Rainer Stamen, Norman Gee
15 March 2007
Analogue Pulse Shape and Timing – using calibration
• FADC DAC scan
– Find the DAC value needed to set the chosen FADC pedestal
• Pedestal Run
– Measure exact FADC pedestal and noise. DAC scan only sets pedestal
to 3-4 FADC counts
• Coarse Timing Scan
– Measure DAQ readout memory offset
• Fine Timing Scan
– Measure fine timing. May change coarse timing setting
• Analogue Pulse Shape
– Measure analogue pulse shapes
• Status: All now store results in COOL database; runs are ~automatic; but
the information is not yet used in module services to set up hardware.
15 March 2007 Rainer Stamen, Norman Gee - Calibration Status 2
FIR Filter and “Electronics ET Calibration”
• FIR Filter coefficients
– Compute the coefficients and “drop bits” settings for each trigger tower.
– Assume we are using matched filter. Aim for best resolution.
• Electronics ET Calibration
– Compute receiver and LUT values to reproduce simulated ET deposits
• All these results should go to database in a “Pulser” settings area.
• Status: Not started. Not hard. BUT requires combined running with
calorimeters which we have not developed – see foil 6.
15 March 2007 Rainer Stamen, Norman Gee - Calibration Status 3
Detector Performance and Beam Corrections
• Physics Pulse Delay
– Latency L1A to analogue pulse is not the same as calibration system
– Will need different software to find coarse & fine timing.
• Physics Pulse Shape
– Particle pulses not identical in shape to charge injection systems.
– Measure particle pulse shapes – not so easy as amplitudes are varying
and usually small, and can have pulses in adjacent timing bins.
– Then convert to FIR coefficients. Initially adjust ET calibration to keep
same area under pulse per GeV?
• Correction factors:
– dead material. One multiplicative factor per tower, from Monte Carlo?
– dead cells. One factor per tower, probably just eta and layer dependent
• Status: all of these not started.
15 March 2007 Rainer Stamen, Norman Gee - Calibration Status 4
• These can all be done with the same L1Calo run mechanisms, but need different
calorimeter conditions and/or different data analysis
• Connectivity – verify the routing of calorimeter towers to PPMs
• Crosstalk – check no significant communication between towers.
• Relative timing of raw cells – check that they are equal for a tower
• Pulse shapes from raw cells – check they are very similar
• Pulse amplitudes from raw cells – check injected energies contribute equally
• Match precision readout to tower readout – beginning of physics calibration
– Connectivity is started. Needs PPMs (coming very soon). Tilecal has
generated the patterns we need. More discussions needed with LAr.
– Other areas not studied. Needs extra calculations on data from the analogue
pulse shape runs, and a few similar types of run.
15 March 2007 Rainer Stamen, Norman Gee - Calibration Status 5
Interface with Calorimeters
• Most of the runs rely on calibration signals from calorimeters
– Electronic calibration (ET in trigger matching injected pulses)
– Also various data patterns are needed to verify crosstalk, signal addition
– Calorimeters are capable (willing!!) to generate patterns
• We also need access to the trigger cables…
• Need to work on combined runs with calorimeters
– Need extra hardware (LTPI) to share triggers, plus software setup &
tests, plus software to analyse and present large numbers of results and
• Status: No actual development yet. Useful discussions with Calo groups,
but no tests. Needs some effort which we are pretty short of.
15 March 2007 Rainer Stamen, Norman Gee - Calibration Status 6
• Automatic procedures to set up digital timing:
– PPM LVDS into CPM & JEM.
– CPM and JEM clock alignment with CAM; CMM & JEM FIO alignment.
– CPM-CMM; JEM-CMM; CMM-CMM; CPM, JEM & CMM G-Link
– Overall system timing alignment.
– Results to go to database.
– Module services to use values from database.
• Status: Procedures within single crates mostly exist, but results do not go
into database (except PPM). No CAM runs yet.
• Timing between crates (LVDS, CMM) and overall optimisation not yet done?
Multi-step runs not viable, so use kicker-style (CMM needs updating).
15 March 2007 Rainer Stamen, Norman Gee - Calibration Status 7
• Analogue pulser measurements just need more channels and cables
• Electronics ET calibration not started but not hard
• Beam pulses (timing, shape) and correction fro dead material need
• Detailed checks of analogue cabling and signal uniformity not done yet
– but can be done mostly with data stored by calibration pulser runs
• Interface with calorimeters coming – but slowly.
– Can be developed in parallel with collecting calib pulser data by hand.
• Digital Timing needs completing
15 March 2007 Rainer Stamen, Norman Gee - Calibration Status 8
15 March 2007 Rainer Stamen, Norman Gee - Calibration Status 9
Comment on Database handling
Trig menu Connectivity
Trigger Menu Connectivity
Module_sim View trends
Module Settings Calibration Results Analysis Results
Hardware Services Current Settings
IOV Dated Dated/Run No/IoV
ROD / VME Calibration
15 March 2007 Rainer Stamen, Norman Gee - Calibration Status 10