Heya!
I want to design a chess board which can ID the various pieces. I’ve read plenty posts here before, and I want to clarify I would like there to be no visible electronics and for the identification to not rely on pre-programmed positions. I’ll be using Arduino 4, but i’ve got a Raspberry accessible too.

I’ve been researching using LC frequencies etc., but seeing as there’s been plenty of threads regarding RFID, without anyone every actually trying it… I think it could be fun to give it a shot!

I’ve considered triangulation with a few receivers, or multiplexing one receiver, but have ultimately decided budget is not a problem here, and I’ll have 64 receivers, one per square of the chess board.

For this purpose i’ve already stumbled upon the PN532, since it’s rather small, and can be wired in series, which sounds like it would work wonders for this project.

But it’s here I must ask, as I’m not entirely sure if NFC or RFID is better for this? I want there to be at least a thin layer of acrylic plastic and veneer across the board, as well as enough room for LED lights under it, so I need at least a few centimeters read-range. However, I don’t want the receivers reading neighboring squares (maybe insulating the grid beneath the board would work?) either way, as I understand NFC is shorter ranger, but does that mean the aforementioned receiver won’t work as it’s an RFID antenna? Or are there other, better options? Any receivers that are compact, and ideally see through would be amazing, though I assume not all can be wired in series, which is presumably preferable.



TL;DR:

64 antennas/receivers to identify chess pieces on a chess board, is RFID or NFC the way to go, any suggestions?

 

Only one RFID receiver that can read an identity, and then one RFID transmitter with the piece name encoded in each piece, and one power pickup coil embedded in each piece, to power the transmitter in that piece, when driven.
THEN, under the board, a transmit coil under each square, wired in a eight by eight X-Y grid, so that all the squares can be scanned one at a time, and the computer will see the response from the receiver as each position is powered. Empty positions will send no signal. All of the fixed circuitry can reside on a circuit board under the board, and it can have a USB link connecting to a nearby computer. The code will actually be rather simple.

 

One RFID receiver that can read an identity, one RFID transmitter with the piece name encoded, and one power pickup coil embedded in each piece, to power the transmitter when driven.
THEN, under the board, a transmit coil under each square, wired in a ten by ten X-Y grid, so that all the squares can be scanned one at a time, and the computer will see the response from the receiver as each position is powered. Empty positions will send no signal. All of the fixed circuitry can reside on a circuit board under the board, and it can have a USB link connecting to a nearby computer. The code will actually be rather simple.

Referring to the start, why would there be a receiver in the pieces? And why a 10x10 grid rather than 8x8, so the crossroads are under each field or?

 

OK, my error on the number of squares, I guessed wrong.
As for the "receiver in each piece, NO, it would be a pick-up coil to get power from the radiating coil beneath that square in the board, to receive power for that RFID transmitter. The coil in each chess piece just picks up enough power to activate the RFID transmitter for a single transmission. Evidently I was not clear enough in describing what I am seeing.

The operation will be that each square is powered in sequence every few seconds, and whatever piece is sitting on that square transmits it's identity. The computer handling the display will deliver the sequence to power the coils in the squares. 8 by 8 is much more convenient, any way. So the computer will know which square it powered, and receive the ID of the piece on that square. So only one receiver needed.
Is this explanation better???

 

OK, my error on the number of squares, I guessed wrong.
As for the "receiver in each piece, NO, it would be a pick-up coil to get power from the radiating coil beneath that square in the board, to receive power for that RFID transmitter. The coil in each chess piece just picks up enough power to activate the RFID transmitter for a single transmission. Evidently I was not clear enough in describing what I am seeing.

The operation will be that each square is powered in sequence every few seconds, and whatever piece is sitting on that square transmits it's identity. The computer handling the display will deliver the sequence to power the coils in the squares. 8 by 8 is much more convenient, any way. So the computer will know which square it powered, and receive the ID of the piece on that square. So only one receiver needed.
Is this explanation better???

Right, I assumed that’s what you meant, but just wanted to make sure!

I like that idea… Originally I just imagined some kind of battery in each piece, but having a coil transmit power wirelessly just for the RFID/NFC tag in the piece to transmit its encoded signal is quite clever…

And there’s no risk of the wireless charger coil interfering with the signal for the antenna? If so, having a layer of antennas, then a layer of wireless power transmitters, then the board (with some room to jam in a few LED’s) sounds quite great…

I haven’t yet found an option better than the PN532, so I just need to find a decent power coil and receiver, as well as the proper RFID chip or NFC tag for this, and we’re good… Thank you!

 

OK, my error on the number of squares, I guessed wrong.
As for the "receiver in each piece, NO, it would be a pick-up coil to get power from the radiating coil beneath that square in the board, to receive power for that RFID transmitter. The coil in each chess piece just picks up enough power to activate the RFID transmitter for a single transmission. Evidently I was not clear enough in describing what I am seeing.

The operation will be that each square is powered in sequence every few seconds, and whatever piece is sitting on that square transmits it's identity. The computer handling the display will deliver the sequence to power the coils in the squares. 8 by 8 is much more convenient, any way. So the computer will know which square it powered, and receive the ID of the piece on that square. So only one receiver needed.
Is this explanation better???

Developing from my last response, I assume I should use some form of ISO14443 tag? Or is one of the small cubic RFID chips better for this? Either way how do I go about powering one of these things? Do I need to wire them to some specific power receiver, or can they just take the power they need directly from the charging coil? Sorry, I’m very new to RFID/NFC, and a lot of the resources online are somewhat overwhelming.

 

In another thread, I proposed a scheme using a colored circle on the bottom of each piece and detector in each square using an RGB LED plus a phototransistor. I still think this would be far simpler than 64 RFID or MFC readers.

No one seemed to like it though, without giving any reasons, so I am possibly missing something. Discriminating 12 colors can’t be very difficult. You would simply need to illuminate with R, then G, then B, and measure the reflection for each. It certainly would not be prone to interference and crosstalk problems.

 

In another thread, I proposed a scheme using a colored circle on the bottom of each piece and detector in each square using an RGB LED plus a phototransistor. I still think this would be far simpler than 64 RFID or MFC readers.

No one seemed to like it though, without giving any reasons, so I am possibly missing something. Discriminating 12 colors can’t be very difficult. You would simply need to illuminate with R, then G, then B, and measure the reflection for each. It certainly would not be prone to interference and crosstalk problems.

On a technological level it’s a great idea (and I’ve read that thread throughly too) what I want to avoid however, is the board having any visible electronics, whilst this solution would rely on there being some level of translucency, whilst I ideally would like a thin layer of wood veneer over the whole thing.

 

Given that RFID transmitters, or transponders, get embedded with every item, they must be small and inexpensive, and able to be programmed with a particular ID. The power from the pickup coil could be fed thru just a diode to a small capacitor to just power the RFID module for long enough to transmit it's ID to the receiver. So not much complexity at all. Certainly the magnetic field will pass thru a wood layer of any reasonable thickness. And the better part is that the transmission is only power, no commands or data. And the sending portion will just need to be coils, nothing else.

 

Given that RFID transmitters, or transponders, get embedded with every item, they must be small and inexpensive, and able to be programmed with a particular ID. The power from the pickup coil could be fed thru just a diode to a small capacitor to just power the RFID module for long enough to transmit it's ID to the receiver. So not much complexity at all. Certainly the magnetic field will pass thru a wood layer of any reasonable thickness. And the better part is that the transmission is only power, no commands or data. And the sending portion will just need to be coils, nothing else.

I really like that… I just need to settle on RFID transmitter/transponder, wrap my head around the charging mechanism, and this sounds like a pretty clever build… It’s elegant, removing the need for batteries, all that’s left is for me to have some wiring skills and we’re golden.

I’m assuming I’d have to wire the power transmitters in a multiplexed grid… But actually, powering one coil at a time removes the risk of neighboring pieces interfering with their signals entirely, and then I suppose I just need to time the antennas to read simultaneously… Unless I simply have them all powered at all times as the coils do a sweep, as the program will now only one coordinate can be active at a time… I’ll have to figure that out later! Hahah, one step at a time!

 

The coils in the chess pieces are passive receivers so they are always ready to receive the power when it is available. THAT is how the computer knows which square that piece is sitting on. And the coils in the chess board must only have one switched on at a time for that reason. Also, only powering one at a time consumes much less power. So the whole system only needs one power generator and one RFID receiver.
Unfortunately I do not yet know exactly what frequency the power delivery system works at. BUT it might be similar to what they use for wireless cell phone charging.

 

The coils in the chess pieces are passive receivers so they are always ready to receive the power when it is available. THAT is how the computer knows which square that piece is sitting on. And the coils in the chess board must only have one switched on at a time for that reason. Also, only powering one at a time consumes much less power. So the whole system only needs one power generator and one RFID receiver.
Unfortunately I do not yet know exactly what frequency the power delivery system works at. BUT it might be similar to what they use for wireless cell phone charging.

Needing only one RFID receiver would be amazing… Then in theory, one wouldn’t even need to multiplex it in any complicated grid-based way, simply make the antenna/antennas large enough to cover the desired board, as only one piece would be active at a time anyways… Though in previous threads on here, I’ve seen that RFID receiver antennas can be a little complicated to set up? Either way, this is an elegant and robust approach… I like it!

 

OK, and certainly the antenna will need to cover the whole play area of the chess board. THAT should not be difficult. Given that the RFID transmitters have some fairly well defined range, that is one of the specifications you will need to know as part of your transmitter selection process. A range of ten inches should be adequate, possibly six inches could work. The receive antenna could be embedded in the chess board, probably.

 

OK, and certainly the antenna will need to cover the whole play area of the chess board. THAT should not be difficult. Given that the RFID transmitters have some fairly well defined range, that is one of the specifications you will need to know as part of your transmitter selection process. A range of ten inches should be adequate, possibly six inches could work. The receive antenna could be embedded in the chess board, probably.

Just wonder how reliable a single longer-range antenna would be at the farther edges of the board (i intend to have more than just chess on it eventually, so some larger games too), especially through a few layers of the internal grid with LED’s etc… But I suppose having for example 4 antennas spread across the board could work too, as I only ever need to read one transmitter at a time anyways, so overlaps won’t be a problem.

Regardless, a multiplexed power-supply system activating only one piece at a time is definitely the way forward, by far the most reliable and simple approach. Curious why nobody has done it before.

 

OK, and certainly the antenna will need to cover the whole play area of the chess board. THAT should not be difficult. Given that the RFID transmitters have some fairly well defined range, that is one of the specifications you will need to know as part of your transmitter selection process. A range of ten inches should be adequate, possibly six inches could work. The receive antenna could be embedded in the chess board, probably.

Additionally it seems long-range RFID readers aren’t that cheap, or easy to come by, found one with a 30cm radius for smaller transmitters, sitting on a 85$ price tag… Seems I’d need to make a custom antenna, or have multiple readers, I need to do more research.

 

You said:" Curious why nobody has done it before. " I have an answer for that, which is that a major portion of my career was mostly finding solutions for problems in manufacturing that others had not come up with a satisfactory solution for. So visualizing a concept and then figuring out the details was a basic activity much of the time. Practicing tends to improve a skill. NOW, it tends to bug folks around me when I suggest solutions to problems. So it is fun to create a concept for the chess board challenge.
This is the second thread about it, I don't recall what I suggested last time.

 

You said:" Curious why nobody has done it before. " I have an answer for that, which is that a major portion of my career was mostly finding solutions for problems in manufacturing that others had not come up with a satisfactory solution for. So visualizing a concept and then figuring out the details was a basic activity much of the time. Practicing tends to improve a skill. NOW, it tends to bug folks around me when I suggest solutions to problems. So it is fun to create a concept for the chess board challenge.
This is the second thread about it, I don't recall what I suggested last time.

I love some good problem solving, and I must give you cheers for an idea that makes me go «that’s so logical, why didn’t I think of that» without me being able to think of that anytime soon, hahah.

The SEN-14066 looks rather promising, claiming long ranges (1-2 feet with onboard antenna, more with external antenna, which seems bulky but is an option), high speed, and other functionality. It is however, stupid expensive. Not sure which approach here is best, but I’m certain I can find some middle ground with these readers. Hopefully.

 

It might be simple to receive the data, I have no information about the format or the modulation scheme. That is a technology I have never used. But possibly others around here have experience in that realm.

 

I just want to point out that NFC tags are powered by the “receiver” in normal operation. That’s just how they operate.

There is a 13.56MHz signal transmitted by the receiver which is inductively coupled to the tag. The tag is powered by it, and to communicate with the receiver it uses load modulation. It can’t transmit so it switches resistors in and out of circuit to vary the load, and so the current draw, from the receiver.

The receiver decodes the changes as data.

 

NFC tags are cheap and easily programmed. That can also be very small (smaller than the base of a medium sized chess piece. The are also very thin, much thinner than a typical PCB.