S. Dasu, University of Wisconsin February Calorimeter Trigger for Super LHC Electrons, Photons,  -jets, Jets, Missing E T Current Algorithms What can be improved? Bottom line Must hold the thresholds low to study electro-weak symmetry breaking physics Only option is to further reduce the backgrounds Electrons/photons/  -jets Dominated by tails of jet fragmentation to  0 Use of tracking in level-1 Pixel only or pixel + outer tracker layers? Mandates higher granularity calorimeter trigger output Jets and missing E T These are real - Only minor improvement in resolution likely Topology with vertex identification to reject pileup Pixel tracker provides vertex?

S. Dasu, University of Wisconsin February Calorimeter Geometry EB, EE, HB, HE map to 18 RCT crates Provide e/  and jet,  E T triggers

S. Dasu, University of Wisconsin February Current Algorithms

S. Dasu, University of Wisconsin February Electron/Photon Algorithm Current Algorithm Input: ~5x5 crystal E T sums (trigger tower sums) and a 1 bit characterization of the fine-grain profile Sliding window: Considers nearest neighbor trigger towers only - slides tower-by-tower E T sharpened by adding in nearest neighbor, 2x1 or 1x2 Isolation achieved by cutting on rest of the neighbors in ECAL and all neighbors in HCAL Result: Max E T e/  reported with “crude” position (0.37  x 037  region instead of  x  trigger towers) Rationale EM showers are narrow - R moliere ~ crystal dimension Adding more than 2 trigger towers is not needed Consider small isolation region to avoid accidentally vetoing in presence of multiple crossing pileup

S. Dasu, University of Wisconsin February e/  Improvements for SLHC Learn from current High Level Trigger Algorithms Improved isolation Gains realized when pileup is small Veto  0 using careful analysis of shower profile Improved Fine-Grain analysis Use track match Pixel only track match Pixel + Outer layer track match This provides most improvement in S 2 /S+B for electrons Must report calorimeter e/  objects in finer  bins

S. Dasu, University of Wisconsin February Improved Fine Grain Info.? Trigger tower size reduction Regroup crystals to reduce tower size in  (but not  due to spread in the B field) Additional fine grain analysis at crystal level Minor gains can probably be achieved using more sophisticated logic - requires change in an ASIC Good improvement requires data exchange between trigger towers Both these changes require changes in on- detector electronics, which is very difficult Access to front-end boards very limited Tied to mechanics, cooling and electrical grounding

S. Dasu, University of Wisconsin February Off-detector Electronics Increased isolation region, e.g., 4x4 instead of 3x3 Any additional data exchange can probably be absorbed by serving more channels per board Must keep the cuts high to avoid accidental vetoing due to higher pileup at SLHC Redesign the system for higher data-exchange between cards - is important for photon trigger. Reporting in finer  x  bins One can report position in trigger tower dimension Improved logic to eliminate potential double counting Necessary to provide linking of e/  to a “track” trigger. Again, this uses more data exchange between cards, but is important for electron trigger.

S. Dasu, University of Wisconsin February  -jet Algorithm Current Algorithm Marks towers above a certain threshold to use in a pattern search for finding narrow objects in 4x4 trigger towers, using memory lookup and discrete logic Uses wide isolation region (12x12 trigger towers) Uses the full jet energy (12x12) for narrow  -jet Rationale Use minimal additional logic to classify jets as  -jets Use of pattern search is easier than using energy sums Patching on the  -jet finding to jet algorithm simplifies logic - adders and sliding windows are expensive

S. Dasu, University of Wisconsin February  -jets Improvements for SLHC Algorithm for  -jets Dedicated narrow jet finding New clustering in 2x2 trigger towers New sliding window on 2x2 sliding by 1x1 Reporting  for  -jets in trigger tower steps Isolation for quark/gluon jets Additional compare circuits Track matching Similar to electron trigger Pixel only or Pixel + outer track layer Allow both 1 prong and 3 prong matches

S. Dasu, University of Wisconsin February Electronics for  -jets New sliding windows + energy sum tree Sums 2x2 sliding on 1x1 Can be integrated into the e/  summation tree New isolation region 6x6 trigger tower region Can be part of jet trigger sums Impact Requires additional hardware Conceptually no different from present system for e/  Can merge in primitives (2x1/1x2,2x2,4x4,6x6,…) for all triggers into a new scheme

S. Dasu, University of Wisconsin February Jet Trigger Current Algorithm Slides on 4x4 trigger towers Sums 12x12 trigger towers Energy resolution set to 8 bits at input Jets reported at (0.37  x 037  region instead of  x  trigger tower resolution) Potential Improvement Option of 6x6, 8x8, 10x10, 12x12 jets Reduced pileup for low E T jets if necessary Report at fine granularity Slide in 1x1 or 2x2 Impact Increased data sharing and adder trees

S. Dasu, University of Wisconsin February Missing E T Current Algorithm E x, E y lookup done using strip sums in bins of  with a granularity  = 0.37 Summed E x, E y used to determine missing ET Potential Improvement E x, E y lookup at  = Improved resolution, 10-bit adders Impact Increased memory lookup Increased data sharing Increased adder count