JPS6148199A - Circulating shift register - Google Patents

Abstract

PURPOSE:To increase cumulative frequency of charge from light receiving means by installing plural skimming means in a circulating shift register and controlling a skimming quantity of charge by means of the skimming means. CONSTITUTION:The first and second charge drawing gates SKG11 and SKG13 discharge prescribed charge respectively. A charge discharging gate SKG12 discharges charge from the charge drawing gates SKG11 and SKG13 to a drain. A floating gate FG detects fluctuation of the sum of charge quantity of two charge accummulating parts and controls charge quantity discharged from a skimming means, and composes part of a charge detecting means. Thus, saturation of the charge accummulating part composing a circurating shift register is loosened, a cumulative frequency of charge from a light receiving means is increased and highly accurate focusing detection can be executed.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a circular shift register that can be used in multiple modes.
'# ff Input the generated signal charge and sequentially cumulatively add the signal charge while cyclically shifting it to obtain the cumulative signal charge for each mode and discharge a predetermined amount of charge from the cyclic shift register. By providing a skimming means, the +s:g i limit on the amount of charge accumulation can be alleviated.
This is related to circular shift registers.

In particular, the present invention is applicable to a focus detection device for a camera 2, etc.
A so-called active type focusing system detects the focus of the camera by emitting a signal light beam from the llll and detecting the position of the spot image of the signal light beam reflected from the object surface using a light receiving means consisting of a plurality of light receiving elements. The present invention relates to a circular shift register suitable for a focus detection device.

Conventionally, as light receiving means in an active type focus detection device, for example, two light receiving elements 1-1 shown in FIG.
．． 1-2.T) There are some things that are configured.

In the figure, P is a spot image formed on the light receiving means 1 after a signal light beam from the camera side is projected onto the subject side and reflected from the subject.

The figure shows what is called a differential type sensor, and the rlt shadow +
The lens is configured to move in the kx direction on the light receiving means l depending on the focus state. When the photographic lens is in focus, the spot image P is located at the center of the two light receiving elements 1 to 1) 1-2, as shown in FIG.
When out of focus, for example, the spot image P is configured to move in the X direction as shown in (-) in the same figure when the front focus is on, and as shown in the same figure ((-K) when the focus is on the rear. Now j
@1 As shown in Figures (a), (b), and (e), the spot image P is X1 + Xo+
Two light receiving elements l-1 when located at X2,
When the output signals from 1-2 are Y□ and Y2, respectively,
Spot image P Position in Ox axis direction and two light-receiving strands 1-
The output from 1゜1-2 (Fig. 2 shows the relationship between fi numbers)
IL).

It will look like (b).

In the same figure, Y and Y are IDs from the camera side respectively.
Output signal from the light receiving element 1-1.1-2 when not emitting a 2D signal beam and then not emitting light, that is, due to ambient light, Y,
This is an output signal at the time of light projection when the signal light flux IL 2L is projected from the camera side. From this, if the output signal, that is, the light emitting component of the signal beam reflected from the object by the light receiving element 1-1.1-2 is Y, Y, then Is28 YIS-YIL-YID Y28''-Y2L-Y2O.

At this time, when the difference 2 between the light emitting components Y and Y from the two light-receiving elements 1-1.1-2 is calculated, it becomes as shown in FIG. 3.

In reality, this G' component 2 is obtained and the in-focus state of the photographic lens is detected from this.

When the position X of spot 1 P is in the front bin area shown in FIG. 1(a), Z)0. When in the focused state shown in FIG. 1(b), 2-0. FIG. 1 (6) When in the rear bin state shown by K, it becomes Z(O).

In such a focusing detector setup, two light receiving elements 1-
1. From 1-2, output signals YIL and Y1 of four modes
D, Y2L, and Y2D are obtained, but what is needed for focus detection is the relative difference between these output signals. If this difference is small or inaccurate, it will cause a decrease in focus detection accuracy.

Voice [For example, if the object distance is far, the projected light component Y□8゜Y2
8 becomes smaller.

Also, output No. 18 Y, Y'5c LD when projecting light due to ambient light
The projected light component Y is removed from the 2D output signal Y and Y.

Even when obtaining IL 2L Y28, accurate light emitting components Y and Y cannot be obtained if the circumferential IUI light changes over time.
2S becomes.

To the main mountain 1) One person in total 1) SI Showo -20776
Is it possible to detect focus when the ambient light changes over time with No. 8?
? Output signals Y and Y when emitting light to prevent the stable from dropping.
and non-IL 2L light emission output signal Y, Y is used repeatedly, 1
0 2D We have proposed a method to eliminate errors caused by fluctuations in ambient light by using cumulative brightness.

At this time, if the cumulative number of output signals increases, the detection of the emitted light component will improve by 1) 1 degree, but the specific charge '
Since the charge storage section for performing the tifi-ate operation can only store a certain amount of charge, a state called saturation occurs.

In order to solve this problem, this 1) 3 applicant filed a patent application in 1982-2.
1) No. 920 with multiple charges # 8 (S circulating and 7'C
In the Kinuta m shift register, a certain amount of charge is transferred to a predetermined amount.
='? A method for alleviating the saturation t- of charge accumulation by providing a skimming means for discharging Tfi' from the junction has been proposed.

However, in the above-mentioned Japanese Patent Application No. 58-21) 920, only one Sugino means was provided, and the amount of charge that could be discharged was constant, so the accumulated charge per charge storage unit was It is difficult to increase the size of the trowel ≠ 1 (ratio.

Kimiori 1 improved the above-mentioned skimming means and created a circular shift register (1-7) to reduce the saturation of the pickled part, and to reduce the accumulation of -i charges from the light receiving means. The purpose is to increase the Siu rate and make it possible to detect highly effective collaborations.

The main characteristic of the circular shift register for achieving the purpose of the present invention is to input the signal 1' which repeatedly occurs in the algebraic mode, and cumulatively add it to each mode 4υ while circularly shifting the signal IB. , the cumulative charge signal for each mode is set to 1, and then the cycle 6 (a part of the shift register is provided with a plurality of skimming means, and a plurality of skimming means are provided in a part of the shift register, and the skimming means This is because the charge of quantification ′ is discharged.

Next, one implementation FIJ of the present invention will be explained with reference to each figure.
The figure is a schematic diagram of a circular shift register of the present invention.

In the figure, SAI to SA4 and 5ISI to SB4 are optical converters, respectively, and there are four photoelectric converters SAI-
At SA4, one light receiving section 4-1 of the light receiving means 4 is formed (1), and the other light receiving section 4-2 of the light receiving means 4 is formed by four photoelectric conversion sections SBI to SB4.

The light receiving sections 4-1, 4-2 are the light receiving elements 1-1.
It corresponds to l-2. The reason why a plurality of photoelectric conversion units are provided in the light receiving unit 4-1, 4-2 is to select and use them according to the brightness of the photographing lens, the object distance, etc.

BG is barrier gate, RAI NRA4, RBI
~RB4 are photoelectric conversion units SAI~SA4, SBI, respectively.
- A charge storage section for temporarily accumulating the charges generated in SB4, 31 (Al~51(A4, 5HBI~5I(
B4 are charge storage parts RAI to RA4, RBI-, respectively.
RB4 K accumulated next charge t-CCD register CRA
, Shift gate for transferring to CRB, MG is CCD
The charge tl from registers CRA and CRB is a charge transfer register for delaying the charge input from one COD register in order to mutually input the charge tl to the charge input register GO.

Ml-M4, G1-G4 are charge storages that form part of a circular shift register that is transferred from the COD register CRA or CRB via the charge input register Go and accumulates the charges while circulating them in the direction of the arrow. Department. For example, the charge accumulating section M1 absorbs 1 from the input register GO and adds the charges transmitted from the section G4.

FG 2 charges t7 i'+(XGl, G2
1 is a floating gate for detecting all fluctuations in the amount of workpiece and controlling the amount of charge discharged from the skimming means described above, and a part of the charge detection means is implanted therein.

lζG is a reset gate for resetting the floating gate FG, -FCl2 is the control 1 signal of IζG (in the reset state), FCDC is the reference voltage when resetting the floating gate FG, and b5 is the reset state. 7
An amplifier for detecting the bottom pressure of the 0-ting gate, O
8 is an amplifier MAMP for the width pressure of the floating gate.
This is the output signal output via the .

CLR1) is a first clear gate that discharges the charge transferred to the charge input register Go, CLR12 is a second clear gate for clearing the charge t of the charge ti part M3 of a part of the cyclic shift register, 5KGII , 5KG1
3 are first and second charge collecting gates for discharging predetermined charges, respectively; 5KG12H1U load collecting gate) 5KGII
, 5KG13'KaraO'C Charge discharge gate for discharging charges to the drain.

, 'f(, 5(n) to (d) are schematic diagrams of the principle of skimming by the skimming means according to the present invention.

The figure (,) shows the initial state in which charges are accumulated in the charge storage section G3, and the figure (b) shows the first state where charges are accumulated in the first pumping gate 5KGII.
Li pressure is applied and a predetermined amount of charge is pumped up to the gate 5KGII.
As you enter, the same figure (C) is the pumping gate 5KGI.
When I returns to the initial state and a predetermined amount of 'Lu charge is pumped out, the same figure (d) shows the point where 7 (L pressure is applied to the discharge gate 5KG12 and the charge is pumped out and fc charge is discharged to the drain. Each is shown.

FIGS. 6(a) to 6(f) are explanatory diagrams showing the principle of '+LE load control n' by the floating gate FG.

The figure (-) indicates the amount of charge A in the charge storage section Gi. * comes in and fC By the way, at this time floating game) 1"
Since Gl is in the reset state, 70-teingge)F
The output os of Gi is vr@f. In the figure (b), the reset of the floating gate is completely released, but since there is no charge in or out of the charge storage section Gi, the output O8b of the float FGi does not change. Same figure (C)
is the charge A from the floating charge storage section Gi after the reset of the floating gate is released. When transferred to the next stage, the charge AD* is discharged from the charge accumulation section Gi, so that the output O8c of the flipping gate FGi changes in potential by an amount corresponding to the charge AL)*. (D) of the same figure shows a floating gate) With the reset of FGi completely released, the E load AL* enters the charge storage part Gi.As a precaution, the output O8 of the floating gate FGi changes the potential by an amount corresponding to the load AL*. do.

As described above, the output voltage O8 increases or decreases depending on the amount of the ・K load into and out of the solder load t'A 4j1 section Gi, so the output voltage O8 changes as shown in FIG. 6 (,), for example. O Output voltage shown in the figure os, os, , as
, os.

are shown in Figure 6 (a), (b), (e), (
The output voltage OS corresponding to d) is shown.

The output voltages A/D and A/L shown in FIG. 6(e) are voltage changes corresponding to the amounts of charges AD* and AL*, respectively.

FIG. 6(f) shows two charges MRf! 15 Gl,
This is an explanatory picture showing the relationship between G2 and 70-Teing Game) FG. FGI and FG2 #'i are each floating gates) Some of the charge storage parts of FG are placed on Gl and Gz. Both floating gates are connected to each other.
It's true.

Floating gates FGI and FG2 detect changes in the charges of -th [accumulators G1 and 02, respectively, and a signal obtained by adding up the amount of change is output as O8.

FIG. 7 shows integration mode 1I to the charge storage section in the present invention.
It is a timing chart of Ii.

SH is a shift gate) By selectively applying 5HAi and 5HBi, the charges ta obtained in the first photoelectric conversion units SAi and SBi are selectively transferred to the CCD register CR.
A, 77 pulses transferred to CRB.

4, φ* is CCD shift register CItA, CRJ
It is a transfer pulse of I and also a transfer pulse of a cyclic 7-foot register. The resistor is the charge storage part Gl.

, G2 is added with 1)43 transfer pulses φt!
A transfer pulse φ* is applied to the ij a-ramming portion M2.

When the knob is moved based on the timing chart shown in FIG. 7, the output AL of the light receiving section 4-1 when emitting light1 the output BL of the light receiving sections 4 to 2 when emitting light1 the output of the light receiving section 4-1 when not emitting C output A of
. , when the light is not emitted, the output BD of the light receiving section 4-2 is 1) 44th order'
tJi 6; Manpower register GOK is transferred.

Outputs of these four modes AL, BL, AD.

BD is repeatedly stopped and added to the signal output of the same mode, which is sent to the charge storage section MIK via the loading register Go, cyclically shifted, and transferred to the charge storage section M1. The addition is not fC1g.
Each time, the load ffida product Bfl Gl r is transferred to 2 at the timing shown in FIG.

In the same figure, ν and ↓ mark indicate that electric power Ki is ton 1:? f Accumulation part Gl
The ↑ mark indicates that the charge has come out from the charge storage section Gi.

The output O8 changes as shown in FIG. 7 due to the inflow and outflow of charges.

Output AL, AD, BL, +3D are cumulative times im
K.

The amount that changes accordingly is now A-ΣAL(n)-AD(n)B-, :, B
t(n) - Bo(n), the signal A required in the differential light receiving element.

B is obtained.

As is clear from the above formula, the output AL and Ap or output 1) t
, and B. Even if the same amount of charge is completely subtracted from
does not receive any shadow IF9.

By the way, in the old method, AL (m) ≧ AD (m).

IIL(m)≧I1))(m>.The amount of skimming at one time is SK, and when AD(m) exceeds the skim level T, skim by NSK from Ap(m) and AL(m). Ma '几13D(m) exceeds the skim level Tt'' fc LABOCm) (!: BL, (m) to amount SK
Only, 1. Kim.

Saturation level of circular shift register'(i:v8fLt
Assuming that the output ranges of A'(m) and B(m) are 'W', the following conditions are required.

ADmax +Rat″′″vsat ' BDmnx
"Rw -"sat...■ Also, 0 (AD, 0 (Bo...00
From the formula, T ・SK ・V, ) 0 ・・
・It becomes ■. ■ is the remainder after taking into account the Sugimu Repel setting error, etc.

By the way, if the input signal AD1, fC, and BD which are added from the amount of skim 1 and 1 increase, AL or 9 will be monotonically added and exceed the Jn level, so the following condition is also required. AO□Maku is ADlm the maximum value of BD□
az * BDlmax, ADlmax ・
BDlmax” SK...■.

Now, there is an ADma engineer that satisfies the conditions of ■.

BDrn□ is ADrnax-”ADlmax・BDmax-”BD
lmax... ■. Therefore, if the Siskim quantity is optimally set according to formulas ■, ■, ■, ■, then SK-ADlmax -BDlmax-+(vsat
-Rv -vr)...■. Therefore, the sequence is V, at-800mV,
If RW-300mV, Vr-20mV K is used, it is best to set the skim amount to 240mV, and in this case, 801max + "Dlmax is z4o@v
becomes. If an external light component exceeding the amount of charge corresponding to this output voltage is accumulated each time, the cyclic shift register becomes saturated.

If the amount of skimming is constant as in the conventional case, the external light level that can be processed by the cyclic shift register is at most a brightness that generates a charge corresponding to +vBW.

As described above, the present invention increases the level of external light that can be processed by providing a plurality of skimming means.

The implementation shown in FIG. 4 has three operating modes: startup mode, integration mode, and readout mode.

First, in the startup mode, the first clear gate CLRII and the second clear gate CLR12'jkON are turned on, and the CCD registers CRA, CRB and circular shift register are operated an appropriate number of times, and at the same time, the shift gates 5HAi, 5HB are activated.
Open i to reset the charge.

The integration mode is carried out in the same manner as described above. In read mode, output A-B signals can be obtained.IJ control signal FGR8t
"Control.

In this implementation sequence, two skimming means are used to collect the charges) 5KGII, 5KG13, so that the same amount of charge SK is discharged. When the skim level Tlft is exceeded, the skim amount f: SK
When the level exceeds 1 skim level T2, the skim "fJ"C is set to 2 x SK.

Then, the above equations ■ and ■ are ADmax +”y −vsat ' BDmaz +
”w −vsat...■' 0 (AD + O(BD...■l
becomes the same as skim level 'rl, T2 K vs.
If the remaining amount is vr, then T2・28K ・T□・SK ・Vr) 0 ・・■
"Dlmhx ' BDlrnax is ADl
max・BDlmax・2sK...■
becomes. By the way, ADmax-max"T2-8K+AD1msx,"
1 +AD1max)... [phase] BDmax-max "2- SK 10 BD1max,"
l +BD1max)...[phase]', but since T2-SK-T□, 80max-
TI +AD1max-BDmax -TI 10BD1
max, and in combination with ■, ■', and ■2, ADl + nax - BDlmaz - 28 = 8 (vs
at -"w-vr)...■. Similarly, if we use n skim gates with the same amount of skimming, we get ADlmax-BDlmaz-nSI(=n+t"'s
at 'w-"r)...0.

When n-2, v3at-800nV, RVr-:30
0trtV.

If V-20mV, SI(-16on, V.

ADlrrl&ABD1maX-320rrL■
Therefore, the level of external light that can be processed is better than that described above.

The skim level is SK + V and 2・SK + V
All you have to do is choose.

r, r Figure 8 shows another embodiment of the present invention 21? ! This is a schematic diagram of 1), in which the same elements as in FIG. 4 are given the same reference numerals.

In this embodiment, the skim means t-2 are arranged in the load storage section.

In the same figure, SKG 21 and 5i (c z 3 are 5KG 22 with 4 loading and 9 gates, each of which is part of the skim means)
.

5KG24 is a charge discharge gate. Mako CL
R21 and CLR22 are charge clear gates, respectively, and are used to clear charges in the startup mode. Apply the skim amount of 5KG21 and 5KG22 with SKI and SK2, and make 5K2-2.
SKI.

When skimming the load in Ruru mode, 5KG21
.

Three levels of skim amount can be selected by skimming 5KG22 one cycle at a time.

In other words, if only 5KG21 is used, the skim amount will be SKI, and if only SI (G22 is used, the skim amount will be SK2-z
・sKi 95KG21 5KG22 (D If you use both, the skim amount will be SKI + SK2-3・SKI. This will have the same effect as warping three gates of the same skim rice. At this time, the skim level T , T2゜T
How to select 3 is to select 3 skim gates with the same amount of skim.
Same as when turning l-, T□-SK+V, , T2-2SK+Vr, T3-
3SK+Vr is fine.

When n-:1, V, t-soomv, R--3
00nV.

v,, -207nV, SKI -120,
V, SK2-2401rLV, AIJlmax-
BDlrrl, L knee becomes 360mV.

The upper part has two skim means and shows the 7C field sweetness, and the next one is 2
i'1). Needless to say, it is okay to set up more than 100 meters. Furthermore, the present invention is not limited to goldfish detection devices, but can be widely used in object detection, collision prevention, etc.

As described above, according to Akiko Mizumoto, a plurality of skim+stages are provided in the circular shift register, and the skimming means allows the amount of charge skimming 'Cflit! It is possible to increase the number of times of accumulation of charge from the light receiving means by controlling K; I'/
degree circular shift register can be achieved.

4 Brief description of the drawings Figure 1<8), (b), and (c) are explanations of conventional differential Enoki light receiving elements (2) and Figure 2 (L) and (b), respectively.
, FIG. 3 is an explanatory diagram of how the signal output from the differential element shown in FIG. 1)1); FIG. 421 and FIG. Figures (a)-(d), F
B 61: gl (fI,) ~ (r), FIG. 5a' ν1 diagram appears.

In the figure, 1.4 is the receiving part, 4-1.4-2 is the light receiving part, SAI = SA4, SBt -SI4 B each is the light army change part, 1tAl-1ζA4, Iζ1) 1 ~
lζ134 are each city, (lii accumulation part, 5HAI −
5IIA4, sf[131-5IIB4 each has 7 footgates, i~1)-M4, Gl~
G4 i, t ′r7r etc. 1; I; 1
) 4-cho 1; 1, ・i, 'l'i 7:B sF
C is 70-f in'f gate, 5KGII, 5KG
13.

5KG2t, CKG23 it each/board gate, 5KG12.

5KG22, F:KG24 are each discharge gate, CI
, R1).

CLR12 and CLR21 are each clear gates.

Claims (3)

[Claims] (1) A plurality of charge accumulators are configured to input signal charges that are repeatedly generated in a plurality of modes, cumulatively add the signal charges for each mode while cyclically shifting the signal charges, and obtain a cumulative charge signal for each mode. 1. A cyclic shift register comprising: a plurality of skimming means provided in a part of said cyclic shift register, and said skimming means discharging a predetermined amount of charge. (2) The cyclic shift register has a detection means for detecting the charge in at least one charge storage section, and a skim control means for controlling the amount of charge discharged from the skim means in accordance with a signal from the detection means. 2. A circular shift register according to claim 1, characterized in that: (3) The cyclic shift register according to claim 1, wherein the plurality of skimming means are arranged in the same charge storage section or different charge storage sections among the plurality of charge storage sections.