“Once upon a time, in a kingdom far far away” – that’s how almost every fairytale starts.

For our fairytale, that time was back in 2014 and that “kingdom” was our hometown of Thessaloniki, indeed far far away from where we are now. That’s when and where we initially came up with the idea of time² and built the first prototype. At that time, we were calling it TimeSquare and it was a lot different than it is today.

We knew it had to be based on wood, but we had absolutely no experience in working with that material at the time. Not only that, but we couldn’t have direct access to a wood working lab and laser cutters where not simple to find at that time and place.

Appearance was the most important aspect of the project and it had to be perfect even on the proof-of-concept stage. The way we saw it, if it didn’t look good, there was no project.

We always loved that contemporary and minimalistic IKEA Lack design, with the strict rectangles and the thick lines, so the initial design was inspired by it. It was easy to follow those principals and design a simple rectangle box to enclose the project and assign its construction to a professional wood workshop. And that’s exactly what we did.

But the biggest challenge was how we were going to carve out the words on the front panel. Initially, misguided by our lack of experience, we were naive enough to think we can do it by ourselves with ordinary tools. We lost countless hours trying to accurately cut plywood using stencils with different wire saws and different bits on a Dremel router. As it should be expected, the results were far from acceptable.

Finally, we found a company that manufactured signs and they owned a laser cutter big enough to fit our 45 x 45 cm front plate. They were nice enough to accept a custom order from us. The result was perfect but, at that time and place, the cost for just cutting our piece of plywood according to our design was higher than the full BOM cost of time² today.

The next task was to create light insulated compartments behind the front plate, one for each word, so that when a word lights up, the light doesn’t also affect the neighboring words. We used slices of plywood for that, and since we didn’t care for the appearance of the internals, we cut them with a jigsaw and installed them using metal 90° angle brackets.

There were 23 individual compartments (words) in the prototype, so we needed to individually control 23 LED strip segments. We also wanted smooth transitions from each clock display to the next one, so just switching on and off the strips wouldn’t cut it. We needed to be able to dim each strip segment. That meant that we needed a controller with at least 23 PWM capable GPIO pins.

Our experience at that time was limited to AVR based Arduinos, but even Arduino Mega didn’t have so many PWM capable GPIO pins. So, we had to find a solution.

When GPIO pins are not enough, the standard solution is to employ shift registers on the design. So that’s what we also did. We used 3 8-port shift registers in series. But hardware PWM is not available when someone uses shift registers, so we needed to develop a software solution to control the 23 LED strip segments through shift registers with support for 8-bit resolution dimming.

This solution also limited us to single color LEDs instead of RGB LEDs. Using RGB LED strip segments would mean that the design would be 3 times more complicated. It would need 9 shift registers, 9 Darlington arrays and the wire jungle from the Darlingtons to the strip segments would be 3 times denser.

Following this solution, now we only needed 3 GPIO pins to control all the strip segments. We didn’t need Arduino Mega. Even Arduino Uno was an overkill. We went with the smallest Arduino with USB, the Arduino Nano.

The last challenge was powering the setup. Arduino GPIO pins operate at 5V and can supply a maximum of 20mA. This may be enough to drive a single indication LED but it’s far for being able to power whole strips of 12V super bright LEDs. To solve this, the outputs of each 8-port shift register was connected to an 8-port Darlington array. Arduino Nano can be powered through an embedded voltage regulator that accepts 12V as its input, so we used a 12V power supply to power the Nano and the LEDs through the Darlington arrays.

When the proof of concept was finished, the project entered a hiatus period until 2018 when we moved to Shenzhen and development of time² began. So why all this?

We like to make and we like to learn through making. TimeSquare was one of our first projects. It was followed by a period of other creative projects like MagnetLight and TimeWall, each with many design and manufacturing challenges and numerous revisions. The development of time² as a commercial product could not have started without the experience gained from dealing with all these challenges.

It’s a very long way from the concept of the idea and the blueprints to actually making a product. It’s not just making it, but making it reliable and efficient, with a clear and easily reproducible manufacturing procedure without sacrificing design principles for convenience, keeping it serviceable and following product safety standards while optimizing materials and assembly cost.

But it wouldn’t be worthy if it was easy. Would it?

“Once upon a time, in a kingdom far far away” – that’s how our fairytale started. But how does it end? Well, that’s up to you.

time², the much more advanced successor of TimeSquare, will be soon featured in a crowdfunding platform near you. And with your help, we are sure it will be a successful campaign, and the fairytale will end with all of us “living happily ever after”.