cb az swings, 5ghz rfi  13jan16

• az 270-630 and back at +/- .35deg/sec
• clear weather, data taken in the afternoon
  
• dome at za=18deg
• mock spectrometer with 7 boxes covered 4200 to 5200 mhz, 150 Mhz spacing
  
• 172MHz bw, 8192 channels per box
• 1 sec sampling
  

• input the data averaging to 1 second
• average pola and b.
  
• interpolate az positions to center of data records.
• For dynamic spectra
  
• compute median band pass, divide this into all spectra
• find rfi free section of band, compute median value  for each spectra and then divide each spectra by these values
• For rfi bands:
• looking at the dynamic spectra:
• find the center and bandwidth of each rfi band
• also find the non-overlapping portion of the rfi band (many rfi bands were overlapping).
• compute the total power in each rfi band using the non-overlapping portion of the rfi band.
• plot this total power vs az for the CW and CCW az swings.
  

Dynamic spectra:

• 4950 Mhz band azswing (.gif)
• the 20 Mhz bands at 4930 and 4950 have been reported by chris  salter.
• these look like u-nni band transmissions.. which would make them illegal
  
• 5100 Mhz band azswing (.gif)
• the u-nni band starts at 5150 Mhz. everything below that is illegal.
• You can see that lots of the 20 Mhz rfi's overlap each other in frequency.

Birdie power vs azimuth.

• Each rfi was plotted in a different color with an offset for display purposes
• For each rfi, the CW and CCW az spins were overplotted
  
• The CW and CCW swings were spaced by at least 20 minutes...
  
• most of the rfi bands show a repeatable azimuth dependence.
  
• This means that the rfi is coming from a stationary azimuth direction
•  (not necessarily the azimuth of the peak since the scattering into the dome can be complex).
• Some of the rfi's look to have similar azimuth peaks.. that means they are probably coming from the same tower
• many of the rfi's have unique azimuth peaks.
  
• I wonder if some of these are coming from people's houses.. which would be spread in azimuth.

Are we seeing intermods in our receiver?

    Is it possible that some of the separate rfi's we see are being created as intermods in our receiver.
To test this, i took 300 seconds of data (15jan16) with the frequencies shifted..
If these are intermods, then the frequency location of the rfi's will shift as the lo's are shifted:

The setups were:   

• data taken 13jan16:
• rf cfr of band: 4650
• individual band centers: 4950, 5100
• az swings 270 to 630 ,za=18deg
  
• data taken 15jan16
• rf cfr of band: 4630 (shifted down by 20 Mhz)
• individual band centers: 4940, 5090 (shifted down by 10 mhz)
• telescope sitting az=200, za=18 degrees, 300 1 second spectra
  

• 4940 Mhz Band (.gif)
• 5090 Mhz Band (.gif)

After changing the lo's all of the visible rfi's showed up at the same frequency

• The 15jan16 data was at a fixed azimuth, so not all rfi's peaked up...
• The rfi's that were seen at the same frequency are labeled verified in the not intermod column of the table.
  

Summary

• The 5 ghz rfi's are not intermods in the system
• The rfi's repeat in azimuth (spaced by  over 20 minutes). So they are coming from a fixed azimuth location (they are not just time variable.
• a number of the rfi's have similar rfi dependence.. these are probably coming from the same towers.
• others have unique azimuth dependence. I wonder if they are coming from internet emanating from peoples houses (the uplink to the tower).