JPS585085A - Dark current correcting circuit of solid-state image pickup device - Google Patents

Abstract

PURPOSE:To suppress variation in dark current by equalizing a dark current component of a horizontal signal line to those of a photodiode and a vertical signal line, and subtracting the dark current component from a signal component. CONSTITUTION:The moment a switch 53 is turned on, a dark current component stored in a horizontal signal line is read out in a spike shape. The output signal of a preamplifier 54 is applied to a clamping circuit 55 for dark current detection and a clamping circuit 56 for signal detection. The output of the clamping circuit 55 is inputted through a sample holding circuit 57 to a differential amplifier 58, and a reference level is inputted to the differential amplifier 58. The output of the amplifier 58 is applied to the clamping circuit 56 to clamp an image pickup signal.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a solid-state imaging device, and more particularly to a rounded signal processing circuit for removing unnecessary dark current components generated by a solid-state imaging device.

For example, as shown in FIG. 1, the solid-state imaging device includes a horizontal scanning circuit (shift register) 1 for 1X position selection, a vertical scanning circuit (shift register) 2 for Y position selection. A photosensitive section 3 in which a photodiode 4 and a transistor 5 (MO8) are arranged in a play shape. and a horizontal switch 8 for sequentially reading out the signals accumulated on the vertical signal line 7. The MO8) transistor (hereinafter referred to as MO8Tr) 5 constitutes a vertical switch that is opened and closed by the vertical scanning pulse output to the signal line 6, and the drains of the vertical switches arranged in the Y direction are shared by the vertical signal line 7. A horizontal switch MO8Ti8 is connected and opened and closed by horizontal scanning pulses as described above.
is connected to output signal #9 via. Therefore, the signals from the two-dimensionally arranged food diodes 4 are transmitted to the output terminal 10 by sequentially opening and closing the switches 5.8 using pulses output from the two horizontal and vertical scanning circuits 1.2.
are read out one after another. FIG. 2 shows the signal waveform read out as described above. FIG. 2(a) shows the vertical scanning period, and 21 and 22 in the figure are vertical retrace periods. Further, 23 indicates a horizontal scanning period, and 24 indicates a horizontal retrace period.

FIG. 2 0) is an enlarged display of one horizontal period.

The nine period indicated by 25 in φ) is the period of Eirin signal, and 24
The period is the horizontal retrace period described above. As is well known, a video signal has a DC component, which must be reproduced correctly. Usually, a pulse clamp circuit is used to regenerate this DC component. Since the pulse clamp circuit is a widely used circuit, a detailed explanation thereof will be omitted. A pulse clamp circuit normally performs direct current regeneration by fixing the horizontal retrace period to a constant direct current level. However, the output of the image sensor includes an unnecessary dark current component 27 in addition to the true video signal component 26. There is no problem if this dark current component is always constant, but the dark current component usually has the property of changing depending on the temperature. The above means that the DC component varies isometrically due to temperature changes. Since the Eirin signal must be subjected to Kr processing and blanking processing (so-called process circuit) after DC reproduction, fluctuations in the DC component are undesirable.

Particularly in color cameras, fluctuations in the DC component are a big problem because they disrupt the color balance and significantly impair color reproducibility. The present invention provides a circuit system that suppresses the temperature fluctuation of the DC component due to this dark current. The present invention will be described in detail below, but prior to the explanation, the cause of dark current generation will be described. In the MO8 image sensor shown in FIG. 1, there are mainly three parts where dark current occurs. The first is the photodiode 4, the second is the vertical signal line 7, and the third is the horizontal signal line 9. FIG. 3 schematically shows, in cross-sectional view, the structure of the elements that pass through until one pixel is read out. The photodiode 4 is a semiconductor substrate 31
(taken as P type in this example) and an inverted ONW layer, which also serves as the source of the vertical switch M O8T 5. Vertical M O'8Tm5's drain/32 is N like the source! ! ! horizontal switch MO
It is connected to the source 33 of 8Tm8 via the vertical signal line 7. Source 33 of horizontal switch MO8T18. It goes without saying that the drain 34 is also made of an N-type diffusion layer. Drain 34 is all horizontal switch MOS
Tm portions are grouped together by a signal output line 9 and connected to a signal output terminal 10. It goes without saying that the drains of transistors for pixels in the vertical direction are connected to the vertical signal line 7 in parallel.

These Nll diffusion layers and the substrate constitute a PN junction, which is always used with a reverse bias. It is known that dark current is generated in a PN junction due to temperature changes. When silicon is used as a substrate, the dark current is said to increase by about one order of magnitude when the temperature rises by 30° C., although it depends on the manufacturing process of the device. An increase of 60°C will increase the temperature by about 100 times, which is a big problem.

As described above, it is unnecessary to explain that the dark current of this PN junction becomes the dark current in the video signal. Since the amount of dark current is proportional to time, photodiode 1
amount of dark current (accumulation time is usually 1/60 to 1/30 seconds)
and a vertical signal line 7 (where many drain diffusion layers are grouped together).
The current situation is that the amount of dark current generated is approximately the same as the amount of dark current with an accumulation time of 6 and 5 μs, and both cannot be ignored.

Next, a specific method for detecting this amount of dark current and suppressing fluctuations in the signal due to changes in dark current will be described.

Dark current Idp generated from the photodiode, I!
The dark current Idv generated from the direct signal line and the dark current Idwt generated from the horizontal signal line have an accumulation time of Idp of 1&7%5.
Idv is 6&4-8. Idw depends on the number of horizontal pixels, but if the number of horizontal pixels is 384, it will be approximately 14 On,S. Although the storage time is different, the temperature dependence of the dark current is the same. Generally, as shown in FIG. 4, a method of suppressing these changes in dark current is to provide a mask 42 that does not optically pass light through a portion of the element light-receiving surface 41, and clamp this position to a constant potential. Methods are being used to avoid the effects of current.

However, in this method, since the optically shielded portion 42 is created, an extra number of pixels are required.

In solid-state imaging devices, increasing the chip area is not preferable because it impairs yield in the device manufacturing process. Therefore, the present invention provides a method for suppressing dark current fluctuations without increasing the number of pixels of an element.

Next, a detailed explanation will be given using a specific example.

An embodiment of the present invention will be described using FIGS. 5 and 6.

The present invention utilizes the property that the dark current of the horizontal output line, the dark current of the vertical line, and the dark current of the photodiode have the same temperature dependence, although their magnitudes differ depending on the storage time. The gist of the present invention is to detect dark current in horizontal signal lines and suppress all other dark currents. The MO8W image sensor 51 is driven by being supplied with horizontal and vertical drive pulses from a synchronizing signal generator 52. At this time, as described above, since the accumulation time of the dark current in the horizontal signal line is short, it is difficult to detect the amount of dark current during the normal signal read period. Therefore, in the present invention, detection is performed during the vertical retrace period. Generally, a vertical retrace period of 21H (H: stands for horizontal scanning period) is provided, and if this period is used as an accumulation period, dark current can be detected with high accuracy.

However, in the image pickup device having the structure shown in FIG. 1, since the preamplifier is always connected to the signal output terminal 10, the dark current is always read out and no storage operation is performed. In the present invention, as shown in FIG. 5(b)K, a switch 53 is provided to separate the horizontal signal line 9 and the preamplifier 54 during the vertical retrace period.
will be established. This switch 53 may be provided on the chip of the image sensor 51, or may be provided externally as a discrete component. FIG. 6 shows a timing chart of a pulse group for driving the image sensor 51. In FIG. As shown in FIG. 6(b), the switch 53 is closed K11M so as to be OFF during most of the vertical blanking period, and becomes KON before the end of the vertical retrace period. The horizontal shift register operates with a start pulse shown in FIG. 6(C). That is, while the switch 53 is OFF, the operation is naturally stopped.

If the above is done, the preamplifier output will be as shown in Figure 6(d).
An output waveform as shown in is obtained (the polarity of the signal is assumed to be positive for convenience of explanation). That is, the dark current amount 61 accumulated in the horizontal signal line is read out in the form of a spike at the moment the switch 53 is turned on. The level 62 after the dark current amount 61 is read out is determined by the very short accumulation time (140
A dark current corresponding to yb8) appears, but this is sufficiently smaller than the vertical retrace period (21H) and can serve as a reference level for a state in which there is no dark current. Next, a start pulse of the horizontal shift register is input and a signal 63 is output.

In FIG. 5, the output signals of the preamplifier 54 are divided into constant circuits, and one of them is applied to a clamp circuit 55 for dark current detection, where it is regenerated as a direct current. The other output is applied to a signal clamp circuit 56. The clamp pulse is shown in Figure 6 (
Shown on the chain.

The DC signal fixed to a constant potential by the dark current detection clamp circuit 55 is applied to two sample and hold circuits 57. One sample-hold circuit samples and holds the dark current amount 61 using the pulse shown in Fig. 6 (e>), and the remaining sample-hold circuit samples and holds the reference level 62 using the pulse shown in Fig. 6 (f). The sample-and-hold output of the quantity is applied to the negative input terminal of the next-stage differential amplifier 58, and the sample-and-hold output of the reference level is applied to the positive input terminal of the differential amplifier 58. That is, the output of the differential amplifier 58 has a dark current. A direct current change opposite to the change in is obtained.Here, we will discuss the amplification degree of the differential amplifier 5B.Now, let the amount of dark current of the photodiode per unit time be Id.
p, the amount of dark current per unit time of the vertical signal line is Idv, the amount of dark current per unit time of the horizontal signal line is Idw, the storage time of the photodiode is tv, and the horizontal scanning period is 111. The accumulation time of the horizontal signal line for dark current detection is ta (approximately 20H)
Then, the amount of dark current Id during the video period of the output signal is Id=IdpXtv+IdvX
−(1). Also, the dark current of the horizontal signal line is valuable.
' is dl=dWXt4 00.
(2) becomes. Id p/ Id w (=α), Id
Ma/Idw (=β) can be easily determined by experiment. Rewriting engineering d using α and β, engineering d = cc engineering dwXty + β engineering d, xtw
``(3).Furthermore, if d is written using Id', then the amplification factor of the differential amplifier 5-8 is the same as the dark current, and an inverse change is obtained in the output. If this DC output is used as the clamp potential of the signal clamp circuit 56, dark current changes in the signal are canceled out, and an output 59 in which the true signal component is always fixed at a constant DC level is obtained as the output of the clamp circuit 56. It will be done.

By doing as described above, dark current correction can be performed without increasing the number of pixels of the element, and K can greatly contribute to not only improvement of image quality but also the manufacturing yield of the element.

[Brief explanation of drawings]

Figure 1 shows MO8m! A block diagram of the solid-state sensor, Fig. 2 is an explanatory diagram of the output signal from the MO811 solid-state sensor, Fig. 3 is the cross-sectional structure of the MO8 solid-state sensor, and Fig. 1 is the conventional black clamp dark. An explanatory diagram of the current correction method, Fig. 5 shows a block diagram of an embodiment of the present invention, and Fig. 6 shows the pulses necessary for Fig. 5.

Claims (1)

[Claims] 1. In a solid-state imaging device in which photosensitive elements are arranged in a play pattern in the horizontal and vertical directions and a video signal is obtained by sequentially scanning each photosensitive element, the dark current of the horizontal signal line is A dark current of a solid-state imaging device comprising means for detecting, means for calculating to make the detected dark current equal to the dark current of a photodiode and a vertical signal line, and means for subtracting the calculation result from the signal. correction circuit. 2. The dark current detection means and the calculation means include means for providing a switch between the output end of the horizontal signal line and the preamplifier, means for closing this switch during the vertical retrace period, and turning on the switch. There is a means to sample and hold the dark current of the horizontal output line obtained at the same time, a means to sample and hold the part where no dark current occurs after reading out the dark current, and an inverse of the change in dark current from the difference between these two outputs. 2. A dark current correction circuit for a solid-state imaging device according to claim 1, further comprising means for obtaining a polarity signal.