The germ, so to speak, for Timestop and the D-20 watch came out of the pandemic. I was running a makerspace, playing D&D, and doing some small electronics projects with my kids. I was spending a lot of time with people who liked games and making things.
The initial idea was extremely broad: some kind of electronic dice. From there, it was an exploration of what was most interesting, possible, and to some degree, hard. We arrived at a segment LCD wristwatch because it was “of the era” (the LCD wristwatch originated in 1973; D&D in 1974), it could have a mind-blowing battery life measured in years, and it seemed impressive if we could do it.
Fittingly, the watch’s STN (Super Twisted Nematic Fluid) display is made using photochemical etching, the same manufacturing process as the PCB that drives it. The segments themselves even act as capacitors.
Initial Sketches
With the general technology chosen, we moved on to how it would work and feel. This was memorialized in rough computer drawings. We wanted a quick way to roll any of the typical polyhedral dice used in TTRPGs (hence “Quickroll” mode), but also wanted something you could actually play the game with - a tool - all the way into multi-dice damage rolls, D100 loot tables, advantage, and disadvantage. Combat Mode.
We wanted a fun animation while rolling, but also an overall subdued look so you could “take the game with you” and even wear the watch at the office. We included day and date complications with a perpetual calendar so it was still a practical timepiece.
But we chose to leave out some things we didn’t like about most digital watches. We replaced squishy, beeping buttons with clicky tactile switches, and killed the alarm completely. Alarms aren’t fun.
Workslike/Lookslike
To prototype the watch, the first thing we did was separate the “workslike” and “lookslike” examples. We had to get a better understanding of the underlying technology, but also had to end up with something we would enjoy wearing.
For the workslike prototype, we used through hole components on a breadboard, a stock display (timekeeping only, no rolling yet) and focused on how the heck to drive an LCD for years on a tiny coin cell battery.
The lookslike prototype was a slab of SLA resin, laser printed face, and stock elastic strap. We chose our “goldilocks” dimensions for a watch: high 30s (mm) width, less than 10mm height, 20mm lugs. More display real-estate than a vintage digital, but smaller than a modern wrist-brick. More like vintage mechanical divers.
Making Room
The next challenge was seeing how small we could make the components and how big we could make the available space within the case.
We had a custom LCD display made so we could not only see its final size, but also meaningfully start to develop the additional roll functionality. The PCB that drove it was just a development board, where we focused on using the final or near-final component packages without yet having to worry about sub-millimeter placement decisions.
The mechanical prototype was also broken down into separate, conceivably manufacturable, parts. The crystal mockup was laser cut acrylic, but the rest of the components were either SLS or SLA 3D printed. We knew most FDM machines would be too coarse for the project, but even found trouble getting reasonable tolerances with the other technologies at this scale.
Convergence
Now that we were adding more firmware functionality, we pretty quickly found a number of problems with our development board (like putting the buttons on non-interrupt IO pins). But it had served its purpose, and on the next board we moved the components to their near-final positions.
This is the point at which the difference between the lookslike and workslike prototypes nearly disappeared. One we kept plugged in for updates and debugging, and the other we assembled and started wearing and using. It was mostly still 3D printed, but with sample final components where 3D printing wouldn’t work, like the buttons (features too small) or the crystal (it’s prettier silkscreened on real glass).
We had a functional, if somewhat temperamental, watch that could tell the time and roll virtual dice. In other words, we were about halfway to a final product. We had a watch but we didn’t have the watch.
The Last Mile
The last step was to simply make everything better. The case was machined in steel with engineering tolerances and o-rings for water resistance. The strap was lengthened and made softer, with the clip made in steel. The interior frame and backlight diffuser were injection molded. We added grippier screws.
The board was tweaked for better battery contact and faster assembly. The display got a brighter reflector. The firmware was overhauled to minimize battery usage and optimize for use in-game and day-to-day.
We couldn’t be more pleased with the result. We wear the watch every day, and use it every game night. We’ll be making an initial production run of the watch this fall, and you can join our mailing list to get updates on progress and availability.