Image sensor

Details Image sensor Image sensor

2001-10-25

2003-11-11

Porta, David (Department: 2878)

Radiant energy

Photocells; circuits and apparatus

Photocell controlled circuit

C348S294000, C348S308000

Reexamination Certificate

active

06646249

ABSTRACT:

This invention relates to an image sensor, and particularly to a pixel design for such an image sensor. Particularly, but not exclusively, the invention concerns image sensor arrays based on thin film transistor (TFT)-photodiode technology.
Optical image sensors typically comprise a pixel array arranged in rows and columns, with row driving circuitry and column reading circuitry being used to address the array of pixels. Typically, the row and column circuits are provided on a separate substrate to the pixel array, so that interconnections must be provided between each row and the row driver circuit and between each column and the column reading circuit. It is known to introduce multiplexer circuits onto the pixel substrate to enable the number of interconnections to be reduced.
Sensors employing TFT-photodiode pixel circuits have been known for some time, and the driving force behind their development has been, and continues to be, their use in medical diagnostic imaging applications. More recently, interest in optical-based fingerprint sensors has increased. Work has initially been based on diode-diode sensor arrays, but now attention is turning to TFT/photodiode technology because of the lower power consumption, faster read-out and higher multiplexer ratios that can be achieved.
FIG. 1
shows a typical structure of a column multiplexer circuit
10
connected to a TFT/photodiode image sensor array
20
. Only a single row in the array
20
is illustrated for simplicity, associated with row conductor
22
. Each pixel in the row comprises a TFT
24
and a photodiode
26
connected in series between a common potential
28
and a respective column conductor
30
. A signal on the row conductor
22
turns on the TFTs
24
of each pixel in the row, which allows the photocurrent produced in the photodiode to flow to the respective column conductor
30
to be read by a charge sensitive amplifier arrangement
40
.
A multiplexer switch
31
in the form of a TFT is connected between each column conductor
30
and the amplifier arrangement
40
. The switches
31
are arranged in groups, with each switch
31
in a group being independently controlled by control lines A, B, C, D. These control lines A to D define four multiplexer channels A to D. Each group is provided with an associated charge measurement device
40
. However, different groups share the control lines. The arrangement shown provides a 4:1 multiplexing function, and requires four additional control lines A to D.
After the array has been exposed to light, signal charges are stored on the capacitances of the photodiodes. At this point, the array can be read out, and this is done by addressing each row in turn by applying a positive pulse to the appropriate row conductor. In an array without a column multiplexer, each column is connected to its own charge-sensitive amplifier, and when the pixel TFTs
24
are turned on, the signal charge from each pixel flows down the column
30
to the respective charge-sensitive amplifier.
However, in an array with a column multiplexer, the situation is more complicated. Consider the situation when the columns connected to multiplexer channel A are to be read out. This is arranged by turning on the multiplexer TFTs connected to control line A, and ensuring that all other multiplexer TFTs are off. When the row pulse is applied, signal charge from the pixels in the columns associated with multiplexer channel A will flow down the column via the multiplexer switch
31
to the respective charge-sensitive amplifier
40
. However, at the same time, the other pixel TFTs
24
will also have been turned on, and signal charge from the photodiodes in those pixels will be transferred to the column capacitance. Hence, the act of reading multiplexer channel A has caused the signal charge from the other pixels to have been lost to the column capacitance.
If a static image is being recorded, for example a fingerprint, the lost charge can be re-created by using multiple exposures. In such a scheme, multiplexer switches A would be turned on, and all rows in the array would be addressed in sequence, thereby reading out charge from all of the pixels connected to columns associated with multiplexer channel A. When this is complete, the array is re-exposed and column multiplexer switches
31
for channel B are turned on. The rows are once again addressed in sequence so that pixel charges from the pixels associated with channel B can be read out. This is repeated for multiplexer channels C and D. While this provides a solution to the charge loss, it suffers from two disadvantages. Firstly, the multiple exposure and read-out lead to a longer image acquisition process which, especially for the fingerprint sensor, is undesirable. The use of multiple exposures is also not appropriate for non-static images, for example in the field of medical diagnostic imaging. Secondly, the data from the array emerges in a column-based sequence. In principle this need not be an issue, but in practice could require the development of bespoke image acquisition and processing software.
To avoid the need for multiple imaging, it is possible for the ‘lost’ pixel charge to be recovered by transferring the charge from the column capacitances. When reading out multiplexer channels B, C and D by turning on the respective multiplexer switches
31
, the column capacitance could be connected to the charge-sensitive amplifier. A number of timing schemes can be devised for such a read-out scheme, but all of them suffer the drawback that the charge is stored on the column capacitance for a period. Further, this period is not the same for all multiplexer channels. The main concern with this type of read-out scheme is the effect of leakage currents from all of the pixels in each column.
A further alternative is to use a complete analogue line store (array of sample and hold circuits) as part of the column multiplexer circuit. With such a circuit in place, the signal charge from the pixels in a given row can be transferred to the line store when the row is addressed. Once this is complete, the charge can be transferred to the charge-sensitive amplifiers via the column multiplexer switches with a timing scheme that is most appropriate for the application. Again, there are several possible implementations, but these basically divide into two types. The first uses a simple switch/capacitor as the sample-and-hold (S/H) circuit, and the second employs a high-gain buffer amplifier as part of the S/H circuit. Both variants have drawbacks: in the first there is charge-sharing between the photodiode capacitor and the column parasitic capacitance; in the second, aside from the added complexity, the issue is the difficulty of implementing a high gain buffer using the same device technology as the pixel TFTs, for example n-channel amorphous silicon technology.
There is a need for an alternative approach which allows column multiplexing to be used with a single exposure image sensor and which is simple to implement. The multiplexing circuitry also needs to be readily implementable using the same technology as the devices of the image sensor pixel, for example n-channel amorphous silicon devices.
U.S. Pat. No. 5,134,489 discloses an image sensor comprising rows and columns of image sensing pixels, each row of pixels being associated with a respective row conductor, and each column of pixels being associated with a respective column conductor, each pixel comprising an image sensing element and a switching device, the switching device enabling a signal of the image sensing element to be provided to the respective column conductor, wherein the switching device is controlled by two inputs, a first input defined by the row conductor, and a second input.
The use of two inputs to the switching device enables an individual pixel within a row, or a group of pixels within a row, to be addressed. In other words, the row address pulse does not result in transfer of charge from the image sensing element of all pixels in the row. The second input is in practice associated with a c

Affiliated with

Also associated with

Rating

Image sensor does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Image sensor, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image sensor will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3118988