Author Options:

Another idea for a piece-tracking chess board Answered

RFID-based chess boards have their merits of course, but would be pretty expensive and could potentially suffer from antenna cross-talk. Furthermore, the relatively immense amount of information carried compared to what's required makes RFID sort of overkill. Reed switches have the opposite problem, only carrying a single bit of information per square (piece present y/n). It seems that what you want is something in-between (six kinds of pieces, two colors, so you only need four bits of entropy per square/piece).

My idea is the following:

  • Each square is a circuit containing a signal generator connected to four bandpass filters at frequencies f1,f2,f3,f4 (multiplexed of course, so that we only need a few signal generators/filters)
  • Each piece has a RLC circuit in the foot (with fairly large inductance)
  • The circuit in each piece is tuned to resonate at frequency f1, f2, f3 or f4, or any combination of the four
  • To poll a square for information, we mux to the square's circuit and generate sin(f1) + sin(f2) + sin(f3) + sin(f4). Now we should be able to discern which piece is on the square based on the frequency response observed at the output of the bandpass filters

If I'm not completely missing something, the resulting PCB should be comparatively cheap to construct. But I'd love to hear if this is a less elegant solution than I think. Are there any problems with this idea?


The forums are retiring in 2021 and are now closed for new topics and comments.

1 year ago

I have a computer chess board with nothing but a momentary push switch on each Square. It works fine. Push the piece you want to move, then move it and again push where it lands.


1 year ago

Hi, Honestly I have no idea how 64 IR sensors and 32 IR transmitters would simplify things.


1 year ago

Well. I've read some more and it seems that the idea is quite good given that it's the one used by DGT which apparently supplies international chess tournaments. It's patented however. Kind of weird how freely patents are granted - it's not exactly a mind blowing concept.


Reply 1 year ago

You can always use a patent for private use, just not commercially ;)

For a different project a few years I considered a slightly different approach to a similar problem.
My case was not 64 positions and the resulting chess pieces, just basic positioning.
The entire multiplexing idea and RF signals is well and good but f you would prefer to stay away from the patented stuff you go optical.
I try to explain my approach while considering the amount of chess pieces and possible positions:
Preferably each square would have a small induction coil to provide wirless power to the chess pieces.
Additionally there will be an IR receiver in each square.
In every peice you have a receiver coil to provide electricity.
The power goes to a chip providing an optical signal code to identify the piece.
The squares only need to kept powered up while a piece is actually present.
If a removal is detected all remaining squares are activated to be able to recognise the new position once the piece has transmitted its code.
Optional: Only activate possible squares based on the type of piece and position it was on.
Wouldn't be too hard to program a small chip to send one of 32 possible ID codes...
And if you make it work properly you still claim a patent on it if you give me some credits for the idea ;)

Don't know enough about induction systems for identification purposes but somehow I think my idea is still a bit easier to realise and for more "elegant" and flexible.
No problem adding a bit more info and use the board for other games ;)
Only real downside might be the optical part.
Position problems can be address with a suitable reflector in the base but dirt or smear not so much.
But then again a proper chess player would keep his set and board nice and clean, would he? (Or she?)