How to Build a Smartwatch: Picking a Chip
ericmigi.com
ericmigi.com
It also powers the Fallout Pip boy and possibly some other stuff from the wand company: https://www.thewandcompany.com/fallout-pip-boy/ See: https://github.com/orgs/espruino/discussions/7577
also everyone watch the Amazon Fallout show if u have the chance its p good
AFAICT, the BLE code is provided as a binary blob. https://github.com/OpenSiFli/SiFli-SDK/tree/6c82a9b15db49871...
Which isn't a problem. But, I wish if something is described as "open source", you could read the source code for it.
Now one might say that's the fault of the person doing the modifications/manipulations, but regulations in various countries require the device to prevent these manipulations.
(N.B. I'm not in the exact business, but that's my take away from looking into the topic some time ago)
Technically they could publish the source in that case but I think some patents would prevent them from doing this. Radio stuff is riddled with patents. Also most likely the "why would we" reason. There is no benefit for the manufacturer.
If it's concerning their own patents, no need to hide the source code. A patent is literally "letters patent" or public description of an invention. Trade secrets on the other hand I could understand.
I'm surprised how much they were able to open source as-is. I think part of that is that the SoCs powering the raspberry are kinda 'old news', definitely not the bleeding edge kind of embedded tech. The first raspberry basically happened because Broadcom had a whole bunch of old chips they weren't able to sell. Only when it took off they started making some actually for the purpose.
> Seems like security by obscurity.
There are legal standards to meet around protecting the system from users modifying things like output power. They don’t literally say that the source code must be private, of course, but keeping it closed source makes it much easier to demonstrate that you were not making it easy for people to exceed the regulatory limits.
If you document the registers for setting output power (for example) then you’re giving the end user a roadmap for changing output power.
There were a few more reasons too, at least at the time. The companies in the space didn't have a culture of open sourcing, and there's probably no perceived commercial upside to releasing code for a chip like that.
Then again, lots of Wi-Fi-enabled devices support simply changing their region setting and will happily let you use different Wi-Fi bands or increase signal power.
On software-defined radios you can often use them way out of spec, way more so than using a forbidden channel. But in a totally different band. A good example is the RTL-SDR stick which was designed to be a TV received but can be used as a wideband SDR these days. That's a receive-only device so it's not that critical to regulators. Once you can transmit, it becomes more of a problem.
An example of a more problematic transmission device is the Raspberry Pi PWM pin. That's been used to transmit all sorts of stuff on many bands. Because it was never designed to transmit anything, it causes all kinds of weird harmonics and artifacts. It's a really bad idea to use it for that. Most people just do it under controlled circumstances.
"We say it's open source because we expect the reader to know that we're not telling the truth"
should be replaced by
"It's open source except for the BLE firmware blob, which can't be open source due to regulatory reasons."
To be fair, the article just repeated the claims made on the GitHub page for the SDK.
The Bluetooth SIG requires that you qualify your device if you advertise that you use Bluetooth IP, similar to what is required for the cellular space. Do you have to do this if you’re just “Bluetooth compatible?” Maybe not. Whatever the case you have to conduct FCC part 15C testing (intentional radiator).
Maybe soon you won’t.
The device needs to operate in spec and if someone reflashes it with out of spec firmware, that's no different than someone soldering a different resistor onto the PCB or feeding the output of the chip into an amplifier and large antenna. It's a modification by the user and so the user is liable for operating out of spec. And all of this really unrelated to source code, just to reflashablitiy. The code could easily be open, but the productions device could be made unflashable if the law really required it. Yet this is not what we see here.
It's IP reasons, and that's fine, but let's not make up additional excuses for them.
Excellent point. Many people (myself included) do assume that open source = redeployable, as in GPL3. I guess an e-fuse for read-only is perfectly OK, both from regulatory standpoint, and for ensuring you're running the same code as what you have the source for. That would be cool.
I would really miss the mod cons I have on my current smartwatch. Payments over NFC, dual-band GPS tracking, 4G LTE connectivity. The Pebble (and repebble) trades all that in for a multi-week battery life. But the minor hassle of once-every-two-days charging on my galaxy watch really isn't enough to forego all these powerful features for me.
Everything is a trade off.
But these days I just have bigger requirements. The whole heartrate thing I don't even care about. Notifications are the main thing, but route tracking, integration with my alarm system and above all payments are just things I can't do without anymore. I had an Amazfit for a while which also had a 3-week battery life (though no apps) but it missed the payment option.
The charging is not a big deal for me because I only wear my watch outside anyway. And it charges in 30 minutes. It makes it easy for me to keep it charged also because it becomes a routine. With the amazfit it was often empty when I needed it.
But yeah it is great that there are devices for everyone.
Maybe you mean types of notifications that I'm not used to using, because notifications were the first thing I would turn off on any smart watch I've used (pebble, Amazfit, galaxy). The last thing I need if to have yet another device (now being worn) vibrating every 5min for some message.
Truth be told I use the least smart feature of the watch: the alarm. I can set quite granular schedules for my alarms and that means no one else needs to wake up but me. The health tracking I barely care, but since I'm wearing it I track it. I could imagine using the payment feature tho, to avoid taking the phone out.
At least on my smartwatch (Samsung one) I can choose which notifications I want to forward to the watch. And I have all the ones I don't care about turned off anyway, even on my phone. For me it replaces a lot of my smartphone (which I only use outdoors anyway, at home I have a real computer). The ability of the galaxy watch to also show pictures in notifications is a big plus (my old pebbles could not do that, not sure about the new repebbles)
The notifications on my watch are the main reason I have it. It avoids looking at my phone every time something comes in. It's not too often because I block so much and I don't use social media.
I don't care much about the health tracking, though I do sometimes use the sleep tracking. Especially the SpO2 tracking is handy because I have apnea. It does bother me that Samsung doesn't make it possible to turn off their health "gamifications" (every day I get nonsense messages like "Will you meet your goal today?", "You're halfway your goal, keep going!"). I still have to see if I can turn that crap off via ADB.
With watches that need charging every one or two days, I get this Sisyphean feeling that I am owning a constantly dying device with a battery indicator that exists solely to be charged. The amount of functionality a watch can provide me (versus a smartphone) just doesn't justify the amount of charging I need to do.
Maybe I'm not a smartwatch power user like OP. But the Pebble and its ilk are a great fit for the niche audience I belong to.
https://www.garmin.com/en-US/p/851039
No one really needs payments on their watch but I'm surprised that anyone wouldn't want it. It's quite convenient, and gives you a backup option in case you forget your wallet or something.
I love having it on my watch, it's just so useful. Especially because I live in a city full of pickpockets. No need to take my phone out.
From the CPU perspective, they are the same
Depends!
If the two chips use UART or SPI for intercommunication, okay, you need two lines between the CPU and two GPIO lines for wakeup, and JTAG can be shared anyway.
But if you use stuff like shared memory, or want to do stuff like updating the display not just from the high-power chip but also from the low-power one, suddenly design becomes much more complex.
Personally these days I would lean towards the ESP32, they continue to iterate on it nicely and it has great community support. I'm personally developing a smart watch platform based on micropython.
Getting down to 10mAh is not so bad. If you're not actively driving the display, you can under-clock significantly [1], if you're not using WiFi you can turn the modem off [2].
[1] https://docs.espressif.com/projects/esp-idf/en/stable/esp32/...
[2] https://docs.espressif.com/projects/esp-idf/en/stable/esp32/...
Also 20 hours of runtime is horrible.
There are many ESP32 variants, depending on what you pick some may be more compelling for your use-case.
Even some newer variant like S3 or C6 only has acceptable power consumption, if what you are after is run-time they are not the best fit.
PineTime, based on NRF52, will get you 4-7 days of practical usage.
Recently they release ESP32P4, with very strong performance, but like you guess, without Radio
I think once we start talking about GPU, MMU, USB, display, etc, we're getting towards a CPU of sorts.
Speaking of a low-end CPU, I want to test out the RV1103 Rockchip, those crazy little chips are running Linux apparently [1], and even able to run Python [2]. Depending on power draw, a Linux-based smart watch could be on the horizon.
[1] http://wiki.osll.ru/doku.php/etc:users:jcmvbkbc:linux-xtensa...
For ex: Bes2700bp and bes2800 has 3d GPU iirc. Their spec is very impressive, too bad that their SDK is kind of limited to non-Chinese vendor
The low power chips can also run in low power mode without BLE running using micro amps, something the ESP can’t match.
I really like ESP32 and I hope they have a low power chip on their roadmap.
Years ago I had a "smart" watch that had a sim card and was a full mobile phone within its own right, I think it was just 10 years or so too early.
And this chip isn't a normal QSPI chip where you read the datasheet. You have to use NRF connect, and Zephyr.
So, this brings up the obvious question: What if I don't want my whole firmware to be Zephyr nRF-connect, just for a Wi-Fi chip?
Most contemporary SoCs will have more memory (and compute power) than that.
Instead there's a massive gap between Linux-capable systems and RTOS-focused systems despite, like you said, the latter now being bigger and better than real shared UNIX systems.
When I saw rePebble be announced, I signed up for it right away. Only later I realized I actually don't want a smartwatch, I want a dumb watch with vibration notifications.
I know I'm in the minority, but it's a niche that has a few very interested people in it [0] [1] [2]
After wearing the Casio F105 for the past 2 years, I can't go back to something larger, heavier or thicker than this. I could accept weekly battery charging for the benefit of having some bluetooth functionality.
So nowadays I'm looking for a super small bluetooth chip that can power a small vibration motor, which can receive all notifications from my iPhone. I would like to glue that chip, motor and a small lithium battery between the two straps of my F105, because in my tests it seems I don't notice if I add a small weight there.
I still remember when I first used my first Mi Band 1, a forgotten fitness band that had no display, just 3 RGB LEDs that could even get specific colors based on the app that sent the notification. I could know right away when I got a blue Messenger chat that I needed to answer now, or a yellow Google Keep reminder that I could ignore until I got back to my computer.
[0] https://www.reddit.com/r/pebble/comments/9xw2j2/im_looking_f...
[1] https://www.reddit.com/r/smartwatch/comments/174hq9x/need_a_...
[2] https://tildes.net/~tech/18nf/smartwatch_primarily_for_notif...
Like you, my needs are simple; vibrating alarm, notifications, but with one key factor; I need a display that I can read in broad daylight that plays nice with my far-sighted eyes. The eInk display on the OG Pebbles hits the mark. Being able to read a text without pulling out my phone is also nice.
Plus they can be got on eBay for about $30 USD and a fresh battery is about $15, so they don't break the bank. The Rebble.io community's work is still functioning well enough for my use, as well.
I'm also a bit scared of the many "charging issues" some people seem to have with them after a few months, but I guess every batch has a few bad devices so I could hopefully return it.
Thanks for the recommendation! I might try one soon.
Overall everyone has been happy though and I haven’t seen Withings come up on HN which is surprising tbh!
Withings smartwatches come do closest, and that's what I have now. Their standout feature is their 30 day recharge, but they still have a ways to go.
They are expensive, but worse they are fragile and no one will repair them in most countries, and the rechargeable battery means they will only live a few years. On top of that I'd like to see them add NFC, payment card and passkey integration, with better security so if they leave you wrist you have to re-authenticate or reset them. None of these things require much (if any) power. Also while the analogue time is great, interacting with the digital interface is clunky. A single button is not enough. The digital interface could be bigger, and could display more - like displaying the last few message.
So my take is while smartwatches are improving, they still have a ways to go. And that way is not the direction Apple appears to be taking, which seems to be to put a power hungry iPhone with a miniature screen on your wrist. I'm hoping the old watch manufacturers will come back with their robust very low power designs, with a few added sensors.
Basically I want a bracelet, if it must be on my arm, but ideally it would be an ankle bracelet. I'm not sure that would provide a convenient spot to get heart rate, or data beyond steps.
It is actually tempting, but I can't support a subscription based hardware product. Just charge me whatever the device cost + profits.
I justify the ~$20/mo the same way I justify a gym membership: it's a bargain if it's compelling you to make positive choices about health/sleep/exercise. If it's not, then yeah, it's just an expensive mood ring.
Of course, there are also smart rings (Oura) and much cheaper devices (FitBit https://store.google.com/category/trackers).
The reason you want Whoop to have a subscription (and why I wear a Whoop now) is because it incentivizes the company to ship great hardware _and_ software. If you don’t have a subscription, your company becomes stuck on the treadmill of needing to have a new device to sell to the public every year (or even twice a year) so that you can continue to fund your business. Pebble found this out, too, and it led to their sale eventually.
Worse, the launch dates are not movable. People are largely not going to wait and buy your new watch in January if they wanted to get a Christmas gift for their spouse. The same goes for Mother’s Day. The scope is also not movable, because it has to have certain things for people to be interested vs last year’s model. We all know what happens when you fix scope and date in the iron triangle — quality suffers.
The subscription model has a great property — you can ship the device to your customers when it’s ready and meets your quality bar, and you can theoretically do it for free, because they are already paying a subscription. I realize Whoop did not take this path with their latest device release, they are clearly trying to goose their revenue for a quarter or two. That said, at the end of the day, their full product offering will either earn your subscription or not, which means that you can be confident that they are aligned with your interests. You generally cannot say that about a pure hardware company, unless they are remarkably disciplined with respect to hiring. I have not seen an example of this in the wearable/health space.
I think you can derive a general principle from this, which is that if a company is incented to sell you crap, they will eventually do so. If instead they are required to repeatedly earn your business, they will either maintain high standards for their products or they will go out of business. I mostly choose to spend my money on products made by companies with the latter model.
Which is exactly what Whoop _didn't_ do. It seems that the subscription model did not actually work for them.
> you can be confident that they are aligned with your interests
Not at all, as this demonstrates.
> If instead they are required to repeatedly earn your business
The trick is that Whoop dropped this requirement for themselves after they got folks to sign up and before they shipped a hardware update. Presumably they think they can keep running that back---lose all their customers, get new ones who don't know, rinse, and repeat. We'll see how that works out for them.
So, I mean, I think you have some great points but it just doesn't seem to work out that way in the real world.
Pebble allows me to walk away from my phone because I will get the notifications (text or phone call) and can see enough to decide to respond.
Pebble is pretty lightweight and if you load up this watchface you'll feel right at home: https://store-beta.rebble.io/app/52f0939b1ac7948708001fc9
As a watch, it does require that you read the manual completely to understand its various interface paradigms, but it’s reasonable given that there is no display per se, only hands, to show all of the information. Also, you may need to synchronize the hands at first setup, that baffled me at first lol.
It’s efficiently solar powered using a super capacitor, with months of battery life if left in the dark, no battery concerns at all if you wear it (unless maybe you are an actual vampire that also cannot tolerate artificial light)
The supercapacitor can be replaced if it wears out, seems to be reliable for about 15-20 years based on their other watches.
I like mine quite a lot, it’s a well built actual timepiece, put it on and forget about it if you want. It doesn’t ask you to do anything for it to just do its job, year after year.
Citizen may have other options as well with BLE, idk.
Id love to see someone reverse engineer the notification / BLE communication protocol and create an open source companion app - the possibilities for hacker-y types would be pretty cool.
Same. Notifications are rarely time-critical. A few years ago I noticed that I kept getting distracted by my phone alerting to some nonsense which really didn't matter, so I completely disabled all notifications. I'll just look at it once every few hours to respond to incoming messages.
A regular smart watch would be pretty pointless for me. Getting stuff like message notifications on my wrist would be a huge anti-feature. However, it would be pretty nice to have a convenient way to set an alarm, which will vibrate on my wrist when it goes off. I want a "leave now to catch the bus" reminder, or a "it's time to take a break" reminder! That's all, nothing more, no need to add any fancy sensors.
Couldn't you just disable notifications for the unimportant apps?
https://github.com/zoobab/FR801xH
You could get smartwatches for 3EUR on Ali with this chip.
Though, can we stop having left-aligned blogs in 2025? Wide screens have been here for a while, it makes it unnecessarily hard to read :(
But it is probably the easiest constraint to get around. I would put this one more towards the end of the list.
1. His team is positively tiny compared to what Pebble used to have, and the less software work that's needed, the better.
2. All of the apps and watchfaces people wrote for the original Pebbles were distributed as compiled ARM binaries, so if you picked an MCU with an entirely different instruction set, you'd lose backwards compatibility. ESP32 would fall into that category, for example (not that it would have been a good choice anyway).