radar rfi apr09

images/plots:
rms/mean by freq channel a2133 apr09 1365-1535 (.gif)
rms/mean by freq channel a2133 apr09 1400-1430 (.gif)
dynamic spectra 1375-1500 (.gif)
dynamic spectra 1400-1430 blowup (.gif)
dynamic spectra 1215-1385 (.gif)
dynamic spectra 1215-1385 and 1400-1430 side by side (.gif)
total power plots vs time (.ps) (.pdf)

    Project a2010 (alfalfa: alfa with 100 MHz centered at 1385, wapps, and 1 second dumps) reported rfi at 1403 and 1425 MHz. It was seen on 17apr09,  18apr09, and maybe as early at 14apr09. Rfi at other frequencies was also reported.

    AUDS (a2133) has been taking data in the early morning daily 05apr09 thru 23apr09. It's setup is:

• alfa with 230 second drift scans.
• Mock spectrometers covering 1225 to 1525 in two 175 MHz bands. One band centered at 1300 MHz and the other centered at 1450 MHz.  8192 channels give a 20Khz resolution.
• 1 second averaging.

rms/mean by frequency channel all of apr09: 

• The images contain 620 230 second scans covering 05apr09 -> 23apr09.
  
• each line of the dynamic spectra is the rms/mean for each of the 8192 channels in a 230 second scan.
• The first image covers the entire upper band
• The second image is a blowup of 1400-1430 MHz.
• the region 1440-1460 MHz was used to flatten the image. This mainly normalized scans that included continuum sources drifting through the beam.
• There was no PFB (butterfly) overflows in this data . These overflows could cause rfi to be spread across the band from a strong source.
  
• There are larger rms's at:
• 1400 .. weak intermittent rfi in the hi res image.
  
• 1403 .. the reported rfi
• 1420 .. HI
• the jumping back and forth in freq in the hi res image is probably from doppler shift..
  
• 1424-1428 .. reported rfi
• 1475 and 1480 MHz.
• 1390-1400 MHz for scans 290-300.
  

18apr09 scan 910800059

• Note that the frequency axis increases right to left.
•  dynamic spectra 1375-1500 (.gif)
• periodic birdies are at: 1403, 1425, 1475, and 1480
•  dynamic spectra 1400-1430 blowup (.gif) 
  
• The 1403 rfi is spaced every 12 seconds (so it's from a radar).
• The 1425 rfi splits into 1424.9 and 1425.1 MHz..
  
• They each have a period of 12 seconds so they are from a radar.
• They differ in time by 4 seconds.
• dynamic spectra 1215-1385 (.gif)
• This is the 172 MHz band with the radars
• 1358,1352,1343.
  
• These are the complex mixer images of the punta salinas radar (about 1300 MHz). They are created in the pdev baseband mixer.
  
• 1330,1350
• The FAA radar.
• 1313.5/1298.6, 1304.5/1289.3.
• These pairs are spaced by 15 MHz just like the punta salinas pairs.
• They have periods of 12 seconds (so they are a radar).
• The synchronize in time with punta salinas
• 1313.5 and 1304.5 are punta salinas frequencies, the other two are not. Maybe there is some extra mixing going on here.
  
• 1270 MHz radar.
• The punta borinquen radar
• 1260-1280.. chinese compass gns satellite (??).
  
• 1232.5/1247 and 1247/1256.5
• The punta salinas radar in mode A.
• The 1.5 second dark gap around each radar pulse is when punta salinas blanks in our direction.
• dynamic spectra 1215-1385 and 1400-1430 side by side (.gif)
• The 1300 MHz band is on the left, the 1400-1430 band is on the right
• seconds 140-210 are shown in the image.
• The dashed horizontal line aligns with the 1425 MHz rfi
• The dotted horizontal line aligens with the 1403 MHz Rfi.
• both of these align with different parts of the punt salinas 1232-1256 radar pulses.
• total power plots vs time (.ps) (.pdf)
• Page 1: rms/mean for the 230 second scan
• Red is polA (same as the images) and green is polB
• The top frame shows the entire band, the bottom frame is a blowup 1400-1430.
• 1403 (polA), 1420(a,b), 1425 (a), 1428.5(b), and 1475(ab) show larger rms's.
• Page 2: power vs time for various frequency bands.
• Top: The power around frequencies that showed radar rfi.
• Each frequencies is color coded.
• vertical colored lines are drawn every 12 seconds (radar rotation period).
  
• Bottom: The power around some of the known radars vs time.
• punta salinas: black,red,green, blue
• faa: dark red
• 1270 borinquen radar: purple
• the same dashed vertical lines as the top frame are plotted.
• Results:
• 1475 birdie is aligned with the faa
• 1403,1424.9,1425.1 aligns with puntas salinas low pulse (before AO blanking)
• 1424.9 also aligns with the punta salinas large pulse (after the AO blanking).

Summary:

• 1475 rfi is from the faa 1350/1330 radar.
  
• Not yet sure how it gets there.
• 1403 rfi is from punta salinas.
  
• I'm not sure how this is getting from the punta salinas freq to 1403.
• A2010 reported the same frequency (1403) as the mock spectrometers.
• A2010 and a2133 use different rf frequencies: 1385 MHz and 1375 MHz) so the lo's are different. This make it look like the 1403 is before the first mixer.
  
• 1424.9, 1425.1 are also from punta salinas
• 1313.5/1298.6, 1304.5/1289.3 are each spaced by 15 MHz so they must be from punta salinas.
• Note:
• Looking through the entire month's worth of  a2133 data there were lots of examples of rfi in the 1450 Mhz band. Some of it right on top of 1420 (looked like a radar).
  
• I haven't gotten any feedback from AUDS (a2133) about rfi in their data (especially the 1450 band).
• Looks like we can't rely on them for a prompt update on problems with their data.
• I guess we'll have to start looking at the data each day...
  

• Need to figure out how these birdies are getting in.. harmonics/intermods, or broadcast over the air.
• Take data at different rf frequencies and see if the rfi moves.
• Take data at higher time resolution to better resolve radar pulse with rfi.