Solid-state imaging device

By using a combination of transfer transistors and accumulation capacitors in a solid-state imaging device, the problems of dynamic range expansion and noise deterioration in the prior art are solved, and more efficient signal processing and noise suppression are achieved.

CN118923125APending Publication Date: 2024-11-08NUVOTON TECH CORP JAPAN NAGAOKAKYO CITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202380028985.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2022-03-25
Filing Date
2023-03-16
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The existing solid-state imaging devices have difficulties in expanding the dynamic range, especially due to the lack of a transfer transistor, which leads to the problem of poor low illuminance noise and unimproved photodiode signal noise.

Method used

Using a solid-state imaging device with a pixel circuit, a detection circuit and a signal processing unit, the photodiode signal is read out through the first transfer transistor, and the overflowing charge is accumulated in the accumulation capacitor, and the detection circuit is used to compare it with the reference value, when the signal reaches the reference value. At the value, the photodiode and the accumulation capacitor are initialized, and the intensity of the incident light is calculated based on the initialization times and the read signal, so as to realize signal processing.

Benefits of technology

It effectively expands the dynamic range of the solid-state imaging device, reduces low illuminance noise, improves the signal noise performance of the photodiode, and improves the accuracy of signal processing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118923125A_ABST
    Figure CN118923125A_ABST
Patent Text Reader

Abstract

A solid-state imaging device (100) is provided with a pixel circuit (3a) that outputs a plurality of pixel signals, a detection circuit (52), and a signal processing unit (70), and the pixel circuit (3a) is provided with a photodiode (PD), a transfer transistor (TG) that reads out a signal of the photodiode (PD) to a charge storage unit (FD0), and a storage capacitor (C2) that stores charges overflowing from the photodiode (PD). The detection circuit (52) compares the signal of the storage capacitor (C2) with a reference signal, initializes the photodiode (PD) and the storage capacitor (C2) when the signal of the storage capacitor (C2) reaches the reference signal, counts the number of initializations, and outputs the signal of the storage capacitor (C2). A signal processing unit (70) calculates a first signal indicating the intensity of incident light on the basis of the number of initializations and a mixed signal of a signal of a photodiode (PD) and a signal of an accumulation capacitor (C2) read from a transfer transistor (TG) to a charge accumulation unit (FD0).
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to a solid-state imaging device. Background Art

[0002] Patent Documents 1 and 2 disclose methods of achieving a wide dynamic range (WDR) in a solid-state imaging device.

[0003] Prior art literature

[0004] Patent Literature

[0005] Patent Document 1: Japanese Patent No. 3878613

[0006] Patent Document 2: Japanese Patent No. 4497366 Summary of the invention

[0007] Problems to be solved by the invention

[0008] According to the solid-state imaging device disclosed in Patent Document 1, a pulse width modulation (PWM) method is disclosed as a method for detecting the light intensity of the signal of the photodiode. The PWM method is a method for detecting the intensity of the incident light as the width of the pulse signal or the frequency of occurrence of the pulse signal. At this time, since the PWM method is only performed on the signal of the photodiode, there is a problem of difficulty in expanding the dynamic range. In addition, since there is no transfer transistor, it is impossible to read the reset component at each initialization, that is, there is a problem that CDS (Correlated Double Sampling) for removing the reset component cannot be performed, and the low-illumination noise deteriorates. In addition, there is a problem that the SN of the photodiode does not change even if it is digitally multiplied by N times, and the SN of the photodiode does not improve.

[0009] The solid-state imaging device disclosed in Patent Document 2 is a system in which a signal overflowing from a photodiode exceeding a predetermined threshold is accumulated in a storage capacitor via an overflow element. However, the storage capacitor is fixed due to process constraints, so it is difficult to further expand the dynamic range.

[0010] Therefore, the present disclosure provides a solid-state imaging device capable of easily expanding a dynamic range.

[0011] Means used to solve problems

[0012] The solid-state imaging device disclosed in the present invention comprises: a pixel circuit that outputs multiple pixel signals; a detection circuit; a pixel control circuit that controls the above-mentioned pixel circuit; and a signal processing unit; the above-mentioned pixel circuit comprises: a photodiode; a first transfer transistor that reads out the signal of the above-mentioned photodiode to a floating diffusion unit; and a storage capacitor that stores the charge overflowed from the above-mentioned photodiode; the above-mentioned detection circuit compares the signal of the above-mentioned storage capacitor with a first reference value, and when the signal of the above-mentioned storage capacitor reaches the above-mentioned first reference value, the above-mentioned photodiode and the above-mentioned storage capacitor are initialized via the above-mentioned pixel control circuit, and the number of initializations is counted; the above-mentioned signal processing unit calculates a first signal representing the intensity of incident light based on the above-mentioned initialization number and a mixed signal of the signal of the above-mentioned photodiode in the above-mentioned floating diffusion unit and the signal of the above-mentioned storage capacitor read out by the above-mentioned first transfer transistor.

[0013] In addition, these general or specific solutions may be implemented by systems, methods, integrated circuits, computer programs, or computer-readable recording media such as CD-ROMs, or by any combination of systems, methods, integrated circuits, computer programs, and recording media.

[0014] Effects of the Invention

[0015] According to the solid-state imaging device of one aspect of the present disclosure, the dynamic range can be easily expanded. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a block diagram showing a configuration example of a solid-state imaging device according to an embodiment.

[0017] Figure 2A It is a block diagram showing a configuration example of a pixel circuit and a logic circuit according to an embodiment.

[0018] Figure 2B It is a block diagram showing a configuration example of a pixel circuit and a logic circuit according to an embodiment.

[0019] Figure 2C It is a block diagram showing a configuration example of a pixel circuit and a logic circuit according to an embodiment.

[0020] Figure 3 It is a block diagram showing a configuration example of a pixel circuit and a logic circuit according to an embodiment.

[0021] Figure 4 It is a block diagram showing a configuration example of a pixel circuit and a logic circuit according to an embodiment.

[0022] Figure 5 This is a diagram for explaining the commonality of logic circuits in a plurality of pixel circuits.

[0023] Figure 6It is a diagram showing a configuration example of a sample-and-hold circuit.

[0024] Figure 7 This is a diagram showing an overview of a method for implementing WDR synthesis within a pixel.

[0025] Figure 8 This is a diagram showing a configuration example of a signal processing unit that performs WDR synthesis using three signals.

[0026] Fig. 9 This is a diagram showing a configuration example of a signal processing unit that performs WDR synthesis using two signals selected from three signals.

[0027] Fig.10 This is a diagram showing a basic readout timing example of a pixel circuit.

[0028] Fig.11 This is a diagram showing the timing of the selective WDR method.

[0029] Fig.12 This is a diagram showing a basic timing example of the PWM method.

[0030] Fig.13 This is a diagram showing a timing example when an AD conversion circuit is used in the PWM method.

[0031] Fig.14 This is a diagram for explaining an example of selecting an HCG signal, an MCG signal, and an LCG signal by giving priority to the largest signal of each color in Bayer units.

[0032] Fig.15 This is a diagram for explaining an example in which an HCG signal, an MCG signal, and an LCG signal are independently selected for each color.

[0033] Fig.16 This is a diagram showing an example in which the solid-state imaging device according to the embodiment is configured as a stacked image sensor.

[0034] Fig.17 This is a diagram showing an example in which the solid-state imaging device according to the embodiment is configured as a stacked image sensor.

[0035] Fig.18 This is a diagram showing an example in which the solid-state imaging device according to the embodiment is configured as a stacked image sensor.

[0036] Fig.19 This is a diagram showing an example in which the solid-state imaging device according to the embodiment is configured as a stacked image sensor.

[0037] Fig. 20 A diagram showing an example of connection between a pixel circuit and a logic circuit according to an embodiment.

[0038] Fig.21 A diagram showing an example of connection between a pixel circuit and a logic circuit according to an embodiment.

[0039] Fig. 22 It is a diagram showing a timing example when the HCG signal and the MCG signal are selected.

[0040] Fig.23 This is a diagram showing a timing example when the MCG signal and the LCG signal are selected.

[0041] Fig.24 FIG. 1 is a diagram showing a timing example when the saturation level of the LCG signal and the residual signal are selected.

[0042] Fig.25 The diagram shows the saturation signal of the HCG signal, the MCG signal, the LCG signal, and the SN of the residual signal.

[0043] Fig.26 It is a diagram showing an example of an imaging device to which the solid-state imaging device according to the embodiment is applied.

[0044] Fig. 27 FIG. 1 is a diagram showing an example of a distance measuring imaging device to which the solid-state imaging device according to the embodiment is applied. DETAILED DESCRIPTION

[0045] A solid-state imaging device of a technical solution disclosed in the present invention comprises: a pixel circuit that outputs multiple pixel signals; a detection circuit; a pixel control circuit that controls the above-mentioned pixel circuit; and a signal processing unit; the above-mentioned pixel circuit comprises: a photodiode; a first transfer transistor that reads out the signal of the above-mentioned photodiode to a floating diffusion unit; and a storage capacitor that stores the charge overflowed from the above-mentioned photodiode; the above-mentioned detection circuit compares the signal of the above-mentioned storage capacitor with a first reference value, and when the signal of the above-mentioned storage capacitor reaches the above-mentioned first reference value, the above-mentioned photodiode and the above-mentioned storage capacitor are initialized via the above-mentioned pixel control circuit, and the number of initializations is counted; the above-mentioned signal processing unit calculates a first signal representing the intensity of incident light based on the above-mentioned initialization number and a mixed signal of the signal of the above-mentioned photodiode in the above-mentioned floating diffusion unit and the signal of the above-mentioned storage capacitor read out by the above-mentioned first transfer transistor.

[0046] When incident light of high intensity (high illumination) is incident, charge overflows from the photodiode and is accumulated in the storage capacitor. However, there are also cases where high-intensity incident light is incident that cannot be fully accumulated in the storage capacitor. Even in such a case, when the signal of the storage capacitor reaches the first reference value, the photodiode and the storage capacitor are initialized and can continue to accumulate charge. At this time, the number of initializations is counted, so the amount of charge of the photodiode and the storage capacitor corresponding to the initialization can be known. In addition, by using a mixed signal of the signal of the photodiode and the signal of the storage capacitor when the signal of the storage capacitor no longer reaches the first reference value, and the number of initializations, that is, the amount of charge of the photodiode and the storage capacitor corresponding to more than one initialization since the incident light was incident, the intensity of the incident light can be calculated even if it is high-intensity. Therefore, the dynamic range can be easily expanded. For example, compared to the case where the PWM method is not used, the dynamic range can be expanded.

[0047] For example, the solid-state imaging device may include a sample-and-hold circuit and an AD conversion circuit.

[0048] For example, it may be that at least one pixel signal among the above-mentioned multiple pixel signals is not held by the above-mentioned sampling and holding circuit, but is directly AD-converted by the above-mentioned AD conversion circuit, and at least one other pixel signal among the above-mentioned multiple pixel signals is held by the above-mentioned sampling and holding circuit, and the held signal is AD-converted by the above-mentioned AD conversion circuit.

[0049] Thus, by directly converting a portion of the pixel signals by an AD conversion circuit, low-light noise can be reduced. For example, by providing a sample-and-hold circuit for each pixel circuit, it is no longer necessary to provide a memory corresponding to each of the plurality of pixel signals. Therefore, this is effective due to the area constraints in a micro unit.

[0050] For example, when the number of initialization times is greater than 1, the pixel circuit outputs a mixed signal of the photodiode and the storage capacitor when the signal of the storage capacitor reaches the first reference value during exposure as a second signal, the photodiode and the storage capacitor are initialized, the second signal is held by the sampling and holding circuit, and the held signal is AD-converted by the AD conversion circuit, and the pixel circuit outputs a mixed signal of the photodiode and the storage capacitor when the signal of the storage capacitor does not reach the first reference value and remains in the photodiode and the storage capacitor as a third signal at the end of exposure, and the third signal is not held by the sampling and holding circuit, but is directly AD-converted by the AD conversion circuit.

[0051] Thus, when the signal of the storage capacitor reaches the first reference value, the mixed signal of the photodiode and the storage capacitor, i.e., the second signal, is held by the sampling and holding circuit, thereby maintaining the amount of charge initialized each time. In addition, when the exposure is completed, the mixed signal, i.e., the third signal, which is the residual mixed signal in the photodiode and the storage capacitor but does not reach the first reference value, is directly converted by the AD conversion circuit because the second signal is already held in the sampling and holding circuit.

[0052] For example, the signal processing unit may calculate the first signal by calculating a sum of a product of the second signal and the number of initialization times and the third signal.

[0053] In this way, the first signal can be calculated by the second signal×the number of initialization times+the third signal.

[0054] For example, the sampling and holding circuit may include a first sampling and holding capacitor element for a reset component and a second sampling and holding capacitor element for a signal component, and when the number of initialization times is greater than 1, during exposure, after the photodiode and the storage capacitor are initialized, the reset component is held in the first sampling and holding capacitor element, and during exposure, when the signal of the storage capacitor reaches the first reference value, the second signal is held as a signal component in the second sampling and holding capacitor element, and the AD conversion circuit performs AD conversion on the pixel signal level by comparing the pixel signal level after the reset component is subtracted from the signal component with the second reference value.

[0055] In this way, the second signal can be correctly read by performing CDS.

[0056] For example, when the above-mentioned initialization times is 0, the above-mentioned pixel circuit outputs the fourth signal in a low illumination mode for low illumination, and outputs the fifth signal in a mode for high illumination higher than the above-mentioned low illumination mode, the above-mentioned fourth signal is not held by the above-mentioned sampling and holding circuit, but is directly AD-converted by the above-mentioned AD conversion circuit, and the above-mentioned fifth signal is held by the above-mentioned sampling and holding circuit, and the held signal is AD-converted by the above-mentioned AD conversion circuit.

[0057] Thus, the fourth signal output in the low illumination mode is directly AD-converted, thereby reducing the low illumination noise. The fifth signal output in the high illumination mode higher than the low illumination mode is held by the sample-and-hold circuit, so although kTC noise is generated during sampling, its influence can be suppressed by performing WDR synthesis with the fourth signal. Thus, WDR with excellent SN at low illumination can be achieved.

[0058] For example, the pixel circuit may include: a second transfer transistor for transferring the charge of the storage capacitor to the floating diffusion; a first amplifier transistor for outputting the signal of the photodiode and the signal of the storage capacitor; a first selection transistor connected to the first amplifier transistor; a second amplifier transistor connected to the storage capacitor; and a second selection transistor connected to the second amplifier transistor; the detection circuit is connected to the storage capacitor via the second selection transistor and the second amplifier transistor; the pixel circuit, in the low illumination mode, outputs the signal of the photodiode from the first amplifier transistor as the fourth signal via the first transfer transistor; the high illumination mode includes a high illumination mode in which the pixel circuit outputs the signal of the storage capacitor from the first amplifier transistor via the second transfer transistor; the sixth signal belongs to the fifth signal.

[0059] Thus, in the low illumination mode, the fourth signal for low illumination is output from the first amplifier transistor via the first transfer transistor, and in the high illumination mode, the sixth signal for high illumination is output from the first amplifier transistor via the second transfer transistor. In addition, since the signal of the storage capacitor can be detected via the second amplifier transistor, it is possible to determine whether the signal of the storage capacitor has reached the first reference value, in other words, the number of initializations can be counted. For example, it is possible to suppress the omission of LED flickering.

[0060] For example, the pixel circuit may include a reset transistor, and the detection circuit controls the reset transistor, the first transfer transistor, and the second transfer transistor via the pixel control circuit based on a comparison result of a signal of the storage capacitor with a first reference value, thereby discharging charges of the photodiode and the storage capacitor and initializing the photodiode and the storage capacitor.

[0061] Thus, the reset transistor, the first transfer transistor, and the second transfer transistor can be reset through the control of the pixel control circuit, and the charges of the storage capacitor and the photodiode can be more reliably discharged and initialized.

[0062] For example, the pixel circuit may include a gain switching transistor for switching the gain of the signal of the photodiode, and the high-illuminance mode may include a medium-illuminance mode in which the pixel circuit outputs the signal of the photodiode as a seventh signal from the first amplifier transistor by controlling the gain switching transistor, and the seventh signal belongs to the fifth signal.

[0063] Thus, in the medium illumination mode, the gain switching transistor is controlled to output the seventh signal for medium illumination from the first amplifier transistor. For example, the step difference of SN at the respective boundaries (transition regions) of the sixth signal for high illumination, the seventh signal for medium illumination, and the fourth signal for low illumination can be reduced.

[0064] For example, the second reference value may be a saturation level of the fourth signal.

[0065] By setting the saturation level of the fourth signal for low illumination to the second reference value, it is possible to determine whether the intensity of incident light is low illumination. Furthermore, based on the determination result, a signal can be selected from a plurality of pixel signals (e.g., the fourth signal, the sixth signal, and the seventh signal), thereby achieving area reduction, high speed, and low power consumption of the solid-state imaging device.

[0066] For example, the signal processing unit may include a WDR processing unit, which generates an output signal from at least the first signal when the number of initialization times is greater than 1, and generates an output signal from at least one of the fourth signal and the fifth signal when the number of initialization times is 0.

[0067] Thus, when the number of initializations is 1 or more, that is, when the incident light is of high illumination, the output signal is generated by the first signal, and when the number of initializations is 0, that is, when the incident light is of medium or low illumination, the output signal is generated by the fourth signal or the fifth signal (for example, the sixth signal or the seventh signal). Therefore, the dynamic range can be expanded from low illumination to high illumination.

[0068] For example, the AD conversion circuit may include a comparator, and when the initialization number is 0, the pixel circuit outputs three or more pixel signals, and the comparator compares one pixel signal among the three or more pixel signals with a second reference value after AD conversion, and generates a selection signal indicating the selection of at least two pixel signals among the three or more pixel signals when the one pixel signal reaches the second reference value, and the pixel control circuit selects the at least two pixel signals from the three or more pixel signals according to the selection signal.

[0069] Thus, by selecting at least two pixel signals from three or more pixel signals output by the pixel circuit, it is possible to achieve area reduction, high speed, and low power consumption of the solid-state imaging device. In addition, when the counter circuit of the AD conversion circuit is a binary counter, since the digital signal is proportional to the level of the pixel signal, at least two pixel signals can be selected, and when the counter circuit of the AD conversion circuit is a Gray code counter, since the digital signal is not proportional to the level of the pixel signal, at least two pixel signals cannot be selected. Therefore, it is particularly effective when using a Gray code counter to achieve low power consumption.

[0070] For example, the AD conversion circuit and the detection circuit may be provided for each of the pixel circuits, and the pixel control circuit may control the pixel circuit for each of the pixel circuits.

[0071] This makes it possible to increase the speed of AD conversion in the AD conversion circuit. In particular, in a global shutter, the waiting time of the charge of the floating diffusion becomes long, which causes the problem of PLS ​​(parasitic sensitivity), but the PLS can be improved.

[0072] For example, the AD conversion circuit and the detection circuit may be provided for each of the plurality of pixel circuits, the pixel control circuit controls the pixel circuit for each single pixel circuit, and the detection circuit synchronizes the corresponding pixel circuit with the pixel control circuit while comparing the signal of the storage capacitor of the pixel circuit with the first reference value.

[0073] Thus, during exposure, the detection circuit and the pixel control circuit operate synchronously with respect to the same pixel circuit, thereby enabling pulse width modulation operation to be realized for each of the plurality of pixel circuits shared by the AD conversion circuit and the detection circuit.

[0074] For example, the pixel control circuit may synchronize control of the first selection transistor, the second selection transistor, and the sample-and-hold circuit of the corresponding pixel circuit.

[0075] Thus, while the control of the first selection transistor, the second selection transistor, and the sampling and holding circuit is synchronized, the pixel signal output from the first amplifier transistor is input to the AD conversion circuit via the first selection transistor and the sampling and holding circuit for AD conversion. In addition, similarly, while the control of the first selection transistor, the second selection transistor, and the sampling and holding circuit is synchronized, the signal of the storage capacitor output from the second amplifier transistor is continuously detected by the detection circuit via the second selection transistor. In this way, the first selection transistor, the second selection transistor, and the sampling and holding circuit in the pixel circuit are synchronized by the control signal from the pixel control circuit, so that the signal level of the pixel signal (especially the pixel signal for high illumination) can be correctly read according to each pixel circuit.

[0076] For example, the plurality of pixel circuits may correspond to pixels arranged in a color filter, and the pixel control circuit may perform the same pixel control on the plurality of pixel circuits in accordance with the maximum signal level of the pixels arranged in the color filter.

[0077] Thus, when WDR is synthesized, even if the color temperature of the subject changes in the unit of the arrangement of the color filters, for example, the unit of the Bayer arrangement composed of 4 colors (R, Gr, B, Gr), the transition area of ​​WDR, i.e., the boundary portion, can be changed in a linked manner by using the gain of the maximum signal level among the 4 colors. Therefore, the gain of each color is the same, and coloring caused by linear deviation of each color can be prevented from occurring before and after the above-mentioned boundary portion.

[0078] For example, the solid-state imaging device may include: a first semiconductor substrate having the pixel circuit; and a second semiconductor substrate having the detection circuit and the AD conversion circuit; and the first semiconductor substrate and the second semiconductor substrate may be stacked.

[0079] When the first semiconductor substrate and the second semiconductor substrate are not stacked, there is a problem that the area of ​​the photodiode becomes smaller. However, by providing a stacked structure, the detection circuit and the AD conversion circuit can be provided without affecting the area of ​​the photodiode.

[0080] For example, the first reference value may be a saturation level of the storage capacitor.

[0081] Thus, when the storage capacitor is saturated, the photodiode and the storage capacitor can be initialized, and the number of initializations can be counted.

[0082] A solid-state imaging device of a technical solution disclosed in the present invention comprises: a pixel circuit having a photodiode, a floating diffusion portion from which a signal of the photodiode is read, a storage capacitor for storing charges overflowing from the photodiode, and a reset unit for initializing the photodiode, the floating diffusion portion, and the storage capacitor; and a detection circuit having a comparator for comparing the signal of the storage capacitor with a first reference value, and an initialization circuit for controlling the reset unit according to the comparison result of the comparator; the pixel circuit is formed on a first semiconductor substrate composed of one to multiple semiconductor layers, the detection circuit is formed on a second semiconductor substrate composed of one to multiple semiconductor layers, and the first semiconductor substrate and the second semiconductor substrate are stacked.

[0083] When incident light of high intensity (high illumination) is incident, charge overflows from the photodiode and is accumulated in the storage capacitor. However, there are also cases where high-intensity incident light is incident that cannot be fully accumulated in the storage capacitor. Even in such a case, when the signal of the storage capacitor reaches the first reference value, the photodiode and the storage capacitor can be initialized and the charge can be continuously accumulated. At this time, when the number of initializations is counted, the amount of charge of the photodiode and the storage capacitor corresponding to the initialization can be known. In addition, by using a mixed signal of the signal of the photodiode and the signal of the storage capacitor when the signal of the storage capacitor no longer reaches the first reference value, and the number of initializations, that is, the amount of charge of the photodiode and the storage capacitor corresponding to the initialization of more than one time since the incident light was incident, the intensity of the incident light can be calculated even if it is high-intensity. Therefore, the dynamic range can be easily expanded. For example, compared to the case where the PWM method is not used, the dynamic range can be expanded.

[0084] For example, the pixel circuit may include a first transfer transistor that reads out a signal from the photodiode to the floating diffusion portion, and a second transfer transistor that transfers a signal from the storage capacitor to the floating diffusion portion.

[0085] Thus, the signal of the photodiode in low illumination exposure (low illumination mode) is controlled by the first transfer transistor, and the signal of the storage capacitor in high illumination exposure (high illumination mode) is controlled by the second transfer transistor. Thus, the former signal and the latter signal can be read out at different timings in time, and these signals can be used to synthesize WDR, which can further expand the dynamic range.

[0086] For example, the detection circuit may include a counter function for counting the comparison result of the comparator.

[0087] Thus, by providing a counter function for counting the comparison result of the comparator in the detection circuit, faster and more accurate counting can be performed. Thus, for example, in a pixel circuit constituting a part of a pixel array, light can be obtained without omission even when light of high illumination is temporarily incident, and high illumination SN can be improved.

[0088] For example, the comparator may further include an AD conversion function formed in the second semiconductor substrate and converting a signal of the floating diffusion into a digital value, and input terminals of the comparator and the AD conversion function may be connected to different output terminals of the pixel circuit, respectively.

[0089] According to the solid-state imaging device disclosed in Patent Document 1, the residual signal of the photodiode after PWM, i.e., the low-light signal, cannot be read out, and the low-light SN deteriorates. However, according to the present invention, the residual signal after PWM, i.e., the low-light signal, can be read out by the AD conversion function and signal processing can be performed, so the low-light SN can be improved.

[0090] For example, it may also include: an AD conversion function, formed on the above-mentioned second semiconductor substrate, converting the signal from the above-mentioned floating diffusion part into a digital value; and a selection circuit, formed on the above-mentioned first semiconductor substrate or the above-mentioned second semiconductor substrate, inputted with the signal from the above-mentioned storage capacitor and the signal from the above-mentioned floating diffusion part, and selecting the output; the output of the above-mentioned selection circuit is inputted to the above-mentioned AD conversion function, and the above-mentioned detection circuit and the above-mentioned AD conversion function share the above-mentioned comparator.

[0091] Thus, the detection circuit and the AD conversion function share the comparator, so that the area of ​​the logic circuit can be reduced.

[0092] Hereinafter, the embodiments will be described in detail with reference to the drawings.

[0093] In addition, the embodiments described below are all general or specific examples. The numerical values, shapes, materials, components, arrangement positions of components, and connection methods shown in the following embodiments are examples and are not intended to limit the present disclosure.

[0094] (Implementation Method)

[0095] [Configuration Example of Solid-State Imaging Device]

[0096] Hereinafter, a solid-state imaging device according to an embodiment will be described.

[0097] Figure 1 : is a block diagram showing a configuration example of the solid-state imaging device 100 according to the embodiment.

[0098] The solid-state imaging device 100 includes a pixel array unit 1, a vertical scanning circuit 14, an input / output signal line 19, a timing control unit 20, a reference signal generating unit 27, a signal processing unit 70, and an output circuit 28. In addition, the solid-state imaging device 100 includes an MCLK terminal for receiving an input of a main clock signal from the outside, a DATA terminal for transmitting and receiving commands or data to and from the outside, a D1 terminal for transmitting image data to the outside, and other terminal types to which a power supply voltage and a ground potential voltage are supplied.

[0099] The pixel array unit 1 has a plurality of pixel circuits 3 arranged in a matrix. The pixel circuits 3 output a plurality of pixel signals. The plurality of pixel signals include, for example, an HCG (High Conversion Gain) signal, an MCG (Middle Conversion Gain) signal, and an LCG (Low Conversion Gain) signal, which will be described in detail later. In addition, there are various structural examples of the pixel circuit 3. As structural examples of the pixel circuit 3, pixel circuits 3a, 3b, and 3c will be described later. In the case where the pixel circuits 3a, 3b, and 3c are not distinguished, they are referred to as pixel circuits 3. The plurality of pixel circuits 3 are Figure 1 The configuration is n rows and m columns.

[0100] The pixel array unit 1 includes a plurality of logic circuits 4 provided corresponding to the plurality of pixel circuits 3 . Figure 1 In the example shown, the pixel array unit 1 includes one logic circuit 4 for one pixel circuit 3. Alternatively, a plurality of logic circuits 4 may be provided for a plurality of pixel circuits 3. The logic circuit 4 has a function of performing AD conversion on a pixel signal output from the pixel circuit 3. The pixel signal has at least two components: a reset component and a signal component.

[0101] The AD-converted pixel signal is output to the signal processing unit 70 via the input / output signal line 19 .

[0102] The vertical scanning circuit 14 controls the horizontal scanning line group (also called row control line group) 15 provided for each row of the pixel circuit 3 in the pixel array unit 1 simultaneously and uniformly in order to implement GS (global shutter). In addition, the output signal from the logic circuit 4 is simultaneously output to the m input / output signal lines 19 (vertical signal lines) and transferred to the signal processing unit 70.

[0103] Figure 1 In the figure, for the plurality of pixel circuits 3, there is provided a horizontal scanning line group 15 (V1, V2, . . . , Vn) consisting of n pieces, each of which includes a control line of a reset transistor, a transfer transistor, and a selection transistor in the pixel circuit 3.

[0104] The timing control unit 20 controls the entire solid-state imaging device 100 by generating various control signal groups. The various control signal groups include control signal groups CN1, CN4, CN5, and CN8 and counter clock signal CK0. The timing control unit 20 obtains the master clock via the MCLK terminal, generates various internal clocks, and controls the vertical scanning circuit 14 and the like.

[0105] The reference signal generating unit 27 generates a reference signal RAMP including a triangular wave, and outputs the reference signal RAMP to a positive input terminal of a comparator 511 in an AD conversion circuit 51 described later. Figure 2A As shown in the figures, the reference signal generating unit 27 includes reference signal generating units 271 and 272. The reference signal generating unit 271 generates a reference signal 1 and outputs it to the logic circuit 4. The reference signal 1 is an example of a first reference value. The reference signal generating unit 272 generates a reference signal 2 and outputs it to the logic circuit 4. The reference signal 2 is an example of a second reference value. The reference signal 2 becomes a saturation level of a reference signal RAMP used in AD conversion and a HCG signal described later.

[0106] The signal processing unit 70 is a processing unit that performs WDR synthesis, and includes a WDR synthesis circuit 76 and a memory 77. The details of the WDR synthesis circuit 76 and the memory 77 will be described later.

[0107] The output circuit 28 outputs a digital pixel signal to the D1 terminal.

[0108] As described above, the pixel circuit 3 has various configuration examples. Figures 2A to 4 Provide explanation.

[0109] Figures 2A to 4 : is a block diagram showing a configuration example of the pixel circuit 3 and the logic circuit 4 according to the embodiment. Figure 2A to Figure 2C FIG. 2 shows a configuration example of a pixel circuit 3 having a horizontal overflow drain (OFD). Figure 3 and Figure 4 A configuration example of the pixel circuit 3 of the vertical OFD is shown.

[0110] [Structural example of pixel structure (horizontal OFD)]

[0111] First, use Figure 2A A configuration example of the pixel circuit 3 of the horizontal OFD will be described.

[0112] Figure 2A , the pixel circuit 3a is shown as one of the plurality of pixel circuits 3 arranged in a matrix in the pixel array unit 1. The solid-state imaging device 100 can further include a sample-and-hold circuit 30 connected to the pixel circuit 3a. This configuration is suitable for a global shutter method.

[0113] The pixel circuit 3 a has a pixel structure of a horizontal OFD. The horizontal OFD is a structure that includes a storage circuit in the horizontal direction.

[0114] The pixel circuit 3a includes a photodiode PD, a transfer transistor TG, a charge storage unit FD0, amplifier transistors SF1 and SF2, a reset transistor RS, and selection transistors SEL1 and SEL2. Furthermore, the pixel circuit 3a includes a storage capacitor C1, a gain control transistor GC1, a storage capacitor C2, and a gain control transistor TGC, which are arranged in a horizontal direction from the viewpoint of charge transfer.

[0115] The photodiode PD is a photoelectric conversion element that converts incident light into signal charges.

[0116] The charge storage unit FD0 is a floating diffusion unit formed as a floating diffusion layer, for example, and holds the signal charge generated by the photodiode PD. In the figure, the storage capacitance of the charge storage unit FD0 is referred to as storage capacitance C0.

[0117] The amplifier transistor SF1 outputs a pixel signal corresponding to the signal charge of the charge storage unit FD0 to the logic circuit 4 via the selection transistor SEL1. The amplifier transistor SF1 is an example of a first amplifier transistor that outputs a signal of the photodiode PD and a signal of the storage capacitor C2.

[0118] The selection transistor SEL1 is a switch element that selects whether to output or not output the pixel signal from the amplifier transistor SF1 to the logic circuit 4. The selection transistor SEL1 is an example of a first selection transistor connected to the amplifier transistor SF1.

[0119] The reset transistor RS is an example of a reset mechanism for initializing the photodiode PD, the charge storage unit FD0, and the storage capacitor C2. The reset transistor RS can perform three types of resets by controlling the gain control transistors GC1 and TGC. That is, the reset transistor RS can reset the charge storage units FD0, FD1, and FD2. In addition, the storage capacitor of the charge storage unit FD1 is set to the storage capacitor C1, and the storage capacitor of the charge storage unit FD2 is set to the storage capacitor C2.

[0120] As indicated by a dashed arrow line A in the figure, the charge overflowing from the charge storage unit FD0 when the transfer transistor TG is in the ON state is transferred to and held in the storage capacitor C1.

[0121] The gain control transistor GC1 controls the connection between the charge storage unit FD0 and the storage capacitor C1. The gain control transistor GC1 realizes an ON state and an OFF state as a switching element. The gain control transistor GC1 is an example of a gain switching transistor that switches the gain of the signal of the photodiode PD.

[0122] The signal charge overflowing from the photodiode PD during exposure is transferred to the storage capacitor C2 and held as indicated by the dashed arrow line B in the figure. Here, an overflow element OF is present between the photodiode PD and the storage capacitor C2, and during exposure, the signal charge overflowing from the photodiode PD is not transferred to the storage capacitor C1 via the transfer transistor TG, but is transferred to the storage capacitor C2 via the overflow element OF. The storage capacitor C2 stores the charge overflowing from the photodiode PD.

[0123] Furthermore, the signal charge overflowing from the storage capacitor C2 is discharged to the power supply via the gain control transistor TGC and the reset transistor RS.

[0124] In addition, as shown by the dotted arrow line B, the signal charge overflowing from the photodiode PD is accumulated in the accumulation capacitor C2 without passing through the charge accumulation unit FD0, so the influence of the dark current component in the charge accumulation unit FD0 can be reduced. In addition, due to the presence of the overflow element OF, the overflow level can be controlled, the depth of the charge channel can be controlled, and the influence of the dark current component generated in the surface layer can be reduced.

[0125] The gain control transistor TGC is a transistor for gain control that controls the connection between the storage capacitor C1 and the storage capacitor C2. The gain control transistor TGC realizes an ON state and an OFF state as a switching element. The gain control transistor TGC is an example of a second transfer transistor that transfers the charge (signal) of the storage capacitor C2 to the charge storage unit FD0.

[0126] In order to hold the signal charge overflowing from the photodiode PD, a switch element and a storage capacitor are set as a set to form a storage circuit for overflow. The first stage storage circuit is a set of gain control transistor GC1 and storage capacitor C1. The second stage storage circuit is a set of gain control transistor TGC and storage capacitor C2.

[0127] For example, in a global shutter method, all pixels are exposed at the same time, and then the signals of all pixels are read out to the sample-and-hold circuit 30 at the same time.

[0128] According to such a configuration example, it is possible to achieve both expansion of the dynamic range and improvement of SN.

[0129] The logic circuit 4 includes an AD conversion circuit 51, a detection circuit 52, and a local pixel control circuit 61. In addition, in the present embodiment, when the AD conversion circuit 51 and the detection circuit 52 are provided for a single pixel circuit 3a or for each of a plurality of pixel circuits 3a, the pixel control circuit that controls the pixel circuit 3a is specifically referred to as the local pixel control circuit 61.

[0130] The AD conversion circuit 51 is an example of an AD conversion function that converts a signal of the charge storage unit FD0 into a digital value. The AD conversion circuit 51 includes a comparator 511 , a counter circuit 512 , and a memory 513 .

[0131] The comparator 511 compares the analog pixel signal output from the pixel circuit 3a with the RAMP waveform (ie, triangular wave) as the reference signal 2 generated by the reference signal generating unit 27, and inverts the output signal indicating the comparison result when the former is greater than the latter.

[0132] The counter circuit 512 counts the time from when the triangular wave in the reference signal RAMP, which is the reference signal 2, changes to when the output signal of the comparator 511 is inverted. Since the time until the inversion is determined by the value of the analog pixel signal, the count value becomes a digitized pixel signal. In this way, AD conversion is performed.

[0133] The memory 513 holds the count value of the counter circuit 512 , that is, the digital pixel signal.

[0134] The detection circuit 52 is connected to the storage capacitor C2 via the selection transistor SEL2 and the amplifier transistor SF2. The detection circuit 52 adopts a PWM method. The detection circuit 52 compares the signal of the storage capacitor C2 with the reference signal 1. When the signal of the storage capacitor C2 reaches the reference signal 1, the photodiode PD and the storage capacitor C2 are initialized almost simultaneously via the local pixel control circuit 61, and the number of initialization times N is counted. The local pixel control circuit 61 is an example of a pixel control circuit that controls the pixel circuit 3a. The detection circuit 52 includes a comparator 521, a counter circuit 522, a memory 523, and an initialization circuit 524.

[0135] The comparator 521 compares the signal of the storage capacitor C2 with the reference signal 1. When the storage capacitor C2 reaches the reference signal 1 as the saturation level, the output signal of the comparator 521 is inverted, and the counter circuit 522 counts the number of inversions, i.e., the initialization number N, and stores it in the memory 523. The counter circuit 522 is an example of a counter function that counts the comparison results of the comparator 521.

[0136] The initialization circuit 524 controls the reset transistor RS according to the comparison result of the comparator 521. Specifically, when the storage capacitor C2 reaches the reference signal 1 as the saturation level, that is, when the inverted output signal of the comparator 521 is input, the initialization circuit 524 outputs the initialization signal to the local pixel control circuit 61 to operate the local pixel control circuit 61, and the local pixel control circuit 61 performs the pixel control required for initialization of the pixel circuit 3a.

[0137] Figure 2A In the circuit shown, input terminals of the comparator 521 and the AD conversion circuit 51 are respectively connected to different output terminals of the pixel circuit 3 a .

[0138] For example, the detection circuit 52 controls the reset transistor RS, the transfer transistor TG and the gain control transistor TGC via the local pixel control circuit 61 based on the comparison result of the signal of the storage capacitor C2 and the reference signal 1, thereby discharging the charge of the photodiode PD and the storage capacitor C2 and initializing the photodiode PD and the storage capacitor C2.

[0139] In the case of a global shutter, the logic circuit 4 can expand the dynamic range by using a PWM method for the pixel circuit 3a having an overflow structure. In addition, as described later, in the AD conversion circuit 51, a redundant read of all the multiple pixel signals is not performed, but a selective read of at least two pixel signals from the multiple pixel signals is performed. By reducing the multiple pixel signals to at least two pixel signals, the number of bits of the memories 513 and 523 can be reduced, the yield can be improved, and the subsequent stage can be accelerated and the power consumption can be reduced. In addition, the signal processing unit 70 can also achieve high speed and low power consumption, and can maintain the same dynamic range as when there are multiple (at least 3) pixel signals.

[0140] In the global shutter method, all pixels are exposed at the same time, and then the signals of all pixels are read out at the same time to the sample-and-hold circuit 30. Alternatively, the signals of all pixels are directly AD-converted by the AD conversion circuit 51. Then, the AD conversion circuit 51 supplies a plurality of pixel signals to the signal processing unit 70.

[0141] Figure 2B and Figure 2C The circuit shown is to Figure 2A The circuit shown in FIG. 5 is modified so that the detection circuit 52 and the AD conversion circuit 51 share one comparator. Figure 2B In FIG. 5 , the detection circuit 52 and the AD conversion circuit 51 share the comparator 521. Figure 2C 2 shows an example in which the detection circuit 52 and the AD conversion circuit 51 share the comparator 511 .

[0142] like Figure 2B and Figure 2C As shown, the solid-state imaging device 100 (logic circuit 4) may include a selection circuit SW1 that receives a signal from the storage capacitor C2 and a signal from the charge storage unit FD0 and selects an output. In addition, the solid-state imaging device 100 (logic circuit 4) may include a selection circuit SW2.

[0143] Figure 2B In the embodiment, the selection circuit SW1 receives the signal of the storage capacitor C2 and the signal of the charge storage unit FD0, and the selected signal is input to one end of the comparator 521. On the other hand, the selection circuit SW2 receives the reference signal 1 of the reference signal generating unit 271 and the reference signal 2 of the reference signal generating unit 272, and the selected signal is input to the other end of the comparator 521.

[0144] Figure 2C In the embodiment, the selection circuit SW1 receives the signal of the storage capacitor C2 and the signal of the charge storage unit FD0, and the selected signal is input to one end of the comparator 511. On the other hand, the selection circuit SW2 receives the reference signal 1 of the reference signal generating unit 271 and the reference signal 2 of the reference signal generating unit 272, and the selected signal is input to the other end of the comparator 511.

[0145] and Figure 2A Similarly, the signal of the storage capacitor C2 is compared with the reference signal 1, and the signal of the charge storage unit FD0 is compared with the reference signal 2. According to the control signal of the selection circuit SW1 and the control signal of the selection circuit SW2, the former comparison operation is performed during the exposure time, and the latter comparison operation is performed exclusively in time during the AD conversion (ADC) time.

[0146] In this way, the detection circuit 51 and the AD conversion circuit 52 use a common comparator, so that the area of ​​the logic circuit 4 can be reduced.

[0147] In addition, Figures 3 to 5 You can also use Figure 2B or Figure 2C Similarly, the detection circuit 51 and the AD conversion circuit 52 share one comparator.

[0148] [Structural Example 1 of Pixel Structure (Vertical OFD)]

[0149] Next, use Figure 3 A configuration example 1 of the pixel circuit 3 of the vertical OFD will be described. In the configuration example 1 of the vertical OFD, a configuration is shown in which the storage circuits of the horizontal OFD described above are connected in two stages in the vertical direction.

[0150] Figure 3, the pixel circuit 3b is shown as one of the plurality of pixel circuits 3 arranged in a matrix in the pixel array unit 1. The solid-state imaging device 100 can further include a sample-and-hold circuit 30 connected to the pixel circuit 3b. This configuration is suitable for a global shutter method.

[0151] Hereinafter, the pixel circuit 3 b will be described mainly focusing on the points that are different from the pixel circuit 3 a .

[0152] In the pixel circuit 3 b , from the viewpoint of charge transfer, the storage capacitor C1 , the gain control transistor GC1 , the storage capacitor C2 , and the gain control transistor TGC are arranged so as to be aligned in the vertical direction.

[0153] The signal charge overflowing from the photodiode PD during exposure is transferred to and held in the storage capacitor C1 as indicated by the dashed arrow line D in the figure. Furthermore, the signal charge overflowing from the storage capacitor C1 is transferred to and held in the storage capacitor C2.

[0154] The storage capacitor C1 holds signal charges overflowed from the photodiode PD. For example, the storage capacitor C1 stores signal charges overflowed from the photodiode PD during exposure via the transfer transistor TG and the gain control transistor GC1.

[0155] The gain control transistor GC1 controls the connection between the charge storage unit FD0 and the storage capacitor C1. The gain control transistor GC1 realizes an ON state and an OFF state as a switching element. In addition, the gate voltage of the transfer transistor TG and the gain control transistor GC1 can be set so that when the exposure is not completely OFF, the signal charge overflowing from the photodiode PD can be transferred from the transfer transistor TG to the storage capacitor C1 via the charge storage unit FD0 and the gain control transistor GC1.

[0156] The storage capacitor C2 holds the signal charge overflowed from the storage capacitor C1. For example, the storage capacitor C2 accumulates the signal charge overflowed from the storage capacitor C1 via the gain control transistor TGC.

[0157] The gain control transistor TGC controls the connection between the storage capacitor C1 and the storage capacitor C2. The gain control transistor TGC realizes an ON state and an OFF state as a switching element. In addition, the gate voltage of the gain control transistor TGC can be set so that it is not completely OFF when exposed, and the signal charge overflowing from the storage capacitor C1 can be transmitted to the storage capacitor C2 via the gain control transistor TGC.

[0158] With such a configuration example, it is possible to achieve both expansion of the dynamic range and improvement of SN.

[0159] Regarding the logic circuit 4, due to the use of Figure 2A The explanation is omitted because it is the same as what has been explained.

[0160] [Structural Example 2 of Pixel Structure (Vertical OFD)]

[0161] Next, use Figure 4 A second configuration example of the pixel circuit 3 of the vertical OFD will be described. The second configuration example of the vertical OFD is a modified example of the first configuration example of the vertical OFD described above.

[0162] Figure 4 , the pixel circuit 3c is shown as one of the plurality of pixel circuits 3 arranged in a matrix in the pixel array unit 1. The solid-state imaging device 100 can further include a sample-and-hold circuit 30 connected to the pixel circuit 3c. This configuration is suitable for a global shutter method.

[0163] The pixel circuit 3 c will be described below mainly focusing on the points that are different from the pixel circuit 3 a .

[0164] In the pixel circuit 3 c , from the viewpoint of charge transfer, the storage capacitor C1 , the gain control transistor GC1 , the storage capacitor C2 , and the gain control transistor TGC are arranged so as to be aligned in the vertical direction.

[0165] The signal charge overflowing from the photodiode PD during exposure is transferred to the storage capacitor C2 as indicated by the dashed arrow line E in the figure and is held.

[0166] The gain control transistor GC1 controls the connection between the charge storage unit FD0 and the storage capacitor C1. The gain control transistor GC1 realizes an ON state and an OFF state as a switching element.

[0167] The storage capacitor C2 holds the signal charge overflowing from the photodiode PD. For example, the storage capacitor C2 accumulates the signal charge overflowing from the photodiode PD during exposure via the transfer transistor TG and the gain control transistor TGC. In addition, the gate voltages of the transfer transistor TG and the gain control transistor TGC can be set so that the signal charge overflowing from the photodiode PD can be transferred from the transfer transistor TG via the charge storage unit FD0 and the gain control transistor TGC to the storage capacitor C2, which is not completely OFF during exposure.

[0168] The gain control transistor TGC controls the connection between the charge storage unit FD0 and the storage capacitor C2. The gain control transistor TGC realizes an ON state and an OFF state as a switching element.

[0169] With such a structure, it is possible to achieve both expansion of the dynamic range and improvement of SN.

[0170] Regarding the logic circuit 4, due to the use of Figure 2A The explanation is omitted because it is the same as what has been explained.

[0171] [Sharing of pixel circuits]

[0172] Figures 2A to 4 In the example described above, the AD conversion circuit 51 and the detection circuit 52 are set for each single pixel circuit 3, and the local pixel control circuit 61 controls the pixel circuit 3 for each single pixel circuit 3. However, it is also possible that the AD conversion circuit 51 and the detection circuit 52 are set for each plurality of pixel circuits 3, and the local pixel control circuit 62 controls the pixel circuit 3 for each single pixel circuit 3. Figure 5 Provide explanation.

[0173] Figure 5 This is a diagram for explaining the commonality of the logic circuit 4 d of a plurality of pixel circuits 3 . Figure 5 , pixel circuits A to D are shown as a plurality of pixel circuits 3, and these are referred to as a pixel circuit group 3d. In addition, any of the pixel circuits 3a, 3b, and 3c may be used as the pixel circuit 3. In addition, the logic circuit 4d will be described with respect to the differences from the logic circuit 4.

[0174] The pixel circuit group 3d (pixel circuits A to D) shares one AD conversion circuit 51 and one detection circuit 52. In addition, a memory 515 included in the AD conversion circuit 51 and a memory 525 included in the detection circuit 52 have areas for each pixel circuit 3. Figure 5 In FIG. 1 , the memory 515 and the memory 525 are described as A to D, and the regions where the pixel circuits A to D are provided in the memory 515 and the memory 525 are schematically shown.

[0175] The local pixel control circuit 62 is an example of a pixel control circuit that controls each of the individual pixel circuits 3 . Figure 5 In FIG. 1 , A to D are written in the local pixel control circuit 62 , schematically showing that the local pixel control circuit 62 controls the pixel circuit 3 for each single pixel circuit 3 .

[0176] The detection circuit 52 compares the signal of the storage capacitor C2 of the pixel circuit 3 with the reference signal 1 while synchronizing the corresponding pixel circuit 3 with the local pixel control circuit 62. For example, the detection circuit 52 compares the signal of the storage capacitor C2 of the pixel circuit A with the reference signal 1 while synchronizing the pixel circuit A with the local pixel control circuit 62. Thus, it is possible to determine whether the storage capacitor C2 has reached the reference signal 1, which is the saturation level, for each pixel circuit 3.

[0177] Furthermore, the local pixel control circuit 62 synchronizes the control of the selection transistor SEL1 and the selection transistor SEL2 of the corresponding pixel circuit 3 with the control of the sampling and holding circuit 30. For example, the sampling and holding circuit 30 includes a capacitor element CR for a reset component and a capacitor element CS for a signal component, and the local pixel control circuit 62 synchronizes the control of the selection transistor SEL1 and the selection transistor SEL2 of the corresponding pixel circuit 3 with the capacitor elements CR and CS. The capacitor element CR is an example of a first sampling and holding capacitor element, and the capacitor element CS is an example of a second sampling and holding capacitor element.

[0178] During exposure, the AD conversion circuit 51, the detection circuit 52, and the local pixel control circuit 62 synchronize the same pixel circuit 3 as the target. For the target pixel circuit 3, the value of the pixel signal is held in the area of ​​the target pixel circuit 3 as the memory 515 of the AD conversion circuit 51, and the number of initialization times N detected by the detection circuit 52 is held in the area of ​​the target pixel circuit 3 as the memory 525. The local pixel control circuit 62 synchronizes the selection transistor SEL1, the selection transistor SEL2, and the sample and hold circuit 30 of the target pixel circuit 3 while independently controlling them. With such a structure, the operation in the PWM mode can be implemented for a plurality of pixel circuits 3 during exposure.

[0179] In addition, for the pixel circuit group 3d, for example, since the detection is performed in the order of pixel circuits A, B, C, and D, the detection timing of the detection circuit 52 is different for each pixel circuit 3, and a delay occurs from the timing when the signal of the storage capacitor C2 exceeds the reference signal 1 to the timing when the detection circuit 52 actually detects it. However, for each pixel circuit 3, at the time when the detection circuit 52 detects that the storage capacitor C2 has reached the saturation level, the signal component of the saturation level is maintained in the sampling and holding capacitor element CS of the sampling and holding circuit 30. Thus, the signal component at the time point detected by the detection circuit 52 can be correctly maintained for each pixel circuit 3. In addition, for each pixel circuit 3, when the storage capacitor C2 and the photodiode PD are initialized, the reset component pre-maintained in the capacitor element CR is used to perform CDS of the signal component, so that the signal of the saturation level of the storage capacitor C2 can be correctly detected. In this way, even if a plurality of pixel circuits 3 share one detection circuit 52, the PWM method can be performed.

[0180] In particular, it is effective when the AD conversion circuit 51 and the detection circuit 52 are not provided for each single pixel circuit 3 in a micro unit.

[0181] [Sample and hold circuit]

[0182] Figure 6 2 is a diagram showing a configuration example of the sample-and-hold circuit 30 .

[0183] The reset component is held in the capacitor element CR by the sample hold signal SHR. The signal component is held in the capacitor element CS by the sample hold signal SHS. By setting the sample hold signal SHRST to a high level, the capacitor elements CR and CS can be discharged.

[0184] At least one pixel signal among the multiple pixel signals including the HCG signal, the MCG signal and the LCG signal is not held by the sampling and holding circuit 30, but is directly AD-converted by the AD conversion circuit 51. At least one other pixel signal among the multiple pixel signals is held by the sampling and holding circuit 30, and the held signal is AD-converted by the AD conversion circuit 51.

[0185] [Expansion of WDR dynamic range]

[0186] Here, the WDR synthesis circuit 76 included in the signal processing unit 70 is described. The WDR synthesis circuit 76 is an example of a WDR processing unit. The WDR synthesis circuit 76 performs WDR synthesis of a plurality of pixel signals output from the pixel circuit 3. As described above, the plurality of pixel signals include HCG signals, MCG signals, and LCG signals.

[0187] The signal processing unit 70 (WDR synthesis circuit 76) calculates a first signal indicating the intensity of incident light based on the number of initialization times N and a mixed signal of the signal of the photodiode PD read into the charge storage unit FD0 through the transfer transistor TG and the signal of the storage capacitor C2.

[0188] When the number of initialization times N is greater than 1, the pixel circuit 3 outputs the mixed signal of the photodiode PD and the storage capacitor C2 when the signal of the storage capacitor C2 reaches the reference signal 1 during exposure as the saturation signal Qsat of the LCG signal, and the photodiode PD and the storage capacitor C2 are initialized. The saturation signal Qsat is held by the sample and hold circuit 30, and the held signal is AD-converted by the AD conversion circuit 51. At the end of exposure, the pixel circuit 3 outputs the mixed signal remaining in the photodiode PD and the storage capacitor C2 when the signal of the storage capacitor C2 does not reach the reference signal 1 as the residual signal of the LCG signal. The residual signal is not held by the sample and hold circuit 30, but is directly AD-converted by the AD conversion circuit 51. The first signal includes the saturation signal Qsat of the LCG signal and the residual signal. The saturation signal Qsat is an example of the second signal, and the residual signal is an example of the third signal. The signal processing unit 70 (WDR synthesis circuit 76) calculates the first signal by calculating the product of the saturation signal Qsat and the number of initialization times N and the sum of the residual signal.

[0189] When the number of initialization times N is 1 or more, the sample and hold circuit 30 holds the reset component in the capacitor element CR after the photodiode PD and the storage capacitor C2 are initialized during exposure. In addition, when the signal of the storage capacitor C2 reaches the reference signal 1 during exposure, the sample and hold circuit 30 holds the saturation signal Qsat as a signal component in the capacitor element CS. In addition, the AD conversion circuit 51 compares the pixel signal level after the reset component is subtracted from the signal component with the reference signal 2 to perform AD conversion on the pixel signal level.

[0190] When the number of initialization times N is 0, the pixel circuit 3 outputs the HCG signal in a low illumination mode for low illumination. The HCG signal is an example of the fourth signal. In addition, when the number of initialization times N is 0, the pixel circuit 3 outputs the MCG signal or the LCG signal in a mode for high illumination compared to the low illumination mode. The LCG signal is an example of the sixth signal. The MCG signal is an example of the seventh signal. The LCG signal and the MCG signal belong to the fifth signal.

[0191] In the low illumination mode, the pixel circuit 3 outputs the signal of the photodiode PD as the HCG signal from the amplifier transistor SF1 via the transfer transistor TG. Specifically, the low illumination mode uses the storage capacitor C0 among the storage capacitors C0, C1 and C2 and is a mode with high sensitivity to incident light of low illumination.

[0192] Compared to the low illumination mode, the high illumination mode includes a medium illumination mode and a high illumination mode.

[0193] In the medium illumination mode, the pixel circuit 3 controls the gain control transistor GC1 to output the signal of the photodiode PD from the amplifier transistor SF1 as the MCG signal. Specifically, the medium illumination mode uses the storage capacitors C0 and C1 of the storage capacitors C0, C1, and C2, and is a mode capable of detecting incident light of medium illumination.

[0194] In the high illuminance mode, the pixel circuit 3 outputs the signal of the storage capacitor C2 as an LCG signal from the amplifier transistor SF1 via the gain control transistors TGC and GC1. Specifically, the high illuminance mode uses the storage capacitors C0, C1, and C2 and is a mode capable of detecting incident light of high illuminance.

[0195] The HCG signal is not held by the sample-and-hold circuit 30 but is directly AD-converted by the AD conversion circuit 51. The MCG signal or the LCG signal is held by the sample-and-hold circuit 30 and the held signal is AD-converted by the AD conversion circuit 51.

[0196] The WDR synthesis circuit 76 generates an output signal based on at least the first signal when the number of initialization times N is greater than or equal to 1. The WDR synthesis circuit 76 generates an output signal based on at least one of the HCG signal, the MCG signal, and the LCG signal when the number of initialization times N is 0. The generated output signal is output to the output circuit 28.

[0197] Although the details will be described later, when performing WDR synthesis, the pixel signal used to generate the output signal can be determined based on only the HCG signal among the plurality of pixel signals. Therefore, the number of pixel signals output from the pixel circuit 3 can be reduced, so that the speed of reading in the logic circuit 4 can be increased. In addition, the number of bits of the memories 513 and 523 of the logic circuit 4 can be reduced, thereby achieving area reduction and suppression of yield reduction.

[0198] Regarding the exposure control of the three frames obtained by high illumination exposure (high illumination mode), medium illumination exposure (medium illumination mode) and low illumination exposure (low illumination mode) for WDR synthesis, in the global shutter mode, they are all exposed at the same timing and use the same pixels, so false color, coloring and time deviation no longer occur. In other words, since each frame is exposed at the same time, there is no time deviation and coloring between frames.

[0199] The pixel signal of the frame in the low illumination area (low illumination exposure) is an HCG signal with a high gain in the pixel, the pixel signal of the frame in the medium illumination area (medium illumination exposure) is an MCG signal with a middle gain in the pixel, and the pixel signal of the frame in the high illumination area (high illumination exposure) is an LCG signal with a low gain in the pixel. By using these pixel signals, noise improvement under low illumination and expansion of dynamic range under high illumination can be performed.

[0200] Figure 7 This is a diagram showing an overview of a method for implementing WDR synthesis within a pixel.

[0201] Figure 8 : is a diagram showing a configuration example of a signal processing unit 70 that performs WDR synthesis using three signals. In this example, the signal processing unit 70 includes an HCG signal generating circuit 71, an MCG signal generating circuit 72, and an LCG signal generating circuit 73. The HCG signal generating circuit 71 generates an HCG signal based on the digital value of a pixel signal in a low illumination mode. The MCG signal generating circuit 72 generates an MCG signal based on the digital value of a pixel signal in a medium illumination mode. The LCG signal generating circuit 73 generates an LCG signal based on the digital value of a pixel signal in a high illumination mode.

[0202] WDR synthesis is performed based on the signal charge Q0 (charge of storage capacitor C0) of the HCG signal of low illumination exposure (low illumination mode), the signal charges Q0 and Q1 (charges of storage capacitors C0 and C1) of the MCG signal of medium illumination exposure (medium illumination mode), and the signal charges Q0, Q1 and Q2 (charges of storage capacitors C0, C1 and C1) of the LCG signal of high illumination exposure (high illumination mode). Figure 7 The horizontal axis of (a) to (d) represents the product of illumination and exposure time, illumination for a certain time, or exposure time under a certain illumination. Figure 7 (a) shows the charge accumulation level. Figure 7 (b) shows the signal potential of the pixel portion. Figure 7 (c) shows the value after AD conversion. Figure 7 (d) shows the SN level.

[0203] Figure 7 (c) and (d) are diagrams for explaining the details of the WDR synthesis circuit 76 constituting the signal processing unit 70. The transition area between the low illumination exposure and the medium illumination exposure in the "illuminance exposure time" is the boundary T1. Figure 7 As shown in (a), the charge of the storage capacitor C0 becomes the charge Q0' just before saturation. Figure 7 As shown in (b) of FIG. 1 , the signal potential of the pixel portion becomes the signal potential FDH. Regarding the transition region between the medium illumination exposure and the high illumination exposure in the “illuminance exposure time”, that is, the boundary T2, as shown in FIG. Figure 7 As shown in (a), the charges of the storage capacitors C0 and C1 become the charge Q0+Q1′ just before saturation. Figure 7 As shown in (b), the signal potential of the pixel portion becomes the signal potential FDM. In addition, at the boundary T3 of the "illuminance exposure time" in the area with the maximum high illuminance exposure, as shown in Figure 7 As shown in (a), the charges of the storage capacitors C0, C1, and C2 become the charge Q0+Q1+Q2′ just before saturation. Figure 7 As shown in (b) of FIG. 8 , the signal potential of the pixel portion becomes the signal potential FDL.

[0204] The WDR synthesis circuit 76 realizes the WDR function by synthesizing the pixel signal of the low illumination frame (low illumination exposure), the pixel signal of the medium illumination frame (medium illumination exposure) and the pixel signal of the high illumination frame (high illumination exposure), that is, the dynamic range is expanded.

[0205] Fig. 9 2 is a diagram showing a configuration example of a signal processing unit that performs WDR synthesis using two signals selected from three signals. In this example, the signal processing unit 70 includes an interpolation circuit 74 .

[0206] The memory 77 stores the gain ratio or sensitivity ratio of the pixel values ​​of the HCG signal, the MCG signal, and the LCG signal corresponding to the illumination and exposure time. The gain ratio or sensitivity ratio is measured in advance and stored in the memory 77. For example, the memory 77 stores Figure 7 Gain 1 (Gain1) and gain 2 (Gain2) are shown in (c). Gain 1 is the gain ratio of the MCG signal to the HCG signal. Gain 2 is the gain ratio of the LCG signal to the HCG signal. In addition, the gain ratio of the LCG signal to the MCG signal can be calculated based on Gain 2 / Gain 1, or a pre-determined value can be stored. Thus, if one value of the HCG signal, the MCG signal, and the LCG signal is determined, the other two values ​​can be respectively calculated by the interpolation processing of the interpolation circuit 74 using the above-mentioned gain ratio. In addition, the memory 77 can also store the sensitivity ratio instead of the gain ratio.

[0207] The interpolation circuit 74 obtains the pixel value of one pixel signal that is not selected by interpolation processing based on two pixel signals selected by the logic circuit 4 among the three pixel signals of the HCG signal, the MCG signal, and the LCG signal. The pixel value of one pixel signal that is not selected by the logic circuit 4 is obtained by interpolation processing in the interpolation circuit 74 and input to the WDR synthesis circuit 76. Interpolation processing means interpolation or estimation.

[0208] The WDR synthesis circuit 76 synthesizes the pixel signal selected by the logic circuit 4 and the pixel signal obtained by the interpolation process of the interpolation circuit 74 from the data of the gain ratio or the sensitivity ratio stored in the memory 77 so that the three pixel signals become linear.

[0209] For example, Figure 7 As shown in (b), regarding the HCG signal and the MCG signal, when the pixel values ​​PH and PM of the pixel signal surrounded by the single-dot chain frame are selected, the LCG signal is obtained by interpolation processing based on the gain ratio or sensitivity ratio stored in the memory 77.

[0210] Likewise, if Figure 7 As shown in (b), regarding the MCG signal and the LCG signal, when the pixel values ​​PM and PL of the pixel signal surrounded by the single-dot chain frame are selected, the HCG signal is obtained by interpolation processing based on the gain ratio or sensitivity ratio stored in the memory 77.

[0211] Thus, the signal processing unit 70 obtains two of the pixel values ​​PH of the HCG signal, the pixel value PM of the MCG signal, and the pixel value PL of the LCG signal. Furthermore, in the signal processing unit 70, the pixel value PH of the HCG signal, the pixel value PM of the MCG signal, and the pixel value PL of the LCG signal are obtained by interpolation using the interpolation circuit 74. The signal processing unit 70 synthesizes the three pixel values ​​obtained in this way so as to be linear, thereby being able to correspond to Figure 7 The mixed charge shown in (a) is obtained Figure 7 (c) shows a linear digital pixel signal.

[0212] Furthermore, at the boundary T1 which is a transition region between low illuminance exposure and medium illuminance exposure, and the boundary T2 which is a transition region between medium illuminance exposure and high illuminance exposure, it is desirable to suppress abrupt changes in the SN of the subject near the boundary as much as possible.

[0213] [Basic readout operation of HCG, MCG, and LCG signals in WDR synthesis]

[0214] Fig.10 1 is a diagram showing a basic readout timing example of the pixel circuit 3 .

[0215] First, from time t1 to time t2, the photodiode PD and the charge storage units FD0, FD1, and FD2 are reset and standby until exposure starts. Then, from time t3 to time t4, the shutter operation is performed and exposure starts. At time t5, the exposure time ends.

[0216] At time t5, the signal charge accumulated in the photodiode PD is transferred to the charge accumulation unit FD1 by turning on the gain control transistor GC1. The reset operation is performed during the period from time t5 to time t6. The reset component of the MCG signal is read during the period from time t6 to time t7.

[0217] Next, a reset operation is performed during a period from time t7 to time t8. A reset component of the HCG signal is read during a period from time t8 to time t9.

[0218] Next, from time t9 to time t10, the signal charge accumulated in the photodiode PD is transferred to the charge accumulation unit FD0 by the transfer transistor TG. The HCG signal component is read out during the period from time t10 to time t11.

[0219] At time t11, the signal charge accumulated in the photodiode PD is transferred to the charge storage unit FD1 by turning on the gain control transistor GC1. From time t11 to time t12, the signal charge accumulated in the photodiode PD is transferred to the charge storage unit FD1 by the transfer transistor TG. During the period from time t12 to time t13, the MCG signal component is read out.

[0220] At time t13, the signal charge accumulated in the photodiode PD is transferred to the charge storage unit FD2 by turning on the gain control transistor TGC. From time t13 to time t14, the signal charge accumulated in the storage capacitor C2 is transferred to the charge storage unit FD2 through the gain control transistor TGC. During the period from time t14 to time t15, the LCG signal component is read out.

[0221] The reset operation is performed during the period from time t15 to time t16. The reset component of the LCG signal is read during the period from time t16 to time t17.

[0222] At time t16, each transistor is turned OFF.

[0223] In this way, the HCG signal, the MCG signal, and the LCG signal are read out.

[0224] [Selective reading operation of two signals, HCG signal, MCG signal, and LCG signal, in WDR synthesis]

[0225] Fig.11 This is a diagram showing the timing of the selective WDR method. Fig.11 , an example is described in which two signals are selected from three signals, namely, the HCG signal, the MCG signal, and the LCG signal, by detecting the level of the HCG signal.

[0226] In the single slope AD conversion circuit 51, if the counter circuit 512 is a binary counter, the digital value is proportional to the HCG signal level, so the most appropriate two signals can be selected. However, in order to reduce power consumption, a gray code counter is more suitable, and since the digital value is not proportional to the HCG signal level, the most appropriate two signals cannot be selected.

[0227] Therefore, in Fig.11 During the AD conversion period of the HCG signal from time t104 to time t106 shown in FIG. 1 , after the HCG signal is AD-converted by the comparison operation of the comparator 511 of the AD conversion circuit 51, Fig.11 During the HCG signal detection period from time t106 to time t107 shown in FIG. 1 , the reference signal 2 is set to the saturation level of the HCG signal ( Figure 7 During this period, the reference signal 2 is the saturation level of the HCG signal, and the pixel signal becomes the HCG signal when it does not exceed the saturation level. Therefore, the comparator 511 outputs a low level as the selection signal S1, and the HCG signal and the MCG signal are selected. On the other hand, when the saturation level (signal potential FDH) is exceeded, the pixel signal is not the HCG signal, so the comparator 511 outputs a high level as the selection signal S1, and the MCG signal and the LCG signal are selected.

[0228] Specifically, the comparator 511 compares the potential of the charge storage part FD0 of the storage capacitor C0 with the potential of the reference signal 2 output by the reference signal generating part 272 via the amplifying transistor SF1 and the selecting transistor SEL1, and outputs a low level when the potential of the charge storage part FD0 of the storage capacitor C0 is high, and outputs a high level selection signal S1 to the local pixel control circuit 61 when the potential is low, so that it controls the pixel circuit 3.

[0229] Thus, when the number of initialization times N is 0, the pixel circuit 3 outputs more than 3 pixel signals, and the comparator 511 compares one pixel signal (for example, the HCG signal) among the more than 3 pixel signals with the reference signal 2 (for example, the saturation level of the HCG signal) after AD conversion, and generates a selection signal S1 indicating the selection of at least two pixel signals (for example, the MCG signal and the LCG signal) among the more than 3 pixel signals when the one pixel signal reaches the reference signal 2. Furthermore, the local pixel control circuit 61 selects at least two pixel signals from the more than 3 pixel signals according to the selection signal S1.

[0230] As a result, two signals can be selected from three signals: the HCG signal, the MCG signal, and the LCG signal.

[0231] [Generation of kTC noise in sample-and-hold circuit]

[0232] When the output of the amplifier transistor SF1 is connected to the sample-and-hold circuit 30 (the so-called Voltage-Domain method), kTC noise is generated when the pixel signal (two components of the reset component and the signal component) is sampled. In this case, the SN at low illumination deteriorates, making it difficult to expand the dynamic range. In order to improve the kTC noise, a high-density and low-leakage capacitor element is required, which has the problem of being difficult to achieve in terms of process.

[0233] Therefore, according to the present disclosure, the HCG signal in WDR synthesis is directly AD-converted to reduce noise, and the data is transferred to the signal processing unit 70. On the other hand, the MCG signal and the LCG signal, which are not critical to noise characteristics, are sampled and held by the sample-and-hold circuit 30, and then AD-converted, and the free memory 513 is used.

[0234] In this way, the memory 513 is mounted only for one of the HCG signal, the MCG signal, and the LCG signal. This can suppress the degradation of the noise characteristics of the WDR synthesis, the increase in the area of ​​the memory 513, and the degradation of the yield.

[0235] [Basic operation of PWM method]

[0236] use Fig.12 A timing example of the PWM method will be described.

[0237] Fig.12 This is a diagram showing a basic timing example of the PWM method. Fig.12 The timing diagram of shows a two-frame imaging operation. The first frame corresponds to relatively strong incident light, and the second frame corresponds to relatively weak incident light.

[0238] The photodiode PD generates charges by photoelectric conversion, and the charges overflowing from the photodiode PD are accumulated in the storage capacitor C2 via the overflow element OF, and the storage capacitor C2 maintains a potential corresponding to the amount of the generated charges. Since the charges are generated according to the amount of incident light, the potential of the charge storage part FD2 of the storage capacitor C2 decreases from the initial voltage over time.

[0239] When the potential of the charge storage unit FD2 of the storage capacitor C2 drops to the reference signal 1, the local pixel control circuit 61 initializes the storage capacitor C2 to an initial voltage.

[0240] Here, the reference signal 1 is preferably Figure 7 The voltage corresponding to the charge Q2' accumulated in the storage capacitor C2 shown in (a) of FIG. 1 is given by Q2' / C2.

[0241] For example, the local pixel control circuit 61 sets the reset transistor RS, the transfer transistor TG, the gain control transistor GC1, and the gain control transistor TGC to a high level when the potential of the charge storage unit FD2 of the storage capacitor C2 drops to the reference signal 1, and initializes the photodiode PD and the storage capacitor C2. The initial voltage may be, for example, a power supply voltage or a predetermined voltage value. By initializing the photodiode PD and the storage capacitor C2, the reset transistor RS, the transfer transistor TG, the gain control transistor GC1, and the TGC are turned off through the local pixel control circuit 61, and the initialization is released.

[0242] The counter circuit 522 starts counting from an initial value of 0. The counter circuit 522 counts the number of initialization times N when the local pixel control circuit 61 initializes the photodiode PD and the storage capacitor C2, and sets a digital value of N times the saturation signal Qsat of the LCG signal representing the intensity of the incident light as the digital output signal OUT1 and outputs a signal corresponding to the number of initialization times N.

[0243] In this way, the solid-state imaging device 100 outputs the digital output signal OUT1 for each pixel unit as the initialization number N. This achieves an expansion of the dynamic range, and even when strong light exceeding the saturation charge amount of the storage capacitor C2 is irradiated to the photodiode PD, the photodiode PD and the storage capacitor C2 are initialized and a digital signal is generated, thereby enabling imaging in a bright scene at a level significantly exceeding the saturation charge amount of the storage capacitor C2.

[0244] [PWM method and AD conversion method]

[0245] Fig.13 This is a diagram showing a timing example when the AD conversion circuit 51 is used in the PWM method.

[0246] The comparator 521 compares the potential of the charge storage unit FD2 of the storage capacitor C2 with the reference signal 1. When they match, the local pixel control circuit 61 turns on the reset transistor RS, the transfer transistor TG, and the gain control transistors GC1 and TGC to perform initialization.

[0247] Specifically, the comparator 521 compares the potential of the charge storage part FD2 of the storage capacitor C2 with the potential of the reference signal 1 output by the reference signal generating part 271 via the amplifying transistor SF2 and the selecting transistor SEL2. When the potential of the charge storage part FD2 of the storage capacitor C2 is lower, a high level is output to the local pixel control circuit 61 as a control signal for initializing the reset transistor RS, the transfer transistor TG, the gain control transistor GC1 and the gain control transistor TGC.

[0248] The initialization circuit 524 generates an initialization signal. The initialization signal may be, for example, a power supply voltage or other voltage values. The initialization signal is supplied as a control signal for initializing the photodiode PD and the storage capacitor C2 via the local pixel control circuit 61 .

[0249] The reference signal generating unit 271 generates a reference signal 1. The reference signal 1 may be, for example, the saturation level of the storage capacitor C2 (the voltage value of the charge storage unit FD2 when the charge held in the storage capacitor C2 is saturated), or a voltage value between the saturation level and the power supply voltage.

[0250] Here, the reference signal 1 is preferably Figure 7 The voltage corresponding to the charge Q2' stored in the storage capacitor C2 shown in (a) of FIG. 1 is given by Q2' / C2.

[0251] The count control signal generating unit 80 generates a count initialization signal INIT and a count stop signal STOP to control the counter circuit 522. The count initialization signal INIT is a control signal for initializing the count value of the counter circuit 522 to 0. The count stop signal STOP is a signal for controlling whether the counting operation of the counter circuit 522 is to be performed or stopped.

[0252] When the initialization signal is at a high level, the local pixel control circuit 61 turns on the reset transistor RS, the transfer transistor TG, and the gain control transistors GC1 and TGC to initialize the photodiode PD and the storage capacitor C2.

[0253] When the count stop signal STOP output by the count control signal generating unit 80 is at a low level, the counter circuit 522 counts the number of initialization times N in which the initialization signal changes from a low level to a high level, and outputs the counting result as a digital output signal OUT1. When the count stop signal STOP is at a high level, the counting is stopped, and when the count initialization signal INIT becomes a high level, the count value is initialized to 0.

[0254] In addition, for example, by making the reference signal 1 a power supply voltage higher than the output voltage VINI of the amplifier transistor SF2, the initialization signal becomes high level, and the photodiode PD and the storage capacitor C2 can be initialized. By making the count initialization signal INIT and the count stop signal STOP of the counter circuit 522 high level, the digital output signal OUT1 is initialized to "0".

[0255] In addition, for example, by making the count initialization signal INIT and the count stop signal STOP low level, and making the reference signal 1 a voltage lower than the output voltage VINI of the amplifier transistor SF2, the initialization signal becomes low level, and the initialization can be stopped. The output voltage of the amplifier transistor SF2 decreases from the output signal VINI at a speed corresponding to the light irradiation amount. When it reaches the reference signal 1, the initialization signal changes to a high level, the digital output signal OUT1 becomes "1", and the photodiode PD and the storage capacitor C2 are initialized, the output voltage of the amplifier transistor SF2 becomes the output signal VINI, and the initialization signal becomes low level.

[0256] In addition, the output voltage of the amplifier transistor SF2 decreases from the output signal VINI at a speed corresponding to the amount of light exposure. When it reaches the reference signal 1, the initialization signal is changed to a high level, the digital output signal OUT1 becomes "2", and the photodiode PD and the storage capacitor C2 are initialized. The output voltage of the amplifier transistor SF2 becomes the output signal VINI, and the initialization signal becomes a low level.

[0257] This operation is continuously performed until the exposure time ends, and the digital output signal OUT1 increases until the exposure time ends. The digital output signal OUT1 at the end of the exposure time corresponds to the number of initialization times N.

[0258] Furthermore, the count stop signal STOP becomes high level, so that the counter circuit 522 continues to hold the count value at that point in time, and generates a digital output signal OUT1 which is a digital signal corresponding to the amount of light irradiated to the photodiode PD for each pixel unit.

[0259] The solid-state imaging device 100 resets when a certain amount of charge is accumulated in the storage capacitor C2, and outputs the number of resets as a digital value. The solid-state imaging device 100 can realize high-speed imaging by generating a digital signal for each pixel unit. Furthermore, even when the photodiode PD is irradiated with strong light to the extent that the storage capacitor C2 cannot perform photoelectric conversion, the photodiode PD and the storage capacitor C2 are initialized and a digital signal is generated, so imaging can be performed even in a bright scene that greatly exceeds the saturation charge amount of the photodiode PD and the storage capacitor C2.

[0260] Furthermore, since the photodiode PD and the storage capacitor C2 can capture images with initialization even if the light receiving area is reduced, the solid-state imaging device 100 can be made smaller and less expensive.

[0261] Next, use Fig.13 An operation example of the AD conversion circuit 51 will be described.

[0262] The AD conversion circuit 51 performs AD conversion on the pixel signal (LCG signal) under high illumination and the residual signal obtained by mixing the photodiode PD initialized last and the FD of the storage capacitor C2, and outputs the AD-converted data as a digital output signal OUT2 indicating the intensity of incident light.

[0263] Specifically, the AD conversion circuit 51 outputs a digital output signal OUT2 corresponding to a potential obtained by mixing the saturation signal Qsat and the photodiode PD and the storage capacitor FD after the counter circuit 512 stops counting.

[0264] The signal processing unit 70 generates a digital output signal OUT3 (ie, a first signal) indicating the intensity of incident light by combining the digital output signal OUT1 and the digital output signal OUT2 .

[0265] exist Fig.13 In the example, the digital output signal OUT1 of the counter circuit 522 is N, and the digital output signal OUT2 of the counter circuit 512 is M. Let the digital value of the saturation signal Qsat after the photodiode PD and the FD of the storage capacitor C2 be the digital value Dsat. At this time, in the signal processing unit 70, the total digital output signal OUT3 becomes "Dsat×N+M".

[0266] The solid-state imaging device 100 performs AD conversion on the residual signal and outputs it digitally according to the number of initializations. This makes it possible to output a digital signal with higher resolution according to the light irradiation amount, and realize good imaging that reproduces brightness and color that are closer to the outside world.

[0267] [Signal selection in units of color filter arrangement]

[0268] Signal selection in units of color filter arrangement will be described by taking four colors (R, Gr, B, and Gb), which is a representative Bayer arrangement, as an example.

[0269] Here, using Figure 7 , illustrating the relationship between the boundary T1 of the HCG signal and the MCG signal and the boundary T2 of the MCG signal and the LCG signal and the color filter.

[0270] Normally, the boundaries T1 and T2 as the transition region change in conjunction with the analog gain or the digital gain.

[0271] For example, in the case of the Bayer arrangement, the sensitivity of the four colors (R, Gr, B, and Gb) is different, Gr and Gb have the highest sensitivity, and B and R have lower sensitivity. Therefore, the boundaries T1 and T2 of Gr and Gb become lower than those of B and R. In conjunction with this, the voltage levels of the boundaries T1 and T2 of each color change.

[0272] There are two methods for synthesizing WDR. Fig.14 and Fig.15 Provide explanation.

[0273] Fig.14 This is a diagram for explaining an example of selecting HCG signal, MCG signal, and LCG signal by giving priority to the largest signal of each color in Bayer units.

[0274] The first is to use the Bayer arrangement in Fig.14 The synthesis method shown in the thick frame in the middle is to synthesize three frames using the maximum signal gain of the four colors (R, Gr, B, and Gb).

[0275] In this case, it is consistent with the color with the highest signal level among the four colors (R·Gr·B·Gb). For example, if the color with the highest signal level is Gr and is detected as an LCG signal, the other three colors (R·B·Gb) are also set as LCG signals. For example, if the color with the highest signal level is R and is detected as an MCG signal, the other three colors (Gr·B·Gb) are also set as MCG signals.

[0276] In this way, the plurality of pixel circuits 3 may correspond to the arranged pixels of the color filter, and the local pixel control circuit 62 may perform the same pixel control on the plurality of pixel circuits 3 in accordance with the maximum signal level of the arranged pixels of the color filter.

[0277] Thus, the same gain is achieved in the units of the color filter arrangement, for example, the Bayer arrangement consisting of 4 colors, and coloration caused by linear deviation of each color that may occur when the gain of each color is different can be avoided. In addition, through AE control, even if the gain of each color changes, the same gain is always achieved. Figure 7 The maximum signal level of the four colors (R, Gr, B, Gb) is used in the SN level shown in (d), and the boundaries T1 and T2 as the transition area can be changed in conjunction. Since the gains of the colors before and after the transition area of ​​WDR, i.e., the boundary part (HCG signal, MCG signal, LCG signal) are the same, coloration caused by linear deviation of each color does not occur.

[0278] Fig.15 This is a diagram for explaining an example in which an HCG signal, an MCG signal, and an LCG signal are independently selected for each color.

[0279] The second is in the case of Bayer arrangement, according to Fig.15 The 4 color (R, Gr, B, Gb) signals shown in the figure use separate gains for each of the 3 frames. Even when the gain of each color is changed by the AE operation, Figure 7In the SN level shown in (d), the transition regions of the four colors (R, Gr, B, Gb), namely, the boundaries T1 and T2, are changed independently. In this case, each color selects the HCG signal, the MCG signal, or the LCG signal according to its signal level.

[0280] Thus, for example, when the white balance changes when the color temperature of the subject changes, WDR synthesis of each frame is possible even if the analog gain or digital gain of each color is variable by AE control when the gain is adjusted for each color.

[0281] Thus, in WDR synthesis, the boundaries T1 and T2 of each color can be maximized, thereby achieving a good SN effect.

[0282] [Structural example of stacked BSI of pixel circuit and detection selection circuit]

[0283] Next, a configuration example of a stacked BSI (Back Side Illumination) of the pixel circuit 3 and the logic circuit 4 will be described.

[0284] Figures 16 to 19 FIG. 1 is a diagram showing an example in which the solid-state imaging device 100 according to the embodiment is configured as a stacked image sensor.

[0285] Each figure ( Figures 16 to 19 ) includes a first semiconductor chip (first semiconductor substrate) and a second semiconductor chip (second semiconductor substrate) or three or more semiconductor chips that are bonded together. The semiconductor chip is an example of a semiconductor substrate. The first semiconductor chip and the second semiconductor chip are bonded and stacked with each other on their respective wiring layer side surfaces. PD in each figure represents a photodiode PD, C represents a sampling and holding capacitor (for example, capacitor elements CR and CS), Ana represents an analog circuit (for example, comparators 511 and 521), Logic and LG represent logic circuits (for example, counter circuits 512 and 522), and MEM represents a memory circuit (for example, memories 513 and 523).

[0286] The pixel circuit 3 is formed on a first semiconductor substrate composed of one or more semiconductor layers, the detection circuit 52 is formed on a second semiconductor substrate composed of one or more semiconductor layers, and the first semiconductor substrate and the second semiconductor substrate are stacked. In addition, the AD conversion circuit 51 is formed on the second semiconductor substrate. In addition, the selection circuits SW1 and SW2 are formed on the first semiconductor substrate or the second semiconductor substrate.

[0287] The first semiconductor chip is referred to as a pixel chip in each figure, and includes a main part of the solid-state imaging device 100 which is a back-illuminated CMOS image sensor. The first semiconductor chip includes a photodiode PD and a storage capacitor C2 of each pixel circuit 3 .

[0288] The second semiconductor chip is referred to as a logic chip in each figure ( Fig.18 The second semiconductor chip includes a detection circuit 52 and an AD conversion circuit 51. The second semiconductor chip includes, for example, a signal processing unit 70 and other main analog circuits (eg, comparators 511 and 521), and logic circuits (eg, counter circuits 512 and 522).

[0289] In such a solid-state imaging device 100, a stacked BSI type CIS (Contact Image Sensor) is provided so that a pixel chip can be equipped with a pixel circuit 3, and a logic chip can be equipped with a logic circuit 4. That is, a photodiode PD can be mounted on each pixel circuit 3 in the pixel chip, and a logic circuit 4 can be formed in the logic chip. In addition, since the pixel signal can be read directly to the logic circuit 4, the logic chip can be formed only of analog circuits and logic circuits, which facilitates integration and reduces the layout area. Furthermore, at least two of the most appropriate signals are selected from the plurality of pixel signals from the pixel array unit 1 through the logic circuit 4. Thus, the number of bits of the memories 513 and 523 of the logic circuit 4 can be reduced and the area can be reduced. Furthermore, the speed of the memories 513 and 523 and the signal processing unit 70 can be increased and the power consumption can be reduced.

[0290] Fig. 20 and Fig.21 It is a diagram showing a connection example of the pixel circuit 3 and the logic circuit 4 according to the embodiment.

[0291] The pixel chip and the logic chip may also be provided with Fig. 20 and Fig.21 The described joints A and B. Fig. 20 Corresponding to Figure 2 to Figure 4 , Fig.21 Corresponds to Figure 5 .

[0292] Fig.16 In the example shown, a capacitor element of a sampling and holding circuit 30 is mounted for each pixel circuit 3 of a pixel chip, and a logic circuit 4 is mounted in a logic chip. The capacitor element that can constitute the sampling and holding circuit 30 in the pixel chip can make the noise tolerance to the surrounding of the digital signal good. The logic chip can be composed of only analog circuits and logic circuits, which is easy to integrate and reduces the layout area. The multiple pixel signals from the pixel array unit 1 are selected by the logic circuit 4 to select at least the two most appropriate signals, and the number of pixel signals is reduced, so the time and power for transferring the digital signal to the signal processing unit 70 can be reduced, and high speed and low power can be achieved.

[0293] Fig.17In the example shown, each pixel is mounted in the pixel chip, and a capacitor element of the sampling and holding circuit 30 is mounted in the logic chip corresponding to each pixel circuit 3. The logic chip is mounted with a logic circuit 4. The capacitor element of the sampling and holding circuit 30 can be configured in the logic chip, which can improve the parasitic sensitivity. In addition, if there is a limitation in the mounting area, the capacitor element of the sampling and holding circuit 30 can be divided into a pixel chip and a logic chip and mounted. Fig.17 The example shown can also be used with Fig.16 The example shown also achieves high speed and low power consumption.

[0294] Fig.18 In the example shown, each pixel is mounted in the pixel chip, a capacitor element of the sampling and holding circuit 30 is mounted in the logic chip A corresponding to each pixel circuit 3, and a memory is mounted in the logic chip B corresponding to each pixel circuit 3. The memory can be configured in the logic chip B, and the output signal from the memory can be accelerated. Fig.18 The example shown can also be used with Fig.16 The example shown also achieves high speed and low power consumption.

[0295] Fig.19 In the example shown, the capacitor element of the sampling and holding circuit 30 of each pixel circuit 3 is mounted in the pixel chip, and the logic circuit 4d is mounted in the logic chip. The capacitor of the sampling and holding circuit 30 can be formed in the pixel chip, and the noise resistance against the surrounding of the digital signal can be improved. The multiple pixel circuits 3 constituting the pixel circuit group 3d share the logic circuit 4d. The logic chip can be composed of only analog circuits and logic circuits, which is easy to integrate and reduces the layout area. Fig.19 The example shown can also be used with Fig.16 The example shown also achieves high speed and low power consumption.

[0296] [Specific action example]

[0297] Next, use Figure 22 to Figure 24 A more specific operation example of the solid-state imaging device 100 using the PWM method and the selective WDR method will be described.

[0298] Fig. 22 It is a diagram showing a timing example when the HCG signal and the MCG signal are selected.

[0299] Fig.23 This is a diagram showing a timing example when the MCG signal and the LCG signal are selected.

[0300] Fig.24 FIG. 1 is a diagram showing a timing example when the saturation level of the LCG signal and the residual signal are selected.

[0301] In addition, Figure 22 to Figure 24 In the description Figure 5 In the case where a plurality of pixel circuits 3 (pixel circuits A to D) share a logic circuit 4d, Fig.11 The timing example of the selective WDR of the selective readout action. Figures 2A to 4 Even when the logic circuit 4 is provided for each single pixel circuit 3 in this manner, the following description can be applied by focusing on one pixel circuit 3 among the pixel circuits A to D.

[0302] In the following, it is assumed that the number of initialization times N is 0 when the incident light has a low to medium illuminance, and the number of initialization times N is 1 or more when the incident light has a high illuminance.

[0303] First, when the incident light is at low illumination and the level of the HCG signal does not exceed the reference signal 2 (the saturation level of the HCG signal), the Fig. 22 At this time, two signals, the HCG signal and the MCG signal, are selected and synthesized in the WDR synthesizing circuit 76 of the signal processing unit 70. The LCG signal is not required.

[0304] From time 102 to time 103, due to low illumination, the charge of the photodiode PD during exposure does not exceed the specified threshold of the overflow element OF. Therefore, no charge is accumulated in the storage capacitor C2, and the detection circuit 52 does not operate. First, after the shutter (SHT) time, the gain control transistors GC1 and TGC are turned on, and the reset component of the LCG signal is sampled and held in the capacitor element CR of the sample and hold circuit 30 through the sample and hold signal SHR. However, since the storage capacitor C2 does not exceed the saturation level during exposure, the detection circuit 52 does not operate, so the sampled and held charge is discharged.

[0305] From time 103 to time 104, the gain control transistor GC1 is turned on, and the reset component of the MCG signal is sampled and held in the capacitor element CR of the sample and hold circuit 30 through the sample and hold signal SHR.

[0306] From time 104 to time 106 , the reset component and the signal component of the HCG signal are AD-converted and stored in the memory 513 .

[0307] exist Fig. 22 In the description of , the following is assumed: from time 106 to time 107, by Fig.11 In the selective readout operation, the level of the HCG signal does not exceed the saturation level (signal potential FDH). In addition, if it exceeds Fig.23 illustrate.

[0308] From time 107 to time 108, the gain control transistor GC1 is turned on, and the signal component of the MCG signal is sampled and held in the capacitor element CS of the sample and hold circuit 30 through the sample and hold signal SHS.

[0309] From time 108 to time 110 , AD conversion of the reset component and the signal component of the sampled and held MCG signal is not performed because the data of the HCG signal has already been accumulated in the memory 513 .

[0310] From time 110 to time 112, the reset component and the signal component of the LCG signal are not read out because WDR synthesis can be performed using two signals, the HCG signal and the MCG signal.

[0311] From time 115 to time 116, the data of the HCG signal in the memory 513 is transferred to the signal processing unit 70 in the next exposure.

[0312] From time 116 to time 118, the reset component of the MCG signal sampled and held in the capacitor element CR of the sample and hold circuit 30 by the sample and hold signal SHR, and the signal component of the MCG signal sampled and held in the capacitor element CS of the sample and hold circuit 30 by the sample and hold signal SHS are AD converted and accumulated in the memory 513 which is idle from time 115 to time 116.

[0313] From time 118 to time 119, the data of the MCG signal in the memory 513 is transferred to the signal processing unit 70 during exposure.

[0314] Next, when the incident light is of medium illumination and the level of the HCG signal exceeds the reference signal 2 (the saturation level of the HCG signal), the Fig.23 Here, two signals, the MCG signal and the LCG signal, are selected and synthesized in the WDR synthesizing circuit 76 of the signal processing unit 70. The HCG signal is not necessary.

[0315] From time 102 to time 103, since the illumination is medium, the charge of the photodiode PD exceeds the specified threshold of the overflow element OF during exposure. Therefore, the charge is accumulated in the storage capacitor C2 but does not reach the saturation level, and the detection circuit 52 does not operate. First, after the shutter (SHT) time, the gain control transistors GC1 and TGC are turned on, and the reset component of the LCG signal is sampled and held in the capacitor element CR of the sample and hold circuit 30 through the sample and hold signal SHR. However, since the storage capacitor C2 does not exceed the saturation level during exposure, the detection circuit 52 does not operate, so the sampled and held charge is discharged.

[0316] From time 103 to time 104, the gain control transistor GC1 is turned on, and the reset component of the MCG signal is sampled and held in the capacitor element CR of the sample and hold circuit 30 through the sample and hold signal SHR.

[0317] From time 104 to time 106 , the reset component and the signal component of the HCG signal are AD-converted and stored in the memory 513 .

[0318] exist Fig.23 In the description of , the following is assumed: from time 106 to time 107, by Fig.11 In the selective readout operation, the level of the HCG signal exceeds the saturation level (signal potential FDH). In addition, for the case where it does not exceed Fig. 22 As explained.

[0319] From time 107 to time 108, the gain control transistor GC1 is turned on, and the signal component of the MCG signal is sampled and held in the capacitor element CS of the sample and hold circuit 30 through the sample and hold signal SHS.

[0320] From time 108 to time 110, the reset component and signal component of the sampled and held MCG signal are AD-converted and stored in the memory 513. At this time, from time 106 to time 107, the data of the HCG signal in the memory 513 is cleared and becomes free.

[0321] From time 110 to time 112, the reset component and the signal component of the LCG signal are sampled and held in the capacitor elements CR and CS of the sample and hold circuit 30 by the sample and hold signals SHR and SHS.

[0322] From time 115 to time 116, the data of the MCG signal in the memory 513 is transferred to the signal processing unit 70 in the next exposure.

[0323] From time 116 to time 118, the reset component of the LCG signal sampled and held by the sample and hold signal SHR in the capacitor element CR of the sample and hold circuit 30 and the LCG signal component sampled and held by the sample and hold signal SHS in the capacitor element CS of the sample and hold circuit 30 are AD converted and accumulated in the memory 513 that was idle from time 115 to time 116.

[0324] From time 118 to time 119, the data of the LCG signal in the memory 513 is transferred to the signal processing unit 70 during exposure.

[0325] Next, when the incident light is of high illumination, and the number of initialization times N is greater than 1, use Fig.24At this time, two signals, the saturation signal Qsat of the LCG signal and the residual signal of the LCG signal, are selected and synthesized in the WDR synthesizing circuit 76 of the signal processing unit 70. The HCG signal and the MCG signal are not necessary.

[0326] From time 102 to time 103, due to high illumination, the charge of the photodiode PD exceeds the predetermined threshold of the overflow element OF during exposure, so that the charge is accumulated in the storage capacitor C2, and the storage capacitor C2 exceeds the saturation level, and the detection circuit 52 operates. Fig.24 This is an example of the case where the number of initialization times N is 2. First, after the shutter (SHT) time, the gain control transistors GC1 and TGC are turned on, and the reset component of the LCG signal is sampled and held in the capacitor element CR of the sample and hold circuit 30 by the sample and hold signal SHR. Then, the storage capacitor C2 exceeds the saturation level, and the detection circuit 52 operates. Then, the charges of the photodiode PD and the storage capacitor C2 are mixed by FD, and the signal component of the saturation level of the LCG signal (saturation signal Qsat) is sampled and held in the capacitor element CS of the sample and hold circuit 30 by the sample and hold signal SHS.

[0327] After initialization, the gain control transistors GC1 and TGC are turned on, and the reset component of the LCG signal is sampled and held by the sample-hold signal SHR in the capacitor element CR of the sample-hold circuit 30. Then, the storage capacitor C2 exceeds the saturation level, and the detection circuit 52 operates. Then, the charges of the photodiode PD and the storage capacitor C2 are mixed by FD, and the signal component of the saturation level of the LCG signal (saturation signal Qsat) is sampled and held by the sample-hold signal SHS in the capacitor element CS of the sample-hold circuit 30. This operation is repeated N times.

[0328] Here, the sample-and-hold circuit 30 does not need to repeatedly accumulate the saturation signal Qsat every time initialization occurs, and may accumulate the saturation signal Qsat only once and continue to hold the saturation signal Qsat accumulated only once during exposure.

[0329] From time 103 to time 104 , the reset component of the MCG signal is not sampled and held by the sample and hold circuit 30 . This is because charge is accumulated in the sample and hold circuit 30 from time 102 to time 103 .

[0330] From time 104 to time 106, the reset component and the signal component of the HCG signal are not AD-converted and are not stored in the memory 513. This is because the HCG signal is not necessary under high illuminance.

[0331] In the period from time 106 to time 107, whether or not Fig.11In the selective readout operation and when the HCG signal level exceeds the saturation level, the processing from time 107 to time 108 is performed.

[0332] From time 107 to time 108, the signal component of the MCG signal is not sampled and held by the sample-and-hold circuit 30. This is because the MCG signal is not required under high illumination.

[0333] From time 108 to time 110, the saturation signal Qsat of the LCG signal (specifically, the reset component and the signal component of the saturation signal Qsat) that is sampled and held is AD-converted and stored in the memory 513. Here, the saturation signal Qsat of the LCG signal may use a signal when the number of initialization times N is 1, or may use a signal when the number of initialization times N is 2.

[0334] From time 110 to time 112 , the charges of the photodiode PD and the storage capacitor C2 are FD mixed, and the reset component and signal component of the residual signal of the LCG signal are sampled and held in the capacitor elements CR and CS of the sample and hold circuit 30 by the sample and hold signals SHR and SHS.

[0335] From time 115 to time 116, the data of the saturation signal Qsat of the LCG signal accumulated in the memory 513 is transferred to the signal processing unit 70 in the next exposure.

[0336] From time 116 to time 118 , the residual signal of the sampled and held LCG signal (specifically, the reset component and the signal component of the residual signal) is AD-converted and stored in the memory 513 that was idle from time 115 to time 116 .

[0337] From time 118 to time 119, the data of the residual signal of the LCG signal in the memory 513 is transferred to the signal processing unit 70 during exposure.

[0338] [Subject illumination dependence of SN]

[0339] Next, use Fig.25 The subject illumination dependency of SN is described.

[0340] Fig.25 The diagram shows the saturation signal Qsat (also referred to as linear saturation (Qsat)) of the HCG signal, the MCG signal, and the LCG signal, and the SN of the residual signal of the LCG signal. Fig.25 The SN of the HCG signal at low illumination, the MCG signal at medium illumination, and the LCG signal at high illumination is shown, which is relative to Figure 7 (d) is a diagram after adding the SN of the first signal (linear saturation of the LCG signal (Qsat)×N+residual signal).

[0341] The SN at low illumination is good because the HCG signal is directly AD-converted by the AD conversion circuit 51. In addition, regarding the maximum subject illumination, as Fig.25 As shown, in the conventional method without PWM, it is "linear saturation (Qsat)", while in the method with PWM of the present disclosure, it becomes "linear saturation (Qsat)×N+residual signal", which can greatly expand the dynamic range.

[0342] Here, in detail, the SN of "linear saturation (Qsat) × N + residual signal" maintains the SN of the linear saturation (Qsat) of the LCG signal. This is because the SN of the linear saturation (Qsat) of the LCG signal is determined by √(linear saturation (Qsat)) of the shot noise, and it does not change only by digitally initializing both the signal S (Signal) and the noise N (Noise) N times by the signal processing unit 70.

[0343] Here, regarding the SN of "linear saturation (Qsat) × N + residual signal", as described above, the SN of "linear saturation (Qsat) × N" is equal to the SN of "linear saturation (Qsat)", and is equivalently represented by the SN of "linear saturation (Qsat) + residual signal". The signal S (Signal) is "S = (Qsat) + residual signal", and the noise N (Noise) is represented by the mean square of each shot noise, so it becomes "N = √ ((√Qsat) ^2 + (√ residual signal) ^2) = √ (Qsat + residual signal)". Therefore, it becomes "SN = √ (Qsat + residual signal)".

[0344] According to the present disclosure, the dynamic range can be expanded even when the storage capacitor C2 is not a high-density storage capacitor C2 but a small-capacitance storage capacitor C2.

[0345] Furthermore, it can be seen that the SN step difference at the boundary of the transition region of the HCG signal, the MCG signal, and the LCG signal is not large, and an ideal characteristic can be achieved.

[0346] In addition, when the PWM method is used for the photodiode PD disclosed in Patent Document 1, the maximum SN is determined by the shot noise of the MCG signal of the photodiode PD when it is saturated and becomes a constant value. This is because the SN of the photodiode PD does not change even if it is digitally multiplied by N. However, according to the present disclosure, the maximum SN is determined by the shot noise of the LCG signal of the photodiode PD and the storage capacitor C2 when it is saturated, so the SN can be improved compared with the conventional method.

[0347] The saturation signal (Qsat) of the LCG signal may be calculated each time the sample-and-hold circuit 30 is initialized, or may be calculated by averaging multiple times. This improves the accuracy of the saturation signal (Qsat) of the LCG signal.

[0348] In addition, it is preferred that the size of the amplifier transistor SF1 is larger than that of the amplifier transistor SF2. By making the size of the amplifier transistor SF1 for reading out the signal larger, noise can be reduced. In addition, by making the size of the amplifier transistor SF2 for reading out the detection signal smaller, the reduction in the area of ​​the photodiode PD can be suppressed.

[0349] [Photographic device using a solid-state imaging device]

[0350] Next, regarding an imaging device using the solid-state imaging device 100, Fig.26 Provide explanation.

[0351] Fig.26 FIG. 2 is a diagram showing an example of an imaging device 200 to which the solid-state imaging device 100 according to the embodiment is applied.

[0352] The imaging device 200 of the figure is a camera system, and includes the solid-state imaging device 100, an imaging optical system 202 including a lens, a signal processing unit 203, a driving circuit 204, and a system control unit 205. In the imaging device 200, the solid-state imaging device 100 is used.

[0353] The drive circuit 204 receives a control signal corresponding to the drive mode from the system control unit 205 and supplies the drive mode signal to the solid-state imaging device 100. In the solid-state imaging device 100 supplied with the drive mode signal, a drive pulse corresponding to the drive mode signal is generated and supplied to each clock in the solid-state imaging device 100.

[0354] The signal processing unit 203 receives the image signal output from the solid-state imaging device 100 and performs various signal processing on the image signal.

[0355] Thus, the imaging device 200 includes the solid-state imaging device 100 , the imaging optical system 202 that guides incident light from a subject to the solid-state imaging device 100 , and the signal processing unit 203 that processes an output signal from the solid-state imaging device 100 .

[0356] [Distance-measuring imaging device using a solid-state imaging device]

[0357] Next, use Fig. 27 A distance measuring imaging device to which the solid-state imaging device 100 is applied will be described.

[0358] Fig. 271 is a diagram showing an example of a distance measuring imaging device 300 to which the solid-state imaging device 100 according to the embodiment is applied. The object 190 is a target for distance measurement.

[0359] The distance measuring imaging device 300 includes a light source driver 150, a light source unit 160, an optical lens 170, a signal processing unit 180, and a solid-state imaging device 100. In the distance measuring imaging device 300, the solid-state imaging device 100 is used.

[0360] The light source driver 150 supplies a driving signal to the light source unit 160 in accordance with a signal instructing the solid-state imaging device 100 to emit light.

[0361] The light source unit 160 generates pulse light for distance measurement according to the driving signal of the light source driver 150 .

[0362] The optical lens 170 is a lens for focusing the reflected pulse light of the object 190 corresponding to the pulse light from the light source unit 160 .

[0363] The signal processing unit 180 obtains the distance to the object 190 through calculation based on the signal received from the solid-state imaging device 100 .

[0364] The solid-state imaging device 100 irradiates the object 190 with background light and near-infrared light from the light source unit 160. The reflected light from the object 190 is incident on the pixel array unit 1 via the optical lens 170. The reflected light incident on the pixel array unit forms an image, and the imaged optical image is converted into a pixel signal. The output of the solid-state imaging device 100 is converted into distance data by the signal processing unit 180, and is also converted into a visible distance image or a brightness image according to the application.

[0365] (Other embodiments)

[0366] The solid-state imaging device 100 of one or more technical solutions of the present disclosure has been described above based on the embodiments, but the present disclosure is not limited to the embodiments. As long as it does not deviate from the main purpose of the present disclosure, various modifications that can be thought of by those skilled in the art to each embodiment or technical solutions constructed by combining constituent elements in different embodiments are also included in the scope of one or more technical solutions of the present disclosure.

[0367] Industrial Applicability

[0368] The present disclosure relates to a solid-state imaging device and a photographing device and a distance-measuring imaging device using the solid-state imaging device as an imaging element, and is suitable for, for example, a digital camera and a distance-measuring system.

[0369] Description of Reference Numerals

[0370] 1Pixel array section

[0371] 3, 3a, 3b, 3c pixel circuit

[0372] 3D pixel circuit group

[0373] 4.4D logic circuit

[0374] 14 Vertical scanning circuit

[0375] 15 horizontal scan line groups

[0376] 19 Input and output signal lines

[0377] 20 Timing control unit

[0378] 27 Reference signal generation unit

[0379] 28 Output Circuit

[0380] 30 Sample and hold circuit

[0381] 51 AD conversion circuit

[0382] 52 Detection circuit

[0383] 61, 62 local pixel control circuit

[0384] 70, 180, 203 Signal Processing Department

[0385] 71 HCG signal generation circuit

[0386] 72 MCG signal generation circuit

[0387] 73 LCG signal generation circuit

[0388] 74 Interpolation Circuit

[0389] 76 WDR synthesis circuit

[0390] 77, 513, 515, 523, 525 memory

[0391] 80 Count control signal generation unit

[0392] 100 Solid-state imaging device

[0393] 150 Light source driver

[0394] 160 Light source

[0395] 170 Optical lens

[0396] 190 Objects

[0397] 200 Photographic Installation

[0398] 202 Camera Optical System

[0399] 204 drive circuit

[0400] 205 System Control Department

[0401] 271, 272 reference signal generating unit

[0402] 300 rangefinder camera

[0403] 511, 521 comparator

[0404] 512, 522 counter circuit

[0405] 524 Initialization Circuit

[0406] C0, C1, C2 storage capacitors

[0407] CR, CS capacitor components

[0408] FD0, FD1, FD2 charge storage unit

[0409] GC1, TGC gain control transistor

[0410] OF overflow element

[0411] PD Photodiode

[0412] RS reset transistor

[0413] TG transfer transistor

[0414] SEL1, SEL2 select transistor

[0415] SF1, SF2 amplifier transistor

[0416] SW1, SW2 selection circuit

Claims

1. A solid-state imaging device, characterized in that: have: A pixel circuit outputs a plurality of pixel signals; Detection circuit; A pixel control circuit, which controls the pixel circuit; and Signal processing unit; The pixel circuit has: Photodiode; A first transfer transistor reads the signal of the photodiode to a floating diffusion portion; and The storage capacitor stores the charge overflowing from the photodiode. The detection circuit compares the signal of the storage capacitor with a first reference value. When the signal of the storage capacitor reaches the first reference value, the photodiode and the storage capacitor are initialized via the pixel control circuit, and the number of initializations is counted. The signal processing unit calculates a first signal indicating the intensity of incident light based on the number of initialization times and a mixed signal of a signal of the photodiode in the floating diffusion unit and a signal of the storage capacitor read out by the first transfer transistor.

2. The solid-state imaging device according to claim 1, wherein The solid-state imaging device includes: a sample and hold circuit; and AD conversion circuit.

3. The solid-state imaging device according to claim 2, wherein: At least one pixel signal among the plurality of pixel signals is not held by the sampling and holding circuit, but is directly AD-converted by the AD conversion circuit. At least one other pixel signal among the plurality of pixel signals is held by the sample-and-hold circuit, and the held signal is AD-converted by the AD conversion circuit.

4. The solid-state imaging device according to claim 2 or 3, wherein: When the number of initializations is 1 or more, The pixel circuit outputs a mixed signal of the photodiode and the storage capacitor as a second signal when the signal of the storage capacitor reaches the first reference value during exposure, and the photodiode and the storage capacitor are initialized. The second signal is held by the sample-and-hold circuit, and the held signal is AD-converted by the AD conversion circuit. The pixel circuit outputs, as a third signal, a mixed signal remaining in the photodiode and the storage capacitor because the signal of the storage capacitor does not reach the first reference value when exposure is completed. The third signal is not held by the sample-and-hold circuit but is directly AD-converted by the AD conversion circuit.

5. The solid-state imaging device according to claim 4, wherein: The signal processing unit calculates the first signal by calculating a sum of a product of the second signal and the number of initialization times and the third signal.

6. The solid-state imaging device according to claim 4 or 5, wherein: The sample-and-hold circuit includes a first sample-and-hold capacitor element for a reset component and a second sample-and-hold capacitor element for a signal component. When the number of initializations is 1 or more, During exposure, after the photodiode and the storage capacitor are initialized, a reset component is held in the first sampling and holding capacitor element. During exposure, when the signal of the storage capacitor reaches the first reference value, the second signal is held as a signal component in the second sampling and holding capacitor element. The AD conversion circuit performs AD conversion on the pixel signal level by comparing the pixel signal level obtained by subtracting the reset component from the signal component with a second reference value.

7. The solid-state imaging device according to any one of claims 2 to 6, wherein: When the above initialization times is 0, The pixel circuit outputs the fourth signal in a low illumination mode for low illumination, and outputs the fifth signal in a mode for high illumination higher than the low illumination mode, The fourth signal is not held by the sampling and holding circuit, but is directly AD-converted by the AD conversion circuit. The fifth signal is held by the sample-and-hold circuit, and the held signal is AD-converted by the AD conversion circuit.

8. The solid-state imaging device according to claim 7, wherein: The above pixel circuit comprises: a second transfer transistor for transferring the charge in the storage capacitor to the floating diffusion; a first amplifying transistor for outputting a signal from the photodiode and a signal from the storage capacitor; A first selection transistor connected to the first amplifying transistor; A second amplifying transistor connected to the storage capacitor; and A second selection transistor is connected to the second amplification transistor. The detection circuit is connected to the storage capacitor via the second selection transistor and the second amplifier transistor. The pixel circuit outputs the signal of the photodiode as the fourth signal from the first amplifier transistor via the first transfer transistor in the low illumination mode. The high illumination mode includes a high illumination mode in which the pixel circuit outputs the signal of the storage capacitor as a sixth signal from the first amplifier transistor via the second transfer transistor. The sixth signal mentioned above belongs to the fifth signal mentioned above.

9. The solid-state imaging device according to claim 8, wherein: The pixel circuit includes a reset transistor. The detection circuit controls the reset transistor, the first transfer transistor and the second transfer transistor via the pixel control circuit based on the comparison result of the signal of the storage capacitor and the first reference value, thereby discharging the charges of the photodiode and the storage capacitor and initializing the photodiode and the storage capacitor.

10. The solid-state imaging device according to claim 8 or 9, wherein: The pixel circuit includes a gain switching transistor for switching the gain of the signal of the photodiode. The high illumination mode includes a medium illumination mode in which the pixel circuit outputs the signal of the photodiode as the seventh signal from the first amplifier transistor by controlling the gain switching transistor. The seventh signal mentioned above belongs to the fifth signal mentioned above.

11. The solid-state imaging device according to any one of claims 7 to 10 as cited in claim 6, characterized in that: The second reference value is a saturation level of the fourth signal.

12. The solid-state imaging device according to any one of claims 7 to 11, wherein: The signal processing unit includes a WDR processing unit. The WDR processing unit mentioned above, When the number of initializations is greater than 1, an output signal is generated from at least the first signal. When the number of initialization times is 0, an output signal is generated from at least one of the fourth signal and the fifth signal.

13. The solid-state imaging device according to any one of claims 2 to 12, wherein: The AD conversion circuit has a comparator. When the number of initializations is 0, The above pixel circuit outputs more than three pixel signals. The comparator compares one pixel signal among the three or more pixel signals with a second reference value after AD conversion, and generates a selection signal indicating selection of at least two pixel signals among the three or more pixel signals when the one pixel signal reaches the second reference value. The pixel control circuit selects the at least two pixel signals from the three or more pixel signals according to the selection signal.

14. The solid-state imaging device according to any one of claims 2 to 13, wherein: The AD conversion circuit and the detection circuit are provided for each single pixel circuit. The pixel control circuit controls each of the pixel circuits individually.

15. The solid-state imaging device according to any one of claims 2 to 13, wherein: The AD conversion circuit and the detection circuit are provided for each of the plurality of pixel circuits. The pixel control circuit controls each of the pixel circuits individually. The detection circuit compares the signal of the storage capacitor of the pixel circuit with the first reference value while synchronizing the corresponding pixel circuit with the pixel control circuit.

16. The solid-state imaging device according to claim 15 as cited in claim 8, characterized in that: The pixel control circuit synchronizes control of the first selection transistor, the second selection transistor, and the sample-and-hold circuit of the corresponding pixel circuit.

17. The solid-state imaging device according to claim 15 or 16, wherein: A plurality of the above pixel circuits correspond to the arranged pixels of the color filter, The pixel control circuit performs the same pixel control on the plurality of pixel circuits in accordance with the maximum signal level of the arrayed pixels of the color filter.

18. The solid-state imaging device according to any one of claims 2 to 17, wherein: The solid-state imaging device includes: A first semiconductor substrate having the above-mentioned pixel circuit; and The second semiconductor substrate has the detection circuit and the AD conversion circuit. The first semiconductor substrate and the second semiconductor substrate are stacked.

19. The solid-state imaging device according to any one of claims 1 to 18, wherein: The first reference value is a saturation level of the storage capacitor.

20. A solid-state imaging device, characterized in that: have: A pixel circuit including a photodiode, a floating diffusion portion from which a signal of the photodiode is read, a storage capacitor for storing charges overflowing from the photodiode, and a reset unit for initializing the photodiode, the floating diffusion portion, and the storage capacitor; as well as The detection circuit includes a comparator for comparing the signal of the storage capacitor with a first reference value, and an initialization circuit for controlling the reset unit according to the comparison result of the comparator. The pixel circuit is formed on a first semiconductor substrate composed of one or more semiconductor layers, and the detection circuit is formed on a second semiconductor substrate composed of one or more semiconductor layers. The first semiconductor substrate and the second semiconductor substrate are stacked.

21. The solid-state imaging device according to claim 20, wherein: The pixel circuit has: a first transfer transistor for reading a signal from the photodiode to the floating diffusion; and The second transfer transistor transfers the signal of the storage capacitor to the floating diffusion.

22. The solid-state imaging device according to claim 20 or 21, wherein: The detection circuit has a counter function for counting the comparison result of the comparator.

23. The solid-state imaging device according to any one of claims 20 to 22, wherein: It also has an AD conversion function formed on the second semiconductor substrate and converts the signal of the floating diffusion into a digital value, The comparator and the input terminals of the AD conversion function are respectively connected to different output terminals of the pixel circuit.

24. The solid-state imaging device according to any one of claims 20 to 22, wherein: Also available: An AD conversion function is formed on the second semiconductor substrate and converts the signal of the floating diffusion into a digital value; and The selection circuit is formed on the first semiconductor substrate or the second semiconductor substrate, receives the signal from the storage capacitor and the signal from the floating diffusion, and selects the output. The output of the selection circuit is input to the AD conversion function. The detection circuit and the AD conversion function share the comparator.