Every day, thousands of satellites pass over our heads almost unnoticed. Unbeknownst to many people, the signals from many satellites can be decoded with the right software. Today, we will go through such a process and download images directly broadcast by the passing weather satellites.
Receiving images from American NOAA satellites can be thought of as a “Hello World” of satellite reception. Although newer NOAA satellites broadcast at different frequencies, older versions, such as the NOAA-14, NOAA-15, NOAA-18, and NOAA-19, broadcast an analog signal containing weather information. With simple equipment, it is quite easy to decode these images as the satellite passes over your head.
The first piece of equipment is the antenna. In this blog, I will focus on the construction of a more advanced antenna, capable of NOAA weather satellite reception. Since I have built a first double-crossed dipole antenna and received my first images, I have had my eyes on a QFH (Quadrifilar Helix Antenna). Same as the double-crossed dipoles, a QHF antenna is circularly polarized with omnidirectional coverage. The design must match the polarization of the signal transmitted from the satellite. In the case of the NOAA family, the signal is RHCP (Right Hand Circularly Polarized). Omnidirectionality means that we don’t need to point the antenna straight at the satellite to detect the signal. Directional antennas have higher gains; however, without a satellite tracking mount, it is very difficult to aim the antenna at the satellite as it moves.
The design of the antenna is outlined in this article. Consequently, I will not be describing the construction process in detail.
I have decided to use 6mm diameter 1mm thickness brass pipes for the construction. Measuring the resistance of the pipes with a handheld multimeter, I was quite surprised to see a reading smaller than 0.6 Ohms per meter. Since I was on a deadline, I could not get soft copper tubing, generally used for such designs.
A benefit as well as the Achilles’ heel of this approach is the rigidity of brass. It proved challenging to bend the tubes into shape. I have tried heating the pipe with a blowtorch as well as a rewound MOT transformer for high currents. Warning: Disassembling microwave ovens can be FATAL! As the effect of softening the tube was minimal, I have decided to omit this step. The tubes were filled with sand, capped at both ends using duct tape, and bent into shape. I have used a radius of 26cm, or about 10 inches. The exact measurement was not crucial, as the tube will be bent further to join the individual pieces. I have bent and cut all the pieces following the tutorial.
Next, I have polished the pipes using steel wool to prepare the surface for soldering. I have used a blowtorch to heat the joints and used the cheapest dollar store solder. It proved beneficial to make a 45 cutout where the two pipes joined, to increase the contact surface. In the future, I would want to build another antenna with proper 90-degree elbow connections, as the bare colder joints are quite fragile.
The only 50-ohm coaxial cable in a local electronics store was 5.5mm in diameter. All the SMA and BNC connectors available, however, were for 5mm cables. This was a huge problem as neither of them fit. When I built my first double-crossed dipole antenna a few years back, I didn’t use crimped connectors. I have found that it severely degraded the signal quality, as I later swapped the connector.
The only viable option was to trim the connector in time under a microscope. This decision was made because if I were to order new connectors, they would not arrive in time. Consequently, each side took multiple hours to crimp.
The first step was to strip the insulation from the coaxial cable. After exposing the outer copper sleeve, it was cut using nail clippers, as can be seen in the photo.
The second step was to trim the foam sleeve using an Exacto knife. This step was critical, as exposing the bare copper will lead to a short with the outer shielding sleeve. First, large chunks of the foam were cut. This has shaped the insulation into a square. Later, the edges were trimmed to further minimize the volume.
The third step was to cut the copper wire in the middle as it was too thick for the crimping pin. Seven strands of wire proved to be the maximum the crimping pin could accommodate. The copper was cut to length and inserted into the pin. Finally, I could crimp it to the cable. I had to do this two times, as the first iteration did not crimp properly.
The last step was to crimp the brass sleeve to lock the entire assembly. The connector body must get under the copper sleeve, which is covered by the brass sleeve. This way, the connector will have good shielding and will hold its shape. The entire process was done two times, once for the SMA end and a second time for the BNC connector.
Image decoding was quite a difficult part, as I have not done it since my last antenna experiment. I have spent multiple hours tinkering with WXImg, Audacity, and Orbiton to get a stable signal.
Recording parameters in SDR# are quite tricky. I have kept the bandwidth at 45000Hz and used the WFM recording. The RF gain was set to 36.4dB, as the antenna had quite a good reception. During the satellite pass, I have adjusted the frequency for the Doppler effect manually; however, this can be done using Orbiton. It is important to resample the audio tracks after recording the signal. The value is usually 11025 Hz, and this step is carried out in Audacity.
