[David] sent us a tip about a company in Belgium, Citronics, that is looking to turn old cellphones into single-board computers for embedded Linux applications. We think it’s a great idea, and have long lamented how many pocket supercomputers simply get tossed in the recycling stream, when they could be put to use in hacker projects. So far, it looks like Citronics only has a prototyping breakout board for the Fairphone 2, but it’s a promising idea.
One of the things that’s stopping us from re-using old phones, of course, is the lack of easy access to the peripherals. On the average phone, you’ve got one USB port and that’s it. The Citronics dev kit provides all sorts of connectivity: 4x USB 2.0, 1x Ethernet 10/100M, and a Raspberry Pi Header (UART, SPI, I2C, GPIO). At the same time, for better or worse, they’ve done away with the screen and its touch interface, and the camera too, but they seem to be keeping all of the RF capabilities.
The whole thing runs Linux, which means that this won’t work with every phone out there, but projects like PostmarketOS and others will certainly broaden the range of usable devices. And stripping off the camera and screen has the secondary advantages of removing the parts that get most easily broken and have the least support from custom Linux distros.
We wish we had more details about the specifics of the break-out boards, but we like the idea. How long before we see an open-source implementation of something similar? There are so many cheap used and broken cellphones out there that it’s certainly a worthwhile project!
Great idea! I would throw in a capacitive touch PCB that could connect to the phone’s interface, to make it easy to add controls to a project. Even if it’s not identical to the original glass, it should send back some kind of consistent data, and a quick calibration step would result in some simple buttons/sliders. If it had a mostly solid silkscreen layer, a custom keyboard layout could be doodled on with a sharpie.
I have thought about this for many years. Unfortunately Linux support and even the ability to unlock the bootloader to install anything other than the official OS image is really bad with most phones.
This seems like the perfect project to apply to old “effectively ewaste” tablets based on allwinner and rockchip chipsets though.
The way these folks are doing it is to start with a model that it’s relatively easy to work with. (In this case, trivial, b/c it’s designed out the gate that way.)
But if you picked a phone with good PostmarketOS support first and then designed the carrier board for that particular model afterwards, I think this could work.
It’s not a one-size-fits-all solution, though. It’s a custom-to-the-phone thing.
The breakout board is certainly a nice bonus; but, especially with newer phones, even ‘just’ the USB port is a lot if the software issues can be dealt with.
Fairly likely to be USB3 (not necessarily capable of actually hitting full rated bus speed; but ‘better than USB2′ counts for a lot given that things like gigabit ethernet fall on the low side of USB3 but over double USB2); sometimes even DP alt-mode; plus, software willing, most old/’dead’ phones are not being discarded because their BT/wifi failed.
I could definitely see a case for both(if nothing else there are a lot of phones that get discarded because of screen damage; which are obvious candidates for something more useful being put in the screen’s place); but for a lot of purposes I’d probably rather have the screen intact and make do with the existing peripherals: SBCs are relatively cheap; but adding a screen that isn’t a little teeny i2c/SPI LCD or an oversized HDMI display tends to be annoyingly pricey; while phones with dead batteries stuck on old android versions have just strikingly nice screens by comparison.
I had some success with this many years ago. I started by porting postmarketOS to an old phone (which was quite easy as I could use the LineageOS kernel sources and just enable some extra kernel options and didn’t care about display output). After this the phone ran a somewhat standard Alpine system which could run Docker IIRC. I remember the benchmarks I tried ran circles around the (at that time) newest Raspberry PI, even though the phone I had was probably 5 years old.
To connect the phone I bought the absolutely cheapest USB OTG usb hubs I could find + some usb to RJ45 adapters. My “broken phone cluster” project fell apart when it turned out all the adapters had the same MAC address…
This is awesome. Would love to know how to do this. I’m building a mini rack with Pi’s, and tinyminimicro nodes. Would love to be able to include the last-gen phones I’ve accumulated from myself, wife, and family. I’ve been avoiding sending them to the questionable ewaste recyclers.
Nice but the board looks very specific, there’s what seem to be battery connectors to fit a specific position on the cellphone board and some spring loaded multi way connectors yet there’s no info as far as I can see about which smartphone boards are supported.
It looks like they’ve worked out one model from one manufacturer which at €150 for the adapter board plus the cost of a second hand (and potentially unreliable, abused) smartphone means you’re better off with one of the established SBCs or more reliable and “proper” industrial controls.
And realistically, who in the industrial world gives a damn about price when reliability is far more critical…
Ya gotta start somewhere, I guess. Maybe they’re working from specific to general, or planning a series of these to cover different phones.
I have a few old phones hanging around. It bugs me not to be able to use them for anything, but doesn’t bug me enough to actually do something about it. And, if the price of the interface board is high enough, I still won’t do anything about it.
I have a collection of old phones, Galaxy Note 2, Samsung s7, I’m wondering how your going to handle so many phone pcbs, brands?
That’s the point, though. They’re not trying to make a universal solution, but rather just picking a starting point and starting.
I imagine that the skills of designing around one phone will help a lot for designing around the next, but since there is very little that’s generic across phones, it’s simply impossible to make something universal, AFAIK.
But if 20 people made phone-specific designs for 20 different phones…
Maybe a simpler way to get there is to design and plug in a USB IO expander, and have a keyboard on that, and an android app to enable it. Most phones will work OK with no battery if you solder a fat decoupling cap in its place and leads to a 4V supply. Then you’ll still have access to the touchscreen. Contrary to previous comments I think most android phones can be jailbroken/allow root access.
https://www.hackster.io/news/one-man-s-trash-is-another-man-s-compute-cluster-b6735c5d8385 There are numerous CPUs in all sorts of devices. There are hacking efforts on many game consoles, but when a new model comes out the efforts start to fade. Same with old routers openwrt has grown beyond wrt54g. Hackaday seems like a great place to keep these project together.
Most cell companies seem to hate letting their customers root phones, my last 3 phones have been unrootable through the methods I have access to.
That’s too bad.
That’s why I went for a non-carrier Google Pixel last time. As twisted as it is Google still releases probably the cleanest and least restricted phones.
It was really easy to swap over to GrapheneOS for a more privacy/security focused OS.
That plus Google having some of the longest guaranteed updates means that it should be useable for like a decade.
They are really the only consistent one, Samsung has been locking down even harder every major release and many others didn’t work well. The real kicker? Ac lot are still run ancient kernels despite Google’s attempts to make it as easy as possible to use something current. Some vendors ship new phones or other devices with 2.6.
I agree… but the methods I’ve seen work lately involve adb debugging and money $ ….it’s actually clever but it works. Why could u root my galaxy s23 U? Jw that would be great
or this one https://www.theverge.com/2015/1/25/7891771/puzzlephone-modular-smartphone-concept-turns-cpus-into-a-supercomputer
Termux
Uses the whole phone as it is. Add some usb periphs…Bob is your uncle.
https://termux.dev/en/
The number one issue I see preventing more reuse is software support. I’ve had devices that despite unlocked bootloaders just fell out of favor enough that LineageOS dropped them and I’m not sure what that status for Linux on old Android devices is but last I looked it was also pretty specialized.
The Fairphone itself is an interesting concept.
Fairphone phones are modular so they are easy to repair. Fairphone commits to supplying replacement parts for the phones for much longer than is usual. Fairphone sources all the components from countries and companies that treat the workers properly.
The Citronics folks probably started with the Fairphone 2 due to the availability of spare parts and complete new modules. Fairphone is on Model 4, now. The support period for the Model 2 is expired (since 2023,) so the Fairphone folks could make a deal to sell off their spares.
Fairphone (as a company) supports bootloader unlocking – no hacks, just ask for an unlock key.
The phones are well made. My Fairphone 3 has held up better than the previous 2 (or three) phones I had that all croaked for various reasons.
Battery replacement would be a breeze if I needed to – pop the back cover off, slap in a new battery. No hot air guns, no fiddly screws, no glue. No trouble with being waterproof, either.
Phones would make great SBCs with the right software support.
I have converted an Exynos S20+ into a general purpose Linux server. But it was one heck of a project (probably would make a great Hackaday article on it’s own) which involved compiling and/or adapting a lot of software to work under existing Android OS without breaking it, since you can’t use a lot of device’s features without Android. And to work around quirks to make things like Docker and LXC to work.
S20+ Exynos has about as much CPU grunt as Raspberry Pi 5 + Pi 4 combined. It is a powerful device despite being quite old.
But it takes a lot of effort to use it as an SBC.
“But it takes a lot of effort to use it as an SBC.”
I think that’s the crux of it, the old joke ‘I wouldn’t start from here if you want to get to x”, there’s any number of smaller, neater SBCs out there which have better support and are, crucially, much cheaper than this adapter board alone.
Plus, an S20? Hells, I’d still be using my S10 if the battery was replaceable ;)
Wow! You really need to pen it down. Maybe take it to twitter or at least share some documentation on how you took it out.
You can run a Linux server instance on nearly ANY Android 5.0+ device if you have root. Using a chroot means there are no other device, model, or hardware requirements.
The Pi Deploy APK installs the Pi-hole ad-blocking DNS server on first-run, but it’s a regular Debian install you can use as a base for just about any project you like: https://github.com/DesktopECHO/Pi-hole-for-Android
Turns it into what??? Bad, bad writing to not define initializims.
We don’t always explicitly define “commonly known” acronyms, and especially not in headlines. I was on the fence about this one, too, so I tossed it in full in the first sentence.
How many other people were confused?
Single Board Computers, the majority of the audience that HAD draws should know the acronym (or at the very least how to google since it’s a super common one in the technical space)
it’s hard to justify, because a purpose-built SBC is so inexpensive. i like to do it on an ad hoc basis, as the phones i have go out of service. but usually, it went out of service for a reason. and on top of that, how many SBCs do you need? there are use cases (hello Foldi-One), but they’re fairly rare (especially compared to the ewaste stream we can harvest) and the ones highlit here are often contrived.
the other thing is, the phone actually has a great set of peripherals. one use case that does often make sense is using an old phone as a wifi-connected camera. which often doesn’t take any big hack at all. there are a few other uses…always-on wall-mounted display, alarmclock, wireless speaker, etc. but once you start getting into those uses where repurposing it is hardly any work at all, you really run into the plethora of reasons you stopped daily-carrying that device in the first place.
very hard to find a middle ground where it’s not so much work to reuse it, but the defects that made it unusable as a phone don’t rain on your parade.
I sent the news to Jean-Luc of CNX back in March, after their CEO was on belgian TV before MWC in Barcelona:
https://www.cnx-software.com/2025/03/02/citronics-built-a-router-based-on-the-fairphone-2-mainboard/
Wow…..we want to stop phones from being recycled, so let’s pick the most obscure and rarely sold phone in the world to make a board for! No way we should make it for a pixel or a Samsung that sold millions of devices, that would actually be useful!
That’s the goal indeed, but there are a lot of parameters that come into play when starting this kind of business. You need to start somewhere, and a phone that can easily be disassembled make that step at least easier in the beginning. Moreover, sourcing the phones was easier since Fairphone takes them back once thei are unused for future recycling so you have access to several thousand phones from one single company. Sourcing other phones can be more difficult since there might not be able to provide a large stock of the same models.