EMCal Jet Trigger Status

1
EMCal Jet Trigger Status

• Hans Muller/CERN
• Peter Jacobs/LBNL

2
Jet Trigger enhancement

• Compare central PbPb to reference measurements
• PbPb peripheral vary system size and shape
• pA cold nuclear matter effects
• pp (14 TeV) no nuclear effects, but different
  energy
• pp (5.5 TeV) ideal reference, but limited
  statistics

All reference systems are required for a complete
systematic study
Includes acceptance, efficiency, dead time,
energy resolution
3
Jet trigger in pp
pp in ALICE L 5x1030/cm2/s 200 kHz ALICE DAQ
limits recording to 200 Hz ? L1 rejection
2000-4000 for pp
Chris Anson
Sharp threshold at required L1 rejection
4
Jet yields one LHC year
Large gains due to jet trigger Large variation
in statistical reach for different reference
systems
5
Jet quenching measurementpPb reference

• With EMCal jet trigger improved jet
  reconstruction provides
• much greater ET reach
• systematically improved response

6
Jet quenching measurementPeripheral PbPb
reference
Without EMCal, significant quenching measurements
beyond 100 GeV are not possible
7
PHOS / EMCal readout and trigger
L1 (6 ms) 2x2 tower analog sum ? TRU ? 4x4 tower
peak finder EMCal 12 FEE/GTL bus ? 1232384
towers/TRU configure as 1624 towers ?Dh x
Df0.22 x 0.34
8
Cartoon of jet patch trigger
Jet patch trigger

Max aggregate input bandwidth 3 Gbit/s
latency 1-2 ms
lt 100 Mbit/s
100 bits/evt
13K towers 30 TRUs
9
First TRU card proto ( PHOS)
10
L0 timing simplest algorithm
Pack and ship every bunch crossing (75 bits/TRU?)
11
L1 timing simplest algorithm
Just the data transfer takes too long!
Simplest approach no bunch filter at L0
Lumi loss due to pileup ? need local L0?
? Need to filter at TRU1 based on global L0
12
Bunch Filter
Level-0 _at_ 1.2 us
9x

• Initially planned major design effort not
  required !
• -gt TMU / SMU cards from TRD can do the job
  lt-
• Major step forward for building JET proto within
  1 year

13
TMU card (Alice TRD detector , KIP Heidelberg )
14
Bunch filter implementation with TMU and SMU
1 optical Output 3.1 Gbit/s
Common LVDS I/O bus
Compact PCI bus 64 bit
SMU
TTC
TMU
Level-0 trigger from TTC _at_ 1200 ns after Bx
L0
FiFo
Transfer selection L0 at 1.2 us, max 200 kHz
Xilinx Virtex-IV
FPGA
Input 11 x 3.1Gbit/s
One SMU /DCS card for n TMU cards ( assume n3)
11 optical inputs ( 1/3rd of TRU) pre-selected
bunch packets from TRU
15
SMU concentrator cards with ALICE DCS ethernet
and TTC link ( LHC clock, bunch count, L0,L1)
16
TMU/SMU crate for EMCal JET
Need space for 2 Compact PCI crates
TRU
Jet Trigger
out
TTC
Ethernet
1 Gbit fiber output
LVDS communication bus
Water-cooled PS
Rocket I/O fiber connections
SYS
empty
99 Gbit fiber inputs from 33 TRU -gt 9 TMU
17
EMCal system overview
Time after Bx
BW
Comment
6000 ns
Level-1 yes/No
Output L1 40 MHz NRZ
V0 (10 Multiplicity bits)
Patch search of JET over V0-defined threshold
TRU
1 GTL branch
Level-1-JET
2000 ns
Input Max 1.35 Gbit/s
CTP
1 optical R/IO link ( 2 spares)
10 x TMU 3 x dcs 1 TRU
L0-triggered forwarding of selected packets
1750 ns
3
3
3 x 3 optical R/IO links
1 TMU Output Max 0.15 Gbit/s
1500 ns
dcs
L0-packets-gtout
L0 lt 200 kHz
TTC
LTC
1200 ns (lev-0)
level-0 selection
DCS, Ethernet
1050 ns
Packet synch -gt FiFo
1 TMU Input 11 x 3.1 Gbit/s
3 x 11
3 x 11
3 x 11
Level-0 3 x Level-1
33 TRU
Xmit 10-bunch Packets 750bit/250 ns 1 link32
TRU channels
33
3 x 3 x 11 optical R/IO links 3.1 Gbit/s
digital
Bunch sync / reset
Trigger-OR
800 ns
Level-0 (600 ns)
1 TRU output Max 9.3 Gbit/s
3
TRU
Bunch Thresh max 2 of 10
320 ns
TRU
3 x Level-1
3x
155 ns
analogue-gtdigital 12 bit
1 TRU Input 96 x 0.48Gbit/s 46 Gbit/s
12
DAQ HLT
12
1 TRU input channel analogue OR of 2x2 towers
22 x RCU 44 branches
22 DDL
RCU
RCU
12 x 8 Analogue Fast OR (100 ns)
150 ns
9 FEE/GTL branch
analogue
396 x FEE
100 ns quad shaper
FEE
FEE
FEE
Ne -gtCSP Steo Voltage
32
36 FEE/SM
4
g -gt Ne conversion
11 Super-Modules
1 Super-Module 12 x 24 Modules
12
g
24
Particle TOF
12672 x APD / CSP / Towers
collision
o
18
Jet trigger summary

• Estimated cost SFr 100K-150K
• Relaxed schedule relative to other project
  issues
• jet trigger effective only for full device and
  high lumi
• current plans commission late 08
• Technical solutions may be in hand now (thanks to
  TRD)
• Participating institutions CERN (H. Muller),
  LBNL, Grenoble, Nantes, ORNL,