Go to the Indico Home Page Oldest event Older event Up to category Newer event Newest event Printable version
  • Compact style
  • Indico style
  • Indico style - inline minutes
  • Indico style - numbered
  • Indico style - numbered + minutes
  • Indico Weeks View
    • Switch to the management area of this event
    Login

    VFC Test Strategy Definition

    → Europe/Zurich
    864/2-B14 - SALLE J.B.ADAMS (CERN)

    864/2-B14 - SALLE J.B.ADAMS

    CERN

    25
    Show room on map
    20180201_Meeting_VFCTestStrategy.pdf
    20180201_Meeting_VFCTestStrategy.pptx
    Hide

    Minutes 1st Meeting “VFC Test Strategy”

    Location: CERN Prevessin, 864-2-B14

    Date/Time: 01.02.2018, 9:30 – 11:45 (2.25 hours)

    Participants: Mathieu Saccani, Volker Schramm, William Viganò, Christos Zamantzas

    • Presentation by V. Schramm “Reliability Testing and Validation of the new VFC-HD for the LHC BLM”
    • Burn-In, slides 4,5:
      • W. Viganò mentioned doubts that 60°C might be too high for burn-in. Maybe 50°C could be a better choice
      • C. Zamantzas added that burn-in should be performed for all 1000 produced VFCs, not only for the 700 used by BI-BL
      • M. Saccani asked about testing the FMC part and suggested to produce a test mezzanine. C. Zamantzas mentioned that commercial mezzanines do exist.
      • C. Zamantzas added that the FMC should be tested, that a test firmware for the FMC exists and that the WhiteRabbit interface cannot be tested
      • W. Viganò pointed out that before burn-in a validation should be performed to be sure that the tested systems are fully functional
      • C. Zamantzas added that a tracking of communication errors can be possible during burn-in to receive a BER bearing in mind the complexity. V. Schramm pointed out the test temperature of 60°C and therefore questioned the usability of the results.
      • C. Zamantzas concluded that the burn-in top priority is to qualify the system so it will perform better than the BLETC
      • To modify the burn-in strategy C. Zamantzas suggested that it could be possible to power 400 VFCs before going into operation in a dedicated crate for a year. It has to be assessed which temperature and testing time is advisable
    • Temperature/Humidity Cycling, slides 7,8:
      • C. Zamantzas suggested to test system parts separate in the climatic chamber, e.g. only the optical adapters inside. Also he suggested to set up a test to compare the reliability of E2000 and LC connectors.
      • C. Zamantzas suggested to add an attenuation outside the climatic chamber to simulate the 2km optical fibre cable from the tunnel to the surface
      • W. Viganò & C. Zamantzas suggested to ask Antonio Gonzalez (EN-EL-FC) which tests were performed to qualify the fibre link, especially if humidity testing was done
      • C. Zamantzas proposed to perform the test for boards equipped with at least 30 SFPs or more. The temperature cycles should also be performed in a pyramid shape. All agreed to extend the test duration and add temperature cycling at constant humidity as well as humidity cycling at decreasing temperatures
      • C. Zamantzas suggested to recheck the MTTF assessment of FTTx Technology for the FT3A05D SFP
      • W. Viganò pointed out the need for a sort of viewer to be able to constantly access the generated test data. C. Zamantzas suggested to use the expert application and save the data local. W. Viganò mentioned that the monitoring of the FMC and DDR memory has to be included. It has to be determined if the expert application provides all needed features. C. Zamantzas added that handling the generated data can also be challenging.
      • W. Viganò and C. Zamantzas suggested to order “button data loggers” and monitor the environment in a rack. In addition Eva Calvo did tests earlier. This data should be checked.
      • C. Zamantzas added that a less steeper slope to change temperatures should be used. The problem is the crystal which cannot handle fast changes and may influence the result
    • Complete test strategy:
      • C. Zamantzas corrected previous assumptions that for testing actually a fan tray module will be needed because it hosts the control part. Nevertheless it might be re-used after testing. What has to be sacrificed for testing will be the backplane and the PSU. A modified backplane will be needed for the P0 test. 30 additional are already ordered for different purposes.
      • About the requirement list C. Zamantzas added that more the amount of needed manpower/time instead of costs shall be pointed out. The possibility to employ a TECH from the University of Stuttgart has been discussed.
      • C. Zamantzas suggested that depending on demand additional tests can be performed.

     

    There are minutes attached to this event. Show them.
    The agenda of this meeting is empty