U.S. patent application number 12/959327 was filed with the patent office on 2011-03-24 for two-dimensional time delay integration visible cmos image sensor.
Invention is credited to Stefan Lauxtermann.
Application Number | 20110068382 12/959327 |
Document ID | / |
Family ID | 39740761 |
Filed Date | 2011-03-24 |
United States Patent
Application |
20110068382 |
Kind Code |
A1 |
Lauxtermann; Stefan |
March 24, 2011 |
TWO-DIMENSIONAL TIME DELAY INTEGRATION VISIBLE CMOS IMAGE
SENSOR
Abstract
A two dimensional time delay integration CMOS image sensor
having a plurality of pinned photodiodes, each pinned photodiode
collects a charge when light strikes the pinned photodiode, a
plurality of electrodes separating the plurality of pinned
photodiodes, the plurality of electrodes are configured for two
dimensional charge transport between two adjacent pinned
photodiodes, and a plurality of readout nodes connected to the
plurality of pinned photodiodes via address lines.
Inventors: |
Lauxtermann; Stefan;
(Camarillo, CA) |
Family ID: |
39740761 |
Appl. No.: |
12/959327 |
Filed: |
December 2, 2010 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11683811 |
Mar 8, 2007 |
|
|
|
12959327 |
|
|
|
|
Current U.S.
Class: |
257/292 ;
257/E31.001; 257/E31.083; 438/73; 438/90 |
Current CPC
Class: |
H04N 5/37206 20130101;
H04N 5/3745 20130101; H04N 5/3743 20130101; H01L 27/14609 20130101;
H01L 27/14603 20130101 |
Class at
Publication: |
257/292 ; 438/73;
438/90; 257/E31.083; 257/E31.001 |
International
Class: |
H01L 31/113 20060101
H01L031/113; H01L 31/18 20060101 H01L031/18 |
Claims
1. An image sensor comprising: first and second regions, the first
region located adjacent to the second region; a first photodiode
located adjacent to the first region; a second photodiode located
adjacent to the second region; a first transfer gate positioned
above the first region, and configured to transfer a charge from
the first photodiode to the first region; and a second transfer
gate positioned above the second region, configured to receive the
charge in the second region, and configured to transfer the charge
from the second region to the second photodiode.
2. The image sensor of claim 1, wherein the first and second
regions are substantially free of any floating diffusion region or
drain region.
3. The image sensor of claim 1, further comprising: a first readout
node located outside the first and second regions, and configured
to collect a first charge from the first photodiode.
4. The image sensor of claim 3, further comprising: a third
transfer gate coupled between the first readout node and the first
photodiode, and configured to facilitate the first node to collect
the first charge from the first photodiode.
5. The image sensor of claim 1, further comprising: a second
readout node located outside the first and second regions, and
configured to collect a second charge from the second
photodiode.
6. The image sensor of claim 5, further comprising: a fourth
transfer gate coupled between the second readout node and the
second photodiode, and configured to facilitate the second node to
collect the second charge from the second photodiode.
7. The image sensor of claim 1, wherein the first transfer gate is
located directly adjacent to the second transfer gate.
8. The image sensor of claim 1, wherein: the first photodiode has a
first Fermi level, the second photodiode has a second Fermi level
substantially the same as the first Fermi level, the first region
has a first region Fermi level substantially higher than the first
Fermi level, and the second region has a second region Fermi level
substantially the same as the first region Fermi level.
9. The image sensor of claim 8, wherein: the first transfer gate
creates a first well in the first region, such that the first well
has a first quasi-Fermi level substantially lower than the first
Fermi level when the charge is transferred from the first
photodiode to the first well, and the second transfer gate creates
a second well in the first region, such that the second well has a
second quasi-Fermi level substantially similar to the first
quasi-Fermi level when the charge is transferred from the first
well to the second well.
10. The image sensor of claim 9, wherein: the first transfer gate
restores the first region Fermi level in the first region after the
charge is transferred from the first region to the second region,
and the second transfer gate restores the second region Fermi level
in the second region to transfer the charge from the second region
to the second photodiode.
11. A method for transferring a charge between a first photodiode
and a second photodiode, comprising the steps of: creating a first
well adjacent to the first photodiode during a first time period;
and creating a second well adjacent to the first well and the
second photodiode during a second time period partially but not
entirely overlapping with the first time period.
12. The method of claim 11, wherein the creating the first well
step includes: adjusting, in a first region adjacent to the first
photodiode, a first quasi-Fermi level substantially lower than a
first Fermi level of the first photodiode.
13. The method of claim 12, wherein the creating the second well
step includes: adjusting, in a second region adjacent to the second
photodiode and the first region, a second quasi-Fermi level
substantially the same as the first quasi-Fermi level and
substantially lower than a second Fermi level of the second
photodiode.
14. The method of claim 11, further comprising the steps of:
collapsing the first well after the first time period; and
collapsing the second well after the second time period.
15. The method of claim 14, wherein: the collapsing the first well
step includes restoring, in a first region adjacent to the first
photodiode, a first region Fermi level substantially higher than a
first Fermi level of the first photodiode, and the collapsing the
second well step includes restoring, in a second region adjacent to
the second photodiode and the first region, a second region Fermi
level substantially higher than a second Fermi level of the second
photodiode.
16. The method of claim 11, wherein: the first time period has a
first portion distinct from the second time period, and the second
time period has a second portion distinct from the first time
period.
17. A method for transferring a charge between a first photodiode
and a second photodiode, comprising the steps of: applying a first
voltage to a first electrode to adjust a first quasi-Fermi level of
a first region located adjacent to the first photodiode during
first and second cycles; and applying a second voltage to a second
electrode to adjust a second quasi-Fermi level of a second region
located adjacent to the first region and the second photodiode
electrode during the second cycle and a third cycle.
18. The method of claim 17, wherein: the first quasi-Fermi level is
substantially lower than a first Fermi level of the first
photodiode, and the second quasi-Fermi level is substantially the
same as the first quasi-Fermi level and substantially lower than a
second Fermi level of the second photodiode.
19. The method of claim 17, further comprising the steps of:
floating the first electrode to restore a first region Fermi level
of the first region after the second cycle; and floating the second
electrode to restore a second region Fermi level of the second
region after the third cycle.
20. The method of claim 19, wherein: the first region Fermi level
of the first region is substantially higher than the first Fermi
level of the first photodiode, and the second region Fermi level of
the second region is substantially higher than the second Fermi
level of the second photodiode.
Description
CLAIM OF PRIORITY UNDER 35 U.S.C. .sctn.120
[0001] This application is a continuation of and claims the benefit
and priority of U.S. application Ser. No. 11/683,811, entitled
"TWO-DIMENSIONAL TIME DELAY INTEGRATION VISIBLE CMOS IMAGE SENSOR,"
filed on Aug. 3, 2007, which is assigned to the assignee hereof and
hereby expressly incorporated by reference herein.
BACKGROUND OF THE INVENTION
[0002] 1. Field of the Invention
[0003] The invention relates generally to a Complementary Metal
Oxide Semiconductor (CMOS) image sensor. More particularly, the
invention relates to two-dimensional time delay integration visible
CMOS image sensor.
[0004] 2. Description of Related Art
[0005] Unmanned Aerial Vehicles (UAVs) are remotely piloted or
self-piloted aircrafts that can carry cameras, sensors, and other
communication equipment. UAVs may be remotely controlled (e.g.
flown by a pilot at a ground control station) or fly autonomously
based on pre-programmed flight plans or more complex dynamic
automation systems. UAVs are typically used for reconnaissance and
intelligence-gathering, and for more challenging roles, including
combat missions.
[0006] Ideally, an image taken from a camera onboard the UAV should
be clear to provide accurate intelligence-gathering and determine
appropriate targets. However, since UAVs shake from wind gusts
during their flight operation, the image received from UAV is not
clear enough to accurately identify targets on the ground.
Consequently, there is a low signal to noise ratio due to wind and
mechanical vibrations of the camera. This problem is compounded
with moving scene imagery.
[0007] To improve signal to noise ratio, prior art stabilizers were
integrated with the gimbal assembly of high speed cameras onboard
the UAVs. The stabilizers reduce interferences caused by wind or
mechanical vibrations. Additionally, the signal to noise ratio may
be improved using Charge-Coupled Devices (CCDs) with Time Delay
Integration (TDI). CCDs with TDI technology allow an image in a
charge domain to move at about the same speed as the moving scene
or target. However, CCDs with TDI are one dimensional and require
multiple chip systems.
[0008] Conventional CMOS integrated circuits can achieve TDI in one
dimension. The CMOS integrated circuits provide TDI using a switch
matrix or a transistor chain CCD equivalent. The switch matrix
typically accumulates additional noise and the signal to noise
ratio improvement is less than proportional to the square root of
the number of TDI channels. The transistor chain CCD equivalent
cannot have high QE photodiode and is not a mainstream CMOS or CMOS
Image Sensor (CIS) process.
[0009] With an ever increasing demand for improved imaging sensors,
there remains a need for a two dimensional TDI visible CMOS image
sensor that allow a charge to move at the same speed and follow a
similar path in the charge domain as the moving image so that more
charge from the scene can be integrated resulting in an improved
signal to noise ratio. If readout noise is dominant, the signal to
noise ratio improvement is proportional to the number of TDI
channels.
SUMMARY OF THE INVENTION
[0010] The present invention fills this need by providing a time
delay integration CMOS image sensor having a first pinned
photodiode and a second pinned photodiode, the first pinned
photodiode collects a charge when light strikes the first pinned
photodiode, the second pinned photodiode receives the charge from
the first pinned photodiode, and a plurality of electrodes in
series located between the first and the second pinned photodiodes,
the plurality of electrodes are configured to transfer the charge
from the first pinned photodiode to the second pinned photodiode.
The plurality of electrodes may be activated consecutively at
different cycles.
[0011] In one embodiment, the time delay integration CMOS image
sensor may include a plurality of readout nodes coupled to the
second pinned photodiode via address lines. The number of readout
nodes may be equal to the number of pinned photodiodes. The
plurality of electrodes, the plurality of readout nodes and the
address lines may form an orthogonal or hexagonal grid around the
perimeter of each pinned photodiode.
BRIEF DESCRIPTION OF THE DRAWINGS
[0012] The exact nature of this invention, as well as the objects
and advantages thereof, will become readily apparent from
consideration of the following specification in conjunction with
the accompanying drawings in which like reference numerals
designate like parts throughout the figures thereof and
wherein:
[0013] FIG. 1 is a prior art pinned photodiode with transfer gate
and floating diffusion.
[0014] FIG. 2 illustrates the charge transport from the prior art
pinned photodiode to the floating diffusion.
[0015] FIG. 3 is a timing diagram of the logic level for the
transfer gate in FIG. 2.
[0016] FIGS. 4-8 illustrate charge transport in a CMOS image
sensor, according to an embodiment of the invention.
[0017] FIG. 9 is a timing diagram of the logic level for the first
electrode in FIGS. 4-8.
[0018] FIG. 10 is a timing diagram of the logic level for the
second electrode in FIGS. 4-8.
[0019] FIG. 11 is a two dimensional time delay integration visible
CMOS image sensor, according to an embodiment of the invention.
[0020] FIG. 12 is a two dimensional time delay integration visible
CMOS image sensor, according to an embodiment of the invention.
[0021] FIG. 13 illustrates lateral charge transport in a two
dimensional time delay integration visible CMOS image sensor,
according to an embodiment of the invention.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0022] Photodiodes are widely used in digital imaging devices to
convert optical signals into electrical signals. Photodiodes may be
arranged in linear or planar arrays with a plurality of
photosensitive sensors, generally designated as pixels, on a
semiconductor chip. Each pixel generates an output signal
representing the amount of light incident on the pixel.
[0023] A pinned photodiode (PPD) is used to produce and integrate
photoelectric charges generated in CCD or CMOS image sensors. FIG.
1 is a prior art pinned photodiode 11 with transfer gate 13 and
floating diffusion 15. The pinned photodiode 11 generates a charge
17 while maintaining a fixed or pinned Fermi level 19. Regardless
of the potential next to the Fermi level 19 of the pinned
photodiode 11, the Fermi level 19 of the pinned photodiode 11 does
not change.
[0024] Using the pinned photodiode 11 with transfer gate 13 allows
for complete charge removal from light sensing area to the floating
diffusion 15. FIG. 2 illustrates the charge transport from the
pinned photodiode 11 to the floating diffusion 15. FIG. 3 is a
corresponding timing diagram of the logic level for the transfer
gate 13. In the first state 21, the charge 17 is collected by the
pinned photodiode 11. The voltage on transfer gate 13 is zero in
the first state 21. Next, in the second state 23, a positive
voltage is applied to the transfer gate 13. This applied voltage
attracts the charge 17 to move underneath the transfer gate 13, as
shown in FIG. 2. Since the applied voltage decreases the
quasi-Fermi level 19 underneath the transfer gate 13, charge 17
cannot move back to the pinned photodiode 11. In the third state
25, the applied voltage on transfer gate 13 is set to zero. Since
the floating diffusion 15 has a quasi-Fermi level 19 that is lower
than the Fermi level 19 of pinned photodiode 11, the charge 17 will
move across to the floating diffusion 15.
[0025] Combining two transfer gates or electrodes in series
provides charge transport from one pixel to the next. FIGS. 4-8
illustrate charge transport in a CMOS image sensor 27, according to
an embodiment of the invention. The CMOS image sensor 27 has two or
more electrodes 29 between pinned photodiodes 31. By using two or
more electrodes 29, charge 36 can be moved from one pinned photo
photodiode 31 to another. Preferably, the charge 36 moves at about
the same speed as a moving image scene.
[0026] FIG. 4 shows a first electrode 32 and a second electrode 34
between pinned photodiodes 31. The CMOS image sensor 27 may have a
plurality of pinned photodiodes 31 with electrodes 32 and 34 in
between. Control logic may be used to operate the first
electrode(s) 32 simultaneously. Control logic may also be used to
operate the second electrode(s) 34 simultaneously and consecutive
to the operation of the first electrode(s) 32. FIGS. 9 and 10 is an
exemplary timing diagram of the logic level for the first
electrode(s) 32 and second electrode(s) 34, respectively.
[0027] In operation, the CMOS image sensor 27 allows charge(s) 36
to travel from one pinned photo photodiode 31 to another.
Initially, in FIG. 4, pinned photodiode 31 collects charge(s) 36
while maintaining a fixed or pinned Fermi level 38. Regardless of
the potential next to the Fermi level 38 of the pinned photodiode
31, the Fermi level 38 of the pinned photodiode 31 does not change.
No voltage is applied to the first and second electrodes 32 and
34.
[0028] Next, in FIG. 5, the first electrode 32 is activated by
applying a voltage for a predetermined period. This voltage
attracts the charge 36 to move underneath the first electrode 32.
Since the applied voltage decreases the quasi-Fermi level 38 of the
first electrode 32 by creating a well, charge 36 cannot move back
to the pinned photodiode 31.
[0029] In FIG. 6, the second electrode 34 is activated by applying
a positive voltage for a predetermined period. The voltage applied
to the second electrode 34 is preferably greater than or equal to
the voltage applied to the first electrode 32. The voltage applied
to the second electrode 34 attracts the charge 36 to move
underneath the second electrode 34 as well. The applied voltage
decreases the quasi-Fermi level 38 of the second electrode 34 to
allow charge 36 to distribute under both electrodes 32 and 34.
[0030] In FIG. 7, the applied voltage for the first electrode 32 is
set to zero. This resets the potential of the first electrode 32
and collapses the well underneath the first electrode 32. Since the
second electrode 34 is still activated, the quasi-Fermi level 38 of
the second electrode 34 will be lower than the quasi-Fermi level 38
underneath first electrode 32 and photodiode 31. Consequently, the
charge 36 that was underneath the first electrode 32 will move
across and remain underneath the second electrode 34.
[0031] In FIG. 8, the applied voltage for the second electrode 34
is set to zero. This resets the potential of the second electrode
34 and collapses the well underneath the second electrode 34. Since
the pinned photodiode 31 has a predetermined fixed Fermi level 31,
the charge 36 underneath the second electrode 34 will move across
to the adjacent pinned photodiode 31. Consequently, lateral charge
36 transport occurs in the CMOS image sensor 27.
[0032] According to an embodiment of the invention, the lateral
charge 36 transport occurs over a 4 cycle period, as shown in FIGS.
9 and 10. A person skilled in the art would appreciate that
different cycles may be used without departing from the spirit of
the invention. In the first cycle, the first electrode 32 is
activated by applying a positive voltage. In the second cycle, the
second electrode 34 is activated as well by applying a voltage. In
the third cycle, the first electrode 32 is deactivated by setting
the voltage applied to the first electrode 32 to zero. In the
fourth cycle, the second electrode 34 is deactivated by setting the
voltage applied to the second electrode 34 to zero.
[0033] As shown in FIGS. 4-8, the CMOS image sensor 27 has a
plurality of pinned photodiodes 31 with at least two electrodes 32
and 34 in between. Electrodes 32 operate at a different phase than
electrodes 34 to allow charge 36 to move from underneath one
electrode to the other. The phase relationship between electrodes
32 and electrodes 34 defines the transport direction of the charge
36. For example, control logic may be used to alternate the phase
shift between the electrodes 34 and 34 such that the charge 36
moves from photodiode 31 adjacent to the second electrode 34, to
underneath second electrode 34, to underneath first electrode 32,
and finally to the photodiode 31 adjacent to the first electrode
32.
[0034] FIG. 11 is a two dimensional time delay integration visible
CMOS image sensor 40, according to an embodiment of the invention.
The sensor 40 has an active array of pixels 42, each pixel 42 may
include a pinned photodiode 44 with four orthogonal electrodes 46,
47, 48 and 49. The pixels 42 are interconnected in a grid with
readout nodes 50 and address lines 52. The address lines 52 control
the voltage on electrodes 46, 47, 48 and 49. Through an additional
transfer gate (not shown) between photodiode 44 and readout node
50, the signal charge can be transferred to the readout node 50 at
the end of a TDI cycle. In one embodiment, the sensor 40 has a
readout node 50 for every photodiode 44.
[0035] With moving scene imagery, pinned photodiode 44 of the time
delay integration visible CMOS image sensor 40 generates a charge
that moves in two dimensions at about the same speed and follows a
similar path as the moving image. Similarly, mechanical vibrations
of a camera cause random walk of any image point on the sensor 40.
FIG. 11 illustrates the two dimensional charge transport directions
54 and 56. The charge moves laterally from one photodiode 44 to
another. This lateral movement of charge provides improved charge
integration from the moving scene. Since there are multiple readout
nodes 50 distributed evenly in the sensor 40, photo-generated
signals may be read at any point in the array closest to the
readout node 50, rather than transporting the charge for readout
down or up stream. This provides high frame rate capability with
improved signal to noise ratio for the sensor 40.
[0036] To better approximate the curved random walk of a scene, the
sensor may be configured to allow for charge transport in three or
more directions. FIG. 12 illustrates charge transport in three
directions 62, 64 and 66 for a two dimensional time delay
integration visible CMOS image sensor 60, according to an
embodiment of the invention. The sensor 60 has an active array of
pixels 68, each pixel 68 may include a pinned photodiode 70 with
six electrodes 72, 74, 76, 78, 80 and 82. The pixels 68 are
interconnected in a polygonal grid, such as a hexagonal grid, with
readout nodes 84 and address lines 86. The address lines 86 control
the voltage on the electrodes 72, 74, 76, 78, 80 and 82. Through an
additional transfer gate (not shown) between photodiode 70 and
readout node 84, the signal charge can be transferred to the
readout node 70 at the end of a TDI cycle. In one embodiment, the
sensor 60 has a readout node 84 for every photodiode 70.
[0037] With moving scene imagery, pinned photodiode 70 of the time
delay integration visible CMOS image sensor 60 generates a charge
that moves in two dimensions at about the same speed and follows a
similar path as the moving image. Similarly, mechanical vibrations
of a camera cause random walk of any image point on the sensor 40.
FIG. 13 illustrates lateral charge transport in sensor 60. Due to
the hexagonal grid configuration, the charge travels in a smooth
path 88 that follows the moving image. This lateral movement of
charge provides improved charge integration from the moving scene.
Since there are multiple readout nodes 84 distributed evenly in the
sensor 60, photo-generated signals may be read at any point in the
array closest to the readout node 84. This provides high frame rate
capability with improved signal to noise ratio for the sensor
60.
[0038] A person skilled in the art would appreciate the potential
applications of the two dimensional time delay integration visible
CMOS image sensor of the present invention. The sensor may be used
for translational image stabilization during single frame
integration time. For example, very high bandwidth of translational
vibrations can be stabilized from about 30 Hz to about 1 MHz. The
maximum translational vibration amplitude may be limited by imager
resolution. The sensor may also be used for rotational image
stabilization during single frame integration time. For example,
very high bandwidth of rotational movement can be stabilized from
about 30 Hz to about 1 MHz. The maximum rotational vibration
amplitude may be limited by pixel size and tolerable
distortions.
[0039] Other applications of the sensor include residue light
photography without tripod or flash, TDI camera with increased
alignment tolerance and flow cytometry for capturing images of
moving cells in fluids. The sensor may also be used, in combination
with a stabilized gimbal, to enhance pointing accuracy to a few
tens of grads. Additionally, the sensor may be used with Inertial
Measurement Unit (IMU) to suppress random motion. Depending on
frame rate, IMU may be replaced with processing algorithm.
[0040] While certain exemplary embodiments have been described and
shown in the accompanying drawings, it is to be understood that
such embodiments are merely illustrative of and not restrictive on
the broad invention, and that this invention not be limited to the
specific constructions and arrangements shown and described, since
various other changes, combinations, omissions, modifications and
substitutions, in addition to those set forth in the above
paragraphs, are possible. Those skilled in the art will appreciate
that various adaptations and modifications of the just described
preferred embodiment can be configured without departing from the
scope and spirit of the invention. Therefore, it is to be
understood that, within the scope of the appended claims, the
invention may be practiced other than as specifically described
herein.
* * * * *