TileCal Electronics

1
TileCal Electronics

• A Status Report
• J. Pilcher
• 17-Sept-1998

2
Outline

• A status report on the front-end and digitizing
  electronics
• Overview of requirements
• Development status
• July 98 test beam results
• Barrel module 0 equipped with 2 superdrawers
• 90 channels
• First system tests with in-drawer digitizers

3
REQUIREMENTS

• Process 10,000 PMT signals
• Located in 256 electronics drawers
• Up to 45 PMT/drawer
• Each module self-contained with own electronics

4
REQUIREMENTS

• Performance
• 16 bit dynamic range
• Up to 2 TeV in single cell
• Must see muons
• Calibration and monitoring
• Enhance muon ID
• Readout resolution should not degrade calorimeter
  energy resolution
• Calorimeter resolution gt 2
• Need readout resolution of a few percent in each
  cell
• Jet populates many channels
• averaging effects

5
REQUIREMENTS

• In situ calibration
• Gives readout conversion factor (pC/count)
• Measures linearity
• Calibrates source integrator
• Slow integrator for PMT current
• LVL1 trigger tower sums

6
ORGANIZATION

• 3-in-1 Card
• One per PMT
• Plugs into PMT anode
• Near-ideal current source
• Pulse shaped signals to digitizers
• Integrator for source calibration and monitoring
  min-bias current
• Gain switching
• Output gating
• Charge injection for electronics calibration
• LVL1 Trigger output
• Gated

7
ORGANIZATION

• Mother Boards
• set of 4 in tandem per drawer
• Services and control signals to 3-in-1
• Digitizer Boards
• set of 4 (or 8) per drawer
• Connections to drawer
• TTC fiber
• S-LINK fiber
• D.C. Power
• CANbus

8
3-in-1 Card Status

• Bigain pulse shaper
• 7-pole Bessel filter (purely passive)
• Exploit current source nature of PMT
• No noise, no power
• Very linear
• Clamping amplifiers and drivers
• Gain ratio 641 for dual 10-bit ADCs

9
3-in-1 Card Status

• Output pulse to digitizers
• Low Gain (1 GeV/mV)
• Full scale signal
• High Gain (16 MeV/mV)
• Muon signal

10
3-in-1 Card Status

• Linearity and calibration
• Residuals lt 1 count over full dynamic range

11
3-in-1 Card Status

• Source integrator
• Essential for Cs calibration and monitoring of
  calorimeter
• See preceding talk
• Cs calibration has short-term reproducibility of
  0.1
• Should be matched by electronics stability

12
Integrator Readout Status

• independent readout for each drawer
• ADC board CANbus
• Multiplexed to individual 3-in-1 cards

13
Source Integrator

• Stability better than 0.1 over 2 months
  (calibrator integrator)

14
Digitizer Status

• Partially equipped Barrel Module 0 (30 channels)
  in July 98
• First system test of in-drawer digitizers
• Two 10-bit 40 MSPS ADCs per channel
• High gain scale 0 - 16 GeV (16 MeV/count)
• Low gain scale 0 - 1000 GeV (1 GeV/count)
• Commercial components
• TTC input on optical fiber
• 40 MHz clock, LVL1 accept, digitizer control data

15
Digitizer Status

• Pipeline delay via custom ASIC
• Digital memory unit (DMU)
• Originally developed for PHENIX TEC
• Output via optical S-LINK
• Read with optical LDC/PMC, RIO processor

16
July 98 Test Beam Results

• Laser calibration
• Measure linearity and stability of PMT and
  electronics
• 3 PIN diodes to monitor laser

17
July 98 Test Beam Results

• Digitized signals

• More pedestal noise on high gain channel
• Digitizing clock not synchronized to beam

18
July 98 Test Beam Results

• Pedestal Noise

• Noise for high gain branch 1.1 counts
• Corresponds to 0.4 photoelectrons in PMT (17
  MeV)
• SPICE simulation predicts 1.2 counts
• Noise for low gain branch 0.5 counts
• SPICE simulation of 3-in-1 card predicts 0.3
  counts
• Digital noise lt 0.4 counts

19
July 98 Test Beam Results

• Muon response for the 3 sampling depths (?90?)
• Pedestal superimposed
• Using signal from empty events
• Width reflects energy algorithm as well as
  electronics
• 10 digitizations used for each measurement (not
  optimized)

• Muon signal well resolved from pedestal

20
July 98 Test Beam Results

• Electron response
• Most energy in a single cell
• Channel-to-channel intercalibration less
  important
• Calibration not yet available
• 50 GeV and 100 GeV electrons

21
July 98 Test Beam Results

• Use e- response to measure readout resolution

• Fit for a
• reflects readout resolution and energy algorithm

22
July 98 Test Beam Results

• Energy resolution gives readout resolution of 0.5
  counts/sample
• ADC quantization error ? noise ? ...
• Noise study gave 0.5 counts
• Well understood result
• Readout will not limit resolution of hadronic
  calorimeter

23
July 98 Test Beam Results

• Pion energy resolution from test beam
• Under analysis
• Needs cell-to-cell intercalibration
• Electron shower largely contained in single cell

24
Future Planning

• Radiation hardness tests this fall
• Design review this fall
• Electronics PRR spring 99
• 3-in-1 production to start spring 99
• Version 2 of digitizer to be demonstrated spring
  99
• Final electronics needed for module calibration
  in 00, 01, 02
• Finish production of electronics in 02
• Before start of installation

25
Conclusions

• TileCal electronics shows good performance
• Achieved required dynamic range with linear
  system
• Very low system noise
• Electron energy resolution used to measure
  readout resolution
• First successful tests of in-drawer digitizers
• No unexpected problems so far
• Still a lot of work to do!
• Expect to start production on schedule