Description:

• The GBT Spectrometer Spigot Card (or Spigot system) is a new pulsar-search backend for the 100-m Green Bank Telescope.
• It uses the extremely powerful and flexible Auto-correlation Spectrometer to perform auto-correlations of the data very rapidly, then dumps the data to a PC with a 2-TB RAID.
• The hardware for the Spigot system is:
  • The spectrometer
  • Two custom digital logic cards designed by NRAO engineer Ray Escoffier. These cards accumulate and sort the data, then pack/decimate it for output.
  • A PCD-60 data-acquisition card sitting in the PC
  • A PC (dual Pentium 2.0-GHz)
  • A 2-TB RAID (using 16 160-GB IDE disks)
• The data format is FITS

Status:

• We achieved first light pulses in November 2002
• Currently, we are developing the observing software and testing observing modes
• Full integration into the GBT observing plan is anticipated in 2004
• Before that, observing may be possible in "Expert" mode
• Testing results:
  • 2003-Aug:
    • Successfully observed a number of pulsars (isolated, binary, millisecond)
    • Observed at 1.4 GHz, 2.0 GHz, and 9.0 GHz
    • Observed a pulsar (B1929+10) simultaneously with the BCPM and measured the same period/TOAs with both
    • Timing analyses of 3 binary msps, and the TOAs do not show any systematic trends
    • A report (ps, pdf) detailed the testing
    • Plots for:
      • PSR J0218+4232 (pulse profile, TOAs; the TOAs show some deviation, but this is due to an uncertain ephemeris)
      • PSR J0613-0200 (pulse profile, TOAs)
      • PSR J1012+5307 (pulse profile, TOAs)
      • PSR B0329+54 (profile at 9.0 GHz)
      • PSR B1929+10 (TOAs for Spigot and BCPM)
      More plots and explanations are in the report.
  • 2003-Sep-15:
    • Modes 1-6 have all been checked out with an artificial noise source
    • Modes 41-46 have all been checked out with an artificial noise source
  • 2003-Sep-22:
    • Double-Nyquist sampling has been tested for modes 1-6. This gives (for these modes) 1024 channels across 400 MHz, instead of 800 MHz. A 200MHz filter is then placed on the data to prevent anti-aliasing, resulting in 512 channels across the 200 MHz of data. See bandpass: the same instrumental bandpass is shown in both panels, but the bottom panel has 0.78 MHz/channel, while the top has 0.39 MHz/channel.
  • 2003-Dec:
    • Got modes 15,16,55,56 to work. These are the first modes with Nlags != 1024
• Software status:
  • Scott Ransom's PRESTO analysis package works with Spigot data
  • Quick-look python scripts are now available

Specifications:

• Full mode table

Collaboration:

• The Spigot system is a collaboration between Caltech, Cornell, and NRAO
• The majority of the software and testing, as well as the original layout and information for these web pages and even the 'logo' is thanks to David Kaplan at the California Institute of Technology.

Links/Documentation:

• List of participants
• Schematic of GBT spectrometer (from J. Yamasaki, 2001-Feb-27)