A Fast Monolithic Active Pixel Sensor with in Pixel level Reset Noise Suppression and Binary Outputs for Charged Particle Detection Y.Degerli1 (Member, IEEE), G.Deptuch2 (Member, IEEE), N.T. Fourches1 (Member, IEEE) A. Himmi2,Y. Li1, P. Lutz1, - PowerPoint PPT Presentation

1 / 1
About This Presentation
Title:

A Fast Monolithic Active Pixel Sensor with in Pixel level Reset Noise Suppression and Binary Outputs for Charged Particle Detection Y.Degerli1 (Member, IEEE), G.Deptuch2 (Member, IEEE), N.T. Fourches1 (Member, IEEE) A. Himmi2,Y. Li1, P. Lutz1,

Description:

A Fast Monolithic Active Pixel Sensor with in Pixel level Reset Noise ... Right : chronogram of the control signals. applied to the pixel ... – PowerPoint PPT presentation

Number of Views:68
Avg rating:3.0/5.0
Slides: 2
Provided by: four71
Category:

less

Transcript and Presenter's Notes

Title: A Fast Monolithic Active Pixel Sensor with in Pixel level Reset Noise Suppression and Binary Outputs for Charged Particle Detection Y.Degerli1 (Member, IEEE), G.Deptuch2 (Member, IEEE), N.T. Fourches1 (Member, IEEE) A. Himmi2,Y. Li1, P. Lutz1,


1
A Fast Monolithic Active Pixel Sensor with in
Pixel level Reset Noise Suppression and Binary
Outputs for Charged Particle Detection
Y.Degerli1 (Member, IEEE), G.Deptuch2 (Member,
IEEE), N.T. Fourches1 (Member, IEEE)A. Himmi2,Y.
Li1, P. Lutz1, F. Orsini11CEA/Saclay,
DAPNIA/SEDI and SPP, 91191 Gif/Yvette Cedex,
France2LEPSI and IRES/IN2P3, 23 rue du Loess,
67037 Strasbourg Cedex 02,France
04/10/04
DESCRIPTION
Left Schematic of the pixel of the first
type (Saclay design) Right chronogram of the
control signals applied to the pixel
Left Synoptic of the MIMOSA8 maps. The first
sub-array of pixels is the Strasbourg design The
second to fourth sub-array is a Saclay design
with decreasing conversion factors
Right layout of our MIMOSA8 design. the
digital control part is on the left side of the
view. the pixel arrays are in the middle. the
digital output block is located at the bottom
together with the discriminators
Above Photograph of the MIMOSA8 in its LCC 84
package. The size of each pixel is 25 µm x 25 µm.
The process used is the TSMC 0.25 µm available
through MOSIS. We chose the CMOS digital version
with 8 µm epitaxial layer. Capacitances were
MOS structures whose advantage is to provide high
values in a reduced area. The discriminators are
of the same design as the one used for MIMOSA6
(Y. Degerli et al. Low-power autozeroed high
speed comparator for the readout chain of a CMOS
monolithic active pixel sensor based vertex
detector, IEEE Transactions on Nuclear Science,
vol. 50, no. 5, October 2003,pp 1709-1717). The
pixel were designed in order to obtain different
conversion factors (CVFs) by adjusting the gain
at the sensitive node of the pixel.
RESULTS A thorough study of the
analog outputs was made before correction. The
figure on the right. shows the observed analog
output signal before column correction. Note the
2 levels for each pixel (VRD and VCALIB). The
useful signal is the difference between these two
levels. Tests without the source show that double
sampling eliminates offset dispersions of the
in-pixel amplifying stage. The offset dispersions
of the output stage is corrected later by the
column readout circuitry (discriminators). The
functionality of the digital output was also
proved here. The typical consumption of each
pixel is reduced to 40 µA.
Above analog signal of the pixels output in a
column. The grey parts of the line are the read
signals and the black parts are the baseline. The
signal is zoomed for clarity. One hit is
distinguishable at the n pixel. A 10 mCi 55Fe
source is used for X ray production.
Pixel array n Sensing element CVF (µV/e-) Output rms noise (_at_fCK100MHz) Input referred noise (ENC) S/N Pixel-to-pixel FPN
2 n-well/p-epi diode 1.2µmx1.2µm 70 1mV 14 e- 120 lt1mV
3 n-well/p-epi diode 1.7µmx1.7µm 60 1mV 17 e- 100 lt1mV
4 n-well/p-epi diode 2.4µmx2.4µm 50 1mV 20 e- 80 lt1mV
Above characteristics of the pixels and their
readouts for the blocks indicated.
Above Output of a pixel column This
corresponds to a block of 32 pixels with the
source on (55Fe,10 mCi) hits of high amplitude
are clearly present.
Above The same as the previous figure with fewer
hits of lower amplitude.
CONCLUSIONS Demonstration has been made
of the functionality of an array of pixels with
in pixel gain and double sampling. X ray
detection ( 5.9 keV and 6.4 keV corresponding
approximately to 1700 e-) is possible with this
pixel design. For a clock frequency of 100 MHz
FPN is lower than 1 mV. Estimated CVFs are
close to the designed ones. Given the present
signal to noise ratio Minimum Ionizing Particle
detection is possible (500 e-). Future work will
be oriented towards MIP detection with full
digital (one bit) operation and increased readout
speed. The 128 rows may be read in 20
µs. Acknowledgements the authors are thankful
to E. Delagnes1, F. Lugiez1, M. Rouger1 ,C.
Colledani2, for their help and advice.
Write a Comment
User Comments (0)
About PowerShow.com