L2 CAL Status

1
L2 CAL Status
Vadim Rusu
For the magnificent L2CAL team
2
The goal

• Improve the L2 MET resolution so that we can run
  at lower MET at L2
• L1 only uses 8/10 bits
• No L2 calculation in the current system
• Change the jet clustering so that we can control
  the rates at high luminosity
• Current L2 clustering pacman style

MET2 JETS
JET40
3
The design
Muon/XTRP/L1/SVT/XCES
Global L2 Decision

• Add new hardware to send the towers straight to
  the L2 PC
• Get new PC for better performance (about time
  anyway)
• Everything to be commissioned parasitically

8-bits
10-bits Energy/Tower
Calorimeter
4
The project

• An international mini-collaboration
• Fermilab
• Chicago
• Madrid
• Padova
• Penn
• Pisa
• Purdue

• Started Oct 2006
• Completed - ASAP

5
The know-how

• Use PULSARs as receivers for the LVDS data from
  DIRAC
• Need to interface the 40 bit cables with the
  PULSAR
• Build Mezzanine card
• Serialize the data in Slink format
• Receive on PC through Slink-PCI boards

6 slink outputs
96 LVDS cables
80 MhZ
96 LVDS cables
Slink merger Pulsars (8)
32-bit _at_ 40 MhZ
6 slink outputs
32-bit _at_ 40 Mhz
80 MhZ
80 MhZ
40-bit word/cable _at_ CDF_clk (132 ns)
6 slink outputs
Clustering timing 10 us
80 MhZ
Data transfer latency on average within 10 us
6
The tool

• L2 CAL mezzanine card

10bits x4 40 _at_ cdfclk (132ns)
input data per cable
Pulsar side
LVDS Connector
40
JTAG
160
CMC Connectors to the pulsar

Pulsar front panel
7
The tool

• Built at UC in record time
• Thanks Henry and e-shop
• Tested and commissioned revision 1.0 over the
  holidays
• Worked out of the box

Ted makes good for the bet he lost.
8
The pieces

• Mezzanine design
• Laura
• Testing/Checkout
• Lucas, Miguel
• Firmware
• Marco, Laura
• FER code
• Massimo, Laura
• Simulation package
• Simone D., Gene F.
• Monitoring
• Giorgio Simone P., Michael Wojtek/Carla
  consulting
• PC infrastructure and algo
• Gene Kristian Hahn consultant
• PC interface
• Anadi
• Preparing the celebrations (parties, happy hours,
  etc)
• myself

9
The status

• Boards arrive over the month of February
• Mezzanine, Slink, AUX (Henry UC)
• Front panels and strain relief (Richard UC)
• Pulsars (Paola Pisa)
• Cables 80 in hand
• Cable map (thanks JDL)
• Software final testings
• Firmware stage 1 ready to go

10
The plan

• Start splitting the cables one DCAS crate at a
  time starting next week
• Have half the DCAS system split and parasitically
  running by the end of February
• March have all system running parasitically
• April new trigger tables
• May Project completed
• 10 bit MET resolution
• New Jet algorithm (cone based)

11
The future

• The L2 CAL trigger will be a very versatile
  system
• What the new L2CAL will provide
• 10 bit MET at L2
• Better L2 jet position
• L2 jet closer to offline jets
• What else?
• Angular correlations (jets, MET)
• Track jet matching
• Dijet mass
• New paradigm Your offline analysis code can be
  implemented in the trigger
• Use the L2 CAL simulation to do studies that will
  improve the trigger selection
• Already starting on this in the context of HTTF

12
THE END