Backscatter Radar on VHF


Basically the idea of doing my own backscatter measurements has come from reading an article in RadCom by VK3OE about Chirp Modulation and Backscatter Radar, March 2012.

The idea is to moitor the backscatter on 6m during openings to find where the multiple hop landing zones are and how many of them there are, Clive G4FVP is especially interested in how often a path is open into the atlantic ocean during the season compared to east/southern Eurpoe (over land) which is the norm from the UK. This is especially difficult to find out as there are not many operators in the middle of the Atlantic! so the backscatter radar system should hopefully give us more of an idea of what is happening.

The system used in the RadCom article uses a HF Transceiver at either end and all software to generate the chirp and de-chirp the signal, this is the easiest way of doing it using the SpectrumLab software but there is no possibility for synchronised timing at present. The replacement solution being worked on at present uses an SDR transceiver which would be quite an expensive solution but very cool by the look of it!
I have decided to build the TX and RX seperately with GPS derived frequnency standard and GPS timing for the start of the chirp so we can rule out the timing errors wherever you are and also any frequnecy discrepencies so measurements can be done over the long term rather than the limited minutes at present, the receiving end could use the original software to do the same job but not as well synchronised.

My system comprises of a transmitter generating (for now) a linear frequency sweep from 50.433500MHz to 50.435500MHz every second to fit in a standard SSB filter bandwidth, the receiver uses the same frequnency sweep but offset by 1KHz as an LO for a direct conversion receiver which feeds the soundcard on a pc to receive the transmit direct signal and the frequency shifted backscatter signals.
If the receiver was in range of the transmitter you would just hear a constant tone at 1KHz from the speaker but on the pc you can see the reflections which are shifted in frequency by tens of Hz, it maybe audible to the ear but we have not tested this yet as the season has not started.. Meteor pings should be noticable though....



What would the reflected frequency shift be....


The Chirp (frequency sweep) ramps up by 2000Hz in one second..
If we have a reflection point at 1500km away the delay for this signal would be the distance of the path multiplied by two (reflected remember) then divided by the speed of an electromagnetic wave? so (1500x2)/300000 which is 10mS. There is also some extra delay due to travelling up to the ionised layer but that is fairly small for one hop, it may be more noticable over multi hops.
so 0.01s multiplied by 2000Hz gives us 20Hz
So at this distance the backscatter signal would appear 20Hz away from the direct signal on the spectrum, easy to see if you narrow the FFT bandwidth on the pc.



Clive who is located some 20 miles from me has a good 6m setup and is interested in doing the receive tests from his end so hopefully the distance will be far enough not to have too much of the direct signal to upset the readings.
The idea is to have initially a controlled 2element antenna at the TX end running 25Werp and control of the rotator to turn in 30deg segments. On the receiving end at each segment the spectrum can be averaged to make the weak reflection more visible, a 5-10minute run will give a 360deg plot of the reflection intensity vs distance.


Spectrumlab has a feature to save the spectrum trace over a specified frequency range so what I am thinking of doing is saving this data (for each segment) and then uploading it to my website and use some code to display it in a radar screen to make it look pretty! With multiple stations receiving it will be interesting to collate the plots to see what happens as these reflections should be seen by other stations many km away in UK and beyond..



A link to the VK forum with more information about the technique is here:

http://www.vklogger.com/forum/viewtopic.php?f=29&t=9002