Interactive LED gaming table

[Daniel Zetterman]

I'm working on a modular board with 4x4 square cells. Each cell contains a RGB led and a touch sensor. Both the RGB led and touch sensor should be possible to control/poll using a serial bus to make it possible to daisy chain at least 16 boards.
With this board it will be possible to assemble it into a wooden table under a tinted glass sheet.
Connecting a microcontroller to this and possibly sound and wifi etc you will have a full fledged gaming table.

My prototype implementation of the above idea is using the smart rgb led apa102c and standard 940nm IR led with a matching photodiode. The Apa102 is controlled using 2 wires, data and clock. The touch sensors are connected to daisy chained piso shift registers which controlled using an enable signal, clock signal and input data signal. The concept works.. see the demo video. I can poll the touch sensors 1000 times per second and more without a glitch.

However the problems are related to the touch sensors:

1. Glass reflection, how to best overcome this?

2. Ambient light sensitivity, how to best overcome this..modulate the IR led?

3. Touch sensing precision. There is a difference between the sensors and howto make it detect only when you touch the glass?

4. General robustness, avoiding noise and false touches.

I'm thankful for anyone wanting to discuss above topics and give me some hints and best practises. I prefer if it comes from own experience of successfull implementation of the same but guesses are also welcome

Below is a simple demo of the prototype.. please ask any questions you have on the design of the board.

Below is the circuit diagram for one Touch sensor for reference.

 

[Harald Kapp]

You may want to have a look at the octolively modules. They use single color LEDs only, but I think it will be an interesting challenge to extend the system to RGB LEDs. The octolivelies are cascadable in two dimensions and thus well suited to a gaming table.
Somewhere in the discussions around octolivelies I remember to have see the issue of covering them with opaque glas/plastic being discussed.

 

[Daniel Zetterman]

Thanks Harald, yes this was one of the projects that inspired me. However the big difference is that it is not centrally controlled, so the overall design and signal flow will be much different.
But it would be very interesting to know how they avoid detecting on the reflection from the glass

 

[GPG]

You need to check the transmissitivity of any cover at 940 especially if its coated/tinted. You could probably use a opto interrupter as a test if its got a big enough gap. Check out plastics as well.