JP5570628B2 - Solid-state imaging device - Google Patents

Description

  The present invention relates to a solid-state imaging device, and more specifically to a dynamic range expansion technique.

  In an amplification type imaging apparatus known by a name such as a CMOS image sensor, there are various electronic shutter technologies capable of controlling the start and end of accumulation simultaneously in the plane without using a mechanical light shielding means.

  In the configuration having the electronic shutter described in Patent Document 1, the charge generated by the photoelectric conversion unit is transferred to the holding unit during the signal charge generation period, and after the exposure is completed, the photoelectric conversion unit is reset and generated by the photoelectric conversion unit. The electronic shutter function is realized by discharging the generated charges.

  The feature of the present technology is that the photoelectric conversion unit is basically specialized in photoelectric conversion, and the function of charge accumulation during exposure is separated by providing the adjacent charge holding unit with the function. This charge holding portion is provided separately from the FD region. Since the number of saturated charges in the photoelectric conversion unit is small, transfer from the photoelectric conversion unit to the charge holding unit can be performed at a low voltage. Compared with a CCD or the like, a simple CMOS manufacturing process or a derivative manufacturing process thereof can be simplified. Can be manufactured in a process.

JP 2006-246450 A

  In Patent Document 1, it is possible to increase the amount of saturated charge by providing a charge holding unit in a pixel. However, since the amount of saturated charge increases, the pixel is provided at the subsequent stage of the photoelectric conversion unit and the charge holding unit. The signal range may be limited by the dynamic range of the readout circuit. As described above, this is a prominent problem because the configuration having the photoelectric conversion function and the configuration for accumulating signal charges are provided as separate configurations, and the amount of charge that can be stored as an appropriate configuration has increased. .

  The present invention has been made in view of the above problems, and is intended to effectively use a signal charge increased by a separately provided charge holding unit as an image signal without being limited by the dynamic range of a readout circuit in the subsequent stage. It is.

  In view of the above problems, the present invention provides a photoelectric conversion unit, a charge holding unit that holds signal charges generated in the photoelectric conversion unit, and a first transfer unit that transfers signal charges from the photoelectric conversion unit to the charge holding unit. A second transfer unit that transfers a signal charge held by the charge holding unit, an amplification unit that amplifies a signal based on the signal charge transferred by the second transfer unit, and the amplification A reset unit that resets the signal charge transferred to the unit, wherein the charge holding unit is arranged in an electrical path between the photoelectric conversion unit and the amplification unit, and the photoelectric conversion With the signal charge held in the unit, a conduction pulse is supplied to the second transfer unit to transfer the signal charge held in the charge holding unit to the amplification unit, and a conduction pulse is supplied to the reset unit. Signal charge transferred to the amplifier After resetting, characterized by transferring the signal charges held in the photoelectric conversion unit by supplying the turning-on pulse to the first and second transfer unit to the amplifying unit.

  According to the present invention, in the configuration in which the charge holding unit is provided in the pixel separately from the photoelectric conversion unit, the signal charge increased by the charge holding unit provided separately is effective without being limited by the dynamic range of the readout circuit in the subsequent stage. It can be used as an image signal.

It is a block diagram of the solid-state imaging device of the present invention. It is an equivalent circuit diagram of the pixel region of the solid-state imaging device of the present invention. It is a drive pulse diagram of a first embodiment. It is a potential diagram of a first embodiment. It is a drive pulse diagram of a second embodiment. It is a drive pulse diagram of a third embodiment. It is a figure which shows distribution of the pixel signal of 3rd embodiment. It is a conceptual diagram which shows the signal processing of 4th embodiment. It is a conceptual diagram which shows the ratio of the noise with respect to the signal of 4th embodiment. It is a drive pulse figure with respect to the signal of 5th embodiment. It is a potential diagram of a fifth embodiment. It is a conceptual diagram of the imaging device of 6th embodiment. It is a conceptual diagram of the imaging device of 7th embodiment. It is a conceptual diagram which shows the relationship between the surface illumination intensity and output of 7th embodiment.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  First, a block diagram of a solid-state imaging device and an equivalent circuit diagram of a pixel that can be commonly applied to the following embodiments will be described with reference to FIGS.

  In FIG. 1, reference numeral 101 denotes a pixel region. A plurality of pixels are arranged in a matrix. Reference numeral 102 denotes a vertical scanning unit. The pixels in the pixel region are scanned every pixel row or every plurality of pixel rows. A shift register or a decoder can be used.

  Reference numeral 103 denotes a column circuit. A desired process is performed on a signal read from the pixel region 101 by scanning of the vertical scanning unit. For example, it may be configured to include a CDS circuit that suppresses pixel noise, an amplification unit that amplifies a signal from the pixel, an AD conversion unit that digitally converts an analog signal from the pixel, and the like.

  Reference numeral 104 denotes a horizontal scanning unit. In order to read out a signal subjected to a desired process in the column circuit to the outside, scanning is sequentially performed for each pixel column or for each of a plurality of pixel columns. Similar to the vertical scanning unit, it is configured using a shift register or a decoder.

  Reference numeral 105 denotes a signal processing unit. Predetermined processing is performed on the signal output from the solid-state imaging device.

  Wiring for transmitting an optical signal, a drive signal, or the like exists between the components, but is omitted here.

  FIG. 2 is an equivalent circuit diagram of pixels arranged in the pixel region. For simplification of description, the pixels included in the pixel region 101 are 9 pixels of 3 rows × 3 columns, but the present invention is not limited to this. Reference numeral 2 denotes a photodiode (PD) that functions as a photoelectric conversion unit. The anode of the PD is connected to a fixed potential (for example, ground), and the cathode is connected to one terminal of the charge holding unit 3 via the first transfer transistor 8 functioning as the first transfer unit. The other terminal of the charge holding unit 3 is connected to a fixed potential (for example, ground). One terminal of the charge holding unit 3 is further connected to the FD region 4 via a second transfer transistor 9 which is a second transfer unit. The FD region is connected to the gate terminal of the amplification transistor 12 that functions as a part of the amplification unit. The gate of the amplification transistor functions as an input unit of the amplification unit. The gate of the amplification transistor 12 is connected to the pixel power supply line via the reset transistor 10 that functions as a reset unit. As each transfer transistor, a MOS transistor can be used.

  In the selection transistor 11 that functions as a selection unit, a drain terminal that is one main electrode is connected to the pixel power supply line, and a source terminal that is the other main electrode is connected to a drain that is one main electrode of the amplification transistor 12. When an active signal SEL is input, both main electrodes of the selection transistor are turned on. As a result, the amplification transistor 12 forms a source follower circuit with a constant current source (not shown) provided on the vertical signal line OUT, and a signal corresponding to the potential of the gate terminal which is the control electrode of the amplification transistor 12 is transmitted to the vertical signal line OUT. Appear in. A signal is output from the solid-state imaging device based on a signal appearing on the vertical signal line OUT, and an image signal is formed through a signal processing circuit unit or the like.

  In FIG. 2, a reset unit, an amplification unit, and a selection unit are provided for each pixel. However, a plurality of pixels can share these configurations. It is also possible to adopt a configuration in which a pixel is selected by the potential of the input portion of the amplifying unit without providing the selecting unit.

  The present invention is applied to a configuration having a charge holding portion between a photoelectric conversion portion and an FD region, taking the configuration as described above as an example.

  Particularly preferably, the photoelectric conversion is performed in a configuration having the following characteristics in the structure of the charge path between the photoelectric conversion unit and the charge holding unit, that is, in a state where a low-level pulse is supplied that makes the first transfer unit non-conductive. It is desirable to use a structure capable of transferring charge from the portion to the charge holding portion.

  For example, as a specific configuration, if the first transfer unit is a MOS transistor, the MOS transistor has a buried channel structure. And even if it is a non-conduction state, it is the structure where the potential barrier has a part where only the part is lowered in a part deeper than the surface. In this case, the charge transfer unit can be in a state in which a constant voltage is supplied without performing active control during accumulation of signal charges. That is, a fixed potential barrier may be provided without having a function as a transfer unit. Then, immediately before the end of accumulation, control is performed to lower the height of the potential barrier, and the signal charge remaining in the photoelectric conversion unit is transferred to the charge holding unit.

  According to such a configuration, almost all signal charges generated by photoelectric conversion when light enters the photoelectric conversion unit can be transferred to the charge holding unit without being accumulated in the photoelectric conversion unit. Therefore, it is possible to align the charge accumulation time in the photoelectric conversion units included in all the pixels. In addition, when the MOS transistor is non-conductive, holes are accumulated on the channel surface, and the channel to which charges are transferred exists at a predetermined depth from the surface. It becomes possible to reduce.

  From another viewpoint, in the period in which the signal charge is accumulated in the photoelectric conversion unit and the charge holding unit, the potential barrier of the charge path between the photoelectric conversion unit and the charge holding unit is between the photoelectric conversion unit and other regions. It can be said that it is lower than the potential barrier of the charge path between them. The potential here is a potential with respect to a signal charge. For example, when the OFD region is provided, the potential may be lower than that between the OFD region.

  In addition, it is particularly preferable that a signal is applied to the surface of the charge holding portion by applying a potential to the counter electrode through the insulating film during the period in which the charge holding portion is formed of a charge coupled device and the signal charge is accumulated in the charge holding portion. Accumulate charges of opposite polarity to the charge. Thus, it is desirable to suppress the generation of dark charges on the semiconductor surface where the charge holding portion is formed.

  According to such a configuration, it is possible to further reduce the dark current caused by the charge holding unit. In addition, since there is no need to perform impurity implantation of the opposite conductivity type as a countermeasure against dark current on the surface of the charge holding portion, the portion in charge of charge holding can be formed shallower near the surface than the photodiode. Therefore, the charge holding capacity per unit volume can be increased, and it becomes possible to have a holding capacity several times that of a configuration in which a conventional photodiode also serves as charge holding.

  Further, from the viewpoint of driving, the signal charge moved from the photoelectric conversion unit to the charge holding unit during one signal charge generation period is held in the charge holding unit and used as an image signal. That is, it can be said that after starting the one-signal charge generation period in the photoelectric conversion unit, a signal is read out of the pixel without going through the reset operation of the charge holding unit. Note that one signal charge generation period is determined in common by each photoelectric conversion unit when an image of one frame is taken.

  Hereinafter, the specific configuration and driving method of the present invention will be described. An example of the pixel configuration at that time is based on a configuration in which the first transfer unit is a buried channel type MOS transistor and the charge holding unit is formed of a charge coupled device. Further, a configuration using transistors as the first and second transfer units will be described as an example.

(First embodiment)
FIG. 3 shows the drive pulse of the first embodiment, and FIG. 4 shows the potential state in each component. In this embodiment, when transferring the signal charge generated during one signal charge generation period to the amplification unit, first, a conduction pulse is supplied only to the second transfer unit, and the signal charge held in the charge holding unit is changed. Transfer to the amplifier. Then, a conduction pulse is supplied to the reset unit to reset the signal charge transferred to the amplification unit. After that, by supplying a conduction pulse to the first and second transfer units, the signal charge held in the photoelectric conversion unit is transferred to the amplification unit. The vertical scanning unit 102 supplies a drive pulse to each element in order to perform these operations. Therefore, the vertical scanning unit can be referred to as a control unit that supplies each driving pulse. Alternatively, a control unit including a timing generator for controlling the vertical scanning unit 102 may be used.

  In FIG. 3, PTX1 is a drive pulse supplied to the first transfer transistor, PTX2 is a drive pulse supplied to the second transfer transistor, PRES is a drive pulse supplied to the reset transistor, and PSEL is a drive pulse supplied to the selection transistor. Respectively. The numbers shown in parentheses indicate pixel row numbers. The imaging apparatus of the present embodiment has a mechanical shutter. Light is incident on the photoelectric conversion unit when the mechanical shutter is open, and light is not incident on the photoelectric conversion unit when the mechanical shutter is closed. The exposure state of the photoelectric conversion unit can be controlled by the mechanical shutter, and the signal charge generation period can be defined. In FIG. 3, the black-painted portion is in the closed state of the mechanical shutter, and the white-painted portion is in the open state.

  PTS is a sampling pulse for capturing a signal in a holding unit for holding an optical signal included in the column circuit, and PTN is a sampling pulse for capturing a signal in a holding unit for holding a noise signal included in the column circuit. The noise signal is an offset of the reset transistor and amplification transistor of the pixel, random noise, or an offset of the column amplification unit when the column circuit includes an amplification unit.

  It is assumed that each transistor conducts or samples by a high level pulse.

  In this embodiment, after transferring the signal charge held in the charge holding unit to the amplification unit, the signal charge transferred to the amplification unit is reset, and then the charge held in the photoelectric conversion unit is changed to the charge holding unit. It is characterized by transferring to the FD area via

  First, at T1, a high level pulse is supplied to the reset transistor and the first to third transfer transistors to be in a conductive state, and the charges in the photoelectric conversion unit, the charge holding unit, and the FD region are reset. At this time, the mechanical shutter is closed.

  At T2, the mechanical shutter is opened and light is incident on the photoelectric conversion unit. At this time, the pulse supplied to each transfer transistor is a low-level pulse that makes each transfer transistor non-conductive.

  At T3, the mechanical shutter is closed.

  At T4, a low level pulse is supplied to PRES in the first pixel row, and a high level pulse is supplied to PSEL. Here, they are performed simultaneously, but may be performed at different times. However, in order to suppress kTC noise in the reset unit, it is necessary that a low level pulse is supplied to PRES at least during a period in which a high level pulse is supplied to a PTN that samples a noise signal.

  At T5, a high level pulse is supplied to PTN, and then a low level pulse is supplied after a predetermined time has elapsed, whereby the noise signal of the pixels in the first row is held in the column circuit.

  At T6a, a high level pulse is supplied to the second transfer transistors in the first pixel row. Thereby, only the charge of the charge holding unit and the charge exceeding the potential barrier between the photoelectric conversion unit and the charge holding unit are transferred to the amplifying unit (first step).

  At T6b, a low-level pulse that is turned off is supplied to the second transfer transistor in the first pixel row.

  At T7, a high level pulse is supplied to the PTS, and a signal based on the signal charge read out in the first step is held in the column circuit.

  At T8, a low level pulse is supplied to PSEL in the first pixel row, and a high level pulse is supplied to PRES. By this operation, the charge transferred to the amplification unit in the first step is reset.

  At T9, in order to read out the same row again, a low level pulse is supplied to PRES of the first pixel row, and a high level pulse is supplied to PSEL.

  At T10, a high level pulse is supplied to PTN, and then a low level pulse is supplied after a predetermined time has elapsed, whereby the noise signal of the pixels in the first row is held in the column circuit.

  In T11a, a high-level pulse that is turned on is supplied to the first transfer transistor and the second transfer transistor in the first pixel row (second step). As a result, the signal charge accumulated in the photoelectric conversion unit is transferred to the FD region via the charge holding unit.

  At T11b, a low-level pulse that is turned off is supplied to the second transfer transistor in the first pixel row.

  At T12, a high-level pulse is supplied to the PTS, and then a low-level pulse is supplied after a predetermined time has elapsed, whereby a signal based on the charge transferred in the second step is held in the column circuit.

  At T13, a low level pulse is supplied to PSEL in the first row, and a high level pulse is supplied to PRES.

  Thereafter, the signal processing unit 105 adds the signals obtained by the first and second steps described above. As a result, most of the charge generated by the photoelectric conversion in the photoelectric conversion unit can be handled as an image forming signal regardless of the dynamic range of the readout circuit arranged downstream of the photoelectric conversion unit and the charge holding unit. Become.

  Then, by repeating the operations from T4 to T13 for each pixel row, it is possible to read out a signal of one frame.

  Here, the first step and the second step are repeated for each row. However, for example, the driving may be performed so that the second step is performed after the first step is performed for all rows. In other words, the driving may be performed so that the data is transferred from the charge holding unit to the FD region once over the entire screen and then returned to the first row and read from the photoelectric conversion unit to the FD region.

  Next, FIG. 4 shows the potential state at the time described in FIG. A black square is displayed at the top of the potential diagram to indicate that the photoelectric conversion unit and the like are shielded from light by the mechanical shutter.

  FIG. 4A shows an operation of resetting the charges of the photoelectric conversion unit and the charge holding unit before the signal charge is accumulated in the photoelectric conversion unit and the charge holding unit. Although not shown in FIG. 4A, the charge transferred to the FD region is discharged by the reset transistor. At this time, the mechanical shutter is in a closed state, and light does not enter the photoelectric conversion unit.

  4B to 4D show potential states during a signal charge generation period in which the mechanical shutter is opened and light enters the photoelectric conversion unit to perform photoelectric conversion.

  FIG. 4B shows a state immediately after the mechanical shutter is opened, and shows a state where no signal charge is generated in the photoelectric conversion unit.

  FIG. 4C shows a state in which a small amount of light is incident, and the photoelectric conversion unit without exceeding the potential barrier between the photoelectric conversion unit and the charge holding unit determined by the peak value of the pulse supplied to the first transfer transistor. In this state, the signal charge is held. Here, although the low level pulse is supplied to the first transfer transistor, the potential barrier is controlled to be relatively low so that the charge generated in the photoelectric conversion unit immediately moves to the charge holding unit. In order to realize such a state, as described above, the first transfer transistor can be realized as a buried channel type MOS transistor.

  FIG. 4D shows a state in which the signal charge generated in the photoelectric conversion unit moves to the charge holding unit beyond the potential barrier formed by the first transfer transistor, and the signal charge is also held in the charge holding unit. . Here, for the purpose of explanation, the number of charges existing in the region below the dotted line portion of the photoelectric conversion unit in FIG. 4D is Q1 = 10000, and the charge existing in the region below the dotted line portion of the charge holding unit. Is Q2 = 60000 and the number of charges existing above the dotted line is Q3 = 40000. Here, the dotted line indicates the height of the potential barrier in a state where a low level pulse is supplied to the first transfer transistor.

  FIG. 4E shows a potential diagram at T6a to T6b in FIG. 3, in which a low level pulse is supplied to the first transfer transistor and a high level pulse is supplied to the second transfer transistor (first level). Step). By this operation, the charges Q3 and Q2 in FIG. 4D are transferred to the FD region.

  FIG. 4F shows a potential diagram at T11a to T11b in FIG. 3, and shows a state in which a high level pulse is supplied to the first and second transfer transistors (second step). By this operation, the signal charge held in the photoelectric conversion unit is transferred to the FD region.

Note that if the number of charges in Q2 and Q3 described above exceeds the retention capacity of the FD region, there is a case where all charges cannot be transferred to the FD region at the timing of T6a to T6b. Such a case will be described with reference to FIG. As indicated by the alternate long and short dash line in the figure, since all charges cannot be transferred to the FD region, some charges that cannot be transferred remain in the charge holding portion. Even in such a case, it is sufficient that the remaining charge is designed to be completely read out to the FD region in the next transfer together with the charge of the photoelectric conversion unit as shown in FIG. Such a condition is that if the number of electrons that can be completely transferred to the FD area at once is QFDMAX,
Q2 + Q3-QFDMAX + Q1 <QFDMAX (Condition 1)
It is expressed by the formula.

If Q1 exceeds QFDMAX, driving of the pixel ends without reading out the remaining charge.
QFDMAX> Q1 (Condition 2)
Is also a necessary condition.

  Note that when the QFDMAX charge is read out to the FD region, a signal based on the transferred charge may not be correctly read out by the source follower or the readout circuit unit in the subsequent stage. That is, when the dynamic range of the readout circuit unit is lower than the dynamic range of the FD region, a part of the optical signal may be lost.

  Examples of readout circuits subject to restrictions include pixel amplification units, column amplification units provided in column circuits, column AD conversion units, final analog amplification units provided for each output channel, AD converters, and the like. Conceivable.

  As an example, the dynamic range of the amplifying unit of the pixel will be described. Here, a case where the amplification unit forms a source follower circuit with an amplification transistor and a constant current source will be described.

  When many signal charges are transferred from the photoelectric conversion unit and the charge holding unit to the FD region, the potential of the FD region decreases. When the potential difference of the FD region, that is, the gate of the amplification transistor and the potential of the source of the amplification transistor becomes lower than the threshold voltage (Vth) of the amplification transistor due to a decrease in the potential of the FD region, the source follower operation is stopped. Then, the signal cannot be read out. Even if a charge smaller than the dynamic range of the source follower circuit is read, if a high gain is applied in the column amplifier, the input dynamic range of the column amplifier circuit is limited to determine saturation in the read circuit. Is done.

  Therefore, QFDMAX is preferably set to a value that can transfer an amount of charge that does not exceed the dynamic range of these readout circuits.

  As a comparison, for example, when the charges of the photoelectric conversion unit and the charge holding unit are simultaneously transferred to the FD region by the transfer operation, and after the charge of the photoelectric conversion unit is transferred to the charge holding unit, the photoelectric conversion unit and the charge holding unit Let us consider a case where the charges are transferred to the FD region together. In these cases, the charge holding portion is provided and the saturation charge amount is increased, so that the dynamic range of the FD region and the readout circuit in the subsequent stage is often exceeded. That is, even if the charge holding unit is provided to increase the saturation charge amount, the increased charge cannot be effectively used. On the other hand, as in the present embodiment, after the charge of the charge holding unit is transferred to the FD region, the charge of the photoelectric conversion unit is transferred to the FD region via the charge holding unit, thereby increasing the charge holding unit. It is possible to effectively use the amount of charge.

  According to the present embodiment, it is possible to perform reading without being limited by the dynamic range of the readout circuit provided at the subsequent stage of the photoelectric conversion unit and the charge holding unit while expanding the dynamic range by providing the charge holding unit.

  This embodiment is particularly effective when the dynamic range including the photoelectric conversion unit and the charge holding unit is larger than the dynamic range of the readout circuit subsequent to the photoelectric conversion unit.

(Second embodiment)
FIG. 5 shows drive pulses according to the second embodiment. In this embodiment, as shown in T14 to T18, the number of transfers by the second transfer transistor from the charge holding unit to the FD region is a new point. A plurality of conduction pulses are supplied only to the second transfer transistor.

  When the conditional expression (1) shown in the first embodiment does not hold, the charge remains in the charge holding unit after transfer from the photoelectric conversion unit to the FD region, and the signal charge cannot be effectively used. For example, this can happen when QFDMAX is small or when the sum of Q1, Q2, and Q3 is larger than 2 × QFDMAX.

  In such a case, reading from the charge holding unit is further performed between the transfer of only the signal charge held by the charge holding unit and the transfer of the signal charge held by the photoelectric conversion unit. That is, a step of reading again only the signal charge remaining in the charge holding unit is added. Thereby, it is possible to suppress the charge from remaining in the charge holding unit after the last transfer is completed.

  This embodiment is particularly effective when the sum of Q2 + Q3 is much larger than QFDMAX.

  The conduction pulse that is additionally supplied to the second transfer transistor may be supplied once or may be supplied a plurality of times. What is necessary is just to determine suitably according to the value of Q2 + Q3.

(Third embodiment)
FIG. 6 shows drive pulses and timings according to the third embodiment. The present embodiment includes a readout circuit having a gain variable amplifier circuit that processes a signal output from a pixel. As indicated by 601, the gain of the amplifier circuit at the second step (the step of reading out charges from the photoelectric conversion unit and the charge holding unit) is switched to G (G> 1) times the gain at the first step. Is a feature. That is, the gain for the signal read at the second step is set higher than the gain for the signal read at the first step.

  The operation of this embodiment will be described with reference to FIG.

  FIG. 7A shows a distribution of charges generated in the photoelectric conversion unit of each pixel by incident light. A dotted line indicates a boundary between saturation in the photoelectric conversion unit and signal charge held in the charge holding unit.

FIG. 7B shows a signal based on the signal charge transferred from the charge holding portion in the portion A of FIG. 7A, that is, in the first step. Here, the random noise in this step can be expressed as N1 = V _RN1 (mVrms).

FIG. 7C is a diagram extracted from a portion B in FIG. 7A, that is, a signal based on the signal charge mainly transferred from the photoelectric conversion unit by the second step. Here, G is a gain in the readout circuit in the subsequent stage. The random noise in this case can be expressed as N2 = V _RN2 (mVrms).

  Thereafter, the signals shown in FIGS. 7B and 7C are combined to generate a single image. In order to make the output gradient with respect to the amount of light, that is, the signal sensitivity constant, the image is read with a high gain. It is necessary to divide the signal 7 (c) by the gain G.

  FIG. 7D shows values after the signal shown in FIG. Although the signal level decreases, the noise level also becomes 1 / G.

The random noise in this case can be expressed as N3 = V _RN2 / G (mVrms).

The random noise after combining these two signals is
((V _RN1 ) ² + (V _RN2 / G) ² ) ^0.5 , noise when gain is not applied ((V _RN1 ) ² + (V _RN2 ) ² ) Low noise compared to ^0.5 Can be read out.

  In this way, by increasing the gain of the amplification circuit with respect to the signal read in the second step, the ratio of noise to the pixel signal, that is, the SN ratio is increased, and as a result, high-sensitivity shooting of a low-luminance subject becomes possible. . As this amplifier circuit, for example, a column amplifier circuit can be used.

  The unique effects of this embodiment will be described. In the case of a low-luminance subject, most of the signal charge generated in the photoelectric conversion unit is held in the photoelectric conversion unit without exceeding the potential barrier between the photoelectric conversion unit and the charge holding unit. That is, in the case of a low-luminance subject, the ratio of the signal charge held by the photoelectric conversion unit is high, and thus the ratio of the signal read in the second step to the entire signal is high. Therefore, by increasing the gain of the readout circuit with respect to the signal based on the signal charge transferred in the second step, the ratio of noise to the pixel signal, that is, the SN ratio is increased, and as a result, high-sensitivity imaging of a low-luminance subject is possible. Become.

  Here, increasing the gain of the readout circuit narrows the input dynamic range. That is, when the gain is increased, the QFDMAX value described above decreases. When the gain is specifically G, the input dynamic range is QFDMAX / G. Therefore, in order to increase the gain of the readout circuit, it is important to read out a small number of charges with good controllability and a narrow dispersion width. On the other hand, in the readout by the second step, since the signal charge is accumulated in the photoelectric conversion unit without exceeding the relatively low potential barrier between the photoelectric conversion unit and the charge holding unit, the dispersion width is also It becomes possible to make it relatively narrow.

  According to the present embodiment, the charge holding unit and the photoelectric conversion unit are divided into a plurality of times and transferred using the potential barrier between them, thereby distributing the first and second signal outputs. Can be reduced. As a result, the charge transferred at one time can be increased, and the gain of the readout circuit can be set high.

Preferably, the saturation charge amount Q1 of the photoelectric conversion unit is
Q1 <QFDMAX / G
It is desirable to be designed to satisfy the following conditions.

(Fourth embodiment)
The present embodiment is characterized in that the signal obtained when only the signal charge held in the charge holding unit is transferred to the FD region is switched between the case where it is used as the image forming signal and the case where it is not used. When the signal output by the signal charge held by the charge holding unit is below a certain threshold, it is possible to prevent the addition of random noise by not using the signal as a signal for image formation Become. The signal to be compared with the threshold value here is a signal obtained when only the signal charge held in the charge holding unit is transferred to the FD region, mainly the signal charge held in the photoelectric conversion unit. Any of the signals obtained when transferred to the area may be used. Furthermore, information regarding the incident light amount may be acquired by an AE sensor provided separately, and addition or non-addition of a signal may be switched based on the incident light amount information. Regardless of which method is used, addition or non-addition may be switched depending on the amount of incident light in each pixel or a region including a plurality of pixels.

  This will be described in more detail with reference to FIG. FIG. 8A shows a signal obtained when only the signal charge held in the charge holding unit is transferred to the FD region, and FIG. 8B mainly shows the signal charge held in the photoelectric conversion unit in the FD region. The signal obtained when transferring to is shown. In FIG. 8A, random noise components occupy most of the signal in the region indicated by the white square, and the S / N ratio deteriorates when addition is performed. On the other hand, as shown in gray in FIG. 8B, the signal has a level to be used as a signal for image formation. For example, in the case of a low-luminance subject, signal charges are accumulated only in the photoelectric conversion unit without exceeding the potential barrier between the photoelectric conversion unit and the charge holding unit. That is, there is almost no signal charge in the charge holding portion. Therefore, even if the charge transfer operation of the charge holding unit is performed in such a state, the amount of noise only increases.

  In order to suppress this, when the signal from the charge holding unit is equal to or lower than a predetermined threshold value, the signal obtained when only the signal charge held in the charge holding unit is transferred to the FD region is used for image formation. This signal is not used. Alternatively, when the signal from the photoelectric conversion unit is equal to or greater than a predetermined threshold value, it may be determined that the signal from the charge holding unit is present and used as a signal for image formation.

  By applying this embodiment to other embodiments, in addition to the effects obtained in each embodiment, it is possible to form an image without deteriorating the S / N ratio when the amount of incident light is relatively low. It becomes.

Specifically, as shown in FIG. 9, the proportion of random noise in the low illuminance side increases, but according to the present embodiment, low noise characteristics that are not affected by the random noise _VRN1 can be realized. Become.

(Fifth embodiment)
FIG. 10 shows drive pulses according to the fifth embodiment. In this embodiment, it is a new point that partial transfer is performed using a partial intermediate level pulse for transfer from the charge holding unit to the FD region. The intermediate level pulse is a pulse having a peak value between a conduction pulse and a non-conduction pulse. This intermediate level pulse can be realized by configuring the control unit 102 to be able to supply the intermediate level pulse.

  First, at T1, a high level pulse is supplied to the reset transistor and the first and second transfer transistors to be in a conductive state, and the charges in the photoelectric conversion unit, the charge holding unit, and the FD region are reset. At this time, the mechanical shutter is closed.

  At T2, the mechanical shutter is opened and light is incident on the photoelectric conversion unit. At this time, the pulse supplied to each transfer transistor is a low-level pulse that makes each transfer transistor non-conductive.

  At T3, the mechanical shutter is closed.

  At T4, a low level pulse is supplied to PRES in the first pixel row, and a high level pulse is supplied to PSEL. Here, they are performed simultaneously, but may be performed at different times. However, in order to suppress kTC noise in the reset unit, it is necessary that a low level pulse is supplied to PRES at least during a period in which a high level pulse is supplied to a PTN that samples a noise signal.

  At T5, a high level pulse is supplied to PTN, and the noise signal of the pixels in the first row is held in the column circuit.

  At T6a, an intermediate level pulse is supplied to the second transfer transistor in the first pixel row, and a potential generated by supplying the intermediate level pulse among the signal charges held in the photoelectric conversion unit and the charge holding unit. Only the signal charge of the part exceeding the barrier is transferred to the FD region (first step).

  At T6b, a low-level pulse that is turned off is supplied to the second transfer transistor in the first pixel row.

  At T7, a high level pulse is supplied to the PTS, and a signal based on the electric charge read out in the first step is held in the column circuit.

  At T8, a low level pulse is supplied to PSEL in the first pixel row, and a high level pulse is supplied to PRES. With this operation, the charge transferred to the FD region in the first step is reset.

  At T9, a high level pulse is supplied to PTN, and the noise signal of the pixels in the first row is held in the column circuit.

  At T10a, a high-level pulse that is turned on is supplied to the second transfer transistor in the first pixel row (second step). Preferably, a pulse having a peak value sufficient to completely transfer the charges in the charge holding portion to the FD region is supplied. At this time, the first transfer transistor in the first pixel row is in a state where a low level pulse is supplied.

  At T10b, a low-level pulse that is turned off is supplied to the second transfer transistor in the first pixel row.

  At T11, a high level pulse is supplied to the PTS, and a signal based on the charge transferred in the second step is held in the column circuit.

  At T12, a low level pulse is supplied to PSEL in the first row, and a high level pulse is supplied to PRES.

  At T13, a high level pulse is supplied to PTN, and the noise signal of the pixels in the first row is held in the column circuit.

  At T14a, a high-level pulse that is turned on is supplied to the first transfer transistor and the second transfer transistor in the first pixel row (third step). As a result, the signal charge held in the photoelectric conversion unit is transferred to the FD region.

  At T14b, a low-level pulse that is turned off is supplied to the first transfer transistor and the second transfer transistor in the first pixel row.

  At T15, a signal based on the signal charge transferred in the third step is held in the column circuit.

  Thereafter, the signal processing unit 105 adds the signals obtained in the first to third steps. As a result, most of the charge generated by the photoelectric conversion in the photoelectric conversion unit can be handled as an image forming signal regardless of the dynamic range of the readout circuit arranged downstream of the photoelectric conversion unit and the charge holding unit. Become.

  By repeating the operations from T4 to T15 for each pixel row, it is possible to read out one frame of signal.

  Further, the transfer (second step) with the intermediate level pulse of T6a to T6b is performed only once, but this may be repeated a plurality of times.

  Next, FIG. 11 shows the potential state at the time described in FIG. A black square is displayed at the top of the potential diagram to indicate that the photoelectric conversion unit and the like are shielded from light by the mechanical shutter.

  FIG. 11A shows an operation of resetting the charges of the photoelectric conversion unit and the charge holding unit before the signal charge generation period. The potential state in the period from T1 to T2 in FIG. 10 is shown. Although not shown in FIG. 11A, the charges transferred to the FD region are discharged by the reset transistor. At this time, the mechanical shutter is in a closed state, and no light enters the photoelectric conversion unit.

  FIG. 11B to FIG. 11D show potential states during a signal charge generation period in which the mechanical shutter is opened and light enters the photoelectric conversion unit to perform photoelectric conversion.

  FIG. 11B shows a state immediately after the mechanical shutter is opened, and shows a state where no signal charge is generated in the photoelectric conversion unit.

  FIG. 11C shows a state in which a small amount of light is incident, and the photoelectric conversion unit without exceeding the potential barrier between the photoelectric conversion unit and the charge holding unit determined by the peak value of the pulse supplied to the first transfer transistor. In this state, the signal charge is held. Here, although the low level pulse is supplied to the first transfer transistor, the potential barrier is controlled to be relatively low so that the charge generated in the photoelectric conversion unit immediately moves to the charge holding unit. In order to realize such a state, as described above, the first transfer transistor can be realized as a buried channel type MOS transistor.

  FIG. 11D shows a state in which the signal charge generated in the photoelectric conversion unit moves to the charge holding unit beyond the potential barrier formed by the first transfer transistor, and the signal charge is also held in the charge holding unit. . Here, for the purpose of explanation, the number of charges existing in the region below the dotted line portion of the photoelectric conversion unit in FIG. 11D is Q1 = 10000, and the charge existing in the region below the dotted line portion of the charge holding unit. Is Q2 = 60000 and the number of charges existing above the dotted line is Q3 = 40000.

  FIG. 11E shows a potential diagram at T6a to T6b in FIG. 10, in which a low level pulse is supplied to the first transfer transistor and an intermediate level pulse is supplied to the second transfer transistor (second phase). Step). By this operation, all of the charge Q3 in FIG. 11D and a part of the charge of Q2 are transferred to the FD region. The amount of charge transferred to the FD region can be arbitrarily set by the peak value of the intermediate level pulse supplied to the second transfer transistor.

  FIG. 11 (f) shows a potential diagram at T10a to T10b of FIG. 10, where a low level pulse is supplied to the first transfer transistor and a high level pulse is supplied to the second transfer transistor (third). Step). By this operation, the remaining charges of the signal charges transferred in the second step among the signal charges held in the charge holding portion are transferred to the FD region. Here, for example, 50000 signal charges are transferred.

  FIG. 11G shows a potential diagram at T14a to T14b in FIG. 10, where a high level pulse is supplied to the first and second transfer transistors (fourth step). By this operation, the signal charge held in the photoelectric conversion unit is transferred to the FD region. Here, 2000 charges are transferred.

  The effects unique to this embodiment will be described. In the first embodiment, it is assumed that all the signal charges of QFDMAX can be handled by the source follower and the readout circuit at the latter stage of the FD region. In other words, when the QFDMAX signal is transferred to the FD area, the value is correctly output to the outside without circuit saturation.

  However, in reality, the charge that can be handled by the circuit may be QFDMAX or lower (hereinafter referred to as QFDMAX2) due to limitations on the power supply voltage used in the amplifier circuit. In such a case, the amount of charge transferred to the FD region must be limited.

  At that time, by performing transfer using an intermediate level pulse, the transfer amount of charges to the FD region can be limited to QFDMAX2 or less, and all charges can be read correctly. That is, the peak value of the intermediate level pulse is set so that the number of signal charges transferred by the intermediate level pulse is less than the maximum value that can be read by the reading circuit.

At that time, the saturation charge amount Q1 of the photodiode is Q1 <QFDMAX2.
It is desirable to be designed so that Of course, by setting Q1 sufficiently lower than QFDMAX2, it is possible to apply a high gain only to the signal from the photoelectric conversion unit and reduce noise in the dark part.

  This embodiment is particularly effective when the dynamic range of the FD region is larger than the dynamic range of the readout circuit at the subsequent stage of the photoelectric conversion unit.

  According to the present embodiment, it is possible to perform reading without being limited by the dynamic range of the readout circuit provided at the subsequent stage of the photoelectric conversion unit and the charge holding unit while expanding the dynamic range by providing the charge holding unit.

(Sixth embodiment)
This embodiment is different from the fifth embodiment in that an intermediate level pulse that has been supplied only once to each pixel row is supplied a plurality of times.

  With respect to the drive pulse diagram, a second intermediate level pulse may be further provided between the intermediate level pulse and the high level pulse in the conductive state with respect to the drive pulse of FIG. Correspondingly, a pulse for resetting the FD region, a noise signal in the column circuit, and a sampling pulse of the optical signal are also added. As for the crest value, an arbitrary value can be selected between a high level pulse for turning on the transfer transistor and a low level pulse for turning off the transfer transistor. Pulses having the same peak value may be supplied, or pulses having different peak values may be supplied.

  According to this embodiment, the transfer conditions from the photoelectric conversion unit to the FD region can be set finely by applying to other embodiments, and the dynamic range of the readout circuit at the subsequent stage than the photoelectric conversion unit is finely supported. It becomes possible to do.

(Seventh embodiment)
The present embodiment is an example in which the intermediate level pulse supplied in the fifth and sixth embodiments is changed according to the temperature. FIG. 12 shows a block diagram of an imaging system including the solid-state imaging device of the present embodiment.

  Reference numeral 1201 denotes a solid-state imaging device. 1202 is a temperature detection unit. Although it is arranged in the solid-state imaging device here, it may be arranged in the vicinity of the outside of the solid-state imaging device. Reference numeral 1203 denotes a CPU. Reference numeral 1204 denotes a control unit. The solid-state imaging device and the like are controlled in accordance with temperature information from the temperature detection unit 1202 and a control signal from the CPU 1203. 1205 is a variable voltage supply source. In accordance with a control signal from the control unit 1204, a voltage corresponding to the temperature is supplied to the solid-state imaging device. By supplying this voltage to the vertical scanning circuit of FIG. 1, the peak value of the intermediate level pulse or the supply time can be varied according to the temperature.

  The operation flow of this embodiment will be described. First, in the signal charge generation period, signal charges are accumulated in the photoelectric conversion unit. After a predetermined time has elapsed, the signal charge generation period ends. Thereafter, the temperature detection unit acquires temperature information of the solid-state imaging device itself or in the vicinity thereof. Then, a lookup table in which voltage information corresponding to the acquired temperature information is recorded is accessed, and control is performed so that a voltage corresponding to the temperature is supplied from the variable voltage supply source. Thereafter, the above read operations are performed.

  The energy of the signal charge varies depending on the temperature, and the number of charges transferred to the FD region differs when the potential barrier generated by the intermediate level pulse is the same height even if the temperature changes. On the other hand, in the present embodiment, a temperature detection unit is provided, and the peak value of the intermediate level pulse or the supply time is changed or switched according to the signal from the temperature detection unit.

  If this embodiment is applied to the fifth and sixth embodiments, in addition to the effects obtained in the fifth and sixth embodiments, also when a temperature change occurs in the vicinity of the solid-state imaging device and the solid-state imaging device. By supplying the intermediate level pulse, it is possible to suppress a change in the number of charges transferred.

(Eighth embodiment)
This embodiment is an example in which the intermediate level pulse described in the above-described embodiment is changed according to the gain of a readout circuit provided in the subsequent stage of the photoelectric conversion unit. Here, as a readout circuit provided in the subsequent stage of the photoelectric conversion unit, an amplification unit provided in the pixel, a column amplification circuit provided in the column circuit, a parallel signal from the column circuit is converted into a serial signal, and then output to the outside. There is an output amplifying section for the purpose.

  FIG. 13 shows a block diagram of an imaging system including the solid-state imaging device of the present embodiment.

  Reference numeral 1301 denotes a solid-state imaging device. Reference numeral 1302 denotes a control unit. Reference numeral 1303 denotes a CPU. Reference numeral 1304 denotes a variable voltage supply source. Switching of the gain of the readout circuit provided in the subsequent stage of the photoelectric conversion unit is performed by the control unit 1302 and the CPU 1303. This gain control is appropriately performed according to, for example, sensitivity switching, a change in reading speed, and the like.

  The operation flow of this embodiment will be described. First, in the signal charge generation period, signal charges are accumulated in the photoelectric conversion unit. The accumulation ends after a predetermined time has elapsed. Thereafter, a look-up table in which voltage information corresponding to the gain of the readout circuit provided at the subsequent stage of the photoelectric conversion unit is accessed, and control is performed so that a voltage corresponding to the gain is supplied from the variable voltage supply source. . Thereafter, the above read operations are performed. In addition, a temperature detection unit may be provided to perform the temperature correction described in the fourth embodiment at the same time.

  The effect of changing the peak value of the intermediate level pulse or the supply time when the readout circuit gain is variable will be described with reference to FIG.

  FIG. 14A shows the incident light quantity dependency (photoelectric conversion characteristics) of the output voltage when the first intermediate level pulse is supplied, and FIG. 14B shows the photoelectric conversion characteristics when the second intermediate level pulse is supplied. FIG. 14C shows photoelectric conversion characteristics when a high level pulse is supplied. FIG. 14D shows output signal characteristics after addition after the charge is transferred to the FD region by each readout pulse. FIGS. 14E to 14H show photoelectric conversion characteristics and output signal characteristics when the pulse peak value is changed in accordance with the gain of the readout circuit. Specifically, when the gain of the readout circuit is changed from the first gain to the second gain that is higher than the first gain, the pulse peak value is made lower than in FIGS. 14 (a) to 14 (c). It is an example.

  When the read circuit gain is variable, the incident light amount (saturated light amount) until the circuit is saturated differs for each gain of the read circuit. Reference numerals 904a, b, and c denote cases where the gain of the readout circuit is low, and reference numerals 906a, b, and c denote cases where the gain of the readout circuit is high. When the gain of the readout circuit is low, the saturation light amount is as high as 905a, b, and c, so that all the charges transferred to the FD region can be used as signals for image formation. However, when the gain of the readout circuit is high, the saturation light amount decreases to 907a, b, and c, so that a part of the charge transferred to the FD region exceeds the input dynamic range of the readout circuit and is lost as information. End up. As a result, as shown in FIG. 14D, as the combined photoelectric conversion characteristics, the characteristics after combining the signals of the regions A, B, and C obtained in FIGS. 14A to 14C are as follows. , 909, and a photoelectric conversion characteristic having a dead zone. For reference, it is preferable to obtain a linear characteristic such as 908 when there is no omission.

  In such a case, as shown in FIGS. 14E to 14G, the peak value of the intermediate level pulse is limited to a lower level than in FIGS. 14A to 14C. Alternatively, the pulse supply time is shortened to limit transient charge movement, and the number of charges transferred when the intermediate level pulse is used is reduced. As a result, photoelectric conversion characteristics as shown in FIGS. 14E to 14G can be obtained even if the pulse is supplied and read out three times in the same manner as FIGS. 14A to 14C. The photoelectric conversion characteristics after combining these signals are shown in FIG. Since there is no missing information as compared with FIG. 14 (d), there is no undesirable dead zone even after the regions A, B, and C are combined.

  Since the number of electrons that can be read at one time is limited, the amount of saturated light after synthesis is reduced. In that case, the decrease in the amount of saturated light can be suppressed by increasing the number of times of reading with an intermediate level pulse as needed and the applied intermediate level voltage.

  If this embodiment is applied to other embodiments, in addition to the effects obtained by each embodiment, a dead zone is also generated when the gain of a readout circuit provided at a stage subsequent to the photoelectric conversion unit changes. It is possible to obtain a signal having continuous photoelectric conversion characteristics even after synthesis.

2 photoelectric conversion unit 3 charge holding unit 8 first transfer unit 9 second transfer unit 10 reset unit 12 amplifying unit 102, 1204 control unit

Claims (15)

A photoelectric conversion unit, a charge holding unit that holds signal charges generated in the photoelectric conversion unit, a first transfer unit that transfers signal charges from the photoelectric conversion unit to the charge holding unit, and held by the charge holding unit. A plurality of pixels having a second transfer unit that transfers the signal charges,
An amplifier for amplifying a signal based on the signal charge transferred by the second transfer unit;
A reset unit for resetting the signal charges transferred to the amplification unit;
A solid-state imaging device comprising:
The charge holding unit is disposed in an electrical path between the photoelectric conversion unit and the amplification unit,
After the mechanical shutter is opened and light is incident on the photoelectric conversion unit, with the mechanical shutter closed, a conduction pulse is supplied to the second transfer unit and the signal charge held in the charge holding unit is supplied to the amplification unit. After transferring and supplying a conduction pulse to the reset unit to reset the signal charge transferred to the amplification unit, the conduction pulse is supplied to the first and second transfer units to be held in the photoelectric conversion unit. A solid-state imaging device, wherein the signal charge transferred to the amplifying unit is transferred.   The operation of supplying a conduction pulse to the second transfer unit and transferring the signal charge held in the charge holding unit to the amplification unit supplies the conduction pulse to the first transfer unit after the mechanical shutter is closed. The solid-state imaging device according to claim 1, wherein the solid-state imaging device is performed without any problem.   The solid-state imaging device according to claim 1, wherein a charge path between the photoelectric conversion unit and the charge holding unit has a buried channel structure.   A potential barrier to a signal charge in a charge path between the photoelectric conversion unit and the charge holding unit during a period in which signal charge is accumulated in the photoelectric conversion unit and the charge holding unit is the photoelectric conversion unit. The solid-state imaging device according to claim 1, wherein the solid-state imaging device is lower than a potential barrier between the first region and the other region.   5. The solid according to claim 1, wherein a plurality of conduction pulses are supplied to the second transfer unit before the conduction pulse is supplied to the first and second transfer units. Imaging device.   The intermediate level pulse having a peak value between the conduction pulse and the non-conduction pulse can be supplied to the second transfer unit. Solid-state imaging device.   The solid-state imaging device according to claim 6, wherein a peak value of the intermediate level pulse can be selected from a plurality of values. Furthermore, it has a readout circuit for processing the signal output from the amplification unit,
8. The solid-state imaging according to claim 7, wherein the peak value of the intermediate level pulse is set so that the number of signal charges transferred by the intermediate level pulse is less than a maximum value readable by the readout circuit. apparatus. Furthermore, it has a temperature detector,
The solid-state imaging device according to claim 7, wherein a peak value of the intermediate level pulse is changed according to temperature information from the temperature detection unit. Furthermore, it has a gain variable amplification circuit for processing the signal output from the amplification unit,
The gain of the amplifier circuit is more than the gain for the signal obtained by transferring the signal charge held in the photoelectric conversion unit to the amplification unit by supplying a conduction pulse to the first and second transfer units. 10. The gain for a signal obtained by supplying a conduction pulse to the second transfer unit and transferring the signal charge held in the charge holding unit to the amplifying unit is higher. The solid-state imaging device according to any one of the above.   A case where a signal obtained by supplying a conduction pulse to the second transfer unit and transferring the signal charge held in the charge holding unit to the amplification unit according to the amount of light incident on the pixel is used for image formation. The solid-state imaging device according to any one of claims 1 to 10, wherein the solid-state imaging device can be switched between when it is not used.   The solid-state imaging device according to claim 1, wherein the amplification unit and the reset unit are shared by a plurality of pixels.   The solid circuit according to claim 1, further comprising: a column circuit that performs processing on signals read from the plurality of pixels, wherein the column circuit includes an AD conversion unit. Imaging device.   A charge is applied to the counter electrode through an insulating film during a period in which the signal charge is accumulated in the charge holding unit, and charges having a polarity opposite to that of the signal charge are accumulated on the surface of the charge holding unit. Item 14. The solid-state imaging device according to any one of Items 1 to 13.   15. An imaging system comprising the solid-state imaging device according to claim 1, further comprising a mechanical shutter.

Images