Watchy: Open-source E-Paper Watch with ESP32
watchy.sqfmi.com
watchy.sqfmi.com
It's probably a great clock and some people seem to have gotten it. But at this stage I'm unsure if I'm going to get it and haven't heard from them
it's an OK excuse for delayed shipment -- but the customer won't ever understand that as a reason if they don't respond to customer service channels.
But everybody else here seems to have gotten theirs, even the person with the same delivery date as me. So I have a bit more faith in it arriving sooner rather than later
It may be misplaced faith, but nothing about the process seemed fly-by-night or in any way indicative that you _wouldn't_ get the product, or it wouldn't be as described.
Another interesting device in this arena, Pine Time: https://www.pine64.org/pinetime/
Of note, PineTime is water proof. Which at least for myself is a big deal.
The breakdown chart on the first link above comparing with other watches is interesting. Watchy seems like it can do a lot more. But I do wonder about battery life.
I don't know who is wearing a bare PCB on their wrist. Especially since the battery (lipo pack) is stuck on the back, so you've got a bare lipo battery sitting against your skin all day.
I got mine a week or two back so I haven't actually had time to run the battery right down (spent too much time plugged into my computer programming and debugging), but the software I've put together so far the battery life seems great.
Over 24 hours, the voltage value I've been displaying (rounded to the nearest tenth of a volt) went from 4.2V to 4.1V. So even under the most unfavourable rounding that's 0.1V over 24h, or still about 95% battery charge. That would be almost 20 days doing some basic processing, screen updating, etc every minute.
I'd agree about the waterproofing, however. Wearing it for a few days and knowing not only that it's not waterproof but exactly how _un_waterproof it is, it's made me a lot more conscientious about handwashing and day-to-day activities.
How're you calculating that? Voltage of (what I assume is) a lithium ion cell is a pretty poor way to measure state of charge - the discharge curve maintains voltage pretty well until 80% capacity before falling off precipitously.
You certainly aren't going to get any work out of the cell below 3V or so without risking permanent damage.
Just googled "lithium polymer discharge curve" and averaged between a few results. Generally saw an inverted S shape where the first few tenths of a volt should be shed pretty quickly before stabilizing for a while, then dropping precipitously as it neared dead (~3V).
I do like the idea of a fully open source watch though, I used to do quite a bit of Android Wear development in its early days [0] [1] [2] but got frustrated with Google breaking my apps every 6 months with new APIs, breaking my Android Studio projects with every new version of Android Studio, and removing them from the Play Store because of new rules. I much prefer hardware and software that just continues to work if you do exactly nothing, which is kind of how all my other appliances work. It's not like my air fryer or rice cooker just stop working if I don't update their software.
Also I absolutely hated Java and would much rather develop in C++, (Micro)Python, or even C, all of which are possible with the ESP32.
[0] https://www.pcmag.com/news/attopedia-for-android-wear-puts-t...
[1] https://www.androidpolice.com/2014/07/15/there-is-now-a-matr...
[2] https://www.androidauthority.com/android-wear-remote-app-404...
I mentioned C/C++/Python because the actual app would be much simpler and less bloated, not because of writing inner functions in those languages.
It is incredible how NDK manages to make me miss Symbian C++, so I get your point anyway.
I now use a Fossil HR Collider. It lasts for 2 weeks probably thanks to e-ink. The HR also isn't useful, but it looks good for a round watch (except for the stage light) which you can make look like a normal smartwatch, and it can control e.g. Spotify and weather. I don't need much more on a smartwatch. The only thing I miss is the ability to put Bluetooth off.
Good thing my Pebble is still ticking!
From the verge review
I've got mine only waking up once a minute using the external RTC as a trigger. In between I put it in hibernation (~2.5uA to the processor). So every minute it's waking up for about a second to do some updates, redraw the screen, and shut itself back down. (It takes quite a while to refresh the screen.)
I haven't had it unplugged more than 24 hours at a time (keep plugging it back in to do some more programming and debugging), but so far the biggest drop I've seen was the battery voltage (rounded to the nearest tenth of a volt) going 4.2 -> 4.1. Even assuming worst case rounding here, that's something like 100% -> 95% battery capacity, so I'm looking at more like 20 days just doing basic time keeping so far.
schneider has a great overview of some of the challenges of (barely) getting that project out the door on time:
I realize it's FOSS hardware but I'm more of a software person and would probably take me too long to figure out how to hack all those things in and get it manufactured ...
It just isn't in my use case. For me, the only use of a watch is on the go. When I'm not on the go, I already have a computer w/me which is more powerful than any smartwatch. Therefore, if I am on the move, I want to blindly use my watch, just like I type this message blind on a keyboard. Physical buttons aid with that. If I am catching a bus, I want to be able to control Spotify. If I am running, I want to be able to control Spotify. Touchscreens are far more cumbersome in that regard.
Apple Watch, yes very powerful computer, beautiful screen, and very expensive (more so if you want the medical features). You have to give it juice every night, and it requires either a 4G connection or you're having to pay for another plan or it uses Bluetooth w/your smartphone. These devices don't have enough storage for my liking. I mean, I am sure I could add and use a microSD with 256 GB or more to it, if it had that option. Instead we gotta rely on very little storage, Bluetooth with smartphone, or another SIM + data plan. All in all a completely different purpose and price class.
The only sensor on this is the Bosch BMA423 3-axis accelerometer. Unfortunately, that has no gyro or magnetometer, which makes it a bit of an odd choice. It's supply constrained, and has an EOL date of Sept 2021 to boot. The other sensor is a DS3231MZ low-power highly-integrated real-time clock. It's also out of stock, and has a 42 week manufacturer lead time.
I've got some experience developing embedded hardware, but the watch form factor is really hard to hack after the fact. It's not particularly friendly to expansion busses and add-on cards...
How accurately do you care about geolocating yourself?
If you only care about <50m accuracy for the watch to be able to provide you location-based behaviors, I've had very good results doing Wi-Fi scans with ESP32 devices (in particular, a FeatherS2) and feeding the results into the Google and Mozilla geolocation APIs. More than enough to pull up the local weather or public transit info for the bus stop in front of you. It could send data requests through your phone when you don't have Wi-Fi.
Obviously it won't have the accuracy and update rate that a GPS would, but on the upside, it does work indoors.
I love my Garmin, but the aggressive price segmentation (even though I bought the device at the top of the list) is vexing and I would love to help build an open-source alternative.
On the note of pushing data through a phone, I suppose I could stick a uBlox, ESP32, and tiny battery in a case on my belt, and push GPS data to the watch that way, but that's basically the same as the watch minus the display.
Just brush it on. Should keep the electronics safe from moisture and maybe survive a quick dunk or two
> The "Respects Your Freedom" certification program encourages the creation and sale of hardware that will do as much as possible to respect your freedom and your privacy, and will ensure that you have control over your device.
https://watchy.sqfmi.com/docs/battery-life/
Yes, I looked, and it's crap. Don't get me started on design/dimensions.
I checked market few weeks ago and there is still nothing better than Amazfit Bip.
I'm genuinely unsure e-ink offers any real benefits at this point.