Surprise! We launched a podcast. We’ll be sharing the inside scoop on how we design our products and what they’re capable of, direct from Pi Towers.
Our inaugural episode takes a close look at Raspberry Pi’s next-generation microcontroller, RP2350. Paul Sherry and Chris Boross from our commercial team walk us through Raspberry Pi’s second franchise, our silicon products, and in particular the RP2350 chip that succeeded RP2040.
Raspberry Pi is probably still best known for our single-board computers and modules, but for over half a decade we’ve been a chip developer too. That journey began when we couldn’t find an off-the-shelf microcontroller that hit the right price and performance for our earliest Raspberry Pi Pico products. We built one of the most talented in-house ASIC teams in the business and designed RP2040 with the aim of using it just for our own products, but other companies soon took an interest. Fast forward to 2025 and we found ourselves, for the first time, selling more microcontrollers over the year than “big” (well, relatively speaking), board-level products like SBCs.
RP2040 launched in 2021, sold millions of units, and, perhaps most valuably, generated loads of customer feedback. Three areas claimed our attention: a lack of super robust security features, which worried some of our commercial customers, together with a demand for more CPU and memory headroom as well as more I/O. We listened and designed RP2350 – secure, powerful, and peripheral-rich – specifically to address this feedback.
A major focus was keeping the new product affordable, at a “disruptive” price point, if you will, and this was probably the biggest ask for our chip architects. RP2350 retails in reels at around 80 cents per unit, only 10 cents more than RP2040, despite roughly doubling the compute on offer.
Chris and Paul get into all the details in the podcast; thank you to the hosts of our first episode! By way of a super tease to encourage you to listen, here’s one of our favourite bits: Chris and Paul call the RISC-V cores in RP2350 “a love letter to CPU enthusiasts,” giving the community a low-cost way to experiment with the architecture. What’s not to [nerdily] love?
Subscribe to the Raspberry Pi Podcast on Spotify, Amazon, or Apple podcasts. We’ll also be releasing each episode on our YouTube channel, where you can watch for free. (Here’s a handy RSS feed for those asking!)
What would you like to listen to next? Leave us a comment requesting a topic and we’ll try to lure the appropriate Pi Towers-dweller behind the microphone for you.
If you’re a Redditor, our upcoming AMA might be worth a look. You can ask our CEO Eben Upton and our CTOs Gordon Hollingworth and James Adams [almost] anything this Thursday 21 May from 3pm UK time until about 5pm, or whenever the comments descend into people telling Eben how much he looks like Jason Statham. Then we will stop.
The post Welcome to the Raspberry Pi Podcast appeared first on Raspberry Pi.
This Maker Monday, we’ve gone big with this Raspberry Pi–powered electric ‘Frankencar’ made up of parts from different vendors. The OVCS (Open Vehicle Control System) team converted an old VW Polo into an electric vehicle that can be driven remotely.
And kids, adults, all of the above, I cannot stress this enough — DO NOT TRY THIS AT HOME.
The electric vehicle (EV) revolution is just about here. With EV charging points found nearly everywhere, and more established car manufacturers introducing electric variants of existing models — or entirely new ones — a petrol-free future seems closer than ever. Of course, as a reader of this blog, you will be aware of how technology companies can be, and won’t be surprised to know that these manufacturers are using a lot of proprietary tech in their vehicles.
“Our project, OVCS [Open Vehicle Control System], aims at breaking the traditional vendor lock-in that you see in cars and other vehicles,” Marc Lainez tells us. “We want to make it possible to interface parts from different brands together as if they were always meant to be working that way. Most car parts have a universal functionality to perform (braking, steering, showing data…), but the language they speak is different. So we thought we could build such a platform that would allow tinkerers like us to extend or swap a vehicle’s functionalities with parts from any brands.”
Marc and his team have developed a prototype of this platform, which uses Raspberry Pi to translate between the different parts.
The team had been looking for a larger-scale hobby project they could sink their teeth into, one that would combine all of their various interests. EVs ended up sitting at the centre of the Venn diagram.
“[We] were growing concerned with the security and reliability of vehicle software platforms,” Marc says. “We thought this was the perfect project to learn a ton about how car parts communicate, how they interact together, and how we could seamlessly integrate parts together using modern languages on off-the-shelf hardware components like Raspberry Pi.”
Raspberry Pi is used in multiple ways in the concept car: first, to power the vehicle management system, which they describe as the brains of the platform.
“It translates messages from the different communication buses (CAN) and routes them to the appropriate ones,” Marc explains. “In total, we have five CAN networks that are being accessed through SPI modules connected to a Raspberry Pi. Without this Raspberry Pi, the car wouldn’t be driveable.”
It’s also used in the infotainment system — something we’ve seen several folks do with Raspberry Pi in cars before. Not only does it show all of the usual info about your vehicle, but it also includes a touchscreen automatic gear shift, as the car is an EV conversion.
Finally, there’s the radio bridge: “[It’s] a component connected to the CAN bus and sends instructions to the VMS to accelerate, brake, and steer,” Marc says. With it, they can control the car remotely.
“From a software perspective, we wanted to have a technology stack that was familiar and at the same time, something ‘batteries-included’ that would allow us to easily build firmware in a high-level language while at the same time making the firmware updates really easy,” Marc continues. “Since we had done quite a lot of Elixir development, we used Nerves. This is an IoT framework built in Elixir and Erlang that relies on Buildroot (Linux build system) and gives you the ability to write your firmware in plain Elixir, a high-level functional language. It made our development cycles much faster/shorter and easier and allowed us to use a language we were already familiar with.”
Getting the various parts to communicate — such as a Nissan Leaf electric motor and parts from a VW Polo — was one of the hardest elements, as manufacturers generally do not publish documentation on how their components communicate.
“We had to reverse-engineer quite a lot of messages in order to make the car functional,” Marc reveals. “To give an example, if you want to know what message gives you the handbrake status (pulled or not), you look at all that is passing on the bus, you pull the handbrake a few times to see what frame is perfectly synchronised with your action to isolate its ID, then you check which bytes change when you pull it… For more complex components, this is a combination of multiple messages and, fortunately, there is a community of car tinkerers who publish their findings on forums online. Sometimes the work was done; sometimes partially and we had to complete it.”
Over the course of 18 months, the team did manage to make their ‘Frankencar’ driveable, which was their main goal — they then went beyond that by making it remote-controlled. They also wanted to document their build, which they’re in the process of completing. After that? “The next goal is for the car to be self-driving,” Marc says.
If you liked this article, there are many more like it in the latest issue of Raspberry Pi Official Magazine. You can purchase a copy from the Raspberry Pi Store in Cambridge. It’s also available from our online store, which ships around the world. And you can get a digital version via our app on Android or iOS.
You can also subscribe to the print version of our magazine. Not only do we deliver worldwide, but those who sign up for a six- or twelve-month print subscription will receive a FREE Raspberry Pi Pico 2 W!
The post OVCS: Raspberry Pi–powered electric car appeared first on Raspberry Pi.
WWF-Pakistan recently got in touch to tell us about the work they and the Lahore University of Management Sciences have been doing to mitigate human–wildlife conflict in the Gilgit-Baltistan mountains. Using Raspberry Pi 4 to power a specially trained AI detection and alert algorithm, the team is helping local communities protect their livestock without damaging the region’s ecosystem.
High in the snow-clad mountains of Gilgit-Baltistan, Pakistan, life is defined by extremes. Bare, rocky ridges stretch across a harsh landscape where communities depend heavily on livestock for survival. In this same terrain lives one of the world’s most elusive predators — the snow leopard.
For local communities, the snow leopard is both a symbol of pride and a source of risk. As natural prey declines and habitats shift under the pressures of climate change and expanding human settlement, livestock depredation has become the primary driver of human–wildlife conflict. In a single predation event, as many as 60–70 animals can be lost — a devastating economic blow for families whose livelihoods depend on their herds.
To address this escalating challenge, WWF-Pakistan partnered with the National Center of Robotics and Automation at the Lahore University of Management Sciences (LUMS) to develop and deploy an AI-powered Predator Early Warning System. At the heart of this innovation lies Raspberry Pi 4.
Since 2022, solar-powered camera traps built around Raspberry Pi 4 single-board computers have been strategically installed in high-risk zones near human settlements. Well adapted for remote, high-altitude environments, the system combines durability with intelligent edge processing.
Each unit captures images of passing wildlife and transmits real-time data to a central monitoring platform through the local 4G network. A trained AI algorithm analyses incoming images to detect the presence of snow leopards and other predators, including foxes, wolves, and Himalayan lynxes.
When a predator is identified near grazing areas or livestock enclosures, the system automatically generates alerts for field teams and community members. This early warning enables villagers to secure their livestock before an attack occurs, shifting their response from reactive to preventative.
The innovation of the solution lies not only in the hardware or the algorithm, but in its impact on community perception. By reducing economic losses, the Raspberry Pi 4–powered system is helping to break the cycle of fear and retaliation that often defines human–wildlife conflict.
Communities are no longer forced to choose between protecting their livelihoods or protecting the region’s wildlife. Instead, they are equipped with information that allows them to do both. This has resulted in a measurable attitude shift, with locals going from viewing the snow leopard solely as a threat to recognising its ecological importance for maintaining healthy mountain ecosystems.
The current model has proven both scalable and adaptable. With support from the Forest, Wildlife and Environment Department of Gilgit-Baltistan, there are plans to expand the system to additional high-conflict areas across the region. The next phase will be for WWF and LUMS to look at how to deter predators from attacking livestock altogether.
This initiative has demonstrated how affordable, flexible computing platforms like Raspberry Pi can power real-world conservation in some of the planet’s most challenging environments. By combining edge computing, AI, solar energy, and mobile connectivity, WWF-Pakistan and its partners are showing that technology can do more than monitor wildlife — it can enable coexistence.
The post Powering coexistence: how Raspberry Pi technology is helping WWF protect wildlife and communities in Pakistan appeared first on Raspberry Pi.
Raspberry Pi is always looking for new and better ways to do things. One of the most significant ways in which AI has affected the internet is the proliferation of chatbots and answer generators, and while we’ll never use AI to author our documentation, we’re curious to find out whether tools like these could help users find technical information easier and quicker.
Something you may have heard about is retrieval-augmented generation (RAG for short), which uses a defined set of documentation to answer questions for you in an informed way. In this blog, we are trialling two different RAG-based documentation tools: InKeep and Kapa. Our website is serving two versions of today’s post to visitors at random; if you reload the page, you’ll get the other tool. Why not ask this one a question about something you’d like to do on a Raspberry Pi and see how it responds?
We’d love you to think about whether the tool did a decent job of answering your question — give the response a thumbs up or a thumbs down and get ready to tell us anything else you want us to bear in mind when you get to the comments section. We take our documentation very seriously, and your thoughts and opinions will help us augment and improve it.
The chatbots are actually using Raspberry Pi documentation, including the html, the white papers and books stored in pip.raspberrypi.com, and some specific GitHub repositories that have documentation built into them (for example, rpi-image-gen). Because of this, it can search within the documentation for specifically relevant chunks of information to feed into the AI model. It can then ask the model your question, and the model will decide whether it actually has relevant information and generate an answer if it feels it does.
When LLMs ingest something, it's first converted into tokens. You can think of a token as a thing that represents a single word, although most tokens represent sub-words ('running' might be broken up into 'run-' and '-ning', for example). The next stage is to convert those tokens into 'embeddings'. Embeddings are actually vectors in a very high-dimensional space (up to a thousand different dimensions), and these vectors represent the "semantic meaning" of the token. Two vectors that are close in value have a similar meaning. Here's the canonical example:
The diagram above shows an example displayed in three dimensions (because we can't even visualise what four spacial dimensions would look like). You could think of the dimensions as terms like 'gender', 'royalty', and '80s music reviews'! If you take the embedding for 'woman' and subtract the embedding for 'man', you get a vector that roughly means "the shift from male to female". Add that same shift to 'king', and you land very close to 'queen'.
This isn't a party trick. It tells you something genuinely useful: nearby coordinates in this space mean similar things, and the same kind of relationship between pairs of words shows up as the same kind of geometric shift. The model has learnt, from billions of words of training data, that the relationship between 'king' and 'queen' is about the same as the one between 'man' and 'woman', and it has encoded that where the points sit.
An extension of this concept is to develop embedding models that can take a sequence of tokens (like a paragraph of, say, 100 words) and create an embedding (vector) from it that gives the semantic meaning of the section of text. This is the type of embedding model used in RAG-based systems, as we want them to find chunks of text that have similar meanings.
Once you notice this property, it's fairly clear what to do with it. Take all your documentation, chop it into reasonably sized chunks — a paragraph or two each — and generate an embedding for every chunk. You now have a database in which similar meanings are close together.
When someone asks a question, the system generates an embedding for that question and goes looking through the database for the embeddings that sit closest to it. The closest ones are, by construction, the chunks of documentation whose meaning is most similar to the question. You haven't had to guess the author's vocabulary; the model has done that work for you.
At this point, the system has a list of the most relevant bits of documentation. It could just display them to the user as is — and plenty of search systems do stop there — but it can go one step further: it can hand those chunks, along with the original question, to a language model and ask it to write an answer using only those chunks.
That's the 'generation' half of 'retrieval-augmented generation'. The retrieval step finds the right documentation, and the generation step turns it into a tidy, plain-English answer. The model isn't drawing on whatever it happens to remember from the training data, which is where language models tend to get themselves into trouble; it's giving an answer from your actual documentation and nothing else. If the answer isn't there, the tool will state that, rather than making something up.
Please have a go and give us feedback on what you found; convenient thumbs-up and -down buttons are included in the chat interface. That information, along with the interaction, will be stored in the system so that we can determine what works and what doesn't.
We will never use AI to create documentation in the first place. Instead, we are hoping to use these tools to help inform us where our documentation has gaps or errors; we put a lot of effort into creating it, and we want it to be as accurate, complete, and straightforward to use as possible. Our documentation team and I are excited to scrutinise the results to discover what they reveal about your needs and how effectively we serve the information you want.
The post Search our documentation by meaning, not keywords appeared first on Raspberry Pi.
This Maker Monday, Phil King shares some of the newer RP2350-based boards that have come to market since we last looked at similar devices back in issue 148 of Raspberry Pi Official Magazine. Just like before, these examples demonstrate the flexibility and versatility of our microcontrollers.
The RP2350 microcontroller is the brains of Raspberry Pi Pico 2, along with a host of third-party boards. RP2350 features dual Arm Cortex-M33 processors running at 150MHz, 520KB of on-chip SRAM, and twelve PIO state machines. So, compared with the RP2040 microcontroller from the original Raspberry Pi Pico, it offers a major performance boost for handling more complex computational tasks. It’s no wonder it’s been used in such a wide range of third-party boards and devices — you can check out the full selection in the Powered by Raspberry Pi product catalogue.
We’ll be taking a look at a few of the most interesting ones here, equipped with a variety of special features such as battery power inputs, extra GPIO pins and connectors, motor/servo controllers, Ethernet ports, IMUs, and even mini LCD touchscreens.
Pimoroni | £21 / $23
Did you know that there are actually two main versions of the RP2350 microcontroller chip? There’s the standard RP2350A used in Raspberry Pi Pico 2 and many third-party boards, and RP2350B, which adds 20 more pin connections — including 16 extra GPIO pins and five more ADC channels.
Pimoroni’s Pico LiPo 2 XL W makes full use of RP2350B with an elongated board that breaks out those 20 extra pins. Helpfully, the rest of the pinout is the same as the standard Raspberry Pi Pico one, and all the (unpopulated) pins are labelled on the top of the board, so it’s easy to find the ones you need. Memory and storage have also been super-sized, with 8MB of RAM and 16MB of flash.
A USB-C connector is used for power and programming. You can also power the board from a LiPo or Li-ion battery (not supplied) via a two-pin JST connector; there’s on-board battery management and charging circuitry.
Other features include connectors for Qwiic and STEMMA QT, debug, and SP/CE (SPI/serial), along with boot/user and power buttons. A Raspberry Pi Radio Module 2 provides Wi-Fi and Bluetooth connectivity.
Extra GPIOs, battery input/charging, and much more.
Pimoroni | £35 / $38
With myriad connections and extra features, this versatile board makes it easier for beginners to get started with coding and electronics. It’s powered by a standard Raspberry Pi Pico 2 W (with built-in wireless connectivity) soldered to the top of the board, surrounded by ports, pins, NeoPixels, and buttons — an identical layout to that of the excellent Inventor 2040 W it supersedes.
For robotics, there are headers for up to six servos, along with an on-board motor driver connected to two JST-SH ports. The latter are for connecting encoder-equipped motors, but you can still wire standard ones up to a four-pin connector.
There are also six GPIO headers (including three analogue inputs), two Qwiic/STEMMA QT ports (for sensors), an I2C/Breakout Garden unpopulated header, a JST-PH input for optional battery power, a two-pin connector for 1A at 1.5W audio output, plus handy ‘User’ and ‘Reset’ buttons. Phew!
Software-wise, the Inventor 2350 W is easy to use, thanks to comprehensive libraries and code examples for MicroPython and C/C++.
A feature-packed board, ideal for electronics and robotics.
Waveshare / The Pi Hut | £23 / $31
Most Pico-style boards lack any form of display, though you can buy add-ons to enable video output. Waveshare, however, produces a range of RP2350-based devices with built-in displays in various shapes and sizes.
This particular model is circular and comes with an optional metal case with cut-outs for the boot and reset buttons, USB-C port, microSD card slot (for extra storage), I2C and UART four-pin headers, and GPIO breakouts on the rear.
With a 466 × 466 pixel resolution, the colour AMOLED screen looks vibrant and has capacitive touch. Combined with the built-in six-axis IMU and an RTC with a battery header, it could easily be used for a wearable or pocket project like a fitness tracker.
While the MicroPython firmware is limited to a handful of code examples, the C/C++ firmware is far more comprehensive, including the use of LVGL (Light and Versatile Graphics Library) to render text, images, and an example GUI complete with animated graphs and a tiny pop-up keyboard.
A very slick RP2350-based touch display with a built-in IMU.
Pimoroni | £17 / $18
We reviewed the original Plasma 2350 back in issue 148, and now there’s a version with built-in wireless connectivity, opening up extra possibilities for controlling a connected strip of WS2812/NeoPixel or APA102/DotStar RGB LEDs. The optional starter kit includes a 10m string of 66 frosted LED stars.
It’s dead easy to get started. Just connect your LED string’s three or four wires to the board’s screw terminals, then use its USB-C port to access the MicroPython firmware, library, and numerous code examples on a computer. Impressive effects include falling snowflakes, alternating/random blinks, sparkles, fire, pulsing, and a sweeping rainbow. More can be found in some community resources, including an impressive demo with 77 effects.
To give the board access to your wireless network, you’ll need to add your Wi-Fi details to a secrets.py file. You can then try out web-based examples, such as setting the colour of the LEDs via the CheerLights IoT system on social media, or running a lighting effect that responds to the weather from Open-Meteo.
An easy and fun way to control a string of addressable LEDs.
Pimoroni | £8 / $9
If you need a really small RP2350-based board for a project where space is very limited, the Tiny 2350 is ideal. Measuring a mere 22.9 × 18mm, it really is tiny — around the same size as a standard UK postage stamp.
Along with a boot select button, the board crams in a handy reset button, RGB LED, and 4MB of QSPI flash storage, though there’s no wireless connectivity. A USB-C port is used to power the board, and to connect it to a computer for firmware installation and coding in MicroPython, CircuitPython, or C/C++.
The Tiny 2350 is available with or without pre-soldered pin headers. One obvious downside is that the number of pins is reduced compared to a standard Raspberry Pi Pico 2. There are 16 in total, including 12 GPIOs (plus another couple on the Qwiic/STEMMA QT port). They include four 12-bit ADC channels, though, and encompass two channels each of the I2C, SPI, and UART protocols.
A teeny-tiny RP2350 board that’s ideal for smaller projects.
WIZnet / The Pi Hut | £11 / $15
While many RP2350-based boards boast built-in wireless connectivity, sometimes you may need a wired Ethernet connection for improved reliability, security, and speed. Possible projects include a low-power HTTP web server, network monitoring, IoT data logging, and industrial devices.
WIZnet produces an Ethernet HAT to use with standard Raspberry Pi Pico boards. Alternatively, you can use one of the firm’s range of Ethernet-equipped RP2040 and RP2350 devices. This RP2350 one is based around WIZnet’s own W6300 chip: a 10/100 Ethernet controller with a hardwired TCP/IP stack that supports both IPv4 and IPv6 — as do the W6100 models, while the W5500-based boards are limited to IPv4.
All of them feature an identical pinout to Raspberry Pi Pico and Pico 2, with 40 pins and 26 multi-purpose GPIOs, so it’s all very familiar. The documentation includes C/C++ firmware/code examples for Ethernet, FreeRTOS, AWS, Azure, and chip performance. There’s no MicroPython firmware available for the RP2350 models yet (unlike the RP2040 ones); hopefully that will come soon.
Perfect for projects requiring a wired network connection.
The post Maker Monday: Some of the best RP2350-based boards appeared first on Raspberry Pi.
Raspberry Pi Connect lets you access your Raspberry Pi devices remotely from anywhere, straight from a web browser. Since we last wrote about Connect, we’ve shipped three updates that we think will make it noticeably more useful — particularly for the growing number of teams using Connect for Organisations to manage fleets of devices.
Once you have more than a handful of Raspberry Pi devices in an organisation, finding the right one quickly starts to matter. With these new updates, you can now apply tags to any device — for example, by location (london, cambridge), by environment (production, staging), or by what the device actually does (point-of-sale, kiosk). Tags appear underneath the device name on both the device page and the device list, and any administrator can add or remove them from the device’s Settings page, or when first linking a device to your account.
The search bar at the top of the device list now combines free-text search with structured filters. Type a qualifier followed by a colon — model:5, memory:4gb, os:raspios-13, or tag:production — and Connect will narrow the list as you type. You can even stack filters in a single query: model:5 tag:production dashboard will find every Raspberry Pi 5 tagged with production that has “dashboard” in its name. Selecting any tag in the device list adds it to your search instantly.
Tags are also exposed through the Management API, meaning you can apply them when you create an authentication key during provisioning — handy if you’re scripting the roll-out of a new batch of devices.
Connect for Organisations now lets administrators require all members to use two-factor authentication (2FA) on their Raspberry Pi ID. It’s a single switch in the new Authentication section of the organisation’s Settings tab, and it adds a meaningful layer of protection against compromised member accounts being used to reach your devices.
Turning it on starts a 14-day grace period. During that window, members without 2FA see a banner showing how long they have left and a link to enable it on their Raspberry Pi ID; everyone else carries on as normal. When the grace period ends, any member still without 2FA is blocked from the organisation until they enable it. They won’t be able to access devices or other organisation resources in the meantime.
If you ever want to relax the requirement, you can turn off 2FA in your organisation settings. (Disabling and re-enabling 2FA resets the grace period, so you can give members another two weeks if you need to.)
Connect’s screen-sharing interface works on phones and tablets as well as desktops, but typing on a touch device was previously only possible if you attached a physical keyboard. The screen-sharing toolbar now includes a dedicated Keyboard toggle alongside the existing buttons for Ctrl, Alt, Esc, and Tab, allowing you to use an on-screen keyboard on mobile devices without any extra hardware attached.
Connect is free for personal use. Connect for Organisations comes with a four-week free trial, after which you’re billed monthly in arrears for the peak number of registered devices. You can find full instructions for tagging and filtering devices, enabling 2FA, and everything else in the Raspberry Pi Connect documentation, or sign in at connect.raspberrypi.com to get started.
The post Raspberry Pi Connect: Device tags, required 2FA, and a mobile keyboard appeared first on Raspberry Pi.
The latest issue of Raspberry Pi Official Magazine has landed. Since our socials are still alight with Raspberry Pi–powered cyberdeck builds, we thought you’d enjoy this article by David Crookes, who seems to have a prophetic sense for what readers want. Please share links to some of the new cyberdecks you’ve seen online in the comments section. We love that this retro tech is enjoying a new lease of life thanks to the cool kids on TikTok.
Maker Jeff Merrick didn’t originally set out to create a cyberdeck inspired by the Alien movie universe. Instead, he sought to modify Penk Chen’s Penkesu retro-style handheld PC by fitting a hot-swap keyboard PCB, but his attempts to produce a writerdeck hit a wall. “I quickly discovered that a 40 percent keyboard isn’t ideal for typing, and a Raspberry Pi Zero isn’t ideal for browser-based text editors,” he says.
Scrapping the idea, he turned his attention to a new build, which he named the Typeframe PX-88. Inspired by the Epson PX-4 computer from 1985 and built around a Raspberry Pi 4, it featured a 65 percent mechanical keyboard and a touchscreen. It worked so well that Jeff decided he wanted to develop more projects with Raspberry Pi. “I had not worked with any single-board computers before (or Linux in years), but found Raspberry Pi’s large and established ecosystem made figuring things out easier,” he tells us.
Jeff had purchased a set of keycaps imprinted with the Semiotic Standard — futuristic symbols created by Ron Cobb for the Alien films. “I didn’t have a clear plan for how I was going to use them, but my wife and I are Alien fans,” he says. “We’ve built some cosplay and props, including the Incinerator Unit from Alien, the pulse rifle from Alien: Romulus, and a Facehugger mask.”
Since he still had his Penkesu build, Jeff decided to resurrect the project using the new keycaps. “I had most of the internals already working from my Penkesu build, and I just wanted to use them in some way,” he explains. “I also knew I wanted to keep it simple after designing my larger writerdeck, so I went with a slate form factor because hinges complicate everything.”
As such, his Penkesu evolved into the PS-85, with a new 3D-printed case and a Raspberry Pi Zero 2 W inside.
The parts and wiring are similar to the Penkesu, but the keycaps set the theme, leading to some fresh ideas. “There’s an exposed GPIO, so I knew I wanted some sort of peripheral that could slot in,” he says. “I found the Adafruit CharliePlex LED Matrix Bonnet, and this led to the addition of a light attachment.”
The build, which includes a Waveshare 7.9-inch touchscreen, wasn’t entirely smooth. “It was a challenge to learn everything on the electronics and hardware side — CAD and soldering,” Jeff says. “But since this was my second project, it went a little smoother than the first.” The results exceeded his expectations. “It’s actually much sturdier than I expected, and it boots up pretty fast,” he adds.
Even so, this build isn’t purely about the internals — a lot of time was spent on the overall look. “The idea of adding a handle came after an initial iteration,” Jeff says. “I also used the same weathering techniques that I had used on some other Alien props: essentially, I painted everything silver, then painted the areas I wanted to look chipped with a liquid mask (I used toothpaste), painted everything white, and finally removed the liquid mask to reveal the silver underneath. Then I added some scratches and an acrylic wash to finish it off.”
The PS-85 looks so good that Jeff hasn’t yet used it for its intended purpose. “My plan was to use it for journalling, and there is a nice command-line application, jrnl (jrnl.sh), that I installed,” he says. “But it’s mostly a display piece right now.”
If you liked this article, there are many more like it in the latest issue of Raspberry Pi Official Magazine. You can purchase a copy from the Raspberry Pi Store in Cambridge. It’s also available from our online store, which ships around the world. And you can get a digital version via our app on Android or iOS.
You can also subscribe to the print version of our magazine. Not only do we deliver worldwide, but those who sign up for a six- or twelve-month print subscription will receive a FREE Raspberry Pi Pico 2 W!
The post Typeframe PS-85 cyberdeck appeared first on Raspberry Pi.
