Speakers
Description
Summary
At the CMS experiment, the event rate at the nominal LHC
luminosity will
approach 1 GHz (about 20 events at each “bunch crossing”
every 25 ns). This
enormous rate will be reduced to the data taking capacity of
100 Hz in a
two-stage process, where the first, hardware-based stage
of the “Level-1
Trigger” has to reduce the rate by a factor of 10,000 to
below 100 kHz. Due
to the limited length of the digital pipelines in the various
subdetectors,
the Level-1 trigger decision must be available within 3.2
microseconds. The
Level-1 trigger and therefore also the L1 Global Trigger will
operate in an
intrinsically dead-time free, synchronous mode, where a
yes/no decision is
calculated for each bunch crossing and becomes available at
a fixed delay
after the event.
Information from calorimeters and muon triggers systems is
used to
calculate this decision for up to 128 different trigger
algorithms, which
are calculated in parallel. Different downscaling factors can
then be
applied to these 128 trigger bits, which are subsequently
combined into a
single decision (‘FinalOr’). Based on this decision, a “Level-1
Accept”
trigger signal is issued if all parts of the CMS detector and
read-out
systems are ready to accept a trigger.
The primary trigger objects are electrons or photons, muons,
taus, jets,
and very weakly interacting particles detected indirectly
through missing
transverse energy. Muons are detected by three systems
built up of drift
tubes, cathode strip chambers, and resistive-plate
chambers. The
information from these three systems is combined in
the “Global Muon
Trigger” (GMT), which also receives information from the
calorimeters to
see if a muon candidate is isolated or not, and if it
corresponds to a
minimum-ionizing particle. Information on all other particles
is yielded by
the electromagnetic and the hadronic calorimeters and
combined first in the
“Regional” and then in the “Global Calorimeter Trigger”.
The hardware of the Level-1 Global Trigger consists of
custom-built VME
modules using FPGA technology, which are housed in one 9-
U VME crate
together with the module of the Global Muon Trigger and the
central trigger
control module. Signals from the muon systems are received
by the GMT
module while signals from the calorimeters are sent
to “Pipeline
Synchronizing Buffer” input modules, which assure that all
signals
referring to one particular event enter the “Global Trigger
Logic” (GTL)
module at the same time. All trigger algorithms are
implemented in the
firmware loaded into the FPGAs of the GTL module and may
thus be changed at
any time. The “Final Decision Logic” module combines the
trigger bits
calculated by the GTL as well as “technical” trigger bits sent
by other
systems (for purposes of calibration etc.) after applying the
individual
downscaling factors and sends its decision to the “Trigger
Control System”
module (TCS), which also receives information on the state
of the various
subsystems (“ready”, “busy”, “overflow warning”, “error
state” etc.) and
then issues a “Level-1 Accept” (L1A) signal if a trigger is
requested and
the detector is ready to accept it. Eight independently
running control
units (PTC) allow to combine subdetectors to groups during
calibration and
test periods.
