Dec 06, 2010 Calibration/Monitoring

From GlueXWiki
Revision as of 18:47, 7 December 2010 by Zisis (Talk | contribs) (Minutes)

Jump to: navigation, search

Teleconference Time: 09:10 CST (Regina) / 10:10 EST / 17:10 GMT+2 (Athens)

  • ESNET preferred, EVO as a backup

Background information

Tentative Agenda

  1. Go over goals/objectives of Monitoring system and review past discussions.
  2. Discuss LED-fibre-light guide concept.
  3. Summarize info, prepare recommendations and discuss it in Working Group to reach conclusion and proceed with Athens Construction MOU.


Attendees: Christina (Athens), Elton (JLab), Zisis (Regina)

  1. System Goals: (see Item 2 October 26 and Item 2 November 23 -- backup info in Nov 30 minutes)
    1. SiPM dead/alive (or on/off) check: Agreed that it is needed but by itself not enough to justify a system.
    2. Time check/offset: Agreed to have and together with previous point starts supporting the use of a system.
    3. Relative gain check: Agreed that this is a main/important goal. Degradation should be monitored continuously (on the scale of hours).
    4. Linearity check: There is no conclusion on this issue and it should be discussed further. Such a test should be done on the bench when evaluating the performance of the SiPMs prior to installation. We need to consider the number of pixels that would need to fail and likely scenarios to see whether such a check is needed. Considering the dynamic range (400) even a few LEDs beaming together would only test a small part of this range.
    5. Temperature stability: Agreed that this is most likely not an issue. Athens LEDs at max light (stable region) show 0.5%/ oC variation. Hall D will have ~1 oC stability at any given point, over time. Inside the solenoid we are planning for ±2 oC, so a 4 oC range for the stability, after thermal equilibrium is reached. SiPM boards will have temp sensors and are only ~10cm away from LED board (when mounted on light guide). SiPM feedback system on temp aims at better than 0.1 oC stability. Question: do we want to monitor the SiPM temp feedback system's functioning? This is an issue but perhaps does not need to be monitored additionally.
    6. Separate flashing of tower rungs: SiPMs will be summed in towers of 3 or 4. Each SiPM in a tower can be flashed separately using some multiplexer system. This can be achieved by both LED+board and LED+fibre systems.
    7. Flash both ends of a module with each LED: This is desired, but has not been shown to work in Athens and Regina tests. Dynamic range is too much for sensors to handle; in PMT tests the PMT HV had to be lowered for the near end LED. Ideas that have been floated are i) use collimator to direct light mostly towards far end, b) use filters or micro-lenses, c) use two LEDs per light guide, one UV (for far end via fibre stimulation) and one blue (for near end stimulation).
  2. Other points
    1. George V. measured light transmission with 1.5mm air gap versus Si cookie. Cookie gave 15% more light. Air gap cuts 40% light from optical contact with grease. Elton stated that this loss has been folded (and increased to 50%) in the calculations. The introduction of a cookie would have to be motivated by string arguments since it would introduce mechanical and thermal conduction issues, and it is becoming late in the game to make such a change.
    2. Use of fibres to transport light. Elton envisages a system with 2-3 boxes containing 3-4 LEDs each plus a temp sensor per each side of each module, and thin (100μm. inexpensive) fibres emanating and fixed (glued) to light guides, coming in with a gentle curvature. Fibres would be mounted on the bench. Christina and Zisis clarified that mechanical issues is the main drawback of this system and fiscal a secondary one. Issues are: fragility, difficulty in replacement, control light losses due to bend radius which is different for each light guide, and others.
    3. George V. will be asked to verify LED dynamic range and stability at low bias, in order to evaluate whether we can have two settings (low/hi) for the LED bias to establish two points for linearity.
    4. Zisis will begin working on a system characteristic/parameter table.
    5. This discussion is mature enough to move to the Thursday December 9 CALWG meeting.
    6. Following the meeting Elton ciculated an email with links to the pertinent information and calculations.