There is a sense of glee associated with discovering that your interests unexpectedly join together, case in point, the 2026 Earth - Venus - Earth bounce, yes, Venus. Much like an Earth - Moon - Earth, or EME bounce, this is the act of transmitting a signal from Earth to Venus and receiving it back on Earth.
While humans have been using our Moon as a reflector, since accidentally in 1944 and on purpose in 1946, Venus is a harder nut to crack at 42 million kilometres distance - at its closest approach. The first radio echo was received in 1958. Radio amateurs have managed to receive a signal off Venus in 2009, 2014, 2025 and now in 2026.
Unlike the earlier achievements where the signal was a carrier, spectacular in its own right, the most recent effort was an actual message, “VENUS ECHO”. A week later on the weekend of 24 and 25 October the volunteers at Dwingeloo and Stockert sent and received a different message.
Here’s where it gets interesting.
The team has published the raw received data for the second message, even included the code they used for their own previous attempt and set a challenge, for anyone to decode the message.
Clicking on the information link opens with the header "Venus MFSK - modulation, coding, demodulation.
When reading that, I started giggling, since I’d been subjected to MPSK not ten minutes earlier in a completely different context, the decoding of an FM RDS signal, a digital signal embedded in a regular commercial FM broadcast. It’s the information your fancy radio uses to tell you which song is playing, for example.
That particular activity came about when I started hunting for example flowgraphs for FM decoders. As you know, I’ve been exploring GNU Radio and recently managed a little breakthrough in getting actual radio audio across the network from an RTL-SDR dongle. Although the audio quality was poor and in glorious mono, it was plenty clear to determine which station I was tuned to without needing to check the frequency.
I started on the journey of discovering precisely what VLC expects as an incoming raw audio stream, which turns out to be completely configurable - which changed my problem into, what does GNU Radio use to stream audio to an audio sink - the block that represents speakers on your computer. I came unstuck with how the audio channels were being dealt with, so I kept hunting.
On my journey I came across an article by “John E. Post”, I kid you not, in the February 2020 edition of Nuts and Volts magazine, titled: “FM Reception with the GNU Radio Companion”.
The article describes in great detail what an FM signal looks like, how to decode it and how to build GNU Radio flowgraphs to make that happen. It also comes with an archive of four separate decoders, including two stereo FM decoders. When I opened the files on my current copy of GNU Radio I discovered that there were several broken and missing blocks. This happens when GNU Radio updates code and makes old blocks obsolete. The blocks show as “Missing Block” in angry red and any blocks connected to one of these also turns red, so plenty of visual indicators that something is seriously wrong.
I attempted to discover these examples on GitHub, or anywhere else outside the archive, but failed, so I decided to create a new repository, called 202002-Post, since that’s what the archive was called, and began the process of translating old blocks into new ones. A partial conversion was achieved using a script called “GNU Radio Companion File Converter” by “M Khanfar”.
I was successful in translating the FM mono receiver and set out to make the RDS one into something that runs inside the current version of GNU Radio. In the middle of the spaghetti of blocks is one called “digital_mpsk_receiver_cc”, as you might guess, angry red, so I set out on the hunt to make it work, when after several hours of spelunking I discovered a decade old GitHub issue where it was depreciated. The promised replacement examples were not in evidence and I was left with a picture of a solution involving two different blocks, a “Costas Loop” and a “Clock Recovery”. I’m not going into either of these today, but together they appear to create at least a partial replacement for the now depreciated MPSK receiver block.
And that’s when I read the Venus decode challenge.
So, what does all these mean?
Consider it a convergence of concepts. On the one hand I’m working on a, some might say, boring, RDS receiver, on the other hand, it appears that the same techniques are being used to bounce signals off Venus. If it wasn’t for my seemingly random exploration of GNU Radio, FM broadcast, RDS and MFSK, there’s no way I’d even link that with a Venus bounce and decode.
Right now, I’m wondering what I don’t yet know to decode the raw signal and I’m keen to figure it out. I’m not ever going to be the first to decode this, since the EVE2026 team have already achieved this, but I’d like to discover if I know enough to do this for myself, or not.
That for me is precisely what this amazing hobby of amateur radio is all about.
If you want to have a go, you can find the details for the Venus decoding challenge on the https://www.camras.nl/ website or search for their callsign PI9RD.
Also, has anyone written any automatic layout tools for GNU Radio blocks, like Graphviz, because I’m getting thoroughly sick of overlapping blocks and really painful, almost “goto” like, constructs.
Oh, before I go, yes, MPSK and MFSK, not the same thing, phase versus frequency, lots to learn.
I’m Onno VK6FLAB

