Receiving and decoding C-Band Aeronautical communications with JAERO

Transcription

1 Receiving and decoding C-Band Aeronautical communications with JAERO I have seen a few YouTube presentations showing receiving and decoding equipment in action. Personally I dislike YouTube presentations, because often they are somewhat narcissistic in nature, in that they demonstrate that the uploader can perform the necessary actions, but they give little or no clue as to exactly how he/she went about obtaining these results. I would like to explain in this article how I went about decoding the C-Band signals and what equipment I used and how I configured it all. If you are an experienced satellite experimenter, stop reading this article, because you will probably find it a little condescending at times. Let us start at the beginning: The Dish I have tried working with a 1.2 m dish, but I have been unable to obtain a useable signal. Be aware that my comments are all based on my experience in Australia using signals from the Inmarsat 4-F1 satellite situated at E. I have read about people being able to receive and decode signals with a small off-set dish and I am amazed that they were successful. I can only assume that the signal strength transmitted from individual satellites varies considerably. In the end I used a 2.0m prime focus dish with a focal length of 73.2cm. The F/D ratio of this dish is 0.36 and at 3.6 GHz I would estimate the gain to be around 35dB. Be aware that the larger the dish diameter, the narrower the beam width and the more critical correct focussing, azimuth and elevation alignment becomes. This image dielectric plate inserted into the feed for circular polarization clearly shows the Aligning the dish Aligning the dish is a bit of a challenge. Because of the large diameter, both azimuth and elevation have to be very close to being absolutely correct. First you need to determine the azimuth and elevation of the satellite you chose as seen from your location. I use David Taylor s excellent programme WXTrack. To set the elevation is relatively simple and there are a number of inclinometer apps which work fine. I use an app called Clinometer. The azimuth is a bit more of a headache, because the steel pole on which the dish is mounted, interferes with the compass. In the end I used a long piece of straight timber, which I placed perpendicular to the base. I placed the

2 compass at the end and manipulated the length of wood until the heading was correct and then aligned the dish with that piece of wood as best as I could. The LNB or feed Initially I tried a Zinwell C-Band Dual Polarity LNBF model ZCF-D218. This worked quite well, except that the vast majority of LNB s are intended for TV satellite reception and as such are designed for very wide bandwidths. As a result they are not very stable and the signal tends to wander all over the place and unattended decoding is out of the question. I decided to fork out the cash required for a PLL (Phase Locked Loop) LNB and I chose another Zinwell product, the Titanium Satellite C1W- PLL Wideband LNBF. If you do not already have an LNB and are contemplating buying one, be carefull! The signals we want to receive are between GHz. Not all C-Band LNB s are created equal and quite a few do not go as low as 3.6GHz and they do in fact cut off at 3.7GHz. This LNB has a very low noise figure 15 K (Kelvin) the lower the noise figure, the better the S/N ratio. The gain is 65dB, which is adequate for most applications. Gain and noise figure are always a compromise generally as gain increases, so does the noise. The frequency stability of this LNB is absolutely amazing. As mentioned earlier most LNB s are designed for satellite TV reception and they are either horizontally or vertically polarized. This poses a problem because the aero transmissions are circularly polarized. Ideally we should use a helical feed, but we have no choice, but to try and convert the linearly polarized signals to circular and again ideally we would use a combiner phasing harness on the horizontal and vertical probes, but of course with an LNB we have no way of getting to the probes. The next best way is rather simple and this method uses a piece of dielectric which is inserted into the feed horn. (Make sure that when you order an LNB the dielectric is included). Circularly polarized signals can be either Right Hand (RHCP) or Left Hand (LHCP) polarized and I believe that the aero signals may be either. The change from RHCP to LHCP is achieved by changing the voltage to the LNB V is for RHCP and 16-20V if for LHCP. The LNB is mounted with a scalar ring, which should have been supplied with the LNB. Insert the Feed-horn into the scalar ring and fasten it at the appropriate F/D ratio for your dish. Also make sure that the LNB is mounted in such a way that focus point of your dish falls approx mm into the feed horn. A complete view of the dish. The square object in the centre of the dish is a RHCP patch antenna which I use for L-Band AERO signals (also using JAERO) No positional data on these, as it is Gnd-Sat

3 Supplying power to the LNB The LNB has to be powered, but under no circumstances should the actual receiver antenna be exposed to DC voltages. This power is used to operate the LNB as well as select polarization. Many experimenters have used a satellite receiver to supply the necessary power, as they have the required bias tee built in. Be aware that if you use the RTL SDR receiver, that although this receiver has a built-in bias tee, under no circumstances must it be used, because it operates on 5V and we need anywhere between 11.5 and 20V. I myself have always loathed to use a piece of equipment costing hundreds of dollars to do something which can easily be achieved with a home brew device. For a few dollars I put a high quality SMD capacitor (about 10pF) and a small SMA inductor (about 100nH) in a small diecast box, to make a bias tee. The homebrew bias-tee

4 I ordered a 35V switching power supply and used two small buck down converters, available on ebay to bring the voltages down to 13V and 20V and used a switch for appropriate voltage selection. Because the input voltage for the down converters is 36V max, I felt this was a bit too close for comfort, so before feeding the input voltage into the converters, I used a 7824 IC to bring the voltage down to a more manageable level. I put two LEDs in the box, one blue and one Red to indicate LHCP or RHCP duhhh The reason for using such a high voltage power supply is because I also have a linear actuator installed on my dish, which allows me to make remote adjustments to the azimuth. The switch which reverses the polarity is in the same box which contains the LNB power supplies. Of course an actuator is not necessary. I used one because I had one handy. The 35V power supply and the box with 2 buck-down- converters and polarity reversing switch The receiver The actual receiving frequency in my case was GHz and GHz. The local oscillator in the LNB operates at 5.150GHz, which means that the SDR should be set to receive at and GHz respectively. (LNB oscillator Freq Freq to be received = = GHz) It is at this stage I must confess to having made a colossal mistake in my initial calculations for the settings of the RTL SDR. Being a radio amateur I am used to heterodyne receivers, hence IF frequencies. Because the receiving frequency of the LNB was GHz and the output frequencies were MHz I concluded quite incorrectly that the IF output of the LNB was = 2.45 GHz WRONG!!! The correct formula is the one shown earlier i.e. LNB Oscillator frequency Frequency to be received. I used the RTL-SDR.COM device, which is based on the RTL2832UR820T2TCXO which also includes a Bias T + HF capability. The dongle plugs into an USB port and instructions for installing the software and driver are available on This device is incredibly stable (for a unit which costs around $20.00 on ebay).the cable coming from the LNB will more than likely be 75Ω impedance and have an F connector. You will need a converter socket/plug from F to SMA. The SMA connector is 50 Ω, but don t worry about this apparent impedance mismatch, as it is irrelevant in a receiving situation. Connect the antenna and make sure that power is inserted into the LNB through a suitable bias tee and that the voltage is correct for the direction of the circular polarization, which in my case is LHCP or 20V. Set the receiver to USB and since we at this stage are mainly interested in 10.5Kb burst signals adjust the bandwidth to around 10,500 Tune from GHz on the SDR, which equates to GHz actual frequency from the satellite. In my case and this applies to signals from the 4-F1 satellite only I receive 10.5 KHz burst signals containing operational data from the aircraft (speed, position, altitude, heading etc.) on GHz and GHz on the SDR. Once you can see the signals, open JEARO and set the programme to burst. I get best results

5 if I adjust the audio frequency, as shown in the right hand bottom corner of JEARO freq: to around Screenshot of the SDR settings and the JEARO panel. In the image above, the SDR uses a dual VFO configuration, to enable simultaneous decoding of data streams from 2 different frequencies. I have noticed that when using the 4-F1 satellite, adjacent to the right of the 10.5KHz burst signal, which is in the shape of a narrow hill there are 2 further signals which look a bit spiky. These are also AERO signals and are also decoded with 10.5KHz burst. However they have a staccato or machine gun like audio sound and do not contain positional data. The data is in fact similar to the data from L-Band AERO signals. These same type of signals are also in evidence to the left of the higher frequency main signal Once you are getting legible messages, rather than BAD T/R messages in JAERO, you might consider plotting the aircraft on Plane Plotter. I am rather computer illiterate, so I will not pretend to be able

6 to explain what settings to use in Plane Plotter. I think that the reader can do this with much more expertise than I can. I only managed with the help of PP list members. The above screen shot shows the PP display of areas N NE from Australia after running the system for about half an hour

7 This screen shot is from exactly the same data stream as the previous one, but it is using a different map, showing areas N N/W of Australia. Note that in both of these screen shots ONLY data received through my system is shown and there is NO shared (or downloaded) data. Is the satellite really geostationary In strict conventional contemporary engineering terms it is and it is supposed to sit right on the equator. However like all geostationary satellites the Inmarsat birds do wander( it seems in latitude only) and the 1-F4 can go up to 3 N to 3 S, which adds up to a 6 variation. This does create some problems for a 2m dish as it has a fairly narrow beam width and you may have to periodically (maybe every 4-5 hours or so) adjust the elevation depending on where the satellite is located.

8 of complete set-up A picture the Acknowledgements: I should not fail to thank Jonti (Jonathan Olds) in New Zealand for his excellent piece of software JAERO His programme is truly amazing and in spite of its ability to do lots of stuff, even I can use it without tearing my hair out (what little there is left). I have a lot of respect for the Kiwis. When I was receiving and decoding WX sat APT images, I used another Kiwi developed programme called WXtoimg. There is a lot of talent in a country of 4.6 million!

9 My thanks also go to PP s John.Satcom who was very helpful and gave me lots of encouragement and clues on how to get stuff working. Finally I should thank David Taylor. Without his Sattrack programme I would have had difficulty aiming my dish and sorting out some problems related to satellite movements / jitter. Disclaimer I have no association or financial interest with the manufacturers and owners of the brand names and model nomenclature and software used in this missive. I have simply cited them to give you some idea as to what I use. I am sure that there are many other devices on the market which will work just as well. Good Luck and enjoy!!! CvL