Happy New Year! We kicked off #CircuitPython2025 over a week ago and are excited to hear from folks. We got big news from @jepler to kick it off. (See below) Please post your vision for CircuitPython by the end of Tuesday, January 14th and email circuitpython2025@adafruit.com to let us know so we can link to it.

@jepler (aka @stylus) posted big news to Mastodon:

My #CircuitPython2025 @adafruit

I’m not going to bury the lede: 2025 is the year I’m hanging up my hat as a core developer of CircuitPython. My involvement with Adafruit & CircuitPython will change from what it has been to just an occasional project. For that reason, this post is more about expressing my gratitude for the past and present rather than a specific plan for the future.

In February 2018 I bought a pair of Trinket M0s because I’d heard about CircuitPython and wanted to try it. In Summer 2019, I did my first paid project for Adafruit, and then in November I left my prior full-time job to make Adafruit & CircuitPython my “main thing”.

After five years and a bit working for Adafruit, I have nothing but love and admiration for the people in this community. I wish I’d learned at the start of my programming career how much more you can create when it’s on a foundation of respect and a desire to help one another. Better late than never.

I grew in other ways too. I remember how nervous I was the first time I went on Show & Tell. Now, being live on YouTube is just a normal Wednesday night. I even gave a talk at a PyCon thanks to Adafruit, which was nothing I’d imagined doing.

What’s ahead for me? For now, I’m wrapping up long term Adafruit projects while trying not to start anything too big. Starting in April I’m going to travel for a few months. When I get back I’m going to focus more on local and volunteer opportunities. I plan to be around on Discord & GitHub to answer questions in cases where I may be the one with the expertise. Or when I have questions while I’m doing my own CircuitPython projects.

I’m proud of what I’ve been able to contribute to CircuitPython and other open source projects, and glad that it’s going to be available effectively forever as open source code. And in a way I’m excited about the void that my absence will create: I think it’s an opportunity for new folks to step up and make their own contributions to CircuitPython. With luck, something wonderful that stems from a fresh viewpoint and a distinct set of strengths. Something I couldn’t have imagined, let alone done. What’s your idea, and what will you create in 2025?

Thanks to Jeff for all of his contributions to CircuitPython and its community. We’re happy you’ll be around from time to time.

#CircuitPython2025 is our annual reflection on the state of CircuitPython. We’d love to hear from you too! See the kick-off post for all of the details. Please post by January 14th.

Many things have combined to make very high-frequency RF gear much more common, cheaper, and better performing. Case in point: [dereksgc] is tearing apart a 20 GHz low-noise block (LNB). An LNB is a downconverter, and this one is used for some Irish satellite TV services.

The scale of everything matters when your wavelength is only 15 mm. The PCB is small and neatly laid out. There are two waveguides printed on the board, each feeding essentially identical parts of the PCB. Printed filters use little patterns on the board that have particular inductance and capacitance — no need for any components. Try doing that at 2 MHz!

The LNB is a single-band unit, so it only needs to worry about the two polarizations. However, [dereksgc] shows that some have multiple bands, which makes everything more complex. He also mentions that this LNB doesn’t use a PLL, and he’d like to find a replacement at this frequency that is a bit more modern.

After the teardown, it is time to test the device to see how it works. If you want to experiment at this frequency, you need special techniques. For example, we’ve seen people try to push solderless breadboards this high (spoiler: it isn’t easy). Maybe that’s why many people settle for modifying existing LNBs like this one.

Since the latest release of ‘piolib’ we can do things like drive NeoPixels on any pin on the Raspberry Pi 5 which rocks, and means we can tackle the next, more complex, project: driving HUB75 RGB Matrix displays these require even MORE timing freakiness: using 10 pins, and ‘manual’ PWM means we have to constantly blit out the color dithering. Historically this was done with mmap’d memory to the GPIO controller bitbanging, which required a full core and could jitter depending on load. But now we can use the PIO peripheral! We can drive massive display arrays at high speeds and color depths using just about any pins. The future is looking bright – video.