Edson Pereira Amateur Radio Homepage PU1JTE : N1VTN : JF1AFN 

Antarctic 20m Spectrometer


Located at -58.4061° West longitude and -62.0743° South latitude (grid locator GC07tw), the Antarctic Spectrometer is an amateur radio experiment to monitor a 48 kHz segment of the 20m Amateur Radio band using a quadrature radio receiver and Fourier analysis software. The spectrometer is installed at the meteorology research lab of the Comandante Ferraz Brazilian Antarctic Base located on the King George Island.

Station Location

The spectrometer antenna is located at the tower near a small patch of snow in the lower right side of the above picture. The green and orange buildings are part of the Brazilian Base. Additional images are available on INPE's online camera.

The Hardware

The selected antenna is a Quadrature Fed Crossed V Dipoles designed by J. K. De Marco, PY2WM, specifically for the application. The antenna was chosen after many hours of simulations. Verticals, sloper dipoles, and V antennas were considered, but the quadrature fed crossed V dipole presented a good compromise in the radiation pattern since an omnidirectional radiation pattern was disired.

A vertical antenna was the initial choice, but after De Marco found that the soil quality for the location was quite poor and after including some approximate values for the soil quality into the antenna simulator, the results obtained were disappointing and the idea of using a vertical antenna was abandoned. The poor soil quality disturbed the antenna gain in low elevation angles due to the pseudo-Brewster angle.

Quadrature Fed Crossed V Dipole Antenna Model

The simulated radiation diagram for the Quadrature Fed Crossed V dipole antenna is shown below. Additional informantion can be found at this web page on PY2WM's web site.

Radiation diagram at 5° elevation for the various antennas that were simulated.
Blue: 45° Phased Double Slopers
Green: Quadrature Fed Crossed V Dipoles
Pink: Single Sloper Dipole
Cyan: ed V Dipole
Yellow: 1/4 Wave Vertical
Red: 180° Phased Double Sloper Dipole

Azimuth Radiation at 15°

Elevation Radiation

Dr. Setzer, Geraldo, and Franco assembling the antenna

The radio receiver is a SDRZero modified for the 20m band. The center frequency is on 14080.0 MHz.

The picture below was taken by William Schauff, PY2GN, before the system was packed and shipped to Antactica. The antenna is in the left lower corner and the SDRZero receiver in the right lower corner of the picture. The computer system used runs Ubuntu Linux and is owned by INPE, the Brazilian Space Research Institute.

The Software

The software used is a combination of generic ALSA utilities and some software and automation scripts developed by myself specifically for the application.

The quadrature signal from the SDRZero receiver is fed into the computer's onboard sound card using the stereo line input. The sound card converts the analog audio to the digital domain and makes the samples available to user software via the ALSA API.

The processing is done by a cron job (script) that peforms the following operations:

  1. Audio samples are captured by the asound ALSA utility and are stored in a .wav audio file containing 3 minutes of samples recorded at a rate of 48 kilo samples per second using 16 bits per sample..
  2. The .wav file is processed by utilities writen in C that performs Fourier analysis of the samples and saves the power spectrum representation in .png and binary files. The Fourier analysis is performed by a complex Fourier transform using the FFTW3 library. The results of the transforms are plotted on a spectrogram image where different color tones between black and white, with intermediate values in blue, represents different power levels. The software automatically adjusts the color contrast level to match the highest detected power and the noise floor. The images are created using the GD library.
  3. The output files are uploaded to the spectrometer web site in São Paulo.

Experiment #1 - Panoramic View of the 20m Band

This experiment processes 1 minute of samples and generates a 48 kHz panoramic view of the spectrum centered at 14080.0 MHz. The process repeats at every 15 minutes. The spectrogram is computed using a complex 4800-point FFT, giving a frequency resolution of 10 Hz per pixel on the final image. The spectrograms can be browsed at the spectrometer web site.

The following images are snap shots of the PSK31 and RTTY sub-band segments. The spectrum scale lines are in 10 dBFS steps.

PSK31 activity on 2008.01.27 07:00 UTC

RTTY activity on 2008.01.27 08:00 UTC

Experiment #2 - QRPp and QRSs Monitor on 14060.0 kHz

This experiment runs a complex 48000-point FFT analysis on the 3-minute .wav file and generates a binary file containing the power levels of 400 FFT bins around the 14060.0 kHz segment. The resulting file is then uploaded to the spectrometer web site where another software utility processes the binary file and generates a .png spectrogram image. Each pixel on the final image represents 1 Hz on the horizontal axis and 1 second on the vertical axis.

Experiment #3 - NCDXF Beacon Network Monitor

This experiment is similar to experiment #2, but the displayed frequency is centered on 14100.0 kHz for receiving the Northern California DX Foundation (NCDXF) Beacon Stations. The final image is time stamped to aid in the identification of the beacons.

NCDXF Beacons on 2008.01.31 04:15 UTC
Beacons from New Hawaii, New Zealand, Russia, and Peru
XS were received in Antarctica at the time of capture

NCDXF Beacons on 2008.02.02 02:30 UTC
A very strong signal from the beacon in Peru with some splatter

The NCDXF Beacon schedule for the 20m band is the following:

4U1UNUnited Nations00:00
W6WXUnited States00:20
ZL6BNew Zealand00:40
VR2BHong Kong01:20
4S7BSri Lanka01:30
ZS6DNSouth Africa01:40


This project was made possible by the invaluable contribution of

  • J. K. De Marco, PY2WM, for designing, simulating, and building the antenna, and modifying and testing a SDRZero for the 20m amateur radio band;
  • William Schauff, PY2GN, for the project conceptualization, logistics, and testing the final system;
  • Dr. Alberto Setzer, PY2ASK, for providing us with the opportunity, space on his meteorology laboratory on the Brazilian Antarctic Base, for providing a host computer, the transportation, and for assembling the system in the ice cold weather of the King George island;
  • The INPE staff based in Antactica (Heber, Gabriel, Geraldo, and Franco) for assisting us with the deployment and for keeping the system running;
  • INPE, CIRM, and PROANTAR for supporting and fostering scientific research programs in Antarctica.

73 de Edson, ewpereira {at} gmail {dot} com

 Home Copyright (C) 2008 Edson Pereira