Radar rfi with serendip 5 and alfa
The seti at home serendip 5 processor has been recording
2.5 Mhz bw centered at 1420 Mhz using 1 bit complex sampling. Some data
from 28mar07 (The first 240 records (50.3 seconds) from file: 28mr07aa)
was used to check for reported glitches in the system.
Beam 0 polA and polB were used for the processing.
The 2.5 Mhz bandwidth was reduced to 300 Khz by averaging 8 samples and
Dynamic spectra were made using +/- 10 kHz bw for the 50 seconds.
The total power over the central 20 Khz was computed and plotted vs time.
Combs were found that repeated every 12 seconds. These are radar rotation
periods. 1 second spectral averages were taken around 3 different
A model of the combs created by the airport faa radar over plotted on the
spectra for polA (.gif):
spectra for polB (.gif):
You can see the comb in the image. The image is scaled
to have maximum intensity at 8 *sigma. On the left edge of the image are
horizontal lines showing where the combs are at. These lines are repeated
every 12 seconds.
of total power and average spectra (.ps) (.pdf):
Page 1 Total pwr vs time: The central 20 kHz of the spectra were averaged
and then plotted vs time. Black is polA and red is polB. The dotted
colored lines correspond to the horizontal dashed lines in the image. They
are repeated every 12 seconds. The bottom shows the time for the first
pulse of each of the 3.
Page 2: 1 second average spectra around the combs. Black is polA, red is
polB. The plots show +/- 10 Khz about DC. The 3 frames are for the 3 different
sets of combs found. The bottom plot (birdies starting at 1.89 seconds
do not look like a comb). The other two do.
Page 3: The same plot as page 2 but now limited to +/- 2 kHz
You can now make out the individual comb elements.
The green dashed vertical lines are the frequency combs created by the
FAA radar using a model. See the faa
radar specs for the input used to make the model. You can also
see the idl routine that generates the data (mkfaarot).
The frequency comb in the top two plots corresponds to the faa radar comb.
The bottom plot does not.
The pulses/comb seen in the seti data corresponds to the faa radar.
There are two 12 second sequences offset in phase by 6.34 seconds (or 190
degrees of rotational phase). This may be the back lobe of the radar that
we are seeing. It also could be a reflection from an airplane that was
landed. If it was an airplane, then one of the 12 second sets should go
The 3rd 12 second sequence did not have an obvious comb.