Overview of the Liquid Argon Front End electronics

1
Overview of the Liquid Argon Front End
electronics

• Esther Ferrer Ribas
• (SPP/DAPNIA-CEA)
• On behalf of the LArg/ATLAS collaboration

CALOR 2002
28th March 2002
2
Liquid Argon Calorimeters Barrel EM 110208
channels End Cap EM 63744 HEC 5888 FCAL
3584 In total 190 K channels
High energy pp collisions _at_ vs 14 TeV Large
luminosity 1033/ 1034 cm-2 s-1 Large dynamic
range 50 MeV 3 TeV
3
Structure/segmentation

• Three segmentations
• Read out channels summed in projective trigger
  towers
• Dh x Df 0.1x 0.1

4
Signal
tdrift 450 ns EM 50 ns FCAL
(smaller gap)
5
Requirements

• Dynamic Range 16 bits (50 MeV 3 TeV)
• Resolution sE/E 10 / vE ? 0.7 (EM calo.)
• ? calibration lt 0.25 whole energy range
• Coherent noise lt 5 of incoherent noise
• Inputs for Trigger Towers Dh x Df 0.1x 0.1
• Limited access ? high reliability
• Radiation levels 5 krad and 1012 N/cm2

6
Radiation tests

• L10 34cm-2 s-1 ? significant radiation
  environment
• Predominantely secondary particles
• 10 years of operations
• COTS (components of the shell) and ASICS
  (application specific integrated circuits) need
  to pass Radiation Tolerance Criteria (RTC) in
  order to be qualified.
• Rad-tolerant tech. for main micro- electronics ?
  DMILL
• Standard test methods
• NIEL (non ionizing energy loss) 1.6 1012
  N/cm2/10 yr
• (neutron irradiation) RTCNIEL 1-3.6
  1013 N/cm2
• TID (total ionizing dose) 5 krad/10 yr
  
• (x-ray irradiation) RTCTID
  50-330 krad
• SEE (single event effects) 7.7 1011 h/cm2/10 yr
  
• (proton irradiation)

7
Design
CALIBRATION Generate 0.2 precision pulses 116
boards _at_ 128 ch
CONTROLLER 40 MHz clock LVL 1 accept
signal Distribute info to control and
configure 116 boards
TOWER BUILDER Analog sommationo of 4 layers for
trigger towers Transmit analog signals to LVL 1
cavern 120 boards _at_ 32 ch
FRONT END Amplify, shape Sum cells within each
depth layer Store signals waiting for LVL 1
Digitise selected pulses, Transmit on optical
fibres 1524 boards _at_ 128 ch
8
(No Transcript)
9
(No Transcript)
10
Front End Crate
38 BOARDS/CRATE 28 FEB 4 CALIBRATION 2 TOWER
BUILDER 2 CONTROLLER 2 MONITORING
Leak free water cooling system
11
Front End Boards (FEB)
Input signals
Preamplifiers
shapers
SCA
ADC
Gain selection logic
SCA controller
SPAC
Output Optical link
For all calorimeters 1600 FEBs
12
FEB-Preamplifiers

• Amplify detector signals above noise level
• 16 bit
• High speed
• Low impedance
• Two types
• EM warm bipolar hybrids (production
  of more than 50)
• Two different input impedances 25/50 W
• en 0.4 nV/ vHz (50 W )
• HEC cold Ga As monolithics
• (production and tests completed
  yield 84 )
• Circuits tested up to 3 kGy and 1014 N

13
FEB-Shapers

• Limit the bandwidth to optimise S/N and match the
  40 MHz
• Bipolar CR-RC2 (to remove signal tail and limit
  bandwidth)
• Optimisation ?pic vs total noise 50 ns
• Tri-gain shapers (1, 10, 100) split dynamic range
  without degrading intrinsic detector resolution)
• Rad tolerant by design (bipolar, static MOS
  switches)
• Summing stage for the LVL1 chain
• Switches for hot cells
• Full production completed.

• Noise high gain850 µV
• medium390 µV
• low250 µV
• Integral non-linearity 0.5

14
Switched Capacitor Array (SCA)

• Samples _at_40Mhz the 3 gains of 4 calorimeter
  channels
• 16 analog channels with 144 cells
• 12 channels to store the signal
• 4 channels store a reference level
• subtraction of closest reference
  channel?reduction of noise
• Dynamic range needed gt 12.5 bits
• Store the 5 samples up to 100 µs
• Read _at_5MHz for ADC conversion
• Simultaneous Write and Read operations
• DMILL

• 50 Krad, 2 1013 Ncm-2
• Noise 300 ?V
• Fixed sequence noise lt 0.2mV
• Dynamic range 13.3 bits
• Cell/cell gain variation lt 0.02
• Drift lt 3 mV/ms
• Integral non linearity lt 0.1
• Starting production

15
Test Beam

• EM
• 1 pre-series module and 4 series modules 2
  barrel, 2 End Cap
• Final electronics tested with 6000 channels
• 50 FEB, 12 calib, 1 TBB tested
• Boards with all functionalities but not
  completely rad-hard
• HEC
• 12 pre-series module and 24 series modules
• 400 cold channels

16
Test Beam EM
Free gain performance

• Final density 128 channels/FEB
• Full dynamic range (50 MeV-3 TeV)
• Noise 140 MeV(high gain) and 240 MeV (medium
  gain) standard electron cluster
• Coherent noise 5-7
• Power dissipation 0.7 W/ch

• Medium gain predefined sample in time lt threshold
  ? high gain
• Threshold in medium gain is respected
• Dynamic range in high gain is well used

Test Beam results ? Dirks presentation
17
CALIBRATION BOARD

• Generates 0.1 precision
• calibration pulses
• 9000 components
• Sensitivity to various COTS
• Injected charge effort? develop op-amp
• Static low-offset op-amp (10 µV)
• DAC in rad-hard technology DMILL

Irradiated to 2kGy and 1014N OK
18
CALIBRATION BOARD
Uniformity 0.11 Crosstalk lt 0.1
19
TOWER BUILDER BOARD

• 32 trigger towers/board (120 boards)
• Correction poles to get all signals at tpeak 35
  ns
• Programmable delays to align pulses before
  summing
• Prediction of the components implies a good
  knowledge of the electronic chain and detector
• ? 3 ns, 5
• Noise 300 MeV saturation
• Irradiated 2 KGy and 3 10 13 N
• Boards validated for production

Analog summation of the different layers to form
LVL1 towers
Ouput cables (70m) Front End to LVL1 cavern
ordered crosstalk lt 0.2
20
CONTROLLER BOARD
Load, update, check different registers and
parameters for all different front end boards
(clock, LVL1, bunch crossing)
Time Trigger Control (TTC) 2 fibers for
reliability Serial control link SPAC
Special bus designed SPAC bus Read and
write lines doubled Speed 10 Mb/s
October 02 2 final boards
21
Optical link
Link to transfer FEBs data to the Read Out
Driver system

• Radiation resistant Gigabit ethernet speed
  digital optical link HP Glink chosen

• Effort on low error rate system
• Front end crate test autumn 02 Glink ready
• Extensive radiation studies 3 kGy, 5 1013N
• SEE 0.47 bit flip/link/hr
• ? 8 hrs dead time in 10 yrs LHC

22
Conclusion and Perspectives

• 6000 channels have been already used in test beam
• Full dynamic range
• Expected noise ? validity of the
  technical choices
• Calibration and
  specifications fulfilled
• System crate test with all final radiation-hard
  boards autumn 02
• Future
• Apr 03 production of 1650 FEB
• Oct 04 assembly complete
• Feb 05 installation complete