Commissioning TileCal with Cosmic Ray Muons

1
Commissioning TileCal with Cosmic Ray Muons
J. Pilcher University of Chicago
2
TileCal Assembly Schedule

• Surface preassembly of barrel
• Complete preassembly December 03
• Start disassembly late February 04
• 2 months possible for commissioning studies
• Cs studies over full height of barrel
• Cosmic ray muons through full readout system
• Underground assembly of barrel
• Begin 11 May 04
• Module assembly complete 10 Sept. 04
• Services complete 26 Nov. 04
• Tests and commissioning 29 Nov. 04 - 27 Dec. 04
• Short time (4 weeks)
• Unpopular time for working
• We must be efficient and well prepared

3
Commissioning Preassembly

• Goals
• System tests with more modules
• Largest system run to date is 6 superdrawers
• In test beam
• Full barrel contains 128 superdrawers
• Intermediate step is essential
• System tests with hardware not yet used
• LV power supply sets in fingers
• Bulk LV power source 80m away
• LVL1 trigger hardware interfacing to TileCal
• Patch panels to remap tower and muon signals
• Receiver boards
• Hardware RODs

4
Commissioning in Pit

• Check that all systems are alive in Dec. 04
• Cs and CIS runs
• Thorough check-out with physics signals will
  take longer
• Attractive to run system on cosmics periodically
  from Dec. 04 - Dec. 06
• Monitor stability
• Get all bugs ironed out
• Provide trigger to other parts of ATLAS
• Likely very little access to pit during this time
• Most work in USA15

5
Estimate Trigger Rates

• Consider rate in back-to-back trigger towers (16
  16 equipped modules)

6
Estimate Trigger Rates

• These triggers are the most useful
• Muon fully traverses single towers
• Could also select these muons with bottom section
  of cal. and scintillator in beam region
• Rate is lower than trigger using larger towers
  (eg. 0.4 x 0.4)
• Return to this case later
• Do simple calculation with Excel

7
Cosmic Ray Flux

• Use numbers from Particle Data Group
• Tabulate in Excel

8
Cosmic Ray Flux

• Integrate over muon energy above a threshold
• Must impose range cut-off, depending on ? (or ?)
• Set up table for linear interpolation
• Gives muons / sec / m2 / steradian

9
Cosmic Ray Flux

• Range information
• Use PDG, Sec. 26, Ref. 1 for range-energy dE/dx
• Article by D.E. Groom et al., Nucl. Data Tables
  78, for muons in iron
• Set up table for linear interpolation
• Use module geometry from CDD drawings
• Solid steel at inner and outer radii (including
  girder, etc.)
• Steel and scintillator mixture between

10
Rate Calculation

• Consider back-to-back cells
• Area from plane mid-way through A-cell on bottom
• Solid angle from corresponding plane on top
• Scale rate by cos2? (zenith angle)
• Integrate over wedges in ?

11
Rate Variation with ? and ?

• Plot fraction of total rate versus ? and ?
• Easy to scale rate to smaller configurations

12
Rate Estimate in UX15 Pit
13
Rate Estimate in UX15 Pit

• Overburden corresponds to 55 m of rock with
  density 2.4 gm/cm3
• Range cut-off now 34 GeV instead of 4 GeV
• Based on CERN Yellow Report 71-18
• SPS study of range-energy and dE/dx of high
  energy muons
• Expected rate reduction is factor of 22
• Zenith angle dependence of overburden not
  included
• Would give 10 reduction
• Vertical access shafts not included
• Would increase rates for certain ? and ?
• Expected rate 90/hr
• Rob McPherson has been asked for a more detailed
  GEANT-based estimate

14
Rate with Alternative Trigger

• Consider trigger on towers of (0.4 x 0.4)
• Same area as before but solid angle increased by
  16
• Rate increased by 16
• 32K events/hr on surface (was 1.9K/hr)
• 1.5K events/hr in pit (was 0.09K/hr)
• Energy spread over several towers
• Trigger less clean
• Events less useful for calorimeter diagnostics
• Much less trigger hardware needed
• Could filter events at LVL2

15
Comments on Rates

• There is a very useful trigger rate both on the
  surface and in the pit
• Actual rate on surface next winter will depend on
  how much electronics is available

16
Hardware Needed

• Electronics drawers (64) - probably OK
• LV power in fingers
• Bulk LV power (200V for USA15)
• Cabling and fibers
• TTC hardware
• LVL1 trigger interface hardware
• Patch panels to separate tower and muon signals
• Receiver boards (64 towers each)
• Trigger logic for cosmic ray running
• LVL2 hardware
• ROD modules
• ROD crate and controller
• Output hardware from ROD crate (Ethernet?)

17
LV System
18
LV System

• Need full scale prototype for 2 fingers with 200V
  bulk supply, cables, splitter boxes
• This summer?
• Need preproduction units
• Use in preassembly commissioning next winter
• 64 of 256 fingers?
• 32 of 256? ? looks reasonable
• Feasibility????

19
LVL1 Interface Hardware

• Cable from barrel drawer delivers 9 tower signals
  and 7 muon signals
• Splitter box needed to collect tower and muon
  signals on separate cables
• Group tower signals as needed for LVL1 receiver
  boards
• Whose responsibility???
• Drawing shows required organization for 1616
  drawers

20
LVL1 Interface Hardware

• Receiver boards produced a single-ended,
  conditioned, analog signal for each tower
• Will go to flash ADC boards in ATLAS
• Could be used for cosmic trigger instead
• Trigger on towers of 0.4 x 0.4 for 88 modules
  requires the following
• add analog signals in groups of 16 towers
  (0.1x0.1 to 0.4x0.4)
• Require energy threshold on sum (discriminate)
• Take 8 2-fold coincidences between back-to-back
  towers
• Take OR of the 8 coincidence outputs

21
LVL2 Hardware

• Data from drawer goes to ROD modules
• 1 module handles 8 drawers
• 88 modules ? 32 drawers or 4 RODs
• Needs ROD crate and ROD Crate Controller
• Might log data from RCC over Ethernet
• Done for other systems in test beam
• ROD schedule?
• 2 prototype modules planned for September
• Can we get 4?
• Production not until next spring

22
Conclusions

• Commissioning with muons is attractive and
  necessary
• The trigger rates look fine
• Try this week to agree on target configuration
  for next winter
• Confirm responsibilities and explore timetable
  with TDAQ groups
• Over next few weeks make more precise list of
  items needed and identify items requiring more
  help
• People willing to help might meet in July to
  agree on division of tasks (phone video
  face-to-face)