JP3089298B2 - Photoelectric conversion device and solid-state imaging device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a solid-state imaging device using a photoelectric conversion device.

[0002]

2. Description of the Related Art Solid-state image pickup devices represented by CCD type image pickup devices tend to be miniaturized and have high resolution in order to be mounted on portable electronic devices such as video cameras and electronic still cameras. In order to prevent the sensitivity from being reduced even when the size is reduced, the effective aperture area needs to be close to 100%. Therefore, an on-chip microlens that collects incident light above the photoelectric conversion element is employed. And
The apparent light receiving area of the photoelectric conversion element is increased.

By the way, even with such a contrivance, S / N, which is one of the most important requirements for determining image quality, is as follows:
Ultimately, it is uniquely determined by the shot noise of the incident light itself, which is a physical limit. S / in this case
The N value is proportional to the square root of p, where p is the total number of incident photons detected in one pixel. Since p is proportional to the product of the accumulation time and the number of incident photons per unit time, the S / N is necessarily improved if p is sufficiently increased by long-time exposure.
In other words, when the light receiving area is the same, the imaging of a dark subject with a small number of incident photons, in other words, a dark subject is theoretically realized only by long-time exposure.

[0004]

However, the current T
As in the NTSC standard system, which is one of the V systems, the storage time per frame / field is usually determined. In addition, in order to secure the dynamic resolution, it is not realistic to increase or decrease the exposure time unnecessarily in accordance with the luminance of the subject.

The present invention provides a solid-state imaging device capable of realizing an improvement in shot noise S / N due to photons in consideration of a limit of S / N caused by light itself having photon number fluctuation due to Poisson distribution. It is an object of the present invention to provide a photoelectric conversion device used therefor.

[0006]

In order to achieve the object, a first aspect of the present invention (corresponding to claim 1) includes a photoelectric conversion unit for generating a signal charge in accordance with incident light;
The signal charge generated by the photoelectric conversion is increased by a predetermined gain.
Amplifying section to be amplified and signal charges amplified by said amplifying section
To be stored while resetting
And the predetermined time of the amplification unit at a predetermined time (t).
Is stored in the memory by the predetermined time (t).
A photoelectric conversion device comprising: a sequential gain control unit that controls according to the accumulated signal charge amount .

[0007] The second invention (corresponding to claim 2) provides:
A first photoelectric conversion unit that generates an electric charge according to incident light, a storage unit that stores the electric charge generated by the first photoelectric conversion unit, and amplifies the electric charge output from the accumulation unit with a predetermined gain. includes an operational amplifier section, and a storage unit for memory the amplified signal charges by the operational amplifying unit, and the predetermined gain of the operational amplifier section, when the definitive during one field period, the gain is applied ( The photoelectric conversion device is controlled based on the amount of charge stored in the storage unit during a predetermined period before t) .

A third of the present invention (Claim 6 corresponds) is arranged on the light receiving portion, a first photoelectric conversion unit that generates charges in response to incident light, the signal charges from the first photoelectric conversion unit And a negative feedback amplifier, wherein the negative feedback amplifier is of a conductivity type opposite to that of the first photoelectric converter, and is adjacent to the first photoelectric converter in the light receiving unit. provided Te, and a second photoelectric conversion unit that generates charges in response to incident light, the first of the same conductivity type as the photoelectric conversion unit, and a charge source for supplying a charge to the storage section, 1 In the photoelectric conversion device, during a field period, a signal charge flowing out of the charge source is modulated by a signal charge generated or accumulated in the second photoelectric conversion unit.

A fourth aspect of the present invention ( corresponding to claim 8 ) includes a photoelectric conversion unit that generates charges in accordance with incident light, and a storage unit that stores signal charges stored in the photoelectric conversion unit. The avalanche photodiode is a part of the photoelectric conversion unit that performs photoelectric conversion, and the avalanche photodiode includes a plurality of band offsets stacked on the storage unit via a plurality of conductive thin films connected to each other with fixed resistors. Avalanche multiplication layer (multiplication layer) films, the compositions of these avalanche multiplication films vary with respect to the laminating direction, and according to the number of accumulated charges accumulated in the accumulation section during one field period. hand,
This is a photoelectric conversion device characterized in that the degree of multiplication of each stacked avalanche multiplication film changes.

A fifth aspect of the present invention (corresponding to claim 15 ) is a photoelectric conversion unit that generates electric charges in accordance with incident light, a storage unit that stores electric charges generated by the photoelectric conversion units, and a storage unit that stores electric charges. A / D converter that digitizes the accumulated signal charges for each destructive or non-destructive readout, and a predetermined number of successive signal charges for each signal charge sequentially output from the A / D converter. Among them, the photoelectric conversion device is characterized by discarding the largest signal charge and the smallest signal charge.

In general, even if the obtained signal charges are photoelectrically converted at a constant quantum efficiency and a constant output amplification is performed by a multiplication means having a constant gain, the signal charge caused by the light shot noise is increased. / N never changes. On the contrary,
In the present invention, the multiplication gain is controlled in accordance with the history of the number of accumulated charges (the number of accumulated charges up to that time). That is, if the number of accumulated charges up to that point is large, the gain is made small, and if the number of accumulated charges is small, the gain is kept large. FIG. 23 is a diagram schematically showing the shot noise of the light. It can be seen that the number of photons fluctuates.
Then, M _{t to} M _{t + 1} , M _{t + 1 to} M _{t + 2} ,. . . . In each of the timing periods, the S / N ratio is improved by changing the gain in the middle and late stages in the period corresponding to the large number of initial photons. FIG. 24 shows the fluctuation distribution, and the present invention means that the S / N ratio is improved by cutting off the hatched portion.

As a result, for example, the average number of incident photons P (pieces /
s), the accumulation period T within one field period
_The shot S of the number of incident photons at the time of _F (s)
Although / N is (PT _F ) ^1/2 , in the case of the present invention, it is possible to obtain an S / N exceeding this value. FIG. 8 shows the S / N with respect to the number of incident photons in one embodiment of the present invention.
FIG. 4 is a simulation diagram comparing a dotted line and a conventional normal S / N (solid line). Details will be described later.

The above logic is applied to an apparatus for imaging not only visible light but also all electromagnetic waves such as infrared rays, ultraviolet rays, and X-rays. Therefore, the present invention can be applied to all these imaging apparatuses.

[0014]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a schematic diagram showing an embodiment of the present invention. In the present embodiment, photoelectric conversion means 1 for generating electric charges in accordance with incident light, amplifying means 2 for amplifying signal charges generated by photoelectric conversion, and memory for storing signal charges amplified with a predetermined gain And a sequential gain control means 4 for controlling the gain of the amplifying means 2 in accordance with the number of accumulated charges. In the operation, when light enters the unit 1 having a photoelectric conversion function, the light is photoelectrically converted and signal charges are generated according to the amount of incident light. The signal charge is amplified by the amplification means 2 having an amplification function with a predetermined gain and stored in the memory 3. The number of accumulated charges is sequentially fed back to the gain control means 4, and the gain control means 4 controls the gain of the subsequent signal charges based on the feedback.

Although FIG. 1 uses an equivalent circuit notation and thus uses analog notation, digital notation using an ADC and DAC can also be used.

The number of accumulated charges fed back is:
It is preferable that the product of the multiplication gain M and the number of accumulated charges per unit time n be a function of a value integrated by the accumulation time. Therefore, the multiplication gain in this case is determined by the following equation (1).

[0018]

(Equation 1)

The amplifying means includes digital signal processing represented by a digital signal processor (DSP);
Various forms such as a negative feedback amplifier circuit and an avalanche photodiode are possible.

When a DSP is provided after the output section of the line sensor or the area sensor, the DSP is made to execute an arithmetic expression for approximating the integral expression for calculating the history of the number of accumulated charges by successive approximation. This is preferable in that the gain can be accurately digitally controlled. However, in order to reduce the tracking error inherent in the sequential calculation due to the internal delay of the digital feedback system, it is necessary to adopt a solid-state image sensor having a reading speed as high as possible.

In the case where a negative feedback amplifier circuit is employed together with the photodiode in the pixel portion, that is, the photoelectric conversion portion, a second photoelectric conversion portion provided in the photoelectric conversion portion with a conductivity type opposite to that of the photoelectric conversion portion; And a charge source of the same conductivity type as that of the second photoelectric conversion unit, wherein the signal charge generated or accumulated in the second photoelectric conversion unit modulates the signal charge flowing out of the charge source. In the case of the DSP, the gain with respect to the signal charge of each pixel had to take a constant value during the period from reading to reading, but when a negative feedback amplifier circuit was introduced into the pixel portion, the gain was determined according to the amount of accumulated charge. Since it changes at any time, it is possible to reduce the following error inherent in the sequential calculation.

In the case where an avalanche photodiode (APD) is employed in the pixel portion, the signal charge is multiplied by the avalanche multiplication by using the avalanche photodiode (APD) in the charge storage mode in the photoelectric conversion portion, and at the same time, according to the number of stored charges. The potential difference between the pn junctions is reduced, and the multiplication gain of the subsequent signal charge is suppressed autonomously. However, conventional Si
In the case of -APD, the multiplication gain fluctuates due to the indeterminate where in the depletion layer the avalanche multiplication phenomenon occurs. This is commonly referred to as excess noise and refers to new noise. The higher the gain setting, the more pronounced this excess noise becomes. On the other hand, the use of the gradient superlattice type is advantageous in that excess noise can be suppressed since the position where avalanche multiplication occurs can be specified. Furthermore, since the multiplication gain can be calculated digitally as proposed in the present embodiment, the gain control can be performed accurately. However, on the other hand, when the gain is increased, the way of suppression in a large gain region greatly fluctuates by a power of two. For example, in the initial state, the gain is 2 10 (1024) to 9 (512).
Suddenly decrease. In order to prevent this, when realizing a large initial gain (for example, 8 times or more), at least one band offset portion of the stacked heterojunction is reduced by 3 /
It is necessary to keep the gain at an appropriate value of 2 · Eg (Eg: gap energy) or less so that the gain does not increase by this power number every time one band offset is exceeded.

Next, specific examples briefly mentioned in the embodiment of the present invention will be described.

Embodiment 1 This is an embodiment using a DSP as a multiplying means. FIG. 2 is a block diagram showing a configuration in a case where the present invention is applied to a line sensor or an area sensor according to this embodiment. The solid-state imaging device 11 includes a high-speed reading CCD 12 in a line sensor or an area sensor, and an A / D converter 121.
And a DSP 1 including a line memory 13 and an operation unit 14
5 and a frame memory 16.

The operation unit 14 stores the following equation 2 for gain control and outputs a signal. Here, M _t : multiplication gain at time t, n: number of accumulated charges per unit time,
M _{t + 1} ; multiplication gain at time t + 1.

[0026]

(Equation 2)

According to the device 11, the signal charge of the m-th line read by the CCD 12 is stored in the line memory 13
, And the signal charges of the m-th and earlier lines are stored in the frame memory 16. From the frame memory 16,
The line at the pixel position corresponding to the line read from the CCD 12 to the DSP 15 is input to the arithmetic unit 14, and the arithmetic unit 14 performs the calculation according to the above equation 2 up to the (n−1) th line, and calculates the line memory 13 with the calculated gain. Is multiplied by the signal charge of the n-th line and output.

That is, at a certain timing, the signal charges are stored in the line memory 13 for each line, and the arithmetic unit 14 reads the signal accumulation state up to that time from the frame memory 16 and calculates the gain by the gain determination table or the determination function. Then, the signal charge in the line memory 13 is multiplied by the gain and stored in the frame memory 16.

The same operation is repeated at the next timing.
The cycle of those timings is, for example, 1/1000 second.

This is defined as one field / frame period (1 /
(30 seconds or 1/60 second), and after the end of one field / frame period, the signal charge is output from the frame memory 6 to the outside.

In other words, during the field / 1 frame period, the gain is changed based on the number of charges stored in the storage unit 16 during a predetermined period before the gain is applied. become.

Embodiment 2 This is an embodiment of the CCD solid-state imaging device of the present invention using a negative feedback type amplifier as a multiplication means. 3 to 5 show the structure of the device of this embodiment. 3 is a plan view thereof, FIG. 4 is a sectional view taken along line A1-A2 of FIG. 3, and FIG. 5 is a sectional view taken along line B1-B2 of FIG.
It is sectional drawing.

The CCD solid-state imaging device 21 includes a band-shaped vertical transfer region 22 made of an N-type semiconductor and N-type vertical transfer regions 22 arranged in the same direction.
A photoelectric conversion unit 23 made of a type semiconductor, a separation unit 24 made of a P-type semiconductor and separating each photoelectric conversion unit 23, and transfer electrodes φV1 to V5 are provided. In the present embodiment, the photoelectric conversion unit 23 has a conventional structure without a negative feedback amplifier,
Although those having the structure of the present embodiment having the negative feedback amplifier are alternately arranged, they may be constituted only by the negative feedback section. Hereinafter, only the structure of the present embodiment will be described.

The photoelectric conversion unit 23 includes a second photoelectric conversion unit 25 made of a P-type semiconductor whose tip is bitten into the inside from the side near the separation unit 24 and the second photoelectric conversion unit 25 along the side far from the vertical transfer region 22. An electron source 26 made of an N ⁺ type semiconductor is provided so as to face the two photoelectric conversion units 25. The second photoelectric conversion unit 25 and the electron source 26 are paired to produce a narrow channel effect, control the electron flow of signal charges from the electron source 26 to the photoelectric conversion unit 23, and function as a negative feedback amplifier.

Various electrodes are provided on the surfaces of the vertical transfer region 22 and the photoelectric conversion section 23 via an insulating layer 27 made of an oxide film. At a position facing the vertical transfer region 22 with the photoelectric conversion unit 23 interposed therebetween, an electrode φVS made of, for example, aluminum is provided in parallel with the vertical transfer region 22.
It is connected to the electron source 26 by a contact hole indicated by a cross. On the surface between the adjacent photoelectric conversion units 23, φV3, φV2 and φV4 opposing so as to partially overlap φV3, and φV1 above φV4 are all formed of an oxide film (not shown) made of polysilicon. And extends over the vertical transfer area 22. In the figure, φV1 is indicated by a thick broken line, φV2 and φV4 are indicated by thin solid lines, and φV3 is indicated by a thin two-dot chain line. φV1 is the second photoelectric conversion unit 2
5 extends above. Further, the vertical transfer region 22 and the photoelectric conversion unit 23 are separated from each other by P ⁺ layers provided in a band shape on both sides of the vertical transfer region 22, but φV1 is a vertical direction not covered by any other electrode. It extends over the transfer region 22 and further overlaps on the edge of the photoelectric conversion unit 23 over the P ⁺ layer. The overlapping portion between φV1 and the vertical transfer region 22 functions as a transfer electrode, and the overlapping portion between φV1 and the photoelectric conversion unit 23 functions as a read gate. φV5 has the same shape as φV1 except that it does not extend above the second photoelectric conversion unit 25.

Next, the operation of this embodiment will be described with reference to FIGS. FIG. 6 is a potential diagram of the electrode at each time,
FIG. 7 is a time chart of the applied voltage for each electrode.

At the accumulation time t1, φV1 and φV5 become 0
V, a negative voltage (for example, −8) is applied to the other electrodes except φVS.
V) is applied, and signal charges (in this case, electrons) are accumulated in the photoelectric conversion unit 23 according to the amount of received light. The holes having the same charge number generated at this time are
Is accumulated in A voltage (for example, 8 V) larger than the depletion voltage of the photoelectric conversion unit 23 is always applied to φVS, but at t1, the voltage is reduced to a voltage (for example, 4 V) that is lower than the depletion voltage of the photoelectric conversion unit. , Electron source 26
Then, electrons flow into the photoelectric conversion unit 23 as second signal charges. However, the amount is modulated by the narrow channel effect caused by the number of holes accumulated in the second photoelectric conversion unit 25.

In the reading time t2, 1 is applied to φV1 and φV5.
When 5 V is applied, the potential of the readout gate, which is a part of φV1 and φV5, decreases, and the second signal charge is added to the signal charge corresponding to the amount of received light, so that the vertical transfer region 22
Is read out. At the time t3 after the reading is completed, the voltages of φV1 and φV5 are set to 0 V, and the reading gate is closed. At time t4, φV2 is also set to 0V, the barrier is removed, and half of the signal charges are transferred to φV2. At time t5, the voltage of φV1 is set to −8 V to form a barrier, and the remaining half of the signal charge is also transferred to φV2. Subsequently, similarly, at time t6, half of the signal charges are transferred to φV3, at time t7, the rest are also transferred, and at time t8, half of the signal charges are transferred again to φV4. Thereafter, this peristalsis (charge t
The data is vertically transferred to a horizontal transfer area (not shown) while repeating the above-mentioned process.

Incidentally, the reset of the holes accumulated in the second photoelectric conversion section 25 is carried out at an appropriate time within one field time.
For example, a large positive voltage of, for example, about 20 V is applied to 1 and the Coulomb repulsion with the positive voltage causes the positive voltage to flow out to GND through a peripheral P ⁺ layer maintained at GND potential.

The above-mentioned photoelectric conversion unit 23, second photoelectric conversion unit 25
The area and the shape of the electron source 26 are calculated according to the above equation (1).
Is designed so that Mt + 1 <Mt is instantaneously repeated together with the accumulated charge amount. In other words, it is designed so that the feedback is made faster so that dt becomes as small as possible.

Therefore, according to the CCD solid-state imaging device 1 of the present embodiment, the light shot noise is reduced by using the conventionally discarded holes for the gain control of the signal charges.

Embodiment 3 This is an example in which an avalanche photodiode (APD) that also functions as a multiplying means is used in the photoelectric conversion unit.

Assuming that a Si-avalanche photodiode designed to satisfy the above equation 1 is incorporated in a CCD solid-state image pickup device as a photoelectric conversion element, an excess due to fluctuation of multiplication gain at the time of avalanche multiplication. Assuming that the noise is small, S
FIG. 8 shows the result of simulating the / N value. The initial gain was 10.

Since the actual multiplication gain is controlled by the applied voltage, it was calculated by substituting the following equation 3 into the above equation 1.

[0045]

(Equation 3)

Moff: Offset gain V: Applied voltage VB: Breakdown voltage In the figure, the dotted line is the result of this example, and the solid line is the simulation result using a normal photodiode. It is clear from FIG. 8 that the light shot noise has been reduced. However, the relational expression between the multiplication gain M and the applied voltage V substituted in Expression 1 is not limited to Expression 3, but may be any expression that satisfies (dM / dV)> 0 at M = f (V).

-Simulation of Embodiment 3 The simulation of Embodiment 3 is a result obtained on the assumption that excess noise caused by fluctuation of the multiplication gain during avalanche multiplication is small. Therefore, this excess noise will be described below, and a structural example for reducing this will be shown.

In the conventional Si-APD, a place where avalanche multiplication occurs is unspecified, so that a statistical variation occurs in the multiplication gain. The reason is that the impact ionization energy required for avalanche multiplication is (3 / 2Eg +
α), where α is indefinite according to the statistical probability. Therefore, new noise caused by this is called excess noise (excess electronic noise when the signal charge is electrons), and S /
It is considered to be a cause of N deterioration.

However, in the avalanche multiplication film having one band offset structure, when the mass of the electrons is m and the velocity of the electrons is v, the value of the band offset ΔE is increased, whereby (3 / 2Eg + α) <(1 / 2mv ² + ΔE) can always be satisfied. That is, by adopting a large band offset to a level not affected by the value of α, impact ionization can be caused only in the band offset portion, and the above-described excessive noise can be reduced as much as possible.

In the case of a structure in which a band offset type avalanche multiplication film is laminated, the state shown in FIG. 9 is taken when the applied voltage is zero. Here, the band offset ΔE is determined by the composition ratio of the material. Only when all layers are completely depleted, a band state having a slope shown in FIG. 10 is obtained for the first time, and impact ionization may occur at the position of the band offset. Such a band offset type avalanche multiplication film and a normal silicon avalanche multiplication film (Si-AP
FIG. 11 shows the difference between the gain characteristics and D). As shown in FIG. 11, the breakdown voltage V _{B in} the case of a normal Si-APD
Although the gain rises sharply in the vicinity, in the case of the band offset type, the gain becomes 2 when the applied voltage is _{Vth_H} or more, and _{Vth_L}
The gain becomes 1 in the following cases. 10, since the number of stacked avalanche multiplication layer is 3 to give the 2 ^three times gain. However, in such a simple laminated structure, the gain changes rapidly from 1 to 8 depending on the applied voltage. Therefore, when used in the charge accumulation mode, the gain does not gradually decrease from 8 → 4 → 2 → 1 along with the accumulation of the signal charge, and the effect of suppressing the optical shot noise aimed at by the present embodiment cannot be expected.

Therefore, in this embodiment, a structure in which the gain is suppressed together with the accumulated charge amount while adopting the band offset type avalanche multiplication film is adopted, thereby reducing both excess noise and optical shot noise.

FIG. 12 shows a film thickness of the photoelectric conversion device of this embodiment.
It is a direction sectional view. The photoelectric conversion device 31 includes the charge storage unit 3
3 band offset type avalanche multipliers on 2
33, 34, and 35 are laminated, and a conductive thin film 3 is interposed between the layers.
6, 37, 38, and 39, and fix the space between the thin films.
Constant resistance R₁, R_Two, R_ThreeIt has the structure connected by. Charge storage
The stacking unit 32 is a normal silicon photodiode (Si-P
D) 32. The thin film 36-39 is made of, for example, indium.
In, polysilicon, etc. Fixed resistance R₁-R_ThreeIs an example
For example, it is made of polysilicon or the like. However, a series of resistors
Is made of the same material with the same sheet resistance and changes the bonding area
The resistance ratio of each other is R₁<R_Two<R _ThreeSatisfy the conditions
It is controlled with high accuracy so as to be a predetermined value to be added.

In this structure, the initial applied voltage is V ₀ ,
The total applied voltage is V _in , the accumulated signal charge is Q, and the accumulated capacitance is C
_{When, = V in = V 0 (} - Q / C) V 1 = {R 1 / (R 1 + R 2 + R 3)} · V in V 2 = {R 2 / (R 1 + R 2 + R 3)} _{· V in V 3 = {R} 3 / (R 1 + R 2 + R 3)} · V in voltage that is applied to each layer. FIG. 13 shows the relationship between the total applied voltage Vin and the voltage applied to each layer. The slope of each straight line corresponds to the ratio of each of R ₁ , R ₂ , and R ₃ . FIG.
In, V _in the V _in = - smaller with increasing Q from (Q / C) relationship. Accordingly, the voltage V _1-3 applied to each layer also decreases. Therefore, the relationship between V _1-3 and Q is as shown in FIG. Each layer gain when the applied voltage gain is less than 2 becomes, V _{th - L} When the above V _{th - H} as described above is 1. Therefore, if this laminated film structure is used in the charge accumulation mode, gain characteristics in which the gain is suppressed in a stepwise manner with an increase in the accumulated charge amount can be obtained as shown in FIG. Therefore, when the primary photoelectric conversion efficiency is constant, the input / output characteristics as shown in FIG.

FIG. 17 is a diagram showing a configuration in which the photoelectric conversion device 31 of the present example is provided for each pixel of a two-dimensional CCD image pickup device. In the figure, _Vref is a common electrode for applying a constant voltage, VCCD is a vertical transfer area, and HCCD is a horizontal transfer area. FIG. 18 is a diagram showing a configuration in which pixels having a normal photodiode and pixels having the photoelectric conversion device of this example are alternately arranged in the vertical transfer direction.

-Third Embodiment 3-In the third embodiment 2 as shown in FIG. 15, the initial gain becomes 8 immediately. Therefore, as shown in FIG. 19, the normal S
An i-APD film 40 is preferably provided. By doing so, although there is S / N degradation due to excessive noise, FIG.
As shown by 0, a continuous initial gain can be obtained.
According to the equation of excess electronic noise of McIntyre, in the case of Si-APD, M is gain, α and β are impact ionization rates per unit distance of electrons and holes, and α / β = k. Then, S / N = M2 {kM + (1-k) (2-1 / M)}, and the S / N becomes worse as the gain is increased. Therefore, when used in combination with the band offset type as in the present embodiment, the same gain can be obtained by reducing excess electronic noise as compared with the case of using only the Si-APD. In this case, the input / output characteristics with respect to the incident light amount are as shown in FIG.

Fourth Embodiment Fourth Embodiment In the above two embodiments, it is possible to change the multiplication gain by a power of two. However, when using in the charge accumulation mode, if the initial gain is increased by increasing the number of layers, the change in the gain according to the incident light amount becomes abrupt and discontinuous, so that fine gain control is impossible. Become. For example, if the number of layers is 10, the multiplication gain is 2 ¹⁰ (= 10
From 24), it suddenly drops discontinuously to 2 ⁹ (= 512).

Therefore, as shown in FIG. 22, by further connecting and using the DSP shown in the first embodiment, the gain can be smoothly controlled from 1024 to 1023 and 1022.

The above embodiments have been described with reference to the CCD type image sensor. However, the present invention is not limited to the image pickup tube, the MOS type image sensor, or any image pickup device which will appear in the future.
Further, as described above, the present invention is, of course, also effective in an apparatus for imaging all electromagnetic waves.

[0059]

As described above, according to the present invention, light shot noise is reduced, so that a high quality image can be obtained even if the exposure time is short.

Specifically, a clear image can be captured at a high resolution even in a very dark place such as a new moon night.

[Brief description of the drawings]

FIG. 1 is a diagram showing a basic circuit configuration of a solid-state imaging device according to an embodiment of the present invention;

FIG. 2 is a block diagram illustrating a configuration of an imaging device according to a first embodiment of the present invention.

FIG. 3 is a plan view illustrating an imaging device according to a second embodiment of the present invention.

FIG. 4 is a sectional view taken along line A1-A2 of FIG. 3;

FIG. 5 is a sectional view taken along line B1-B2 of FIG. 3;

FIG. 6 is a diagram showing potentials of electrodes for each time of the imaging apparatus according to the second embodiment of the present invention.

FIG. 7 is a time chart of an applied voltage for each electrode of the imaging apparatus according to the second embodiment of the present invention.

FIG. 8 is a graph showing the number of incident photons versus the S / N value by simulation of the imaging apparatus according to the third embodiment of the present invention.

FIG. 9 is a diagram showing a band state of the band offset type avalanche multiplication film at the time of zero bias.

FIG. 10 is a diagram showing a band state of the band offset type avalanche multiplication film at the time of reverse bias;

FIG. 11 is a graph showing gain characteristics of a band offset type avalanche multiplication film and a normal Si-avalanche multiplication film with respect to an applied voltage.

FIG. 12 is a sectional view showing a second photoelectric conversion device according to a third embodiment of the present invention.

FIG. 13 is a graph showing the relationship between the voltage applied to the photoelectric conversion device according to Example 2 of the present invention and the voltage applied to each layer.

FIG. 14 is a graph showing the relationship between the voltage applied to each layer of the photoelectric conversion device according to Example 2 of the present invention and the amount of accumulated charges.

FIG. 15 is a graph illustrating the relationship between the number of accumulated charges and the gain in the photoelectric conversion device according to the third embodiment of the present invention.

FIG. 16 is a graph showing the relationship between the number of incident photons and the number of accumulated charges in the second photoelectric conversion device according to the third embodiment of the present invention.

FIG. 17 is a diagram illustrating a configuration in which the second photoelectric conversion device according to the third embodiment of the present invention is applied to a two-dimensional CCD imaging device.

FIG. 18 is a diagram showing a configuration in which the two photoelectric conversion devices according to the third embodiment of the present invention and a conventional photoelectric conversion device are used in combination with a two-dimensional CCD imaging device.

FIG. 19 is a sectional view showing a third photoelectric conversion device according to a third embodiment of the present invention.

FIG. 20 is a graph showing the relationship between the number of accumulated charges and the gain in the third photoelectric conversion device according to the third embodiment of the present invention.

FIG. 21 is a graph showing the relationship between the number of incident photons and the number of accumulated charges in the third photoelectric conversion device according to the third embodiment of the present invention.

FIG. 22 is a diagram showing a configuration in which a two-dimensional CCD imaging device incorporating the photoelectric conversion device according to the third embodiment of the present invention is combined with the DSP according to the first embodiment.

FIG. 23 is a diagram illustrating the principle of the present invention.

FIG. 24 is a diagram illustrating the principle of the present invention.

[Explanation of symbols]

1, 23 photoelectric conversion unit 2 multiplication (amplification) means 3 memory 4 gain control unit 11, 21 solid-state imaging device 12 CCD 13 line memory 14 operation unit 15 DSP 22 vertical transfer area 24 separation unit 26 charge (electron) source 27 insulation Layers 31, 41 Photoelectric conversion devices 32, 42 Charge storage units 33, 34, 35 Band offset type avalanche multiplication films 36, 37, 38, 39 Conductive thin films 40 Si-avalanche multiplication films φV1 to φV5, φVS electrodes

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H04N 5/30-5/335

Claims (18)

(57) [Claims] 1. A photoelectric device that generates a signal charge according to incident light.
A conversion unit, and the photoelectric conversion unit
Amplifying section for amplifying the signal charge with a predetermined gain;
Resets the signal charge amplified by the
At a predetermined time (t)
The predetermined gain of the amplifying unit at the predetermined time
Depending on the signal charge amount stored in the memory by (t)
And a sequential gain control section for controlling the photoelectric conversion. 2. A first photoelectric conversion unit for generating electric charges in accordance with incident light, a storage unit for storing electric charges generated in the first photoelectric conversion unit, and a charge output from the storage unit for a predetermined period. an operational amplifier for amplifying at a gain, and a storage unit for memory the amplified signal charges by the operational amplifying unit, and the predetermined gain of the operational amplifying unit is definitive in one field period, its gain A photoelectric conversion device characterized in that the photoelectric conversion device is controlled on the basis of the amount of charge stored in the storage unit during a predetermined period before the time point (t) to be applied. 3. Photoelectric conversion for generating electric charge according to incident light
And A / D conversion of charges generated in the photoelectric conversion unit
An A / D converter, and storing signal charges from the A / D converter.
A line memory to be stacked and output from the line memory.
Calculation unit that amplifies the accumulated charge with a predetermined gain, and the calculation unit
And a frame memory for storing the amplified signal charges.
The predetermined gain of the arithmetic unit is during one field period.
Before the time (t) at which the gain is applied
During the period, the stored power stored in the frame memory
Photoelectric conversion characterized by being controlled based on the load
apparatus. 4. An A / D converter for converting an analog signal output from the storage unit into a digital signal, wherein the amplification unit and the storage unit perform digital processing. Photoelectric conversion device. 5. The number of stored charges previously stored is determined by the gain M changing and the number of stored charges n per unit time.
Product The photoelectric conversion apparatus according to claim 2 or 4, wherein it is a function of the integrated value in the storage time until it with. 6. An electric power supply according to claim 1 , wherein said electric power supply is arranged in a light receiving section.
A first photoelectric conversion unit that generates a load, and the first photoelectric conversion unit
A storage section for storing signal charges from the section, a negative feedback amplification section,
Wherein the negative feedback amplification unit includes the first photoelectric conversion unit and
Is a reverse conductivity type, and is connected to the first photoelectric conversion unit in the light receiving unit.
The second, which is provided adjacently and generates charges in accordance with incident light
And the same conductivity type as the first photoelectric conversion unit.
And a charge source for supplying a charge to the storage unit.
During one field period, the second photoelectric conversion unit
The signal charge generated or accumulated in
Note that the signal charge flowing out of the charge source is modulated.
Photoelectric conversion device. 7. The modulation means that the supply of charges from the charge source to the storage unit decreases when the signal charges stored in the storage unit are large, and the signal charges stored in the storage unit are small. 7. The photoelectric conversion device according to claim 6 , wherein the supply of the charge from the charge source to the storage unit increases. 8. Photoelectric conversion for generating electric charge according to incident light
Unit and accumulates the signal charges accumulated in the photoelectric conversion unit
A portion that includes a storage unit and performs photoelectric conversion of the photoelectric conversion unit.
Is an avalanche photodiode.
Shed photodiodes are connected to each other with fixed resistors
Via a plurality of conductive thin films, laminated on the storage unit
Multiple band offset avalanche multiplication (multipli
cation layer), and these avalanche multiplication films
Each composition changes based on the lamination direction,
During one field period, the data stored in the storage unit
Each avalanche multiplication according to the number of accumulated charges
A photoelectric conversion device characterized in that the degree of multiplication of a film changes.
Place. 9. The composition refers to (amorphous silicon a-Si) / (amorphous silicon carbon a-Si)
C) a or (amorphous silicon a-Si) / (amorphous silicon germanium a-SiGe), photoelectric of the ratio of the carbon C and silicon Si is claimed claim 8, characterized in that sequentially decreasing or increasing Conversion device. 10. The photoelectric conversion device according to claim 8 , wherein the resistance value of the fixed resistor increases as it goes away from the incident light side. 11. The photoelectric conversion device according to claim 8 , wherein said fixed resistor is made of polysilicon. 12. A silicon-avalanche multiplication film is interposed between the plurality of band-offset avalanche multiplication films and the storage section.
12. The photoelectric conversion device according to any one of 8 to 11 . 13. The photoelectric conversion device according to claim 6 , wherein the first photoelectric conversion unit and the storage unit are the photoelectric conversion device according to any one of claims 6 to 12. Photoelectric conversion device. 14. A photoelectric conversion device according to any one of claims 6 to 13, the solid-state imaging device characterized by comprising a transfer region for transferring the photoelectric signal charges accumulated. 15. A photoelectric conversion device that generates an electric charge according to incident light.
Conversion unit and an accumulation unit that accumulates charges generated in the photoelectric conversion unit.
Unit and the signal charge accumulated in the accumulation unit
Is an A / D converter that digitizes every non-destructive readout
And each signal charge sequentially output from the A / D converter
About a predetermined number of successive signal charges.
And their values are the maximum signal charge and the minimum signal charge.
A photoelectric conversion device characterized by being abandoned . 16. The photoelectric conversion device according to claim 15, wherein said predetermined number of successive signal charges are sequentially selected without overlapping at all with previous successive signal charges. 17. The photoelectric conversion device according to claim 15, wherein said predetermined number of continuous signal charges are sequentially selected so as to partially overlap each of the preceding continuous signal charges. . 18. The light according to any one of claims 15 to 17,
A power conversion device, and amplifies the electric charge output from the storage unit with a predetermined gain.
Amplifying section and memory for signal charges amplified by the amplifying section.
Storage unit , wherein the predetermined gain of the amplification unit is within one field period.
At a specified time period before the gain is applied.
Change based on the number of stored charges stored in the storage unit
A photoelectric conversion device characterized by being modified.