A DIY Watch You Can Actually Wear

(hackster.io)

84 points | by sarusso 2 days ago

15 comments

  • jblezo 2 hours ago
    That's more a programmable watch than a DIY one :-)

    I build mine from scratch, including the PCB and a 3D printed case.

    For sure, that's not at all the same level of customability, programmability, capacity, nor quality. But It is really a DIY one.

    For anyone interested: https://github.com/jblezoray/hpdl1414-watch

    • Guestmodinfo 1 hour ago
      I like your hpdl watch better than the LilyGo watch. As a lover of analog and casio watches I can tell you that your watch can soon build you a loyal following in its current form if you only build 3 such watches and put them on Amazon, etc. and price them at 50$. Next batch of 3 you could try at 100$. Don't worry about glass cover, our local watch repair guy can easily slap one in it
      • hrimfaxi 1 hour ago
        The HPDL-1414 appears to be discontinued though still available from some suppliers in varying MOQs and leads.
  • pickleglitch 16 minutes ago
    This is almost $80. The PineTime watch is less than half that price. Obviously the specs are different but that's quite a difference.
  • oritron 2 hours ago
    I like a good smart watch and I appreciate open source, but an ESP32 isn't a great pick when low power consumption is important and the device is going to be communicating regularly. I'm surprised LILYGO went that direction in a watch form factor.
    • smlacy 1 hour ago
      What would you suggest instead?
      • oritron 1 hour ago
        Nordic Semi, or maybe ST Micro. I've got an STM32WB on my bench at the moment with sensitive coulomb counting and it looks very promising but without all those radios. Of course with all those radios (ie, if you need LoRa on a watch... which is a design decision I'm also skeptical of) then Nordic has a good track record.
        • jkestner 1 hour ago
          Different use case (environmental data recorder) but our current product uses an STM that turns on an ESP32 (not trusting that sleep mode) when it needs radio, to run on 2 AAs forever.
          • rylando 48 minutes ago
            Do you have any info you can share on this recorder? Sounds really interesting!
    • hrimfaxi 1 hour ago
      Most of their lora devices are ESP32.
  • Retr0id 2 hours ago
    It's cool that the firmware is hackable but I think "DIY" is an imprecise way to describe that.
  • briandw 1 hour ago
    No mention of battery life? I guess it depends on the software that you run. But it would be nice to have a benchmark for how long it would last in normal watch mode.
  • MrDrMcCoy 1 hour ago
    There are only a few features I care about in a smartwatch:

    1. O2 monitoring. I have sleep apnea and live at high altitude, so this matters to me.

    2. Motion sensor. Also mostly for tracking sleep.

    3. Vibrator for notifications.

    4. A screen backlight.

    5. Battery life longer than a week.

    6. Waterproof enough to survive a splash in the shower/rain.

    I consider GPS, cellular, AI, touchscreens, cloud-only sync and control apps, and just about everything else to be anti-features. There are no devices that really cover all this that I've found. A few Garmin and Amazfit/Zepp devices come close, but they have enough drawbacks for me to not be happy with them. The new Pebble is nearly perfect, but the lack of an O2 sensor is a dealbreaker for me :(

    • Findecanor 1 hour ago
      The Sensor Watch circuit board [1] inside the case of a Casio F91-W / A158W / A159W satisfies 2, 4, 5, and 6. Accelerometer or thermometer available as daughterboard. Battery life measured in months, if not years. Although the simple LED backlight and the segmented LCD leaves a bit to be desired, and there is no wireless connectivity for notifications. Open source firmware.

      The Ollee watch circuit board [2] is similar, better backlight but closed-source firmware and configuration over BLE in a smartphone app. Still no notifications over BLE though.

      I'd think combining 1 and 6 (O₂ monitoring and waterproofing) would be difficult.

      [1]: https://www.sensorwatch.net/

      [2]: https://www.olleewatch.com/

      • doctorpangloss 6 minutes ago
        my experience with a sensor watch has been terrible.

        imagine breaking a $3 watch that is not quite as indestructible as people think it is, but it is nonetheless pretty robust, and then trying to shove something 100x glitchier and 5x as expensive into its case...

      • curiousgal 59 minutes ago
        The Pro version has a better screen (still segmented but more), RGB LEDs and an infrared sensor.
  • gitowiec 2 hours ago
    This device looks capable of a lot of features and possibilities. Unfortunately nothing comes to my mind because I'm not good with diy hardware (once connected raspberry pi zero with led strips). Could someone tell examples of interesting and/or useful projects one can implement with this watch?
  • inasio 1 hour ago
    Does anyone know if this has an accelerometer? I recently got a nice sports-oriented smartwatch (non-Garmin), to use it mostly for rowing, but it doesn't track the rowing-rate. It should be pretty easy to program one if the watch has accelerometers, but couldn't tell from the spec sheet (maybe that means no?)
    • fjfaase 1 hour ago
      It had a Bosch motion sensor with AI abilities it says in the description.
  • HardwareLust 2 days ago
    LILYGO site shows pre-orders of all 3 versions are sold out unfortunately.
  • gamerslexus 3 hours ago
    s/Watch/Smartwatch

    Regular DYI watches aren't big news...

    (I would be over the moon for a DIY smartwatch with zero AI and e-ink screen.)

    • stackghost 2 hours ago
      I would consider a DIY mechanical/analog watch to be far bigger news/more impressive than a smartwatch.
      • bloggie 2 hours ago
        To be honest there is not much to it, you buy the movement, put it in a case, and put the hands on it. you can get everything from aliexpress. it's easier and often cheaper to just buy a normal watch if you need one.
        • Avicebron 2 hours ago
          It's impressive you start with a lathe and make the movement yourself!
          • saltcured 1 hour ago
            Not nearly as impressive as designing and fabricating your own integrated circuits and display!
        • NooneAtAll3 2 hours ago
          buying movement is like buying whole PCB

          DIY analogy would probably be about acquiring individual gears

          • gamerslexus 2 hours ago
            Is it different with a smartwatch? You buy the kit, it's not like you solder much as far as I understand.
      • gamerslexus 2 hours ago
        I thought so too, but after quick research apparently there are kits. For various values of "DIY", I guess...
        • stackghost 2 hours ago
          Sure but buying a movement kit is no different than buying a pcb. Writing code is not impressive any more.
  • jwr 2 hours ago
    This does look very cool. Every peripheral one could think of, even LoRA!
    • hrimfaxi 1 hour ago
      That stood out to me, too. Garmin should take a hint.
  • avipars 55 minutes ago
    trun on and trun off
  • dariosalvi78 55 minutes ago
    No heart rate sensor
  • ImPostingOnHN 2 hours ago
    Preorders sold out already!
  • ck2 3 hours ago
    have wished for decades now there was an open-source Garmin on the level of Cyanogenmod / LineageOS for Android

    not sure if it will happen this decade but definitely next decade

    proper running/cycling metrics are hard as demonstrated by how many well-funded competitors are somewhat close but not there 100% yet (Coros, Amazfit, etc)

    someone once hacked and decompiled older Garmins but newer ones are encrypted/signed/locked-down

    • m463 2 hours ago
      > newer ones are encrypted/signed/locked-down

      I have a garmin watch and didn't know this.

      That said, I just used it out of the box, and never (on purpose) hooked it to wifi, bluetooth, garmin connect, etc. Can't do that with an apple watch.

    • rjsw 2 hours ago
      Have you looked at the specs for the upcoming PineTime Pro [1]?

      [1] https://pine64.org/2026/03/28/pinetime_march_2026/

      • xrd 2 hours ago
        I'm very excited about this. GPS was the final piece of the puzzle.

        I love(d) my bangle.js. Such a true hacker device. Really fun to use WebUSB and push JavaScript files as apps.

        But the GPS on that device was a mess, honestly. I know this is a complicated problem but having to synchronize to satellites and recalibrate all the time was beyond me.

        I really wanted it to work because I built my own toy run tracker visualization tool.

        I am curious about this new lilygo device because it sounds like it has an alternative location sensor: "A u-blox MIA-M10Q GNSS module provides accurate location tracking..."

        I'll need to look that up. Anyone have a summary on what's the difference between that and regular GPS?

      • branon 55 minutes ago
        Oh nice, didn't realize they were doing a second one. Loved the original but I took mine rock climbing and cracked it :(
    • mghackerlady 3 hours ago
      I have a garmin from the late 90s and am saddened by the lack of FOSS software to even sync a new map onto it
      • ck2 2 hours ago
        not sure if this will help you but there is a neat website that allows you to build free maps for older Garmin models that didn't have them at first like Fenix5

        https://garmin.bbbike.org/

        1990s is going way back though, they didn't even have mass-storage mode then, it was their proprietary "garmin mode" for usb which only things like BaseCamp can talk to

        • mghackerlady 47 minutes ago
          oh bud, mine doesn't even mention USB in the manual. I got the thing for like a dollar at goodwill haha