A common design language for watches has evolved ever since they first started popping up in the 1500s. Whether worn on the wrist or in a pocket, watches are relatively slim front to back, with the display mounted on the face. That’s understandable given the imperatives of human anatomy. Still, it’s not the only way to arrange things, as this very cool LED matrix watch with an edge-mounted display demonstrates.
True, the unique form factor of this watch wasn’t really the point of the whole project. Rather, [Vitali]’s design was driven by a couple of things. First off were the extremely cool Hewlett Packard HDSP-2000 displays, with four 5×5 LED matrices shining through the clear cover of a DIP-12 package. Also visible through the cover are the shift registers that drive the matrices, complete with gold bonding wires.
The main attraction for [Vitali], though, was the challenge of working within the limits of the ATtiny85 he chose to run the watch. The MCU’s limited IO made hardware multiplexing necessary, no mean feat given the limited resources and real estate available. He still managed to pack everything in, with the unique edge-mount display coming from the LEDs bridging the space between the two main PCBs. Everything fits into a nice wood veneer case, although we think it looks just fine without it. [Vitali] puts it through its paces in the short video below.
Hats off to [Vitali] for a great-looking project that pushed his limits. We just love these displays, too; of course, it’s not the first time we’ve seen them put to similar use.
Printed circuit boards are typically only something you’d find in a digital watch. However, as [IndoorGeek] demonstrates, you can put them to wonderful use in a classical analog watch, too. They can make the perfect watch dial!
Here’s the thing. A printed circuit board is fundamentally some fiberglass coated in soldermask, some copper, maybe a layer of gold plating, and with some silk screen on top of that. As we’ve seen a million times, it’s possible to do all kinds of artistic things with PCBs; a watch dial seems almost obvious in retrospect!
[IndoorGeek] steps through using Altium Designer and AutoCAD to layout the watch face. The guide also covers the assembly of the watch face into an actual wrist watch, including the delicate placement of the movement and hands. They note that there are also opportunities to go further—such as introducing LEDs into the watch face given that it is a PCB, after all!
It’s a creative way to make a hardy and accurate watch face, and we’re surprised we haven’t seen more of this sort of thing before. That’s not to say we haven’t seen other kinds of watch hacks, though; for those, there have been many. Video after the break.
Every now and then, a commercial product aims to help you in your life journey, in a novel way, making your life better through its presence. Over the years, I’ve been disappointed by such products far more often than I have been reassured, seeing each one of them rendered unimaginative and purposeless sometimes even despite the creator’s best intentions. The pressures of a commercial market will choke you out without remorse, metal fingers firmly placed on your neck, tightening with every move that doesn’t promise profit, and letting money cloud your project’s vision. I believe that real answers can only come from within hacker communities, and as we explore, you might come to see it the same way.
This is the tip of the iceberg of a decade-long project that I hope to demonstrate in a year or two. I’d like to start talking about that project now, since it’s pretty extensive; the overall goal is about using computers to help with human condition, on a personal level. There’s a lot of talk about computers integrating into our lives – even more if you dare consult old sci-fi, much of my inspiration.
Tackling a gigantic problem often means cutting it down into smaller chunks, though, so here’s a small sub-problem I’ve been working on, for years now, on and off: Can you use computers to modify your sense of time?
The Time Question
Ever start your day thinking you will hack on a project, and in the evening, realize you’ve instead done something else entirely? Sometimes you find something cool while distracted, and sometimes, getting distracted comes to haunt you.
maybe one day I will assemble these
This has been a staple of my days as long as I remember my conscious life, and at some point, I started wondering just how much this could be modified. Do you remember one particular project we’ve seen a couple people build, a vibration-based compass build that gives you a sense of where north is? Ironically, I have made PCBs for building my own version of this project – they were designed in 2022, I finally ordered them last year in 2023, and I haven’t gotten to assemble them still.
So, you can give yourself a sense of “where’s north” – something that humans are missing, generally. Technically, humans are also missing a source of time, which is why we always supplemented it with wrist-worn watches and pocket clocks. Having compared my day plans to what actually happens on that day for two decades, I can see that I need something more than that. It’s traditionally been common for me to mis-estimate when exactly I could get something done – I would give an estimate that felt correct, then start doing part of the task and forget about the flow of time, minutes passing by me.
So, there are two problems here. One of them is that, despite having been alive for a fair bit of time, my database of “how much it takes for me to do X” is inaccurate. This makes sense: keeping such references is a conscious effort that might not extend to, and day-to-day situations are highly variable. Still, if someone is relying on me, it would be nice to be aware enough to at least notify that person, and to learn to plan ahead. Another is that it’s easy for me to get and forget about the flow of time. It sure helps me concentrate on articles, but it doesn’t help when someone is waiting on me.
At some point, this started to screw with my sense of self. Really, just how much can you rely on some aspects of your mind if it continuously fails you and people you care about, in a manner that you are expected to “just figure out already”? You have to learn to distrust certain basic aspects of your cognitive processes; again and again, something as “simple” as time planning is weighted down by all the instance of letting people down with zero intention to do so. This is a pretty uncomfortable position to be in, if being honest with yourself is a priority of yours. Unsurprisingly, it also made things pretty difficult when talking about employment or real-life obligations. Something had to be done.
Well, could you give yourself a sense of time, say, with vibromotors? Apparently, you can, but there’s nuance to it. Let me tell you about two projects I’ve built to attempt this, and some basic concepts I learned about human-computer integration.
The Not-A-Bomb Wearable
My first project in this vein grew out of a purpose-less experiment, funnily enough: a project literally called I Made This And I Don’t Know Why – a simple board I built to make use of seven-segment displays our hackerspace had a dozen of. ESP8266, dynamic indication with a shift register, and MicroPython – writing firmware for this board was a nice challenge in writing non-blocking code and finding portions of code to optimize. Soon, the board found a good few purposes – among them, a time tracker.
I decided to solve a simple problem – building a mental database on the amount of time does it take me to get from “start” to “finish” for an arbitrary task. Tracking that was tricky – say, I want to check the length of a bicycle ride from my house to a certain point. I’d need to check my phone at the exact time when I left the house, keep that time in mind, and then, once I’ve arrived to my destination, check again. Both of these require some time to execute and some memory, so, I decided to make an automatic countdown timer. Glancing at my wrist felt significantly easier, so, after some cutting, sewing, and hotglue work, I made one of the IMTAIDKW boards into an oversized watch, and used one of my universal power source designs to power it from a 18650.
There were some setbacks during – notably, this countdown timer required me to patch MicroPython’s ESP8266 port, due to an obscure bug making the time.time() function seriously imprecise; an inaccurate countdown timer wasn’t in my plan. Still, it was a nice experiment – relying on something that you build yourself is always fun, and I’ve added features like adjusting the start time. It was also automatic enough to be useful, with digits large enough and bright enough to be noticeable, still, making for an unobtrusive device, and pretty cool to wear.
The main problem was that I forgot to put it on and start the countdown. It was a purpose-built device, and I only needed it a couple times a day at its very most, so most of the time it stayed off my wrist, and I would even lose track of it sometimes. Another problem was remembering to check the time of arrival, unsurprisingly – looking at my wrist was easy enough, so most of the time I could notice the time difference and go “oh interesting”, but even then, it was easy to forget. The last, main problem, was actually keeping a mental database – turns out that when you need to remember pretty similar datapoints, it’s easy to confuse them. Does it normally take me 15 minutes to get to the city center, or was it the electronics store? This turned out to be pretty easy to mix up.
The lessons from this iteration: decreasing resistance to use is good, collecting data is good, and, you should automate the data collection process if at all possible. I wouldn’t stop here, of course – some time later, I found an even nicer wristband to hack on.
Unconventional Battery Upgrades
The TTGO (or was it Lilygo?) T-Wristband is a fun product – with an ESP32 at its heart, a good few sensors, a 160 x 80 IPS LCD, and a single capacitive button. It’s an old device by now, but when I bought it in the beginning of 2020, it was fun to hack on, and hack it I did, making it run MicroPython. I didn’t know what exactly to do with it, but soon I remembered about the “sense of time” project. At the time, I wanted to tap into my life minute-by-minute and see if I could build a device able to help me notice when I’m distracted. The minimum viable prototype idea was very simple – adding a vibromotor to the watch, then having it vibrate exactly every minute, having it be an “am I currently spending my time correctly” reminder.
The problem was, by the time I came up with that, a good few months passed where the wristband was sitting in a drawer with the battery fully discharged – hurting its capacity a bit, which, at 80 mAh, was already not great. Also, I wanted to be able to keep adding features to the code without carefully balancing sleep modes or having to charge my watch multiple times throughout the day; I just wanted to run code and charge the battery every night at most. So, it got a battery upgrade – a Samsung phone battery glued to, ahem, yet another wristband, and a devboard with vibromotor driver taped on top. After the hardware tweaks, the code itself was seriously easy to write.
despite the added bulk, it was surprisingly fun to wear. at some points, I even added features like remote PC control and a gesture interface!
Whenever I’d notice it vibrating, I’d ask myself – “am I doing the right thing right now?” And, to my surprise, it did catch some distraction moments every now and then, for sure! Oftentimes, I wasn’t doing the right thing, in one way or another, and a reminder about being supposed to do something else was quite welcome. Other times, when I was focused on something, the “am I doing the right thing” question would get a “yes” in my mind, and, it felt good to think that.
It wasn’t as comfortable in times when I wasn’t expecting me to be on top of things – while I’d be resting, the every-minute feedback of the watch would feel annoying and needlessly distracting; soon, I implemented a vibration toggle with the capacitive button, and a few other things. My guess is that the annoyance factor and generally getting used to the vibrations has made me less sensitive to the vibromotor’s signal, which in turn made the wearable less effective at its goal. Apart from that, the battery wire kept breaking every so often, taking the watch out of commission, which made it hard to start properly relying on it.
On the upside – itturnedoutthat this idea has been floating in collective unconscious for a while now, to the point that it was the point of a watch worn by one of the characters in Mr. Robot, and a relatable one at that. It’s pretty good to get external independent confirmation that an idea of yours has merit! In particular, the video above reminds me a lot of my experiences – I spent less time on my phone and generally less time doing things I didn’t want to do, I was getting up and walking around more often, and, I had add a small feature that mutes the watch when I go to sleep.
It All Worked Out Despite The Plan
Lessons here? If you can hook your device’s signals into producing a thought in your brain, that helps massively – checking for “am I doing the right thing” every minute came to me naturally, and a lot quicker than I expected it to. Context sensitivity is a must for self-help devices- the wearable would’ve had been way more effective if I had some ways to detect that I’m likely to be distracted, as opposed to having it vibrate indiscriminately every minute. In general, make sure your device is not annoying to you in any bad way – it’s supposed to be helping you, so any reason you’re annoyed by it, is a problem for the device’s primary usefulness.
On the hardware side, make your device reliable – building habits takes an ongoing effort, and you want it to be consistent. At the same time, consider building your device as a playground for developing your idea further; this could require a bigger battery, or more space inside the case, or an expansion socket. Reality is to plans what pure oxygen is to paper, and getting things done is typically way more important than getting them right the first time. Last but by no means least, wires suck – I’ve been saying this, and I will repeat that as much as needed.
In the end, I have mostly solved my original problem by tweaking my personal approach to time over the years, learning to over-estimate estimates, and ultimately putting myself in less situation where I am under time pressure – it turned out that was the bigger problem. It would’ve been nice if I could’ve noticed that sooner, but, the devices I’ve built certainly have helped. Today, I still have some sense-of-time solutions I rely on, but they are new, designed with these lessons in mind, and they’re a part of a multi-faceted system that I can only tell you about in the next articles – stay tuned!
You might think of the smartwatch era as beginning with Apple, relatively recently. Or, you might think back to those fancy Timex models with the datalink thing going on in the 1990s. Seiko can beat them all, though, with its UC-2000 smartwatch that debuted all the way back in 1984.
The UC2200 was the bigger docking station of the two.
The UC-2000 very much looks cutting edge for its era, and absolutely ancient today. It featured a 4-bit CPU, 2 kilobytes of RAM, and 6 kilobytes of ROM. Display was via a simple 10×4 character LCD in a rectangular form factor, with four buttons along the bottom. Branded as a “personal information processor,” it was intended for use with the UC-2100 dock. This added a full physical QWERTY keyboard that interacted with the UC-2000 when the two were combined together. Alternatively, you could go for the UC-2200, which not only had a keyboard but also a thermal printer to boot. Oh, and ROM packs for Microsoft Basic, games, or an English-to-Japanese translator.
What could you do on this thing? Well, it had basic watch functions, so it told the time, acted as a stop watch, and an alarm, of course. But you could also use it to store two memos of up to 1000 characters each, schedule appointments, and do basic calculations.
The one thing this smartwatch was missing? Connectivity. It couldn’t get on the Internet, nor could it snatch data from the ether via radio or any other method. By today’s measures, it wouldn’t qualify as much of a smartwatch at all. Moreso a personal organizer that fit on the wrist. Still, for its day, this thing really was a whole computer that fit on your wrist.
If something has a “smart” in its name, you know that it’s talking to someone else, and the topic of conversation is probably you. You may or may not like that, but that’s part of the deal when you buy these things. But with some smarts of your own, you might be able to make that widget talk to you rather than about you.
Such an opportunity presented itself to [Benjamen Lim] when a bunch of brand X smartwatches came his way. Without any documentation to guide him, [Benjamen] started with an inspection, which revealed a screen of debug info that included a mysterious IP address and port. Tearing one of the watches apart — a significant advantage to having multiple units to work with — revealed little other than an nRF52832 microcontroller along with WiFi and cellular chips. But the luckiest find was JTAG pins connected to pads on the watch face that mate with its charging cradle. That meant talking to the chip was only a spliced USB cable away.
Once he could connect to the watch, [Benjamen] was able to dump the firmware and fire up Ghidra. He decided to focus on the IP address the watch seemed fixated on, reasoning that it might be the address of an update server, and that patching the firmware with a different address could be handy. He couldn’t find the IP as a string in the firmware, but he did manage to find a sprintf-like format string for IP addresses, which led him to a likely memory location. Sure enough, the IP and port were right there, so he wrote a script to change the address to a server he had the keys for and flashed the watch.
So the score stands at [Benjamen] 1, smartwatch 0. It’s not clear what the goal of all this was, but we’d love to see if he comes up with something cool for these widgets. Even if there’s nothing else, it was a cool lesson in reverse engineering.
Conectar
