How to see 56 MHz of bandwidth...

Foundations of Amateur Radio

The other day I stumbled on a project called Maia SDR by Daniel EA4GPZ. Maia, spelled Mike Alpha India Alpha, is a star in the Pleiades cluster. The Maia SDR project homepage proclaims that it is "An open-source FPGA-based SDR project focusing on the ADALM Pluto".

Now, I can completely understand if that collection of words is gibberish to you, but take it from me, it's not, let me explain.

PlutoSDR or Pluto is the common name of a piece of hardware which is officially called the ADALM-PLUTO Evaluation Board. It's a sophisticated device made by Analog Devices that provides a radio platform with some very interesting properties. Specifically it's both a radio transmitter and receiver with the ability to use frequencies between 70 MHz and 6 GHz. It runs embedded software you can tinker with because it's all Open Source and it's all very well documented.

Many people have used the Pluto as a remote transceiver by controlling the on-board radio with a USB cable. While that's neat, it's not what I have been wanting to do for a number of reasons.

The Pluto has the ability to sample data at a rate of 61.44 mega samples per second or MSPS. That translates to a bandwidth of 56 MHz. A typical amateur radio has a bandwidth of 2.5 kHz.

This bandwidth comes at a price. For starters, USB on the Pluto isn't fast enough to handle 56 MHz of data, so if you're using it as a remote radio over USB, you need to lower your expectations.

However, the hardware itself can process data at that rate, as long as it stays inside the radio. So, if you had a way to process data inside the radio and a way to show what you did with the data across USB, you could use all of the 56 MHz at once.

The Maia SDR project does exactly that. It processes the data and presents it to the world as a waterfall image, like the one you might have seen in WSJT-X, fldigi or SDR++. If you've seen the voice version of my podcast on YouTube, you'll also have seen a waterfall. It's an image that scrolls vertically, showing frequencies left to right, and signal strength by colour, traditionally, a rainbow that uses blue for low power and red for high power. Every time period the image scrolls adding another row representing the radio spectrum at that time. It's a very useful way to show massive amounts of radio spectrum data in close to real-time.

The waterfall that WSJT-X produces is about 2.5 kHz wide. The waterfall that Maia SDR produces is 56 MHz wide. To give you some context, the entire HF spectrum, between 2200m and 6m easily fits within 56 MHz.

Now, there's a wrinkle. As I said, the Pluto frequency range starts at 70 MHz, so that means we can't use it to listen to HF. Well, not without the help of another gadget, called a transverter. Essentially it moves a set of frequencies from one range to another. The gadget I have, a SpyVerter 2 HF Upconverter, translates anything between 1 kHz and 60 MHz and moves it to between 120 MHz and 180 MHz.

If you combine the Pluto with Maia SDR and a SpyVerter, you can plug your antenna into the SpyVerter, connect that to the Pluto, connect to the Maia SDR website that's running on your Pluto, tune it to 120 MHz, and see 56 MHz of HF bandwidth scrolling past as fast or slow as you want. You'll find the 10m band at 148 MHz, the 15m band at 141 MHz and the 20m band at 134 MHz.

Now if that's not cool enough for you, Maia SDR is as I said Open Source. This means that the project publishes all of the code that makes this happen. The Pluto comes with a number of devices on-board that process information. At the antenna end is an AD9363, essentially a chip that converts RF into digital and back. The digital information is processed by a device called an FPGA, a Field Programmable Gate Array. Field Programmable means that mortals like you and I can change the software that it runs.

Essentially an FPGA is a programmable circuit board used for information processing. To scratch the surface of what that means, you could for example program an FPGA to behave like a microprocessor, or you could use it to do accelerated matrix multiplications used for neural networks like you can with a graphics chip, or in this case, a device that does all of the digital signal processing. Finally the Pluto has a dual core ARM processor. You'll find those inside most Android phones and Raspberry Pi's to name a few. It's used to extract data from the FPGA and present it on a web page. Oh, and there's a progressive web app for your phone, so you can see this waterfall on your mobile phone if you want.

So, thank you to Daniel EA4GPZ for sharing your project, it's very much appreciated!

There are some caveats. The Pluto is easily overwhelmed by strong signals, so you probably need filters. I'm using a wide 2m band pass filter between the SpyVerter and the Pluto, just so that my local WiFi network doesn't overwhelm the whole thing. You're receiving between 0 and 56 MHz, so you'll need an appropriate antenna. The frequency response for the Pluto isn't linear, so the same colour on two bands might not be the same signal strength. You need to update the firmware of the Pluto, so make sure that you have a copy of the official firmware before you start because some of the FPGA functionality has been removed by Maia SDR to make this stuff work, most notably, the ability to use the Pluto across USB as a remote radio which is restored if you re-install the official firmware.

It's all documented really well and I'd encourage you to have a go if you're so inclined. If you're a software developer, Maia SDR aims to encourage FPGA development in the radio sphere using Amaranth, the project About page has more details.

As random Internet searches go, Maia SDR was a lovely surprise and I can't wait to dig deeper, but that will have to wait until my computer stops processing something like 6 billion WSPR records, which it's been doing for the past two weeks.

What have you found worth sharing?

I'm Onno VK6FLAB