• Need Joe to help setup to use subreflector during maintenance. Normal setups will fail. Need Antenna in scan coordinator. For Antenna manager: doscan=1, make sure in gregorian model, xyz click, CCU in monitor mode, SCU in NRAO control and enabled, Subreflector drives enabled, Az and El drive disabled.
• We verified software that VEGAS is passing the integration time properly such that the system knows the proper switching period when using nodUnit='integrations'. Note for tinit=1.5 sec, the Spectrometer is slightly less than 1.5sec and VEGAS is slightly longer than 1.5s.
• For long integration times (1.5s), the Spectrometer has more time recorded for the beam positions than VEGAS, but VEGAS scans are more symmetric.
• Need to run VEGAS with short integration times to improve observing efficiency (e.g., tinit=0.2) for SubBeamNod
• Comparison between Spectrometer and VEGAS using standard spectrometer parameters of tint=1.5 and nodlength = 5 integrations -- for 60sec scan, 40 integrations. Spectrometer has more time at position 1 than position -1, while VEGAS has ~equal integration time per beam position. For spectrometer scans, 2 integrations are flagged during move from pos1 to pos-1, while 1 integration is flaggged during move from pos-1 to pos1. For VEGAS 2 integrations are flagged for movement in both directions. In this example, the Spectometer observed onsky (at pos1 or pos-1) for 72.5% of the time, while VEGAS only observed for 65% of the scan time.
  

  Subreflector Position (0=moving) (1,-1 are the two beam positions)	Spectrometer Number of Integrations	VEGAS Number of Integrations
  1	5	4
  0	2	2
  -1	3	3
  0	1	2
  1	4	3
  0	2	2
  -1	3	3
  0	1	2
  1	4	3
  0	2	2
  -1	3	3
  0	1	2
  1	4	3
  0	2	2
  -1	3	3

• Comparions of observing efficiency for VEGAS SubBeamNod as function of integration time for fixed nodlength~10sec.
• scan 32 tint=0.2 sec nodlength=50 integrations (10sec)

• Pos1 43.7%
• Pos-1 41.6%
• Pos-0(moving) 12.8%
• Blanked data 1.9%

• scan 35 tinit=0.5sec nodlength=20 (10sec)

• • Pos1 40.9%
  • Pos-1 39.7%
  • Pos-0(moving) 18.2%
  • Blanked data 1.25%

• scan 38 tint=1sec nodlength=10 (10sec)

• • Pos1 39.8%
  • Pos-1 37.3%
  • Pos-0(moving) 21.6%
  • Blanked data 1.25%

• The effective observing efficiencies for VEGAS SubBeamNod (10sec nodding) is 85.3%, 80.6%, and 77.1% for tint=0.2,0.5, and 1sec respectively. I did not test for 0.1sec, since I would expect limited gains since the subreflector move times are >>0.1 sec. The move time and settling time is about 1.2 sec.