Imaging device

Abstract

To provide an imaging apparatus for which the area efficiency of a semiconductor device substrate is improved while provided with a global shutter function.SOLUTION: An imaging apparatus includes pixels, each having a photoelectric conversion unit, a storage circuit capable of simultaneously storing voltage signals on the basis of electric charges generated by the photoelectric conversion units in all pixels, and having m storage sections (m being an integer of 2 or more) on each pixel, and n (n is an integer smaller than m) AD conversion circuits arranged on each row of pixels, for sequentially AD converting a plurality of voltage signals stored in the m storage sections.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present invention relates to an imaging device. [Background technology]

[0002] In a CMOS solid-state imaging device, a global shutter function that simultaneously accumulates all pixels is known as a function for eliminating distortion of a captured image that occurs when a moving object is captured.

[0003] Meanwhile, in recent years, in CMOS solid-state imaging devices, a backside illumination technology that receives light on the surface opposite to the surface on which the pixel circuits are formed, and a technology for laminating semiconductor substrates to form a stacked structure in the backside illumination imaging device have been developed. Patent Document 1 proposes a structure using this stacked structure to provide a charge retention memory required for the global shutter function for each pixel on a substrate different from the substrate on which the pixel circuits are provided.

[0004] Meanwhile, imaging devices capable of simultaneously obtaining a focus detection signal and an imaging signal using a phase difference method that detects the amount of defocus by dividing the pupil of an imaging optical system and receiving light are becoming widely used. Patent Document 2 discloses that all pixels have multiple photoelectric conversion units so that the pupil of the optical system can be divided under one microlens, and output a signal from at least one photoelectric conversion unit and an imaging signal obtained by adding all the signals from the multiple photoelectric conversion units. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2010-219339 A [Patent Document 2] JP 2013-211832 A Summary of the Invention [Problem to be solved by the invention]

[0006] However, as in Patent Document 1, if it is assumed that in addition to the charge retention memory for each pixel, the AD conversion circuit is also provided on a second semiconductor substrate different from the first semiconductor substrate on which the photoelectric conversion unit is arranged, the area of ​​the second semiconductor substrate increases, reducing the effect of reducing the cost of the chip itself. Also, if multiple photoelectric conversion units are provided in each pixel and configured to be able to output a phase difference focus detection signal and an image signal, not only memory but also an AD conversion circuit is required for the multiple photoelectric conversion units provided in each pixel, further increasing the area of ​​the second semiconductor substrate.

[0007] The present invention has been made in view of the above-mentioned problems, and an object of the present invention is to provide an imaging device having a global shutter function and improving the area efficiency of a semiconductor substrate. [Means for solving the problem]

[0008] An imaging device according to the present invention includes a plurality of pixels each including a plurality of photoelectric conversion units, and outputting a first voltage signal based on charges of some of the plurality of photoelectric conversion units and a second voltage signal based on charges combining the charges of the plurality of photoelectric conversion units, and a plurality of holding circuits provided in a one-to-one correspondence with the plurality of pixels and holding the voltage signals based on the charges generated by the plurality of photoelectric conversion units, each of the plurality of holding circuits including a first holding unit that holds the first voltage signal and a second holding unit that holds the second voltage signal. and an amplifying transistor Equipped with The number of amplifying transistors provided in each of the plurality of holding circuits is less than the number of holding units provided in each of the plurality of holding circuits. It is characterized by: Effect of the Invention

[0009] According to the present invention, it is possible to provide an imaging device having a global shutter function and improving the area efficiency of a semiconductor substrate. [Brief description of the drawings]

[0010] [Figure 1] 2 is a diagram showing the configuration of a light receiving surface of a first substrate in the imaging device according to the first embodiment of the present invention. [Diagram 2] FIG. 2 is a diagram showing the configuration of a surface opposite to a light-receiving surface of a first substrate in the first embodiment. [Diagram 3] FIG. 4 is a diagram showing the configuration of a second substrate in the first embodiment. [Figure 4] 2 is an equivalent circuit diagram showing the configurations of a pixel, a pixel circuit, and a storage circuit according to the first embodiment. [Diagram 5] 1 is an equivalent circuit diagram showing a configuration of an AD conversion circuit according to a first embodiment. [Figure 6] 4 is a timing chart showing a method of driving the imaging device according to the first embodiment. [Figure 7] FIG. 4 is a diagram showing the configuration of a pixel according to a first modified example of the first embodiment. [Figure 8] FIG. 13 is a diagram showing a configuration of a second substrate in a second modified example of the first embodiment. [Figure 9] FIG. 11 is a diagram showing the configuration of a storage circuit according to a second embodiment. [Figure 10] 10 is a timing chart showing a method of driving the imaging device according to the second embodiment. [Figure 11] FIG. 11 is an equivalent circuit diagram showing the configuration of a storage circuit according to a third embodiment. [Figure 12] FIG. 13 is a diagram showing the configuration of a second substrate in the third embodiment. [Figure 13] FIG. 13 is an equivalent circuit diagram showing the configurations of a pixel, a pixel circuit, and a storage circuit according to a fourth embodiment. [Figure 14] FIG. 13 is a diagram showing the configuration of a pixel according to the fifth embodiment. [Figure 15] FIG. 13 is an equivalent circuit diagram showing the configurations of a pixel, a pixel circuit, and a storage circuit according to a fifth embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Hereinafter, the embodiments will be described in detail with reference to the attached drawings. Note that the following embodiments do not limit the invention according to the claims. Although the embodiments describe a number of features, not all of these features are essential to the invention, and the features may be combined in any manner. Furthermore, in the attached drawings, the same reference numbers are used for the same or similar configurations, and duplicated descriptions are omitted.

[0012] <First embodiment> 1 is a diagram showing the configuration of a light receiving surface of a first semiconductor substrate 101 provided with a photoelectric conversion unit (photodiode, hereinafter referred to as PD) in an image pickup device 100 according to a first embodiment of the present invention. The image pickup device 100 of this embodiment has a layered structure in which a plurality of semiconductor substrates including the first semiconductor substrate 101 are layered.

[0013] In Fig. 1, the reference numeral 10(pq) denotes a pixel located in the pth row and the qth column. In Fig. 1, 6 rows and 8 columns of pixels are arranged on the light receiving surface, but in reality, a large number of pixels amounting to tens of millions are arranged two-dimensionally. In addition, in order to receive light by dividing the pupil region of the imaging optical system (not shown), all pixels 10(pq) have two photodiodes PDA and PDB that are decentered with respect to one microlens, as shown in pixel 10(00).

[0014] 2 is a diagram showing the configuration of the surface opposite to the light-receiving surface of the first semiconductor substrate 101 in the imaging device 100 of this embodiment. On this non-light-receiving surface, pixel circuits 11(pq) corresponding to the pixels 10(pq) are two-dimensionally arranged. A voltage signal output from the pixel circuit 11(pq) is transferred to another semiconductor substrate (hereinafter referred to as a second semiconductor substrate) 102 (see FIG. 3) constituting a laminated structure via an electrical contact (hereinafter referred to as CONT) 18 for each pixel.

[0015] The CONT 18 is also connected to the vertical scanning circuit 20 of the second semiconductor substrate 102, and a global shutter operation is realized by supplying a control signal from the vertical scanning circuit 20 to the photodiodes PDA, PDB, etc. on the first semiconductor substrate 101. The drive timing of the photodiodes PDA, PDB, etc. will be described later.

[0016] Here, for example, an N-type silicon substrate is used for the first semiconductor substrate 101. In order to store electrons of electron-hole pairs generated when light is received, the photodiodes PDA and PDB may be N-type semiconductor regions. Also, the photodiodes PDA and PDB are separated from each other by a P-type semiconductor region.

[0017] 3 is a diagram showing the configuration of the second semiconductor substrate 102 of the imaging device 100 of this embodiment. The storage circuits 12(pq) are arranged two-dimensionally in correspondence with the pixel circuits 11(pq) of the first semiconductor substrate 101. The pixel circuits 11(pq) are connected to each other through the contacts CONT18, and therefore the voltage signals output from the pixel circuits 11(pq) can be stored in the storage circuits 12(pq) all at once (all pixels can be stored at once). This makes it possible to realize a global shutter.

[0018] The accumulated voltage signals are scanned row-sequentially by a vertical scanning circuit 20 that scans the storage circuits 12(pq), and are output to n AD conversion circuits ADq0 (n is an integer) provided for each column via a column output line VLq0. The AD conversion circuits ADq0 convert the input signals into AD signals. The AD-converted digital signals are scanned column-sequentially by a horizontal scanning circuit 30, and are output to an output section 50 via a row output line HL.

[0019] The output section 50 includes a known parallel-serial conversion circuit (hereinafter, P / S conversion circuit) and sequentially converts the digital imaging signals output from the row output lines HL into a high-speed serial transmission format such as LVDS. The output section 50 may also include a circuit for performing correction processing for defective pixels, etc.

[0020] In the imaging device 100 of this embodiment, one column output line VLq0 is arranged for each column, and is shared by all rows of the same column. Also, one AD conversion circuit ADq0 is arranged for each column. The pixel signal output to the column output line VLq0 and a reference signal RAMP that varies in proportion to time are input to the AD conversion circuit ADq0.

[0021] The structures of the AD conversion circuit ADq0, the storage circuit 12(pq), the pixel 10(pq), and the pixel circuit 11(pq) will be described later using equivalent circuit diagrams. A timing generating circuit 40 generates predetermined drive timing signals that the vertical scanning circuit 20 provides to the storage circuit 12(pq) in row sequence, as well as control signals for the AD conversion circuit ADq0 and horizontal scanning timing signals. The drive timing will be described later using timing charts.

[0022] Next, FIG. 4 is a diagram showing an example of an equivalent circuit of the pixel 10(pq), the pixel circuit 11(pq), and the storage circuit 12(pq) in the imaging device 100 of this embodiment.

[0023] First, the pixel 10 (pq) has photodiodes PDA and PDB. The pixel circuit 11 (pq) has transfer transistors TXA and TXB that transfer the charges generated in the photodiodes PDA and PDB to a charge-voltage conversion section (floating diffusion section, hereinafter referred to as FD section), a first amplification transistor (hereinafter SF1) that connects the FD section to its gate and configures a source follower circuit with a current source (not shown), a reset transistor RES that resets the FD section with a predetermined power supply VDD, and a batch transfer transistor GS that connects the source of SF1 to its drain.

[0024] Of the above transistors, the gates of the reset transistor RES, the transfer transistors TXA and TXB, and the batch transfer transistor GS can be controlled by the vertical scanning circuit 20. Then, readout control of a signal corresponding to the charge generated only in the photodiode PDA, and readout control of a signal corresponding to the charge generated and added in both the photodiodes PDA and PDB are performed for all pixels at once. Note that an overflow drain may be provided to discharge unnecessary charges from the photodiodes PDA and PDB.

[0025] The source of the batch transfer transistor GS is connected to the storage circuit 12(pq) via the CONT 18. The storage circuit 12(pq) includes m (m is an integer of 2 or more) storage capacitors (storage units) that store the voltage signals of the pixel circuits 11(pq). In this embodiment, the storage circuit 12(pq) includes three storage capacitors CN, CA, and CAB.

[0026] The storage capacitor CN stores the voltage of the FD of the pixel circuit 11(pq) after the reset is released (hereinafter, referred to as the N signal). The storage capacitor CA stores the voltage of the FD that has dropped in response to the signal charge of the photodiode PDA (hereinafter, referred to as the A signal). The storage capacitor CAB stores the voltage of the FD that has dropped in response to the combined signal charge of the photodiodes PDA and PDB (hereinafter, referred to as the imaging signal).

[0027] These storage capacitors may be capacitive elements with an expanded surface area using a trench structure. Also, high-capacitance elements formed between wirings in a portion of the second semiconductor substrate 102 where a high-dielectric-constant material is used between the wiring layers may be used. Furthermore, the transistor gate oxide film of the second semiconductor substrate 102 may be used in part. Increasing the storage capacitance in this way can reduce thermal noise and improve image quality.

[0028] Moreover, the storage circuit 12(pq) includes memory write transistors MWN, MWA, and MWAB for writing voltage signals to the above three storage capacitors, respectively.

[0029] The storage capacitors CN, CA, and CAB are connected to the gates of the second amplification transistors SF2N, SF2A, and SF2AB, respectively, and form a source follower circuit using a current source (not shown). Selection transistors SELN, SELA, and SELAB are provided so that the source voltages of these second amplification transistors can be selectively transmitted to the column output line VLq0. A driving method and sequence for outputting the N signal, A signal, and imaging signal by these selection transistors will be described later using a timing chart, but it goes without saying that these selection transistors are also used for the above-mentioned row-direction scanning.

[0030] Next, FIG. 5 is a diagram showing an example of an equivalent circuit of the AD conversion circuit ADq0 in this embodiment.

[0031] The AD conversion circuit ADq0 includes a comparator COMPq0 that receives the pixel signal VLq output from the column output line VLq0 and a reference signal RAMP as input signals, and a counter COUNTERq0 whose stop is controlled by the output polarity of the comparator COMPq0. The comparator COMPq0 compares the voltages of the pixel signal VLq and the reference signal RAMP, and when the value of the pixel signal VLq is higher, the output polarity becomes Hi, and when the value of the pixel signal VLq is lower, the output polarity becomes Lo. When a reset signal (not shown) is released, the counter COUNTERq0 continues counting when the output polarity of the comparator COMPq0 is Hi, and stops when it is Lo.

[0032] As a result, for example, when the voltage of the pixel signal VLq, which has dropped in response to the signal charge of the photodiode PDA, falls below the reference signal RAMP, which drops in proportion to time, the counting can be stopped, and the voltage of the pixel signal VLq can be AD converted. Although this will be described in detail later using a timing chart, the image pickup signal has a wider voltage range than the A signal, and the A signal has a wider voltage range than the N signal, so even if the AD conversion circuits can be arranged in parallel, it is more time efficient to sequentially AD convert a plurality of types of signals with different voltage ranges using a single AD conversion circuit.

[0033] Next, a method for driving the imaging device of this embodiment will be described. Fig. 6 is a timing chart showing a method for driving the imaging device of this embodiment. In Fig. 6, the symbols PRES, PTXA, PTXB, PGS, PMWN, PMWA, PMWAB, PSELN, PSELA, and PSELAB respectively indicate control signals applied to the gates of the transistors RES, TXA, TXB, GS, MWN, MWA, MWAB, SELN, SELA, and SELAB, where the transistors are turned ON at Hi and OFF at Lo. Also, the symbol RAMP represents the voltage of the reference signal RAMP supplied to the AD conversion circuit ADq0, the symbol COMP represents the output polarity of the comparator COMPq0, and the symbol COUNTER represents the count value of the counter COUNTERq0.

[0034] 6, the period from time t600 to time t611 represents a so-called global shutter operation in which a voltage signal corresponding to the charge accumulated in all pixels is transferred to the second semiconductor substrate 102 in a lump. The period from time t612 to time t623 represents an operation in which the signal of the storage circuit 12(0q) arranged in the 0th row of the second semiconductor substrate 102 is sequentially AD converted. The vertical scanning circuit 20 executes the drive from time t612 to time t623 for the storage circuit 12(0q) arranged in the 0th row, the storage circuit 12(1q) arranged in the 1st row, and so on up to the final 5th row, and signals from all pixels are output.

[0035] First, from time t600 to time t601, the control signal PRES is set to Hi to reset the FD to the power supply VDD.

[0036] Next, at time t602, the control signal PGS is set to Lo, and at time t603, the control signal PMWN is set to Lo, thereby releasing the reset of the FD and writing the statically settled voltage, i.e., the N signal, to the storage capacitance CN of the storage circuit 12 (pq). Since the N signal, which is a voltage signal, is written to the storage capacitance CN when the control signal PMWN becomes Lo following the control signal PGS, the control signal PGS and the control signal PMWA may be set to Hi from time t600.

[0037] Next, from time t604 to time t605, the control signal PTXA is set to Hi, and the signal charge of the photodiode PDA is transferred to the FD.

[0038] Next, at time t606, the control signal PGS is set to Lo, and at time t607, the control signal PMWA is set to Lo, so that the FD after the reset is released drops and settles in response to the signal charge from the photodiode PDA. Then, the settled voltage, i.e., the A signal, is written to the storage capacitance CA of the storage circuit 12 (pq). As with the N signal, the control signal PGS and the control signal PMWA may be set to Hi from time t604.

[0039] Next, from time t608 to time t609, the control signals PTXA and PTXB are set to Hi, and the signal charges of the photodiodes PDA and PDB are transferred to the FD.

[0040] Next, at time t610, the control signal PGS is set to Lo, and at time t611, the control signal PMWAB is set to Lo, so that the FD after the reset is released drops and settles in response to the signal charges from the photodiodes PDA and PDB. Then, the settled voltage, i.e., the image signal, is written to the storage capacitance CAB of the storage circuit 12 (pq). As with the N signal, the control signal PGS and the control signal PMWA may be set to Hi from time t608.

[0041] Up to this point, the FD has not been reset except from time t600 to time t601. By using such a configuration and driving method of the pixel circuit 11(pq) and the correlated double sampling of the imaging signal and the N signal, which will be described later, it is possible to reduce noise added to the imaging signal.

[0042] The voltage signals written to the storage capacitors CN, CA, and CAB are AD converted in the following procedure during the Hi periods when the vertical scanning circuit 20 sequentially sets the control signals PSELN, PSELAB, and PSELA to Hi.

[0043] That is, from time t612 to time t615, the control signal PSELN is set to Hi, and the N signal appearing on the column output line VLq0 is compared by the comparator COMPq0 with the reference signal RAMP that starts to fall from time t613. If the magnitude relationship between the signals is inverted at any time up to time t615 (time t614 in FIG. 6), the polarity of the comparator COMPq0 changes to Lo, and in response, the count value of the counter COUNTERq0 stops.

[0044] Here, COUNTERq0 starts counting in synchronization with time t613 when the reference signal RAMP starts to fall, and the count value of the N signal is recorded in a latch circuit (not shown) from time t615 to time t616. In some cases, the N signal may be counted down. In that case, correlated double sampling can be easily performed by counting up during the AD conversion of the imaging signal, which is performed from time t616 to time t619.

[0045] At time t615, the reference signal RAMP is reset to the same voltage as before time t613.

[0046] From time t616 to time t619, the control signal PSELAB is set to Hi, and AD conversion of the imaging signal (A+B signal) is performed by the same operation as from time t612 to time t615. Also, from time t620 to time t623, the control signal PSELA is set to Hi, and AD conversion of the A signal is performed by the same operation as from time t612 to time t615.

[0047] After time t623, the horizontal scanning circuit 30 sequentially scans the latched AD conversion results of the N signal, the image capture signal, and the A signal in the column direction, and transfers them to an image processing unit (not shown) via row output lines HL and the output unit 50. Correlated double sampling of the image capture signal and the N signal, and the A signal and the N signal is performed by subtraction processing in the output unit 50. Also, down-counting of the N signal may be utilized as described above.

[0048] Here, the AD conversion period of the image signal has the same slope of the reference signal RAMP as that of the N signal, but is longer than the AD conversion period of the N signal. Similarly, the AD conversion period of the A signal is shorter than that of the image signal and longer than that of the N signal. Since the A signal corresponds to the signal charge generated by receiving light that has passed through part of the pupil of the imaging optical system, it has a smaller voltage range than the image signal that has received light that has passed through the entire pupil of the imaging optical system, and therefore requires a shorter AD conversion period. Since the N signal, which does not basically include a received light signal, has an even smaller voltage range than the A signal, it goes without saying that it requires a shorter AD conversion period than the A signal.

[0049] As described above, by sequentially performing AD conversion of the N signal, the imaging signal, and the A signal, it is possible to share the AD conversion circuit provided for each column as shown in Fig. 3 in the time direction. Therefore, in this embodiment, by providing only one AD conversion circuit for each column, it is possible to suppress an increase in the area of ​​the second semiconductor substrate. The improvement of time efficiency by sequential AD conversion will be described in the second modified example of this embodiment.

[0050] In addition, in an image processing unit (not shown), a B signal is generated from the difference between the imaging signal and the A signal, and the defocus amount of the photographing optical system is calculated by a known correlation calculation between the A signal and the B signal. For such correlation calculation, signals of all rows are not necessarily required, and the A signal may be thinned out in the row direction. In such a case, the AD conversion period of the A signal from time t619 to time t623 is shortened, and the AD conversion of the N signal of the next row is performed earlier, thereby improving the frame rate.

[0051] If there is a margin in the frame rate, the power supply of the AD conversion circuit ADq0 and the current source of the column output line VLq0 can be stopped from time t619 to time t623 to reduce power consumption. Also, the processing of the next row onwards may be sequentially accelerated to perform AD conversion, and power supply stopping to reduce power consumption may be performed after the final row. This makes it possible to improve the frame rate and reduce power consumption compared to the case where the A signals of all rows are AD converted.

[0052] In addition, taking into consideration the AD conversion performed by thinning out the A signal in the row direction, the AD conversion of the imaging signal is started immediately after the AD conversion of the N signal. In this way, the time interval of the correlated double sampling by the AD conversion circuit common to the N signal and the imaging signal can be shortened. In addition, since the time interval of the correlated double sampling can be aligned with the row where the AD conversion is performed by thinning out the A signal in the row direction, the noise difference between the rows in the imaging signal can be reduced. In addition, after repeating the AD conversion of the N signal and the imaging signal from time t612 to time t619 up to the last row and outputting the N signal and the imaging signal from all the storage circuits 12(pq), the AD conversion of the A signal may be performed from time t619 to time t623 for the rows where the A signal is required.

[0053] Furthermore, by utilizing the fact that the imaging device of this embodiment can perform global shuttering for simultaneous accumulation of all pixels including the A signal, the division direction of the photodiodes PDA and PDB may be made different from that of pixel 10(pq) in Fig. 1. Fig. 7 is a diagram showing a pixel 14(pq) of an imaging device in a first modified example of this embodiment. The difference from pixel 10(pq) is that the split photodiodes PDA and PDB are divided in the row direction of the imaging device.

[0054] Even if all pixels of the imaging device are configured like pixel 14(pq), the A signal and imaging signal of 14(pq) required for the correlation calculation of the qth column are prepared as information at the same time regardless of p, so that the correlation calculation in the vertical direction can be performed stably even if the subject moves. In addition, both pixel 10(pq) and pixel 14(pq) may be arranged. By using both the horizontal correlation calculation using pixel 10(pq) and the vertical correlation calculation using pixel 14(pq), it is possible to calculate the defocus amount regardless of the orientation of the spatial frequency component of the subject.

[0055] Incidentally, in order to improve the frame rate, the AD conversion in the row direction from the storage circuits 12(pq) may be parallelized. FIG. 8 is a diagram showing the configuration of a second semiconductor substrate in a second modified example of this embodiment. The difference from FIG. 3 is that the column output lines are VLq0, VLq1, and VLq2, which are three for each column, and the AD conversion circuits are ADq0, ADq1, and ADq2, which are three for each column, to parallelize the AD conversion. Accordingly, the column output line VLq0 and the AD conversion circuit ADq0 are connected to the 0th and 3rd rows, the column output line VLq1 and the AD conversion circuit ADq1 are connected to the 1st and 4th rows, and the column output line VLq2 and the AD conversion circuit ADq2 are connected to the 2nd and 5th rows to AD convert the voltage signals of the storage circuits 12(pq).

[0056] In this case, the number of circuits such as the AD conversion circuit and the current source for each column output line increases, and the second semiconductor substrate becomes larger. However, the efficiency in the time direction is improved compared to the case where, for example, three column output lines and AD conversion circuits for each column are connected to all rows and the N signal, the imaging signal, and the A signal are AD converted in parallel. This is because AD conversion is possible sequentially as soon as the AD conversion of each signal is completed without being limited to the imaging signal with the widest voltage range (the generation period of the reference signal RAMP is long), and the other two sets of column output lines and AD conversion circuits can be used in parallel for transmitting and AD converting the voltage signal of another row.

[0057] This is possible because, in the case of having a storage circuit 12(pq) that acquires a plurality of voltage signals from one pixel, the voltage signals are sequentially AD converted via a common column output line, as shown in the equivalent circuit diagram of Fig. 4. That is, according to this embodiment, it is possible to improve the operating efficiency during the AD conversion period and also improve the area efficiency of the second semiconductor substrate.

[0058] <Second embodiment> According to the technical concept of the present invention, it is only necessary that the column output lines and AD conversion circuits be shared by signals from a plurality of storage capacitors per pixel, and other configurations are also possible.

[0059] In this second embodiment, the second amplification transistor SF2 of the storage circuit 12(pq) is used in common for signals of multiple storage capacitances. Fig. 9 is a diagram showing the configuration of the storage circuit 12(pq) of the imaging device of this embodiment. The difference from the storage circuit 12(pq) of the imaging device in the first embodiment shown in Fig. 5 is that the storage capacitances CN, CA, and CAB can be input to the gate of one second amplification transistor SF2 via the memory transfer transistors MTN, MTA, and MTAB.

[0060] Furthermore, one selection transistor SEL is provided to the column output line VLq0. Also, there is no second amplification transistor dedicated to each of the storage capacitors CN, CA, and CAB, and they cannot form independent source follower circuits. Therefore, when signal voltages are sequentially transmitted from these storage capacitors to the AD conversion circuit via the column output line VLq0, it is necessary to reset the gate of the amplification transistor SF2 each time, and therefore a second reset transistor RES2 is provided.

[0061] Next, a method for driving the imaging device of this embodiment will be described. Fig. 10 is a diagram showing a timing chart illustrating a method for driving the imaging device of this embodiment. The difference from the first embodiment shown in Fig. 6 is in the storage circuit 12 (pq), so the all-pixel simultaneous storage period (from time t600 to time t611 in Fig. 6) for realizing a global shutter is the same as that of the first embodiment, and a description thereof will be omitted. Also, the name of each transistor in the equivalent circuit diagram of Fig. 9 is prefixed with P to indicate the Hi or Lo of its gate control signal. When the control signal is Hi, the transistor is ON, and when it is Lo, it is OFF.

[0062] To start reading out the 0th row from time t1011, the control signal PSEL is set to Hi to turn on the selection transistor SEL of the storage circuit 12(pq).

[0063] First, from time t1011 to time t1012, the control signal PRES2 is set to Hi to reset the gate of the amplifying transistor SF2 to the power supply voltage VDD.

[0064] Next, at time t1012, the control signal PMTN is set to Hi, and AD conversion of the N signal is performed until time t1015. This AD conversion is the same as that up to time t615 in FIG.

[0065] Next, from time t1015 to time t1016, the control signal PRES is again set to Hi to reset the gate of the amplifying transistor SF2 to the power supply voltage VDD.

[0066] Next, at time t1016, the control signal PMTAB is set to Hi, and AD conversion of the imaging signal (A+B signal) is performed until time t1019. This AD conversion is the same as that up to time t619 in FIG.

[0067] Next, from time t1019 to time t1020, the control signal PRES is again set to Hi to reset the gate of the amplifying transistor SF2 to the power supply voltage VDD.

[0068] Next, at time t1020, the control signal PMTAB is set to Hi, and AD conversion of the A signal is performed until time t1023. This AD conversion is the same as that up to time t623 in FIG.

[0069] It should be noted that the present embodiment can also obtain the same effects as the first embodiment through modifications similar to those of the first embodiment.

[0070] <Third embodiment> In the technical concept of the present invention, it is not necessary to transmit all of the voltage signals of the multiple storage capacitors per pixel to the AD conversion circuit via a common column output line. In this embodiment, a configuration will be described in which some of the multiple storage capacitors per pixel are transmitted to the AD conversion circuit via a common column output line and AD converted.

[0071] Fig. 11 is a diagram showing the configuration of a storage circuit 12(pq) in the imaging device of this embodiment, and Fig. 12 is a diagram showing the configuration of a second semiconductor substrate of the imaging device of this embodiment.

[0072] The difference from Figures 5 and 3 is that only the A signal and the image signal (A+B signal) stored in the storage capacitors CA and CAB are sequentially AD converted via a common column output line, and the N signal stored in the storage capacitor CN is AD converted in parallel with the image signal and A signal via an independent column output line.

[0073] In addition, the same effect can be obtained by sequentially AD converting the N signal and the image signal via a common column output line and independently AD converting the A signal, or by sequentially AD converting the N signal and the A signal via a common column output line and independently AD converting the image signal.

[0074] <Fourth embodiment> The technical concept of the present invention is also applicable to cases where the photoelectric conversion unit of a pixel is not necessarily divided, such as the pixel 10(pq) and the pixel 14(pq).

[0075] In this embodiment, the pixel 15(pq) does not have a divided photoelectric conversion unit, but the storage circuit 12(pq) has multiple storage capacitors. Fig. 13 is an equivalent circuit diagram showing the configurations of the pixel 15(pq), pixel circuit 11(pq), and storage circuit 12(pq) of the imaging device of this embodiment.

[0076] The difference with FIG. 4 is that pixel 15(pq) does not have a divided photoelectric conversion unit like pixel 10(pq) and pixel 14(pq), and the storage circuit 12(pq) has two capacitances CN and CS that store the voltages of the N signal and the imaging signal.

[0077] 4, the transistors that write the voltage of the image signal (A+B signal), the storage capacitance that stores it, the transistor that amplifies it, and the transistor that selects it are respectively referred to as MWAB, CAB, SF2AB, and SELAB. In contrast, in this embodiment, the transistors that write the voltage of the image signal, the storage capacitance that stores it, the transistor that amplifies it, and the transistor that selects it are respectively referred to as MWS, CS, SF2S, and SELS.

[0078] Therefore, although it is necessary to read the control signal PMWAB as PMWS and the control signal PSELAB as PSELS in FIG. 6, the imaging device of this embodiment can be driven in the same manner as from time t600 to time t603 and from time t608 to time t619.

[0079] In this way, by sequentially AD converting the N signal and the imaging signal, which are stored in two storage capacitors for each pixel and have significantly different voltage ranges, it is possible to obtain the same effects as in the first embodiment.

[0080] <Fifth embodiment> The technical concept of the present invention is applicable not only to pixels having a photoelectric conversion unit divided into two, such as pixel 10 (pq) and pixel 14 (pq), but also to pixels having a photoelectric conversion unit divided into more than two parts.

[0081] In this embodiment, the pixel 16(pq) has a photoelectric conversion unit divided into four, and the storage circuit 12(pq) has three storage capacitors. Fig. 14 is a diagram showing the configuration of the pixel 16(pq) of the imaging device of this embodiment. Fig. 15 is an equivalent circuit diagram showing the configurations of the pixel 16(pq), pixel circuit 11(pq), and storage circuit 12(pq).

[0082] 14, the photoelectric conversion unit of the pixel is divided into a cross shape, and if the A signal corresponding to the signal charges of the photodiodes PDA and PDB and the imaging signal corresponding to the signal charges of all the photoelectric conversion units can be output, a correlation calculation in the vertical direction can be performed and a captured image can be generated. Also, if the A signal corresponding to the signal charges of the photodiodes PDA and PDC and the imaging signal can be output, a correlation calculation in the horizontal direction can be performed and a captured image can be generated.

[0083] Furthermore, the signal charge accumulation times of the photodiodes PDA and PDD are set long, and those of the photodiodes PDB and PDC are set short. In addition, by outputting a long-term signal corresponding to the former signal charge and a short-term signal corresponding to the latter signal charge, an image processing unit (not shown) can synthesize the images to generate an image signal with an expanded dynamic range.

[0084] In either case, the storage circuit 12(pq) can be configured with three storage capacitors including one for the N signal, as in the previous embodiments. The difference between Fig. 15 and Fig. 4 is that the pixel 16(pq) includes four photodiodes PDA, PDB, PDC, and PDD, and transfer transistors TXA, TXB, TXC, and TXD that transfer the signal charges to FD.

[0085] Next, a method for driving such an imaging device will be described with reference to Fig. 6 of the first embodiment. First, when signals used for horizontal correlation calculation and imaging are sequentially AD converted, the control signal PTXC may be controlled at the same timing as the control signal PTXA in Fig. 6, and the control signal PTXD may be controlled at the same timing as the control signal PTXB.

[0086] In addition, in order to sequentially AD convert the signals used for vertical correlation calculation and imaging, the control signals PTXA and PTXB may be controlled at the same timing, and the control signals PTXC and PTXD may be controlled at the same timing.

[0087] Furthermore, when the short-time accumulation signal (the voltage of the A signal corresponding to the signal charges of the photodiodes PDA and PDD) and the long-time accumulation signal (the voltage of the imaging signal corresponding to the signal charges of the photodiodes PDB and PDC) are sequentially AD-converted, the control signal PTXD is controlled at the same timing as the control signal PTXA during the period from time t604 to time t605 in Fig. 6. Also, the control signal PTXC is controlled at the same timing as the control signal PTXB during the period from time t608 to time t609 in Fig. 6, and the control signal PTXA is set to Lo during the period from time t608 to time t609, and the interval from time t606 to time t608 is further extended.

[0088] In either case, the same effect as in the first embodiment can be obtained by sequentially AD converting the N signal, the imaging signal (A+B signal), and the A signal, which have significantly different voltage ranges and are stored in the three storage capacitors of each pixel.

[0089] (Other embodiments) The present invention can also be realized by a process in which a program for implementing one or more of the functions of the above-described embodiments is supplied to a system or device via a network or a storage medium, and one or more processors in a computer of the system or device read and execute the program. The present invention can also be realized by a circuit (e.g., ASIC) that implements one or more of the functions.

[0090] The invention is not limited to the above-described embodiments, and various modifications and variations are possible without departing from the spirit and scope of the invention. Accordingly, the following claims are appended to apprise the public of the scope of the invention. [Explanation of symbols]

[0091] 10: pixel, 11: pixel circuit, 12: storage circuit, 20: vertical scanning circuit, 30: horizontal scanning circuit, 40: timing generation circuit, 50: output section, VLq0: column output line, RAMP: reference signal, ADq0: AD conversion circuit, COMP: comparator, COUNTER: counter, HL: row output line, PD: photodiode, C: storage capacitance

Claims

1. A plurality of pixels each including a plurality of photoelectric conversion units, each outputting a first voltage signal based on charges of some of the plurality of photoelectric conversion units and a second voltage signal based on a combined charge of the charges of the plurality of photoelectric conversion units; a plurality of holding circuits provided in a one-to-one correspondence with the plurality of pixels, the holding circuits holding voltage signals based on charges generated by the plurality of photoelectric conversion units; each of the plurality of holding circuits includes a plurality of holding units including a first holding unit that holds the first voltage signal and a second holding unit that holds the second voltage signal, and an amplification transistor; 11. An imaging device, comprising: a first holding circuit configured to hold a first signal from a first amplifier transistor; a second holding circuit configured to hold a second amplifier transistor;

2. A plurality of pixels each including a plurality of photoelectric conversion units, each outputting a first voltage signal based on charges of some of the plurality of photoelectric conversion units and a second voltage signal based on a combined charge of the charges of the plurality of photoelectric conversion units; a plurality of holding circuits provided in a one-to-one correspondence with the plurality of pixels, the holding circuits holding voltage signals based on charges generated by the plurality of photoelectric conversion units; an AD conversion circuit that sequentially AD-converts the voltage signals held in the plurality of holding circuits; each of the plurality of holding circuits includes a plurality of holding units including a first holding unit that holds the first voltage signal and a second holding unit that holds the second voltage signal; an imaging device, characterized in that the number of the AD conversion circuits arranged for each column of the pixels is smaller than the number of holding units provided in each of the holding circuits.

3. 3. The imaging device according to claim 1, wherein each of the plurality of holding circuits further includes a third holding section configured to hold a third voltage signal obtained after a reset operation of each of the plurality of pixels is released.

4. 4. The imaging device according to claim 3, wherein the plurality of holding sections included in each of the plurality of holding circuits are three, namely, the first holding section, the second holding section, and the third holding section.

5. 5. The imaging device according to claim 3, wherein the first voltage signal, the second voltage signal, and the third voltage signal have different voltage ranges.

6. 6. The image pickup device according to claim 1, wherein each of the plurality of holding circuits outputs the second voltage signal before the first voltage signal.

7. 7. The image pickup device according to claim 1, wherein each of the plurality of pixels includes a microlens, and the plurality of photoelectric conversion units receive light by dividing a pupil region of an image pickup optical system.

8. 8. The imaging device according to claim 1, wherein the imaging device is configured by stacking a plurality of semiconductor substrates, the plurality of pixels are disposed on a first semiconductor substrate of the plurality of semiconductor substrates, and the plurality of holding circuits are disposed on a second semiconductor substrate of the plurality of semiconductor substrates.

9. 9. The imaging device according to claim 1, wherein the number of the holding sections provided in each of the plurality of holding circuits is greater than the number of the photoelectric conversion sections provided in each of the plurality of pixels.

Images