Notkia: Building An Open And Linux-Powered Numpad Phone

Two Nokia 1860 phones side by side - a Notkia-modded phone on the right, and an unmodified Nokia phone on the left

Many of us hackers have a longing for numpad-adorned mobile phones. We also have a shared understanding that, nowadays, such a phone has to be open and Linux-powered. Today’s project, Notkia, is the most promising and realistic effort at building a keypad phone that fits our requirements. Notkia is a replacement board for Nokia 168x series phones, equipped with an improved display, USB-C, WiFi, Bluetooth, and LoRa — and [Reimu NotMoe] of [SudoMaker] tells us this project’s extensive story.

The Notkia effort started over two years ago, because of [Reimu]’s increasing dislike for modern smartphones — something every hacker is familiar with. Her first-hand experience with privacy violations and hackability limitations on Android phones is recounted in detail, leading to a strong belief that there are fundamental problems with phones available nowadays. Building new hardware from the ground up seems to be the way forward. Two years later, this is exactly what we got!

When it comes to this phone’s physical form-factor, reusing an existing shell is the most economical solution out there, and the Hackaday.io page describes a journey towards finding a shell that fits. In the end, Nokia 1680 series phones turned out to be the perfect candidate. These phones are small and fit easily in your hand, there’s plenty of space inside the shell, and replacement shells and batteries are easy to get nowadays — at least, the kind of phone that you might want to get, anyway.

This replacement motherboard packs quite a few features. The old and laggy 128×160 display is replaced with an IPS screen with visible area of 220×280 pixels. They couldn’t find a small enough 4G module, but Notkia uses a LoRa module instead. There’s WiFi, Bluetooth, a Yamaha MA-3 music synthesizer, a USB Type-C port for charging and OtG, an RGB LED, an SHT20 sensor, and the 1680 version supports a 5MP camera. Such a feature set makes Notkia’s ambitious goal of producing a usable phone quite achievable.

Just like with the X1501 project we’ve covered, the Ingenic X1000 CPU has freely available and open datasheets. This phone runs Linux already – from here, software support work is ongoing, with an easy path to features like full disk encryption. There’s a series of demos: keypad input, LCD backlight dimming, LVGL music player, and of course, Bad Apple – with a USB audio adapter through USB-OtG. Drop tests were conducted, too. Interested in getting a Notkia board? [Reimu] aims to launch it on CrowdSupply – until then, there’s an email signup list to get project updates. If you’re interested in helping with one of the software priorities, it seems possible to get in early, too.

It’s reassuring to see a Linux phone with this much production potential. Projects to reuse old phone shells to get a viable feature phone have appeared every now and then. These Nokia 3310 and 3210 rebuilds have a few good ideas to borrow, and the WiPhone has successfully delivered on the ESP32 front with SIP calls. And if you’re looking to go even more DIY, you can always try to sandwich a Pi Zero between a few boards, or build an ATMega-powered phone with a PCB case!

42 thoughts on “Notkia: Building An Open And Linux-Powered Numpad Phone

    1. 3G/4G calls, to be clear – not with current hardware. The creators couldn’t find suitable 4G modules with low enough footprint, as the writeup describes; in other words, there’s just not enough space inside the phone for that, as it stands. Thankfully, there’s LoRa and WiFi for SIP, which is quite a powerful combo. Plus, I imagine it’s quite possible to make an external addon for 4G, like with an GPS addon being planned: https://twitter.com/ReimuNotMoe/status/1536695582693003265

        1. To make it clear, I’m not this project’s creator, I just happen to tinker with phones and related technologies a lot, personally, so here’s what I see.

          First, you are missing the LoRa part – it does provide for WiFi-less messaging! Messaging is what I personally use my phone for, most of the time. As for voice calls whenever they’re needed, a portable WiFi hotspot would satisfy my own needs. I can see how this would be limiting for some folks who rely on voicecalls more than I do; I carry a WiFi hotspot for my other devices anyway, so if I really need to talk to someone in a call, I will just use that.

          Personally, a Linux phone’s potential outweighs the LoRa and WiFi for me, and I’m eager to learn stuff like SIP (and write about it for everyone else!) if that moves me away from the numerous shitty aspects of cellphone technology.

          1. I’m not familiar with LoRa; Does LoRa have the coverage of 3/4/5G? Does it give full internet connectivity? Can it send messages via SMS? Or is this just going to work if your friend is within 10 miles with their LoRa phone?

          2. (I do not get a Reply “link” in the correct comment, #comment-6484893, nesting limit reached?)

            To Reimu NotMoe : 2G to 5G is provided by a telco. Internet wired service is provided by… a telco. Probably the same company, as in some countries you contract packages that include xDSL/cable/FTTx (with voice calls over that) and mobile service (which again includes voice calls and data transfers). Contracts that required ID in some places, even for prepaid. A Wifi hotspot would depend on such services to talk to “the world” (unless Arsenijs Picugins or anyother can say how their hotspots do not depend of cellular/wired service provided by a telco). Using free/whatever Wifi means more telcos, or random companies like hotels or shops, getting their nose in the middle. Contracting VoIP-only means a telco is involved, maybe just a different one. No idea who uses LoRa where I live, but if it is to just contact a telco provided end point, we are back to square one.

            Soooooo… in all cases there is “centralization” when trying to talk to “others”, just less or more layers over it. Extra hops for nothing are a waste, or even increase the failure points. Maybe you can stay not-centralized with modern equivalent of walkie-talkies, but then you are not contacting random “others”. Or maybe there is some not-centralized thing that lets me contact others (and be contacted back), but I never heard about. If so, please tell.

            Myself, and I guess some of the commenters, so far want a terminal we control, that works over the avaliable networks, and to be used to contact others that are using such networks and protocols (“plain” “old” voice calls, SMS, as well as Internet traffic). And a device we can make sure goes into “off” completly, or selectively (wifi off, bt on, for example).

            It’s OK if Notkia is not that terminal, or you can explain how it fits into the above. Meanwhile we will keep on searching, or compromising by using one terminal plus some kind of gateway (smartphone in hotspot mode, proper hotspot, etc), instead of a single device.

          3. Centralized somewhere, you forgot to mention the fact that the backbone infrastructure of the Internet belongs or is operated by various states, usually via state-owned companies. This basically means that governments can shut down the Internet inside their borders. And this actually happened a few times…

            The only decentralized communications network that can be out of state control is network of ham radio operators. And even these can be tracked and arrested by using radiolocation…

      1. I find it hard to believe that modern phones almost never have antenna sockets. This would be such a good feature and I’d be using my DJI drone to transport said antenna to 120 metres above my head to get reception for a 20 minute call or data connection in a cell signal dead spot.

        1. This might work if you’re near a cell tower but blocked by geology, but it’s not going to work if you’re just too far away from a cell tower. Cell antennas are optimised for reception on or near the ground, otherwise they’d be wasting transmission power blasting into space, and raising the noise floor for reception.

        1. Indeed!

          Here at Hackaday we are tolerant, accepting, and some might even suggest – celebratory, towards lameness, pointlessness, time-wasting, fail-boating, money wasting, amateurishness, rank incompetence, and even a little bit of self aggrandisement.
          As long as you got off your arse and did _something_ we’re good with it.

      2. Between you mentioning Python where the author quite clearly is using C and LVGL, referring to a 2G modem when the discussion is about 3G/4G ones, or referring to the author as “he” given I’m using “she” … it’s apparent you haven’t read either my article or the project description.

        I don’t know what to answer to this. Are you sure you’re in the right comment section?

      3. It seems you missed the reasons for the project entirely, perhaps go read the project build log.

        As for “no idea about hardware and software development for embedded devices”, have you considered that the best and brightest embedded systems developers learn by doing and don’t fall into the trap of reading textbooks and dismissing ideas out of hand because of biases you’ve picked up from external sources.

        Is this project without fault? No, clearly the lack of a cell network radio is holding it back for ‘some’ use cases. However, there is always room for improvement and a next iteration. Unless of course one chooses to do as you do and dismiss things out of hand because you “know better”, in which case the project never even gets started.

        For most geeky-types, ie the ones that would build such a device, LORA, WiFi, and BLE is more than enough connectivity.

        And finally, as developers, it is useful to abstract away things that don’t matter to your application and the choice of Linux here is actually a benefit. Using Linux means access to well written and well reviewed code for networking. If you goal is privacy, security and explicit understanding of how your data is used on a device, then Linux offers a good starting point without falling into the trap of having to be an expert in all things x, y, and z in order to accomplish w as well. Also, lets not forget that C and Linux are intimately connected, Linux != Python and in my opinion this is a really myopic way to look at systems in the real world.

      4. Pretty sure GSM is dead in the US thanks to AT&T being done with their towers this month…. So how does GSM work as a form factor?

        Not to be negative, I’m just making sure that the comments line up with the realities and are not misplacing time and location.

      5. I’m the creator of the project. First of all, I learned C and x86 assemble by myself when I was in middle school. I have my own SDK and toolchain recipe for Microchip PIC32 MCUs. Our company, SudoMaker, is Microchip’s Authorized Design Partner. Currently I’m working on the missing Linux drivers for this project. I think you don’t have good feeling of scales and dimensions – please at least Google this Nokia and see how small it is. I guess you are living in a country that still have a bunch of 2G networks running. Do you know the proper method to multiplex PPP data and AT commands on the same UART without Googling? At last, please don’t assume people’s gender without asking them. Thank you very much. :)

      6. I mean, most of this comment is just flat out wrong and/or stupid, but the thing that really sticks out to me is “why bother?”

        My dude, do you realize what site you’re on? If you’re asking “why bother”, I think the better question to ask yourself is “why am I here looking at this site and interacting with this community”, because you seem to have entirely missed the point.

    2. Yes. It can make decentralized phone calls using LoRa. Just imagine a better walkie-talkie with digital voice encoding, FEC, encryption, and without the PTT button. LoRa is nothing more than a modulation scheme, just like the OFDM in 802.11.

      1. Have you done this in practice or is this just theory? I’m genuinely interested – I’m currently successfully using LoRa to transmit GPS correction data for agri-robots.

        1. Yes we have. We’ve got bi-direction realtime audio working on a single frequency, with only ONE conventional 433M LoRa module on each side – not LoRaWAN! Also we’ve got ethernet over LoRa working, ping can go through without an issue. Actually our LoRa networking solution has already been used in some industrial situations.

      2. Thanks, I think I understand. It’s a cool project, and I LOVE the form-factor, but I think calling it a phone is misleading if it doesn’t handle make/receive voice calls or SMS to a phone number.

        A lot of people are looking for an open-source device which does that, which is where a lot of the comments are coming from.

    3. well, I totally get this phone

      if we can get more LoRa satellite repeaters in orbit then we can have an SMS-like messaging com ability that totally bypasses the proprietary, totally locked down cell phone networks. And solar-powered LoRa repeaters could be relatively cheaply put up around communities or community areas (just start putting a repeater on the roof or in the attic) to where there’s a free, public com messaging system. And LoRa repeaters could bridge into the Internet and tie communities together across great distances. And LoRa repeaters could be quickly put up that have a bridge to BlueTooth to where conventional cell phone users could then participate too if they install the appropriate app on their smart phone that establishes a Bluetooth mesh network (that gateways into LoRa).

      IOW, the cell phone infrastructure will always be locked down, proprietary controlled com medium, but LoRa could be a basis for establishing an open, text messaging wireless network. Maybe all this doesn’t exactly exist (i.e., Bluetooth mesh networking protocols and gateways of Bluetooth to LoRa), or a solar powered LoRa repeater to mount upon on the roof or nearby tall tree), but that just means there are more cool projects to spin up. And maybe there could be crowdfunding efforts to get more LoRa satellites in orbit.

      IOW, ordinary people could, over time, bit by bit, build out their own public wireless messaging network that ultimately spans the globe.

  1. Hey this is a cool project! I love it when people repurpose existing cases with new electronics.

    It feels like mechanical design is one of the hardest things to get down as a home hacker. Vendors have had a long time to optimise their designs mechanically, and good mechanical design doesn’t age, so why reinvent it?

    1. Mechanical design is a right bugger. I wish I could find a 3D printer tool that had a library of common “orifices” such as USB A, DB9 and so on that could then be positioned on the face of a cube as desired. That would cover 90% of my enclosure designs; the approach would fit quite well with openscad. I am pretty sure something like this must already exist and I am just ignorant…

      1. I like that idea, in fact you inspired me to look for something like that. Seems some of the OpenSCAD libraries could provide a promising solution.

        Particularly this one:
        https://github.com/nophead/NopSCADlib/blob/master/readme.md
        Although I’m sure there are others as well. Looking over the github repository it does seem to support USB A, although I don’t see DB9s (although, you only have to add them once and you can keep reusing them).

  2. Oh this is nice! There are some frankly depressing and rude comments about this device. Balls to them. This is a great hack nicely designed and if it suits your purposes, perfect!

  3. I’m an old school curmudgeonly old gray haired ham operator.
    I have to say this is impressive. Excellent design.
    In ham radio, there a a lot of digital modes for data.
    Two for digital voice are D-STAR and DMR. If there are no 3/4G modules,
    perhaps maybe a 5G one? I see here in the USA that a lot of networks
    such as Verizon, TMobile AT&T etc. are going 5G these days.
    I have an LG G2 smart phone and basically only use it for phone calls.
    I don’t do internet on it. It replaced a Motorola RAZR which was my
    very first cellphone. That phone was tiny, did exactly what I needed a
    phone to do. Make and receive phone calls nothing more. Once 5G has a hold
    on everything I expect the 3/4G networks to shut down, although I’m not
    sure why those frequencies couldn’t be reused for 5G instead of using
    another band. People way above my pay grade can answer that one.
    Hams that participate in emergency communications for when the zombie
    apocalypse happens or whatever have what’s known as a go bag.
    One thing I have in mine believe it or not is an old Princess Trimline
    rotary phone. I can hook it up to any phone jack and while that may be
    very very old school, I’m of the school of thought that simplicity is best.

  4. The phones started sucking about five or more years ago. Now all phones really suck hard!

    No physical number/key pad, having all the needless glitter and eye-candy features! Removing the number/keypad was obviously a cost saving measure versus anything of necessity! Touch screens are horrid to use! Funny, phones still do not include a clicker for simple counting, but do all usually include calculators!

    The recent supposed 4G upgrade now sucks 4-5x’s more battery power from my current mobile/cell phone. And with each firmware upgrade, the phone tends to suck more power down it’s synthesized copper tubes.

    If you can get 4G going, you’ll likely see an exodus of (mostly intelligent) people buying your product, until a few of those currently exploiting mobile/cell (or any type of) phones gets mad.

    Surprising landline phones, after 20+ years of the age of the computer, still do not incorporate a USB port for synchronizing caller ID or phone number directories!

Leave a Reply to Arya VoronovaCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.