U.S. patent application number 12/946449 was filed with the patent office on 2011-03-10 for solid state imaging device with horizontal transfer paths and a driving method therefor.
Invention is credited to Katsumi IKEDA, Hirokazu KOBAYASHI, Kazuya ODA, Hiroyuki OSHIMA.
Application Number | 20110058083 12/946449 |
Document ID | / |
Family ID | 38558311 |
Filed Date | 2011-03-10 |
United States Patent
Application |
20110058083 |
Kind Code |
A1 |
OSHIMA; Hiroyuki ; et
al. |
March 10, 2011 |
SOLID STATE IMAGING DEVICE WITH HORIZONTAL TRANSFER PATHS AND A
DRIVING METHOD THEREFOR
Abstract
In a solid state imaging device, signal charges are branched to
be output to in the form of one or plural outputs. At a horizontal
transfer speed not lower than a predetermined transfer speed, the
imaging device transfers signal charges of color attributes
classified by a branching section, to plural horizontal transfer
paths, where the signal charges are converted into analog voltage
signals, which will be output synchronously. At a horizontal
transfer speed lower than the predetermined transfer speed, the
analog voltage signal converted is output from, e.g. the horizontal
transfer path which has been selected. Output amplifiers arranged
on the horizontal transfer paths are differentiated in
sensitivities in detecting signal charges, depending on color
attributes of signal charges supplied, and output the analog
voltage signals.
Inventors: |
OSHIMA; Hiroyuki;
(Asaka-shi, JP) ; KOBAYASHI; Hirokazu; (Asaka-shi,
JP) ; ODA; Kazuya; (Asaka-shi, JP) ; IKEDA;
Katsumi; (Kurokawa-gun, JP) |
Family ID: |
38558311 |
Appl. No.: |
12/946449 |
Filed: |
November 15, 2010 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11727139 |
Mar 23, 2007 |
7855742 |
|
|
12946449 |
|
|
|
|
Current U.S.
Class: |
348/311 ;
348/E5.091 |
Current CPC
Class: |
H01L 27/14868 20130101;
H04N 5/3728 20130101; H04N 9/045 20130101; H01L 27/14621 20130101;
H04N 5/369 20130101; H04N 5/3653 20130101; H04N 5/37213 20130101;
H01L 27/14843 20130101; H04N 9/04557 20180801 |
Class at
Publication: |
348/311 ;
348/E05.091 |
International
Class: |
H04N 5/335 20110101
H04N005/335 |
Foreign Application Data
Date |
Code |
Application Number |
Mar 30, 2006 |
JP |
2006-94407 |
Mar 30, 2006 |
JP |
2006-94567 |
Mar 30, 2006 |
JP |
2006-95198 |
Mar 30, 2006 |
JP |
2006-95374 |
Claims
1. A solid state imaging apparatus comprising: a color filter for
color separating the incident light from a field being viewed into
a plurality of colors; a plurality of photosensitive cells for
photo-electrically transducing the light transmitted through said
color filter, said photosensitive cells being arranged in register
with the colors; a first transfer circuit for transferring signal
charges read out from said photosensitive cells in a first
direction; a second transfer circuit for transferring signal
charges transferred in said first transfer circuit in a second
direction; a branching circuit arranged at an output end of said
second transfer circuit for distributing the signal charges
transferred to a plurality of output destinations; a plurality of
third transfer circuits connected as the output destinations to
said branching circuit; and output circuits connected to an output
end of said third transfer circuits; said plural colors being
divided into a plurality of groups; said second transfer circuit,
said branching circuit and said third transfer circuits
transferring signal charges read out from the photosensitive cells
related with colors of the same group, and subsequently
transferring signal charges read out from the photosensitive cells
related with colors of a different group or groups.
2. The solid state imaging apparatus in accordance with claim 1
wherein the plural colors are three colors of red, green and blue,
and the groups are two groups, one being a group composed of red
and blue and the other being a group composed of green.
3. The solid state imaging apparatus in accordance with claim 1
wherein, in case there are a plurality of colors belonging to one
of said groups, the signal charges read out from the photosensitive
cells related with the colors are transferred using a plurality of
said third transfer circuits; in case there is one color belonging
to one of said groups, the signal charges read out from the
photosensitive cells related with the color being transferred using
a specified one of said third transfer circuits.
4. The solid state imaging apparatus in accordance with claim 1
wherein, in case there are a plurality of colors belonging to one
of said groups, the signal charges read out from the photosensitive
cells related with said colors are transferred using a plurality of
said third transfer circuits; in case there is one color belonging
to one of said groups, the signal charges read out from the
photosensitive cells related with the color being also transferred
using a plurality of said third transfer circuits.
5. The solid state imaging apparatus in accordance with claim 3
wherein, in case there are a plurality of colors belonging to one
of said groups, the signal charges read out from the photosensitive
cells related with the colors are transferred using a specified one
of said third transfer circuits from one color to another.
6. The solid state imaging apparatus in accordance with claim 1
further comprising a correction circuit for correcting the
difference in characteristics among said output circuits.
7. The solid state imaging apparatus in accordance with claim 6
further comprising: a holding circuit for holding data for
correcting the difference in characteristics among said output
circuits; said correction circuit correcting the difference in
characteristics using data owned by said holding circuit.
8. The solid state imaging apparatus in accordance with claim 1
wherein said first transfer circuit transfer the signal charges,
readout from the photosensitive cells, related with the color
belonging to the same group, to said second transfer circuit, and
subsequently transfer signal charges read out from the
photosensitive cells related with the color belonging to the
different group.
9. An imaging method comprising the steps of: color-separating
incident light from a field being imaged by a color filter;
photo-electrically transducing the light transmitted through said
color filter by a plurality of photosensitive cells related with
the colors; transferring signal charges read out from said
photosensitive cells by first transfer circuit in a first
direction; transferring the signal charges transferred by said
first transfer circuits, by a second transfer circuit in a second
direction; distributing the signal charges transferred to a
plurality of output destinations, by a branching circuit arranged
at an output end of said second transfer circuit; and outputting
the signal charges by a plurality of third transfer circuits,
connected as output destinations to said branching circuit, and by
output circuits arranged at an output end of said third transfer
circuits; said plural colors being grouped into a plurality of
groups; and said second transfer circuit, said branching circuit
and the third transfer circuits transferring the signal charges
read out from the photosensitive cells related with the color
belonging to the same group, and subsequently transferring signal
charges read out from the photosensitive cells related with the
color belonging to the different group.
10. An imaging method comprising the steps of: color-separating
incident light from a field being imaged by a color filter;
photo-electrically transducing the light transmitted through said
color filter by a plurality of photosensitive cells related with
the colors; transferring signal charges read out from said
photosensitive cells by first transfer circuit in a first
direction; transferring the signal charges, transferred by said
first transfer circuit, by a second transfer circuit in a second
direction; and outputting signal charges via an output circuit
arranged at an output end of said second transfer circuit; said
plural colors being grouped into a plurality of groups; and said
second transfer circuits transferring the signal charges read out
from the photosensitive cells related with the color belonging to
the same group, and subsequently transferring signal charges read
out from the photosensitive cells related with the color belonging
to the different group.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] This application is a Divisional of co-pending application
Ser. No. 11/727,139 filed Mar. 23, 2007 the entire contents of
which are hereby incorporated by reference and for which priority
is claimed under 35 U.S.C. .sctn.120.
BACKGROUND OF THE INVENTION
[0002] 1. Field of the Invention
[0003] The present invention relates to a solid-state imaging
device and a driving method therefor. More particularly, the
present invention relates to a solid-state imaging device such as a
CCD (Charge-Coupled Device) imaging device for transducing incident
light into signal charges and horizontally transferring the charges
to output an image signal through an output amplifier. The
invention also more specifically relates to a driving method of
driving the imaging device for horizontal transfer.
[0004] 2. Description of the Background Art
[0005] Japanese patent laid-open publication No. 50409/1995
discloses bifurcating a sole transfer path, that is, a shift
register, and providing an amplifier at the distal end of each
bifurcated transfer path. Each amplifier is a floating diffusion
amplifier (FDA) and differs from one another in charge detection
sensitivity or charge-to-voltage conversion efficiency. When a
subject being imaged is of lower luminance, the higher-efficiency
amplifier outputs the output signal. This amplifier selection
improves the voltage conversion efficiency for lower signal charges
obtained on light reception, and raises the sensitivity. When the
subject being imaged is of higher luminance, the lower-efficiency
amplifier outputs the output signal. With the use of this
amplifier, the image generated by the output signal is of a wider
dynamic range. Another Japanese patent laid-open publication No.
298626/1996 is substantially similar to the '409 publication.
[0006] Still another Japanese patent laid-open publication No.
244340/1993 also teaches bifurcating a sole transfer path, or a
shift register. However, the signal charges obtained on light
reception or optical sensing are alternately switched at the
bifurcating section and transferred to the respective amplifiers.
The frequency of the driving pulse supplied to the transfer path
not bifurcated is twice as high as the driving pulse supplied to
the bifurcated transfer paths. In other words, a driving frequency
one-half the usual driving frequency suffices for driving the
bifurcated transfer paths. Thus, the amplifier renders it possible
to raise the transfer speed as the driving frequency is maintained
within the frequency band prescribed as its operating
characteristics.
[0007] In the '409 and '626 publications, there is simply disclosed
guiding the signal charges to a selected transfer path. The '340
publication discloses alternately outputting signal charges. If
those publications are simply combined together, it would indeed be
possible to output images that satisfy the requirements for high
sensitivity and wide dynamic range.
[0008] In producing a color image, color attributes are allocated
to the respective signal charges. However, there is neither
suggestion nor disclosure as to how the signal charges are
transferred and distributed to the bifurcated transfer paths from
color to color. Thus, with those prior art publications combined
together, it would not be possible to provide an image reduced in
noise based on a white-balance gain with respect to color.
[0009] If with the conventional solid state imaging device a
subject with lower color temperature, for example, is shot, then
the amounts of signal charges obtained in the photosensitive cells
of the imaging device are larger for red (R) pixels and smaller for
blue (B) pixels. If the signal charge are horizontally transferred
from the transfer path not bifurcated in the sequence of an R
pixel, a first green (G1) pixel, a B pixel and a second green (G2)
pixel, the amount of charges left over by the forward R pixel and
mixed into the rear side G1 pixel is greater than the amount of
charge mixing between the forward side B pixel and the rear side G2
pixel. Thus, the G1 and G2 pixels which are of the same color
signal differ in the signal quantity, thus affecting the finished
image as a fixed pattern noise.
[0010] In addition, deterioration in the transfer efficiency of
signal charges is caused with the solid state imaging device in
which signal charges are not branched optimally at a branching
electrode such that signal charges to be sent to one of the
horizontal transfer paths are intruded into the other transfer
path.
[0011] More specifically, such a case is now considered in which
signal charges are transferred from a sole horizontal transfer path
in the sequence of a G1 pixel, an R pixel, a G2 pixel and a B
pixel, and are branched at the branching electrode, signal charges
of the pixels G1 and G2 are sent to one of the horizontal transfer
paths and those of the pixels R and B are sent to the other
transfer path. If the ambient temperature at the time of imaging is
low, part of signal charges of the pixel G1 is mixed into signal
charges of the pixel R.
[0012] Such deterioration in the transfer efficiency, i.e.,
transfer deterioration, of signal charges is caused not only when
the ambient temperature at the time of imaging is low, but also
when the color temperature of a subject being imaged is low or the
ISO (International Organization for Standardization) sensitivity is
high. In particular, in case of a subject with a low color
temperature, signal charges are mixed in different quantities, even
if the signal charges are of the same color. More specifically,
supposing that signal charges of pixels G1, R, G2 and B obtained on
imaging a subject of a low color temperature are transferred in
that order, the signal charges are mixed in the pixel R from the
pixel G1 in larger quantity than the quantity of the signal charges
mixed from the pixel G2 into the pixel B. The result is that signal
quantities of the pixels R and B become different from each other,
with the difference in signal quantity then being visualized as
noise in the image.
[0013] In the meantime, in those prior art publications, there is
neither suggestion nor disclosure as to the sequence of readout of
color signals or as to how signal charges are to be split to the
branched transfer paths depending on the colors. It is noted that,
if the branching section receives signals of different colors, and
cannot correctly transfer the signal charges to the branches, part
of those signals of different colors may be mixed with each other
on the branches.
SUMMARY OF THE INVENTION
[0014] It is an object of the present invention to provide a solid
state imaging device and a driving method therefor with which it is
possible to branch signal charges to output the so branched signal
charges on either one or plural paths, as well as to achieve the
noise reduction based on the white balance gain.
[0015] It is another object of the present invention to provide a
solid state imaging device and a driving method therefor in which
the ill effect of transfer deterioration ascribable to the
bifurcation may be moderated even with the use of an imaging device
having its horizontal transfer path bifurcated partway to output
signal charges in either one or plural paths, thereby enabling an
optimum image to be produced.
[0016] It is still another object of the present invention to
provide a driving method for a solid state imaging device which is
free from deterioration in the transfer efficiency at the branching
electrode and is capable of producing an optimal image, as well as
to provide a driving method for use in an imaging apparatus
including the solid state imaging device.
[0017] It is yet another object of the present invention to provide
a solid state imaging apparatus and an imaging method, in which,
even in case the branching section cannot correctly transfer signal
charges, signal mixing between different colors is minimized.
[0018] In accordance with the present invention, there is provided
a solid state imaging device including a two-dimensional array of
photosensitive cells, a first horizontal transfer circuit for
transferring signal charges, a branching circuit, a plurality of
second horizontal transfer circuits, and a plurality of output
circuits. The photosensitive cells are supplied with incident light
from a field being imaged, via color filter segments adapted for
color separating the incident light. The photosensitive cells
operate for transducing the light transmitted through the color
filter segments into electrical signals depending on the volume of
transmitted light. The first horizontal transfer circuit operate
for transferring signal charges having color attributes, read out
from each of the photosensitive cells and transferred in a vertical
direction, in a horizontal direction perpendicular to the vertical
direction. The branching circuit is arranged adjacent to an output
end of the first horizontal transfer circuit for transiently
holding the signal charges transferred and for distributing the
signal charges to a plurality of output destinations related with
the color attributes of the signal charges. The second horizontal
transfer circuits are connected to the branching circuit as
branching destinations. The output circuits are each provided at an
output end of each of the second horizontal transfer circuits. The
output circuits each operate for converting the signal charges into
an analog voltage signal responsive to detection of the signal
charges to amplify the analog voltage signal. The signal charges
supplied to the branching circuit at a horizontal transfer speed
during horizontal transfer not lower than a predetermined transfer
speed are classified depending on the color attributes and
transferred to each of the second horizontal transfer circuits. The
signal charges supplied from the output circuits are converted into
analog voltage signals which are output. The analog voltage signals
are output from selected horizontal transfer circuit with transfer
at a horizontal transfer speed lower than the predetermined
transfer speed. The plural output circuits exhibit differential
detection sensitivities for the signal charges depending on the
color attributes of the signal charges supplied. The output
circuits output the analog voltage signals.
[0019] With the solid state imaging device according to the present
invention, signal charges of the color attributes, classified in
the branching circuit, are transferred to each of a plural number
of horizontal transfer means for horizontal transfer with a
horizontal transfer speed not lower than a predetermined speed. The
analog voltage signals, converted from the signal charges, are
output. For horizontal transfer with the horizontal transfer speed
lower than the predetermined transfer speed, the analog voltage
signals converted from the signal charges are output from the
selected horizontal transfer circuit. As each of the output
circuits associated with plural horizontal transfer means has
differential sensitivities in detecting the signal charges,
depending on the color attributes of the signal charges supplied
thereto, it is possible to modulate the sensitivities of the red
and blue signals with respect to those of the green signals to
suppress unneeded gain and hence reduce the noise.
[0020] In accordance with the present invention, there is also
provided a method for driving a solid state imaging device for
transferring signal charges having color attributes. The solid
state imaging device includes a plurality of vertical transfer
circuits and horizontal transfer circuit. The vertical transfer
circuits operate for reading out signal charges having color
attributes from each of photosensitive cells each adapted for
transducing incident light into an electrical signal. The vertical
transfer circuits transfer the read-out signal charges having the
color attributes, in the vertical direction. The horizontal
transfer circuit transfers the signal charges transferred from the
vertical transfer circuits, in the horizontal direction. The solid
state imaging device also includes a plurality of horizontal
transfer circuits branched from a transfer region adapted for
branching signal charges from the horizontal transfer circuit
having color attributes. The plural horizontal transfer circuits
exhibit differential detection sensitivities to signal charges
having the color attributes. The method comprises classifying the
signal charges, supplied to the branching transfer region with high
speed driving with the horizontal transfer speed not less than a
predetermined transfer speed, according to the color attributes,
sending the signal charges of the color attributes on each of the
plural horizontal transfer circuits, converting the signal charges
of the color attributes into analog voltage signals, outputting the
analog voltage signals, outputting the analog voltage signals from
selected horizontal transfer circuit with low speed driving at a
horizontal transfer speed lower than the predetermined transfer
speed, and adjusting the branched output destination of the signal
charges having color attributes by changing the phase of a driving
pattern in horizontal transfer, before branching, with respect to
the driving in at least one branched horizontal transfer.
[0021] With the method for driving the solid state imaging device,
the plural horizontal transfer circuits exhibit differential
detection sensitivities to signal charges having the color
attributes. The branched output destination of the signal charges
having color attributes may be adjusted by changing the phase of a
driving pattern in horizontal transfer, before branching, with
respect to the driving in at least one branched horizontal
transfer. This renders it possible to freely change the sensitivity
in charge detection for signal charges supplied and to set the gain
flexibly in keeping with imaging conditions on hand. In this
manner, data of high accuracy may be obtained to provide a
high-quality image.
[0022] In accordance with the present invention, there is also
provided a solid state imaging apparatus including a solid state
imaging device. The solid state imaging device includes a plurality
of photosensitive cells, a plurality of color filter segments, a
vertical transfer circuit, a first horizontal transfer circuit, a
branching circuit, second and third transfer circuits and first and
second output circuits. The photosensitive cells are arrayed in a
matrix of rows and columns for photo-electrically transducing
incident light from a field being imaged into signal charges. The
color filter segments are arranged in register with the
photosensitive cells for color separating the incident light into
plural colors to cause the light of the plural colors to be
incident on the photosensitive cells. The vertical transfer circuit
operates for vertically transferring the signal charges read out
from the photosensitive cells. The first horizontal transfer
circuit operates for receiving the signal charges vertically
transferred from the vertical transfer circuit to transfer the
received signal charges horizontally. The branching circuit is
arranged at an output end of the first horizontal transfer circuit
for distributing the horizontally transferred signal charges to an
optional one of a plurality of output destinations. The second and
third transfer circuits are adapted for receiving the signal
charges distributed from the branching circuit to further
horizontally transfer the signal charges. The first and second
output circuits are arranged at output ends of the second and third
transfer circuits, respectively. The solid state imaging apparatus
comprises transfer efficiency measuring means for measuring the
transfer efficiency on the second and third transfer circuits in
the course of transfer of the signal charges from the branching
circuit through the second and third transfer circuits to the first
and second output circuits. The driving start time of a first drive
signal for driving the first horizontal transfer circuit or a
second drive signal for driving the second and third horizontal
transfer circuits is changed depending on the result of measurement
by the transfer efficiency measuring circuit. One of the second and
third horizontal transfer circuits is preferentially used to
transfer the signal charges.
[0023] Thus, with the solid state imaging apparatus according to
the present invention, when the signal charges obtained by the
imaging unit from the photosensitive cells, during high speed
driving, are transferred via a plural number of vertical transfer
paths to a first horizontal transfer path, and the signal charges
horizontally transferred from the first horizontal transfer path
are branched and transferred to the second and third horizontal
transfer paths, the phase of the horizontal timing signal
controlling the driving of the horizontal transfer path of the
imaging unit is offset from that of the initial driving condition,
depending on the results of measurement by the transfer efficiency
measurement unit. By so doing, it is possible to interchange the
signal charges transferred to the second horizontal transfer path
and those transferred to the third horizontal transfer path, by way
of reversing the branching.
[0024] Moreover, with the solid state imaging apparatus according
to the present invention, signal charges of red and blue pixels are
transferred on one of the second and third horizontal transfer
paths which has a higher transfer efficiency, during high speed
driving, while those of green pixels are transferred on the
remaining horizontal transfer path. By so doing, the quantity of
residual transfer charges from the red and blue pixels to the green
pixels may be reduced to provide for an optimum image.
[0025] With the solid state imaging apparatus according to the
present invention, all signal charges are transferred during low
speed driving to the second or third horizontal transfer paths
which has a higher transfer efficiency, depending on the results of
measurement by the transfer efficiency measurement unit. By so
doing, the quantity of residual transfer charges of the pixels may
be decreased to moderate any adverse effect of transfer
deterioration of the entire image to provide for an optimum
image.
[0026] Furthermore, with the solid state imaging apparatus
according to the present invention, the imaging unit switches
between the output destinations of the second and third horizontal
transfer paths, depending on the transfer efficiency. By so doing,
there is no necessity of changing the wiring of the external
circuit of the imaging unit even if the output site of the color
signals is changed between the second and third horizontal transfer
paths.
[0027] In accordance with the present invention, for accomplishing
the above other object of the present invention, the duty cycle
and/or the period of the drive signal driving the transfer circuit
before branching, which is an electrode directly upstream of the
branch electrode in the solid state imaging device, is changed, or
the duty cycle and/or the period of the drive signal driving the
horizontal transfer path transferring branched signal charges
signal charges is changed, in order to provide for transfer time
for signal charges from the electrode before branching to the
branch electrode longer than the usual transfer time, or for
transfer time from the branch electrode to the horizontal transfer
path transferring branched signal charges longer than the usual
transfer time.
[0028] In more detail, attention is directed in the present
invention to the fact that, if part of signal charges of red and
blue pixels is intruded into signal charges of the green pixels,
the result is the noise represented on the image. Thus, the
transfer time of signal charges of red and blue pixels to the
branch electrode from the electrode before branching or the
transfer time of transfer to the horizontal transfer path of
branched signal charges of red and blue pixels from the branch
electrode, is set so as to be longer than the usual value.
[0029] The usual transfer time is the transfer time prior to
changing the duty cycle or the period of the drive signal, that is,
the transfer time in which no transfer efficiency deterioration has
occurred, viz, the transfer time in which the transfer efficiency
is maintained. Meanwhile, the duty cycle is the temporal relation
between the periodically alternating high and low levels, and
specifically is the ratio of the high level time of a signal in a
signal period, or a duty ratio.
[0030] By providing for the transfer time during transfer of the
signal charges of the red and blue pixels longer than the usual
transfer time, it is possible to secure sufficient shifting of the
signal charges. Hence, it is possible to prevent part of the signal
charges of red and blue pixels from being left over and mixing into
the signal charges of the green pixel.
[0031] The processing for providing for the transfer time longer
than the usual transfer time may be changed in dependence upon the
temperature, color temperature of the subject, sensitivity or the
rate of reading out electrical signals. How much the duty cycle or
the period is to be changed may be found by calculating the
transfer efficiency, which transfer efficiency may be calculated
using a reference signal charge.
[0032] According to the present invention, it is possible to
prevent deterioration of the transfer efficiency in the branch
electrode and to obtain an optimum image free of noise.
[0033] In accordance with the present invention there is also
provided a solid state imaging apparatus comprising a color filter,
a plurality of photosensitive cells, a first transfer circuit, a
second transfer circuit, a branching circuit, a plurality of third
transfer circuits, and output circuits. The color filter operates
for color separating the incident light from a field being viewed
into a plurality of colors. The photosensitive cells operate for
photo-electrically transducing the light transmitted through the
color filter. The photosensitive cells are arranged in register
with the colors. The first transfer circuit operates for
transferring signal charges read out from the photosensitive cells
in a first direction. The second transfer circuit operates for
transferring signal charges read out from the first transfer means
in a second direction. The branching circuit is arranged at an
output end of the second transfer circuit for distributing the
signal charges transferred to a plurality of output destinations.
The third transfer circuits are connected as the output
destinations to the branching circuit, and the output circuits are
connected to an output end of the third transfer circuits. The
plural colors are divided into a plurality of groups. The second
transfer circuit, the branching circuit and the third transfer
circuits transfer signal charges read out from the photosensitive
cells related with colors of the same group, and subsequently
transfer signal charges read out from the photosensitive cells
related with colors of a different group or groups.
[0034] According to the present invention, in case the second
transfer circuit, for example, the horizontal transfer path, is
branched partway, outputting is to be made at plural output parts,
and the colors belonging to different groups, for example, red (R)
and blue (B) on one hand and green (G) on the other hand, may be
read out in a state completely isolated from each other. For
example, in transferring G signals and RB signals contained in the
same line or field, it is possible to transfer the G signals
contained in a given line or field, and subsequently to transfer
the RB signals contained in the same line or field. That is, the G
signals and the RB signals contained in the same line or field, may
be separately read out as if these G signals and RB signals belong
to a distinct line of field.
[0035] Hence, there is no fear that different colors alternately
pass through the branching circuit, so that it is possible to
prevent G signals from mixing into R and B signals or to prevent R
and B signals from mixing into the G signals, so as to provide an
image in an optimum state. The reason signal mixing is likely to be
produced in branching circuit is that it is necessary for the
branching circuit to be able to transfer signal charges in plural
directions and hence is larger in size or of a special profile
other than a square with the result that signal charges cannot be
transferred unobjectionably. For example, it may occur that the
signal at a terminal part of branching circuit cannot be
transferred.
[0036] According to the present invention, the signal charges
transferred in a lump as belonging to the same group are not
necessarily the signal charges of one line or field, and may be
signal charges of more or less than one line or field. A plural
number of lines or fields, one frame, a plural number of frames or
one-half line of signal charges may be transferred in a lump.
[0037] In the present invention, if the plural colors are three
colors of red, green and blue, it is preferred to provide two
groups, namely a group of red and blue and a group of green. The
reason is that, even though signal mixing occurs between red and
blue, this state can be corrected rather easily.
[0038] If there are plural colors belonging to one of the groups,
the signal charges read out from the photosensitive cells related
with the respective colors are preferably transferred from color to
color using specified one of third transfer circuits. Since output
circuits are determined from color to color, it is possible to
prevent differential intensities (step differences) between the
same colors ascribable to individual differences of the output
parts.
[0039] In case signal charges are transferred using a plural number
of third transfer circuits, it is preferred to use correction
circuit for correcting the difference in characteristics between
the plural output circuits in order to correct the difference in
characteristics between the plural output circuits. This corrects
the step difference of colors ascribable to individual differences
of the output parts.
[0040] The present invention may be applied to an imaging method
not exploiting branching circuit. That is, the present invention
may be applied to an imaging method comprising color-separating
incident light from a field being imaged by a color filter,
photo-electrically transducing the light transmitted through the
color filter by a plurality of photosensitive cells related with
the colors, transferring signal charges read out from the
photosensitive cells by first transfer circuit in a first
direction, transferring the signal charges transferred by the first
transfer circuit, by second transfer means in a second direction,
and outputting signal charges via output circuits arranged at an
output end of the second transfer circuit. The plural colors are
divided into a plurality of groups. The second transfer circuit
transfers the signal charges read out from the photosensitive cells
related with the color belonging to the same group, and
subsequently transfer signal charges read out from the
photosensitive cells related with the color belonging to the
different group.
[0041] According to the present invention, there may be provided a
solid state imaging method and apparatus in which signal mixing may
not be produced among different colors even in case branching
circuit cannot unobjectionably transfer signal charges.
BRIEF DESCRIPTION OF THE DRAWINGS
[0042] The objects and features of the present invention will
become more apparent from consideration of the following detailed
description taken in conjunction with the accompanying drawings in
which:
[0043] FIG. 1 is a schematic block diagram showing a preferred
embodiment of a two-line readout CCD implemented as a solid state
imaging device according to the present invention;
[0044] FIG. 2 is a schematic block diagram showing a preferred
embodiment of a digital camera employing the solid state imaging
device of FIG. 1;
[0045] FIG. 3 is a schematic block diagram showing drivers shown in
FIG. 2;
[0046] FIG. 4, part (A) is a partial plan view looked from above,
and showing the schematic constitution of a horizontal transfer
path in the solid state imaging device of FIG. 1, part (B) is a
cross-sectional view of the transfer path, taken along a section
line IV-IV, and parts. (C) and (D) show how the potential changes
in various parts of the horizontal transfer path;
[0047] FIG. 5 is a timing chart showing the timing of drive signals
supplied to respective electrodes of FIG. 4, parts (A) and (B);
[0048] FIG. 6, part (A) continuing from the lower part of FIG. 4 is
a cross-sectional view of the transfer path, and parts (B), (C) and
(D) show how the potential changes;
[0049] FIG. 7, part (A) is a partial plan view, looking from above,
showing the schematic constitution of the horizontal transfer path
in the device of FIG. 1, part (B) is a cross-sectional view of the
transfer path, taken along a section line VII-VII, and parts (C)
and (D) show how the potential changes in various parts of the
horizontal transfer path;
[0050] FIG. 8, part (A) continuing from the lower part of FIG. 7 is
a cross-sectional view of the transfer path, and parts (B), (C) and
(D) show how the potential changes;
[0051] FIGS. 9A to 9E are schematic views for illustrating the
transfer of signal charge with color attributes over horizontal
transfer paths in the solid state imaging device of FIG. 1;
[0052] FIGS. 10A and 10B schematically show the difference in gate
capacitances in the output amplifier of FIG. 1;
[0053] FIGS. 11A and 11B schematically show the difference in film
thickness formed in the floating diffusion in the output amplifier
of FIG. 1;
[0054] FIGS. 12A and 12B schematically show the difference in the
surface area in the floating diffusion in the output amplifier of
FIG. 1;
[0055] FIGS. 13A and 13B schematically show the presence and the
absence of a film formed in the floating diffusion in the output
amplifier of FIG. 1, respectively;
[0056] FIG. 14 is a partial plan view of an array of offset pixels
and color filter segments in the solid state imaging device of FIG.
1;
[0057] FIGS. 15 and 16 are timing charts showing re-arraying of
signal charges of the first and second fields during the horizontal
blanking period in connection with horizontal transfer of the
device of FIG. 14, respectively;
[0058] FIG. 17 is a timing chart showing the relationship of the
drive signals supplied for the first field and the output signals
in connection with horizontal transfer of the device;
[0059] FIG. 18 is a timing chart showing the relationship of
low-speed readout drive signals ad output signal in the horizontal
transfer of the device;
[0060] FIG. 19 is a schematic block diagram showing a three-line
readout solid state imaging device applied to the imaging unit of
FIG. 2;
[0061] FIG. 20 is a timing chart showing the relationship between
the drive signals and the output signals as applied to the
horizontal transfer for the device of FIG. 19;
[0062] FIG. 21 is a schematic block diagram showing a four-line
readout solid state imaging device applied to the imaging unit of
FIG. 2;
[0063] FIG. 22 is a timing chart showing the relationship between
the drive signals and the output signals as applied to the
horizontal transfer for the device of FIG. 21;
[0064] FIG. 23 is a schematic block diagram showing an alternative
embodiment of a solid state imaging apparatus according to the
present invention;
[0065] FIG. 24 is a timing chart useful for understanding an
operational sequence of horizontal transfer consistent with
horizontal timing signals of the initial driving condition at the
time of high speed driving in the solid state imaging device of
FIG. 1;
[0066] FIG. 25 is a timing chart useful for understanding an
operational sequence of horizontal transfer consistent with
horizontal timing signals of the inverted-branching driving
condition at the time of high speed driving in the device of FIG.
1;
[0067] FIG. 26 is a timing chart also useful for understanding an
operational sequence of horizontal transfer at the time of
low-speed driving in the device of FIG. 1;
[0068] FIG. 27 schematically shows signal charges being transferred
responsive to the horizontal timing signals of the initial driving
condition at the time of transfer efficiency measurement on the
horizontal transfer path in the device of FIG. 1:
[0069] FIG. 28 schematically shows signal charges being transferred
responsive to the horizontal timing signals of the
inverted-branching driving condition at the time of transfer
efficiency measurement on the horizontal transfer path in the
device of FIG. 1;
[0070] FIG. 29 is a graph showing the relationship between the
quantity of residual transfer charges and the quantity of reference
signals on two branching transfer paths in the device of FIG.
1:
[0071] FIG. 30 is a graph showing the relationship between the
transfer efficiency and the quantity of the reference signals;
[0072] FIG. 31 is a timing chart useful for understanding the
operational sequence of mixing of horizontal pixels on a horizontal
transfer path before branching in the device of FIG. 1;
[0073] FIGS. 32A through 32I showing how the potential level
changes which is formed in the respective transfer elements by
horizontal pixel mixing on the horizontal transfer path before
branching in the device of FIG. 1;
[0074] FIGS. 33 and 34 are block diagrams schematically showing
changes in the output destination, consistent with the transfer
efficiency on the branching horizontal transfer paths in the device
of FIG. 1;
[0075] FIG. 35 schematically shows horizontal transfer paths in the
device of FIG. 1, looked from above;
[0076] FIG. 36, parts (A) and (B) are a schematic plan view and a
schematic cross-sectional view of one of the horizontal transfer
paths shown in FIG. 35, respectively;
[0077] FIG. 37, parts (A) and (B) are a schematic plan view and a
schematic cross-sectional view of the other of the horizontal
transfer paths shown in FIG. 35, respectively;
[0078] FIG. 38 is a timing chart showing the timing of drive
signals supplied to the respective electrodes shown in FIG. 35;
[0079] FIG. 39 is a schematic potential diagram showing the state
of transfer of signal charges on the horizontal transfer path in
FIG. 36;
[0080] FIG. 40 is a schematic potential diagram showing the state
of transfer of signal charges on the horizontal transfer path in
FIG. 37;
[0081] FIGS. 41A to 41E schematically show the state of transfer of
signal charges on the horizontal transfer path in FIG. 35;
[0082] FIG. 42 is a schematic timing chart showing an example of
the timing of drive signals supplied to the respective electrodes
shown in FIG. 35;
[0083] FIG. 43 is a schematic timing chart showing another example
of the timing of drive signals supplied to the respective
electrodes shown in FIG. 35;
[0084] FIG. 44 is a flowchart useful for understanding illustrative
processing for calculating the transfer efficiency;
[0085] FIGS. 45 and 46 schematically show illustrative processing
for calculating the transfer efficiency on one and the other of the
horizontal transfer paths, respectively;
[0086] FIG. 47 is a graph schematically showing the residual charge
quantities detected from one reference signal to another;
[0087] FIG. 48 is a graph schematically showing the transfer
efficiency calculated from the residual signal quantities shown in
FIG. 47;
[0088] FIG. 49 is a flowchart showing illustrative processing of
calculating the transfer efficiency by the processing sequence
shown in FIG. 44 and for setting a variable value of the duty cycle
using the transfer efficiency calculated;
[0089] FIG. 50 is a block diagram schematically showing a further
alternative embodiment of a digital camera employing the device of
FIG. 1;
[0090] FIG. 51 is a plan view looked from above, showing a
horizontal transfer path in the device of FIG. 1;
[0091] FIG. 52 is a partial cross-sectional view showing essential
part of a horizontal transfer path of FIG. 51;
[0092] FIG. 53 is a partial plan view of an array of offset pixels
and color filter segments as applied to the device of FIG. 1;
[0093] FIGS. 54A through 54E and 55A through 55E schematically show
transfer of signal charges R and B, and signal charges G on the
horizontal transfer path of FIG. 1, respectively;
[0094] FIG. 56 is a timing chart showing the supply timing of drive
signal to the electrode of FIG. 51;
[0095] FIGS. 57A through 61B schematically show the potential
levels generated on the horizontal transfer path when the drive
signal shown in FIG. 56 is applied;
[0096] FIG. 62 is a timing chart showing the timing of drive
signals supplied to the electrodes of FIG. 55 on transferring the
signal charge G;
[0097] FIGS. 63A through 64B schematically show the potential level
generated on applying the drive signal shown in FIG. 62;
[0098] FIGS. 65 and 66 are timing charts showing drive signals for
providing the same color of the signal charges output from the same
output amplifier in the first and second lines; and
[0099] FIGS. 67A to 67E schematically illustrate the transfer of
signal charges G in two horizontal transfer paths.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0100] A solid state imaging device according to the present
invention will now be described with reference to the accompanying
drawings. With reference to FIG. 1, in a preferred embodiment of a
solid state imaging device 44, signal charges of attributes of
colors classified by a branching section 54 are transferred with a
horizontal transfer speed or rate higher than a predetermined
transfer speed, on both of the horizontal transfer paths 56 and 58,
and converted analog voltage signals 82 and 84 are output
simultaneously. For transfer with a horizontal transfer speed lower
than the predetermined transfer speed, the converted analog voltage
signal 82 is output from the selected horizontal transfer path,
such as path 56. The output amplifiers 60 and 62 arranged on
plural, e.g. two, horizontal transfer paths 82 and 84 exhibit
differential sensitivities for signal charge detection, depending
on the attributes of colors of signal charges supplied. By
outputting the analog voltage signals 82 and 84 under this
condition, it is possible to modulate the sensitivity of red and
blue signals with respect to the green color to suppress unneeded
gain to suppress noise.
[0101] In the present embodiment, the solid state imaging device of
the present invention is applied to a digital camera 10. The parts
or components not directly pertinent to understanding the present
invention are not shown nor described.
[0102] Referring to FIG. 2, the digital camera 10 includes an
optical system 12, an imaging unit 14, an amplifier power supply
16, a biasing circuit 18, drivers 20, a pre-processor 22, a memory
24, a signal processor 26, a system controller 28, an operating
unit 30, a timing signal generator 32, a media interface (I/F)
circuit 34, media 36 and a monitor display 38, which are
interconnected as illustrated.
[0103] The optical system 12 has an automatic focusing (AF)
function of receiving incident light 40 from a field being imaged
to focus an image of the field on the imaging surface of the
imaging unit 14 responsive to an operation from the operating unit
30. The optical system 12 adjusts the angle of view or focal
length, responsive to a zooming operation or a half-stroke
depression of a shutter release button, not shown, on the operating
unit 30. The optical system 12 also has the function of adjusting
the diaphragm or iris stop for the incident light 40 to a value in
keeping with the manipulation on the operating unit 30 in the
imaging unit 14. The optical system 12 adjusts the incident light
40 to a light beam 42 based on these functions to focus the light
beam 42 onto the imaging unit 14.
[0104] The imaging unit 14 includes a solid state imaging device
44, which is implemented by a charge-coupled device (CCD) shown in
FIG. 1 with the illustrative embodiment. The imaging device 44
includes color filter segments, as shown in FIG. 14, arranged in
the incoming direction of the incident light beam 42 according to
the mounting positions of photosensitive cells. The device 44 has
the function of color-separating the incident light beam 42,
converting the light of the color components resulting from the
color separation into signal charges by photosensitive cells 46 and
outputting corresponding electrical signals. The device 44 reads
out signal charge accumulated in proportion to light exposure, to a
vertical transfer path 48 for transferring sequentially in the
vertical direction. The device 44 includes a horizontal transfer
path 50 extending in a direction substantially perpendicular to the
vertical transfer path 48. The signal charges, transmitted
vertically, are supplied to the horizontal transfer path 50.
[0105] The horizontal transfer path 50 of the present embodiment
has its output end 52 including a branching section 54. From the
branching section 54, there are formed horizontal transfer paths 56
and 58 in a branching fashion. Separate output amplifiers 60 and 62
are provided at output ends of the horizontal transfer paths 56 and
58, respectively. The output amplifiers 60 and 62 are floating
diffusion amplifiers having the function of converting signal
charges into a corresponding analog voltage signal. To the output
amplifiers 60 and 62 are connected power supply lines 64 and 66,
respectively. The power supply lines 64 and 66 are connected
independently from the power supply 16. The output amplifiers 60
and 62 are supplied from the drivers 20 with reset signals 68 and
70, respectively. The output amplifiers 60 and 62 thus supplied
with the reset signals may run independently. In the description to
follow, signals are denoted by reference numerals of connection
lines on which they appear.
[0106] The branching section 54 is supplied with a bias signal 72
from the biasing circuit 18. With the bias signal thus supplied,
the signal charges from the horizontal transfer path 50 are
branched to the horizontal transfer path 56 or 58. A horizontal
drive signal 74 is supplied to the horizontal transfer path 50,
while a horizontal drive signal 76 is supplied to each of the
horizontal transfer paths 56 and 58. The horizontal drive signal 76
has a frequency equal to, for example, one half the frequency of
the horizontal drive signal 74. By driving the horizontal transfer
paths 56 and 58 in this manner, high speed readout may be enabled
even though the design frequency ranges of the output amplifiers 60
and 62 are halved. The device 44 is also supplied with an overflow
drain (OFD) pulse 78 and with a vertical drive signal 80.
[0107] In this manner, the device 44 outputs two channels of output
signals 82 and 84 from the output amplifiers 60 and 62,
respectively to the pre-processor 22. The horizontal transfer in
the device 44 will be described in detail subsequently.
[0108] Reverting to FIG. 2, the amplifier power supply 16 has the
function of supplying the supply power to the output amplifiers 60
and 62 arranged in the device 44. The amplifier power supply 16
supplies the supply power depending on whether the device 44 is to
supply a one-channel output or a two-channel output. This power
supply is controlled by a control signal 86 supplied from the
signal processor 26 to the amplifier power supply 16.
[0109] The biasing circuit 18 has the function of supplying a bias
signal 72 to the branching section 54. The bias signal 72 is
applied as a bias voltage which prescribes the gain. The biasing
circuit 18 is controlled by a control signal 88 supplied from the
signal processor 26.
[0110] The drivers 20 have the function of supplying a variety of
drive signals for driving the device 44. The drivers 20 are
supplied with plural timing signals 90 from the timing signal
generator 32. The drivers include an OFD pulse output circuit 92, a
vertical (V) driver 94, a horizontal series (HS) driver 96, a
horizontal parallel (HP) driver 98, and a reset (RS) driver 100, as
shown in FIG. 3. The OFD pulse output circuit 92 outputs an OFD
pulse 78 to the device 44. The V driver 94, HS driver 96 and HP
driver 98 output a vertical drive signal 80, the horizontal drive
signal 74 and the horizontal drive signals 76 to the device 44,
respectively. The horizontal drive signal 76 has a period double
that of the horizontal drive signals 74. The RS driver 100 outputs
the reset signals 68, 70 to the device 44.
[0111] Reverting to FIG. 2, the pre-processor 22 has an analog
front-end (AFE) function. This function removes the noise by
correlated double sampling (CDS) for the analog electrical signals
82, 84 supplied, while digitizing the noise-free analog electrical
signals by analog-to-digital (A/D) conversion. The pre-processor 22
is supplied from the timing signal generator 32 with a timing
signal or with sampling signals 106 and 108 for performing
pre-processing on the input signals of respective channels for
noise removal and A/D conversion. It is noted that two channels of
analog electrical signals 82 and 84 are supplied to the
pre-processor 22. In case only one of the two channels is supplied
with an input, and one of the sampling signals 106 and 108 for the
input channel is supplied for the sole channel thus receiving the
input, it may be sufficient to activate only one channel of the
operation of the CDS sampling and A/D conversion, thereby reducing
power consumption. The pre-processor 22 is responsive to the supply
of the sampling signals 106 and 108 to output one or two of
channels of digital signals 110 and 112 to the memory 24.
[0112] The memory 24 has the function of temporarily storing the
digital signals 110 and 112 supplied and outputting the stored
signals. Specifically, a line memory is provided for each channel
in the memory 24. The memory 24 has its input and output controlled
by a control signal 116 supplied over a bus 114. The memory 24 is
responsive to the control signal 116 to output the input digital
signals 110 and 112 as a digital signal 118 over the bus 114 and
signal line 120 to the signal processor 26.
[0113] The signal processor 26 has the function of processing the
digital signal 118 supplied to generate a control signal. The
signal processor 26 includes a power supply control 122, a gain
control 124, an AF control 126, an automatic exposure (AE) control
128, an automatic white balance (AWB) control 130 and an data
converter 132. The power supply control 122 has the function of
generating a control signal 86 depending on high speed readout or
low speed readout, based on, for example, scene discrimination in
the system controller 28. The power supply control 122 outputs the
generated control signal 86 to the power supply 16.
[0114] The gain control 124 has the function of generating a
control signal 88 depending on which of the horizontal transfer
paths 56 and 58 the signal charge from the horizontal transfer path
50 via the branching section 54 are to be supplied to. The gain
control 124 outputs the control signal 88 to the bias control 18.
The bias control 18 routes the bias signal 72 to the branching
section 54. The AF control 126 has the function of adjusting the
focus based on produced image data. The AE control 128 has the
function of finding an evaluation value based on the produced image
data, for adjusting the diaphragm and the shutter speed. The AF
control 126 and the AE control 128 are responsive to adjustment to
transmit a control signal, not shown, over signal line 120, bus 114
and signal line 134 to the system controller 28. The AWB control
130 has the function of adjusting the white balance based on
produced image data. The data converter 132 has the function of
converting the image data, obtained on high-speed readout as
two-channel data, into, e.g. dot sequential image data
corresponding to the array of the color filter segments, and
forming one frame of image. In case the output from the
pre-processor 22 is obtained by low speed readout, the data
converter 132 is to cope with one-channel output, and accordingly
performs the positional conversion for a one-channel output.
[0115] The signal processor 26 also has the function, not shown, of
simultaneously outputting the three-color data supplied, and using
the image data of the three primary colors obtained simultaneously
to produce luminance/chrominance (Y/C) signals, which is sometimes
referred to as synchronization. The signal processor 26 also has
the function of converting the produced Y/C signals into
displayable signals, such as signals appropriate for a liquid
crystal display monitor. The signal processor 26 also has the
function of compressing the Y/C signals produced, depending on the
recording modes, and decompressing the compressed signals for
restoring and reproducing the image data. The recording modes may
be exemplified by JPEG (Joint Photographic Experts Group), MPEG
(Moving Picture Experts Group) and raw data modes. The signal
processor 26 transmits the image data processed in accordance with
the recording modes over a signal line 120, a bus 114 and a signal
line 136 to the media I/F circuit 34. The signal processor 26 also
outputs a liquid crystal monitor signal 138 to the monitor display
38.
[0116] The system controller 28 has the function of generating a
variety of control signals responsive to an operating signal 140
from the operating unit 30, as later described. The system
controller 28 includes a setting and operational control functional
unit, not shown. This setting and operational control functional
unit receives the control signal 140 from the operating unit 30 as
a setting condition to generate a control signal 142 depending on
the setting condition. The setting and operational control
functional unit generates the control signal 142 which controls
whether an output of horizontal transfer is to cause a double
output or a single output. Thus, the system controller 28 verifies
whether or not the horizontal transfer is adapted for high speed
readout, depending on moving picture mode setting, speed setting of
repeated shooting, scene decision and a depressing operation for a
shutter release button, not shown. The system controller 28 then
outputs the control signal 142 generated to the timing signal
generator 32. In addition, the system controller 28 controls the
memory 24, the signal processor 26 and the media I/F circuit
34.
[0117] The operating unit 30 includes a power supply switch, a zoom
button, a menu display selector switch, a select key, a moving
picture mode setting unit, a repeated shooting speed setting unit
and a shutter release button which are not shown. The power supply
switch is used for turning on or off the power supply of the
digital camera 10. The zoom button controls the angle of view of
the field being imaged, inclusive of a subject being imaged, and
adjusts the focal length of the subject responsive to the control.
The menu display selector switch is used for selecting the menu
displayed on the liquid crystal monitor display to cause movement
of a select cursor, and may, for example, be a cross switch. The
selector key is used for selecting the items of the menu
displayed.
[0118] The moving picture setting unit sets whether or not a moving
picture is to be displayed on the liquid crystal monitor display,
using, e.g. a flag. By this setting, the digital camera 10 displays
an image of the field captured on the monitor display 38 in the
form of through-image. The moving picture setting unit includes
items for setting the resolution, the number of displayed frames
and the repeated shooting speeds. The items for resolution are
those for selecting the resolution of, for example, HDTV
(High-Definition Television) standard/the reference VGA (Video
Graphics Array). The number of displayed frames is an item for
selecting one of 30 and 15 (30/15).
[0119] The repeated shooting speed setting unit sets the speed in
repeated shooting from plural repeated shooting speeds as set. It
sets the speed in repeated shooting depending on two inputs/one
input. The number of times of repeated shooting is set for an image
with a predetermined number of pixels. One input or two inputs is
selected, respectively, in driving the solid-state imaging device,
depending on that the number of times of repeated shooting is less,
or equal to or greater than a threshold value of the number.
[0120] The shutter release button has the function of selecting the
operational timing or the operational mode of the digital camera 10
responsive to half-stroke/full-stroke depressions. The shutter
release button initiates the AE and AF operations responsive to the
half-stroke depression. The AE and AF operations use an image
obtained on moving picture display to find out a diaphragm, a
shutter speed and a focal length which may be optimum. The shutter
release button also sends the recording start/recording end timing
to the system controller 28, as an operating signal 140, by the
full-stroke depression, to define an operational timing responsive
to the setting mode of the digital camera 10. The setting modes may
be exemplified by, for example, a still image recording mode and a
moving picture recording mode.
[0121] The timing signal generator 32 has the function of
generating a variety of timing signals for the device 44 of the
imaging unit 14. These timing signals may be exemplified by
vertical and horizontal synchronous signals, field shift gate
signals, vertical and horizontal timing signals, OFD pulse signals
and reset signals. This function generates a variety of timing
signals 90, 106 and 108 responsive to the control signal 142 from
the system controller 28. The timing signal generator 32 outputs a
variety of the timing signals 90 to the drivers 20. The timing
signal generator 32 has the function of generating a reference
clock signal and, in particular, generates the horizontal timing
signal. The timing signal generator 32 frequency-divides the
horizontal timing signal to generate two horizontal timing signals
of frequencies different from each other. The timing signal
generator 32 is responsive to the control signal 142 from the
system controller 28 to output one of the sampling signals 106 and
108 as one-channel output. In this manner, it is possible to
suppress power consumption of the digital camera 10.
[0122] The media I/F circuit 34 has the interface control function
of controlling the recording and/or reproduction of image data
depending on the recording media which may be in use. The media I/F
circuit 34 is able to control recording/readout of image data 144
for a PC (Personal Computer) card as a semiconductor recording
medium. The media I/F circuit 34 is also able to control
recording/readout of image data 136 supplied over bus 114 under
control by an enclosed USB (Universal Serial Bus) controller. There
are a variety of standards for the semiconductor memory cards used
for the recording media 36.
[0123] For the monitor display 38, a liquid crystal monitor
display, for example, may be used. The image data 138 supplied from
the signal processor 26 is displayed on the monitor display 38.
[0124] With the above-described configuration, the digital camera
10 may be run satisfactorily as signal charges from the horizontal
transfer path 50 are read out as two inputs at a high speed or as
one output at a low speed.
[0125] The electrode structure in which the horizontal transfer
path 50 is bifurcated at the branching section 54 into the
horizontal transfer paths 56 and 58, and the transfer of signal
charge responsive to a drive signal, will now be described. For
facility in description, the side of the horizontal transfer path
56 and that of the horizontal transfer path 58 are separately
described. For each of the horizontal transfer paths 50, 56 and 58,
each transfer element is formed by paired two electrodes of
polycrystalline silicon (polysilicon) and paired two impurity
layers in the vicinity of a silicon substrate. A stepped potential
gradient is formed by applying a drive signal of the same potential
to the two electrodes.
[0126] From the right-hand side towards an electrode HSL of the
branching section 54, there are sequentially formed polysilicon
electrodes HS4, HS3, HS4, HS1, HS2, HS3, HS2 and HS1 on the
horizontal transfer path 50 with each polysilicon electrode being
formed by an electrode pair. Referring to FIG. 4, four polysilicon
electrodes HP1, HP2, HP1 and HP2, and an OG (Output Gate) electrode
are formed from the electrode HSL of the branching section 54
towards an output amplifier 60. Adjacent to the left side of the OG
electrode, there is formed a floating diffusion (FD) layer.
Adjacent to the left end of the FD layer is formed a reset
electrode (RS). Ultimately, next to the left end of the reset
electrode is formed a reset drain (RD) layer.
[0127] Next, the above electrodes are imaginarily sectioned,
beginning from the reset drain RD on the left end to the electrode
HP1 on the horizontal transfer path 56 and further from the
branching section 54 to the electrode HS2 on the horizontal
transfer path 50 for illustrating the sectioned surfaces as shown
by a chain-dotted line IV-IV. As may be seen from these sectioned
surfaces, an impurity layer is formed directly below each electrode
on a P-type substrate not shown. A plural number of the impurity
layers are formed in register with the respective electrodes. For
generating the respective impurity layers, impurities are doped by
using, e.g. an ion implantation method. The sorts and the
concentrations of the impurities doped characterize the magnitudes
of the potential gradients. The potential gradients of preset
magnitudes are determined in dependence upon the voltage level of
the drive signals applied to the electrodes formed directly on the
top of the impurity layers.
[0128] A variety of drive signals supplied to the respective
electrodes will now be described. A drive signal .phi.HS2 is
supplied to the electrode HS2. A drive signal .phi.HSL is supplied
to the electrode HSL. The drive signal .phi.HSL is a constant bias
voltage. Drive signals .phi.HP1 and .phi.HP2 are supplied to the
electrodes HP1 and HP2, respectively. Drive signals .phi.OG,
.phi.RS and .phi.RD are supplied to the electrodes OG, RS and the
reset drain RD, respectively.
[0129] FIG. 5 shows the timing for these drive signals. As regards
the phase of each of the drive signals, the drive signals .phi.HS1
and .phi.HS3 of FIG. 5, part (A), are two phase drive signals
phase-reversed by 180.degree. from the drive signals .phi.HS2 and
.phi.HS4 shown in part (B). The drive signal .phi.HP1 of part (C)
and the drive signal .phi.HP2 of part (D) are phase-reversed from
each other and are each a two-phase drive signal.
[0130] Turning to the period of the drive signal, the drive signals
of the sets of FIG. 5, parts (A) and (B), are of the period
one-half that of the drive signals of the sets of parts (C) and
(D). That is, frequency of the drive signals of the sets of parts
(A) and (B) is twice that of the drive signals of the sets of parts
(C) and (D). The drive signal .phi.RS supplies a level "H" at
timings of, e.g. t=1, t=5, . . . , that is, at timing of t=4n+1,
where n is a variable inclusive of zero. Output signals OS1 and OS2
are supplied as indicated in part (F).
[0131] Reverting to FIG. 4, since the drive signal .phi.HSL is
supplied in a manner not shown, there are generated, in a region
directly below the electrode HSL supplied with the drive signal
.phi.HSL, a potential level of a perpetually fixed reference level
146 and a potential level or barrier 148 prohibiting reverse flow
of signal charges supplied from the horizontal transfer path
50.
[0132] The potential generated responsive to the drive signals
supplied, and the movement of signal charges accompanying the
potential generation will now be described. Signal charges having
attributes of colors red (R), green (G) and blue (B) are termed
signal charge R, G and B, respectively. Referring to FIG. 5, part
(B), the drive signal .phi.HS2 is at a level "L", and the drive
signals .phi.HSL and .phi.HSP1 are supplied. The drive signal
.phi.HS1 is at a level "L". When the drive signals are supplied in
this manner, the signal charge R is retained in the branching
section 54. The level "L" is supplied at this time to the impurity
layer of the electrode HP1, not shown, adjacent to the electrode
HSL, so that there is generated a potential or a barrier which is
just high enough to prohibit, for example, the signal charge R from
mixing into the horizontal transfer path 56, as indicated by a
broken line 150.
[0133] The drive signal .phi.HP2 of the level "H" is supplied to
the other electrode HP2 neighboring to the branching section 54.
This generates a potential level 152 lower than the reference level
to permit the signal charge R to flow into the horizontal transfer
path 58. At this time, the signal charge R is present in both
packets of the reference level 146 and the potential level 152.
[0134] There are formed impurity layers 154 and 156 directly
underneath the electrodes HP2 and HP1. When the level "H" is
supplied, there are generated stepped potential levels composed of
a level lower than the reference level 146 and the lowest level.
When the level "L" is supplied, there are generated stepped
potentials composed of a level higher than the reference level 146
and the same level as the reference level 146. Hence, the potential
generated becomes sequentially lower in a stepped fashion along the
transfer direction of the signal charges. At a timing t=1, signal
charges G are stored at every second electrodes on the horizontal
transfer path 56.
[0135] At the next timing t=2, the drive signal .phi.HS2 at a level
"H" is applied to the electrode HS2 as shown in FIG. 5, part (B).
This causes the impurity layers of the electrode HS2 to generate a
potential level 148 and the reference level 146. This allows the
electrode HS2 to generate a packet between it and the electrode
HSL. This packet holds the signal charge G. The rear side
electrodes are supplied with the drive signals of the same level as
at the timing t=1. Hence, the potential levels are the same as
those at timing t1. In the interim, the signal charge R at the
electrode HSL is moved to the electrode HP2 on the horizontal
transfer path 58 on a lower side not shown, on the figure sheet.
The signal charge R in this state is indicated by a broken
line.
[0136] At the next timing t3, the drive signal .phi.HS2 of the
level "L" is applied to the electrode HS2 as shown in FIG. 6. This
sets the potential levels at the timing t1. The signal charge G
held in the packet generated at the timing t=2 is moved to the
reference level 146 at the branching section 54. At this time, the
drive signal .phi.HP1 of the level "H" is supplied to the electrode
HP1 of the horizontal transfer path 56 neighboring to the electrode
HSL. The potential generated in the impurity layer associated with
the electrode HP1 is higher in level than the reference level 146,
as indicated by a broken line 158. As a result, the signal charge G
is present in both packets of the reference level 146 and the
potential level 160. At this time, the drive signal .phi.HP2 of the
level "L" is supplied to the electrode HP2 of the horizontal
transfer path 58. This generates the potential level of a broken
line 158 in the electrode HP2. This potential level prohibits the
signal charge G from mixing into the horizontal transfer path 58.
The signal charge G at the branching section 54 is moved to the
packet formed at the electrode HP1 on the horizontal transfer path
56 on an upper part of the drawing sheet.
[0137] The drive signal .phi.HP2 of the level "L" is applied to the
electrode HP2 neighboring to the electrode HP1. This generates the
potential level 148 and the reference level 146 in the impurity
layers 154, 156, respectively. The drive signal .phi.HP1 of the
level "H" is applied to the electrode HP1 neighboring to the
electrode HP2. This generates a level lower than the reference
level 146 and the lowest potential level in the impurity layers 154
and 156, respectively. The potential level 148 and the reference
level 146 are generated by the level "L" supplied to the
neighboring electrode HP2. As a result, the signal charge G held in
the packet at t=2 is moved to and retained in the packet generated
in the electrode HP1.
[0138] The signal charge G, retained in the packet, generated at
the timing t=2, is moved towards the output side, and thence
transferred via electrode OG because of the increment of the
potential level.
[0139] Next, at a time t=4, the drive signal .phi.HS2 of level "H"
is supplied to the electrode HS2, so that, at this electrode, the
potential which is the same as that at time t=2 is generated. The
signal charge B is retained in the packet generated at this time.
The signal charge G at the branching section 54 is moved to the
packet formed directly underneath the electrode HP1 on the
horizontal transfer path 56. A drive signal of the same level as
that at time t=3 is supplied to all of rear side electrodes. Hence,
the potential levels generated are the same as those at time
t=3.
[0140] Next at time t=5, the impurity layers in register with the
electrode HS2 generate the same potential levels as at time t=1.
This generates a potential level 158, directly underneath the
electrode HP1, neighboring to the branching section 54, thus
providing a potential barrier against the signal charge B. The
signal charge B may be made not to be color-mixed into the
horizontal transfer path 56. The branching section 54 causes the
incoming signal charge B to be moved further to the horizontal
transfer path 58. The horizontal transfer path 56 is supplied with
the drive signals which are of the same level as at time t=1.
Hence, the potential levels generated are the same as that at time
t=1. At time t=4, the signal charge G supplied to the FD layer is
converted into an analog voltage signal which is output to the
output amplifier 60.
[0141] Next, the above electrodes are imaginarily sectioned,
beginning from the reset drain RD on the left end to the electrode
HP2 on the horizontal transfer path 58 and further from the
branching section 54 to the electrode HS2 on the horizontal
transfer path 50, for illustrating the sectioned surfaces. As may
be seen from these sectioned surfaces, an impurity layer is formed
directly underneath each electrode on a P-type silicon substrate.
The P-type silicon substrate is not shown. A plural number of the
impurity layers are formed in register with the respective
electrodes. In generating the impurity layers, the concentration of
each of the impurity layers is adjusted to generate preset
potential levels depending on the voltage levels of the drive
signals. The horizontal transfer path 58 is featured by having one
more electrodes than the horizontal transfer path 56.
[0142] At time t=1, the drive signal (.phi.HS2 of the level "H",
the drive signal .phi.HSL of the constant bias voltage and the
drive signal .phi.HP1 of the level "L" are supplied to the
electrodes of the horizontal transfer path 58, as shown in FIG. 5,
part (B). When the drive signals are applied in this manner, a
signal charge R is retained in the branching section 54. At this
time, the potential level generated by the impurity layer of the
electrode HP2, not shown, adjacent to the electrode HSL, on
applying the drive signal .phi.HP2, is lower by one step than the
reference level 146. The potential level 150 generated directly
underneath the electrode HP1 on the horizontal transfer path 56,
operates as a potential barrier, and prevents mixing of the signal
charge R.
[0143] A sum total of four electrodes, that is, the electrodes HP1
and HP2 are provided next to the electrode HP2. Hence, the number
of the electrodes provided on the horizontal transfer path 58 is
one more than that on the horizontal transfer path 56. The impurity
layers 154 and 156 of FIG. 4, for example, are sequentially
provided, when looking from the right side, as the impurity layers
lying directly underneath the four electrodes. Since the drive
signal .phi.HP1 of the level "L" is supplied to the electrode HP1,
the potential level 148 and the reference level 146 are formed
directly underneath the electrode HP1. The drive signal .phi.HP2 of
the level "H" is supplied to the electrode HP2. This generates the
level by one step lower than the reference level 146 and the lowest
potential level directly underneath the electrode HP2.
[0144] At time t=1, the drive signals are supplied as described
above, and hence a packet is generated directly underneath the
electrode HP2. The signal charge R and G are retained in packets
sequentially from the branching section 54.
[0145] Next, at time t=2, the drive signal .phi.HS2 of the level
"H" is applied to the electrode HS2 as shown in FIG. 5, part (B).
This causes the impurity layers of the electrode HS2 to generate
potential levels which are of the same level as that at time t=2 of
FIG. 4 to generate a packet. The signal charge G is retained in
this packet. From this time on, the drive signals of the same level
as at time t=1 are supplied to the electrodes of the horizontal
transfer path 58. Hence, the potential levels generated are the
same as those at time t=1.
[0146] The potential levels at t=3 are shown in FIG. 8, part (B).
At time t=3, the drive signal .phi.HS2 is applied at level "L" to
the electrode HS2. This sets the potential levels which are the
same as those at time t=1. The signal charge G retained in the
packet directly below the electrode HS2 at time t=2 is moved to the
branching section 54 of the reference level 146. At this time, the
drive signal .phi.HP2 is applied at level "L" to the electrode HP2
on the horizontal transfer path 58 neighboring to the electrode
HSL. The potential level for the electrode HP2 is slightly higher
than the reference level 146 as indicated by broken line 158, due
to the impurity layer associated with the electrode HP2 which is
not shown because of cross-section. That is a potential barrier is
formed to prevent the mixing of the signal charges G into the
horizontal transfer path 58. On the other hand, the potential
indicated by a broken line 160 is formed by the potential level "H"
supplied to the electrode HP1 of the horizontal transfer path 56.
This causes movement of the signal charge G along the direction
perpendicular to and into the drawing sheet, as indicated by arrow
162. There is generated a packet by the potential 160 directly
underneath the electrode HP1 supplied with the drive signal
.phi.HP1 on the horizontal transfer path 56 as indicated at time
t=1 in FIG. 4.
[0147] On the horizontal transfer path 58, the packet generated on
the electrode HP2 at time t=2 is formed on the electrode HP1,
responsive to the supply of the level "H" of the drive signal
.phi.HP1. The signal charges R and B are retained in the packets of
the electrodes HP1 sequentially from the branching section 54. The
signal charge R on the electrode HP2, retained in the packet,
generated at time t=2 is shifted towards the output with rise in
the potential level, and transferred to the FD layer via the
electrode OG.
[0148] Next at time t=4, there are generated the same potential
levels as those at time t=2. The signal charge B is retained in a
packet then generated. The drive signals of the same levels as
those at time t=3 are supplied to the rear side electrodes. Hence,
the potential levels generated are the same as those at time t=3.
The potential level directly underneath the electrode HP2
neighboring to the electrode HSL is higher than the reference level
146 as indicated by a broken line 158. On the other hand, the
potential level formed directly underneath the electrode HP1
neighboring to the electrode HSL is lower than the reference level
146 as indicated by a broken line 160.
[0149] Then, at time t=5, the same potential level as that at time
t=1 is generated. The signal charges R and B are retained in the
packets of the electrode HP2 sequentially from the branching
section 54. At time t=4, the signal charge R supplied to the FD
layer is converted into an analog voltage signal which is output
from the output amplifier 60.
[0150] The operating principle of horizontal transfer responsive to
supply of drive signals is shown in FIGS. 9A to 9E. In horizontal
transfer, the signal charges R_G1_B_G2 supplied at time t=1 from
the horizontal transfer path 50 to the branching section 54 are
distributed at the branching section 54 to the horizontal transfer
paths 56 and 58. The symbol _ denotes a potential barrier region.
It may be seen that, on the horizontal transfer path of FIG. 9, the
potential barrier separating the signal charge is generated for the
length of one electrode. The horizontal transfer path 56 transfers
only the signal charges G responsive to the drive signals supplied.
At the above time point, the potential barrier is formed at the
electrode HP1 of the horizontal transfer path 56 neighboring to the
branching section 54 to prevent the signal charge R from being
mixed into the horizontal transfer path 56. The horizontal transfer
path 58 transfers the signal charge R and B responsive to the drive
signals supplied.
[0151] The horizontal transfer path 50 is operated at a frequency
double that of the horizontal transfer paths 56 and 58. Thus, at
time t=2, the horizontal transfer path 50 horizontally transfers
one packet of the signal charge it holds, towards the branching
section 54, responsive to the drive signals supplied. Conversely,
on the horizontal transfer paths 56, 58, there is no change in the
transfer of signal charges, because the drive signals undergo no
level changes. However, the signal charge R in the branching
section 54 is moved to a packet generated in the electrode HP2,
because the potential level at the electrode HP2 is lower than the
reference level 146.
[0152] At time t=3, the horizontal transfer path 50 horizontally
transfers the signal charges it holds, by one packet each towards
the branching section 54. A signal charge G1 is retained in the
packet generated directly underneath the branching section 54 and
the electrode HP1 of the horizontal transfer path 56 neighboring to
the branching section 54. At this time point, there is generated a
potential barrier in the electrode HP2 on the horizontal transfer
path 58 neighboring to the branching section 54 to prohibit the
signal charge G1 from mixing into the horizontal transfer path 58.
The horizontal transfer paths 56 and 58 horizontally transfer the
signal charge it holds, towards the output amplifiers 60 and 62 on
the packet-by-packet basis. This transfers the signal charge G and
B to the FD layers of the output amplifiers 60 and 62 on the
horizontal transfer paths 56 and 58.
[0153] Then, at time t=4, the horizontal transfer path 50
horizontally transfers the signal charges it holds, towards the
branching section 54 by one packet. The signal charge G.sub.1 is
moved to a packet directly underneath the electrode HP1 of the
horizontal transfer path 56 neighboring to the branching section
54. The signal charge R is moved to a packet directly underneath
the electrode HP1 of the horizontal transfer path 58 neighboring to
the branching section 54.
[0154] At time t=5, the horizontal transfer paths 50, 56 and 58
horizontally transfer the signal charges by one packet towards the
output side. Thus, the output amplifiers 60, 62 simultaneously
convert the signal charges of the colors G and B into analog
voltage signals, which are then output as output signals OS1 and
OS2. This eliminates difference generated in the processing of the
output signals OS1 and OS2 with lapse of time.
[0155] Meanwhile, if the difference in the processing with lapse of
time is tolerable, the output signals OS1 and OS2 may be output
alternately.
[0156] By the above sequence of operations, it is possible to
classify signal charges having color attributes to transfer and
output the signal charges without color mixing. In general, it is
required of the solid state imaging device to read out signal
charge generated, at a high speed in order to cope with the
increasing number of pixels. This demand affects the frequency band
in the output amplifiers on the horizontal transfer path. The solid
state imaging device is difficult to drive at a frequency higher
than a preset frequency due to shortage in the frequency band.
However, with the device 44 of the instant embodiment, it is
possible to read out output signal charges from color to color
within a preset frequency band by bifurcating an output and
increasing the number of output channels even though the driving
frequency of the horizontal transfer path 50 is increased in order
to cope with the increasing number of pixels. That is, an improved
signal charge readout speed may be achieved.
[0157] The sensitivity of charge detection in the output amplifiers
60 and 62 will now be described. The output amplifier is divided
into a floating diffusion section FD and an amplifier section. The
sensitivity of charge detection basically depends on the parasitic
capacitance C.sub.fd of the floating diffusion section FD. This
parasitic capacitance C.sub.fd in turn depends on the sum of five
capacitances. These five capacitances are the PN
(Positive-Negative) junction capacitance C.sub.sub between the
floating diffusion section FD and the substrate, the parasitic
capacitance C.sub.o with the output gate OG terminal, the
capacitance of the reset RS terminal C.sub.r, the gate-drain
capacitance C.sub.d of an MOS (Metal Oxide Semiconductor)
transistor in an output amplifier connected to the section FD as a
source follower amplifier, and a gate-to-source capacitance C.sub.s
of the MOS transistor.
[0158] It is noted that the gate-to-source capacitance C.sub.s
appears to be smaller due to the source follower gain G. Thus, the
parasitic capacitance C.sub.fd may be expressed by
C.sub.fd=C.sub.sub+C.sub.o+C.sub.r+C.sub.d+C.sub.s (1-G).
[0159] If desired to provide a difference in sensitivity in charge
detection of the output amplifiers 60 and 62, several conditions
are involved. The first condition is to provide a difference in the
gate capacitances of the output amplifiers. For providing the
difference in the gate capacitances, the difference in the
sensitivity is varied mainly by the gate-drain capacitance C.sub.d.
The channel widths and channel lengths in the MOS transistors of
the output amplifiers 60 and 62 are W, w, L and l, respectively, as
shown in FIGS. 10A and 10B. Neither the ratio of the channel width
to the channel length W/L of the MOS transistor shown in FIG. 10A
nor the ratio of the channel width to the channel length w/l of the
MOS transistor shown in FIG. 10B is changed. By so setting, the
frequency response or the gain G is not changed appreciably.
However, the gate capacitance may be varied significantly. By this
variation, the difference in sensitivity in charge detection may be
afforded to the output amplifiers 60 and 62. That is, the output
amplifier 60 may have low sensitivity in charge detection, while
the output amplifier 62 may have high sensitivity in charge
detection.
[0160] The second condition is providing a difference in thickness
of a silicon nitride film (SiN) 166 formed on an N.sup.+ layer 164
in the floating diffusion section (FD). The thickness of the
nitride film 166 gives rise to the connection capacitance of the
nitride film 166 and the N.sup.+ layer 164, in a manner different
from the PN junction capacitance Csub. If the thickness of the
nitride film 164 is thicker than the nitride film 168 of FIG. 11B,
the junction capacitance is increased. The difference in
sensitivity in charge detection may be afforded to the output
amplifiers 60 and 62 by taking advantage of this feature. That is,
the output amplifier 60 is lowered in sensitivity in charge
detection, while the output amplifier 62 is raised in sensitivity
in charge detection.
[0161] The surface area in the floating diffusion section FD
represents a third condition in mainly varying the PN junction
capacitance C.sub.sub. This third condition takes advantage of the
fact that the surface of the PN junction is proportional to
parasitic capacitance. The output amplifier 60 may be decreased in
sensitivity in charge detection in proportion to the increase in a
surface area 170 of the nitride film as shown in FIG. 12A, while
the output amplifier 62 may be increased in sensitivity in charge
detection in proportion to the decrease in a surface area 170 of
the nitride film as shown in FIG. 12B.
[0162] In addition, a fourth condition is provided as a special
condition of the second condition in prescribing the junction
capacitance between the nitride film 166 and the N.sup.+ layer 164.
The fourth condition is the presence or absence of the nitride
film. With the output amplifier 60, the sensitivity in charge
detection is increased by forming the nitride film 166 for the
floating diffusion section FD based on this condition as shown in
FIG. 13A. With the output amplifier 62, the sensitivity in charge
detection is decreased by not forming the nitride film 166 for the
floating diffusion section FD, that is, by forming only the N.sup.+
layer 164 as shown in FIG. 13B.
[0163] The imaging device 44 of the present embodiment is of a
so-called honeycomb array. More specifically with reference to FIG.
14, the photosensitive cells 46 are arrayed in the same row
direction at a pitch PP and in the same column direction at the
same pitch PP, while the photosensitive cells 46 of a row or a
column neighboring to a given row and a given column of the
photosensitive cells 46, respectively are arrayed with a shift of
one-half pitch in both the row and column directions. A color
filter formed on the incident light side of the photosensitive
cells 46 is of three primary colors R, G and B, and constituted by
plural color segments R, G and B. The color segments G are arrayed
in a square pattern, while the color segments R and B are arrayed
in a complete RB checkered pattern. That is, the filter array is a
so-called G-square RB-checkered pattern. With the pixels or
photosensitive cells 46 arrayed with offset as described above, a
plural number of the vertical transfer paths 48 are formed
meandering such as bypassing the pixels.
[0164] The signal charges as read out are transferred on the
vertical transfer paths 48 towards the horizontal transfer path 50
not shown in FIG. 14, responsive to eight-phase drive signals
.phi.V1B, .phi.V2, .phi.V3B and .phi.V4 to .phi.V8. The signal
charges are transferred towards the horizontal transfer path 50 by
using line memory LM. Although not shown in FIG. 14, electrodes
HS1, HS2, HS3, HS2, HS1, HS4, HS3 and HS4, . . . are provided on
the horizontal transfer path 50, when looking from its left end.
With the honeycomb array and the G-square RB checkered pattern,
signal charge are re-arrayed or re-positioned in an output sequence
by taking advantage of the line memory LM.
[0165] This re-arraying may be achieved using a drive signal
.phi.LM in FIG. 15, part (A), supplied to the line memory LM and
drive signals .phi.HS1 to .phi.HS4 in parts (B) to (E), supplied to
the electrodes of the horizontal transfer path 50. For this
re-arraying, the drive signals .phi.HP1 and .phi.HP2 not temporally
changed in level, as shown in FIG. 15, parts (F) and (G), and FIG.
16, parts (F) and (G) are supplied to the horizontal transfer paths
56 and 58. This does not activate the horizontal transfer paths 56
and 58.
[0166] The timing chart of FIG. 15 shows re-arraying or
re-positioning of the first field during the horizontal blanking
(HBL) period. Initially, the drive signal .phi.LM of FIG. 15, part
(A), becomes "L" in level at time 174. At this time, only drive
signal .phi.HS2 of part (C) is at the level "H". Signal charges are
transferred from the line memory LM to the packet generated
directly underneath the electrode HS2 of the horizontal transfer
path 50 supplied with the drive signal .phi.HS2.
[0167] The drive signal .phi.HS1 of FIG. 15, part (B), supplied to
the electrode HS1, then goes "H" in level. This generates a packet
directly underneath the electrode HS1 to cause movement of the
signal charge. The drive signal .phi.HS4 of part (E), supplied to
the electrode HS4, then goes "H" in level. This generates a packet
directly underneath the electrode HS4 to cause movement of the
signal charge. The drive signal .phi.HS3 of part (D), supplied to
the electrode HS3, then goes "H" in level. This generates a packet
directly underneath the electrode HS3 to cause movement of the
signal charge. The drive signal .phi.LM of part (A) goes "L" at
time 176. Only the drive signal .phi.HS1 of part (B) goes "H" in
level. The signal charge are supplied to and retained in this
manner in the packet.
[0168] The timing chart of FIG. 16 shows re-arraying of the second
field during the horizontal blanking (HBL) period. Initially, the
drive signal .phi.LM of FIG. 16(A) goes "L" at time 178. Only the
drive signal .phi.HS4 of FIG. 16, part (E) is "H" in level. The
signal charges are supplied from the line memory LM to a packet
generated directly underneath the electrode HS4 on the horizontal
transfer path 50 supplied with the drive signal .phi.HS4. The level
"H" is then supplied in the sequence of parts (B), (C), (D), (C)
and (B), depending on the electrode array. This causes movement of
the signal charge with the movement of the packet generated. That
is, the first signal charge is sequentially moved in the sequence
of the electrodes HS4, HS1, HS2, HS3, HS2 and HS1. The first signal
charge is transiently retained in the electrode HS1.
[0169] The drive signal .phi.LM of FIG. 16, part (A), becomes "L"
in level at time 180. At this time, only the drive signal .phi.HS3
of part (E) is at a level "H". The second signal charge is
transferred from the line memory LM to the packet generated
directly underneath the electrode HS3 on the horizontal transfer
path 50 supplied with the drive signal .phi.HS3. The level "H" then
is supplied in the order of parts (C) and (B) depending on the
electrode array. In this time sequence, the packet of the second
signal charge readout at time 180 is moved to the electrode HP2.
Substantially simultaneously with this charge movement, the first
signal charge retained in the second field is moved to the
electrode HS4. The first signal charge thus retained by the
electrode HS4 is then moved to the electrode HS3.
[0170] By this re-arraying, the two rows of signal charges RGBGRGBG
. . . , read out from the lowermost end in FIG. 14 are put into
order as the first field. The two rows of signal charges BGRGBGRG .
. . , read out from above the two bottom rows are put into order as
the second field.
[0171] This re-arraying is a technique used for the G-square RB
complete checkered pattern in a honeycomb array. Any other arraying
pattern may give rise to unneeded re-arraying or differential
timing. This re-arraying may be used for routine square pixels.
[0172] After the re-arraying, the drive signals .phi.HS1, .phi.HS3
and .phi.HS2, .phi.HS4 shown in FIG. 17, parts (A) and (B) are
supplied to transfer the signal charges retained on the horizontal
transfer path 50 to the horizontal transfer paths 56 and 58. The
drive signals .phi.HP1, .phi.HP2 shown in parts (C) and (D) are
supplied to the horizontal transfer paths 56 and 58. The start
position is the position at which the potential is initially
changed from the horizontal blanking period. The start positions of
the drive signals .phi.HP1, .phi.H2 are compared to those of the
drive signals .phi.HS1, .phi.HS3, .phi.HS2 and .phi.HS4 and the
drive signals .phi.HP1, .phi.HP2 are started within the period
intervals of the drive signals .phi.HS1, .phi.HS3, .phi.HS2 and
.phi.HS4 with a delay of one-half periods. Directly before
outputting, the reset signal .phi.RS shown in part (E) is applied,
whereby output time domains 182 to 188 are obtained.
[0173] The horizontal transfer path 50 transfers, as signal
charges, a dummy D1, a dummy D2, an optically black pixel OB1, an
optically black pixel OB2, R, G, B, G, . . . , in this order.
[0174] The solid imaging device 44 converts the signal charges into
analog voltage signals to output the dummy D3, optically black
pixel OB2, G, G, . . . , in the output time domains 182 to 188, as
output signals OS1 of FIG. 17, part (F). The device 44 also outputs
the dummy D1, optically black pixel OB1, R, B, in the output time
domains 182 to 188, as output signals OS2 of part (G). By supplying
the timings in this manner, the output signal OS1 on the horizontal
transfer path 56 outputs the color G, while the output signal OS2
on the horizontal transfer path 58 outputs the color R/B.
[0175] If desired to switch the outputs, the drive signals .phi.HS1
and .phi.HS3 shown in FIG. 17, part (H), and the drive signals
.phi.HS2 and .phi.HS4 shown in part (I) may be supplied with a
delay of start positions corresponding to one period interval 190
of these drive signals. The device 44 outputs the dummy D1,
optically black pixel OB1, R, B, in the output time domains 182 to
188, as output signal OS1 of part (J). The device 44 also outputs
the dummy D2, optically black pixel OB2, G, G, . . . , in the
output time domains 182 to 188, as output signal OS2 of part
(K).
[0176] In the present embodiment, the technique of delaying the
drive signal .phi.HS is used. This technique is not to be
interpreted as restrictive and the supply of the drive signal
.phi.HS may be started at a time earlier by one period
interval.
[0177] By this operation, an output signal may readily be changed
over without regard to prevailing driving modes.
[0178] With the device 44, the output amplifiers 60 and 62 are
afforded with differential sensitivities in charge detection. Thus,
during normal imaging, signal charges of the color G are supplied
to and output from the output amplifier 60 of low detection
sensitivity, while signal charges of the colors R/B are supplied to
and output from the output amplifier 62 of high detection
sensitivity. The device 44 thus prevents saturation of signal
charges of the color G having the highest sensitivity, while
amplifying output signals of the colors R/G with respect to the
output signal for the color G. In this manner, it is possible with
the device 44 to suppress the white balance gain to a smaller value
to improve the S/N (signal to noise) ratio of the output
signal.
[0179] In addition, the output signal of the color G is exploited
in calculating for AE/AF control. In this case, the output signal
for the color G may be switched so as to be output from the output
amplifier 62 of high detection sensitivity. In this case, the
device 44 is able to amplify the signal quantity of the color G and
to output the so amplified signal for the color G. This may improve
the accuracy in calculating AE/AF control.
[0180] The operational timing for low speed driving in the device
44 will now be described. With the low speed driving, the drive
signals .phi.HS1, .phi.HS2, .phi.HS3 and .phi.HS4 shown in FIG. 18,
parts (A) and (B) are of the same frequency as that of the drive
signals .phi.HP1 and .phi.HP2 shown in parts (C) and (D). In case
the drive signal .phi.HS2 arriving at the last electrode HS2 of the
horizontal transfer path 50 is at level "L", the drive signal
.phi.HP supplied to the electrode HP1 is at a level "H", while the
drive signal .phi.HP2 supplied to the electrode HP2 is at a level
"L". Hence, the signal charges are transferred at all times via
branching section (HSL) 54 towards the electrode HP1, that is,
towards the horizontal transfer path 56.
[0181] After application of the reset signal .phi.RS shown in FIG.
18, part (E), the output signal OS1 shown in part (F) is output.
The output signal OS2 shown in part (G) is at "L" level, as a
result of which the output signal OS1 is a single line output.
Since the output signal OS2 is not used, the power supply for the
output amplifier 62 may be turned off. In this case, supply of
power for the amplifier power supply 16 is controlled depending on
the control signal 86 from the power supply control 122. The power
supply 66 is turned off.
[0182] The phases of the drive signal .phi.HP1 and .phi.HP2 may be
reversed from those shown, or the phase of the set of the drive
signals .phi.HS1 and .phi.HS3 and that of the set of the drive
signals .phi.HS2 and .phi.HS4 may be reversed from those shown, so
that, with the device 44, only the output amplifier 62 will be in
operation. By so doing, only the horizontal transfer path 58 will
be in operation (one-line outputting). It is possible in this
fashion to switch between one-line outputting and two-line
outputting to enable free selection of the outputs. With low
sensitivity, if the dynamic range is prioritized, it may be
preferred to select the output amplifier with low sensitivity. On
the other hand, if the high sensitivity is prioritized, it may be
preferred to select the output amplifier with high sensitivity.
[0183] Referring to FIG. 19, the device 44 includes trifurcated
horizontal transfer paths 192, 194 and 196 for three line readout
after a branching section (HSL) 54 provided from a horizontal
transfer path 50. Output amplifiers 198, 200 and 202 are provided
at output sides of the horizontal transfer paths 192, 194 and 196,
respectively. The output amplifiers 198, 200 and 202 may exhibit
differential sensitivities in charge detection. On the horizontal
transfer path 192, electrodes HP3, HP2, HP1 and HP2 are provided
adjacent to the branching section 54. On the horizontal transfer
path 194, electrodes HP2, HP4, HP2, HP4 and HP2 are provided
adjacent to the branching section 54. The number of the electrodes
on the horizontal transfer path 194 is larger by one than that on
the horizontal transfer path 192. On the horizontal transfer path
196, electrodes HP1, HP2, HP3, HP2, HP1 and HP2 are provided
adjacent to the branching section 54. The number of the electrodes
on the horizontal transfer path 196 is larger by one than that on
the horizontal transfer path 194.
[0184] The driving timing in the horizontal transfer for three line
output will now be described. On the horizontal transfer path 50,
drive signals .phi.HS1 to .phi.HS4 are supplied in the same way as
above, as shown in FIG. 20, parts (A) and (B). On the horizontal
transfer paths 192, 194 and 196, there is supplied a drive signals
.phi.HP1 to .phi.HP4. The drive signal .phi.HP2 of part (D) is the
same as the drive signal .phi.HP2 of FIG. 5, part (D). The drive
signal .phi.HP4 of FIG. 20, part (F), of the present embodiment is
the same as the drive signal .phi.HP1 of FIG. 5, part (C). The
frequency of the drive signals .phi.HP1 and .phi.HP3 of FIG. 20,
parts (C) and (E) is one-half that of the drive signals .phi.HP2
and pHP4. The rising edge of the drive signal .phi.HP1 of FIG. 20,
part (C), is synchronized with the rising edge of the drive signal
.phi.HP4. As for the timing relationship of FIG. 20(E), the rising
edge of the drive signal .phi.HP3 is synchronized with the falling
edge of the drive signal .phi.HP2.
[0185] On the horizontal transfer paths 192, 194 and 196, output
signals are generated in the output time domains 204, 206, 208
responsive to the reset signal .phi.RS of FIG. 20, part (G),
supplied. The output signals OS1 and OS3 shown in part (H) are
output at every second output time domains, that is, at output time
domains 204, 208. The output signal OS1 is an output signal of the
color R, while the output signal OS3 is an output signal for the
color B. The output signal OS2 shown in part (I) is output in every
output time domain, that is, at the output time domains 204, 206,
208. The output signal OS2 is an output signal for the color G.
[0186] By outputting from color to color in this fashion, it is
possible to improve the degree of freedom in color designing.
[0187] A device 44 shown in FIG. 21 includes horizontal transfer
paths 210, 212, 214 and 216 for four line readout connected to the
horizontal transfer path 50 via branching section (HSL) 54. The
horizontal transfer paths 210, 212, 214 and 216 are provided at
output ends with output amplifiers 218, 220, 222 and 224,
respectively. The output amplifiers 218 and 222 are afforded with
differential sensitivities in charge detection. However, the output
amplifiers 222 and 224 reading out the same colors G.sub.r and
G.sub.b are desirably adjusted to the same sensitivity for charge
detection. The colors G.sub.r and G.sub.b mean that the color
attributes of the color filter segments of the photosensitive cells
46 neighboring to the color G are the colors R and B.
[0188] In the horizontal transfer path 210, electrodes HP4, HP1,
HP2, HP3 and HP4 are provided next to the branching section 54. In
the horizontal transfer path 212, electrodes HP3, HP4, HP1, HP2,
HP3 and HP4 are provided next to the branching section 54. The
number of the electrodes of the horizontal transfer path 212 is one
more than that of the horizontal transfer path 210. In the
horizontal transfer path 214, electrodes HP2, HP3, HP4, HP1, HP2,
HP3 and HP4 are provided next to the branching section 54. The
number of the electrodes of the horizontal transfer path 214 is one
more than that of the horizontal transfer path 212. In the
horizontal transfer path 216, electrodes HP1, HP2, HP3, HP4, HP1,
HP2, HP3 and HP4 are provided next to the branching section 54. The
number of the electrodes of the horizontal transfer path 216 is one
more than that of the horizontal transfer path 214.
[0189] The driving timing in the horizontal transfer of the
four-line output will now be described. To the horizontal transfer
path 50 are supplied drive signals .phi.HS1 to .phi.HS4 which are
the same as those described above, as shown in FIG. 22, parts (A)
and (B). To the horizontal transfer paths 210, 212, 214 and 216
following the branching, drive signals .phi.HP1 to pHP4 shown in
parts (C) to (F) are supplied. The frequency of the drive signals
.phi.HP1 to .phi.HP4 is one-fourth that of the drive signals
.phi.HS1 to .phi.HS4. The drive signals .phi.HP2 to .phi.HP4 are
phase-shifted 90.degree., 180.degree. and 270.degree. with respect
to the rising edge of the drive signal .phi.HP1. On the horizontal
transfer paths 210 to 216, output signals OS1 to 0S4 are obtained
in the output time region 226 at the same time in keeping with the
reset signal .phi.RS of part (G).
[0190] It is noted that, even in the device 44 for horizontal
transfer for three and four line readouts, it is possible to select
two lines and to change the driving pattern responsive to signal
charges of the color attributes to change over the output
destinations. This suppresses the WB gain to improve the S/N
ratio.
[0191] By this multi-line readout, it is possible to read out
signal charges with a further lower driving frequency to improve
the degree of freedom in color designing.
[0192] An alternative embodiment of the present invention will now
be described with reference to the drawings. Referring now to FIG.
23 showing a solid state imaging apparatus 10a of the present
modification, the timing signal generator 32 controls the transfer
timing on the horizontal transfer path of the imaging unit 14,
according to the measurement by a transfer efficiency measurement
unit 500 of the signal processor 26. In the description, like
components are designated with the same reference numerals and
description thereon will not be repeated for simplicity.
[0193] In the present solid state imaging apparatus 10a, the speed
of horizontal transfer of signal charges in the imaging unit 14 may
be varied depending on an image shooting mode, such as a still
image mode, a moving picture mode or a repeated shooting mode, or
on the result of scene decision on an image being imaged. More
specifically, with the instant alternative embodiment, signal
charge transfer may be made at a high or a low speed. In the
low-speed transfer, the imaging unit 14 outputs a sole shot image
via a sole output circuit by way of one-line outputting. In the
high-speed transfer, the imaging unit 14 outputs a sole shot image
via two output circuits by way of two-line outputting.
[0194] The device 44 transmits the drive signal .phi.HS2, supplied
from the drivers 20, to respective electrodes of the transfer
device HS2 on the horizontal transfer path 50 to transmit the drive
signal .phi.HSL supplied from the biasing circuit 18 to each
electrode HSL of the branching section 54. The drive signal
.phi.HSL is a constant bias voltage. The device 44 routes the drive
signals .phi.HP1 and .phi.HP2 to transfer elements HP1 an dHP2 on
the horizontal transfer elements 56 and 58, respectively. The
device 44 also transmits drive signals .phi.OG, .phi.RS and .phi.RD
supplied from the drivers 20, to the electrodes OG and RS and to
the reset drain RD, respectively. The electrode OG is supplied with
a preset voltage by this drive signal .phi.OG, while the reset
drain RD is supplied with a preset power supply voltage by the
drive signal .phi.RD.
[0195] The signal processor 26 has the function of generating a
control signal responsive to a digital signal 118. In the present
alternative embodiment, the signal processor 26 includes, in
addition to the transfer efficiency measurement unit 500, the power
supply control 122, gain control 124, power supply control 122, AF
control 126, AE control 128, AWB control 130 and data converter 132
like the embodiment shown in and described with reference to FIG.
2.
[0196] The transfer efficiency measurement unit 500 measures the
transfer efficiency for signal charge proceeding from the branching
section 54 through the horizontal transfer paths 56 and 58, that
is, the transfer efficiency on the horizontal transfer paths 56 and
58, based on the digital image signal 118.
[0197] The transfer efficiency measurement unit 500 measures the
horizontal transfer efficiency of the horizontal transfer paths 56
and 58 in advance, for example, at the time of shipment from the
plant. Preferably, it is verified whether the horizontal transfer
efficiency of one of the horizontal transfer paths is satisfactory,
responsive to the results of the measurement, and the result of
decision is stored in a memory circuit, not shown. The transfer
efficiency measurement unit 500 may store the results of
measurement, that is, the horizontal transfer efficiency itself in
the memory, for having the system controller 28 check and verify
the horizontal transfer efficiency.
[0198] The system controller 28 may also have the function of
verifying, in case the transfer efficiency measurement unit 500 has
stored the measured results in a memory, which of the horizontal
transfer paths has the optimum transfer efficiency, based on the
result. The system controller 28 sends a control signal
representing the results of decision by the system controller 28 or
by the transfer efficiency measurement unit 500 to the timing
signal generator 32.
[0199] In the present alternative embodiment, the timing signal
generator 32 may receive the result of decision indicating which of
the horizontal transfer paths 56 and 58 has the optimum horizontal
transfer efficiency, from the transfer efficiency measurement unit
500 or the system controller 28, and vary the driving conditions
for the horizontal timing signal for the horizontal transfer path
50, based on the results of decision. For example, the timing
signal generator 32 may shift the driving start time in the initial
driving condition for the horizontal timing signal to make a
relative shift from the driving start time in the initial driving
condition of the horizontal timing signal for the horizontal
transfer paths 56 and 58.
[0200] When a horizontal timing signal of the initial driving
condition for high speed driving is supplied by the timing signal
generator 32 to the drivers 20 to control the horizontal transfer
path 50 and the branching section 54, green signal charges are
transferred to the horizontal transfer path 56, while alternately
red and blue signal charges are transferred to the horizontal
transfer path 58. If the horizontal timing signal of the initial
driving condition is supplied to the drivers 20, as the horizontal
timing signal is relatively offset so as to be delayed or advanced
by, for example, one period, such as to control the horizontal
transfer path 50 and the branching section 54, the signal charges
of the red and blue colors may be alternately transferred to the
horizontal transfer path 56, while the green signal charges may be
transferred to the horizontal transfer path 58.
[0201] With the timing signal generator 32 of the present
alternative embodiment, it is possible to determine the driving
condition for reversed branching, as the horizontal timing signal
of the initial driving condition is offset as described above, in
order to reverse signal charges transferred on the horizontal
transfer paths 56 and 58 by invert electrical signals output to the
output amplifiers 60 and 62.
[0202] The timing signal generator 32 may determine desired driving
conditions for high speed driving, responsive to the results of
decision of the horizontal transfer efficiency, to transfer desired
signal charges, such as red and green signal charges, to one of the
horizontal transfer paths 56 and 58 where the horizontal transfer
efficiency is optimum.
[0203] Moreover, during low speed driving, the timing signal
generator 32 transfers all signal charges only to one of the
horizontal transfer paths 56 and 58. It is therefore possible to
determine desired driving conditions for low speed driving,
responsive to the result of decision of the horizontal transfer
efficiency, and to transfer all signal charges to the horizontal
transfer efficiency 56 or 58 having the optimum horizontal transfer
efficiency. The horizontal transfer in the initial driving
conditions is carried out as described above with reference to
FIGS. 4 to 9E.
[0204] The operation of horizontal transfer under an initial
driving condition during high speed transfer in the solid state
imaging apparatus 10a will now be described with reference to the
timing chart of FIG. 24.
[0205] FIG. 24, parts (A) to (G) illustrate the operation in which
the timing signal generator 32 transmits the horizontal timing
signal of the initial driving condition to the drivers 20. In the
figure, parts (A) and (B) show drive signals .phi.HS1 and .phi.HS3,
and drive signals .phi.HS2 and .phi.HS4, which the HS driver 96
outputs to the horizontal transfer path 50, respectively. Parts (C)
and (D) show drive signals .phi.HP1 and .phi.HP2, the HS driver 98
outputs to the horizontal transfer paths 56 and 58, respectively.
Part (E) show the reset signal .phi.RS, the RS driver 100 outputs
to the output amplifiers 60 and 62, while parts (F) and (G) show
output signals OS1, OS2, output from the output amplifiers 60 and
62, respectively.
[0206] It is now assumed that, on the horizontal transfer path 50
of the present alternative embodiment, the signal charges
transferred from each vertical transfer path 48 are re-arrayed in
the output sequence, during the horizontal blanking period. It is
also assumed that, at a time t202, the dummy pixels D1, D2,
optically black pixels OB1, OB2, R pixel, G pixel, B pixel, G
pixel, . . . are stored in this order, in the respective transfer
elements transferred, beginning from the end neighboring to the
branching section 54 to the opposite end.
[0207] In the alternative embodiment, the horizontal drive signals
74 actuated at time t204 is generated by the HS driver 96 in the
drivers 20 as shown in FIG. 3, responsive to the horizontal timing
signal of the initial driving condition. As this drive signal 74,
the drive signals .phi.HS1 and .phi.HS3 and the drive signals
.phi.HS2 and .phi.HS4 are transmitted to the horizontal transfer
path 50. There is also generated the horizontal parallel drive
signal 76 actuated at time t206 in the HP driver 98 in the drivers
20. As the drive signal 74, the drive signals .phi.HP1 and .phi.HP2
are supplied to the horizontal transfer paths 56 and 58,
respectively.
[0208] First, on the horizontal transfer path 50, the signal
charges are transferred horizontally towards the branching section
54. In the alternative embodiment, the signal charges are
transferred in the sequence of the dummy pixels D1, D2, optically
black pixels OB1, OB2, R pixel, G pixel, B pixel, G pixel, . . .
.
[0209] The signal charges are alternately transferred at this time
from the branching section 54 to the horizontal transfer paths 56
and 58, in a branched fashion. The branched transfer will be
described subsequently. The signal charges are initially sent to
the horizontal transfer path 58 supplied with a horizontal parallel
drive signal .phi.HP2 which is at level "H" at time t204. The
signal charges are then sent to the horizontal transfer path 56
supplied with the horizontal parallel drive signal .phi.HP1 which
is at level "H" at time t206. Thus, in the alternative embodiment,
the dummy pixel D1, the optically black pixel OB1, R pixel and the
B pixel are transferred in this order towards the horizontal
transfer path 58, while the dummy pixel D2, the optically black
pixel OB2, G pixel and the G pixel are transferred in this order
towards the horizontal transfer path 56.
[0210] The signal charges sent to the horizontal transfer paths 56
and 58 are transferred to the output amplifiers 60 and 62,
respectively, where they are converted into analog electrical
signals which are then output. The output amplifier 60 of the
instant alternative embodiment sequentially outputs the dummy pixel
D2, the optically black pixel OB2, G pixel and the G pixel, as
output signal OS1, that is, as analog electrical signal 82, during
the output periods t182, t184, t186 and t188, respectively, as
shown in FIG. 24, part (F).
[0211] On the other hand, the output amplifier 62 sequentially
outputs the dummy pixel D2, dummy pixel D1, optically black pixel
OB1, R pixel and the B pixel, as output signal OS2, that is, as
analog electrical signal 84, during the output periods t182, t184,
t186 and t188.
[0212] Thus, if the imaging unit 14 executes horizontal transfer
under the initial condition, the output signal OS1 indicating the
color signal of the G pixel is output from the horizontal transfer
path 56 and the output amplifier 60, while the output signal OS2
indicating the color signal of the R and B pixels is output from
the horizontal transfer path 58 and the output amplifier 62.
[0213] The operation of shifting the horizontal timing signal from
the initial driving condition for high-speed transfer in the solid
state imaging apparatus 10a of the present alternative embodiment
will now be described with reference to the timing chart shown in
FIG. 25.
[0214] In FIG. 25, parts (A) to (G), there is shown the operation
in which the timing signal generator 32 sends to the drivers 20 a
horizontal timing signal of the driving condition of inverted
branching, having a delay of one period, e.g. a period t190 from
the initial driving condition. In the figure, parts (A) and (B)
show the drive signals .phi.HS1 and .phi.HS3, and the drive signal
.phi.HS2 and .phi.HS4 which the HS driver 96 outputs to the
horizontal transfer path 50, respectively. Parts (F) and (G) show
output signals OS1 and OS2 output by the output amplifiers 60 and
62, respectively. It is noted that the drive signals .phi.HP1 and
.phi.HP2 and the reset signal .phi.RS may be the same signals as
those shown in FIG. 24, parts (C), (D) and (E).
[0215] In the present alternative embodiment, it is again assumed
that the dummy pixels D1 and D2, optically black pixels OB1 and
OB2, R pixel, G pixel, B pixel and the G pixel, . . . , are stored
at time t202 in the respective transfer elements in the horizontal
transfer path 50 in the same manner as above.
[0216] In the instant alternative embodiment, the horizontal
parallel drive signal 76 actuated at time t206 is generated by the
HP driver 98 in the same manner as above. As this drive signal 76,
the drive signals .phi.HP1 and .phi.HP2 are supplied to the
horizontal transfer paths 56 and 58, respectively. Moreover, the HS
driver 96 generates the horizontal parallel drive signal 74
actuated at time t208 responsive to the horizontal timing signal of
the driving condition for inverted branching. As this drive signal
74, the drive signals .phi.HS1 and .phi.HS3 and the drive signals
.phi.HS2 and .phi.HS4 are supplied to the horizontal transfer path
50.
[0217] Next, on the horizontal transfer path 50, the signal charges
are horizontally transferred towards the branching section 54. In
the present alternative embodiment, the signal charges in the order
of the dummy pixels D1, D2, optically black pixels OB1, OB2, R
pixel, G pixel, B pixel, C pixel, . . . .
[0218] At this time, the signal charges are alternately transferred
from the branching section 54 to the horizontal transfer paths 56
and 58 in the sequence reversed from that of the horizontal
transfer of the initial driving condition. The signal charges are
initially sent to the horizontal transfer path 56 supplied with the
horizontal parallel drive signal .phi.HP1 which is at level "H" at
time t208, and are then sent to the horizontal transfer path 58
supplied with the horizontal parallel drive signal .phi.HP2 which
is at level "H" at time t210. Thus, in the present alternative
embodiment, the dummy pixel D1, optically black pixel OB1, R pixel
and the B pixel are sent in this sequence to the horizontal
transfer path 56, while the dummy pixel D2, optically black pixel
OB2, G pixel and the G pixel are sent in this sequence to the
horizontal transfer path 58.
[0219] The signal charges on the horizontal transfer paths 56 and
58 are then sent to the output amplifiers 60 and 62, respectively.
The output amplifier 60 outputs, as the output signal OS1, the
dummy pixel D1, optically black pixel OB1, R pixel and the B pixel
during the output periods t182, t184, t186 and t188, respectively,
as shown in FIG. 25, part (F). On the other hand, the output
amplifier 62 outputs, as the output signal OS2, the dummy pixel D2,
optically black pixel OB2, G pixel and the G pixel, during the
output periods t182, t184, t186 and t188, respectively, as shown in
part (G).
[0220] Thus, in case the imaging unit 14 executes horizontal
transfer under the driving condition of inverted branching, the
horizontal transfer path 56 and the output amplifier 60 output the
output signal OS1 indicating the color signals of the R and B
pixels, while the horizontal transfer path 58 and the output
amplifier 62 output the output signal OS2 indicating the color
signals of the G pixels.
[0221] The operation of low-speed horizontal transfer by the solid
state imaging apparatus 10a of the present alternative embodiment
will now be described with reference to the timing chart of FIG.
26.
[0222] FIG. 26, parts (A) to (G), show the operation in case the
timing signal generator 32 sends a horizontal timing signal of the
initial driving condition to the drivers 20. Specifically, in the
figure, parts (A) and (B) show horizontal serial drive signals
.phi.HS1, .phi.HS3 and .phi.HS2, .phi.HS4 which the HS driver 96
outputs to the horizontal transfer path 50. Parts (C) and (D) show
the horizontal parallel drive signals .phi.HP1 and .phi.HP2, output
by the HP driver 98 to the horizontal transfer paths 56, 58,
respectively. Part (E) shows the reset signal .phi.RS the RS
deriver 100 outputs to the output amplifiers 60 and 62, and parts
(F) and (G) show output signals OS1 and OS2 output from the output
amplifiers 60 and 62, respectively.
[0223] In the present alternative embodiment, the horizontal serial
drive signals .phi.HS1, .phi.HS3 and .phi.HS2, .phi.HS4 are output
at the same frequency as that of the horizontal parallel drive
signals .phi.HP1 and .phi.HP2. Referring to FIG. 26, when the drive
signal .phi.HS2 supplied to the last electrode HS2 of the
horizontal transfer path 50 becomes "L" in level at time t212,
while the drive signal .phi.HP1 supplied to the electrode HP1
becomes "L" in level. The signal charges at the trailing end
electrode HS2 are transferred at all times via branching section
(HSL) 54 to the electrode HP1, that is, towards the horizontal
transfer path 56.
[0224] With the device 44, the output amplifier 60 supplies an
output signal OS1 indicating the color signal as shown in FIG. 26,
part (F), while the output amplifier 60 supplies an output signal
OS2 indicating the level "L" as shown in part (G).
[0225] Thus, during low speed driving, the device 44 supplies only
the output signal OS1 on a single line without using the output
signal OS2. Hence, the power supply of the output amplifier 62 may
be turned off. It is preferred in this case to control the supply
of the supply power of the amplifier power supply 16, responsive to
the control signal 86 from the power supply control 122, to turn
off the power supply 66.
[0226] During low-speed horizontal transfer, the system controller
28 and the timing signal generator 32 may control the drivers 20 to
reverse the phase of the drive signals .phi.HP1 and .phi.HP2 or to
reverse the phase of the drive signals .phi.HS1, .phi.HS3 and the
phase of the drive signals .phi.HS2, .phi.HS4. By so doing, the
device 44 may actuate only the horizontal transfer path 58 and the
output amplifier 62 to output the output signal OS2 by sole line
output.
[0227] By so doing, it is possible to switch between the one-line
outputting and the two-line outputting extremely readily to make
free output selection. If dynamic range preference is selected in
case of low sensitivity, it is preferred to select the output
amplifier with low sensitivity in charge detection, whereas, if
sensitivity preference is selected in case of high sensitivity, it
is preferred to select the output amplifier with high sensitivity
in charge detection.
[0228] A further alternative embodiment of the operation of
measuring the transfer efficiency in the solid state imaging
apparatus 10a will now be described.
[0229] In the present apparatus 10a, signal charges are mixed on
the horizontal transfer path 50 by horizontal pixel mixing, for
measuring the transfer efficiency on the branching section 54 and
on the horizontal transfer paths 56 and 58. The drivers 20 control
the driving of the imaging unit 14 in order to provide for
two-channel outputting on the branching section 54 and on the
horizontal transfer paths 56 and 58.
[0230] By this horizontal pixel mixing, the horizontal transfer
path 50 accumulates a pixel of a reference signal, obtained on
mixing plural signal charges, and void pixels, deprived of signal
charges by this mixing. The horizontal transfer path 50 repeatedly
generates pixel groups 250, each composed of plural pixels, more
specifically, each composed of the pixel of the reference signal
Sig followed by two or more void pixels Emp1 and Emp2. The
horizontal transfer path 50 horizontally transfers the pixel groups
250, each made up of the pixel Sig, Emp1 and Emp2, towards the
branching section 54.
[0231] In case the branching section 54 and the horizontal transfer
paths 56 and 58 are driving-controlled, responsive, e.g. to the
horizontal timing signal of the initial driving condition, the
branching section 54 sends the pixel Sig and the second void pixel
Emp2 to the horizontal transfer path 56, while sending the first
void pixel Emp1 to the horizontal transfer path 58.
[0232] In this case, the signal charge left over at the time of
transfer of the pixel Sig of the reference signal in the course of
horizontal transfer on the horizontal transfer path 56 is intruded
into the second void pixel Emp2, while the signal charge left over
at the time of transfer of the pixel Sig of the reference signal in
the course of horizontal transfer from the branching section 54
onto the horizontal transfer path 58 is intruded into the first
void pixel Emp1.
[0233] The amount of signal charges combined from the first void
pixel Emp1 and the second void pixel Emp2, that is, the amount of
the residual charges 232, left over on the horizontal transfer path
56 until the pixel Sig gets to the output amplifier 60, as it
travels from the branching section 54 through the horizontal
transfer path 56, is varied with the pixel Sig of the reference
signal, that is, with the quantity of the reference signal
charges.
[0234] The output amplifier 60 then outputs the output signal 82
including the pixel Sig of the reference signal and the second void
pixel Emp2. The pre-processor 22 generates a digital signal 110 by
processing this output 82 to store the digital signal in the memory
24. In similar manner, the horizontal transfer path 58 and the
output amplifier 62 output the output signal 84 including the first
void pixel Emp1, and the pre-processor 22 generates a digital
signal 112 by processing this output 84 to store the digital signal
112 in the memory 24.
[0235] The signal processor 26 reads out the pixel of the reference
signal Sig and the first and second void pixels Emp1 and Emp2 from
the memory 24, as digital signal 118, over bus 114 and signal line
120.
[0236] In the signal processor 26, the transfer efficiency
measurement unit 500 may acquire a residual transfer charge
quantity 232 on the horizontal transfer path 56, based on the first
void pixel Emp1 and the second vacant pixel Emp2. The transfer
efficiency measurement unit may then calculate the horizontal
transfer efficiency HTR11 pertinent to transfer from the branching
section 54 to the horizontal transfer path 56, of the horizontal
transfer efficiency (HTR) on the horizontal transfer path 56, based
on the charge of the reference signal Sig and the residual transfer
charge Emp1. The transfer efficiency measurement unit also may
calculate the horizontal transfer efficiency HTR12 pertinent to the
transfer on the horizontal transfer path 56 up to the electrode OG
and the section FD, based on the quantity of the charge of the
reference signal Sig and that of the residual transfer charge Emp2.
The transfer efficiency measurement unit 500 may calculate the
horizontal transfer efficiency HTR11, using an expression:
HTR11=(Sig-Emp1)/Sig.times.100, while it may calculate the
horizontal transfer efficiency HTR12, using an expression:
HTR12=(Sig-Emp2)/Sig.times.100. The horizontal transfer
efficiencies HTR11 and HTR12 are varied responsive to the charge
quantity of the reference signal Sig as shown in FIG. 30.
[0237] In case the branching section 54 and the horizontal transfer
paths 56 and 58 are driving-controlled under the driving condition
of reversed branching, with the horizontal timing signal being
offset from that of the initial driving condition, the branching
section 54 sends the first void pixel Emp1 to the horizontal
transfer path 58, while sending the pixel of the reference signal
Sig and the second void pixel Emp2 to the horizontal transfer path
56.
[0238] The signal charge left over at the time of transfer of the
pixel Sig of the reference signal, in the course of horizontal
transfer on the horizontal transfer path 58 from the branching
section 54 is intruded into the first void pixel Emp1, while the
signal charge left over at the time of transfer of the pixel Sig of
the reference signal in the course of horizontal transfer from the
branching section 54 onto the horizontal transfer path 58 is
intruded into the second void pixel Emp2. The amount of signal
charges combined from the void pixels Emp1 and Emp2, that is, the
amount of the residual charge 232 left over on the horizontal
transfer path 58 until the pixel Sig gets to the output amplifier
62, as it travels from the branching section 54 through the
horizontal transfer path 58 is varied with the quantity of the
reference signal charges as shown in FIG. 29.
[0239] The output amplifiers 60, 62 then output the output signal
82 including the first void pixel Emp1, and the output signal 84
including the pixel of the reference signal Sig and the second void
pixel Emp2. The pre-processor 22 stores digital signals 110, 112,
derived from the output signals 82 and 84, respectively, in the
memory 24.
[0240] The signal processor 26 reads out the first void pixel Emp1,
pixel of the reference signal Sig and the second void pixel Emp2
from the memory 24 as digital signal 118, over a bus 114 and a
signal line 120.
[0241] In the signal processor 26, the transfer efficiency
measurement unit 500 may acquire a residual transfer charge
quantity 234 on the horizontal transfer path 58, based on the first
vacant pixel Emp1 and the second vacant pixel Emp2. The transfer
efficiency measurement unit may then calculate the horizontal
transfer efficiency HTR21 pertinent to transfer from the branching
section 54 to the horizontal transfer path 58, out of the
horizontal transfer efficiency (HTR) on the horizontal transfer
path 58, based on the charge of the reference signal Sig and the
residual transfer charge Emp1. The transfer efficiency measurement
unit also may calculate the horizontal transfer efficiency HTR12
pertinent to transfer on the horizontal transfer path 58 up to the
electrode OG and the section FD, based on the charge of the
reference signal Sig and the residual transfer charge Emp2. The
transfer efficiency measurement unit 500 may calculate the
horizontal transfer efficiency HTR21, using an expression:
HTR21=(Sig-Emp1)/Sig.times.100, while it may calculate the
horizontal transfer efficiency HTR22, using an expression:
HTR22=(Sig-Emp2)/Sig.times.100. The horizontal transfer efficiency
HTR21 and the horizontal transfer efficiency HTR22 are varied
responsive to the charge of the reference signal Sig as plotted in
FIG. 30.
[0242] The transfer efficiency measurement unit 500 may measure one
or both of the horizontal transfer efficiencies HTR11 and HTR12,
while it may also calculate one or both of the horizontal transfer
efficiencies HTR21 and HTR22, and use the measured results for
determining the horizontal transfer efficiencies of the horizontal
transfer paths 56 and 58. In the present alternative embodiment, it
is particularly preferred to decide on the horizontal transfer
efficiency based on the amount of residual charges left over from R
and B pixels to the G pixel at the branching section 54, and to
measure only the residual transfer charge attributable to the
branching section 54, that is, the horizontal transfer efficiencies
HTR11 and HTR22, for use in determining the horizontal transfer
efficiency.
[0243] The transfer efficiency measurement unit 500 may compare the
horizontal transfer efficiencies HTR11 and HTR21 to verify which of
the horizontal transfer paths 56 and 58 is better in the transfer
efficiency. It is noted that the transfer efficiency measurement
unit 500 may compare the horizontal transfer efficiencies HTR11 and
HTR21 based on the quantity of the charges at a sole point, or may
calculate evaluation values of the horizontal transfer efficiencies
HTR11 and HTR21 based on the quantities of the charges at plural
points.
[0244] The present apparatus 10a captures a preset subject in
advance to measure the transfer efficiency. In the present
alternative embodiment, the apparatus may capture plural different
subjects to obtain reference signal charge quantities and residual
charges at plural points. The transfer efficiency measurement unit
500 may then calculate the horizontal transfer efficiencies HTR11
and HTR21 at plural points for comparison and verification.
[0245] In the present alternative embodiment, a plural number of
groups 250, each made up of a pixel of a reference signal Sig
followed by two or more void pixels Emp1 and Emp2, may be generated
in succession by horizontal pixel mixing on the horizontal transfer
path 50. To this end, the horizontal transfer path 50 may be driven
in a horizontal eight pixel mixing system to formulate the pixel
group 250 each made up of a pixel of a reference signal Sig
followed by three void pixels Emp1, Emp2 and Emp3.
[0246] The operation of driving the horizontal transfer path 50 in
accordance with the horizontal eight pixel mixing system, in the
solid state imaging apparatus 10a of the present alternative
embodiment, will now be described with reference to the timing
chart of FIG. 30 and potential transition diagrams of FIGS. 32A
through 32I.
[0247] In the potential transition diagrams of FIG. 32A through
32I, there are shown the potential levels and signal charges
retained on the horizontal transfer path 50. Transfer elements HS4,
HS1, HS2, HS3, HS2, HS1, HS4, HS3 and HS4 are formed in the
transfer path 50 in this sequence from its end neighboring to the
branching section 54 towards its opposite end. The signal charge of
each transfer element is transferred towards left, that is, towards
the forward side transfer element.
[0248] On the horizontal transfer path 50 of the present
alternative embodiment, horizontal eight pixel mixing is carried
out during the horizontal blanking period such that signal charges
of the group 250 composed of the transfer elements H51, HS2, HS3,
HS2, HS1, HS4, HS3 and HS4 are mixed together. Referring to FIG.
31, part (A), signal charges are transferred to the horizontal
transfer path 50, responsive to the drive signal .phi.LM supplied
to the line memory LM. The signal charges are mixed together on the
horizontal transfer path 50, under driving control by the
horizontal serial drive signals 74, such as drive signals .phi.HS1,
.phi.HS2 and .phi.HS3, .phi.HS4, supplied from the HS driver 96, as
shown in FIG. 31, parts (B) to (E).
[0249] When the drive signal .phi.LM shown in FIG. 31, part (A),
becomes "H" in level at time t302, signal charges are transferred
from the line memory LM to the packets generated directly
underneath the electrodes of the transfer elements HS1 to HS4 on
the horizontal transfer path 50, because the drive signals .phi.LM
and the signals .phi.HS1 to .phi.HS4 shown in FIG. 31, parts (B) to
(E) are all at level "H". Referring to FIG. 32A, the potential
level of each of the transfer elements HS1 to HS4 in the state of
time t302 is reference level 300.
[0250] If the level "L" drive signals .phi.HS2, .phi.HS4 are then
supplied at time t304 to the transfer elements HS2 and HS4, the
potential level becomes higher at the transfer elements HS2 and HS4
as shown in FIG. 32B, so that the signal charges are transferred to
the forward side transfer elements HS1 and HS3.
[0251] At time t306, the drive signal .phi.HS4 at level "H" is
supplied to the transfer element HS4, so that the potential level
reverts to the reference level 300.
[0252] If the level "L" drive signal .phi.HS3 is then supplied at
time t308 to the transfer element HS3 as shown in FIG. 32D, the
potential level becomes higher at the transfer element HS3. Thus,
if the forward side transfer element is HS2, which is of the same
potential level, the signal charge is maintained at the transfer
element HS3. However, if the forward side transfer element is HS4,
which is low in potential level, the signal charge at the transfer
element HS3 is transferred.
[0253] At time t310, the drive signal .phi.HS4 at level "L" is
supplied to the transfer element HS4 to raise the potential level
as shown in FIG. 32E, so that the retained signal charges are
transferred to the forward side transfer element HS1.
[0254] At time t312, the level "L" drive signal .phi.HS1 is
supplied to the transfer element HS1 to raise the potential level
as shown in FIG. 32F. The level "H" drive signal .phi.HS2 is
supplied to the transfer element HS2 so that the potential level
reverts to the reference level 300. The signal charge is
transferred from the high potential level transfer element HS1 to
the forward side low potential level transfer element HS2. A signal
charge is also transferred from the rearward high potential level
transfer element HS3 to the transfer element HS2 which has become
lower in potential level.
[0255] At time t314, the level "L" drive signal .phi.HS2 is
supplied to the transfer element HS2 to raise the potential level
as shown in FIG. 32G. The level "H" drive signal .phi.HS3 is
supplied to the transfer element HS3 so that the potential level
reverts to the reference level 300. The signal charge is
transferred from the high potential level transfer element HS2 to
the forward side low potential level transfer element HS3.
[0256] At time t316, the level "H" drive signal .phi.HS2 is
supplied to the transfer element HS2, so that the potential level
reverts to the reference level 300 as shown in FIG. 32H. The level
"L" drive signal .phi.HS3 is supplied to the transfer element HS3
to raise the potential level. The signal charge is transferred from
the high potential level transfer element HS3 to the forward side
low potential level transfer element HS2.
[0257] At time t318, the level "H" drive signal .phi.HS1 is
supplied to the transfer element HS1, so that the potential level
reverts to the reference level 300 as shown in FIG. 32I. The level
"L" drive signal .phi.HS2 is supplied to the transfer element HS2
to raise the potential level. The signal charge is transferred from
the high potential level transfer element HS2 to the forward side
low potential level transfer element HS1.
[0258] In this manner, the signal charges of the group 250 composed
of the transfer elements HS1, HS2, HS3, HS2, HS1, HS4, HS3 and HS4,
are transferred to the foremost transfer element HS1, as the signal
charges undergo the horizontal eight pixel mixing process.
[0259] The pre-processor 22 may be composed of a G pixel processor
352 suited for processing G pixels, and an RB pixel processor 354
suited for processing R and B pixels as shown in FIG. 35. It is
necessary in this case to supply an electrical signal of the G
pixel and an electrical signal of the R and B pixels, out of the
electrical signals 82 and 84 supplied from the imaging unit 14, to
the G pixel processor 352 and to the RB pixel processor 354,
respectively.
[0260] In the present alternative embodiment, output units 362, 364
are connected to the output amplifiers 60 and 62 of the imaging
unit 14. A connection unit 372 and another connection unit 374 are
connected to the connection line 82 outputting the aforementioned
electrical signal, and a connection unit 376 is connected to the
connection line 84 outputting the aforementioned electrical signal.
The electrical connection between the connection units and the
output amplifiers is changed responsive to measured results by the
transfer efficiency measurement unit 500.
[0261] The transfer efficiency of the imaging unit 14 is measured
by, e.g. a test on a silicon wafer (probe test). Based on the
results of measurement of the transfer efficiency, it is verified
which of the horizontal transfer paths 56 and 58 has a better
transfer efficiency. If it is the horizontal transfer path 58 that
has a better transfer efficiency, the output unit 362 may be
connected to the connection unit 372, while the output unit 364 may
be connected to the connection unit 374 as shown in FIG. 35. If
conversely the horizontal transfer path 56 has the better transfer
efficiency, the output unit 362 may be connected to the connection
unit 374, while the output unit 364 may be connected to the
connection unit 376 as shown in FIG. 36.
[0262] By so doing, when the present apparatus 10a is operated with
high-speed driving, the output signal 82 of the imaging unit 14 is
at all times for the G pixel, which is processed by the G pixel
processor 352, while the output signal 84 is at all times for the R
pixel and the B pixel, which are processed by the RB pixel
processor 354. It is therefore possible to avoid switching of the
processing or the electrical connection outside the imaging unit
14.
[0263] In the imaging unit 14, the output units 362 and 364 and the
connection units 372, 374 and 376 may be formed by bonding pads,
while connection lines 382, 384 interconnecting the connection and
output units may be formed by wires the connections of which may be
changed as desired. It is necessary for these wires 382, 384 to be
connected without physical contact or intersection in order to
avoid cross-talk in the transmitted signals.
[0264] In case the present apparatus 10a is used for capturing a
subj ect with low color temperature, for example, the signal charge
obtained at the R pixel photosensitive cell is great and that at
the B pixel photosensitive cell is small. If the signal charges are
transferred in the order of the R pixel, G1 pixel, B pixel and the
G2 pixel on the horizontal transfer path 50 towards the branching
section 54, the amount of the charge left over by the forward side
R pixel and intruded into the rear side G1 pixel is greater than
the amount of charge mixing that occurs between the forward side B
pixel and the rear side G2 pixel. This is because the amount of
charge mixing, that is, the amount of residual transfer charges, is
increased with increase in the signal quantity as shown in FIG. 29.
Hence, there is produced the difference between the signal quantity
of the G1 pixel and that of the G2 pixel, even though the pixels
are of the same color, this difference in the signal quantity
affecting the ultimate image as a fixed pattern noise.
[0265] With the solid state imaging apparatus 10a of the present
invention, in which, when the signal charges are sent from the
branching section 54 via horizontal transfer paths 56 or 58 and
output from the output amplifier 60 or 62, the R and B pixels, in
particular, are output on the horizontal transfer path with higher
transfer efficiency, it is possible to prevent charges from being
left over from the R and B pixels to prevent charge mixing into the
G pixel.
[0266] A further modification of the present invention will now be
described. In this modification, deterioration in the transfer
efficiency in the branching electrode is to be precluded. In the
present alternative embodiment, two horizontal transfer paths 56
and 58 are connected to the branching section 54, which branching
section divides signal charges into two parts which are to be
supplied to the horizontal transfer paths 56 and 58. The present
invention is, however, not limited to this configuration. That is,
the number of the horizontal transfer paths provided in the
branching section may be optionally set, and branching may
optionally be made in keeping with the number of the horizontal
transfer paths.
[0267] In the solid state imaging device 44 including the branching
section 54, analog electrical signals are read out, for example, in
the following manner. FIG. 35 schematically shows the horizontal
transfer paths of the device in a plane view. FIG. 36, part (A),
schematically shows the horizontal transfer paths 50 and 56 shown
in FIG. 35, to an enlarged scale. FIG. 36, part (B), schematically
shows, in cross-section, taken along a chain-dotted line
XXXVI-XXXVI in part (A). FIG. 37, part (A), schematically shows the
horizontal transfer paths 50 and 58 shown in FIG. 35, to an
enlarged scale. FIG. 37, part (B), schematically shows, in
cross-section, the horizontal transfer paths 50 and 58, taken along
a chain-dotted line XXXVII-XXXVII in part (A).
[0268] In the present alternative embodiment, signal charges are
transferred on the horizontal transfer path 50 in the order of the
G, R, G, B in terms of the color attributes. In the branching
section 54, the signal charges with the color attributes of R and B
are branched to the horizontal transfer path 56, while the signal
charges with the color attributes of G are branched to the
horizontal transfer path 58. A plural number of the transfer
elements are formed on the horizontal transfer paths 50, 56 and 58.
Each transfer element is made up of two electrodes of
polycrystalline silicon (polysilicon) and two impurity layers in
the vicinity of the surface of the silicon substrate. The two
impurity layers lying underneath the two electrodes differ from
each other in constitution. Thus, when the equi-potential drive
signals are applied, stepped potential levels are generated. The
branching section 54 is similarly a transfer element including two
electrodes. In the following, the transfer element and the two
electrodes included in the element are denoted by the same
reference numeral. For example, the branching section 54 denotes
the transfer element, while the electrode 54 denotes two electrodes
of the branching section 54.
[0269] On the horizontal transfer path 50, there are formed
polysilicon electrodes HS2, HS1, HS4, HS3, HS4, HS1, HS2 and HS3 in
this order from the right side towards the electrode HSL of the
branching section 54 lying on the left side as shown in FIG. 35.
This set of the polysilicon electrodes constitutes a repetitive
unit. An electrode HL is provided adjacent to the right side of the
electrode HSL, that is, adjacent to the left side of the electrode
HS3 on the output end of the horizontal transfer path 50 in FIG.
35. The electrode HL is equivalent to the electrode HS2 on the
right side of the electrode HLS in FIGS. 4 and 6 to 8. The
electrode HL will be described in detail subsequently.
[0270] On the horizontal transfer path 56, there are sequentially
formed four polysilicon electrodes HP1, HP2, HP1, HP2 and an OG
(output gate) electrode, from the electrode HSL of the branching
section 54 towards the output amplifier 60, as shown in FIGS. 37
and 38.
[0271] On the horizontal transfer path 58, there are sequentially
arranged five polysilicon electrodes HP2, HP1, HP2, HP1, HP2 and an
OG (output gate) electrode, from the electrode HSL of the branching
section 54 towards the output amplifier 62, as shown in FIGS. 37
and 39.
[0272] Directly below the electrodes within the P-type silicon
substrate not shown, there are formed impurity layers, as shown
from imaginary cross-sectional surfaces of the left end reset drain
RD to the electrode HP1 of the horizontal transfer path 56 and
thence further to the electrode HL of the horizontal transfer path
56, as indicated from a chain-dotted line XXXVI-XXXVI of FIG.
36.
[0273] Turning to the drive signal supplied to the respective
electrodes, drive signals .phi.HS1, .phi.HS2 and .phi.HS3 and
.phi.HS4 are supplied to the electrodes HS1, HS2, HS3 and HS4,
respectively. The drive signal .phi.HSL is supplied to the
electrode HSL, in a manner not shown, and is a constant bias
voltage. The drive signal .phi.HL is supplied to the electrode HL.
The timing for these drive signals is shown in FIG. 38.
[0274] The flow of signal charges transferred horizontally by those
drive signals will now be described. FIGS. 41A through 41E and 42
show the potential levels generated on the horizontal transfer
paths 50, 56 and 58 when the drive signals are applied thereto.
FIG. 43 shows the state of signal charge transfer at this time, as
seen from above the horizontal transfer paths. The timings shown in
FIGS. 41A through 41E, 42 and 43 correspond to those shown in FIG.
38. For example, the timing of FIG. 41A, FIG. 42, part (A) and 43A
corresponds to time t=1 in FIG. 38. The same applies for the other
timings as well.
[0275] FIG. 39 shows the potential for the horizontal transfer
paths 50 and 56. A simplified diagram of FIG. 36, part (B), is also
shown in FIG. 39 for showing the potential level positions.
Similarly, FIG. 40 shows the potential for the horizontal transfer
paths 50 and 58. A simplified diagram of FIG. 37, part (B), is also
shown in FIG. 40 for showing the potential level positions. FIGS.
41A through 41E, 42 and 43 are substantially the same as FIGS. 4
and 6 through 9E and hence the description therefor is dispensed
with.
[0276] Heretofore, the solid state imaging unit 44 suffers from the
problem that, if the transfer efficiency is deteriorated in the
branching section 54, the signal charges left over affect signal
charges of the remaining pixels, with the signal charges thus left
over becoming a fixed pattern noise in the image formed.
[0277] Specifically, assuming a case in which signal charges
transferred in the sequence of color attributes G, R, G, B, are
branched at the branching section 54, and the signal charges R and
B are transferred on the horizontal transfer path 50, while the
signal charge G is transferred on the horizontal transfer path 58,
as shown in FIGS. 41A through 41E, 42 and 43. When the
deterioration in the transfer efficiency, that is, transfer
deterioration, has occurred on the branching section 54, part of
the signal charge R left over is mixed into the next signal charge,
that is, the signal charge G. In particular, if the signal charges
are those obtained on shooting a subject having a low color
temperature, the signal charge R mixed in the signal charge G is
increased, while the signal charge B mixed in the signal charge G
is decreased. Hence, a difference in the amounts of signal charges
G transferred on the horizontal transfer path 58 is produced and
represented as a fixed pattern noise on the image.
[0278] Thus, in the present example, an electrode HL is provided
directly in front of the branching section, and adapted for being
driven independently. The duty cycle and/or the cycle of the drive
signal .phi.HL supplied to the electrode HL is changed by the
timing signal generator 32 in order to provide for longer transfer
time of the signal charges from this electrode to the branching
section. Moreover, in the present example, the duty cycle and/or
the cycle of one or both of the horizontal drive signals 76a and
76b driving the horizontal transfer paths 56 and 58, respectively
is varied in the timing signal generator 32 in order to provide for
a longer signal charge transfer time from the branching section to
one of the horizontal transfer paths longer than the usual transfer
time.
[0279] The usual transfer time is the transfer time prior to
changing of the duty cycle and/or the cycle of the drive signal,
and means the transfer time in case no transfer efficiency
deterioration has occurred, that is, in case the transfer
efficiency is maintained. In terms of the duty cycle, for example,
the usual transfer time in the present example means the transfer
time in case the duty cycle is 50%, that is, in case the high level
time is about equal to the low level time. It is noted that the
usual transfer time is not limited to that for the duty cycle of
50%, insofar as the transfer efficiency is maintained.
[0280] In FIG. 36, the electrode HL is a transfer element, provided
one in front of the electrode HSL of the branching section, and
transfers the signal charge received from the electrode HS3, to the
electrode HSL. In the present example, the electrode HL is a pair
of polysilicon electrodes arranged as a set, as are the other
electrodes. In addition, the electrode HL may be actuated
independently. In the present example, a drive signal .phi.HL is
supplied to the electrode HL as shown in FIG. 7.
[0281] The drive signal .phi.HL shown in FIG. 38 is the drive
signal .phi.HL supplied to the electrode HL during the usual
driving time. In the present example, the drive signal .phi.HL
during the usual driving is of the same signal waveform as the
drive signals .phi.HS2, .phi.HS4. The reason is that, since the
electrode of the right side neighbor of the electrode HL is the
electrode HS3, it is necessary to provide for the same potential as
that of the electrodes HS2 and HS4 during usual driving. However,
the present invention is not limited to this and the drive signal
to be supplied to the electrode HL may be optionally set depending
on the horizontal transfer path 50. For example, the drive signal
.phi.HL during the usual driving may be the signal of the same
waveform as the drive signals .phi.HS1 and .phi.HS3.
[0282] If, in the present example, there is fear of transfer
deterioration, as in the case of a low color temperature of the
subject, the timing signal generator 32 changes the duty cycle of
the drive signal .phi.HL as shown in FIG. 42. By so doing, the
signal charge transfer time from the electrode HL to the electrode
HSL may be made longer than the transfer time for the usual
transfer time, that is, the transfer time for the case of not
changing the duty cycle.
[0283] FIG. 42 depicts a timing chart schematically showing the
timing of the drive signals to be supplied to the respective
electrodes shown in FIG. 35. Specifically, FIG. 42 schematically
shows the processing in which the duty cycle of the drive signal
supplied to the electrode HL on the occasion of deterioration in
the transfer efficiency is changed so that the signal charge
transfer time from the electrode HL to the electrode HSL is made
longer than that prior to duty cycle change, such as to prohibit
deterioration in the transfer efficiency.
[0284] In FIG. 42, the drive signal .phi.HL has a low level time
longer than the high level time. Referring more specifically to
FIG. 42, the high level time and the low level time of the drive
signal .phi.HL are both time Ta, that is, one-half period, in the
previous example. In the present example, the low level time is the
time Tb, while the high level time is the time Tc. Since the time
Ta is the one-half period, the relationship, time Tb>time
Ta>time Tc, is established.
[0285] In the example shown in FIG. 42, the period is not
modulated. The drive signals .phi.HS1 to .phi.HS4 also are not
modulated. It is because the transfer efficiency on the horizontal
transfer path 50 is to be maintained. However, the present
invention is not limited to this and, for example, the drive
signals .phi.HS1 to .phi.HS4 may be modulated. This drive signal
modulation becomes possible by the system controller 28 controlling
the timing signal generator 32, as an example.
[0286] The signal charges are transferred from the electrode HL to
the electrode HSL when the drive signal .phi.HL is low in level and
the drive signal .phi.HP1 is high in level. Thus, if the duty cycle
of the drive signal .phi.HL is varied as described above to provide
for the high level time Tb of the drive signal .phi.HL longer than
its low level time Tc, it becomes possible to set the signal charge
transfer time from the electrode HL to the electrode HSL from time
Ta to time Tb which is longer than time Ta. The result is that the
signal charge transfer from the electrode HS to the electrode HSL
may be better and the amount of charges left over untransferred may
be decreased to eliminate the problem of deterioration in the
transfer efficiency.
[0287] By changing the duty cycle of the drive signal .phi.HL in
this manner, the high level time in the drive signal .phi.HL may be
shorter, so that the transfer time to the electrode HL from the
electrode as the left side neighbor of the electrode HL may be
shorter. For example, in the example of FIG. 35, the signal charge
transfer time from the electrode HS3 to the electrode HL becomes
shorter. However, there is sufficient allowance in frequency
characteristics in transferring signal charges from the electrode
HS3 to the electrode HS, so that there is sufficient margin in
decreasing the transfer time. Hence, transfer may be achieved
unobjectionably by applying this margin to the transfer time from
the electrode HS to the electrode HSL.
[0288] It is noted that the present invention is not limited to
changing the duty cycle of the drive signal .phi.HL. More
specifically, the duty cycle and the period of each of the drive
signal .phi.HL, horizontal drive signal .phi.HP1 and the horizontal
drive signal .phi.HP2 may be changed by the timing signal generator
32, as shown for example in FIG. 43, in order to provide for a
longer signal charge transfer time from the branching section to
one of the horizontal transfer paths.
[0289] FIG. 43 is a timing chart schematically showing another
timing of the drive signals supplied to the electrodes shown in
FIG. 35 in case of transfer efficiency deterioration. Specifically,
FIG. 43 schematically shows the processing for eliminating the
transfer efficiency deterioration, according to which the duty
cycle and the period of the drive signal .phi.HL, horizontal drive
signals .phi.HP1 and .phi.HP2 are changed so that the transfer time
of signal charges from the electrode HSL to the horizontal transfer
path 56 in case of transfer efficiency deterioration will be longer
than that before such change, such as to eliminate the transfer
efficiency deterioration. In FIG. 43, the same reference numerals
as those used in FIG. 42 denote the same or equivalent component
parts.
[0290] In the example shown in FIG. 43, not only the duty cycle of
the drive signal .phi.HL but also that of each of the drive signals
.phi.HP1 and .phi.HP2 is changed. More specifically, the duty cycle
of the drive signals .phi.HP1 and .phi.HP2 is changed in FIG. 43 so
that, in the drive signal .phi.HP1, the high level time is changed
from time Td to time Tp and the low level time is changed from time
Te to time Tq. On the other hand, since the drive signal .phi.HP2
is reverse-phased from the drive signal .phi.HP1, its low level
time and high level time are set to time Tp and Tq, respectively.
Meanwhile, since the pre-change time Td is about equal in length as
pre-change time Te, and is equal to one-half the period of each of
the drive signals .phi.HP1 and .phi.HP2, the time Td, Te, Tp and Tg
are related to one another by Tp>Td, Te>Tq.
[0291] In keeping with such changes in the drive signals .phi.HP1
and .phi.HP2, the period of the drive signal .phi.HL is changed, so
that the signal part with the period equal to Tp and that with the
period equal to Tq will appear alternately. In the example shown in
FIG. 43, the duty cycle of the drive signal .phi.HL is set so that
the low level period is loner than the high level period as in FIG.
42. For example, in the present example, in a signal part with one
period equal to time Tp, the low level time is time Tl, while the
high level time is time Tm, with Tl>Tm, whereas, in a signal
part with one period equal to time Tq, the low level time is time
Tn, while the high level time is time To, with Tn>To.
[0292] In the present example, even though the period of the drive
signal .phi.HL differs from one signal part to the next, the duty
cycle of the signal part with the period equal to time Tp is made
equal to that of the signal part with the period equal to time Tq.
In more detail, if, in the signal part where the period is equal to
Tp, the low-level time Tl is 60% of time Tp and the high-level time
Tm is 40% of time Tp, the low-level time Tn is 60% of time Tq and
the high-level time To is 40% of time Tq. However, the present
invention is not limited to this and may be modified optionally.
For example, the duty cycle may be changed for the signal part with
the period equal to Tp and for the signal part with the period
equal to Tq. Alternatively, the duty cycle may be set to 50% for
both of the signal parts and the lengths of both the low level time
and the high level time may be equal to one half period, only by
way of illustration.
[0293] The period of each of the drive signals .phi.HS1 to .phi.HS4
is also changed with change in the drive signal .phi.HL.
Specifically, one period of each of the drive signals .phi.HS1 to
.phi.HS4 in register with the signal part of the drive signal
.phi.HL with the period of Tp is set to time T3 which is about
equal to time Tp. In similar manner, one period of each of the
drive signals .phi.HS1 to .phi.HS4 in register with the signal part
of the drive signal .phi.HL with the period of Tq is set to time T4
which is about equal to time Tq. It is noted that, in each cycle,
the low and high levels are each of one-half cycle, without the
duty cycle being changed.
[0294] Suppose that the duty cycle of the drive signals .phi.HP1
and .phi.HP2 is changed in this manner, so that the high level time
of the drive signal .phi.HP1 is made longer. The signal charge
transfer time from the electrode HSL to the horizontal transfer
path 56 may then be made longer, since the signal charge is
transferred from the electrode HL to the horizontal transfer path
56 during the high level time of the drive signal .phi.HP1.
Moreover, if the period of the drive signal .phi.HP is changed to a
longer period, the time during which the drive signal .phi.HL is
low in level and the drive signal .phi.HP1 is high in level may be
made longer, even in case the duty cycle is 50%, whereby it is
possible to provide for longer signal charge transfer time from the
electrode HL to the electrode HSL.
[0295] In particular, if the duty cycle of the drive signal .phi.HL
is also changed, as in the present alternative embodiment, the time
during which the drive signal .phi.HL is low in level and the drive
signal .phi.HP1 is high in level may be made longer, whereby it is
possible to provide for longer signal charge transfer time from the
electrode HL to the electrode HSL. Meanwhile, whether or not the
duty cycle is to be changed when the period of the drive signal
.phi.HL is changed may optionally be determined depending on the
state of transfer then prevailing in the transfer section.
[0296] When the duty cycle of the drive signals .phi.HP1 and
.phi.HP2 is changed, the transfer time from the electrode HSL to
the horizontal transfer path 58 becomes shorter from Te to Tq (time
Te>time tq). It is noted that, in case the signal charges of the
pixels R and B are being transferred to the horizontal transfer
path 56 and the signal charges of the pixel G are being transferred
to the horizontal transfer path 58, as in the present alternative
embodiment, the signal charge of the pixel G tends to be mixed into
those of the pixels R and B, since the transfer time from the
branching section to the horizontal transfer path 58 becomes
shorter. However, since the signal quantities of the pixels R and B
differ from each other, the adverse effect of mixing of signal
charges, if any, is only small.
[0297] In case the duty cycle of the drive signals .phi.HP1 and
.phi.HP2 is changed as shown in FIG. 43, the reset level Tr and the
feed-through level Ts of the output waveforms OS1, OS2 become
shorter, and the data level Tt becomes longer. Thus, in case the
noise is removed in the rear side pre-processor 22 in accordance
with the correlated double sampling, it becomes necessary to change
the phase of the sampling pulse in keeping with the change in the
drive signals .phi.HP1 and .phi.HP2.
[0298] In the present alternative embodiment described above, the
duty cycle or the period of the drive signals .phi.HL, .phi.HP1 and
.phi.HP2 is changed to provide for longer transfer time from the
electrode HL towards the electrode HSL or longer transfer time from
the electrode HSL towards the horizontal transfer path 56 to
provide for transfer of a sufficient quantity of signal charges, as
shown in FIGS. 44 and 45. Consequently, the transfer efficiency may
be prevented from being deteriorated in the electrode HSL. In
addition, since the duty cycle or the period may be changed by
changing the timing signal generated by the timing signal generator
32 shown for example in FIG. 2, it is possible to prevent
deterioration of the transfer efficiency without requiring
redundant elements.
[0299] The above-described driving with variable duty cycle or
period of the drive signals .phi.HL, .phi.HP1 and .phi.HP2 for
eliminating the deterioration of transfer deterioration may be
effected depending on, for example, the temperature of the device
44, color temperature of the subject, ISO (International
Organization for Standardization) sensitivity or the driving speed.
For example, the transfer efficiency tends to be deteriorated under
low temperature in the device 44. This deterioration may be coped
with by driving with variable duty cycle or period.
[0300] Meanwhile, the temperature of the device 44 may be measured
by known temperature measurement means, such as a thermometer or a
sensor, as mounted in an optional location of the solid state
imaging apparatus, such as in the imaging unit 14 or in the system
controller 28. If the measured temperature is lower than a preset
value, the timing signal generator 32 may be controlled by, for
example, the system controller 28, in order to change the duty
cycle or the period of each drive signal, whereby it becomes
possible to prevent deterioration of the transfer efficiency under
low temperatures, only by way of example.
[0301] Meanwhile, if the detected temperature is higher than a
preset value, the transfer efficiency becomes higher. Hence, usual
driving is preferably used. Specifically, with a high detected
temperature, the timing signal generator 32 routes to the drivers a
usual timing signal in which the duty cycle is 50%, with the low
level time and the high level time each being one-half period.
[0302] In case the color temperature is drastically high or low, an
ill effect caused by mixing of the signal charges of the pixel R
and the pixel B into those of the pixel G is increased. The duty
cycle or the period may then be changed to eliminate transfer
deterioration to combat the ill effect caused by mixing.
[0303] The color temperature may be said to be drastically high in
case it is higher than 6000 Kelvin, as an example. The color
temperature may be said to be drastically low in case it is lower
than 3000 Kelvin, as an example. The present invention is not
limited to these numerical values. The color temperature may be
detected using known techniques.
[0304] In driving at high ISO sensitivity, that is, in case the
imaging optical sensitivity is higher than the usual imaging
sensitivity, the subject is low in luminance and hence the quantity
of signals generated may be low. Hence, the ill effect caused by
mixing tends to be increased, the duty cycle or the period may then
be changed to eliminate transfer deterioration to combat the ill
effect caused by mixing.
[0305] During high speed driving, the transfer time becomes shorter
than during usual driving. Hence, it is feared that the quantity of
signal charges left over is increased. Thus, by varying the duty
cycle or the period for driving, it becomes possible to prevent
mixing to generate an optimum image. Meanwhile, during the low
speed driving, it is preferred to revert to usual driving, because
sufficient transfer time may then be secured. However, the present
invention is not limited to these cases. The duty cycle or the
period may be changed in an optional case where there is fear of
deterioration in the transfer efficiency in the branching section
54, in order to eliminate deterioration.
[0306] In the processing shown in FIGS. 44 and 45, it is possible
to freely set how much the duty cycle is to be changed. That is,
the range of variations of the duty cycle may be set freely. This
setting may be made by measuring, in a situation where
deterioration in the transfer efficiency is likely to be produced,
the quantity of signal charges which may be left over to the rear
side, and by calculating the transfer efficiency using the so
measured quantity of signal charges left over to the rear side.
[0307] FIG. 44 is a flow chart showing typical processing of
calculating the transfer efficiency, by measuring the residual
transfer quantity, for setting the magnitude of variation of the
duty cycle. In FIG. 44, the system controller 28 captures a light
source of a predetermined light volume in the imaging unit 14 to
generate a reference signal (step S1). Under such control, the
imaging unit 14 shoots the light source of a predetermined light
volume and, as shown in FIGS. 31 to 34, mixes eight pixels in the
horizontal direction on the horizontal transfer path 50 to generate
a reference signal pixel 400 and at least three void pixels 402 to
406 consecutive to the reference signal pixel on its rear side.
Meanwhile, it is sufficient that there are at least two pixels on
the rear side of the pixel 400, so that the present alternative
embodiment is merely an illustration and is not restrictive.
[0308] In case the reference signal pixel 400 and three consecutive
void pixels are generated in this manner in rear of the reference
signal pixel 400, the branching section 54 of the device 44
bifurcates the reference signal pixel 400 and the void pixels, so
that the reference signal pixel 400 and one of the three void
pixels are supplied to one of the horizontal transfer paths, and
the remaining two void pixels are supplied to the other horizontal
transfer path (step S2).
[0309] FIG. 45 schematically shows how the reference signal pixel
400 and the three consecutive void pixels in rear of the reference
signal pixel 400, generated by the processing shown in FIGS. 31 to
34, are being transferred from the branching section 54 to the
horizontal transfer paths 56 and 58. Specifically, FIG. 45
schematically shows the exemplary processing for calculating the
transfer efficiency. In FIG. 45, the same reference numerals as
those of FIG. 35 denote the same component parts. In FIG. 45, the
reference signal pixel 400 and void pixels 402, 404, 406 are
supplied from the horizontal transfer path 50 to the branching
section 54. As a result of bifurcation at the branching section 54,
the reference signal pixel 400 and the void pixel 404 are
sequentially supplied to the horizontal transfer path 56. To the
horizontal transfer path 58, the void pixel 402 directly in rear of
the reference signal pixel 400 on the horizontal transfer path 50,
and the void pixel 406 are sequentially supplied.
[0310] The reference signal pixel 400 and the void signal 404 are
sequentially transferred on the horizontal transfer path 56, to the
output amplifier 60 which then outputs the signal 82 composed of
the reference signal pixel 400 and the void signal 404. In similar
manner, the void pixels 402, 406 are sequentially transferred on
the horizontal transfer path 58 to the output amplifier 62 which
then outputs the signal 84 composed of the void pixels 402, 406.
The signals 82, 84 are then processed by the pre-processor 22 to
generate digital signals 110, 112, which are then stored in the
memory 24.
[0311] At the time of branching to the horizontal transfer path 56
from the branching section 54, if there are left-over signal
charges, these left-over signal charges are admitted into the
branched void pixel 402. The signal processor 26 reads out the
reference signal pixel 400 and the void pixel 402 from the memory
24, as digital signal 118, over bus 114 and signal line 120, and
calculates the transfer efficiency HTR.sub.HSL1 at the time of
transfer from the branching section 54 to the horizontal transfer
path 56, in accordance with the expression (1) (step S3):
( S - T ) .times. 100 S ( 1 ) ##EQU00001##
where S denotes the signal quantity of the pixel 400, T the
quantity of left-over signal charges detected from the void pixel
402, that is, the quantity of residual signal charges. Meanwhile,
the present invention is not limited to this processing. The
invention can deal with any pixel.
[0312] After branching, signal charges left over to the rear side
on the horizontal transfer path 56 are admitted into the void pixel
404. Thus, in the present alternative embodiment, the signal
processor 26 detects the quantity of residual signal charges
present in the void pixel 404 to calculate the transfer efficiency
HTR.sub.OS1 (step S4). It is similarly possible to calculate the
transfer efficiency HTR.sub.OS1 by setting the quantity of residual
signal charges detected from the void pixel 404, as the variable T
in the above expression (1). In particular, the residual charges
left over in the ultimate stage on the horizontal transfer path 56,
that is, between the output gate and the floating diffusion
amplifier shown in FIG. 36 are admitted into the void pixel 404. It
is therefore useful in improving the ill effect, such as image
deterioration, to maintain the transfer efficiency calculated with
the use of the void pixel 404.
[0313] In this manner, the transfer efficiency in branching from
the branching section 54 to the horizontal transfer path 56 and the
transfer efficiency on the horizontal transfer path 56 are
calculated. The transfer efficiency HTR.sub.HSL2 and the transfer
efficiency HTR.sub.OS2 on the horizontal transfer path 58 may
similarly be calculated as shown in FIG. 46. Specifically, the
transfer efficiency HTR.sub.OS2 on the horizontal transfer path 58
may be calculated by supplying the reference signal pixel 400,
prepared by the processing of FIGS. 31 through 32I, to the
horizontal transfer path 58, and by subsequently detecting signal
charges present in the void pixel 404, as shown in FIG. 46. The
transfer efficiency HTR.sub.HSL2 in branching signal charges from
the branching section 54 to the horizontal transfer path 58 may
also be calculated by detecting the signal charges present in the
void pixel 404.
[0314] The transfer efficiency calculated may be that for the
reference signal pixel 400 of a preset signal quantity, as an
example. However, since the residual charge quantity is varied with
the signal quantity of the reference signal pixel 400 as shown in
FIGS. 47 and 48, it is also possible to provide several reference
signal pixels 400 of different signal quantities and to calculate
the transfer efficiency from one signal quantity to another. It is
noted that the present invention is not limited to calculating the
transfer efficiency by the technique shown in FIGS. 46 to 48 and
any suitable known technique may also be used.
[0315] FIG. 47 schematically shows measured results of the residual
charge quantity for different signal quantities of the reference
signal pixel 400. FIG. 48 schematically shows a transfer efficiency
calculated from the residual charge quantity as shown in FIG. 47.
In FIG. 47, the abscissa and the ordinate represent the signal
quantity (mV) of the reference signal pixel 400 and the residual
charge quantity (mV), respectively. FIG. 47 shows the results of
detection of the residual signal charges in the branching section
54. In this figure, a curve 232 stands for residual charges on
bifurcating signal charges from the branching section 54 to the
horizontal transfer path 56, that is, the signal quantity detected
from the void pixel 402 of FIG. 45. Another curve 234 stands for
residual charges on bifurcating signal charges from the branching
section 54 to the horizontal transfer path 58, that is, the signal
quantity detected from the void pixel 402 of FIG. 46. On the other
hand, there are shown in FIG. 48 the transfer efficiency
HTR.sub.HSL1 and the transfer efficiency HTR.sub.HSL2 as calculated
from the respective values shown in FIG. 47. The present invention
is not limited to calculating the transfer efficiency at the
branching section 54. That is, the transfer efficiencies in the
horizontal transfer paths 56 and 58 may similarly be calculated
from one signal quantity of the reference signal pixel 400 to
another.
[0316] FIG. 49 is a flowchart showing exemplary processing for
calculating the transfer efficiency by the solid state imaging
apparatus 10 in accordance with the sequence shown in FIG. 44 and
for setting variable values of the duty cycle with the use of the
so calculated transfer efficiency. In FIG. 49, the transfer
efficiency of a section, for which are set the variable values of
the duty cycle or the period, is calculated in accordance with the
sequence shown in FIG. 44 for setting the variable values (step
S11). For example, in determining the variable values used in
increasing the transfer time from the branching section 54 to the
horizontal transfer path 56, in the present alternative embodiment,
a reference signal pixel 400 of a certain signal quantity is
generated, and the transfer efficiency HTR.sub.HSL1 and the
transfer efficiency HTR.sub.HSL2 are calculated as shown in FIG.
48.
[0317] Meanwhile, the transfer efficiency calculated at this time
is used as a reference in setting the variable values. Thus, in
calculating the transfer efficiency at the step S11, it is
preferred to set the duty cycle of each of the drive signals
.phi.HL, .phi.HP1 and .phi.HP2 to 50% and to set a constant period,
for usual driving. However, this is not meant for restricting the
present invention.
[0318] Once the transfer efficiencies in bifurcating the signal
charges from the branching section 54 to the horizontal transfer
paths 56 and 58 are obtained, it is verified whether or not one of
the transfer efficiencies exceeds a reference value (step S12). For
example, in the present alternative embodiment, it is verified
whether or not the transfer efficiency HTR.sub.HSL1 of the
horizontal transfer path 56 exceeds the reference value.
[0319] For the reference value, an optionally set value may be
used. For example, it may be set based on the transfer efficiencies
HTR.sub.HSL1 or HTR.sub.HSL2'' as calculated in the step S11, or
may be empirically set. However, this is not meant to restrict the
present invention. For example, in the present alternative
embodiment, the reference values are set as indicated by dotted
lines 242, 244 in FIG. 48. In this figure, the dotted line 242 is a
reference value of the transfer efficiency in branching from the
branching section 54 to the horizontal transfer path 56, while the
dotted line 244 is a reference value of the transfer efficiency in
branching from the branching section 54 to the horizontal transfer
path 58.
[0320] It is noted that an image generated is affected severely by
transfer deterioration that should occur in branching from the
branching section 54 to the horizontal transfer path 56. Thus, in
the present alternative embodiment, the reference value of the
transfer efficiency in branching from the branching section 54 to
the horizontal transfer path 56 is set more severely than that in
branching from the branching section 54 to the horizontal transfer
path 58. The present invention is not limited to this and the same
reference value may be set for branching from the branching section
54 to the horizontal transfer path 56 and for branching from the
branching section 54 to the horizontal transfer path 58.
[0321] As a result of decision in the step S12, if the transfer
efficiency HTR.sub.HSL1 is higher than the reference value,
processing transfers to a step S13 in order to verify whether or
not the other transfer efficiency, specifically the transfer
efficiency HTR.sub.HLS2, for branching from the branching section
54 to the horizontal transfer path 58 exceeds the transfer
efficiency (step S13).
[0322] As a result of decision in the step S13, if the transfer
efficiency HTR.sub.HSL2 also exceeds the reference value, the
processing for setting the variable value is finished. If the
result of decision in the step S13 indicates that the transfer
efficiency HTR.sub.HSL2 is lower than the reference value, the duty
cycle or the period of each drive signal is adjusted (step S14).
Specifically, the duty cycle or the period of each drive signal is
adjusted. It may optionally be set how much the duty cycle or the
period is to be changed. For example, the duty cycle or the period
may be set depending on how much the transfer efficiency
HTR.sub.HSL1 exceeds the reference value, or on how much the
transfer efficiency HTR.sub.HSL2 falls below the reference value.
The duty cycle or the period may also be set as the amount of
change for each adjustment event, for example, the value of
increase of the duty cycle, such as 5%, is set and the adjustment
is then made based on this change amount. However, this is not
meant for restricting the present invention.
[0323] After the adjustment in the step S14, processing reverts to
the step S11 to calculate the transfer efficiencies HTR.sub.HSL1,
HTR.sub.HSL2 again. It is then verified whether or not the transfer
efficiencies HTR.sub.HSL1, HTR.sub.HSL2 exceed the reference
values. If the transfer efficiencies HTR.sub.HSL1, HTR.sub.HSL2
exceed the reference values, processing is terminated. Why the
processing reverts to the step S11 is to verify whether or not the
transfer efficiency HTR.sub.HSL2 in switching from the branching
section 54 to the horizontal transfer path 58 has been improved,
and also is to verify whether or not the transfer efficiency
HTR.sub.HSL1 in switching from the branching section 54 to the
horizontal transfer path 56 has become lower than the reference
value as a result of change in the duty cycle or the period in the
step S14.
[0324] As a result of decision in the step S12, if the transfer
efficiencies HTR.sub.HSL1 has become lower than the reference
value, processing transfers to a step S15 to verify whether or not
the other transfer efficiency HTR.sub.HSL2 has exceeded the
reference value (step S15). If the result of decision indicates
that the other transfer efficiency HTR.sub.HSL2 has also become
lower than the reference value, it is determined that the variable
value cannot be set (step S16) to terminate the processing.
[0325] If the transfer efficiency HTR.sub.HSL2 has exceeded the
reference value, the duty cycle or the period of each drive signal
is adjusted so that the transfer efficiency HTR.sub.HSL1 will
become higher (step S17). How much the duty cycle or the period is
to be changed may optionally be set, as in the step S14. After the
change, processing reverts to the step S11 to verify how much the
transfer efficiency HTR.sub.HSL1 has been improved by the change in
the step S17, and whether or not the other transfer efficiency
HTR.sub.HSL2 has become lower than the reference value.
[0326] The variable value is set as the transfer efficiencies
HTR.sub.HSL1, HTR.sub.HSL2 are measured, as described above. It is
therefore possible in this manner to avoid the problem that, while
the transfer efficiency on one of the horizontal transfer paths may
be improved as a result of changes in the duty cycle or the period,
the transfer efficiency on the opposite side horizontal transfer
path is deteriorated, so that it is possible to obtain the variable
value which may be optimum case by case. Although it is the
variable value in branching from the branching section 54 to the
horizontal transfer paths 56, 58 that is set in the present
alternative embodiment, it is possible to set the variable value in
changing the transfer time from the electrode HL before branching
to the electrode HSL of the branching section 54 in similar
manner.
[0327] The variable values may preferably be set under a condition
which may aggravate the transfer efficiency, such as at elevated
temperatures, higher sensitivity, high speed readout or at an
extremely high or extremely low color temperature, since it is then
possible to obtain the variable value which will improve transfer
deterioration more effectively. It is also preferred to calculate
the transfer efficiency or set the variable values at the time of
shipment of the device 44 or the solid state imaging apparatus 10
since then it is possible to take up individual differences of the
solid state imaging apparatus 10 from one solid state imaging
apparatus to another. Of course, the present invention is not
limited to this and the variable values may be set under optional
conditions and at optional stages.
[0328] A further alternative embodiment of the present invention
will now be described. FIG. 50 shows an alternative embodiment of
the device 44 according to the present invention. In case signal
charges for plural colors are transferred on the horizontal
transfer path 50, the signal charges are bifurcated by the
branching section 54 from color to color, to each of the plural
horizontal transfer paths 56 and 58, and analog voltage signals 82,
84 converted from the signal charges are output simultaneously. In
case signal charges for one color are transferred on the horizontal
transfer path 50, the analog voltage signal 82 converted from the
signal charges are output from the selected horizontal transfer
path 56, as an example.
[0329] In the present alternative embodiment, the color filter
separates the incident light into three colors of red, green and
blue. The three colors are divided into a group consisting of red
and blue, and another group consisting of green. Of the three
colors of red, green and blue constituting a line, green is read
out separately from the red and blue. Then, of the three colors of
red, green and blue constituting the next line, green is read out
separately from the red and blue. This operation is repeated for
all lines making up the entire pixels (one frame). That is, the
readout system is progressive.
[0330] The signal charges for green color read out from the
photosensitive cells are transferred by using only the horizontal
transfer path 56. The signal charges for red and blue colors read
out from the photosensitive cells are transferred by using the
horizontal transfer paths 56 and 58. The red color and the blue
color are transferred over the horizontal transfer paths 56 and 58,
respectively.
[0331] The digital camera 10 includes the optical system 12, the
imaging unit 14, amplifier power supply 16, biasing circuit 18,
drivers 20, pre-processor 22, memory 24, signal processor 26,
system controller 28 and timing signal generator 32 as shown in
FIG. 50.
[0332] The branching section 54 is supplied from the biasing
circuit 18 with a bias signal 72 as a fixed voltage. The branching
section 54 causes signal charges from the horizontal transfer path
50 to be branched to one of the horizontal transfer paths 56 and
58, depending on the colors, in a manner which will be described
later.
[0333] The HS driver 96 outputs the horizontal drive signal 74 to
the device 44. The HP driver 98 outputs the horizontal drive signal
76 to the device 44. The horizontal drive signal 76 has a period
equal to or twice the period of the horizontal drive signal 74,
depending on which color is being transferred. In the present
alternative embodiment, the period of the horizontal drive signal
76 is constant, and the period of the horizontal drive signal 74 is
equal to or half the period of the horizontal drive signal 76,
depending on the colors. The RS driver 100 outputs the reset signal
68 and 70 to the device 44.
[0334] The signal processor 26 has the function of performing
signal processing on the digital signal 118 supplied to generate a
control signal. The signal processor 26 includes the power supply
control 122 and the bias control 124. The signal processor 26 shown
in FIG. 50, of course, includes those components corresponding to
the AF control 126, AE control 128, AWB control 130 and data
converter 132 shown in and described with reference to FIGS. 2 and
23, which are not shown merely for the purpose of simplicity. The
power supply control 122 has the function of generating the control
signal 86 responsive to high speed readout or low speed readout.
The power supply control 122 outputs the control signal 86
generated to the amplifier power supply 16.
[0335] The bias control 124 outputs the control signal 88 to the
biasing circuit 18. The biasing circuit 18 applies the bias signal
72 to the branching section 54.
[0336] The constitution of the horizontal transfer path 50, the
branching section 54 and the horizontal transfer paths 56 and 58,
and the method for horizontal transfer of signal charges will now
be described. The following description will be made separately for
transfer on the branching section 54 and the horizontal transfer
path 56 and for transfer on the branching section 54 and the
horizontal transfer path 58. There are provided plural transfer
elements as described above, on each of the horizontal transfer
paths 50, 56 and 58. In the following, the sole transfer element
and the two electrodes contained therein are denoted by the same
reference numeral. For example, the `branching section 54` denotes
a transfer element, and the `electrode 54` denotes two electrodes
of the branching section 54.
[0337] On the horizontal transfer path 50, there are formed
polysilicon electrodes HS2 and HS1, in this order, from right to
left, towards the branching section 54 (electrode HSL), as shown in
FIG. 51. This set of the polysilicon electrodes constitutes a
repetitive unit. On the horizontal transfer path 56, there are
provided six polysilicon electrodes HP1, HP2, HP1, HP2, HP1, HP2,
from the branching section 54 towards the output amplifier 60. On
the horizontal transfer path 58, there are provided seven
polysilicon electrodes HP2, HP1, HP2, HP1, HP2, HP1, HP2, from the
branching section 54 towards the output amplifier 62. The
horizontal transfer paths 56 and 58 are of the same structure as in
FIGS. 4 and 7, as may be understood from FIGS. 51 and 52.
[0338] The drive signals applied to the respective electrodes will
now be described. The drive signals .phi.HS1 and .phi.HS2 are
supplied to the electrodes HS1 and HS2, respectively. The drive
signal .phi.HSL is supplied to the electrode HSL. The drive signal
.phi.HSL is a constant bias voltage. The drive signals .phi.HP1 and
.phi.HP2 are supplied to the electrodes HP1 and P2, respectively. A
drive signal .phi.OG is supplied to the electrode OG. The drive
signal .phi.OG is a constant bias voltage. The drive signal .phi.RS
is supplied to the reset drain RS. The drive signal .phi.RS is a
constant bias voltage.
[0339] The flow of signal charges, transferred horizontally by
these drive signals, inclusive of vertical transfer preceding
horizontal transfer, will now be described. Initially, the signal
charges are generated in the device 44 of the imaging unit 14. The
array of pixels in the device 44 is shown in FIG. 53. In the device
44 of the present alternative embodiment, shown in FIG. 53, the
filter array is generally of the so-called G-square RB-complete
checkered pattern.
[0340] The charge transfer sequence on the vertical transfer path
48 is lines 201, 252, 203 and 254. On the line 201, the signal
charges are arrayed in the order of R, B, R, B, . . . . On the
lines 252, 254, the signal charges are arrayed in the order of G,
G, G, G, . . . . On the line 203, the signal charges are arrayed in
the order of B, R, B, R, . . . . This sequence is maintained on the
horizontal transfer paths 50, 56 and 58, so that the signal charges
are transferred in the sequence of R, B, R, B, . . . of line 201,
G, G, G, G, . . . of line 252, B, R, B, R, . . . of line 203 and G,
G, G, G, . . . of line 254.
[0341] It is noted that R, B, R, B, . . . on the line 201 and B, R,
B, R, . . . on the line 203 are transferred on the two horizontal
transfer paths 56 and 58, while G, G, G, G, . . . on the lines 252,
254 are transferred on the sole horizontal transfer path 56.
Moreover, as regards the R, B, R, B, . . . on the line 201 and B,
R, B, R, . . . on the line 203, these are separated into R and B
signals, by the branching section 54, so that the R signals and B
signals are transferred at all times on the horizontal transfer
paths 56 and 58, respectively.
[0342] This is shown in FIGS. 54A through 55E. FIGS. 54A through
54E show how the R, B, R, B, . . . on the line 201 and B, R, B, R,
. . . on the line 203 are transferred horizontally at times t=1, 2,
3, 4 and 5. FIGS. 55A through 55E show how the G, G, G, G, . . . on
the lines 252, 254, transferred next to lines 201 and 203,
respectively, are transferred horizontally at times t=1, 2, 3, 4
and 5. The times t=1, 2, 3, 4 and 5 are those at which signal
charges are transferred on the horizontal transfer path 50 towards
left by one transfer element at a time. Since the horizontal
transfer paths 56 and 58 are driven in FIG. 54 at a frequency equal
to one-half that for the horizontal transfer path 50, the transfer
speed on the horizontal transfer paths 56 and 58 is one-half that
on the horizontal transfer path 50. On the other hand, the
horizontal transfer paths 56 and 58 are driven in FIG. 55 at a
frequency equal to that for the horizontal transfer path 50, and
the transfer speed on the horizontal transfer paths 56 and 58 is
equal to that on the horizontal transfer path 50. The method for
implementing this will be described in detail subsequently.
[0343] In the signal processor, next following the horizontal
transfer paths, processing is carried out on the assumption that a
first line is formed by signal charges on the lines 201 and 252 and
that a second line is formed by signal charges on the lines 203 and
254. Meanwhile, RGBGRGBG . . . and BGRGBGRG . . . are stated as
first and second lines on the horizontal transfer path 50 shown in
FIG. 51. This indicates in which locations on the horizontal
transfer path 50 the signal charges descend in the drawing from the
vertical transfer paths 48, but does not indicate that the signal
charges are transferred in the sequence of RGBGRGBG . . . and
BGRGBGRG . . . on the horizontal transfer path 50.
[0344] Reverting to FIGS. 55A through 55E, the signal charges read
out from the photosensitive cells 46 to the vertical transfer path
48 are transferred on the vertical transfer path 48 towards the
horizontal transfer path 50 by eight-phase drive signals .phi.V1 to
.phi.V8 supplied to the vertical transfer path 48. On the
horizontal transfer path 50, there are provided the electrodes HS1,
HS2, HS1, HS2, . . . are provided from its left end, as described
previously.
[0345] With the honeycomb array and with G-square RB-complete
checkered pattern, as in the present alternative embodiment, the G
signals and the RB signals descend on separate lines, that is,
lines 201 and 252, respectively. However, with a routine array of
photosensitive cells and the color filter segments, the G signals
and the RB signals do not necessarily descend in separated states.
In such case, it becomes necessary to re-array the G signals and
the RB signals. To this end, it is sufficient to provide a line
memory between the vertical transfer path 48 and the horizontal
transfer path 50 for re-arraying the signals and subsequently
transfer the re-arrayed signals to the horizontal transfer path
50.
[0346] The drive signals on the horizontal transfer paths 50, 54,
56 and 58 will now be described. Initially, the case of
transferring R, B, R, B, . . . on the line 201 and B, R, B, R, . .
. on the line 203 will be described. FIG. 56 shows drive signals
for transferring R, B, R, B, . . . on the lines 201 and 203. If
attention is directed to the phase of the drive signals, the drive
signal .phi.HS1 of FIG. 56, part (A), is a two phase drive signal
phase-shifted by 180.degree. from the drive signal .phi.HS2 of part
(B). The drive signal .phi.HP1 of part (C) and the drive signal
.phi.HP2 of part (D) are phase-reversed from each other and are two
phase drive signals.
[0347] If attention is directed to the periods of the drive
signals, the drive signals of FIG. 56, parts (A) and (B) are each
of a period equal to one-half the period of parts (C) and (D). That
is, the frequency of each of the drive signals of parts (A) and (B)
is twice the frequency of each of the drive signals of parts (C)
and (D). The drive signal .phi.RS as shown in part (E) becomes "H"
in level at, e.g. the time t=1, . . . , t=5, that is at time
t=4n+1, where n is an integer including zero. Output signals OS1
and OS2 are output as shown in part (F).
[0348] FIGS. 57A through 61B show the potential generated on the
horizontal transfer paths 50, 54, 56 and 58 in case the above drive
signals are applied to the horizontal transfer paths 50, 54, 56 and
58. Those figures also depict the structure shown in FIG. 52. A
simplified diagram of FIG. 52 is also shown for indicating the
potential positions. FIGS. 57A and 57B are for time t=1 of FIG.
54A, FIGS. 58A and 58B are for time t=2 of FIG. 54B, and FIGS. 59A
and 59B are for time t=3 of FIG. 54C. FIGS. 60A and 60B are for
time t=4 of FIG. 54D, and FIGS. 60A and 61B are for time t=5 of
FIG. 54E. FIGS. 57A, 58A and 59A, and 60A and 61A stand for the
horizontal transfer paths 50, 54 and 56. FIGS. 57B, 58B and 59B,
and 60B and 61B stand for the horizontal transfer paths 50, 54 and
58. Thus, as may be seen from the figures, the potential levels of
FIGS. 57A, 58A, 59A, 60A and 61A, and of FIGS. 57B, 58B, 59B, 60B
and 61B are the same insofar as the horizontal transfer paths 50
and 54 are concerned.
[0349] Referring further to FIGS. 56 through 61B, horizontal
transfer will be described. Since the drive signal .phi.HSL is
supplied, in a manner not shown, there are generated a potential
level of the reference level 146 always fixed, and a potential
level (barrier) 148 which prohibits reverse flow of signal charges
supplied from the horizontal transfer path 50, in a region directly
below the electrode HSL supplied with the drive signal
.phi.HSL.
[0350] The shifting of the signal charges by the variable potential
levels generated responsive to the drive signals supplied, and the
constant potential levels 146, 148 will now be described. The
signal charges corresponding to the colors R, G and B are referred
to below as signal charges R, G and B. Initially, the transfer of
the signal charge R on the horizontal transfer path 56 will be
described with reference to FIGS. 57A, 58A, 59A, 60A and 61A.
[0351] The reason the signal charges may be distributed in the
branching section 54 in the present alternative embodiment will now
be described. There are provided impurity layers directly
underneath the electrodes HP2 and HP1 of the horizontal transfer
paths 56 and 58 next following the branching section 54. Thus, when
the electrodes HP2 and HP1 are supplied with the level "H", there
are generated stepped potential levels, that is, a level lower by
one step than the reference level 146 and the deepest level. The
reference level 146 is generated at all times in the branching
section 54 by the constant bias voltage. This is shown for example
in FIG. 57B. When the electrodes HP2 and HP1 are supplied with the
level "L", there are generated a potential level higher by one step
than the reference level 146 and a potential level which is the
same as the reference level 146. This is shown for example in FIG.
57A. Hence, the potential level generated is sequentially lowered
in steps along the signal charge transfer direction. Based on the
above, the operation with lapse of time will now be described.
[0352] At time t=1 in FIG. 56, the drive signals .phi.1, .phi.HS2,
.phi.HSL and .phi.HP1 are supplied. The drive signals .phi.HS2 and
.phi.HP1 are "L" in level. In case the drive signals are supplied
in this manner, the signal charge B is retained in the branching
section 54. At this time, there is generated, in the impurity layer
directly underneath the electrode HP1 of the horizontal transfer
path 56 neighboring to the electrode HSL, by the above supply of
the level "L", a potential level 148 or a barrier which is just
enough to prevent the signal charge B from mixing into the
horizontal transfer path 56.
[0353] The electrode HP2 on the horizontal transfer path 58,
neighboring to the branching section 54, is supplied with the
potential level "H" of the drive signal .phi.HP2. By this, a
potential level 152 lower than the reference level is generated as
shown in FIG. 57B such as to permit the signal charge B to flow
into the horizontal transfer path 58. At this time, the signal
charge B is held in both packets of the reference level 146 and the
potential level 152. At time t=1, signal charges R are retained at
every other electrode on the horizontal transfer path 56.
[0354] Then, at time t=2 in FIG. 54B, a drive signal .phi.HS2 at
level "H" is applied to the electrode HS2 on the horizontal
transfer path 52. With the drive signal, thus applied, there are
generated the potential level 148 and the reference level 146 below
the electrode HS2. With these potential levels applied, a packet is
generated between the electrodes HS2 and HSL. In this packet is
retained the signal charge R. To the electrodes of the horizontal
transfer paths 56 and 58 are supplied drive signals of the same
level as that at time t=1. Hence, the potential levels generated is
not changed as from time t=1. In the interim, the signal charge B
is shifted from the electrode HSL to the electrode HP2 on the
horizontal transfer path 58, as shown in FIG. 58B.
[0355] Then, at time t=3 in FIG. 54C, there is generated the drive
signal .phi.HS2 at level "L" at the electrode HS2. By this drive
signal applied, the potential level is as shown at time t=1. By
this potential level, the signal charge R retained in the packet
generated at the electrode HS2 at time t=2 is moved to the
reference level 146 of the branching section 54, as shown in FIG.
59A. At this time, the level "H" drive signal .phi.HP1 is supplied
to the electrode HP1 of the horizontal transfer path 56 neighboring
to the electrode HSL. The potential level generated underneath the
electrode HP1 is a low level 152 which is lower than the reference
level 146. As a result, the signal charge R is held in both packets
of the reference level 146 and the potential level 152. At this
time, the level "L" drive signal .phi.HP2 is supplied to the
electrode HP2 on the horizontal transfer path 58. Hence, a
potential level 148 is generated at the electrode HP2, as shown in
FIG. 59B. This potential level 148 prohibits the signal charge R
from mixing into the horizontal transfer path 58. The signal charge
R of the branching section 54 is moved towards the packet generated
at the electrode HP1 on the horizontal transfer path 56, as shown
in FIG. 59A.
[0356] The level "L" drive signal .phi.HP2 is supplied to the
electrode HP2 on the horizontal transfer path 58 neighboring to the
electrode HSL, as described above. Thus, the potential level 148
and the reference level 146 are generated below the electrode HP2,
while the level "H" drive signal .phi.HP1 is applied to the
electrode HP1 neighboring to the electrode HP2. This generates a
potential level one step lower than the reference level 146 and the
deepest potential level below the electrode HP1. In addition, the
potential level 148 and the potential level of the reference level
146 are generated by the potential level "L" supplied to the
neighboring electrode HP2. As a result, the signal charge B
retained by the packet at time t=2 is moved to and retained by the
packet generated at the electrode HP1.
[0357] The signal charges R, B at the electrode HP2 retained in a
packet generated just in rear of the output parts 60, 62 of the
horizontal transfer paths 56, 58, at time t=2, are moved towards
the output side with rise in the potential level, and transferred
via electrode OG to the section FD.
[0358] Then, at time t=4, the level "H" drive signal .phi.HS2 is
supplied to the electrode HS2, so that there is generated a
potential which is the same as that at time t=2. The signal charge
B is retained in the packet generated at this time. The signal
charge Rat the branching section 54 is moved to the packet formed
directly below the electrode HP1 of the horizontal transfer path
56. The downstream side electrodes on the horizontal transfer paths
56 and 58 are supplied with drive signals of the same level as that
at time t=3. Hence, the potential levels generated are the same as
those at time t=3.
[0359] Then, at time t=5, the same potential as that at time t=1 is
generated at the impurity layer in register with the electrode HS2.
This generates a potential level 148 directly below the HP1
neighboring to the branching section 54. The potential level 148
thus generated proves a potential barrier against the signal charge
B. This signal charge B may be prohibited from mixing into the
horizontal transfer path 56. The branching section 54 shifts the
signal charge 148 transferred thereto further to the horizontal
transfer path 58, as the branching section generates a packet. The
drive signal of the same level as that at time t=1 is supplied to
the horizontal transfer path 56. Hence, the potential level
generated is the same as that at time t=1. At time t=5, the signal
charges R, B supplied to the section FD are converted into analog
voltage signals, which are then output to the output amplifier
60.
[0360] The transfer on the horizontal transfer path 58 will now be
described. On the horizontal transfer path 58, there are formed a
plural number of impurity layers, directly underneath the
respective electrodes on a P-type substrate, in the same way as on
the horizontal transfer path 56. The impurity layers are obtained
on partitioning a sole impurity layer in keeping with the sizes of
the electrodes of polycrystalline silicon. Each of the impurity
layer has the impurity concentration adjusted, for generating
preset potential levels as later described, in keeping with the
voltage levels of the drive signals applied. The horizontal
transfer path 58 is featured by having one more electrode than the
number of the electrodes on the horizontal transfer path 56.
[0361] At time t=1, the level "H" drive signal .phi.HP2, the drive
signal .phi.HSL of the low bias voltage and the level "L" drive
signal .phi.HP1 are supplied to each of the electrodes on the
horizontal transfer path 58 as shown in FIG. 54A. With these drive
signal applied, the signal charge B is retained at the branching
section 54. With the drive signal .phi.HP2 applied, the potential
level at the impurity layer of the electrode HP2 on the horizontal
transfer path 58 neighboring to the electrode HSL is a level 152
one step lower than the reference level 146. The potential level
148 generated below the electrode HP1 of the horizontal transfer
path 56 operates as a potential barrier and prohibits mixing of the
signal charge B.
[0362] Since the level "L" drive signals .phi.HP1 is supplied to
the electrode HP1, there are generated the potential level 148 and
the potential level of the reference level 146 directly below the
electrode HP1. The level "H" drive signals .phi.HP2 is supplied to
the electrode HP2. This generates a potential level one step lower
than the reference level 146 and a potential level of the lowest
level directly below the electrode HP2.
[0363] At time t=1, when the drive signals are generated as
described above, there are generated packets directly below the
respective electrodes HP2. The signal charges B are retained in
these packets.
[0364] Then, at time t=2, the level "H" drive signal .phi.HS2 is
applied to the electrode HS2, as shown in FIG. 54B. With the drive
signal applied, there is generated a potential level in the
impurity layer below the electrode HS2, as shown in FIGS. 58A and
58B, to generate a packet. In this packet is retained the signal
charge R. The drive signals of the same level as that at time t=1
are supplied to the further downstream side electrodes on the
horizontal transfer path 58. Hence, the potential levels generated
are the same as those at time t=1.
[0365] FIG. 59B shows the potential at time t=3. The level "L"
drive signal .phi.HS2 is applied at this time t=3 to the electrode
HS2. With the drive signal applied, the potential level is as at
time t=1. The signal charge R retained by a packet directly
underneath the electrode HS2 at time t=2 is moved to the branching
section 54 which is at reference level 146. At this time, the level
"L" drive signal .phi.HP2 is supplied to the electrode HP2 on the
horizontal transfer path 58 neighboring to the electrode HSL. The
potential level of the impurity layer underneath the electrode HP2
becomes a potential level 148 which is higher than the reference
level 146. That is, a potential barrier is generated. With the
barrier produced, the signal charge R is not mixed into the
horizontal transfer path 58. On the other hand, a potential level
152 is generated by the level "H" supplied to the electrode HP1 on
the horizontal transfer path 56. This shifts the signal charge R as
indicated by an arrow 162. Directly below the electrode HP1 on the
horizontal transfer path 56 supplied with the drive signal
.phi.HP1, there is generated a packet by the potential level 152 as
shown in FIG. 59A.
[0366] On the horizontal transfer path 58, there is generated a
packet below the electrode HP1, at time t=3, by the level "H"
supplied to the drive signal .phi.HP1. The signal charge B is
retained by the packet at the electrode HP1. With rise in the
potential, the signal charge B retained by the packet, generated at
time t2, is moved towards the output side and transferred towards
the section FD via electrode OG.
[0367] Then, at time t=4, the same potential as that at t=2 is
generated directly underneath the electrode HS2. The signal charge
B is retained in a packet generated at this time. The drive signals
of the same level as that at time t=3 are supplied to further
downstream side electrodes on the horizontal transfer path 58.
Thus, the potential levels generated are of the same level as those
at time t=3. The potential level generated directly underneath the
electrode HP2 neighboring to the electrode HSL is in the state of
the potential level 148 which is higher than the reference level
146. The potential level generated directly underneath the
electrode HP1 neighboring to the electrode HSL is in the state of
the potential level 152 which is lower than the reference level
146.
[0368] Then, at time t=5, the same potential as that at time t=1 is
generated. The signal charge B is retained in a packet of the
electrode HP2. The horizontal transfer shown in FIG. 54 is achieved
as described above. The horizontal transfer operation will now be
described with reference to FIGS. 54A through 54E and further with
reference to FIGS. 57A through 58B.
[0369] In the horizontal transfer, the signal charges supplied at
time t=1 from the horizontal transfer path 50 to the branching
section 54, such as R, B, R, B, are distributed by the branching
section 54 to the horizontal transfer paths 56 or 58. As may be
seen from FIGS. 54A through 54E, the signal charges are retained at
every other transfer element. Only the signal charge R is
transferred on the horizontal transfer path 56, responsive to the
drive signal supplied. On the other hand, only the signal charge B
is transferred on the horizontal transfer path 58, responsive to
the drive signal supplied. Since the potential barrier wall is
generated at this time point at the electrode HP1 on the horizontal
transfer path 56 neighboring to the branching section 54, the
signal charge B is prevented from mixing into the horizontal
transfer path 56.
[0370] The horizontal transfer path 50 is run at twice the
frequency that of the horizontal transfer paths 56 and 58. Thus, at
time t=2, the signal charges held on the horizontal transfer path
50 are horizontally transferred by one transfer element towards the
branching section 54, responsive to the drive signals supplied. On
the horizontal transfer path 56, no signal charge transfer occurs
because the drive signals supplied are not changed in level. On the
horizontal transfer path 58, no signal charge transfer occurs as
well because the drive signals supplied are not changed in level.
However, the signal charge B on the branching section 54 is moved
to the packet generated below the electrode HP2 because the
potential level, lower than the reference level 146 at the
branching section 54, has been formed below the electrode HP2.
[0371] At time t=3, the signal charges held on the horizontal
transfer path 50 are horizontally transferred by one transfer
element towards the branching section 54. The signal charges R are
retained in the packets formed on the branching section 54 and
directly underneath the electrodes HP1 on the horizontal transfer
path 56 neighboring to the branching section 54. At this time
point, a potential barrier is generated at the electrode HP2 on the
horizontal transfer path 58 neighboring to the branching section
54. Hence, the signal charge R is prohibited from mixing into the
horizontal transfer path 58. The signal charges retained on the
horizontal transfer paths 56 and 58, are horizontally transferred
towards the output sections 60, 62, each by one transfer element,
depending on the levels of the drive signals supplied. This sends
the signal charges R and B to the output amplifiers 60 and 62 on
the horizontal transfer paths 56 and 58.
[0372] Then, at time t=4, the signal charges retained on the
horizontal transfer path 50 are horizontally transferred, each by
one transfer element, towards the branching section 54, depending
on the drive signals supplied. The signal charge R is moved to the
packet generated directly below the electrode HP1 on the horizontal
transfer path 56 neighboring to the branching section 54.
[0373] Then, at time t=4, the signal charges, held on the
horizontal transfer path 50, are horizontally transferred, each by
one transfer element, towards the branching section 54, depending
on the drive signals supplied. The signal charge R is moved to the
packet, generated directly underneath the electrode HP1 of the
horizontal transfer path 56 neighboring to the branching section
54.
[0374] At time t=5, the signal charges held on the horizontal
transfer paths 50, 56 and 58 are horizontally transferred towards
the output side, each by one transfer element. The signal charges
of the colors R and B are then transformed simultaneously into
corresponding analog voltage signals, so as to be output as output
signals OS1 and OS2 from the output amplifiers 60 and 62,
respectively. These output signals OS1 and OS2 are processed by
completely parallel processing. This eliminates differential signal
intensities attributable to processing in the time domain of the
output signals OS1 and OS2. Meanwhile, in case the differential
signal intensities attributable to processing in the time domain
are tolerable, the output signals OS1 and OS2 may be output
alternately.
[0375] By the above processing, the signal charges may be
transferred and output without mixing. In general, in keeping with
increasing numbers of pixels, it is required of a solid state
imaging device to read out the signal charges at a high speed. In
order to meet this demand, it is necessary to raise the frequency
range of the output amplifiers on the horizontal transfer paths.
The solid state imaging device is difficult to drive at a frequency
higher than a preset frequency. This is due for example to shortage
in the frequency band of the output amplifier. With the device 44
of the instant alternative embodiment, it is possible to read out
output signal charges, from color to color, without the frequency
of the output amplifier increased, by bifurcating an output and
increasing the number of output channels, even though the driving
frequency of the horizontal transfer path 50 is raised to cope with
the increasing number of pixels. That is, an improved signal charge
readout speed may be achieved.
[0376] The drive signals on the horizontal transfer paths 50, 54,
56 and 58 for transferring signal charges of the same color G, G,
G, G, . . . on the lines 252 and 254 will now be described. For
signal charges of the same color only the horizontal transfer path
56 out of the horizontal transfer paths 56 and 58 is used in the
present alternative embodiment. In this case, the drive signals
.phi.HS1 and .phi.HS2 shown in FIG. 56, parts (A) and (B), are of
the same frequency as the drive signals .phi.HP1 and .phi.HP2 shown
in parts (C) and (D). That is, the drive signals .phi.HS1 and
.phi.HS2 are lower in speed than in FIG. 56.
[0377] At time t=1, the drive signal .phi.HS2 supplied to the last
stage HS2 on the horizontal transfer path 50 is "L", while the
drive signal .phi.HP1 supplied to the last stage HP1 is "H", and
drive signal .phi.HP2 supplied to the final stage HP2 is "L".
Hence, the signal charge is transferred via branching section (HSL)
54 to the electrode HP1, that is, to the horizontal transfer path
56.
[0378] The potential levels generated on the horizontal transfer
paths 50, 54, 56 and 58, when the above drive signals are applied
thereto, are shown in FIGS. 63A through 64B. FIGS. 63A through 64B
show the structure shown in FIG. 52. A simplified form of FIG. 52
is also shown for indicating the potential positions. FIGS. 63A and
63B are for time t=1 of FIG. 62, and FIGS. 64A and 64B are for time
t=2 of FIG. 62. FIGS. 63A and 64A stand for the horizontal transfer
paths 50, 54 and 56, and FIGS. 63B and 64B stand for the horizontal
transfer paths 50, 54 and 58.
[0379] At time t=2, the drive signals .phi.HS2 supplied to the last
electrode HS2 of the horizontal transfer path 50 is at level "H".
The drive signals .phi.HP1 supplied to the electrode HP1 is at
level "L" and the drive signal .phi.HP2 supplied to the electrode
HP2 is at level "H". Thus, the signal charge is at the transfer
element HS2. The signal charge is also transferred from the
transfer element HP1 to the transfer element HP2.
[0380] The state at time t=3 is the same as that at time t=1, while
the state at time t=4 is the same as that at time t=2. The reason
the states at time t=1 and those at time t=2 are repeated is that
the drive signals .phi.HS1, .phi.HS2, .phi.HP1 and .phi.HP2 are of
the same frequency and the phase adjustment has been made so that
only the two states will be generated.
[0381] As for the output parts, the reset signal .phi.RS shown in
FIG. 62, part (E), is applied, while the output signals OS1, OS2
shown in FIG. 62, parts (F) and (G), are output. As for the
transfer of the signal charges G, the device 44 is in one-line
outputting state for outputting only the output signal OS1, while
the output signal OG2 is not used. Hence, the power supply for the
output amplifier 62 may be turned off. In this case, the power
supply by the amplifier power supply 16 is controlled by the
control signal 86 from the power supply control 122 to turn off the
power supply 66.
[0382] It is also possible to reverse the phase of each of the
drive signals .phi.HP1, .phi.HP2, .phi.HS1 and .phi.HS2, in order
to run only the output amplifier 62. By so doing, only the
horizontal transfer path 58 may be run by way of one-line
outputting. Thus, in the instant alternative embodiment, it is
readily possible to switch between one-line outputting and two-line
outputting or to freely select the output amplifiers.
[0383] If the sequence of R and B pixels in the first line and that
in the second line are compared to each other, the sequence on the
first line 201 is R, B, R, B, while that on the second line 252 is
B, R, B, R, as shown in FIG. 53. Thus, the first pixels of the
lines differ from each other. If the same horizontal transfer drive
signals are used for the first and second lines, the pixels output
from the output amplifier 60 differ from line to line. The same may
be said of the output amplifier 62. The output amplifiers 60 and 62
slightly differ from each other in characteristics, such as gain,
so that step differences are produced at the output stages for the
same color. For avoiding this, the same color is desirably output
from the same output amplifier. The method for implementing this
will now be described.
[0384] The horizontal transfer drive signals for horizontal
transfer on the lines 201, 203 of the first and second lines are
shown in FIGS. 65 and 66, respectively. Initially, the horizontal
transfer on the line 201 will be described with reference to FIG.
65. From the line 201, the signal charges are transferred through
the vertical transfer paths to the horizontal transfer path 50,
beginning from the line 201, in the sequence of the dummy pixels D1
and D2, optically black pixels OB1 and OB2, R, B, . . . .
[0385] The drive signals .phi.HS1 and .phi.HS2 shown in FIGS. 65
and 66 are supplied to transfer the signal charges held on the
horizontal transfer path 50, to the horizontal transfer paths 56
and 58. To these horizontal transfer paths 56 and 58, the drive
signals .phi.HP1 and .phi.HP2 shown in FIG. 62, parts (C) and (D),
are supplied. The transfer start position is such a position where
the potential of the drive signal is changed for the first time
from the lapse of the horizontal blanking period 260. In
outputting, the reset signal .phi.RS shown in FIG. 62, part (E), is
supplied directly before the signal charge is supplied to the
output stage, and the output signal is subsequently output in the
output period 182.
[0386] The output stage converts signal charges into an analog
voltage signal, and outputs the dummy D2, optical black OB2 and R,
R . . . , in each output time domain 182 as output signal OS1 of
FIG. 62, part (F). The output stage also outputs the dummy D1,
optical black OB1 and B, B, . . . , in each output time domain 182
as output signal OS2 of FIG. 62, part (G). That is, the output
signal OS1 on the horizontal transfer path 56 outputs the color R,
while the output signal OS2 on the horizontal transfer path 58
outputs the color B.
[0387] On the other hand, in horizontally transferring the second
line 203, outputting is switched to the first line 201. Hence, the
drive signals .phi.HS1 and .phi.HS2 shown in FIG. 66, parts (A) and
(B), respectively, are used. The other signals are the same as
those of the first line. On the second line, as compared to the
first line, the start positions of the drive signals .phi.HS1 and
.phi.HS2 are faster by one period of these drive signals.
[0388] Hence, the output stage outputs the dummy D1, the optical
black signal OB1 and R, R in each output time domain 182, as output
signal OS1 of FIG. 66, part (F). The output stage also outputs the
dummy D2, the optical black signal OB2 and B, B . . . , in each
output time domain 182, as output signal OS2 of FIG. 66, part (G).
In this manner, for the first and second lines, the same colors may
be output from the same output stages.
[0389] In the present alternative embodiment, the technique of
phase-shifting the drive signal .phi.HS is used. The present
invention is not limited to this and outputting may be to the
horizontal transfer path 50 after re-arraying employing a line
memory. FIGS. 65 and 66 show a case where the drive signals
.phi.HS1 and .phi.HS2 are fast. However, the same technique may be
used in case the speed of the drive signals .phi.HS1 and .phi.HS2
is low. In either case, the output signals may be changed over
extremely readily.
[0390] In FIGS. 55A through 55E, the signal charges G are
transferred only on the horizontal transfer path 50. Alternatively,
the signal charges G may also be transferred using the horizontal
transfer paths 56 and 58 as shown in FIGS. 67A through 67E. For the
horizontal drive signals, it is sufficient to use the signals shown
in FIG. 56. In this case, high speed transfer becomes possible with
the signal charges G. However, since the same color is output from
the different output stages, step differences for the same color,
attributable to the output stages, may be produced.
[0391] As techniques for coping with the step difference, the
following technique, for example, may be used. Data for correcting
the difference in characteristics among plural output stages, such
as gain difference, are acquired prior to shipment of the camera 10
from a plant. As the method for acquiring the data, a light source
of a preset light volume is shot by the camera, and measurement is
made of the output value of the signal charges G of the output
stages 60, 62. Coefficient data for providing for the same output
values of the output stages 60 and 62 are calculated and held in a
non-volatile memory in the camera 10. In post-shipment correction,
the signal processor 26 reads out the coefficient data stored in
the memory, and multiplies the coefficient data by output values of
the output stages 60 and 62 to correct the gain difference. This
provides for the same output values of the signal charges G.
[0392] The entire disclosure of Japanese patent application Nos.
2006-094407, 2006-094567, 2006-095198 and 2006-095374, all filed on
Mar. 30, 2006, including the specifications, claims, accompanying
drawings and abstracts of the disclosure is incorporated herein by
reference in the entirety thereof.
[0393] While the present invention has been described with
reference to the particular illustrative embodiments, it is not to
be restricted by the embodiments. It is to be appreciated that
those skilled in the art can change or modify the embodiments
without departing from the scope and spirit of the present
invention.
* * * * *