Using the Graves radar to check elevated antenna lobes on 2m

By now, most VHF-ers have heard of the Graves radar on 143.050MHz in JN27si, France. The radar transmits very high effective radiated power (ERP) CW towards the southern sky at various azimuths switched in sequence and, apart from its intended purpose of tracking satellites for the French Air Force, provides a very useful beacon for radio hams. More details can be found on PE1ITR’s excellent webpage, here.

The ERP from Graves is high enough that it’s easy to receive moonbounce signals from it. In the past I have had excellent results using just a small 2m antenna, with enough signal to noise margin to suggest even smaller is possible – shown below is a 25 minute waterfall of signals captured from IO91wv in 2008, using just a 4el yagi fixed at around 100 degrees azimuth:

Signals from Graves radar

Signals from Graves radar using 4el yagi, as received in IO91wv.

Graves’ moonbounce signal is reliable enough that, by tracking the Moon’s azimuth with an antenna fixed at, or near, zero elevation, it’s possible to get some idea of what that antenna’s pattern looks like at higher elevations. There are (at least) a few caveats: Firstly the earth-moon-earth path is not 100% reliable or stable, so there will be signal variations that are unrelated to antenna performance; Secondly, 143.050MHz is some way away from the design frequency of a 2m yagi, so the antenna pattern will probably be skewed slightly; and Thirdly, Graves’ antenna pattern appears to be focused between 15 and 40 degrees elevation, so measurements need to be made when it sees the Moon within that window, which might preclude being able to test one’s own antenna at all (especially low) elevations. I decided to have a go at measuring it anyway, and was quite pleased with the results…

A quick run down of how my system is configured:

    • My 2m antenna is an 11el F9FT, rotatable but with a fixed elevation of about 10 degrees. The rotator is computer-controllable, with an Easy Rotator Controller fitted to its control box.
    • The antenna can track the Moon’s azimuth, using F1EHN’s superb free EME System software.
    • Audio from my FT857D is fed to a Windows XP PC, via an isolated interface (bought from ebay, cheaper than I could have built it for).
    • The audio is analysed, in this case, using DL4YHF’s excellent free Spectrum Lab software.

Spectrum lab isn’t necessarily the easiest thing to set up but it does provide powerful audio analysis features, which, incidentally, were hardly touched during this test! For this measurement I used Spectrum Lab’s plotting feature (Main menu->View/Windows->Watch List & plot window), with “watches” set for finding the peak signal (amplitude and frequency) within a small range of audio frequencies, targeted at where the signal from Graves was expected to be. Watches were also set to measure the base audio noise amplitude, and the peak-minus-noise amplitude.

With Spectrum Lab running as described above, I set the antenna tracking the Moon and started logging results from 15:26z on 13th November 2013. At this time the Moon was at 14 degrees above the horizon at JN27si, i.e. almost within the antenna pattern at Graves. Having to wait for the Moon to enter Graves’ beam pattern meant that the Moon was already almost 5 degrees above my horizon, already above my first ground gain peak: I don’t see that as a problem because I already know that the antenna works fine at very low elevations, I’m more interested in what happens higher up!

Running the measurements through to a little beyond 18:30z, when the Moon’s elevation at Graves went above 40 degrees, gave the following result:

Graves moonbounce signal

Graves radar moonbounce signal plotted against time

The top graph shows the peak amplitude of the Graves moonbounce signal (green) and the average base noise amplitude (red) against time. The bottom graph shows the audio frequency of the Graves moonbounce signal with dial set to 143.049MHz USB, within a few tens of Hz (blue) and the difference between the two signals from the top graph (mauve). The slowly changing doppler as the Moon rises can be clearly seen on the blue trace, with a few “spikes” where the signal is too weak to get a lock.

As previously mentioned, I already know that I get a good ground-gain peak between 0 and 5 degrees, which is not shown on the above graphs. However, the graphs do show two clear strong peaks, between approximately 16:05z-16:35z and 17:03z-17:32z, with a later insignificant peak. At my location on 13th November, these equate to elevations of around 10-15 degrees and 19-23 degrees respectively. The lower window is no surprise to me, agreeing well with previous results off the Moon. What is surprising is the higher window which appears to be equally as good: In the past I’ve always assumed that I might as well stop looking when the Moon’s gone above 15 degrees, but this shows that it might be worth looking again a little higher up! I shall certainly give it a try 🙂

2 thoughts on “Using the Graves radar to check elevated antenna lobes on 2m

  1. Werner, DK1KW

    Hi Mark,
    very nice report about GRAVES. I like GRAVES very much and I am amazed how well I can follow the signals with just a 3 El. HB9CV and IC-7100 on my bacony.
    I have observed the up and down of the moon reflections as well and always have regarded it as a matter of polarisation change during the orbit of the moon as my small antenna might not have significant side lobes. However I am not skilled enough to figure it out.
    Btw, did You make Your nice graphs just with Spectrum Lab? I am using this software too but just the basic features as I didn’t pick up the energy to read into the whole possibilities of that great software.
    Vy 73, Werner, DK1KW

    Reply
    1. Mark Turner Post author

      Hi Werner,

      thanks for your comments, we are very lucky to have such a signal close to the 2m band! The measurements and graphs are all from Spectrum Lab but, as you say, it requires time to find some of the features…

      Of course, you are correct about the possibility of changing polarisation – there are two sources: Firstly “spacial rotation”, caused by the angular geometry changing between two locations as the Moon’s azimuth changes (because we live on a sphere but mostly use simple az/el rotation systems); and secondly “Faraday rotation”, caused as the signal passes through the ionosphere. Spacial rotation is easy to calculate at any time, but Faraday rotation changes somewhat randomly and is also frequency dependent. Most people think of Faraday rotation as “a bad thing” but, in fact, it can sometimes help by cancelling out spacial rotation – so it’s not always bad!

      It’s difficult to find any exact information, but I suspect the signal transmitted from Graves is circularly polarised – It would certainly make sense from a system design point of view, to eliminate unnecessary signal variations. Assuming that’s the case, then we do not need to worry about any source of polarisation rotation, even if we are using linearly-polarised antennas to receive.

      Even if the Graves signal is linearly polarised, I suppose it would be possible to average out the effects of spatial rotation by making measurements on different days with the Moon at different declinations, because the spatial rotation will change at different rates. For best results, perhaps measuring with the Moon at high northerly declination would be best (in the northern hemisphere) because the spatial rotation then changes very little as the Moon rises up into the sky very quickly? Having said that, at 144MHz the effects of Faraday rotation are usually averaged out over fairly small timescales and, for these measurements, Faraday actually helps by cancelling any effects of spacial rotation too!

      For interest, I have repeated the measurements several times and always get a similar result, but it would be interesting to hear about other results too 🙂

      Best regards, Mark

      Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.