Terminology

• pattern: this is a single position on,off with calon,off measurement (150 secs on, 150 sec off, 10sec cal on,off). There were typically 21 patterns per day. Normally a source had only 1 pattern per day.
  

Intro:

• rcvr:Cband: 4500 to about 5500
• telescope: on, off position swithcing (150 sec on) with cals
• backend: mock spectrometer:
• both groups used giving 14 frequency bands per integration.
• Of the 14 bands, some were 11 Mhz and others were 86 Mhz.
  
• the digital filtering and offsetting was used.

History/problems:

• 01-17nov10 : gortex cover comes loose from cband horn mouth. moisture get ins horn, causes tsys to go up
  
• 22feb11:  the wideband H2CO line did not line up with the narrow band  H2CO line. The narrow band is correct, the wide band sees the line  4.03 Mhz low.
• ran mocksp version of cima. had been using svnwork.. not sure if this was the problem
• ran calibration, then executed command file, but skipped the line with the load config..
  
• My guess is that this was the problem.
• The header info looks correct, but the data is off. Must be that some of 2nd IF synths were not set correctly.
• My guess is that synth 5-7 (b1,b3,b5) probably didn't get restored correctly after the calibration run.
• might have been a band flip problem with mocksp version since svnwork has some fixes in it.. but band flipped wouldn't have put the line where it was seen in the wideband.
  
• next day, ran with svnwork and didn't skip the load line in the command file and things looked ok.
• The plots for 22feb11 below have jumps across the band in Tsys.
  

The data:    

Plot each band of each source with baseline removed

• Each source has 14 plots.
  
• the narrow band plots are on 1 page followed by the wideband plots.
• A linear baseline has been removed from each band.
• 10nov10 (.ps) (.pdf)
• 11nov10 (.ps) (.pdf)
• 12nov10 (.ps) (.pdf)
• 13nov10 (.ps) (.pdf)
• 14nov10 (.ps) (.pdf)
• 22feb11 (.ps) (.pdf) trouble
  
• 23feb11 (.ps) (.pdf)

Plot all bands of a source together. No baseline removed

•   Each source had 1 plot covering the 1 Ghz
• The data was smoothed and decimated by 51 channels (so the plot file was maneagable.).
• 10nov10 (.ps) (.pdf)
• 11nov10 (.ps) (.pdf)
• 12nov10 (.ps) (.pdf)
• 13nov10 (.ps) (.pdf)
• 14nov10 (.ps) (.pdf)
• 22feb11 (.ps) (.pdf) trouble
• 23feb11 (.ps) (.pdf)

Tsys and cal values:

• az,za pos
• this shows the az, and za postion for each cal on vs pattern number
• The bottom two plots show how far the az and za moved between the calOn and the cal off (since it tracks the off source during the cal on,off
• cntsToK and Tsys
• Page 1: Attenuator settings
  
• this has the rf, if, and downstairs attenuator,gain settings for each pattern. Normally they would be contstant for the entire nite.
• Page 2,3: Conversion factor counts to kelvins for each of the 14 bands.
• Plotted a % of the median value. If the attenuator settings changed, then there will be a large jump in this value.
• Black is polA, red is polB
• pages 3,4: Fractional change Tsys (cal Off) versus pattern number.
• The Tsys for each band was normalized to the median value for the nite.
• pages 5,6: Fractional change Tsys (cal off) vs za
• The fractional change is now plotted vs the za of the cal Off.