Note: Descriptions are shown in the official language in which they were submitted.
CA 02497465 2005-02-28
FIELD OF THE INVENTION
The present invention relates to an on-pixel readout circuit for detectors,
and
more particularly to an active pixel sensor (APS) circuit for radiation
detection,
capable of providing stable and predictable on-pixel amplification in the
presence
of device degradation and/or mismatch, and variation in environmental factors
like temperature and mechanical strain. This invention is applicable to
various
detector technologies such as flat panel imagers, sensor arrays, x-ray
detectors,
CMOS cameras, etc.
BACKGROUND OF THE INVENTION
2. Description of the Prior Art
The following description will be based on an example of a flat panel x-ray
imaging system, although the invention is not restricted to this invention.
Flat-
panel detectors have gained significant interest recently in view of their
ability to
provide an immediate, high quality radiographs after the exposure. They
require
less handling, more convenient management, immediate image viewing,
computer aided analysis and more convenient storage on computer disks rather
than archive film stacks [1].
Flat-panel detectors convert incident radiation images to charge images. In
the
case of x-rays, for instance, the radiated x-rays may be converted to charge
directly [2] by using an x-ray sensitive photoconductor such as amorphous
selenium, a-Se, or indirectly [3] by using a phosphor screen to convert an
incident x-ray image to light, which is then converted to a charge
distribution. In
both approaches the final charge image is stored on the pixel capacitors. This
stored image is then read out using a large area integrated circuit backplane
usually referred to as active matrix. Active matrix addressing involves a
layer of
backplane electronics, based on thin-film transistors (TFTs) fabricated using
amorphous silicon (a-Si:H), polycrystalline silicon (poly-Si), CMOS, organic,
polymer, or other transistor technologies, to provide on-pixel amplification
and
read out capability in each imaging pixel. In order to utilize the flat panel
for
reliable radiation detection, it is necessary to decrease the electronic noise
or
CA 02497465 2005-02-28
increase the radiation signal. One basic approach to increase the signal to
noise
ratio would be on-pixel amplification of the input signal, In this approach,
the
TFTs will act as both analog amplifiers and switches compared to the
conventional case where the transistors were used just as switches [3].
However, transistor threshold voltage and mobility instability, and mismatch
issues are a major hurdle in flat panel active matrix applications with on-
pixel
amplification. This is because the amplifying circuits are far more sensitive
to
TFT instability. As an example of instability, Fig. 1 shows the threshold
voltage
shift in amorphous silicon TFTs. In a flat panel, the instability would mean
that
the amplifier gain and the output DC current would vary across the array
and/or
decrease over time, which is unacceptable.
One solution to this stability problem is to use current programmed pixel
circuits.
The circuit shown in Fig. 2 is a current programmed pixel circuit, which
automatically compensates for any shift or mismatch in the threshold voltage
of
the drive TFT to ensure that the output DC current and circuit gain do not
decrease over time.
CA 02497465 2005-02-28
SUMMARY OF THE INVENTION
Accordingly, the object of the present invention is to provide pixel circuits
for
radiation detectors, capable of providing stable and predictable on-pixel
amplification, despite device degradation and/or mismatch, and change of
environmental factors like temperature and mechanical strain.
In order to achieve these goals, a family of pixel circuits for radiation
detection is
presented. The present invention comprises a plurality transistors in CMOS or
thin film technology for amplifying the x-ray generated signal in each pixel
of the
flat panel. These pixel circuits are automatically capable of compensating
shifts
and mismatches in the characteristics properties of the transistors in a
pixel.
The basic architecture comprises of a plurality of switching transistors, an
amplifying transistor, and a sensor for detection of radiation (such as
optical
signal, x-ray signal, etc.).
CA 02497465 2005-02-28
BRIEF DESCRIPTION OF THE DRAWINGS
The above objects and features of the present invention will become more
apparent by the following description of the preferred embodiments with
reference to the attached drawings. Although they directly apply to TFT
circuits in
amorphous silicon (a-Si) flat panel technology, they can be readily extended
to
other technologies like polysilicon, organic, polymer, CMOS, etc. for visible
and
x-ray detection.
Fig. 1- Threshold voltage shift as a function of stress voltage of a discrete
a-Si
TFT.
Fig. 2- Schematic of current mode active pixel sensor circuit with TFT
parameter
compensation capability along with the driving waveforms.
Fig. 3- Output current versus threshold voltage of the AMP TFT presented here.
Fig. 4- Input-output transfer characteristics demonstrating the current gain.
Fig. 5- An array structure of the pixel presented in Fig. 2.
Fig. 6- Another version of the circuit in Fig. 2 along with the driving
waveforms.
CA 02497465 2005-02-28
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Fig. 2 and Fig. 2 show the presented pixel circuit along with its control
signals. It
comprises four transistors T1, T2, READ and AMP, a storage capacitor CSroRe
and a radiation sensor (SENSOR). The pixel circuit is connected to a common
ground, a select line (SELECT), an input bias signal fine (IBIAS) and an
output
signal line. The four transistors T1, T2, READ and AMP can be a-Si:H,
nano/micro crystalline silicon, poly silicon, organic thin-film transistors
(TFT),
transistors in standard CMOS, or other technologies. AMP is the amplifying
transistor while the other three act as switches. Central to the circuit is a
common
source circuit comprising of AMP and READ TFTs, which produces a current
output to drive an external charge amplifier. The active matrix array
architecture
is assumed to be column-parallel, i.e., one charge amplifier per column so
that
an entire row can be read out simultaneously. The APS circuit operates in
three
modes:
Programming: the SELECT control signal becomes high; consequently the pixel
is selected and transistors T1 and T2 turn on. If we assume that initially
there is
no charge on the storage capacitance (CsroRE), the AMP TFT is off and
therefore
the constant programming current Ig,AS will flow through T1 and T2, which
charges up the storage capacitor. Simultaneously, the drain-source current of
the
AMP TFT will increase until eventually it becomes equal to IB~AS. After that
there
would be no further increase in the voltage of CSTORE, which is the same as
the
voltage across the gate-source terminal of the AMP TFT. In this stage, the AMP
TFT operates in its saturation mode where the current-voltage is given by
~ _ ~~.U ~Ci '~ W ymn I vGS -vT ~~ - K I\VGS VT la
CU ~' e(t
(1)
Here, ,u~ is the effective device mobility, ~' is a material parameter, C; the
gate
dielectric capacitance, a a coefficient that ranges between 2 and 2.4,
Ysat - ~~ + ~~~DS
(2)
CA 02497465 2005-02-28
and ~, is the channel length modulation parameter. The voltage across CgTORE
stabilizes at the point when all of IB~AS flows through T2 and the AMP TFT,
and
none through T1. This process is independent of the parameters of the AMP
TFT.
Integration: The SELECT line is switched to low. The voltage stored in CSTORE
remains constant, except for the initial drop due to charge feed-through
stemming from the gate-source capacitance of T1 when it turns off.
Subsequently, the integration starts, during which the incident x-ray photons
will
generate electron-hole pairs in the x-ray sensitive a-Se layer. These electron-
hole pairs are separated and driven by an applied electric field to the
surfaces of
the photoconductor where they will form a latent charge image. In the
integration
period T,N,. , the input signal, h v , generates carriers that discharge
CSTORE bY
~QP and increase its potential by a small-signal voltage of v,n
Readout mode: After integration, the READ TFT is turned ON. As a result, a
common source topology in which AMP TFT is the amplifying transistor and
READ TFT acts as a passive resistor, will be constructed. The output current
of
the circuit at this point would be
I BIAS + gm X vin
(3)
Where gm is the transconductance of AMP TFT and v,n is the small signal
voltage
generated during the integration period.
The on-pixel gain of the pixel is the change in the output current dlo"T ,
with
respect to the incident x-ray illumination h v
g = d ~~our ~ - d ~~QP ~ , d ~~V~ ~ d ~01 ovr
d~hv~ d~hv~ d~~QP~~ d~tlV~~
(4)
CA 02497465 2005-02-28
Here, dQP is the input charge signal due to pixel irradiation and dv~ is the
corresponding change in the gate voltage of the AMP TFT. The value of the
first
term is identified by the characteristics of the x-ray detector and the amount
of
change in the charge of CSroRE it gives with changing h v . The second term
depends upon the voltage change across CsroRE due to change in its charge
dQp = dYG.CSTORE
(5)
For the last term to be constant, the linear small signal condition on the AMP
TFT
should be imposed during operation
d Y~ « 2(V~ - Vr
(6)
Here Y~ and VT are the gate bias voltage of the AMP TFT and its threshold
voltage, respectively.
As shown in Fig. 2, the output current of the APS circuit will flow through an
off-
chip charge amplifier where it can be converted into voltage and further
amplified. Assuming constantdIoUT, the change of the charge amplifier output
voltage dVour will be
Ts
dVOUT - 1 j~OUTdt - ~OUTTS
CF 0 CF
O'mdVc ~T'S ~gm~in ~Ts
- CF - CF
(7)
where v;n represents the small signal voltage at the gate of the AMP TFT due
to
x-ray induced electron-hole pairs and TS is the pulse width during the Readout
mode. Using (4) and (7), the total charge gain G,~, defined as the change in
the
charge across the feedback capacitance of the charge amplifier over the change
in the charge of pixel capacitance is
CA 02497465 2005-02-28
~Qour W ouT ~Ts ~ W our y's
Gror = _
~QP ~~P ~yGCSTORE
_ (g,n .TS
CsroRE
(8)
Using (7) and (8), the charge gain G,oi can be related to the voltage gain
fli, ~' ~ vOUT ~ vin ~ a$
CF
Gu~r = Av .
CsroRs
(9)
Fig. 6 is another version of the presented method. The major difference is
that in
the readout mode the circuit becomes a common-source with active load.
Consequently, this circuit has higher voltage gain compared to the circuit in
Fig.2.
References:
1. Chan HP, Doi K, Galhotra S, Vborny CJ, MacMahon H, Jokich PM, " Image
feature analysis and Computer-Aided Diagnosis in Digital Radigraphy. 1.
Automated Detection of Microcalcifications in Mammography."MedPhys 14:538-
548, 1987.
2. W. Zhao and J. A. Rowlands, "X-ray imaging using amorphous selenium:
Feasibility of a flat panel self-scanned detector for digital radiography,"
Med. Phys. 22, 1595-1604, 1995.
3. L. E. Antonuk, J. Boudry, W. Huang, J. L. McShan, E. J. Morton, J.
Yorkston, M. J. Longo, and R. A. Street, "Demonstration of megavoltage
and diagnostic x-ray imaging with hydrogenated amorphous silicon arrays,"
Med. Phys. 19, 1455-1466, 1992.
