traditionally, spectroscopic measurements have often been performed using devices that combine a spectroscopic element and a light-receiving element. A grating, a spectral filter, etc. are used as the spectral element, and a spectral filter is suitable for downsizing the device.

On the other hand, as a light receiving element, an SPD (silicon photodiode) array or a self-scanning image sensor is used. In the case of SPD, the dynamic range of photoelectric conversion is very wide and is sufficient for optical measurement. However, when using SPDs arranged in an array for spectroscopic measurements, it is necessary to connect an amplifier circuit to each SPD, and if you want to perform spectroscopy at a deep blue pitch, the amplifier circuit requires a huge amount of space. The disadvantage is that the amount is large. In this regard, COD
Self-scanning image sensors such as the above have the advantage that even if the number of pixels increases, there is only one output section, so only one amplifier circuit is required. However, solid-state imaging devices such as CCDs generally have a narrow dynamic range and may require means such as integration time switching, which makes them difficult to use.

Japanese Patent Laid-Open No. 59-92318 discloses the use of COD as a solid-state imaging device for spectroscopic measurements, but it does not disclose anything about expanding the dynamic range. Furthermore, although JP-A No. 62-76765 describes expanding the dynamic range, the dynamic range expansion method proposed there provides a knee characteristic for floating diffusion in the output section to prevent saturation. It's something you try to suppress. Japanese Patent Publication No. 63-6392
No. 8 attempts to expand the dynamic range by changing the integration time and capturing images multiple times.

Since the above-mentioned Japanese Patent Application Laid-Open No. 59-92318 provides only a narrow dynamic range, it cannot be said to be sufficient for highly accurate spectroscopic measurements. JP-A No. 62-76765 not only has the disadvantage of losing output linearity, but also
There is a drawback that uniform characteristics cannot be obtained due to variations between elements. Furthermore, Japanese Patent Application Laid-Open No. 63-63928 has the disadvantage that the control method is complicated and that it cannot cope with instantaneous changes in light.

The present invention has been made in view of the above points, and an object of the present invention is to provide a spectroscopic measurement device that has a simple control method and can obtain a wide dynamic range with good linearity.

The spectroscopic measurement device of the present invention that achieves the above 6 objects in order to perform the above-mentioned tasks includes a spectroscopic means for spatially dispersing the light of each wavelength of the incident light, and a light-receiving section for each of the dispersed light of each wavelength. It is comprised of a plurality of solid-state imaging devices having different sensitivities arranged to receive light as shown in FIG.

In that case, the output selection means can be configured to extract the output of a highly sensitive solid-state image sensor as long as it is not saturated.

Furthermore, in order to eliminate the influence of device variations, it is preferable to make the dynamic range of a solid-state image sensor with high sensitivity and the dynamic range of a solid-state image sensor with low sensitivity to be small (or to partially overlap with each other).

Effect: According to this configuration, when the amount of incident light is small, the output of the image sensor with high sensitivity can be selected, and when the amount of incident light is large, the output of the image sensor with low sensitivity can be selected and extracted. The dynamic range will be expanded. The dynamic range can be further expanded by providing a large number of image pickup elements with different sensitivities and increasing the difference between the minimum sensitivity and the maximum sensitivity.

Note that when selecting the output of the solid-state image sensor, if the output of the solid-state image sensor with high sensitivity is selected as long as it is not saturated, an output that is not affected by noise can be obtained.

Furthermore, if the dynamic range of a high-sensitivity solid-state image sensor and the dynamic range of a low-sensitivity solid-state image sensor overlap at least in part, it can be used even if the saturation output level fluctuates due to manufacturing variations. The area can be secured reliably.

Embodiments Hereinafter, embodiments of the present invention will be described with reference to the drawings. 1st
The figure shows a conceptual diagram of the spectral sensor used in this example. In the figure, the thickness of the spectral filter (1) in the X direction is gradually increased along the X direction so that the dominant wavelength of spectral transmission changes continuously along the X direction, and the thickness of the spectral filter (1) is gradually increased along the Y direction. The thickness is formed to be constant so that the dominant wavelength of spectral transmission is uniform. This spectral filter (1
) A linear output CCD solid-state image sensor (2) having a plurality of pixels divided in the X direction is disposed in the light receiving section. Light (3) incident on the spectral sensor from the X direction is separated by the spectral filter (1), then reaches the CCD solid-state image sensor (2), where it is photoelectrically converted by each pixel of the CCD solid-state image sensor (2). Ru.

FIG. 2(a) shows a cross-sectional configuration diagram of the spectroscopic sensor taken along the X direction, and FIG. 2(b) shows the output characteristics of each pixel. In FIG. 2(a), incident light (3) enters the spectral filter (1). Pixel of CCD solid-state image sensor (4a
) is passed through the spectral filter (1) located above it.
), and the light incident on the pixel (4b) is light with a dominant wavelength λ2 that has been separated by the spectral filter (1). The output of each pixel is shown in FIG. 2(b), where the output (5a) for pixel (4a) corresponds to the spectral output of light with dominant wavelength λ1, and the output (5b) for pixel (4b) corresponds to the spectral output of light with dominant wavelength λ2. corresponds to the spectral output of light.

Next, the principle of expanding the dynamic range in this embodiment will be described. Figure 3 (a) shows the relationship between the amount of incident light and the output of a conventional CCD solid-state image sensor. Region ■ corresponds to the case where the amount of light is extremely small, and only the dark output appears, and the output is constant. This is the output. This region cannot be used as a spectral sensor because no output corresponding to the intensity of light can be obtained. On the other hand, region ■ corresponds to a case where the amount of light is extremely large, and the charge is accumulated to the maximum amount that can be accumulated in the pixel and is saturated, so the output is still constant, and like region I, it is used as a spectroscopic sensor. This is an unusable area. Therefore, in the end, the area that can be used as a spectroscopic sensor is only ■, and this width is the dynamic range, but this width is usually about 2 to 3 digits and often does not satisfy the specifications required for spectroscopic measurement. .

In this embodiment, the dynamic range is expanded by using pixels with different sensitivities as shown in FIG. 3(b). Graph (6a) in Figure 3(b)
represents the output of a pixel with high sensitivity, and graph (6b) represents the output of a pixel with low sensitivity.

In this case, the area (2) is an area where both high-sensitivity pixels and low-sensitivity pixels are output during dark, and is an unusable area. The area 11-a is a dark-time output area for low-sensitivity pixels, but it is a usable area for high-sensitivity pixels where an output corresponding to the intensity of incident light can be obtained. Region (b) is a usable region where both low-sensitivity pixels and high-sensitivity pixels can obtain outputs corresponding to the intensity of incident light.

Here, the usable area of the two pixels overlaps so that it can be used by both the low-sensitivity pixel and the high-sensitivity pixel.The reason is to ensure the usable area even if the saturation output level fluctuates due to process variations. This is to do so. The area 11-c is a saturated output area for high-sensitivity pixels, but is a usable area for low-sensitivity pixels where an output corresponding to the intensity of incident light can be obtained. Area (2) is an area where both high-sensitivity pixels and low-sensitivity pixels have saturated outputs, and is an unusable area. As is clear from comparing Figures 3 (a) and (b), by using high-sensitivity pixels and low-sensitivity pixels, it is possible to expand the area that can be used as a spectroscopic sensor, and it can also be used for highly accurate spectroscopic measurements. A spectral sensor is obtained.

Next, a configuration in which high-sensitivity pixels and low-sensitivity pixels are used together will be described. FIG. 4 shows a first embodiment. Spectral filter (1
) is designed so that the spectral transmission dominant wavelength changes continuously in the X direction, and is uniform in the X direction, as in FIG. A CCD solid-state image sensor (7) having high-sensitivity pixels and a CCD solid-state image sensor (8) having low-sensitivity pixels are arranged for this spectral filter (1).

Here, (9) and (13) are overflow drains, respectively, and photodiodes (10) and (
14) has the function of discharging excess charge that is generated and can no longer be stored.

(11) and (15) are transfer gates, respectively, and have the function of transferring the charges generated and accumulated in the photodiodes (10) and (14) to the transfer registers (12) and (16).

The transfer registers (12) and (16) transfer the transferred output charges of each photodiode and sequentially output them. As is clear from the figure, the location X0 where the spectral filter (1) is located
The dominant wavelength of spectral transmission λ. The light enters both the photodiode (10) and the photodiode (14). In this embodiment, by making the area of the photodiode (10) larger than the area of the photodiode (14), the pixels of the CCD solid-state image sensor (7) are made to have high sensitivity, and the pixels of the CCD solid-state image sensor (8) are made to have low sensitivity. ing.

(19) is an output selection in which the output of the CCD solid-state image sensor (7) given through the line (17) and the output of the CCD solid-state image sensor (8) given through the line (18) are selected and outputted by a method described later. It is assumed that the circuit includes an element having an arithmetic function, such as a microcomputer. (20) is an output terminal as a spectral sensor (21). Figure 5 shows each CCD solid-state image sensor (7)
(8), in which (a) is the output of the CCD solid-state image sensor (7) with high-sensitivity pixels, and (b) is the output of the CCD solid-state image sensor (8) with low-sensitivity pixels. is shown. Here, V and H are high-sensitivity CCD solid-state image sensors (
7) represents the i-th pixel output. Similarly, V i L
represents the i-th pixel output of the low-sensitivity CCD solid-state image sensor (8).

Now, the light in the area ub shown in Fig. 3 is incident on the i-th pixel of each CCD solid-state image sensor (7) (8) (i + 1)
The light in the area If-c shown in FIG. 3 is incident on the pixel number i.
The method of selecting data by the output selection circuit (19) will be explained with reference to the flowchart of FIG. 6, assuming that the light of the area 1-a shown in FIG. 3 is incident on the +2) pixel. First, in step (I1), i=1 as the pixel number.
Set. Next, in step (111), the output ViL of the low-sensitivity pixel is read, and in step (1112), the outputs ■ and H of the high-sensitivity pixel are read. Next, in step (12), the output ViH of the high-sensitivity pixel and the saturated output value V, □ are compared.

This saturated output value V sit shall be measured in advance and kept in stock.

In the case of the output shown in FIG. 5, since V, , <V, □, the process proceeds to step (114). In step (I4), the high-sensitivity pixel output V i H is divided by the sensitivity Rib of the high-sensitivity pixel to normalize the data. At this time, V i L has an output equal to or higher than the dark output V air; however, ■i>v! L
Therefore, option (2), which is less susceptible to noise, etc., is selected. Next, proceeding to step (I5), it is checked whether the above selection has been completed for all pixels.

If NO here, the pixel number is incremented by one in step (I6) and the process returns to step (I2). Step (
Next, in step I2), a comparison is made between Vfi and +111 and Vt□.

In the case of the output shown in Fig. 5, V (i.
3) is the low sensitivity pixel output Vli. Data is normalized by dividing I, L by the sensitivity R (i, 1.L) of the low-sensitivity pixel. By repeating the above process, all pixels of the CCD solid-state image sensor (7) (8) Data is obtained. Each data is in the areas II-a, I[-b,
Since it is compatible with II-c, data with an extremely wide dynamic range can be obtained, making it effective as a spectroscopic sensor for high-precision measurement.

Next, a second embodiment will be described using FIGS. 7 and 8. This embodiment is configured so that a transfer register is shared by a high-sensitivity CCD solid-state image sensor and a low-sensitivity CCD solid-state image sensor. In Figure 7, (22) (
28) is an overflow drain, (23) is a photodiode that constitutes a high-sensitivity CCD solid-state image sensor (31), (27) is a photodiode that constitutes a low-sensitivity CCD solid-state image sensor (32), (24) (26) are respective transfer gates, and (25) is a common transfer register. FIG. 8 shows a detailed view of the COD solid-state image sensor portion of the embodiment shown in FIG. 7.

In FIG. 8, the pitch of the transfer register (25) is a percentage of the pitch of the photodiodes (23) and (27), and the number of stages is the sum of the number of high-sensitivity photodiodes and the number of low-sensitivity photodiodes. (A1) to (A,) indicate high sensitivity photodiodes, and (B1) to (B,) indicate low sensitivity photodiodes. (33) is an aluminum film for light shielding, and is a transfer register (25).

It covers the transfer gates (24) (26) and a portion of each photodiode. When a bias is applied to the transfer gates (24) and (26), the charges generated and accumulated in the photodiode (A1) are transferred to one stage (C) of the transfer register (25).
1). Also, the charge generated and accumulated in the photodiode (B,) is transferred to the other one of the transfer registers (25).
(C3°). Photodiode (A2)
The same applies to ~(8,) and (B2) ~(B,). In the manner described above, all charges of each photodiode are transferred to the transfer register (25). Thereafter, the transfer register (25) transfers the charges of each stage and sequentially outputs them. Furthermore, thereafter, the output selection circuit (29) selects and normalizes the output data of each photodiode in the same manner as in the first embodiment, and only the selected output is delivered to the output terminal (30). According to the second embodiment, since only one CCD transfer register is required, the chip size reduction yield can be improved, and the number of external circuits can be reduced, leading to cost reduction.

In each of the embodiments described above, the transfer register is driven in two or four phases, but any other number of phases may be used. Further, in this embodiment, charges are directly read out from the photodiode, but a charge storage section formed of an MO3 capacitor may be provided between the photodiode and the transfer gate, or other structural requirements may be added.

Furthermore, although the present embodiment is used for spectroscopic measurement, it can also be used for other optical measurements.

Further, in this embodiment, a spectral filter is used as the spectral element, but of course a grating or another spectral element may also be used.

As described above, according to the present invention, the dynamic range is expanded by using a plurality of solid-state image sensors with different sensitivities and by selecting their outputs, so there is no charge accumulation as in the conventional case. The control method is simpler than those that control time, and it has the advantage of being able to respond to instantaneous changes in light. Also, you can expect good nearness.

At this time, if the output selection means is configured to take out the output of the solid-state image pickup device, which has high sensitivity as long as it is not saturated, it is possible to obtain an output that is not affected by noise.

In addition, if the dynamic range of a solid-state image sensor with high sensitivity and the dynamic range of a solid-state image sensor with low sensitivity overlap at least in part, 1. Even if the level of saturated output changes due to manufacturing variations, a usable range can be ensured.

[Brief explanation of the drawing]

FIG. 1 is a conceptual diagram of a spectral sensor used in the present invention,
FIG. 2 is a diagram explaining the principle. FIG. 3 is a diagram for explaining the dynamic range expansion method according to the present invention in comparison with a conventional example. FIG. 4 is a configuration diagram of a spectroscopic measurement device embodying the present invention, FIG. 5 is a diagram showing an example of the output of each solid-state image sensor, and FIG. 6 is a flowchart showing the operation flow of the output selection circuit. FIG. 7 is a block diagram of another embodiment of the present invention, and FIG. 8 is a block diagram of its solid-state image sensor section. (1) -m-spectral filter. (7) (8) (31) (32)-, CCD solid-state image sensor. (10) (23) --- High sensitivity photodiode (16
) (27) - One low sensitivity photodiode. (11) (15) (24) (26)・Transfer gate. (12) (25)--One transfer register. (19(29)--Output selection means. Figure 1 Figure 2

Claims (3)

[Claims] (1) A spectroscopic means that spatially disperses light of each wavelength of incident light, and a plurality of solid-state image sensors with different sensitivities arranged so that each light receiving part receives the dispersed light of each wavelength in the same way. , and output selection means for selectively extracting the outputs of the plurality of solid-state image sensors according to the light of each wavelength. (2) The spectroscopic measurement device according to claim 1, wherein the output selection means extracts the output of a highly sensitive solid-state image sensor as long as it is not saturated. (3) The spectroscopic measurement device according to claim 1, wherein the dynamic range of the solid-state image sensor with high sensitivity and the dynamic range of the solid-state image sensor with low sensitivity overlap at least in part.