Shared pixel with fixed conversion gain

Abstract

An active pixel sensor comprising: a plurality of pixels, wherein each pixel includes a light sensitive element and a transfer gate, and the plurality of pixels have at least one floating diffusion region; and a plurality of processing circuits associated with the plurality of pixels; wherein each processing circuit comprises a charge amplifier.

Description

TECHNICAL FIELD

The present disclosure relates to image sensors, and more particularly, to a 3D stacked active pixel sensor with shared pixels.

BACKGROUND

An active pixel sensor (APS) combines a photodiode with processing and amplification circuitry in each pixel to form an image sensor. In some implementations, the associated circuitry is implemented with four transistors, and may be referred to as a 4 T APS. Such sensors are commonly found in devices such as cell phone cameras, digital cameras, and web cameras.

Smaller APS and smaller pixels are enabled by the improvements in semiconductor processing. A consequence of smaller pixel size is that the pixel's photodiode must also shrink to accommodate the smaller APS area. Less light is then able to fall on each pixel, and as a result image quality is reduced. Conventionally, as pixel size is getting smaller, a plurality of pixels share transistor(s) so to increase fill factor, which represents the percentage of the pixel area that is consumed by the actual photodiode, and is a higher-is-better (HB) metric.

However, when a plurality of pixels share a transistor, each pixel contributes additional parasitic capacitance to the pixel sensor. The higher capacitance of the pixel sensor results in a lower conversion gain (CG), which is undesirable.

It is desirable to maintain conversion gain when more pixels share electrical circuitry. Further, it is desired to have high CG, especially for high sensitivity under low illumination conditions.

SUMMARY OF THE INVENTION

In an embodiment, an active pixel sensor comprises a plurality of pixels and a plurality of processing circuit. Each pixel includes a light sensitive element and a transfer gate. The plurality of pixels include at least one floating diffusion region. The plurality of processing circuits are associated with the plurality of pixels. Each processing circuit comprises a charge amplifier.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention is illustrated in an exemplary manner by the accompanying drawings. The drawings should be understood as exemplary rather than limiting, as the scope of the invention is defined by the claims.

FIG. 1 is a plan view of a sensor die included in an image sensor having two stacked semiconductor dies according to an embodiment of the invention.

FIG. 2 is a cross-sectional view of two stacked semiconductor dies forming an image sensor of FIG. 1.

FIG. 3 is a circuit diagram illustrating an image sensor according to an embodiment of the invention.

FIG. 4 is a circuit diagram illustrating an image sensor according to an embodiment of the invention.

FIG. 5 is a diagram illustrating the timing sequence of the first reset signal and the second reset signal.

FIG. 6 is a circuit diagram illustrating an image sensor according to an embodiment of the invention.

FIG. 7 is a circuit diagram showing a high-level structure including the image sensor according to an embodiment of the invention.

FIG. 8 is a circuit diagram showing a high-level structure including the image sensor according to another embodiment of the invention.

DETAILED DESCRIPTION

As will be discussed below, the disclosed embodiments allow for larger light sensitive elements in spite of shrinking image sensor circuit features through the use of a stacked configuration of sensor and circuit dies. Further, some of the disclosed embodiments achieve a stacked image sensor without requiring a floating diffusion region on the circuit die. Floating diffusion region is a combination of diffusion areas like drain of transfer transistor, source of reset transistor and parasitic capacitors associated therewith.

FIG. 1 is a plan view of a sensor die portion of an image sensor having two stacked semiconductor dies according to an embodiment of the invention. Image sensor 100 is implemented as an active pixel sensor (APS), such as, for example, a Complementary Metal-Oxide-Semiconductor Transistor (CMOS) image sensor. Image sensor 100 includes pixel regions 102 arranged in an array of rows and columns. Image sensor 100 can have any number of pixels regions, such as, for example, 1280 columns by 960 rows of pixel regions.

FIG. 2 is a cross-sectional view of two stacked semiconductor dies forming an image sensor of FIG. 1 along the section line Z-Z′. Image sensor 100 includes a sensor die 202 and a circuit die 204. In this view, it can be seen that the circuit die 204 is communicatively coupled to the sensor die 202. For example, the circuit die 204 may underlie the sensor die 202.

FIG. 3 is a circuit diagram illustrating an image sensor according to an embodiment of the invention. An image sensor 30 utilizes a structure features that a plurality of pixels share a processing and amplification circuits 300 shown in the dashed line. As shown in FIG. 3, the image sensor 30 includes a first photodiode PD1, a first transfer gate M_Tx1, a second photodiode PD2, a second transfer gate M_Tx2, a third photodiode PD3, a third transfer gate M_Tx3, a fourth photodiode PD4, and a fourth transfer gate M_Tx4. Although only four pixels PD1, PD2, PD3 and PD4 are shown in FIG. 3, an image sensor 30 can include additional pixels in an embodiment in accordance with the invention. The image sensor 30 further includes the processing and amplification circuit 300. The processing and amplification circuit 300 comprises an amplifier AMP, a first capacitor CF, and a reset gate M_RST. The processing and amplification circuit 300 can also be called a charge amplifier, as the processing and amplification circuit 300 act a charge-to-voltage converter. Both the first reset gate M_RST and the first capacitor CF are connected in parallel between an input INN of the amplifier AMP and an output Vout of the amplifier AMP. A reference signal is fed to a positive input INP of the amplifier AMP. REF represents a signal bus and is connected to a reference generator circuit in peripheral.

For examples, we have an array size 400×100=40K pixels. There will be 40K micro VIAs just for pixels. If 4 pixels share an amplifier AMP for example, the total micro VIAs to connect pixels will be 40,000/4=10K. Therefore shared pixel reduces number of micro VIAs.

However shared pixel causes higher parasitic capacitance in floating node, which translates to lower conversion gain. For example, transfer gate has drain area (diffusion) in floating node. There is parasitic capacitance between transfer gate poly gate and drain diffusion. In another example, each diffusion area has capacitance. Diffusion area cannot be shares when multiple pixels share floating area. The additional diffusion increases parasitic capacitance. The connection wires between multiple diffusion areas add more parasitic capacitance. Conversion Gain (CG) is a measure of voltage changes on floating node in response to the charge transferred from photoelectric converting device.

Lower conversion gain means that higher gain is needed to achieve required signal at same lighting condition, or higher noise. For example, amplifier AMP will generate noise, say “N_AMP.” We use CG=100 uV/e⁻ and 200 uV/e⁻ as examples. The signal is 1000 electrons and an output signal 400 mV is desired. When CG=100 μV/e⁻, signal at the output bus Vout=100×10⁻³×1000+N_AMP=100+N_AMP (mV). Additional 4 times gain is needed to achieve 400 mV signal. Signal=4×(100+N_AMP)=400+4×N_AMP with noise “4×N_AMP”. When CG=200 μV/e⁻, signal at the output bus Vout=200×10⁻³×1000+N_AMP=200+N_AMP (mV). Additional 2 times gain is needed to achieve 400 mV signal. Signal=2×(200+N_AMP)=400+2×N_AMP with noise “2×N_AMP”. Assume 2× or 4× amplifier (not show in FIG. 3) does not introduce additional noise, it can be seen that under same situation except different conversion gains, in order to achieve the same output signal, the lower the conversion gain, the higher the noise will be generated.

Note that although FIG. 3 only shows one processing and amplification circuit 300 in the dashed block, those having ordinary skill in the art can understand that in actual application, an image sensor comprises a plurality of processing and amplification circuits.

Floating diffusion node FD is virtual ground if the amplify AMP is ideal, that is if the amplifier gain is infinite, which makes the voltage difference between the positive and negative inputs zero. The floating diffusion node FD will keep at constant voltage REF (the same as the voltage of negative input INP of the amplifier AMP) in this case. Note that in FIG. 3 the amplifier AMP has differential inputs INN and INP, those having ordinary skill in the art can understand that the amplifier may include a single-ended input instead of differential input ports. In the single-ended input scenario, the floating diffusion node FD is connected to the input of the amplifier. The floating node parasitic capacitance C_FD will have no impact if the voltage doesn't change. When charge is transferred from photodiode PD to the floating diffusion region FD, and FD voltage is constant (i.e., charge maintains constant in parasitic capacitors), the equal number of opposite charge must come from somewhere else to the floating diffusion region FD. Those opposite charges come from the first capacitor CF. Conversion gain means how much voltage will be with one electron. So conversion gain can be represented as

$\mathrm{CG}=\frac{q}{C_F}=\frac{1.6\times {10}^{-19}}{C_F}.$ Therefore, the conversion gain for this embodiment can be represented as

$\mathrm{CG}=\frac{e^{–}}{C_F}.$ Wherein e⁻ represents an electron, and C_F represents the capacitance of first capacitor CF.

FIG. 4 is a circuit diagram illustrating another embodiment of the invention. In addition to the first capacitor CF, the first reset gate M_RST1, the processing and amplification circuit 400 further comprises a second capacitor C2, a third capacitor C1 and a second reset gate M_RST2, wherein the first reset gate M_RST1 is connected between the input of the amplifier AMP and the output of the amplifier AMP, the second capacitor C2 is connected in series with the first capacitor CF between the negative input INN of the amplifier AMP and the output Vout of the amplifier AMP, the second reset gate M_RST2 is connected in parallel with the second capacitor C2, and the third capacitor C1 is connected between a connecting point A of the first capacitor CF and the second capacitor C2 and a fixed reference voltage. The fixed reference voltage may comprise the ground GND as shown in FIG. 4, or a VDD. The conversion gain in this scenario will be discussed in further details with reference to the following FIG. 5.

FIG. 5 is a diagram illustrating the timing sequence of the first reset signal and the second reset signal. As shown in FIG. 5, at time T, both RST1 and RST2 are off. Reset finishes, and assume no charge injections are generated from the two reset transistors M_RST1 and M_RST2, then in conjunction with the circuit diagram of FIG. 4,V(FD)=REF, V(A)=REF, V(Vout)=REF,wherein REF represents the voltage of the reference signal bus.Q(FD)=C_FD×REF, Q(A)=C₁×REFThen, assume a charge “q” is transferred to floating node, thenV(FD)=REF, V(A)=VA, V(Vout)=VoutAnd there are constant charge numbers in both nodes “FD” and “A,” thenQ(FD)=C_FD×REF+q+C_F×(REF−VA)=C_FD×REFQ(A)=C_F×(VA−REF)+C₂×(VA−Vout)+C₁×VA=C₁×REFFrom the two equations above, we have

$\mathrm{Vout}=\mathrm{REF}+\frac{q}{C_F}\times \frac{C_1+C_2+C_F}{C_2}.$ Therefore, for the circuit shown in FIG. 4, the conversion gain CG can be represented as

$\mathrm{CG}=\frac{e^{–}}{C_F}\times \frac{C_F+C_1+C_2}{C_2}.$

With the above embodiment, Photo Response Non Uniformity (PRNU) can be improved. Basically, the signal should be same if same amount of photons hit a pixel. The variation of each pixel in response to the photon is called PRNU. The matching of CF becomes worse when CF is smaller, then the variation of conversion gain is bigger, i.e., PRNU is getting worse. There are many processing and amplification circuits 400 in pixel array. CF in every processing and amplification circuit 400 will be slightly different due to manufacturing variations. The smaller the CF, the higher percentage of the difference. For example, CF depends on the area of CF. Assume CF size is 10×10 and process variation (in each direction) is 1, the CF area ranges from 9×9 to 11×11. Max/Min==11²/9²=1.49. Alternatively, when CF is increased to 100×100 with same process variation, CF area ranges from 99×99 to 101×101. Max/Min=101²/99²=1.04. Increased CF improves PRNU.

Further,

$\frac{C_1}{C_2}$ ratio can be adjusted to meet the CG requirement. The

$\frac{C_1}{C_2}$ ratio can be adjusted based on process variations. For example, the minimum area to achieve reasonable uniformity is 100. CG is 10 when use area 100 capacitor using circuit in 300 of FIG. 3. In order to achieve CG=100, CF(area) needs to be 10. Area 10 is too small that will cause large non uniformity. Therefore in order to achieve CG=100 with reasonable uniformity, an effective capacitance is needed which is 10 times smaller which uses 400 of FIG. 4. Therefore, CF(area)=100, and C2(area)=100 are picked. As

$\mathrm{CG}=\frac{e^{–}}{C_F}\times \frac{C_F+C_1+C_2}{C_2},\begin{array}{c}{\mathrm{CF}}^{\prime }\left(\mathrm{area}\right)=\frac{C_F\times C_2}{C_F+C_1+C_2}\\ =\frac{100\times 100}{100+100+C_1}\\ =10,\end{array}$ therefore C1=800.

Therefore in FIG. 4, when CF(area)=100, C1(area)=800, C2(area)=100, CG=100. Since area of each capacitor is above 100, the uniformity is better.

Alternatively or in addition, the floating diffusion region is covered with metal layer. Assume many pixels share one amplifier. Then all drain nodes of transfer gates are connected together. In the case we have a very strong light hit one pixel (including the drain node of transfer gate), the entire floating node bus is constantly exposed. It is like pixels connected to this floating node bus are all very bright, which is called smear. Those of ordinary skill in the art can understand that when the floating diffusion region is covered with metal layer, which means that each floating node is covered with metal light shielding, the smear effect can be eliminated or alleviated.

FIG. 6 is a circuit diagram illustrating an image sensor according to an embodiment of the invention. In FIG. 6, same reference signs represent same elements in the active pixel sensor 40 shown in FIG. 4 the detail descriptions of which are omitted. The dotted line 66 represents a die border between two dies 62 and 64. As shown in FIG. 6, all of the reset signal bus RST and the transfer signal buses TX1, TX2, TX3 and TX4 are located in the circuit die 64, so as to save space on the sensor die 62 for accommodating a larger photo detector on the sensor die 62. In other words, the transfer nodes TXN1, TXN2, TXN3 and TXN4 on the sensor die 62 each is interconnected to its corresponding transfer signal bus TX1, TX2, TX3 and TX4 in the circuit die 64. The transfer gates M_TX1, M_TX2, M_TX3 and M_TX4 are in the same die as the photodiodes PD1, PD2, PD3 and PD4 so to avoid contact requirements. The photodiodes PD1, PD2, PD3 and PD4 each is surrounded with pure silicon crystal so to avoid defect which may lead to dark current, or defect pixel, etc. The ground bus in FIG. 6 is located in the circuit die 64. There is ground connection between the sensor die 62 and the circuit die 64.

FIG. 7 is a circuit diagram showing a high-level structure including the image sensor according to an embodiment of the invention. In FIG. 7, each charge amplifier shown as a block is followed by an analog-to-digital converter (ADC) and there is no need for a separate switch.

FIG. 8 is a circuit diagram showing a high-level structure including the image sensor according to another embodiment of the invention. In FIG. 8, each charge amplifier shown as a block is followed by a switch. The switch can be implemented by a N-channel metal-oxide-semiconductor field-effect transistor (NMOS).

Features and aspects of various embodiments may be integrated into other embodiments, and embodiments illustrated in this document may be implemented without all of the features or aspects illustrated or described. One skilled in the art will appreciate that although specific examples and embodiments of the system and methods have been described for purposes of illustration, various modifications can be made without deviating from the spirit and scope of the present invention. For example, embodiments of the present invention may be applied to image sensors having different types of light sensing devices, such as photodiodes, photogates, pinned photodiodes, and equivalents. Moreover, features of one embodiment may be incorporated into other embodiments, even where those features are not described together in a single embodiment within the present document. Accordingly, the invention is described by the appended claims.

Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. Even if certain features are recited in different dependent claims, the present invention also relates to an embodiment comprising these features in common. Any reference signs in the claims should not be construed as limiting the scope.

Note that any and all of the embodiments described above can be combined with each other, except to the extent that it may be stated otherwise above or to the extent that any such embodiments might be mutually exclusive in function and/or structure.

Although the present invention has been described with reference to specific exemplary embodiments, it will be recognized that the invention is not limited to the embodiments described, but can be practiced with modification and alteration within the spirit and scope of the appended claims. Accordingly, the specification and drawings are to be regarded in an illustrative sense rather than a restrictive sense.

Claims

1. An active pixel sensor comprising: a plurality of pixels, wherein each pixel includes a light sensitive element and a transfer gate, and the plurality of pixels have at least one floating diffusion region; anda plurality of processing circuits associated with the plurality of pixels;wherein each processing circuit comprises a charge amplifier, wherein the charge amplifier comprises a feedback path, the charge amplifier is used for conversion gain, and the conversion gain is determined by the ratio of two capacitors in the feedback path, wherein the conversion gain equals 100 uV/e−,wherein the charge amplifier comprises an amplifier, a first reset gate, a first capacitor, a second capacitor, a third capacitor and a second reset gate, wherein the second capacitor is connected in series with the first capacitor between the input of the amplifier and the output of the amplifier, the first reset gate is connected between the input of the amplifier and the output of the amplifier, the second reset gate is connected in parallel with the second capacitor, and the third capacitor is connected between a connecting point of the first capacitor and the second capacitor and a fixed reference voltage,wherein the fixed reference voltage includes a power supply.

2. The active pixel sensor of claim 1, wherein a processing and amplification circuit is shared by an amount of pixels.

3. The active pixel sensor of claim 2, wherein the processing and amplification circuit and the amount of pixels are within a die.

4. The active pixel sensor of claim 2, wherein the amount of pixels are in a sensor die, and the processing and amplification circuit is in a circuit die different from the sensor die.

5. The active pixel sensor of claim 4, wherein a ground of the sensor die is connected to a ground signal bus located in the circuit die.

6. The active pixel sensor of claim 1, wherein the floating diffusion region is covered with metal layer.

7. The active pixel sensor of claim 1, wherein the light sensitive element comprises a photodiode.

8. The active pixel sensor of claim 1, wherein the plurality of pixels share a floating diffusion region.