Presentation on theme: "Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop1 L1 Configuration and Future Prospects DCZ cuts and new GLT objects L1 rate diagnosis Can an improved."— Presentation transcript:
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop1 L1 Configuration and Future Prospects DCZ cuts and new GLT objects L1 rate diagnosis Can an improved EMT help ? Bhabha veto at Level 1 ? Su Dong
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop2 Perceptions, Questions and Desires We should at least maintain our current physics efficiency while trying to reduce L1 rate. The rate dominating 2 particle triggers are not sustainable by EMT and is mainly carried by DCT. Can we make them pure DCT and even open up kinematics ? (gain kinematically and shave geometrically with DCZ is a favorable trade off). EMT E,G,Y with higher energy cuts are quite robust against background, but what does G serve particularly ? While DCZ is reducing track triggers, what about EMT ? In particular the 3M related triggers ?
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop3 GLT Trigger Object Changes There are 17 GLT trigger objects for forming the trigger lines. The current first attempt in incorporating DCZ: Old DCT: A’,AM,BM,A’M,BMX replaced by New DCZ: Z’, Z, Zt, Zk, EZ The EZ object is E and Z back-back for e.g. ISR events. The match objects are considered expendable as the matching was primarily used as an effective Pt cut (takes 250 MeV Pt to reach EMC). DCZ ZPD objects have controllable Pt cuts in themselves. This is a first attempt. We may well end up deciding on something different. It’s a fairly localized GLT firmware change.
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop4 ZPD single track efficiency from 2-prong events
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop5 ZPD single track efficiency in hadronic events. Fortunately we normally only need 1 or 2 tracks among many. Z cut range at 10-15cm as expected from design simulation.
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop6 L1 Rate with Trial DCZ Config Significant changes: Old New 3A&B* 3B&2A&2Z B*&1A&1A’ B*&1A&1Z’ 2M&1A&1A’ 2M&1A&1Z’ 2BM&2M 2Zt&1A/2 M*&1B/4 M*&1Z 1Z alone is ~10Khz ?
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop7 L1 Trigger Rate with Trial DCZ Config For L1 triggers rejected by L3 ( and not firing L3 Bhabha script )
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop8 Compare with DCZ design simulation study CDR study Sep/01. Simple 1Z cut 2-prong reduction close to expectation, but 3-prong reduction is less in data ?
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop9 EMT Trigger Issues E,G signals and 4M are resilient to background. Can we do something to the 3M related lines ? If we had 3D EMT clusters instead of strips, would that push some physics 3M to 4M ? A brief look was done sometime ago and surprisingly not that much. 3M is actually taking a wide mixture of physics: hadrons, tau, radiative Bhabha/mu etc. so that anything serious on this needs careful study.
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop10 L1 Bhabha Veto ? The L1 fiducial Bhabha is ~53nb. This is 500Hz at 10 34, 1.5Khz at 3x10 34. It’s worth considering once past 1.5-2x10 34. A study was done ~ 2 years ago: –Steal EMT G = a strip Esum>2.4 GeV signal –1-prong veto: nG=1 & nY=1 & 1<=nM<=2 & nA’=1 & nA=1 & nB=1 & nBM=1 –2-prong veto: nG*=1 & nM=2 & nB=2 & nB*>0 & nBM=2 & nAp>0 (can probably do better still with more plays of cuts) - Only apply veto to 2 particle L1 lines There is at least means of firmware this down in GLT decision FPGAs.
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop11 Not too bad but more careful looks for more physics channel. The L3 and Lumi bookkeeping will also be complicated.
Dec/02/04 Su DongCaltech Trigger/DAQ/Online workshop12 Summary L1 configuration study using DCZ just starting and the ZPD tracks will be the main hope for background reduction. Some work is needed in deciding what can be done to the EMT 3M related triggers. There is a reasonable prospect for vetoing Bhabhas at L1, but only worth doing at very high luminosity. There are other improvements such as TSf lookup table calibration should also be explored.