-- TaoHuang - 2019-01-12

This twiki is Taohuang's note for OTMB related studies and includes Run2 and post Run2 developments. Work is mostly completed by CSC local trigger group (TAMU+UCLA and others)

Plan, 2019 Jan.

The following changes that are correlated to OTMB firmware are expected:

  • GE1/1 installation is during LS2
  • GE2/1 will be commissioned during YEST 2021,2022
  • New OTMB board designed by Jason will be deployed for ME1/1 and ME2/1 and previous OTMBs for ME1/1 during Run-2 operation will be moved to ME3/1 and ME4/1

Several versions of OTMB FW to accommodate these changes in endcap forward region in CMS

  1. CSC-alone OTMB FW for ME2/1, with old OTMB. the changes on top of OTMB FW for ME1/1 (old OTMB board) is to remove ME1/a part and replace 7 fibers by 5 fibers. The fiber input order in patch panel for ME2/1 is probably different from the one for ME1/1.
  2. CSC-alone OTMB FW for ME3/1, ME4/1, with old OTMB
  3. CSC-alone OTMB FW for ME1/1, with new OTMB board. The changes is mainly on pin assignment file(ucf) compared to FW for previous OTMB
  4. CSC-alone OTMB FW for ME2/1, with new OTMB board.
  5. GE1/1-ME1/1 integrated local trigger OTMB FW
    • Phase1: GEM+CSC communication and GEM+CSC DAQ data handling. Was demonstrated by UCLA group during 2016-2017. Recap these development sine Jan. 2019
    • Phase2: GEM+CSC integrated local trigger by combining GEM hits and CSC hits. the Matching between GEM and CSC incorporates the alignment effect
  6. GE2/1-ME2/1 integrated local trigger OTMB FW
    • on top of GE1/1-ME1/1 OTMB FW

Current OTMB documents

OTMB algorithm upgrade for high luminosity

DN-13-016, DN for high lumi OTMB algorithm

As LHC plans to move to high luminosity era in future and proton-proton scattering per bunching cross will also increase from ~40 (Run2) to 140-200, the forward region in endcap becomes more and more challenging. Local trigger would seriously suffer this harsh environment and starts to become less performance. To improve the local trigger in high luminosity and enhance the muon triggering, the CSC alone OTMB algorithm upgrade and new detection installation were proposed. The CSC alone OTMB algorithm upgrade project is focused on improving CSC local trigger by implementing some key features and changing the operation configurations :

  • Localized dead time zone in pre-CLCT trigger
  • Enable CLCT reuse
  • Smarter LCTs sorting algorithm
  • Reducing ALCT-CLCT matching window to reject the pileup
  • Other changes in ALCT part

Simulation studies using single muon gun sample showed that improved OTMB algorithm could give a few percents local trigger efficiency increase and it does not require any changes in hardware, which means a low cost upgrade. Simulation studies were documented in DN-13-016.

OTMB firmware development for high luminosity upgrade, 2018 OTMB FW

Back to 2016, Yuriy and Jason started to add new features studied in DN-13-016 in OTMB FW. Later starting from 2017, Tao and Jason continued this project by validating and testing the FW. Some Bugs were fixed and FW code was improved. The changes on OTMB FW and operation were categorized in 4 features:
  • Localized dead time zone in pre-CLCT trigger preCLCT trigger is fired when a valid pattern of comparator hits is found with least 3 layer hits. Every time one halfstrip, is fired by preCLCT trigger, the old OTMB algorithm would freeze the whole chamber for a few BX (~4 or 5BX) from next preCLCT trigger. In other word, each preCLCT trigger could cause ~4 or 5BX dead time in whole chamber. The new algorithm removed this dead time constraint in pretrigger state machine, and to void retriggering on same muon, a local zone around fired halfstrip is prohibited from triggering. let us say halfstrip N is fired in preCLCT trigger and the local dead zone is [N-4, N+4] if N is in central region of chamber, other the zone is limited by chamber boundary.
  • Enable CLCT reuse In old algorithm, the CLCT is always matched with first ALCT in term of timing and then is expired for next ALCT which might be the right. New algorithm could allow CLCT to be valid in whole matching window and to match with more than one ALCT.
  • Smarter LCTs sorting algorithm This feature is almost included in old OTMB algorithm and ideal is to send out the LCT built from smaller ALCT and CLCT timing difference if more than two LCTs presence. The supporting reason is that LCT originated from real muon usually has close ALCT and CLCT timing while the mismatch ALCT-CLCT pair has larger timing difference.
  • Reducing ALCT-CLCT matching window to reject the pileup In high pile up environment, the mismatch ALCT and CLCT could happen quite often as stubs from pile up are mixed with the one from real muon. To avoid the mismatch and maintain high efficiency of right match, reducing the Run-2 nominal 7BX wide window to 5BX or even 3BX could help decrease the mismatches.

Above features are motivated primarily from DN-13-016 and during OTMB FW development and tests, some other changes are proposed and tested (A collaboration in CSC community, especially Jay Hauser, Andrew Peck, Cameron Bravo, Will Nash from UCLA + Tao and Jason from TAMU):

  • Fix the bug in DCFEB readout in OTMB FW
  • Reduce DCFEB edge effect by expanding the adjacent zone for readout from 5HS to 6 , 7 or 8
  • Reduce L1A -TMB matching window from 7BX to 5BX or even 3BX to avoid mismatch between L1A and TMB signal
  • Extend comparator hit persistence from 4BX to 5BX to increase CLCT efficiency
  • Loose ME1/1 and ME2/X CLCT layer requirement from >=4 layer hits to >=3 layer hits

The 2018 OTMB FW was tested building 904 ME1/1 test stand (cosmic), GIF++ test stand (test beam) and Point5 (pp collision + heavy ion collision). Some conclusion from OTMB FW tests:

  • OTMB FW tests at b904 showed that no significant difference between old and new OTMB FW in term of performance and LCT timing. Observed differences are expected and understood. And a long cosmic run was token and it showed that new algorithm with localized dead time zone feature could allow two sequential trigger within 5BX while old algorithm never reports two sequential trigger less than 5BX at b904
  • OTMB FW tests at GIF++ also showed no significant difference between old and new OTMB FW OTMB FW was tested at GIF++ 3 times.
    • First test was 2018 May and one bug was found with OTMB when CLCT+L1A readout was enabled. Bug was later fixed and new version FW is built
    • Second test was 2018 June and ALCT-CLCT match priority bug was found (also confirmed at b904 test, P5 test). Bug is observed in CLCT_BXN and L1A BXN monitor plot and it is due to wrong assignment in ALCT-CLCT matching priority. Bug was also later fixed.
    • The last OTMB test at GIF++ in 2018 is end of August and it confirmed that new algorithm works perfectly
  • OTMB FW tests at Point5 showed that new algorithm could dramatically improve local trigger efficiency by ~5% in ME1/a region and also improve offline segment reco efficiency by a few percents. The stable new OTMB FW was released on 2018-08-20. The improvements from new OTMB FW were measured and justified by CSC efficiency group using collision data (Isabelle De Bruyn, Stephen Trembath-Reichert, UW-Madison) and data-emulation study. Data-emulation study with event display also showed that the localized dead time zone is working as expected and allow two sequential trigger within 5BX.

The summaries of Point5 tests were documented here: https://twiki.cern.ch/twiki/bin/viewauth/CMS/CSCOTMB2018

GEM-CSC integrated local trigger algorithm, simulation

DN-14-018, DN for GEM-CSC algorithm

OTMB firmware development for GEM-CSC algorithm, GE1/1-ME1/1

GEM+CSC OTMB FW development is a big project during LS2 and the final goal of this firmware is to achieve the full GEM+CSC integrated local trigger(ILT) simulated in CMSSW. GEM+CSC ILT is to ensure that CSC trigger info is efficiently sent to EMTF for L1 muon finding and GEM-CSC bending is supposed to be calculated at EMTF with better resolution as EMTF receives both OTMB (CSC info) and CPT7(GEM info) trigger data.

The dedicated GEM+CSC OTMB FW twiki: https://twiki.cern.ch/twiki/bin/view/Main/GEMCSCOTMBFWDevelopementLS2

OTMB firmware development for ME2/1, M3/1, ME4/1

OTMB firmware development for GEM-CSC algorithm, GE2/1-ME2/1

Test stands at building 904 and Point 5, CERN

list of test stands for different purposes:
   
GEMCSCTeststand b904.jpeg
GE1/1 and ME1/1 test stand at b904
GEMCSCteststand sketchup b904.png
GE1/1 and ME1/1 test stand at b904

In addition to ME1/1 test stand at b904, the one GEM chamber is placed on top of ME1/1 chamber and a test stand for GEM related development was set up. A joint ME1/1-GE1/1 test stand integration is expected in future

OTMB Test stand at TAMU

Twiki created by Jay and it should be updated!!! https://twiki.cern.ch/twiki/bin/view/CMS/TAMUTestStandME11

New OTMB design, Jason Gilmore

Topic attachments
I Attachment History Action Size Date Who Comment
PNGpng CMSView_GE11GE21ME0.png r1 manage 399.3 K 2019-02-08 - 17:13 TaoHuang  
PDFpdf DN-13-016_temp.pdf r1 manage 6428.9 K 2019-01-15 - 23:06 TaoHuang  
PDFpdf DN-14-018_temp.pdf r1 manage 1191.6 K 2019-01-15 - 23:06 TaoHuang  
JPEGjpeg GEMCSCTeststand_b904.jpeg r1 manage 241.4 K 2019-02-08 - 17:51 TaoHuang  
PNGpng GEMCSCteststand_sketchup_b904.png r1 manage 98.9 K 2019-02-08 - 17:57 TaoHuang  
PDFpdf OTMB_Firmware_flowchart_yuriy2016.pdf r1 manage 553.8 K 2019-02-08 - 13:30 TaoHuang  
PDFpdf OTMB_Mez_PRRslides.pdf r1 manage 9593.2 K 2019-02-08 - 13:30 TaoHuang  
PDFpdf cms_upg_o_g_b_ni_gem_re1_me0_grid_160229.pdf r1 manage 56.1 K 2019-02-08 - 17:10 TaoHuang  
PDFpdf tmb2005_spec_v5p07.pdf r1 manage 3333.1 K 2019-01-15 - 23:06 TaoHuang  
Edit | Attach | Watch | Print version | History: r6 < r5 < r4 < r3 < r2 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r6 - 2019-02-08 - TaoHuang
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Main All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback