JP5943853B2 - TDI linear image sensor and driving method thereof - Google Patents

Description

  The present invention relates to a linear image sensor used in fields such as remote sensing and a driving method thereof.

  Many image sensors have been developed in which a large number of photodetectors are arranged in an array on a semiconductor substrate, and signal charge readout circuits and output amplifiers from the photodetectors are provided on the same substrate. Also, in remote sensing, which is a means for remotely measuring an object, for example, imaging the ground surface from an artificial satellite, a linear image sensor in which photodetectors are arranged in a one-dimensional array is mounted on the artificial satellite, etc. A two-dimensional image of the earth's surface is photographed by making the direction perpendicular to the direction coincide with the traveling direction of the satellite.

  In order to improve the image resolution, it is desirable to reduce the pixel pitch as much as possible. However, there is a problem that the incident light amount is reduced by the reduction in the area of the photodetector and the S / N is deteriorated. As a clever means for improving the S / N, a TDI (Time Delay and Integration) image sensor has been developed. The TDI method uses an FFT (full frame transfer) CCD (Charge Coupled Devices), which is a two-dimensional image sensor, to integrate the signal charge by synchronizing the charge transfer timing with the movement timing of the subject image, This is a readout method for improving S / N. In the case of remote sensing, the TDI operation can be realized by matching the vertical charge transfer in the image with the moving speed of the satellite. When an M-stage TDI operation is performed with a vertical CCD, the accumulation time is effectively M times, so that the sensitivity is improved M times and the S / N is improved to √M times.

In the case of a general two-dimensional image sensor, the charge transfer in the vertical direction is performed during the horizontal blanking period, and one stage of vertical transfer is performed every time one horizontal line signal is read. However, when this transfer method is applied to a TDI linear image sensor used for remote sensing, there arises a problem that an obtained image is blurred, that is, MTF (Modulation Transfer Function) is deteriorated. On the other hand, if the vertical transfer is performed by dividing into a plurality of times in the horizontal transfer period, the MTF deterioration is suppressed, but there arises a problem that spike-like coupling noise is superimposed on the signal output due to the interference of the vertical transfer clock.
In order to solve this problem, for example, Patent Document 1 discloses a method of temporarily stopping the horizontal transfer clock for a certain period before and after the rising / falling of the vertical transfer clock.

Japanese Patent No. 3866699

    However, according to the method for avoiding the coupling noise shown in Patent Document 1, since the horizontal transfer is stopped a plurality of times while the signal charge is horizontally transferred, it is transferred from one end of the horizontal CCD to the other end. The time required may vary depending on the charge packet. As a result, there is a problem that the dark output generated in the horizontal CCD becomes uneven for each charge packet, and the output FPN (Fixed Pattern Noise) increases.

  The present invention has been made to solve the above-described problems, and suppresses MTF deterioration without increasing the FPN and avoids the influence of coupling noise caused by a vertical transfer clock. An object is to provide a sensor and a driving method thereof.

In order to achieve the above object, the present invention is configured as follows.
In other words, the TDI linear image sensor according to an aspect of the present invention includes an oblique readout unit and a charge storage unit each including a plurality of vertical CCDs in the horizontal direction between the pixel region and the horizontal CCD. The pixel area is divided into a plurality of pixel blocks in the horizontal direction, and the number of horizontal CCD stages is larger than the number of horizontal pixels. Further, in the connection portion between the oblique readout section and the horizontal CCD, a plurality of CCDs are arranged between the pixel blocks as non-connection transfer sections in the horizontal direction where the oblique readout section is not connected.
In addition, the TDI linear image sensor driving method according to the above aspect uses the TDI linear image sensor according to the above aspect, performs vertical transfer divided into a plurality of times during the horizontal transfer period, and in the signal output blank period. The vertical transfer clock is switched.

In other words, the TDI linear image sensor according to an aspect of the present invention includes a plurality of photoelectric conversion units that form pixels and transfer pixel signal charges in a vertical transfer direction by a time delay integration operation. A TDI linear image sensor including a horizontal transfer unit that transfers pixel signal charges sent out by the photoelectric conversion unit in a horizontal transfer direction;
A charge transfer unit that is provided between the photoelectric conversion unit and the horizontal transfer unit and that transfers pixel signal charges from the photoelectric conversion unit to the horizontal transfer unit;
The photoelectric conversion unit and the charge transfer unit are divided into N blocks in the horizontal transfer direction, where N is a positive integer of 2 or more,
The horizontal transfer unit includes a connection transfer unit to which a block in the charge storage unit is connected, and a non-connection transfer unit that is not connected to any of the two adjacent blocks in the charge transfer unit. The non-connection transfer unit and the connection transfer unit are alternately arranged in the transfer direction, and the pixel signal charge transfer in the horizontal transfer direction is continuously performed without pausing, where the non-connection transfer unit is A pixel blanking period that is arranged correspondingly between adjacent blocks and in which the pixel signal charge does not exist within one imaging period;
It is characterized by that.

According to the TDI linear image sensor in one aspect of the present invention, the photoelectric conversion unit and the charge transfer unit are divided into N blocks in the horizontal transfer direction, and the connection transfer unit and the non-connection transfer unit are alternately arranged in the horizontal transfer unit. The non-connection transfer unit generates a pixel blanking period. Therefore, according to the TDI linear image sensor, it is not necessary to temporarily stop horizontal CCD transfer in the horizontal transfer unit in order to provide a pixel blanking period between signal output periods of a certain pixel block and the next pixel block. . Therefore, the horizontal CCD dark output component of each pixel signal becomes constant, and an increase in FPN is suppressed.
Further, according to the driving method of the TDI linear image sensor, the vertical transfer clock in the photoelectric conversion unit is switched in the pixel blanking period of the signal output from the horizontal transfer unit, so that the influence of coupling noise due to the vertical transfer clock can be avoided. .

1 is an element plan view showing a schematic configuration of a TDI linear image sensor according to a first embodiment of the present invention. FIG. 2 is a layout diagram in which a part of a pixel area of the TDI linear image sensor shown in FIG. 1 is enlarged. 2 is an output waveform schematically showing output for one horizontal line when uniform light is incident on the entire pixel region in the TDI linear image sensor shown in FIG. FIG. 2 is a timing diagram schematically showing timings of a vertical transfer clock and a signal output period for explaining a method of driving the TDI linear image sensor shown in FIG. 1. It is an element top view which shows schematic structure of the TDI system linear image sensor by Embodiment 2 of this invention. FIG. 6 is a timing chart schematically showing timings of a vertical transfer clock and a signal output period for explaining a method of driving the TDI linear image sensor shown in FIG. 5. FIG. 2 is a layout diagram mainly showing a charge transfer unit in the TDI linear image sensor shown in FIG. 1. FIG. 2 is a timing diagram for explaining a method of driving the TDI linear image sensor shown in FIG. 1. It is a figure which shows the CCD transfer method of the TDI system linear image sensor shown in FIG. FIG. 6 is a layout diagram illustrating a planar structure of the entire element of a conventional TDI image sensor. FIG. 11 is a layout diagram in which a part of a pixel region of the TDI image sensor shown in FIG. 10 is enlarged. It is a figure which shows the CCD transfer method of the conventional TDI system linear image sensor. It is a figure which shows the relationship between the driving method of the conventional TDI system linear image sensor, and (phi) V coupling noise.

  A TDI linear image sensor and a driving method thereof according to an embodiment of the present invention will be described below with reference to the drawings. In each figure, the same or similar components are denoted by the same reference numerals. In addition, in order to avoid the following description from becoming unnecessarily redundant and to facilitate understanding by those skilled in the art, a detailed description of already well-known matters and a duplicate description of substantially the same configuration may be omitted. . Further, the contents of the following description and the accompanying drawings are not intended to limit the subject matter described in the claims.

In order to facilitate understanding of the above-described embodiment, first, the configuration and driving method of a conventional TDI image sensor will be described with reference to the drawings.
FIG. 10 is a layout diagram showing a planar structure of the entire element of a conventional TDI image sensor using a four-phase drive CCD as a vertical CCD. FIG. 11 is an enlarged layout view of a part of the pixel area of the TDI image sensor shown in FIG.
In the layout of FIG. 10, the pixels 1 are arranged in a two-dimensional array on the surface of the Si substrate. Here, the signal charges integrated with time delay by the TDI method are transferred toward the horizontal CCD 2 in the vertical direction (downward in the drawing), and further transferred in the horizontal direction (to the right in the drawing) by the horizontal CCD 2 to be output by the output amplifier 6. Read from. A charge storage unit 4 is provided between the pixel 1 and the horizontal CCD 2. Further, a charge discharge drain 5 for discharging unnecessary charges is provided on the opposite side (upper side of the drawing) of the horizontal CCD 2 with the array of pixels 1 interposed therebetween. In the example of FIG. 10, a four-phase drive CCD is used for vertical TDI transfer.

  As shown in FIG. 11, transfer electrodes 9a, 9b, 9c and 9d (sometimes collectively referred to as transfer electrodes 9) made of polysilicon are arranged on a Si substrate, and transfer channels (see FIG. Not shown). This transfer channel is electrically isolated by an isolation region 10 made of an impurity region having a conductivity type opposite to that of the Si substrate. In the four-phase drive CCD, the pixel 1 is formed by a set of four transfer electrodes 9a to 9d. CCD transfer clocks φV1 to φV4 are applied to the transfer electrodes 9a to 9d from the input pins 8a to 8d through the metal wiring 7.

Next, driving of the image sensor configured as described above will be described.
FIG. 12A is a schematic diagram of a cross-sectional structure along the transfer direction of the four-phase drive CCD, and a diagram showing the change in potential of the transfer channel in time series, and FIG. The clock waveform given to the 4-phase drive CCD is shown.
When the transfer clocks φV1 to φV4 shown in FIG. 12B are applied to the transfer electrodes 9a to 9d of the four-phase drive CCD from the input pins 8a to 8d, the potential distribution of the transfer channel at times t1 to t5 is shown in FIG. It becomes like a). Here, the high voltage of the CCD transfer clock is H, and the low voltage is L. As shown in FIG. 12A, the potential well 11 is formed below the transfer electrode 9 to which the voltage H is applied among the transfer electrodes 9. Here, when the signal charge generated by the light incidence on the image sensor is 12, the signal charge 12 is charged to the right side of the drawing as the potential well 11 moves to the right side of the drawing by the transfer operation of the CCD. Transferred. At this time, the TDI (time delay integration) operation can be realized by matching the moving speed of the subject image with the CCD transfer speed, and the signal charge 12 generated by the light incidence by the subject image is a potential well corresponding to the image position. 11 is transferred vertically.

In the case of a general two-dimensional image sensor, charge transfer in the vertical direction is performed during the horizontal blanking period, and one stage of vertical transfer is performed every time one line of signals is read in the horizontal direction.
On the other hand, when this general transfer method is applied to a TDI linear image sensor used for remote sensing, the positional relationship between the subject image projected on the detector surface and the potential well vertically transferred is 1 at the maximum. There is a problem that the line portion shifts and the MTF along the direction of travel of the artificial satellite deteriorates, that is, the image is blurred. This is due to the fact that the movement of the subject image is continuous and smooth, whereas the movement of the potential well by CCD transfer is stepwise.

  In order to suppress the MTF degradation by minimizing the difference between the subject image and the potential well, for example, the effective electrode length of each phase (that is, within the transfer direction length of the transfer electrode) The electrode lengths excluding gate overlap and the like (hereinafter simply referred to as “electrode lengths”) are made equal to each other, and the time t1 to t5 are set at equal intervals in the transfer clock shown in FIG. Good. At this time, the vertical transfer is divided into a plurality of times (in this example, four times) during the signal readout period (that is, the imaging cycle), whereby the difference between the subject image and the potential well can be reduced, and the MTF deterioration is reduced. It is suppressed.

However, if charge transfer in the vertical direction is performed during the signal readout period, another problem such as coupling noise superimposed on the signal output occurs. The pattern will be described below with reference to the drawings.
FIG. 13 is a timing chart schematically showing a vertical transfer clock and signal output timing of a TDI linear image sensor using a four-phase drive CCD for vertical transfer. As shown in FIG. 13A, vertical transfer is performed four times at a quarter step in one imaging cycle, thereby suppressing degradation of MTF. At this time, as shown in FIG. 13B, the output waveform of the sensor spikes in the signal output due to interference of these drive clocks because the vertical transfer clock is switched at times N1, N2, and N3 during the signal output period. The problem of overlapping coupling noise occurs.

As a countermeasure against such a problem, Patent Document 1 proposes a method for avoiding coupling noise superimposition by temporarily suspending the horizontal transfer clock for a certain period before and after the rising / falling of the vertical transfer clock. Yes.
However, according to the method for avoiding the coupling noise shown in Patent Document 1, since the horizontal transfer is stopped a plurality of times while the signal charge is horizontally transferred, it is transferred from one end of the horizontal CCD to the other end. The time required may vary depending on the charge packet. As a result, the dark output generated in the horizontal CCD becomes non-uniform for each charge packet, and the output FPN increases.

  The TDI type linear image sensor and the driving method thereof in the present embodiment described below have a configuration for solving such a problem.

Embodiment 1 FIG.
FIG. 1 is an element plan view showing a schematic configuration of a TDI type linear image sensor 101 according to Embodiment 1 of the present invention. The linear image sensor 101 includes a photoelectric conversion unit 111, a horizontal transfer unit 112, and a charge transfer unit 115 as basic components, and further includes the above-described conventional charge discharge drain 5 and output amplifier 6.
In the element plan view of FIG. 1, the pixels 1 are arranged in a two-dimensional array on the surface of the Si substrate. A portion in which the pixels 1 are arranged in a two-dimensional array corresponds to the photoelectric conversion unit 111, and pixel signal charges are transferred in the vertical transfer direction by a time delay integration (TDI) operation. The photoelectric conversion unit 111 includes the metal wiring 7 and the input pin 8 described above. In the example of Embodiment 1 of the present invention, the photoelectric conversion unit 111 uses a four-phase drive CCD for vertical drive, and the pixel 1 is composed of a set of four transfer electrodes, as in the case described above. The four-phase CCD transfer clocks φV1 to φV4 are supplied from the input pins 8a to 8d through the metal wiring 7. As described above, the vertical transfer direction corresponds to the moving direction of the image sensor 101.

FIG. 2 is an enlarged layout view of a part of the pixel area of the TDI linear image sensor 101 shown in FIG. The pixel layout of the TDI linear image sensor 101 shown in FIG. 2 is the same as that of the conventional case shown in FIG.
As shown in FIG. 2, transfer electrodes 9a to 9d made of polysilicon are alternately arranged on a Si substrate, and a transfer channel (not shown) is formed thereunder. The transfer channel is electrically isolated by an isolation region 10 made of an impurity region having a conductivity type opposite to that of the Si substrate. As described above, in the four-phase driving CCD, the pixel 1 is formed by a set of four transfer electrodes 9a to 9d. CCD transfer clocks φV1 to φV4 are applied to the transfer electrodes 9a to 9d from the input pins 8a to 8d through the metal wiring 7. The vertical transfer operation is the same as the vertical transfer operation of the conventional TDI linear image sensor.

  Therefore, in the photoelectric conversion unit 111, the signal charge of each pixel, which is time delay integrated by the TDI method, is transferred in the vertical direction (downward in the drawing) toward the horizontal transfer unit 112. The transferred pixel signal charge is transferred to the horizontal transfer unit 112 via the charge transfer unit 115, further transferred in the horizontal direction (right side in the drawing) by the horizontal transfer unit 112, and read from the output amplifier 6. Become.

  Further, in the linear image sensor 101 according to the first embodiment, the pixels 1 arranged in a two-dimensional array in the photoelectric conversion unit 111 are divided and divided into four pixel blocks 111a, 111b, 111c, and 111d in the horizontal transfer direction. Yes. The number of blocks to be partitioned and separated is not limited to four in the present embodiment, and may be a positive integer value of 2 or more. Further, in relation to the number of phases of the CCD transfer clock for driving the photoelectric conversion unit 111, the number of blocks can be set to be equal to or greater than the number of clock phases, for example.

In the linear image sensor 101, a charge transfer unit 115 is provided between the photoelectric conversion unit 111 and the horizontal transfer unit 112.
The charge transfer unit 115 includes an oblique readout unit 113 composed of a plurality of stages of vertical CCDs, and a charge storage unit 114.
Here, with respect to the oblique readout unit 113 and the charge storage unit 114, four sets of the oblique readout unit 113a and the charge storage unit 114a, in the horizontal transfer direction, corresponding to the four pixel blocks 111a to 111d of the photoelectric conversion unit 111. The oblique readout unit 113b and the charge storage unit 114b, the oblique readout unit 113c and the charge storage unit 114c, and the oblique readout unit 113d and the charge storage unit 114d are divided. Therefore, if the number of divisions in the photoelectric conversion unit 111 changes, the number of divisions in the oblique readout unit 113 and the charge storage unit 114 is changed accordingly.
The charge storage units 114a to 114d are connected to the horizontal transfer unit 112, respectively.

The horizontal transfer unit 112 is composed of a CCD, and the horizontal CCD 112b to which the pair of the oblique readout unit 113 and the charge storage unit 114 is not connected is between the horizontal CCD 112a to which the pair of the oblique readout unit 113 and the charge storage unit 114 is connected. In this embodiment, a plurality of stages are formed. That is, the horizontal CCD 112b is connected to both adjacent blocks in the charge transfer unit 115, for example, the blocks of the oblique readout unit 113a and the charge storage unit 114a and the oblique readout unit 113b and the charge storage unit 114b. This corresponds to a non-connection transfer unit that does not perform (hereinafter may be referred to as “non-connection transfer unit 112b”), and is arranged corresponding to adjacent blocks. As will be described later, the non-connection transfer unit 112b generates a pixel blanking period in which no pixel signal charge exists within one imaging cycle. Further, in the present embodiment, a plurality of non-connection transfer units 112b are provided, but at least one may be used according to the number of pixel blocks in the photoelectric conversion unit 111.
The horizontal CCD 112a to which the set of the oblique readout unit 113 and the charge storage unit 114 is connected may be referred to as a connection transfer unit 112a.
In this manner, the connection transfer unit 112a and the non-connection transfer unit 112b are alternately and continuously arranged in the horizontal transfer direction, and constitute one horizontal transfer unit 112 as a whole.

  Further, a charge discharge drain 5 for discharging unnecessary charges is provided at an end (upper side of the drawing) opposite to the horizontal transfer unit 112 of the pixel 1 array.

  FIG. 7 is a layout diagram of the above-described charge transfer unit 115 including the photoelectric conversion unit 111 and the horizontal transfer unit 112 in the TDI linear image sensor 101 according to the first embodiment of the present invention shown in FIG. Show. In particular, FIG. 7 shows details of the layout in the vicinity of the connection portion from the two adjacent oblique readout portions 113 and the subsequent two charge storage portions 114 to the connection transfer portion 112a and the non-connection transfer portion 112b of the horizontal CCD. FIG.

In FIG. 7, photoelectric conversion is performed in the pixel 1 portion of the photoelectric conversion unit 111. Although not shown, light from the subject is not incident on the oblique reading unit 113, the charge storage unit 114, the horizontal transfer unit (horizontal CCD) 112, and other peripheral circuit regions other than the pixel 1. A light shielding film is formed on the element surface side.
In the pixel 1, a region sandwiched between the separation regions 10 forms a signal charge transfer channel. In the oblique reading unit 113, the non-connection transfer unit 112b can be formed between the connection transfer units 112a of the two sets of horizontal CCDs by bending the transfer channel and laying it in an oblique direction. In general, a two-layer drive CCD is often used as the horizontal CCD. In the example of FIG. 7, two transfer electrodes φH1 and φH2 are used as one set (four gates as gates) to form a single horizontal CCD. ing. In this example, one set of unconnected transfer units 112b is formed of five stages of horizontal CCDs.
The connection transfer unit 112a in the horizontal CCD is formed by one horizontal CCD per pixel. That is, when the photoelectric conversion unit 111 is divided into a plurality of blocks, the number of horizontal pixels per block is equal to the number of horizontal CCD stages of the connection transfer unit 112a of the horizontal CCD connected thereto. Therefore, when viewed as the whole horizontal CCD, the total number of horizontal CCD stages is larger than the total number of horizontal pixels by the number of stages of the non-connection transfer unit 112b.

The operation, that is, the driving method of the TDI type linear image sensor 101 according to the first embodiment configured as described above will be described below.
As already described, the photoelectric conversion unit 111 transfers the pixel signal charge of each pixel 1 in the vertical transfer direction by the time delay integration operation in accordance with the supply of the four-phase CCD transfer clock in this embodiment. The charge transfer unit 115 sends the transferred pixel signal charge to the horizontal transfer unit 112. Since the horizontal transfer unit 112 is provided with the non-connection transfer unit 112b between the connection transfer units 112a as described above, the pixel signal charge of one horizontal line is divided into four blocks and output. Thus, an image blanking period occurs between the blocks. The pixel signal charge output operation will be described in more detail below with reference to FIGS.

  FIG. 3 is an output waveform schematically showing output for one horizontal line when uniform light is incident on the entire pixel area in the TDI linear image sensor 101 shown in FIG. The signal from the pixel 1 is read out from the output amplifier 6 in time series. As described above, the photoelectric conversion unit 111 and the charge transfer unit 115 are divided into a plurality of blocks in the horizontal transfer direction, and the horizontal transfer unit 112 The connection transfer unit 112a and the non-connection transfer unit 112b are alternately and continuously arranged in the horizontal transfer direction. Accordingly, the pixel signal charges of one horizontal line are divided into four blocks 111a, 111b, 111c, and 111d and output in this example, and an image blanking period 152 occurs between the blocks. Here, the output signals 151a, 151b, 151c, and 151d are output signals that are generated in the pixel blocks 111a, 111b, 111c, and 111d, respectively, and are time-delay integrated. That is, the signal 151d, blanking, signal 151c, blanking, signal 151b, blanking, during one imaging cycle, that is, until the reading of pixel signals of one horizontal line is completed, that is, during one horizontal transfer period. Pixel signals are output in time series from the side closer to the output amplifier in the order of the signal 151a.

  In this readout operation, the horizontal transfer unit 112 is continuously driven without being temporarily stopped during one imaging cycle, that is, until readout of pixel signals of one horizontal line is completed. Therefore, the dark output component of the horizontal CCD superimposed on each pixel signal is constant. Further, by continuously driving the CCD of the horizontal transfer unit 112, the horizontal transfer clock may be a normal clock, and there is an effect that a complicated pattern is not required.

FIG. 4 is a timing chart schematically showing the timing of the vertical transfer clock and the signal output period for explaining the driving method of the TDI linear image sensor 101 shown in FIG. The TDI linear image sensor 101 according to the first embodiment uses a four-phase drive CCD for vertical transfer, and performs CCD transfer using four sets of vertical transfer clocks φV1, φV2, φV3, and φV4. Here, by setting the timings t1, t2, t3, and t4 at which the high level and the low level of the vertical transfer clock are switched at equal intervals in time, it is possible to suppress the MTF deterioration in the TDI operation.
At this time, as shown in FIG. 4, the pixel block 111a in the photoelectric conversion unit 111 is set so that the times t1, t2, t3, and t4 fall within the pixel blanking period 152 in the pixel signal output output from the output amplifier 6. By setting the division width in the horizontal transfer direction of ~ 111d, it is possible to avoid coupling noise from being superimposed on the pixel signal output.

Next, a signal transfer operation from the photoelectric conversion unit 111 to the horizontal transfer unit 112 will be described.
8 shows a vertical transfer clock (φV1, φV2, φV3, φV4) and a control clock for the charge storage unit 114 (φV1, φV2, φV3, φV4) for driving the TDI linear image sensor 101 according to the first embodiment of the present invention shown in FIG. It is a clock timing diagram showing (φST, φSC). In FIG. 8, t indicated by a dotted line in the drawing represents the time at which the vertical transfer clock is switched, and each corresponds to one of t1 to t4 shown in FIG.
FIG. 9 is a potential diagram along the vertical transfer channel for explaining the charge transfer method of the TDI linear image sensor 101 shown in FIG. In FIG. 9, a <b> 1 to a <b> 8 correspond to the time indicated by the dotted line in FIG. 8, and schematically represent the state of the signal charge transferred to the potential well at each time.
As shown in FIG. 9, the charge transfer of the oblique readout unit 113 is performed in exactly the same way as the vertical transfer of the pixel portion is performed by the vertical transfer clocks φV1, φV2, φV3, and φV4. First, at time a1, potential wells are formed under the φV2 gate and the φV3 gate, and charges are accumulated therein. At this time, a potential well is also formed under the φST gate, but the well is empty. Next, at time a2, the potential wells of the photoelectric conversion unit 111 and the oblique readout unit 113 are shifted below the φV3 gate and below the φV4 gate. At this time, the potential well under the φST gate is continuous with the potential well under the adjacent φV4 gate, and the signal charge spreads over the entire well. Next, at time a3, the potential well is shifted by one gate, and further, at time a4, the charges transferred through the oblique reading unit 113 are stored in the potential well below the storage gate φST. Next, when the accumulation control gate φSC is opened at time a5, the signal charge accumulated under φST moves to the horizontal CCD, and finally, transfer to the horizontal transfer unit 112 is completed at time a8. Although not shown in FIG. 9, the pixel signal for one line is output to the output amplifier by the horizontal CCD operation within the period from the completion of the transfer of the charge to the horizontal CCD until the next time the storage control gate φSC is opened. Is transferred and read.

In the above-described first embodiment, the vertical CCD constituting the photoelectric conversion unit 111 is a four-phase drive TDI type linear image sensor. However, the photoelectric conversion unit 111 is driven with another number of phases such as three-phase drive or five-phase drive. The same applies when a vertical CCD is used. That is, as described above regarding the relationship between the number of phases and the number of blocks, in the case of three-phase driving, the layout and the driving clock may be set so that the pixel region is divided into three or more, for example, three blocks. In the case of phase driving, the pixel region may be divided into five or more, for example, five blocks.
With this configuration, it is possible to increase the number of appearances of the blanking period of the pixel signal formed per imaging cycle than the number of switching timings of the vertical transfer clock per imaging cycle. As a result, the vertical transfer clock switching timing can be set within the pixel blanking period even when a vertical CCD driven by another phase number such as three-phase driving or five-phase driving is used. Superimposition of coupling noise on the output can be avoided.

  As described above, according to the TDI linear image sensor 101 according to the first embodiment, it is not necessary to temporarily stop the transfer operation in the horizontal transfer unit 112 in order to provide the pixel blanking period 152 in the signal output period. The horizontal CCD dark output component of each pixel signal becomes constant, and the increase in FPN is suppressed.

Further, according to the driving method of the TDI linear image sensor 101 according to the first embodiment, the vertical transfer clock is switched in the pixel blanking period 152 of the signal output, so that the coupling by the vertical transfer clock is suppressed while suppressing the deterioration of the MTF. The influence of noise can be avoided.
Therefore, in the imaging system using the TDI linear image sensor 101 and the driving method thereof according to the first embodiment, image correction for removing the influence of φV coupling noise is facilitated, and the image processing unit is downsized. It is also possible to save power.

Embodiment 2. FIG.
FIG. 5 is an element plan view showing a schematic configuration of the TDI type linear image sensor 102 according to the second embodiment of the present invention. FIG. 6 is a timing chart schematically showing a vertical transfer clock and a signal output period timing for explaining a method of driving the TDI linear image sensor 102 shown in FIG.

The difference between the TDI linear image sensor 101 according to the first embodiment described above and the TDI linear image sensor 102 according to the second embodiment is that pixel blocks are divided in the horizontal transfer direction of the pixel area as shown in FIG. In the second embodiment, the pixel block is divided into eight pixel blocks. Other configurations are the same as those of the TDI linear image sensor 101 of the first embodiment.
Although not shown, the pixel layout in the TDI linear image sensor 102 of the second embodiment is the same as the pixel layout in the TDI linear image sensor 101 of the first embodiment, and a four-phase drive CCD is used for vertical transfer. Used.

Next, a driving method of the TDI linear image sensor 102 according to the second embodiment will be described with reference to FIG.
In the driving method of the TDI linear image sensor 102 according to the second embodiment, at least two adjacent vertical transfer clocks among the four vertical transfer clocks φV1, φV2, φV3, and φV4 are completely at a high level at all timings. The vertical transfer clock is driven so that For example, in two sets of a set of vertical transfer clocks φV1 and φV2 and a set of vertical transfer clocks φV2 and φV3, both sets become H level between time t4 and time t5.
With this driving method, the length in the transfer direction of the potential well formed below the transfer gate is always secured by two or more gates, so that the maximum transfer charge amount of the vertical CCD constituting the photoelectric conversion unit 111 can be increased. it can. At this time, the electrode lengths of the four-phase set of transfer gates are made equal to each other, and the vertical transfer clock switching timings t1 to t8 are equally spaced in time to suppress MTF deterioration in the TDI operation. it can.

At this time, as shown in FIG. 6, the timing at which the high level and the low level of the vertical transfer clock are switched appears eight times per imaging cycle, but photoelectric conversion is performed in the horizontal transfer direction according to the element layout shown in FIG. 5. By dividing the unit 111 and the charge transfer unit 115 into eight pixel blocks, the pixel blanking period 152 of the signal output sent out by the output amplifier 6 also appears eight times. Therefore, by setting the division width of the pixel block in the photoelectric conversion unit 111 so that the switching timing of the vertical transfer clock matches the pixel blanking period 152, it is possible to avoid coupling noise from being superimposed on the signal output. .
This method is effective when it is desired to secure the maximum transfer charge amount of the vertical CCD in a TDI linear image sensor with a small pixel pitch.

As described above, according to the TDI linear image sensor 102 according to the second embodiment, an increase in FPN can be suppressed as in the TDI linear image sensor 101 according to the first embodiment.
Further, according to the driving method of the TDI linear image sensor 102 according to the second embodiment, the vertical transfer is performed while suppressing the deterioration of the MTF, similarly to the driving method of the TDI linear image sensor 101 according to the first embodiment. The influence of coupling noise due to the clock can be avoided. Therefore, even in the imaging system using the TDI linear image sensor 102 and the driving method thereof according to the second embodiment, image correction for removing the influence of φV coupling noise is facilitated, and the image processing unit is small. And power saving can be achieved.

  In addition, according to the TDI linear image sensor 102 and the driving method thereof according to the second embodiment, since the frequency of appearance of the signal blanking period 152 per imaging cycle is increased, the maximum transfer charge amount of the vertical CCD Even in the case of a driving method in which the value is increased, the influence of coupling noise due to the vertical transfer clock can be avoided.

1 pixel, 6 output amplifier,
101, 102 TDI type linear image sensor, 111 photoelectric conversion unit,
112 horizontal transfer unit, 112a connection transfer unit, 112b non-connection transfer unit,
113 oblique readout section, 114 charge storage section,
115 charge transfer unit, 152 pixel blanking period.

Claims (5)

A plurality of photoelectric conversion units that form pixels and transfer pixel signal charges in the vertical transfer direction by time delay integration, and a horizontal transfer unit that transfers pixel signal charges sent from the photoelectric conversion units in the horizontal transfer direction TDI linear image sensor,
A charge transfer unit that is provided between the photoelectric conversion unit and the horizontal transfer unit and that transfers pixel signal charges from the photoelectric conversion unit to the horizontal transfer unit;
The photoelectric conversion unit and the charge transfer unit are divided into N blocks in the horizontal transfer direction, where N is a positive integer of 2 or more,
The horizontal transfer unit includes a connection transfer unit to which a block in the charge transfer unit is connected, and a non-connection transfer unit that is not connected to any two blocks in the charge transfer unit. The non-connection transfer unit and the connection transfer unit are alternately arranged in the transfer direction, and the transfer of the pixel signal charges in the horizontal transfer direction is continuously performed without pausing,
Here, the non-connection transfer unit is disposed corresponding to adjacent blocks, and generates a pixel blanking period in which the pixel signal charge does not exist within one horizontal transfer period .
This is a TDI linear image sensor. The time delay integration operation in the photoelectric conversion unit is driven by an M-phase vertical transfer clock signal,
The number of blocks N in the photoelectric conversion unit and the charge transfer unit is equal to or greater than M, where M is a positive integer equal to or greater than 2.
The TDI linear image sensor according to claim 1.   The charge transfer unit is provided corresponding to each block divided in the photoelectric conversion unit, and N oblique readout units to which the blocks are connected, and an oblique readout unit provided corresponding to each oblique readout unit. The TDI linear image sensor according to claim 1, further comprising N charge storage units that send the pixel signal charges that have passed through to the horizontal transfer unit. In the driving method of the TDI type linear image sensor according to any one of claims 1 to 3,
The pixel signal charge generated by the time delay integration operation in the photoelectric conversion unit provided in the linear image sensor is transferred from the photoelectric conversion unit to the horizontal transfer unit via the charge transfer unit,
The horizontal transfer unit is continuously driven, and the horizontal transfer of the output signal in which a pixel blanking period not including a pixel signal is generated between the pixel signal charges transmitted by the divided blocks in the photoelectric conversion unit and the charge transfer unit. Send from the department,
A driving method of a TDI type linear image sensor.   In the vertical transfer clock signal supplied to the photoelectric conversion unit because the photoelectric conversion unit performs a time delay integration operation, the timing at which the voltage level of the vertical transfer clock signal changes is matched within the pixel blanking period. The driving method of the TDI type linear image sensor according to claim 4.

Images