Speaker
Description
Summary
For the readout of such a highly granular Silicon-Tungsten calorimeter, there are
several options available. From detailed simulation of the Tera-Pixel ECAL we know,
that most pixels are only hit once per event, if one chooses a pixel size of 50 x 50
µm. We can employ a simple binary readout using a comparator instead of an analog
readout, which simplifies the pixel layout. We have then designed and fabricated a
CMOS Monolithic Active Pixel Sensor (MAPS) in the novel INMAPS process. The INMAPS
process is a standard 0.18 micron CMOS image-sensor technology with a high energy
“deep-Pwell” implant located beneath the active circuits. A conventional MAPS design
will experience charge sharing between the sense-node(s) and any PMOS active devices
in the pixel which can dramatically reduce the efficiency of the pixel. By implanting
the “deep-Pwell” in the pixel regions containing active circuits, charge deposited in
the epitaxial layer is reflected and conserved for collection at only the exposed
collection diode nodes. The pixels contain four N-well diodes for charge-collection;
analog front-end circuits for signal pulse shaping; comparator for threshold
discrimination; digital logic for threshold trim adjustment and pixel masking. Pixels
are served by shared row-logic which stores the location and time-stamp of pixel hits
in local SRAM, at the target 150ns beam bunch crossing rate of the ILC. The sparse
hit data is read out from the columns of logic in the quiet time between bunch
trains. A prototype sensor consisting of 8 units of 42x84 pixels with 6 million
transistors in total has been produced. The data acquisition requirements for such a
system with 10^12 readout channels are driven by the noise. Even with a noise level
of 10^-6, there will be 1 million hits per event and the occupancy is entirely noise
driven. Therefore the required DAQ bandwidth is around 500 Gbit/s. Another system
issue is the power consumption. For a MAPS detector with a 1 % duty cycle we obtain a
power consumption of 40 µW/mm^2. The test sensor will allow us to explore options to
further reduce this. A clear advantage of the MAPS approach is the fact, that it can
be manufactured in a industry standard process and will be cheaper to produce than
the combination of high resistivity silicon sensor and a readout chip. Parallel to
the design work on the sensor itself, we have worked extensively on the physics
simulation of a MAPS-based ECAL. For an accurate sensor simulation the Sentaurus
package was used for optimizing the design layout and to study the charge spread
within the MAPS pixels. To study the physics performance, the MAPS based calorimeter
was implemented in the MOKKA detector simulation, which is based on GEANT4. The
simulation output has been used to test Particle Flow Algorithms and some first
results of using Particle Flow with a highly granular MAPS-based ECAL are presented.
