GLAST Large Area Telescope

1
GLAST Large Area Telescope Engineering
Meeting 23 November 2004 Charge Injection
Calibration James Swain Stanford Linear
Accelerator Center jswain_at_slac.stanford.edu 650-9
26-2721
2
Goal

• To gather enough information to produce
• DAC -gt ADC curves for each discriminator in the
  front ends
• Measurements of cross-talk between front end
  channels
• Time over threshold curves
• Need subsystem input
• How should we collect TOT data? One channel at a
  time?
• What pulse patterns do we need to make available?
• FSW is proposing providing two pulse patterns,
  each of which can optionally be masked

3
Operation

• Configure
• Use LATC to perform initial configuration.
• Perform reconfiguration (DACs, thresholds etc)
  within calibration loops as determined by LCI
  configuration files.
• Pulse
• Number of times determined by LCI configuration
  files
• Read
• Accumulate
• Package
• Preserve as much of the raw data as possible.
• Send
• Dont Analyse
• Dont Apply

4
Calibration Styles
Subsystem Calibrations Bit scan ADC scan
ACD PHA (ADCs) CNO Accepts Hits Pedestals X X X X X
CAL ADCs Low Energy Trigger High Energy Trigger Accepts Pedestals X X X X X
TKR Accepts Time over Threshold X X
5
Calibration Styles (contd)

• Bit scan
• For charge injection DAC setting
• For each of threshold setting
• Pulse N times and record the number of accepts.
• Simple version just uses a fixed number of DAC
  settings
• Intelligent version focuses around the
  interesting threshold region.
• ADC scan
• Fixed number of injection DAC settings
• Record ADC response to injection N times.

6
Volume and Time Estimates
Subsystem Projected Time Uncompressed Volume Compressed Volume
CAL 10 s 420 Mbits 150 Mbits
TKR 640 s 8 440 Mbits 420 Mbits
ACD 10 s 30 Mbits 10 Mbits

• Assumptions
• Pulsing at 10 kHz
• Pulsing 1 channel per TFE, 12 TFE per TRC
• Cross-talk measurements will take longer
• Auto-range readout for the CAL
• Intend to use auto-ranging for most of the DAC
  values
• Use four-range readout for overlap regions and
  keep two ranges so the overlap regions will
  increase data volume

7
FSW needs

• What results would FSW like to be uploaded after
  calibration?
• ACD pedestals
• Might be nice, but not necessary
• CAL linear correction
• Two values per range per log end required.