Silicon Vertex Trigger (SVT) working group
status report at collaboration meeting
7 March 1997
(Horst Wahl)

OUTLINE:

• Introduction (motivation, history)
• working group and its mission
• outline of activities
• beam position stability
• Monte Carlo events
• CFT vs SVT
• tracking in SVT
• progress in design
• benefit for physics
• conclusions

INTRODUCTION

• SVT = L2 preprocessor
  • finds hit clusters in SVX
  • finds tracks in SVX in roads provided by CFT
  • determines track parameters ( , impact parameter)
  • ships information to global L2 processor for decision

• motivation:
  • tagging of b's, c's, 's at the trigger level;
  • b-physics, t-physics, Higgs searches, SUSY searches,

• previous working group (Brian Winer & Co.) made preliminary study of feasibility and physics benefits of SVT;
• findings summarized in DØNote 2984;

MAIN CONCLUSIONS OF PREVIOUS WORKING GROUP:

• SVT improved momentum resolution at trigger level
  sharper trigger turn-on, lower trigger rate;
• SVT has potential to benefit many physics topics
  (improved efficiency and S/N ratio accumulate more interesting events)
  • b physics (CP violation, mixing, rare decays)
  • top physics (all jets, + jets), for both and single top
  • light Higgs from associated production with W or Z (signature similar to single top)
  • SUSY: soft multileptons (e.g. wino, zino)
  • SUSY: stop jets with b-tag
• alignment of SVX detector critical for good operation of trigger
• conceptual design, estimated cost $800k;
• more detailed and more realistic studies needed
  

new working group formed in September 1996;

MISSION OF THE SVT WORKING GROUP:
To develop a proposal based on the work by the previous vertex trigger study group, as summarized in DØNote 2984.

Main tasks:

1. do more realistic simulations to assess the benefit for the physics potential;
   simulations should include
   • effects due to real (rather than ideal) detector, noise, inefficiencies;
   • effects due to algorithms in trigger (pattern recognition,..)
   • effects due to other parts of the trigger system
2. make progress in practical design;
   (only conceptual design given in DØNote 2984)
3. understand requirements on beam position stability, and possibilities of accelerator to satisfy these requirements
4. Manpower and cost estimate
5. write proposal

• first meeting: 13 September 1996, total of 12 meetings so far
• minutes can be found on Webpage
  <tex2html_verb_mark><tex2html_verb_mark>
  
• people who have contributed and/or shown interest:
  • Bob Angstadt,
  • Fred Borcherding,
  • Cecilia Gerber,
  • Rafael Gomez,
  • Phil Gutierrez,
  • Terry Heuring,
  • John Hobbs,
  • Marvin Johnson,
  • Boaz Klima,
  • Patrick Ledu,
  • Arthur Maciel,
  • Meenakshi Narain,
  • Doug Norman,
  • Henryk Piekarz,
  • Eric Smith,
  • Kathy Streets,
  • Andrzej Zieminski,
  • Armand Zylberstejn.

OUTLINE OF ACTIVITIES:

• implement latest SVX geometry into GEANT [MN]
  (change to 6 barrel geometry, include H-disks, 90 stereo, digitization,..)
• provide software package to access SVD information in GEANT output [MN]
• include all material (support, cables, electronics,..) in simulation [Greg Landsberg, TH]
• generate MC events + GEANT (7-barrel geometry) for physics channels of interest [BK, MN]
• study alignment and beam position stability issues [Paul Derwent, HP]
• study efficiency of CFT for finding tracks from displaced vertex [FB,RG,MJ]
• study fake track candidates in CFT [FB, KS]
• define tracking algorithm for SVT [PG, ES]]
• implement tracking code in trigger ntuples [CG,PG,KS]
• single track studies to debug trigger ntuple code
• physics studies
• technical feasibility studies, prepare detailed design [FB,MJ,HP]

BEAM POSITION STABILITY QUESTIONS
(Henryk Piekarz, Paul Derwent as consultant))

• beam size at IP (interaction point)
• expected resolution on impact parameter
• need precise alignment (position and direction) of beam relative to detector.
• CDF studies during last run show:
  store to store variations:
  • horizontal offset typical, up to possible
  • vertical offset typical, up to possible
  • angle variations typical
  change of position during the store:
  • horizontal and vertical typical,
  • angle typically too small to be measurable
• CDF: feedback beam control system, using measured beam position to set dipole corrector magnets so as to stabilize beam position and angle
  • range of variation determined by range of magnet powersupply ( A)
    beam direction: range
    IP position range .
  • tests performed by CDF and accelerator division in June 1995 showed:
    with 5 minute updates, beam position could be kept stable to within , and beam angle to within
  • test in Feb. 1996: beam moved vertically, horizontally without change elsewhere around ring
    can control beam position at DØ and BØ independently
• initial alignment at installation:
  Using BPM's and flying wires installed temporarily for this purpose, CDF plans to measure absolute beam location in collision hall very precisely during commissioning period at beginning of run
  goal: install track detector with axis aligned to better than , center to better than
  motorized support to allow position adjustment.

EVENT GENERATION
(Boaz Klima)

Monte Carlo events available for physics studies:

• or or , where

• (all decay channels)
• (all decay channels):
• ( GeV)
  (SUGRA input: GeV, )
• , where and

• (for initial studies of background rates)
  = 50900., 5280., 532., 37., 1.8, 0.054 b for
  = 2-5, 5-10, 10-20, 20-40, 40-80, 80-500 GeV

• 
• 

INTERPLAY BETWEEN CFT AND SVT
(Fred Borcherding, Marvin Johnson, Rafael Gómez, Kathy Streets)

• SVT looks for tracks in roads given by L1 CFT;
• CFT track candidates found by 8-fold coincidences between fiber doublets
• coincidences (``equations'') determined assuming track comes from primary vertex
• question: efficiency of CFT in finding tracks coming from displaced vertex;
• 2 studies have been done, in agreement
• FredB.:
  • CFT efficiency varies quadratically with displacement,

    

    where D = vertex displacement in mm.

  • 
  • but for most of physics that we care about, D < 1 mm;
  • from direct production:
  • from top decay: ;

• Marvin's and Rafael's study (see DØNote 3119)
  • use samples of b decays from direct production and from top decay
  • consider only tracks with 1.5 GeV,
  • 99.7% of tracks from directly produced b have impact parameter < 1 mm;
  • 95.3% of tracks from b's from top decay have impact parameter < 1 mm;
  • use analytical method to define CFT roads and to calculate road acceptance vs impact parameter trigger efficiency vs impact parameter
  • efficiency > 95% for impact parameter < 1 mm.
  • folding b impact parameter distribution with trigger efficiency vs impact parameter, find:
    • efficiency for direct b production is 99.9%;
    • efficiency for b's from top decays is 99%;
• number of track candidates, fake tracks, optimization of CFT algorithm under study [Fred Borcherding, Kathy Streets]

TRACK FINDING AND FITTING IN SVT
(Phil Gutierrez, Eric Smith)

• get track candidates from CFT
• project tracks into Silicon Detector
  • projection equation
    • is curvature;
    • ;
    • r is distance from nominal beam line to detector element.
• fit track using fiber tracker and silicon hits
  • fit equation
    • b is the impact parameter.

• parameters determined by minimization;
• equations can be linearized by expanding parameters about a central track (Note the expansion exact since comes from linear least squares fit);
  • ;
  • central track currently taken as CFT track with projection into SVX
  • reduces parameter fits to products and sums;
  • derivatives can be calculated ahead of time - in LUT

• Figure 1:
  distribution of SVX clusters from projected track (all Si layers combined)
  

• Figure 2:
  Example of fit track
  • Dashed line-projection from CFT to SVX; Solid line-fit to hits;
  • Closed dots-hits used in fit; Open dots-all other hits;
  • .
• Figure 3:
  Impact parameter distribution for single muon track:
  • Track momentum ;
  • ;
  • Estimate of impact parameter resolution ;

SOME DESIGN CONSIDERATIONS FOR THE SILICON VERTEX TRIGGER
(Marvin Johnson)

Parameters defined by other parts of the DØ system:

1. Input data is obtained from the silicon by splitting the optical cable.
   This means that :
   1. Data format is that of the SVX II chip plus port card number.

   2. Data input rate is 53 Mhz

   3. Serial format requires 850 nm receiver and HP Glink chip.

   4. Average readout time is around 7 micro seconds
2. Level 1 input from fiber trigger is the same as the muon format
   1. 6 tracks per sector

   2. Momentum ordered with highest first

   3. Fiber number of the outer (and hopefully) inner barrel

   4. Sign of the track

3. If L2 fiber data is needed, it will only be available by splitting the optical signal; conditions as in item 1 above.

4. System will be located on the second floor of MCH (no other space)

5. Each trigger crate will be a geographic sector and will need to conform to DØ specifications for geographic sectors. This includes communication with the trigger framework for trigger information.

6. Data transfer to L2 will be by a specification that is now being finalized.

7. Data transfer to L3 will be via a VBD.

8. Total processing time must be less than 50 microseconds including data readout from SVX II and formatting for L2 Global.

9. Although not absolutely required, each trigger board should process a multiple of 4 readout fibers. The port card and VRB both are 4 fiber modules.

10. Modules must support VME 32.

Some design suggestions (not required)

1. The optical receiver design is done for the readout board so it could be copied.

2. preliminary design indicates that cluster finding can be done at 53 Mhz with an FPGA. This would eliminate storing all the data and then processing it. (see DØNote 3168)

3. it appears to be possible to do all the track fitting in commercial DSP boards.

4. Selecting data in a fiber road can be done with one FPGA for 4 fibers.

5. VME 64 can do 30 or more Mbytes/second so it is a good candidate for a bus

Some questions:

1. How to arrange fiber trigger signals relative to Si signals?
   Silicon is 6 fold symmetric. CFT trigger divided into 80 sectors. The fibers determine the roads so one needs to include all the silicon under a group of fiber roads. Would like to have Si trigger divided into sectors so that sector fits into one crate; at 4 fibers per chip, estimate that about 1 fifth of the total fit into one crate , i.e. fiber trigger has 16 sectors per silicon sector so maximum number of trigger tracks is 96.
   But could also envision division into six SiV trigger sectors (economy vs convenience).

2. How many tracks need to be processed per 4 fiber group?

3. How long does the fitting algorithm take in a fast DSP?

4. How many hits are in each road? This sets the speed required for VME transfers if one goes in this direction.

PHYSICS BENEFIT

• main beneficiary: b-physics
  (lower pt threshold increase acceptance/efficiency)
• other beneficiaries:
  physics channels which at present (without SVT) are bandwidth-limited at L2;
  i.e. efficiency could be increased with additional background rejection at L2;
  examples:
  • Higgs + W associated production
    see New Phenomena triggerfest study by John Hobbs
    
  • SUGRA to trileptons
    see New Phenomena triggerfest study by Doug Norman
    
    

SUMMARY

• good indications that SVT could substantially improve sensitivity to interesting physics;
• required beam position stability appears to be possible, but need to extend contacts with accelerator people
• correct positioning of central tracker at installation, and maintaining position afterward is critical issue;
• GEANT implementation of latest detector version (including dead material) in progress,
  
• present CFT L1 trigger is more than 99% efficient for b and top;
• fake track candidates -- studies in progress
• trigger tracking algorithm available, implemented in trigger ntuples;
• progress in understanding design issues, hitfinding
• we are very close to having all the tools needed to do quantitative assessment of physics benefits of SVT.
• Saclay physicists are showing interest in designing and building it.
• estimated time to build: about 2.5 years