JPH0437628B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a charge transfer device used in one-dimensional image sensors and the like. In particular, it relates to the structure of an output section in which signal charges transferred by two CCD shift registers are taken out by one reading section.

[Technical Background of the Invention] Examples of one-dimensional image sensors using conventional CCD shift registers are shown in FIGS. 1, 2, and 3. FIG. 1 is an overall view, FIG. 2 is a detailed view of the output section, and FIG. 3 is a longitudinal sectional view thereof. In FIG. 1, a pixel column 2 is formed by arranging a plurality of photosensitive pixels in a straight line on a semiconductor substrate 1 of one conductivity type. Shift electrodes 3 along both sides of the pixel row 2
is provided. This shift electrode 3 is connected to the pixel column 2
It controls the transfer of the signal charges accumulated in the CCD shift register 4 to the CCD shift register 4, which will be described later. moreover,
CCD soft registers 4 are provided in parallel on both sides of the shift electrode 3, respectively. And these
Transfer electrode 4- of the final stage of CCD shift register 4
An output gate electrode 5 set at a constant potential is provided adjacent to 1 and 4'-1. A floating junction type charge readout section (hereinafter referred to as floating junction section) 6 is provided adjacent to the output gate electrode 5 for extracting signal charges sequentially and serially transferred by the CCD shift register 4. Reference numeral 7 indicates a reset electrode for periodically resetting the potential of the floating junction 6 to the same potential as the output drain 8. A potential change caused by the transfer signal charge occurring at the floating junction 6 is outputted to the outside in the form of a voltage from the output signal terminal OS via the output circuit (generally a source follower circuit) 9.

Next, as shown in FIG. 2, each transfer electrode 4-1, 4-2,...4-n,
4'-1, 4'-2, . Here, the final stage transfer electrode 4-1,
It should be noted that the arrangement is such that the signal charges read out from 4'-1 are concentrated on one side of the floating junction 6. In the figure, a broken line indicates a transfer scheme effective for transferring signal charges.

Next, FIG. 3 shows the cross-sectional structure of each of the above parts.
In FIG. 3, each transfer electrode 4-1, 4-2,...
..., 4-n, the output gate electrode 7, and the reset electrode 7 are formed on the semiconductor substrate 1 via the insulating oxide film 10, and the surface of the semiconductor substrate 1 below each electrode has a conductivity opposite to that of the semiconductor substrate 1. A shaped impurity region 11 is formed, forming a so-called buried channel type CCD shift register structure. The floating junction portion 6 and the output drain portion 8 are impurity regions having a conductivity type opposite to that of the semiconductor substrate 1. FIG. 4 shows the state of potential change of each electrode as described above.

In FIG. 4, the output drain section 8 is reverse biased to a constant potential with respect to the substrate 1, and when a high level pulse is applied to the reset electrode 7,
The floating junction 6 has the same potential as the output drain portion 8. Next, when the potential of the reset electrode 7 becomes low level "L", the floating junction 6
is in a floating state. In this state, when the final electrode 4-1 becomes low level "L", the signal charge
Q _sig flows under the output gate electrode 5 into the floating junction 6 and changes its potential by V _sig . This potential change _Vsig is taken out to the outside via the output circuit 9. This potential change V _sig is expressed by the signal charge amount Q _sig and the capacitance C _F of the floating junction 6 as V _sig =Q _sig / _CF . Therefore,
To increase the output voltage for a certain amount of signal charge (increase the charge/voltage conversion gain), use capacitance.
_CF needs to be reduced. "H" indicates high level.

Note that in order to prevent signal charges from the final electrodes 4-1, 4'-1 from mixing with each other, the final electrodes 4-1, 4'-1
Pulses having mutually opposite phases are applied to 4'-1 and read out.

[Problems with background technology]

The problems with the conventional structure of the output section described above are that high-speed driving is difficult and there is a limit to reducing the area of the floating junction.

In other words, the final transfer electrodes 4-1 and 4'-1 of the two CCD shift registers are only under the electrodes.
It is necessary to accumulate the maximum amount of charge that can be transferred by the CCD shift register, so the area of each final transfer electrode 4-1, 4'-1 is L ₁ ×W ₁ (see Figure 2).
The repetitive transfer electrode part 4 of the CCD shift register
It is necessary to make the area of n... larger than L ₀ ×W ₀ .
However, since it is necessary to apply clock pulses of opposite phases to both final transfer electrodes 4-1 and 4'-1, there is a limit to the distance L between the two, and it is not possible to make them very close to each other. The spacing L is usually designed to be several μm or more. On the other hand, as can be seen from V _sig =Q _sig / _CF shown above, the amplitude of the output signal can be increased if the area of the floating junction 6 is as small as possible.

For this reason, in the past, the channel width under the output gate electrode 5 was designed to be gradually narrower as it progressed in the transfer direction, as shown by the broken line in FIG.

However, with such a configuration, the length of the signal charge transfer path under the output gate electrode 5 becomes long, which poses a problem when driving at high speed. Furthermore, if the transfer channel width at the end of the output gate electrode on the floating junction 6 side is L ₂ (see Figure 2), then one side of the floating junction 6 must always have a length of 2L ₂ or more. This hinders the miniaturization of the floating joint portion 6.

[Purpose of the invention]

Therefore, the present invention has an output part structure that can shorten the transfer channel under the output gate electrode and reduce the area of the floating junction, thereby realizing high-speed drive and high charge-voltage conversion gain. The purpose of the present invention is to provide a charge inversion device with a built-in charge conversion device.

[Summary of the invention]

In order to achieve the above object, a charge transfer device according to the present invention includes two columns of charge transfer registers that transfer signal charges transferred from a photosensitive pixel column in the same direction, and a charge transfer register that transfers signal charges transferred from the last stage of each register alternately. In a charge transfer device having a floating junction type charge readout section for reading out, the floating junction type readout section is formed in a shape having at least two adjacent sides, and the last stage of one column of the charge transfer register is The last stage of the other column of the charge transfer registers is arranged to face one side of the two sides, and the last stage of the charge transfer register is arranged to face the other side of the two sides. The present invention is characterized in that the signal charges are transferred from a vertical direction.

[Embodiments of the invention]

An embodiment of a charge transfer device according to the present invention will be described below with reference to the drawings.

FIG. 5 shows the structure of the output section of the charge transfer device according to the present invention. In FIG. 5, parts having the same functions as those shown in FIGS. 1 to 3 are given the same names and symbols and will be described below.

As shown in FIG.
has a square shape and are arranged in a diamond shape. An inverted dogleg (or angle) output gate electrode 5 is provided adjacent to two mutually perpendicular sides on the final transfer electrode side. A final transfer electrode 4-1 and a final transfer electrode 4'-1 are arranged in parallel and adjacent to one part of the output gate electrode 5 and the other part, respectively. Therefore, the directions of signal charges flowing into the floating junction 6 from each final transfer electrode 4-1, 4'-1 are at right angles to each other, and unlike in FIG. The water flows in from two different sides of the section 6. Note that 12 is the output circuit 9
A contact hole 13 indicates an Al wiring. Final transfer electrodes 4-1, 4'-1 and repetitive transfer electrode section 4 arranged along pixel row 2
-n and 4'-n, a curved transfer path may be formed using, for example, a wedge-shaped electrode.

In this way, by arranging the final transfer electrodes 4-1 and 4'-1 so that signal charges are transferred to the floating junction 6 from two different orthogonal directions, the following effects can be obtained. can.

(1) The relative spacing between the final transfer electrodes 4-1 and 4'-1 can be increased, and even in this case, the shape of the transfer channel under the output gate electrode 5 is not restricted.

(2) If the transfer channel width L ₂ (Fig. 5) on the side of the floating junction 6 under the output gate electrode 5 is the same as that in Fig. 2, the area of the floating junction 6 is changed from that in Fig. 2. Therefore, the charge-voltage conversion gain can be increased, and a large output signal can be obtained.

(3) Generally, the shape of the transfer channel (broken line) under the output gate electrode 5 is such that the width of the transfer channel at the end on the floating junction 6 side is L2, as shown in FIGS. 1 and ₂ . , the width of the transfer channel on the opposite side is L ₁ , in order to minimize the area of the floating joint 6.
It is common to design so that L ₂ <L ₁ .
In this case, under the conventional structure, CCD
The only way to reduce the width of the transfer channel from each electrode of the shift registers 4, 4' to the floating junction 6 (broken line in Figure 2) is to reduce the width of the transfer channel at only one edge _CH1 . In case the transfer channel is a floating junction 6
Since it enters from a direction perpendicular to each other, it can be equally narrowed at both edges CH ₁ and CH ₂ (Fig. 5). As a result, the edges of the transfer channel below the output gate electrode 5 CH ₁ ,
The length of CH ₂ can be made shorter than before,
Therefore, delays in transfer time can be prevented, allowing high-speed driving.

〔Effect of the invention〕

As described above, according to the present invention, the final stage of the charge transfer register is configured such that signal charges are transferred to the floating junction type charge readout section from two different directions substantially orthogonal to each other. The length of the transfer channel at the bottom of the output gate electrode can be shortened, and two transfer channels contact one floating junction charge readout section from different directions. The area of the section can be reduced. This makes it possible to drive at high speed and obtain a high conversion gain.

[Brief explanation of drawings]

Fig. 1 is an overall view showing an example of the structure of a one-dimensional image sensor using a conventional CCD shift register, Fig. 2 is a partially enlarged view showing the detailed structure of the output section of the charge transfer device of the conventional image sensor, and Fig. 3 4 is an explanatory diagram showing the potential state of the output section, and FIG. 5 is a partially enlarged view showing an example of the output section of the charge transfer device according to the present invention. be. 1... One conductivity type semiconductor substrate, 2... Photosensitive pixel column, 4, 4'... CCD shift register, 4-1,
4-2,...,4-n...Transfer electrode, 4'-1,
4'-2,...,4'-n...transfer electrode, 5...output gate electrode, 6...floating junction type charge readout section.

Claims (1)

[Claims] 1. Two columns of charge transfer registers that transfer the signal charges transferred from the photosensitive pixel columns in the same direction, and a floating junction type charge readout that reads out the signal charges alternately from the final stage of each register. In the charge transfer device, the floating junction type readout section is formed in a shape having at least two adjacent sides, and the final stage of one column of the charge transfer registers is arranged to face one side of the two sides. The last stage of the other column of the charge transfer registers is arranged to face the other sides of the two sides, and the signal charge is transferred to each of the two sides from a direction substantially perpendicular to the sides. A charge transfer device characterized in that: 2. The charge transfer device according to claim 1, wherein the floating junction type readout section has a square shape.