180727-180803 looking at the sun
There is an interest in developing a solar radar
at AO. This could possibly use the highest band of the hf
transmitters (20+Mhz). A test was done to see if we could detect the
sun using the 8.175 MHz dipoles.
What we did:
- The mesh was in the 8Mhz position (15 feet below the 5 MHz
- the reflected power cables were connected to the tx side of
the load/ant switch.
- This brought the dipole signal back to the hf ctrl room.
- data acquisition equipment:
- the anritsu spectrum analyzer was run continuously with:
- 2 Mhz span, 10KHz rbw, cfr 8.175 Mhz, 25 sweep averages.
- data was taken with the preamp off and later with preamp
- 8Mhz dipole 1 was plugged into the anritsu
- this is a radial, positioned 60 degrees east of north.
- 2 sdr's were used to record i/q data (more
- sdr 1: a lowbandwidth of 240KHz then later 480KHz bw. cfr
of 8.175Mhz then 8.249 Mhz
- sdr 1:a wider bandwidth of 2mhz was used. cfr=8.175Mhz
- What a detection would look like.
- The sun transits around 12:33pm
- For these days, it was passing within .5 deg of overhead.
- the Beamwidth of the 8Mhz system should be about;
- using hpbw=K*lambda/diameter gives
- hpbw=1.2 * 36.7/300 = .147 Radians= 8.4 deg.
- It is probably less than that because of the blockage of
- So lets use 10deg for the beamwidth
- the earth moves at 4 minutes/degree
- So the sun will move through the beam in about 40 minutes
(to the half power points)
- we used only 1 dipole rather than all 3 pairs.
- Felix modeled this and said the beamwidth stayed
about the same, but the sidelobes went up.
- The bandwidth of the balun/dipole is supposed to be about
- So taking 100KHz total power about 8.175 Mhz, we should see:
- the signal will start to rise around 12:15
- peak at 12:33
- and go back down to the initial level around 12:50.
- Data acquired:
- 180727 18:30 to 180730 08:30
- anritsu run continuously, the preamp was off.
- telescope was in use during this time by regular
- 180731 to 180803 around lunch time. the anritsu preamp was
- for transit, the azimuth was at 270 deg. It wasn't till
the last day that i realized we needed to disable the drives
with the emergency stop button.
- 180731 09:55 to 13:00.
- 180801 11:20 to 13:20
- 180802 11:20 to 13:15
- 180803 11:05 to 13:00. Used emergency stop to try to force
vertex shelter off.
- Using the ionosonde, fof2 was 5-6 MHz at noon
during the observations.
Making the plots and
- Making the dynamic spectra
- they were flattened by:
- compute median bandpass for data set
- fit an 18th order harmonic function to this
- divide this into each spectra.
- this will remove the bandpass structure, but variations in
the spectra will be seen.
- The green horizontal line is the sun transit (12:33)
- the red vertical line is 8.175 MHz.. the center of the
- Making the plots
- Page 1: average and median spectra for the dataset
- some days have a second set (in different colors) where
the rfi was lower
- Page 2: 100KHz total power about 8.175 MHz.
- This should be the center of the baluns
- colored lines are flagged where sets of spectra are
overplotted in the lower panel.
- This was looking to see if the bandpass shape changed.
- Page 3: azimuth position, torques, and 100KHz total power
- top; azimuth position during data set.
- some red lines show where the azimuth state changed (or
15 minutes after a state change)
- 2nd: azimuth motor torques vs time.
- when the motors are disabled, the torques go to zero,
- 15 minutes after disable, a small current is run
through the motors to keep them warm.
- this is what is causing the rfi even when the motors
- 3rd: 100KHz total pwr vs time. The red flags show where
the rfi goes away when the motors are disabled and then
returns 15 minutes later when the heating current is
- 4th: 100KHz total pwr vs azimuth.
- For 03aug18 this shows how the azimuth dependence of
the power coupling into the dipole
The dynamic spectra
|27_30 (.ps) (.pdf)
- telescope moving the entire times
- median noise floor about -85 dbm in 10KHz rbw
- plot notes:
- page 3: red * show where azimuth was not moving. n
spectra for entire set
- telescope stationary for most of run.
- median noise floor all data:-82dbm
- median noise floor 11.3 to 11.56: -90dbm
- page 1: avg and median bandpass
- all data (black,red)
- 11.3 to 11.56 (green,blue) when rfi was lowe
- Page 3. az pos,torques
- at 11.5 motor currents low for 15 minutes.
- 100KHz totPwr -75dbm -> -90dbm .. but doesn't
got all the way down to -105dbm (value at 10.7 hrs).
11:20 - 13:20
- telescope stationary till 13:12
- median noise level (-87dbm @ 8.175Mhz)
- Not sure what caused the power drop around 12.7
11:15 - 13:15
- telescope stationary for entire time
- median noise level (-87 dbm @8.175Mhz)
- extra noise present for entire time
11:05 - 13:00
- telescope stationary during transit
- median noise level (-105dbm @ 8.175MHz)
- Used emergency stop button to try and keep warming
current off. Looks like it worked
- page 2: the purple trace shows the bandpass shape
when the strong motor rfi was present.
- The fwhm of the trace is about 100KHz.
- The signal is centered around 8.29 Mhz rather than
- it falls off by 10db in about 3 octaves
- The last page shows the azimuth dependence of the
- the radial dipole is at an az of 60 (or 240) deg.
- The minimum coupling is when the azimuth is aligned
with the dipole.
Measuring S11 from the hf ctrl room
On 02aug18 S11 for the 8MHz radial dipoles was
measured from the hfctrl reflected power panel (The signal output)
the Plots show the S11 results
- The keysight network analyzer was used.
- +/- 1 Mhz about 8.175 MHz was swept.
- No calibration was done (sorry dana...)
- The mesh was at the 8Mhz position .
- Each radial dipole is a different color
- black: dipole 1 (at az=60deg)
- red : dipole 3 (at az=180 deg)
- green: dipole 5 (at az=300 deg)
- a blue line is at 8.175 MHz. The dashed blue lines are +/-
- The minimum is at 8.3 MHz.
- The broadband noise from the motors was also centered around
- The 8 Mhz dipole 1 (radial at az=60deg) was used to try and
detect the sun.
- The sun was within .5 deg of overhead during the observations.
- The sun was not seen in any of the data.
- 27-30Jul18 data was taken while the telescope was moving.
- this caused a large amount of broadband rfi.
- 31Jul18 - 02Jul18 - data was taken with the telescope stopped,
but we still had lots of broadband rfi.
- The problem with rfi when the motors are disabled come from:
- After the motor currents are off for 15 minutes, the vertex
control system send a low current through the motor windings
to help with condensation.
- For this current to flow, the master contactor has to be
closed (to supply the power).
- Pushing emergency stop will force the master contactor to
remain open. This will prevent the warming current from
- 03Jul18: the vertex emergency stop button was used to prevent
the motor warming current to flow.
- the broad band rfi was not present.
- at sun transit the 100KHz total power was -105 dbm with no
sign of the sun.
- the azimuth dependence of the motor noise was a
maximum when the azimuth aligned with the radial dipole.
- S11 was measured for each dipole (with no calibration).
- the band was centered at 8.3 MHz rather than 8.175 MHz..
<- page up