The CDF silicon vertex trigger Page: 2 of 4
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to Digital Library by the UNT Libraries Government Documents Department.
The following text was automatically extracted from the image on this page using optical character recognition software:
orders of magnitude, the total inelastic cross-
section at the Tevatron is about 50 mb, while
the b-quark cross-section within CDF's accep-
tance (transverse momentum PT > 6 GeV, ra-
pidity JyI < 1) is about 10 pb, and the t-quark
cross-section is about 5 pb. At luminosities above
0.35 x 1032 cm-2s-1, the mean number of interac-
tions per beam crossing exceeds 1. Reducing the
1.7 MHz beam-crossing rate to CDF's 70 Hz DAQ
output rate implies a trigger rejection of 25000.
Good background rejection in the trigger re-
quires fast identification of distinctive signal sig-
natures. In the CDF trigger, many important sig-
natures exploit fast charged-particle track recon-
struction in the bending plane of the spectrom-
eter, transverse to the beam axis. The trigger
matches drift chamber tracks with EM calorime-
ter showers, muon chamber stubs, and silicon
detector data, respectively, to identify electrons,
muons, and b and c daughters.
CDF uses a three-level trigger. On each
beam crossing (396 or 132 ns), the entire front
end digitizes (silicon samples and holds). A
5.5 ps pipeline of programmable logic forms axial
drift chamber tracks and can match these with
calorimeter and muon-chamber data. On Level 1
accept, front-end boards store the event to one
of four buffers (silicon digitizes and transmits to
the silicon trigger and event builder). Level 2
processing, with about 30 ps latency, adds fast
silicon tracking, calorimeter clustering, and EM
calorimeter shower-max data. The final Level 2
decision is made in software on a single-board
computer, so a wider range of thresholds and de-
rived quantities is possible (e.g. transverse mass
of muon track pairs), even for information that is
in principle available at Level 1. On Level 2 ac-
cept, front-end VME crates transmit to the event
builder. At Level 3, a farm of 250 commodity PCs
runs full event reconstruction. This is the first
stage at which three-dimensional tracks (e.g. for
invariant mass calculation) are available. Events
passing Level 3 are written to disk.
While some optimization remains to be done,
the maximum output at L1/L2/L3 is approxi-
mately 35000/350/70 Hz. Each of these rates
is an order of magnitude higher than in CDF's
1992-96 running period. In addition, drift cham-
ber tracking has moved from L2 to L1, and sili-
con tracking has moved from offline to L2. These
three changes allow CDF to collect large sam-
ples of fully hadronic bottom and charm decays,
by requiring two drift chamber tracks at L1, re-
quiring each track to have a significant (at least
120 pm) impact parameter at L2, and perform-
ing full software tracking at L3 to confirm the
hardware tracking. The samples made possible
by CDF's front-end, trigger, and DAQ upgrades
have yielded novel physics results  at an early
stage of Run 2.
CDF's Level 1 drift chamber hardware track
processor, XFT , is a cornerstone of the CDF
trigger. For every bunch crossing, with 1.9 ps
latency, it finds tracks of PT > 1.5 GeV with
96% efficiency. XFT obtains coarse hit data (two
time bins) from each axial drift chamber wire,
finds line segments in the 12 measurement lay-
ers of each axial superlayer, then links segments
from these four superlayers to form track candi-
dates. XFT's resolutions, o(-1) = 1.7%/GeV
and -(0o) = 5 mrad, are only about a factor of
10 coarser than those of the offline reconstruction.
2. SVT TRACK PROCESSING
For each event passing Level 1, the Silicon Ver-
tex Trigger (SVT) [6-8] swims each XFT track
into the silicon detector, associates silicon hit
data from four detector planes, and produces
a transverse impact parameter measurement of
35 pm resolution (50 pm when convoluted with
the beam spot) with a mean latency of 24 ps,
9 ps of which are spent waiting for the first sili-
con data. SVT's impact parameter resolution for
PT ~ 2 GeV is comparable to that of offline tracks
that do not use Layer 00 (mounted on the beam
pipe), which is not yet available in SVT.
For fiducial offline muon tracks from J/0 de-
cay, having PT > 1.5 GeV and hits in the four
silicon planes used by SVT, measured SVT effi-
ciency is 85%. The most suitable definition of
efficiency in a given context depends on what one
aims to optimize: restricting the denominator to
PT > 2 GeV increases the efficiency to 90%, while
relaxing the requirements on which layers contain
offline silicon hits reduces the efficiency to 70%,
Here’s what’s next.
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Ashmanskas, B.; Barchiesi, A. & Bardi, A. The CDF silicon vertex trigger, article, June 23, 2003; Batavia, Illinois. (digital.library.unt.edu/ark:/67531/metadc737597/m1/2/: accessed March 20, 2018), University of North Texas Libraries, Digital Library, digital.library.unt.edu; crediting UNT Libraries Government Documents Department.