LAT FSW System Checkout TRR

1
GLAST Large Area Telescope LAT Pre-Shipment
Review LAT Test Results J. Eric
Grove NRL Stanford Linear Accelerator Center
2
Purpose / Contents

• Demonstrate LAT functionality
• Verify performance requirements
• Note majority of sell-off done during/after
  environmental test
• Monitor LAT performance via CPT and calibrations
  throughout Commissioning and Environmental test
• CPT
• Detector subsystems
• Copper paths, interfaces
• Calibrations
• Detector subsystems

3
Data products, analysis, and reports

• Online analysis Mobile Computing Rack
• Data converted from CCSDS to LSF to LDF
• Same data format used in LATTE era (i.e. prior to
  FSW)
• Analysis scripts and reports from LATTE
• Offline analysis SLAC computer farm
• Two reports produced for all runs, corresponding
  to data products
• Digi report (basic detector quantities)
• Trigger info Total rate Rates of trigger
  types Arrival times
• ACD Occupancy Threshold info Hit map
• CAL Occupancy
• TKR Number of strips and layers hit Time over
  threshold
• Reconstruction report (derived quantities)
• Number of tracks
• Reconstructed energies, directions, positions
• TKR-CAL alignment, track extrapolation to ACD
• Full GLAST community has access and tools for
  offline products

4
Test matrix
5
Script/Test Status

• Liens against LAT CPT
• Detector CPTs
• Data collected and analyzed trending is in
  process
• No known script issues in CAL, TKR CPTs
• ACD CPT
• AcdTciRegRange (front end gain and linearity
  test) shows systematically lower gain (5-10
  lower) than GSFC test. Likely cause change in
  ACD timing config.
• Full AcdVetoCal (30 min runtime) is in use
  should abbreviate it.
• LAT-22x (Science Ops Demo)
• Needs to exercise onboard event filter
• Needs additional analysis steps (data and trigger
  aliveness). Coding is in process.
• LAT-23x (GRB Handling)
• Test is being coded
• LAT-14x (SIIS test)
• SIIS brought up and checked out. Test case is
  being coded

6
Test status

• Data integrity issues
• Dropped word on 256-byte boundary within packet
• Rare loss of data word, likely caused by dropped
  read from the SDI board in the VSC.
• Fixed? To be verified
• Firmware upgrade and change in a configuration
  parameter in the SBC made on Mon 24 Apr.
• Dropped word at end of packet
• Last word is dropped when length of packet is
  some particular modulo value. Affects many LPA,
  LCI runs.
• Cause is VHDL bug in EBM output chip.
• Workaround in FSW B0-6-8 add one cell of
  padding if packet has this particular length
• Need B0-6-8 installed and operating on LAT before
  final CPTs
• Compression error in LCI
• Occasional error in generation of compression key
• Rare but repeatable (e.g. affects 1 of 64 configs
  within a single CAL script)
• Still under investigation

7
TDF performance

• TDF functionality
• Trigger system demonstration in CPT
• Trigger primitive aliveness
• Scheduler, Engine, and Messaging functionality
• Trigger and veto efficiency
• Collected with LATTE, analysis on-going
• Collected with FSW, analysis on-going
• Detector subsystems timed-in (calibration)
• TREQ aligned
• ACD veto and TKR time-aligned with CAL-LO
• TACK delay optimized
• ACD and CAL MIP peaks maximized
• TKR hit occupancy maximized

8
Trigger summary
9
ACD performance

• Subsystem performance
• Minimum-ionizing particle (MIP) peaks
• Veto rates
• Veto thresholds
• Electronics
• Pedestal centroids and widths
• Width is measure of noise
• Gain
• Calibration quantities
• MIP peaks
• Veto thresholds

MIP spectra (PHA)
10
ACD trending

• Pedestal trend
• Centroid and width difference
• Two epochs
• Apr 06 CPT
• GSFC preship
• Typical Veto threshold 100 bins
• Pedestal changes are small

• Veto occupancy
• Aliveness test for all Veto signals
• GEM Veto List
• Muons during LAT CPT show all channels are alive

11
CAL performance

• Subsystem performance
• Gain (MeV/bin)
• Linearity
• Electronic performance
• Pedestal centroids and widths
• Width is measure of noise
• Gains
• Front-end linearity
• Threshold DAC gains
• Calibration quantities
• Gain
• Linearity

Hit occupancy
12
CAL trending

• Time evolution of pedestal width
• Typical FWHM 4 bins (HEX8)
• Plot shows stability of 32 channels

• Front-end linearity
• Typical avg non-linearity 0.2
• Stable to 0.02 or better

13
TKR performance

• Subsystem performance
• Dead strip list
• Noisy strip list
• Electronics
• Noise occupancy
• Calibration quantities
• Dead strip list
• Noisy strip list
• TOT gain

Hit occupancy
14
TKR trending

• Disconnected channel trend
• Increase in disconnects lt 1.5 since delivery of
  TKR towers.

15
Calibration

• Calibration
• Datasets collected with cosmic muons and test
  charge injection
• Both under LATTE and FSW
• Collections with FSW are on-going
• Test cases LAT-07x, LAT-70x, and LAT-71x
• No known problems with scripts
• Analysis
• Combination of online (near real time) and
  offline processing
• Online processing of FSW data is complete
• Offline processing in progress
• Output is set of calibration constants for each
  detector subsystem
• Trending system for constants is under
  development
• Constants need review by subsystems

16
Sample events
Gamma ray converting in TKR
Cosmic muon traversing multiple towers
17
Summary

• LAT assembled and operating under FSW
• LAT CPT in place
• Except GRB Handling
• Liens Minor mods to speed acquisition or add
  analyses
• Calibrations
• Data collected with LATTE are analyzed and in use
• Data collected with FSW are under analysis
• Calibration constants maintained in SAS/ISOC
  database
• Performance trending
• Trending processes exist, and subsystems are
  involved
• Large data volume to consume
• Status
• Detectors are operating daily with flight
  settings
• Test results with FSW show (approximately) same
  instrument that we assembled and tested with
  LATTE
• Leave SLAC with a baseline reference for entry
  into environmental test program