Oct 26, 2010 Calibration/Monitoring

Teleconference Time: 09:00 CST (Regina) /18:00 GMT+2 (Athens)

• Direct Polycom point-to-point connection

Items for followup from previous meeting(s)

1. Athens test results on Wico vs Trap light guides

Tentative Agenda

1. Followup items
2. Goals/objectives of BCAL calibration/monitoring system
3. Calibration methods
4. Discussion of light transport by fibres or direct LED mounting
5. Athens MOU
6. Other issues

Minutes

Attendees: Christina, George V., Pavlos (Athens), George L., Zisis (Regina)

1. Followup item; Wico vs Trap light guides
   1. GV's September measurements shows that within error the Wico and Trap light transmission is the same. Also, simple contact of LG to PMT vs optical grease in between resulted in a 40% loss of light for both Wico and Trap configurations. GV will repeat the measurements this time with a 1mm air gap in between LG and PMT. It was felt that there is no pressing need to mask the PMT's window to the SiPM areal size, but if time permits GV will have a look at it.
   2. The Athens and Regina groups feel that any sacrifice of light for mechanical expedience should be seriously considered as this would impact the reaction thresholds at low energies, particularly in light of the expected SiPM degradation with time due to neutron radiation exposure.
2. Goals/objectives of BCAL calibration/monitoring system
   • The Athens-Regina group supports the following objectives for the proposed system:
   1. Dead/alive quick look for SiPMs
   2. Timing check, towards constant timing offset monitoring. LED system has 0.25-0.5 ns play in electronic signal and pulses are similar in timing to those of a PMT (blue LED rise/recover time of 5ns/11ns and green has 8ns/15ns).
   3. Relative gain calibration
      • Flashing both SiPMs at either end has not been resolved. Tests in Athens and Regina show that the dynamic range is too large. This can be overcome by an on/off system for the LEDs, always using the far LED for illumination check.
      • The LEDs can be studied to extract their response to bias and use this to inject quantized levels of light to check linearity of SiPM.
3. Calibration methods
   • An absolute energy/gain calibration is beyond the scope of this system and can be extracted from a) cosmics in situ, b) pi0 and eta reconstruction (constrained by the invariant mass) and possibly by omega decay; charged partyicles (pions and protons) can also provide helpful information, as demonstrated by Andrei/Irina in their MC sims earlier this year.
4. Discussion of light transport by fibres or direct LED mounting: fibres have been suggested so that the LEDs (and fewer of them) can be away from the readout zone where temperature shifts would affect them.
   • The Regina and Athens groups discussed this issue and are not in favour of using fibres. The reasons are:
   1. The Athens LEDs from Kingsbright or Agilent have a light output shift of 0.5%/degree C (and not 2%/degree as other systems).
   2. The LEDs will be calibrated for their temperature response and a thermistor could be added on their board, read out and used in a feedback system.
   3. Fibre splitters were investigated by Athens in 2007. Prices then were $2000 for 16 units.
   4. Fibres are fragile mechanically and difficult to install in the tight space around the SiPMs.
   5. Coupling is an issue: the system needs to be designed for removal and replacement of SiPMs. Using SMA connectors is expensive.
   6. What is the radiation hardness of such fibres (probably several hundred micro diameter)?
   7. Will they be sheathed or naked in an optically-tight enclosure?
5. Athens MOU: on hold until a final decision is taken on whether to have LEDs directly mounted on LGs or use fibres for light transport.
6. Other issues
   1. (Small) RF pickup from an LED system with its LPM board, ribbon cables and connection to LPCM control board is not exected to cause problems. What needs to be verified is that this system will not cause pickup in the SiPMs and their boards.
   2. Noise from the magnet is usually low frequency and can be filtered out.