Intro  (top)

• The feed:
• ETS-Lingdgren Open Boundary Quad-Ridged horm.
• model 3164-06 300-MHz to 6 Ghz
• feed data sheet (.pdf)
• feed user manual (.pdf)

15may13 rfi scan using spectrum analyzer (top)

    The output of the mit lna was connected to an agilent spectrum analyzer. Data was then acquired from 150 MHz to 3 GHz to look at the rfi environment.
The setup was:

• the dewar had 50 db of gain
• their was a 10db pad between the dewar and the spectrum analyzer
• the preamp on the spectrum analyzer was off
• The resolution bandwidth was set to 2 Mhz.
  
• 20 second peak holds were done at each setting.
• The spectrum analyzer had 401 points.

• Each measurement of the spectrum analyzer is plotted in a different color (some measurements overlapped).
• Top: The value measured at the spectrum analyzer
• Bottom: The power level at the output of the mit feed (i removed the 40db from lna and 10db pad).

17may13 Pointing patterns on quasars and gps satellite  (top)

    The mit equipment tracked 2 quasars and one gps satellite on 17may13.
The plots show the az,za positions for each pattern (.ps) (.pdf):

• a star pattern was done:
• az motion, za motion, and then two 45 degrees strips.
• The star pattern is added to the normal tracking of the source across the sky.
  
• each * is spaced by 1 second.
• The az,za have had the model removed from the measured encoder values.
  
• the az is the feed azimuth. the source is 180 degrees from this
• Colors show the different strips. the start time for each strip is labeled on the left of each plot (hh:mm:ss UTC-4).
• The black * show the motion from the end of 1 strip to the start of the next.
  
• Page 1: Quasar 3C48
• We did two patterns on this using the standard heiles calibration scan
• The each strip was 24 arcminutes long (6 beams of 4 arcminutes)
  
• Page 2: Quasar  3C138
• We did two patterns on this using the standard heiles calibration scan
• The each strip was 24 arcminutes long (6 beams of 4 arcminutes)
• Page 3: gps satellite:
• This was done with a script that mike nolan wrote to send 1 second updates on top of the ephemeris tracking.
• In this mode, the recorded start times from mike's log file is probably 2 seconds behind when the offset was applied.
• each strip was 1.2 degrees long.
  

18may13: rfi from mit power on ,off switch  (top)

• 10 Ghz transceiver. fiber in, copper out.
• power strip with Ethernet  in (copper)
• All other equipment was plugged into the power strip.

•  the fiber transceiver and Ethernet switch used to turn the other equipment off was making a lot of rfi.
• This transceiver and Ethernet switch were not getting turned off (if they were off, then you couldn't communicate with them to turn the other stuff off).

• The spectra were taken with the mock spectrometers:
• 1 second dumps, 172 MHz bw, 8192 channels.
• The dashed lines (around 11.1 hours) show when the switch was turned off, back on, and finally off.
• The images have  been flattened by the median of the spectra with the switch off.
• 1115-1286 dynamic spectra
• 1265-1425 dynamic spectra
• 1415-1585 dynamic spectra
• 1565-1735 dynamic spectra

• The average with the switch on is plotted in red.
• The average with the switch off is plotted in black
• plots with red spikes and no corresponding black spikes come from the switch.
  
• The spectra have been normalized to the switch off (Tsys)
• A 31 order harmonic function was fit to the 8192 channels of the off.
  
• The vertical scale is in units of Tsys.
• Page 1: full scale, log plots
  
• The largest birdies occur at:
• 1250 MHz : 10000 Tsys
• 1375 MHz:300 Tsys
• 1687 MHz  200 Tsys
• Page 2: blow vertical axis, linear scale

Summary:

• The mit lincoln lab power on/off switch/transceiver caused lots of rfi at lband.
• The switch was first turned on tues 14may13 and remained on until sat 18may13 (11am). All of their equipment was removed on 20may13.
  
• The strongest lband birdies were:
• 1250: 10000 times Tsys
• 1375: 300 * tsys
• 1657: 200*Tsys.
• Weaker biridies covered the entire band
• Some users reported weaker birdies at  at 1422 and 1425 MHz.
• I did not check other bands. The rfi probably extended into s and maybe cband.
  
• The mit people have said they will try to shield this transceiver/switch pair on their next run here. If their shielding is not sufficient, then we will have to manually turn the equipment off after each run.

20may13 rfi azswing (top)

    An azimuth swing was done on 20may13 starting around 9:30 ast  to look at rfi using the lincoln lab wideband feed. The setup was:

• Feed azimuth 180 to 540 degrees at .4 deg/second
• dome za at 10 degrees
• receiver chain:
• 40 db pad
• lna with 50 db of gain
• 16db coupler
• 20 db amp
• spectrum analyzer
  
• spectrum analyzer setup:
• 2 MHz rbw, preamp on, 401 points, peak hold
• about 15 spectra were taken during the swing. 7 of these covered 150 to 1000 MHz.
  

• I've only plotted 150 MHz to 1000 MHz. The spectra higher than this did not show many birdies above the noise floor.
• Top: over plot the spectra during the swing.
• The dbm units are those from the spectrum analzyer.
  
• The dashed color lines are the rfi peaks that are plotted in frame 2 and 3. 
• Middle: rfi peak values vs time during the swing.
• Each colored line is a different rfi frequency (see labels on the right).
• The dbm units are from the spectrum analyzer.
• The spectrum analyzer time was not synchronized with real time.I'm not sure how to map time to azimuth position.
• The left side of the plot should be az=180deg,  the right should by 540 Deg, and the telescope moved at .4 degs/sec.
• (except that the time duration is > than the 15 minute swing.).
  
• Bottom: rfi peak values during swing scaled to the output of the antenna.
• The amplitude scale should match the output of the antenna
• The receiver chain had:
• -40db pad, +50db amp, -16db coupler, + 20 db amp = 14db added by chain.
• I then removed this from the middle plot
• The lna gain and the 20db amp gain is probably not flat across the 150 to 1000 MHz band.
  

• Col 1: hh:mm:ss ast
• Col 2: feed azimuth (deg). This is 180 degrees from the source azimuth. The encoder model has not been removed.
  

 home_~phil