links to plots, images,summary

• 230221: az,el:340,30 deg
  
• Dynamic spectra polA (.png)
  
• Dynamic spectra polB (.png)
• 70 sec average spectra and rms/mean by freq channel (.ps) (.pdf)
• Voltage sampling:total power vs time (.ps)  (.pdf )  (used 8bit output data, hit 8bit limit)
  
• 230222: az,el 302,31 deg
  
• Dynamic spectra polA (.png)
  
• Dynamic spectra polB (.png)
• 70 sec average spectra and rms/mean by freq channel (.ps) (.pdf)
• Voltage sampling, total power vs time during rfi peak (.ps) (.pdf) 16bit output data
  
• Voltage sampling:Plot power, voltage, & phase vs time during pulse (.ps) (.pdf)
• 230223: az,el: 340,20 deg
• Dynamic spectra polA (.png)
  
• Dynamic spectra polB (.png)
• 70 sec average spectra and rms/mean by freq channel (.ps) (.pdf)
• 7x172MHz total power vs elevation (.ps) (.pdf)
• 230224: azswings az el=40 deg
• Dynamic spectra CW az swing (az loc vs freq). PolA and PolB averaged (.png)
• Blowup of dynamic spectra around rfi azimuth peak (.png)
• Total power vs azimuth for the 7 x172 MHz bands. (.ps) (.pdf)
• Summary
  

---

Intro

    The 12meter xband receiver has seen rfi in the 8730 to 9000 MHz range. The rfi can saturate the xband rf.

    Previous pages covering this rfi:

• jun,jul2022: mapping cygnus at xband show xb saturation.
• 220830: xb azimuth swings at low elevation

---

Data taken:

    on 21 thru 24feb23 data was taken to learn more about the rfi

• 230221: 
• 172MHzz x 7bands of spectra dumped at .1 sec covering all of xband
• Az=340 , el=30 deg.
  
• Dynamic spectra polA (.png)
  
• Dynamic spectra polB (.png)
• The rfi repeats every 4 seconds
• the spectra are averaged to .1 seconds.
  
• the dark horizontal lines are where the xband receiver saturates.
• polB is saturating more than polA.
• 70 sec average spectra and rms/mean by freq channel (.ps) (.pdf)
• Black is polA, red is polB, blue lines are spaced every 20MHz starting at 8732MHz.
  
• Top: 70 sec avg freq. The bandpass has not been removed.
• bottom:rms/mean by freq channel.
• 5 MHz bandwidth voltage data taken
• the mock  setup was:
• 160MHz clock, decimate by 32 =5MHZ. We got  complex samples every .2 useconds.
  
• adjpwr 30cnts=1sigma (12bit input data),output 8bits
• cfr 8800MHz
• when the rfi was strong,  we hit the 8 bit limit.
• 60 seconds of data taken.
  
• total power vs time (.ps)  (.pdf )
• Top: total power vs time. 2 seconds of data  smoothed to 10 usec resolution
• You see the spikes occurring.
  
• around 7.8 seconds, the rfi saturated.
• you also see negative going spikes below this time period.
  
• The red dotted line is blown up in the bottom plot.
• Bottom: total power vs time blowup with 1 usecond resolution.
• the pulse lasts for about 22 useconds.
• processing: x101/230221/xbspec.pro, xbvoltsmptest.pro
  
• 230222:
• 172MHzz x 7bands of spectra dumped at .1 sec covering all of xband
• az=302,el=31 deg (maintenance was using the telescope so i couldn't move it).
  
• Dynamic spectra polA (.png)
  
• Dynamic spectra polB (.png)
• the spectra are averaged to .1 seconds.
  
• the dark horizontal lines are where the xband receiver saturates.
• polB is saturating more than polA.
• The red and green lines show where the 5MHz voltage data was taken.
  
• 70 sec average spectra and rms/mean by freq channel (.ps) (.pdf)
• Black is polA, red is polB, blue lines are spaced every 20MHz starting at 8732MHz.
• Page 1:
• top: 70 sec average spectra (bandpass not removed).
• The blue lines are spaced every 20MHz (starting at 8723 MHz)
• Bottom: rms/mean by freq channel covering all of xband.
• Page 2: blowup of rms/mean by freq channel.
• The blue lines are spaced every 20MHz
• the green lines are 16 MHz wide
• The rfi is 16 MHz wide and has a 4 MHz guard band.
  
• 5 MHz bandwidth voltage data taken
• the mock  setup was:
• 172.032 MHz clock, decimate by 32 =5.4 MHZ. We got  complex samples every .186 useconds.
  
• adjpwr 20cnts=1sigma (12bit input data), output 16bits (tried to give a bit  more dynamic range).
  
• 5MHz band centered at  8762 MHz
• 5 MHz band centered at 8798 MHz.
• 20 seconds of data taken
• Total power vs time during rfi peak (.ps) (.pdf)
• used the  5MHz voltage data centered at 8762 MHz band
• Page 1: total power vs time 8762 MHz Band
  
• Black is polA,red is polB
  
• Top: totPwr vs time with 10Usec resolution. .5 secs of data
• purple lines: the positive spikes are spaced by 14.5 milliseconds.
• the rfi returns to the same freq every 14.5 milliseconds.
  
• green lines: rfi spikes the strongest, also has negative going spikes.
• The negative spikes are when the rfi is strong but outside our 5MHz band,
  
• It saturates the entire xband rf band.
• The shape looks like a rotating radar sweeping through our beam.
• the negative spikes last for about 2.97 milliseconds.
• with a 4 second rotation period, the radar beamwidth woudl be < 2.7 deg.
  
• Middle: blowup around negative going spikes.
• The negative spikes spaced by 1.209ms. this is probably the ipp of the radar.
• the green lines are blown up in the bottom plots.
• bottom: blowup around a single positive peak.
• the pulse lasts for about 21.5 usecs.
• This is how long the pulse stays in our 5MHz band.
  
• Page 2: over plot 8762 and 8798 MHz bands
• Top: over plot 8762  (black) and 8798 (purple) total powers. 3Isec resolution.
  
• The +/- spikes are spaced at 1.2096 ms.
  
• The 8798 + spike occurs 2 ipps after the 8762 MHz spike
• the dynamic spectra showed 16 MHz bands spaced by 20MHz
• the 8798 band is 2  20MHz band from 8762.
  
• looks like the rfi is stepping through each 20MHz band every ipp.
• Middle: blowup with 1 usec resolution.
• When a band goes positive, the other goes negative (rf saturation).
• Bottom: blowup .186 usecond resolution
• The positive pulse lasts for 22usecs. This is how long the rfi stays in our 5MHz band
  
• The negative saturation lasts for 53.8 Usec. The xband rf covers all of the rfi freqband.
  
• 53.8/22*5MHz = 12.2 MHz.
  
• The dynamic spectra showed a 16MHz band.
• We don't see the entire 16 MHz  band probably because the freq edge may not be saturating the xband rf.
• Plot power and voltage vs time during pulse (.ps) (.pdf)
  
• Page 1: totPwr and phase vs time (8762 MHz band)
  
• Top: total power vs time for pulse (time resolution: 1 usec)
• black is polA, red is polB
• the green lines show 21usecs used to plot the phase
• middle: Voltage vs time polA
• bottom: voltage vs time polB
• black is the real component, blue is the imaginary value
• This is a chirped signal
• the chirped freq goes to 0 when it passes through the center of our 5MHz band.
• the voltage looks like it is getting clipped at 12 bits.
  
• Page 2: voltage magnitude and phase vs time
• black : polA, red: polB
  
• Top: voltage magnitude vs time during pulse
• Bottom: phase vs time
• a linear fit was also done to the phase vs time
• polA: chirpRate: 15.7 rd/sec
• polB; chirpRate: 17.0 Rd/sec.
• The data was probably saturated so these numbers may not be exact.
• This chirp rate probably extends over the enter 16 MHz of the rfi pulse.
• processing: x101/230222/xbspec.pro,xbvoltsmptest.pro
  
• 230223: Look for elevation dependence
  
  • 172MHzz x 7bands of spectra dumped at .1 sec covering all of xband
  • az=340,el=20 deg
    
  • Dynamic spectra polA (.png)
    
  • Dynamic spectra polB (.png)
  • the spectra are averaged to .1 seconds.
    
  • PolA shows no saturation at .1 sec integration. You can see the dark lines in polB .
  • The red and green lines show where the 5MHz voltage data was taken.
    
  • 70 sec average spectra and rms/mean by freq channel (.ps) (.pdf)
  • Black is polA, red is polB, blue lines are spaced every 20MHz starting at 8732MHz.
  • Page 1:
  • top: 70 sec average spectra (bandpass not removed).
  • The blue lines are spaced every 20MHz (starting at 8723 MHz)
  • Bottom: rms/mean by freq channel covering all of xband.
  • 172MHz x 7bands. move in elevation 6 to 85 degs at .5 degs/sec, dump spectra at .1 seconds.
  • total power vs elevation (.ps) (.pdf)
  • 4 172MHz bands did not contain the 8732 to 9000 MHz rfi
  • 2 172 MHz bands contained the 8723 to 9000 MHz rfi.
  • The positive and negative spikes are spaced by 4 seconds.
  • at .5deg/sec it shows up every 2 degrees.
    
  • Page 1: total power vs elevation and atm vs el fits.
  • Top& 2nd frame: polA, polB total power vs el for bands without the rfi.
  • the purple lines are a fit to Atm contribution to Tsys vs elevation.
  • 3rd & bottom frames: PolA, PolB total power vs el for bands with rfi.
  • the purple lines are fits to Atm contribution to Tsys vs elevation.
  • Page 2: Remove atm fit to Tsys  vs el
  • frame 1,2: bands with no rfi
  • frame 3,4: bands with rfi
  • The saturation and rfi peaks have maximum around an elevation of 38 degrees.
  • processing: x101/230223/xbspec.pro, xbrfi_elswing.pro
    
• 230224: Azimuth swings at el=40 deg.
  
• 172MHzz x 7bands of spectra dumped at .1 sec covering all of xband.
• Clockwise (CW) azswing az= -90 to +270 deg moving at 1 deg/sec
• Counter clockwise (CCW) azswing moving 270 to -90 at -1deg/sec
• Dynamic spectra CW az swing (az loc vs freq). PolA and PolB averaged (.png)
• this shows -90 to +270 deg az
  
• Blowup of dynamic spectra around rfi azimuth peak (.png)
• covering az=205 to 265 deg.
• the peak is at az=235 deg
• Total power vs azimuth of the 7 x172 MHz bands. (.ps) (.pdf)
• Black is the CW az swing, red is the CCW swing
• The tp in the swings do not sit at exactly the same azimuth because the rfi repeats at 4 seconds.
  
• Top frame: 4 x 172 MHz bands  outside the rfi freq range ( i excluded the band with the 8330MHz rfi)
  
• The negative spikes show the saturation.
• The max is at az=235 deg
• between 0 and 90 deg az there is no saturation
• Bottom: 2  x 172 MHz bands that include the rfi.
• the peaks is again at az=235.
• the rfi remains strong for most of the az swing
  
• between 0 and 90 deg the signal is weaker (hence no saturation seen in the above plot).
• processing: x101/230224/xbazwing.pro
  

---

Summary:

• freq range	8732 to 8992 MHz
  16 MHz bands spaced be 20 MHz	16MHz signal, 4 MHz guard band
  pulsed radar	ipp:1.2096 milliseconds
  rotation rate	4 seconds.
  chirped pulse:	about 16 Rd/usec Probably chirping over 16 MHz
  pulse width:	measured: 53.8 usecs (using saturation) probably: 70 usecs to cover the entire 16MHz
  freq hopping	step by 20 MHz every ipp
  measured pk	peak at: az=340, el=38 but strong through most of azswing except weaker at az 0 to 90 deg
  When seen	early morning (3am a4004) 8-10 am (rfi tests) normally not see after 11am through evening.
  Signal saturated xband rf	This is probably occurring in synth module of the UDC