I will try to design, build and program a smart chessboard that is as accessible to learn from and use as possible. This will mean I will have to tackle complexity, price and quality all at the same time. I hope I will be able to allow others to enjoy and learn from this project just as much as I do.
Started designing physical layout
First I tried to get all my ideas onto paper. That didn’t work out well and I realised that I was going to have to build a physical prototype anyway so might as well just do that to clearly communicate the design. When doing so I realised that making the surface of the board from cardboard would take very long and I can probably find an easier solution. Note: this is also part of the reason I decided to make this specific project, because I like finding elegantly simple solutions to problems I usually see being overkilled.
Sorry this Devlog is brief because it is more of a bug encounter than a working part of a program. When I finish the prototype or hit another mile stone, I can explain a lot more.
The image attached is a snapshot of my current progress.
Figured out how to power the devboard and led strip.
First I thought of different ways of powering it like using rechargeable batteries or directly from a wall socket. They both had their own cons: I think adding batteries and their management modules would add a lot of complexity; wall sockets are much simpler but plugging your chess board into a wall doesn’t feel right and that means that you cannot play outdoors. I then had the idea to use a power bank as that means that I can use the portability of a battery without needing to handle the complexities especially as I am trying to keep it as simple as possible for beginners to build. Also expecting power from a power bank (through a USB port) would also mean that it can also be powered from a wall socket phone charger.
Next I researched and figured out how I can power and signal the WS2812B LED strip especially when it was expecting 5V and the devboard (RP2040 based) was signalling on 3.3V. I could have used a level shifter but then I found that I can use a ‘sacrificial LED’ trick. This involves stepping voltage down from 5V to 4.3V using a diode’s natural 0.7V drop. Because the WS2812B chip works on a voltage as low as 3.5V and the signal should be at least 70% of the power voltage, supplying 4.3V to it would allow it to work on a signal of (70% * 4.3V =) 3.01V which my 3.3V is well above. The sacrificial LED would step up the signal voltage to it’s power voltage (4.3V) which is more than enough to signal the rest of the strip (which will be powered from a full 5V). I then tested this on the few WS2812B LED strips I have at home and it worked perfectly.
Lastly I was going to use an external USB type-C female connector to power because I thought that pulling 600mA that I calculated will be the max current draw (when the queen is in the middle of the board I need to light about 30 LEDs) will be too much current for the built in USB port on the devboard to handle and may cause heat issues. But after a quick search I found that it should be completely fine as long as I take the current from the right pin. So I have decided to use the VSYS pin or the equivalent from the devboard to power the LED strip.
The digital diagram below shows the circuit (it does not include how I plan to receive input and track the chess pieces; it only shows the output and process part of the system). I have also included the drawings I made as I was thinking and planning.
