CMOS image sensors are known.
An active pixel sensor is disclosed in U.S. Pat. No. 6,744,068. In an active pixel sensor of this type, charge is accumulated in a photoreceptors, e.g., the area under a photogate. The charge is transferred from the photogate to a floating diffusion node, and to a source follower output amplifier. The output amplifier drives a voltage signal across a signal line e.g. a column line.
In a column parallel device, rows of pixels are read out sequentially as the pixels in a row drive their respective column lines. This is for example described in U.S. Pat. No. 7,369,166. After row has been completely read-out, some other row can be selected for reset. The row that is selected for reset is in general not the next row in the read-out sequence. Therefore, the reset row has a shortened integration time relative to the frame as a whole.
The cycle may continue with the next row being read out. When the next row is read and the process continues, the result is a so-called rolling shutter effect.
The rolling shutter can cause imaging artifacts. For example, it can distort objects relative motion within the image frame, and cause a “smearing” effect.
A snapshot pixel image sensor integrates all pixels simultaneously for a period less than the frame period. Several different techniques have been proposed for implementing such a snapshot sensor, for example see U.S. Pat. Nos. 6,486,503 and 7,388,239. In the case of '503, multiple high efficiency charge transfers are required. '239 introduces kTC noise into the readout of the charge storage element. This allows all pixels to be integrated simultaneously for a period less than the frame readout.
The present application reduces a need for multiple high efficiency charge transfers and still maintains the suppression of kTC noise in a pixel sensor that can integrate all pixels simultaneously for a period less than the complete frame period.
In the Drawings:
An embodiment is shown in
The node 121 of floating diffusion 120 is connected to the base of the source follower transistor 125 which buffers the output signal. A pixel select transistor 130 uses a select line which is brought active to transfer the voltage from the source follower 125 to the read-out line 135.
A resettable column amplifier 140 drives a sample and hold circuit 145.
The embodiments use feedback to reset the charge on the floating diffusion to a nominal reset level. The use of this feedback forms a portion of the feedback loop, operating to reduce the KTc noise, since the reset is based on the amount of sampled charge.
In operation, signal photocarriers are integrated under the photogate PG. At the end of the integration period, the carriers are transferred by the transfer gate by pulsing the transfer gate high, or reducing the voltage on the photogate terminal.
The photo generated carriers are transferred to the floating diffusion 120. The time when this happens, called T1, marks the beginning of a frame cycle. The photogate and transfer data are then returned to their integrating levels, to again receive photo generated charge. While the photogate is integrating the charge from a next cycle (or ready to integrate), the charge from the previous cycle is held in the floating diffusion. The source follower SF 125 buffers the signal. The voltage indicative of the value in FD is presented at its output node 126.
The photogenerated charge is sampled from the floating diffusion by selecting a row or pixel of the sensor using the SEL signal.
A charge mode trans-impedance amplifier 175 can be shared among a number of different pixels. This amplifier outputs an amount of charge proportional to its input voltage from the readout of the pixel value. Each pixel is activated to use the trans-impedance amplifier by the select signal.
The read value is applied to amplifier 140, the other input of which receives the reset value Vrst. The output of the amplifier is a feedback voltage which is driven through the rest FET 155 by selecting the reset level RST.
This feedback value causes the output of the source follower transistor 125 to become equal to the reset level Vrst. The amount of charge that is required to maintain the output of the source follower at that level is integrated on the feedback capacitor 160. When RST is released, this charge level appears as a change in voltage at the output of the amplifier 140, and is then sampled and held by sample and hold circuit 145 as the readout value for the pixel. However, the floating diffusion has been consistently reset to the reset level Vrst, so that each floating diffusion is always actively reset to this value. This prevents stray charge in the components from causing KTc noise.
The feedback capacitor 160 can then be reset by a switch 165. This can zero out any voltage across the feedback resistor.
Well known techniques can be used to reduce the fixed pattern noise due to threshold variations in source follower. For example, this can be removed by digital techniques or off chip. Another embodiment may use a switched capacitor to null the column wise amplifier outputs
According to another embodiment in
An alternative embodiment shown in
The floating diffusion storage node 120 is protected against incident photons by a light shield 170 which can be placed over all areas which might receive light. However, the storage node may still be susceptible to carriers which are generated elsewhere that diffuse to its vicinity. Hence, to further protect the floating node 120, a sub collector n region 400 as shown in
In another embodiment, a heavier p level doping area 300 can be used that will create an energy barrier 305 to reflect the errant photo carriers. The barrier should typically be larger than KT/q. In another embodiment, both of these techniques may be used together as shown in
Although only a few embodiments have been disclosed in detail above, other embodiments are possible and the inventors intend these to be encompassed within this specification. The specification describes specific examples to accomplish a more general goal that may be accomplished in another way. This disclosure is intended to be exemplary, and the claims are intended to cover any modification or alternative which might be predictable to a person having ordinary skill in the art. For example, other components can be used. While the above describes operation in charge mode, it should understood that this system can operate in current mode or any other mode. Also, while this describes MOS pixel sensors, these techniques can be used with other sensors.
Another embodiment may use similar techniques to those described herein, along with backside illumination.
Any of the embodiments described herein may reverse the doping types relative to the types described.
These embodiments may be located in a semiconductor substrate, whose top or bottom surface receives the illumination. That illumination receiving surface may also have color filter arrays, to allow full color operation, and may have microlenses and other optical parts.
In addition to the circuits described herein, this integrated circuit may include many other conventional parts, including A/D converters and noise reduction parts (such as correlated double sampling circuits), latches, counters, decoders, row and/or column drivers, and signal conditioning circuits.
Also, the inventors intend that only those claims which use the words “means for” are intended to be interpreted under 35 USC 112, sixth paragraph. Moreover, no limitations from the specification are intended to be read into any claims, unless those limitations are expressly included in the claims.
This application claims priority from provisional No. 61/145,690, filed Jan. 19, 2009, the disclosure of which is herewith incorporated by reference.
Number | Date | Country | |
---|---|---|---|
61145690 | Jan 2009 | US |