Status Update for the Photodetector Test Facilities

1
Status Update for thePhotodetector Test
Facilities

• Status Update for the PDTF
• Aims and Strategy
• Test Program
• Infrastructure
• Software
• New Developments since PRR
• PDTF Meeting 26.01.
• Critical Issues
• Answers to Referees
• Quality Assurance
• Timing of delivery batches
• Feed back time on delivered batches
• Conclusions

substantial overlap with HPD PRR presentation
Stephan Eisenhardt University of Edinburgh
RICH meeting, CERN, 04.02.2004
2
Aims and Strategy

• verify specs of 550 HPD tubes and characterise
  them for selection over 1½ years
• delivery batches of 30 HPD per month
• 8 weeks from delivery to accept/reject tubes
• Photo Detector Test Facilities (PDTF)
• two sites Glasgow Edinburgh
• test one HPD / work day / site (i.e. 40
  HPDs/month possible)
• two test stations per site (i.e. 4 DAQ chains in
  total)
• one test station/site for mandatory tests to
  accept/reject tubes
• second test station/site for extra
  characterisation and redundancy
• automation mandatory

3
Test Program

• Mandatory tests (one test station / site)
• mechanics (sizes and tolerances)
• pixel chip register check DAC linearity scan
• threshold working point, scans, TP noise
  threshold maps
• DAQ timing
• LED test dead pixel map, sensor position,
  demagnification
• HV scan gain curve
• Bias V scan depletion IV-curve
• dark count rate ion feed back rate

4
2nd Test Station

• Extra characterisation (on subset of HPDs)
• Photoelectron efficiency test with backpulse
  measurement
• Quantum Efficiency test with Sr90 source
• electron optics distortion map
• Additional studies (on subset of HPDs)
• swap tubes between sites and re-test for
  cross-check
• re-measure for long-term stability
• Redundancy
• immediate spare if first test station fails
• parallel mandatory tests, if we have to catch up
  with schedule

5
Infrastructure

• 4 sets of mech. setups finished
• 2 setups to be shipped to Glasgow next week
• arrangement fits on one laboratory table
  (excluding DAQ PC)
• usable for prototype and production HPDs
• additional cabling in production
• available in Feb 05
• electronics HPD training sessions at CERN
• attended by Edinburgh Glasgow physicists
• test programme will be shared between 3-4
  people/site
• physicists technicians
• HPD storage
• in DEP boxes plastic bags under N2 atmosphere
• shipment
• from DEP to Edinburgh, Glasgow CERN possible

updated info
6
Software

• DAQ PC setup complete (4PCs)
• Win XP, service applications, HW drivers,
  Labview installed and tested
• PDTF software package
• adoption of PTS software to our setup done
• PTS software runs now also with parallel port
  JTAG interface done
• stand-alone steering/reading for done
• Matsusada HV supply
• PT100 temperature sensors
• LED delay
• bias voltage supply
• pA meter
• integration under top level VI work in progress
• test programme separated into tasks
• automated logging of tasks, steering parameters
  results
• automated task flow with flexibility to divert
• self-restoring monitors (HV, Bias V, T)

updated info
7
New Developments since PRR

• Operation of prototype HPDs at Edinburgh (thanks
  to Thierry!)
• 2 prototype HPDs 2 HV dividers received
• recorded photons _at_ 20kV, up to 80V bias voltage
• performed threshold scans
• found fix for shift in delay between TP strobe
• Setup of LED light source
• single LED multimode light fibre or LEDs inside
  dark box
• synchronised light emission to DAQ
• PC steered scan over strobe to HPD with
  calibrated delay

8
Photo Gallery Jan05
HV interlock
light sources
HV supply
HV distribution
HPD
I/O board
independent PMT
HV interlock
sealed divider chain
LED
HV terminal
fibre mount
prototype HPD
9
PDTF Meeting 26.01.

• at Edinburgh to discuss status, organisation and
  work plan
• answers to referees
• Database
• store summary data of each HPD test at PDTF in
  central database
• include DEP data in summary data
• raw PDTF data stored locally for possible
  reanalysis
• backup of local data by disk mirrors
• Distribution of software development
• Glasgow
• mechanism and SW implementation for PDTF access
  to database
• (Andrew is link to Sajan)
• integration of external analysis SW (cygwin,
  root, C based code) to Labview
• Edinburgh
• top-level VI for test station monitors, data
  logging, bookkeeping
• framework and flow control for test tasks
• data structures for storing raw and summary data
  backup

10
Critical Issues

• FLIC Cards unreliable
• one FLIC card from Edinburgh (no.31) still gives
  the odd unsynchronised data record ? unreliable
  data
• ? under investigation at CERN today
• even if this is fixed we only will have 4 FLIC
  cards
• already one FLIC from Glasgow died and was
  replaced by a repaired FLIC in Dec04
• we need at least one, better two additional
  (spare) FLIC cards!
• potential to jeopardise the whole project
• HPD availability
• need 4 preproduction HPDs in March (2 by end of
  Feb05), to have 2 HPDs at each PDTF site in
  March05
• needed for proper commissioning and
  cross-calibration (PRR referees remarks)
• conflict with spring-test ? argue for priority
  to PDTF!

11
Answers to Referees I

• Organisation
• will have a Quality Assurance Manager (Franz
  M.) responsible for
• quality assurance according to schedule, storing
  of QA parameters in central database
• tracing HPDs from arrival at PDTF sites to
  delivery at CERN
• efficient communication between PDTF team, HPD
  and RICH project leaders and DEP, e.g. regular
  PDTF meetings, fast reporting of problems
• synchronising test protocols between test sites
• will have two Site Managers (Andrew P. and
  Stephan E.) responsible for
• day-to-day work at each PDTF site
• training of technical staff
• maintaining test protocols
• reporting to QA manager, close contact between
  sites
• flagging problems / bottlenecks
• possible in future one person can be QA and Site
  Manager
• plan to visit DEP together with Thierry in March
  05 to establish links

12
Answers to Referees II

• Timing of delivery batches
• DEP is interested to make multiple shipments per
  month
• to get paid as soon as possible
• get early feed-back on quality
• we also are interested in
• get early feed-back on quality
• minimum administrative overhead in shipping
  bookkeeping
• direct shipments from DEP to Glasgow and
  Edinburgh
• proposed solution (to be discussed with DEP)
• 2 shipments per month from DEP with 15 HPDs each
• alternating between Edinburgh and Glasgow
• (e.g. middle of month to Edi and end of the
  month to Gla)
• this is the optimum for feed-back time and
  bookkeeping
• we disfavour weekly shipments from DEP
• Delivery schedules in holiday seasons (summer
  Christmas)
• to be discussed with DEP
• arrange for cover at PDTF sites

13
Answers to Referees III

• Feed back time on delivered batches
• proper test of HPD needs one full work day
• preparation, mounting, first (slow) HV ramp to
  20kV and functionality tests need at least 0.5
  work days
• testing 2 HPDs/day would introduce a 50 overhead
  in time,
• ? unaffordable! (a month does not have 6 weeks!)
• would force to unmount a HPD just when it gets
  interesting
• increases the risk of damage due to hurry and
  additional handling
• conclusion stick with 1 HPD/day test scheme
• have a tube fully tested at the end of each day
• three categories good, to be checked again (on
  second station), failed
• this ensures QA feed-back in reasonable response
  time
• 5 out of 15 HPD tested within one week of
  delivery
• sub-sample sufficient to reveal major production
  problems

14
Conclusions

• setup of photodetector test facilities at
  Edinburgh and Glasgow is well underway
• 1 DAQ setup is operational successful HPD
  operation
• 4 setups by end Feb05
• stand-alone software for all components is
  working
• further software integration for automated test
  programme underway, to be finished in Mar/05
• critical issues
• unreliable FLIC cards
• HPD availability conflict with spring-test
• answers to PRR referees
• Quality Assurance
• Timing of delivery batches
• Feed-back time on delivery batches

15
Spare Slides
16
DAQ System

• based on CERN HPD DAQ system
• same system as used for chip testing
• hardware status
• 4 full DAQ chains
• 2 spares for custom electronics
• only 4 FLIC cards (bottle neck)
• all tested OK, but problems with FLIC
• 2 of 4 HV dividers/terminals 2 prototype HPDs
• 4 pre-series HPDs by March05 needed (conflict
  with spring-test)
• improved automation
• no more hand-steered/recorded parameters
• Silicon Bias Voltage steering
• Silicon Bias current read back
• LED delay setting wrt. DAQ clock
• Temperature monitors

17
Photo Gallery Dec04
LED trigger
bias pick-up
fibre
HV supply
LED pulse
HV divider
pA meter
I/O
bias supply
MB/FB supply
dark box
HPD
PP JTAG
flat pointing light source
to ZIF board
MB FB
PCI bus