The concept of a smartwatch was thrown around for a long time before the technology truly came to fruition. Through the pursuit of miniaturisation, modern smartwatches are sleek, compact, and remarkably capable for their size. Companies such as Apple and Samsung throw serious money into research and development, but that doesn’t mean you can’t create something of your own. [Electronoobs] has done just that, with this Arduino-based smartwatch build.
The brain of the watch is that hacker staple, the venerable ATmega328, most well known for its use in the Arduino Uno and Nano platforms. An FTDI module is used for USB communication, making programming the board a snap. Bluetooth communication is handled by another pre-built module, and a smartphone app called Notiduino handles passing notifications over to the watch.
This is a build that doesn’t do anything crazy or difficult to understand, but simply combines useful parts in a very neat and tidy way. The watch is impressively thin and compact for a DIY build, and has a host of useful functions without going overboard.
We’ve seen other DIY builds in this space, too – such as this ESP8266-based smartwatch. Video after the break.
Both this and the ESP8266 has the same flaw: There are parts on the back of the board, that press, rub and get corroded by the sweat on your arm.
Love the ideas in both, both are flexible and very hacker friendly. Just build /up/ from the top, not both sides, so you can put tape, silicon or just about anything else over the back(MAYBE a few passives on the back, but ideally nothing).
Oh.. And please, once your dick of the pretty but power hungry OLED, use a Sharp Memory LCD.. It’ll solve your battery life problem.. :)
Seems the hard part is finding wearable project cases that size. Watch cases are usually custom. Maybe 3D printing can help here?
I think 3d-printing could indeed help. Actually, I saw that someone else already made a smart-watch ( https://www.youtube.com/watch?v=LfmwhabTVB4 ), and he also 3d-printed the case. Only thing about that watch is that it’s a bit bulky…
I suspect that it’s actually far easier to do a double sided board with components on both sides is more hacker friendly than trying to stuff a board where the parts are all on one side.
Also if you do have solder joints on the back side, make sure you use RoHS for your health.
Unmodified Arduino framework isn’t the best choice for low power.
Quote: “And please, once your dick…”
Are you really talking about my dick?
or did you mean YOU’RE sick?…
Damn lack of edit button.. Will teach me for typing while half asleep!
Jesus this thing needs some strain relief on the attachment point for the wrist band! As it is, it will probably break after a day’s normal use! Else it looks great :)
One way to get around the components on the back being a problem would be to coat the back in epoxy resin.
Maybe even make a mold and set the whole thing in epoxy. That creates some repairbility issues – but whose really going to repair the hardware anyway…
How about wearing a wrist band under the watch?
Or just have a second blank FR4 as the backing. The FR4 could be a thin one e.g. 0.5mm so it doesn’t increase the thickness too much. With 2 PCB and some careful planning, one could use regular watch bands. I bought $3 metal band replacement for my watch from China and they are quite nice looking.
I was thinking the same thing. But then I got to worrying how to protect the moving parts, ie switches, from the epoxy? I decided petroleum jelly might do. But then how to rinse it away after? You could use some non-polar solvent (non-conductive, of course). But that might start to melt the plastic away. Assuming the other steps so far haven’t wrecked anything. I was going with clay to make the positive part, then cast that using fine plaster, covered with, say, petroleum jelly.
If no petroleum jelly, maybe margarine? Then use distilled (will dionised do?) hot water and detergent to wash it off. Except if any water stays behind eventually ions from dirt and sweat will get dissolved in it.
That’s assuming epoxy is OK with all this. What was the “potting compound” they used to pot stuff with way back? It was an attractive pink colour, actually, horrible to look at. Colour it in with a marker? Not spray paint, of course.
See, it’s actually a more complicated issue than you’d think. I haven’t worried about the heat from curing epoxy, yet. Or possibly expansion.
I would like to come up with all sorts of smart suggestions, like why is the buzzer at the back (pointing towards the arm) and not to the front, why is the FTDI chip on the watch (taking up boardspace and energy) it could easily be done in the form of custom cable. Why an RTC chip, when there is already a microcontroller onboard? With some extra effort the microcontroller could be made to run “accurate enough”, this should save a lot of boardspace. The wristband slots on the PCB look like the can snap off any minute. The lack of a case is just a detail, I’m sure that can be fixed, it must be, because sweat/rain/etc. are not really an issue, one bump against the coffee table while reaching for the paper underneath and the OLED display shatters or scratches the table in a very nasty way. If the solderjoints don’t catch the thread of you expensive sweater/shirt first.
Anyway room for improvement or not, in the mean time he has a DIY smart watch to show to his friends, while I have nothing to show at all (except my wise ass suggestions). Anyway, fun project!
Looks like its not from the web site, he has a USB-serial module he plugs in. I think what your seeing on the watch is the battery charger.
I think I’d start with these for a watch
https://tinycircuits.com/pages/tinyduino-overview
Tinyduino are kinda pricey for what you get. An adafruit bluefruit board with an oled module would probably get you closer to a smart watch for cheaper.
https://www.adafruit.com/categories/943
Neat idea, but if memory is an issue, is there any reason not to go with a Cortex M0 or M4 or something instead of the atmega 328?