Intro:

•  
• 161019 - lots of rfi seen
• processing the data
• plotting the data
• identifying the rfi
• dynamic spectra
• summary
  
• 
  

161019: 300 sec position switching.

The setup was:

• lbw receiver. Covering the entire band
• wapp backend, 100 MHz bands, cover entire band. 1 second dumps, 3 level sampling.
  

Processing the data:

• The average spectra for each on source was computed.
  
• Each 300 sec on/off pair was processed
• data was scaled to Tsys (didn't want rfi in cal to contaminate things)
• For each 300 sec on scan the rms/mean by channel was computed
• this does a better job of showing channels that have rfi.
• the tracks for different GNS (Global Navigation Satellites) across the dish were plotted for this time range
• galileo - 4 satellites visible
• glonass - 9 satellites visible
• beidou - 3 satellites visible (chinese gns constellation).
• gps - 10 satellites visible
• The angle between each gns satellite and the telescope beam was also plotted.

Plotting the data:

    There were 9 on/off pairs  of data taken on source.
Each wapp band is plotted in a different color.

Average 300 sec on scan spectra (.ps) (.pdf):

• 3 on scans per page
• Each band was normalized to its median value. No band pass correction was done.
• I've averaged polA and polB
  

300 sec On/off position switched averages (.ps) (.pdf)

• 3 on/off pairs per page
• The vertical scale is in units of Tsys
• PolA and polB  have been averaged.
• The expected rms Noise in a 300 second integration, 100 MHz bw, 4096 channels, hanning smoothing, polAdding, sinx/x channels, 3 level sampling:
• bwEff=100E6/4096 *1.208(sinx/x)*2.25(hanning)=66 Khz
• 3level noise: 1.23
• npolAdd=2
• on/off division : sqrt(2).
• tmOn=300. (on or off)
  
• deltaT/T =sqrt(2.)*  1.23/sqrt(bwEff * npol*tmOn) = .0003 Tsys
• The vertical scale is +/- .1 Tsys to show the rfi

rms/mean by channel for each on scan (.ps) (.pdf)

• rms/mean was done for each freq channel of each on scan.
• The expected rms (deltaT/T) -
• 1.23/sqrt(bwEff*1.sec * 2pol)= .003 Tsys.
• For some of the cleaner bands, the rms was .0035 Tsys.

• The two line elements sets from 21oct16 were used (so they should be pretty close).
  
• each visible satellite is plotted in a separate color.
• The vertical axis is  the AO za for the satellite.
  
• The horizontal axis is the AO source azimuth of the satellite.
• points are separated by a few minutes in time (but the spacing is not uniform in time).
• The satellite name is on the right. the date yymmdd and time hhmmss (ast) for the lowest AO za is printed on the right
• Page 1: Galileo satellites. 4 visible
• Page 2: Glonass satellites:  9 visible
• Page 3: Beidou satellites: 3 visible
• Page 4: GPS satellites: 10 visible.

The angle between the satellites and the pointing direction for the AO telescope (.ps) (.pdf)

• This was just done for the on scans.
• The difference was computed from the satellite ra,dec J2000 and the J2000 pointing direction of the telescope.
• The angles are great circle.
• I clipped the vertical scale to 40 degree separation. the table below shows some of the closest approaches:
•  

• constellation	sat	smallest ang (deg)	time(ast) hh:mm	freq used:
  Galileo	0204	19	03:50	E5A 1164-1189 E5B  1189-1214 E6A,E6B 1260-1300 E2 1559-1563 L1 1563-1587 E1 1587-1592
  Glonass	2460	7.5	03:31	1242-1252 1598-1605
  	2461	2.0	04:54
  Beidou	M2	5.0	03:25	1207.14 1268.52 1561.098
  	M4	6.0	05:00
  GPS	27	9	03:00	1174 1227 1381 1575
  	08	9	04:05

Dynamic spectra of scan 54: 04:43:51 ast

• the on scan was used (300 1 sec recs)
• I split the display into 2 separate images (bands 1,2,3) and bands (4,5,6,8)
• This scan included gps L3 at 1381
• the images were flattened by the median bandpass.
• The images were scaled to about 2 sigma (to show weaker things).
  

dynamic spectra bands 1,2,3 (.gif)

dynamic spectra bands 4,5,6,8 (.gif)

Identifying the rfi:

• Satellites:
  
• 1174MHz. gps  or galileo L5
  
• 03:26 strongest. probably gps 27. present at a low level entire time.
  
• 1207: beidou or galileo
• 03:26 strongest: beidou  M2 satellite.  5 degrees from beam.
• 1367: beidou
• 03:26 strongest. beidou M2 satellite. 5 degrees  from beam.
• 1381: gps l3
  
• 1575: gps or galileo
• 03:26 strongest. probably gps 27
• 1602: glonass
• 04:55 strongest. cosmos 2461. within 2deg of beam.
• 1620: iridium
  
• Radars:
• 1228,1243,1279,1294 : punta salinas mode A
• 1252,1257 1344,1349 FAA
• 1269.4,1275.6,1327.4, 1332.5 Punta borinquen radar.
  

161019: Summary:

• strong rfi was seen from the chinese beidou satellite at 1207 and 1268MHz,. It passed within 5 degrees of the main beam.
• The dynamic spectra showed that the gns satellites plus the radars is making it very difficult to do measurements 1170-1300 MHz.
• It would be helpful to pick an observing time when satellites are not close to the main beam.
• the wapps 100 MHz bw use a 3 level sampler. strong signals will spread in frequency more because of this.
• The mocks (with their 12bit samplers and pfb will do a better job of limiting the spread of strong rfi (but of course they still have their problem of jumps a power of 2^n from butterfly underflow)

 home_~phil