VEGAS Team Meeting

Attendees: Amanda, Anish, Paul, Joe, Marty, Richard, Mark, Randy, Dan, John, Hong

Agenda:

  1. DONE OGP calibration; status of ADC calibration in general
  2. DONE Latest issue: Usage of MMCM in low bandwidth mode
  3. DONE ADC calibration routines and documentation for DIBAS
  4. DONE Mode 10-19 (ghost frequencies issue)
  5. DONE Blanking of last switched integration
  6. DONE Status of “Manager writes to many sub-bands in 8 subband mode” problem (release and test)
  7. DONE What is content of August 'release'?
  8. DONE Observer documentation
  9. DONE Use of test time (including overnights)
  10. DONE Status of other “current issues”: https://safe.nrao.edu/wiki/bin/view/CICADA/CICADAGreenBankSpectrometerCurrentIssues

Meeting notes:

  1. OGP calibration; status of ADC calibration in general
    • Release new VEGAS bank manager. This manager now reloads the appropriate OGP for the ADCs everytime the mode is changed and reloads the INL for the ADCs on newlyReady.
      • commit 8a50fb7f491ba201449544aae379acf7029ed836
    • load the INLs on startup to the ADCs
      • Hong to investigate
      • Not an effect on August release
  2. Latest issue: Usage of MMCM in low bandwidth mode VEGAS, unfortunately, has some modes which clock the ADC at very low sample rates, which forces us to use low bandwidth mode
      • possible solutions proposed to date:
      1. decimate by two (use every other ADC sample)
      2. double the clock rate and switch ADC modes to use 2 inputs, tossing out the one which isn't actually connected
      3. for lower sample rates another ADC could be used. It's very difficult support a very wide range of sample rates.
      4. change the VEGAS spec and not allow low sampling rates - at least set a temporary lower limits until resolved
        • Since data rates are not selectable by the Observer, this means the affected modes are not usuable
  3. ADC calibration routines and documentation for DIBAS
    • ready to send the python code and it is documented. SHAO will need to duplicate setup.
    • Jason to do brief overall
    • Send 'git' address along with some changes that Ray is making to bof files
  4. Mode 10-19 (ghost frequencies issue)
    • testing a matrix to find commonalities but has not yielded success
    • a couple bof still has image
    • Could this be spectrum jitter? Modification to dump one spectra still had 'ghost' (mode 10) so not jitter-related
    • mode 12 worked in LBW8 mode, but anything higher has a memory issue - needs additional time to work on it
    • capture FPGA time-domain output data stream in small windows and investigate offline for indicators of the image
    • This remains Joe's top VEGAS priority with a break to release Blanking of last switched integration
  5. Blanking of last switched integration
    • Still in pre-release testing - finish
  6. Status of “Manager writes to many sub-bands in 8 subband mode” problem (release and test)
    • Should be fixed but not deployed
  7. Observer documentation
    • Amanda collecting information in coordination with Ron & others for interactions
  8. Use of test time (including overnights)
    • Unscheduled request(s)
      • about an hour of time to run the Q-band ripple test for the whole system (i.e., including everything from receiver on down). Don't need any engineers or telescope movement, so it could happen at night. Might want to avoid sunrise/sunset.[Kepley]
      • Release fits file changes [Marganian]
    • Tuesday
      • 08:00 – 16:30; this time is available for anyone who wants it
    • Wednesday
      • 8:00am – 1pm; again available
      • Data monitoring
      • Q-band ripple test for the whole system
      • Overnight- Mixed mode testing [Prestage]
    • Thursday
      • 8:00am – 1pm; again available
      • Could be used for MUSTANG Calibration
    • Monday
      • overnight Monday through Wednesday is available
  9. Status of other “current issues”: https://safe.nrao.edu/wiki/bin/view/CICADA/CICADAGreenBankSpectrometerCurrentIssues

Other

  1. Please make sure to properly code your timesheets for the current time period and going forward.
    • Software engineers should be using a GBT ops software CM&E account - Account 893252
    • Electronic engineers should be using a GBT ops electronics CM&E account - Account 432510
    • Commissioning activities should be getting charged to hardware support and testing - Account 432510
    • Scientists assisting with shared-risk commissioning should be charging an observer support account
      • 432540 for the most part. Accounts 142020 (GB-based staff) and 142010 (CV-based staff) for when you are helping with overarching observing support (not problem solving, VEGAS specific)

-- MartyBloss - 2014-06-17
Topic revision: r2 - 2014-06-17, MartyBloss
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding NRAO Public Wiki? Send feedback