How electronic chessboards works
Cesar A. K. Grossmann
I'm curious to know how the DGT Chessboard works
(http://www.dgtprojects.com/eboard.htm). Can someone explain this to me?
In great details, please.
[]s
--
.O. Cesar A. K. Grossmann ICQ UIN: 35659423
..O http://www.LinuxByGrossmann.cjb.net/
OOO Quidquid Latine dictum sit, altum viditur
Mike N.
Thanks.
Mike
"Cesar A. K. Grossmann" <cesa...@bol.com.br> wrote in message
news:c0g8hg$16ko4n$1...@ID-210972.news.uni-berlin.de...
Noah Roberts
> Hi!
>
> I'm curious to know how the DGT Chessboard works
> (http://www.dgtprojects.com/eboard.htm). Can someone explain this to me?
> In great details, please.
>
> []s
--
"I'm a war president. I make decisions here in the Oval Office
in foreign policy matters with war on my mind." - Bush
David Richerby
> Cesar A. K. Grossmann wrote:
>> I'm curious to know how the DGT Chessboard works
>> (http://www.dgtprojects.com/eboard.htm). Can someone explain this to
>> me? In great details, please.
>
I think it's very simple RFID but I'm not at all sure. If that is the
case, it works essentially by the board emitting an electromagnetic field
which interacts with the different shaped aerials in the pieces in
different ways. The baord can then detect the changes in the field with
its own aerials.
Dave.
--
David Richerby Impossible Robot (TM): it's like a
www.chiark.greenend.org.uk/~davidr/ high-tech robot but it can't exist!
Cesar A. K. Grossmann
> Noah Roberts <nrob...@dontemailme.com> wrote:
>
>>My guess is magnets or RFID chips of some sort.
>
> I think it's very simple RFID but I'm not at all sure. If that is the
> case, it works essentially by the board emitting an electromagnetic field
> which interacts with the different shaped aerials in the pieces in
> different ways. The baord can then detect the changes in the field with
> its own aerials.
http://home.planet.nl/~jeroenvandorp/chess/dgtboard/dgtboard.htm suggest
RFID too, but how it works? I find this article
http://electronics.howstuffworks.com/smart-label.htm that gives some
hints about it.
I was thinking that all you need is to have the same is used in home
alarms, an magnetic and a switch that closes when the magnetic is near
(a magnetic under every chess piece, and a "magnetic" switch under every
cell of the chessboard). You mount the board and "tells" it that the
initial position is set up, and from that time the chessboard simply
tracks the conneting/disconnecting switches to know what piece goes to
what position, e. g., if in the first move the key at e2 opens and the
key at e4 closes, then the chesspiece that was in e2 was moved to e4,
the board then consults what it thinks was the current position before
the move, and learns that on e2 there was a pawn, so the move is a pawn
move from e2 to e4. Can this work?
Noah Roberts
What if two pieces are picked up and only one set down? What if two
pieces are picked up and both set down? I think these scenarios pose a
problem.
Cesar A. K. Grossmann
>
> What if two pieces are picked up and only one set down?
Like in a capture? Logic can deal with it with ease... Suppose that a
black and a white piece are picked, and a piece is set down in the place
of the black piece. It was a capture done by white. The same applies
when the piece set down is not placed in the place of any of the pieces
that where picked. It was an "en passant" capture.
> What if two
> pieces are picked up and both set down? I think these scenarios pose a
> problem.
Like in the castling? Again, the logic of the program can deal with it
with ease. Suppose that the rook at h1 and the king at e1 where picked,
and two pieces where placed on f1 and g1. It was the king-side castling.
But it's necessary that the chessboard "knows" the rules of chess, and
can detect what is happening.
Doom & Gloom Dave
Plus promotion!! Promotion is the real killer for this scheme.
Noah Roberts
>
>>
>> What if two pieces are picked up and only one set down?
>
>
> Like in a capture? Logic can deal with it with ease... Suppose that a
> black and a white piece are picked, and a piece is set down in the place
> of the black piece. It was a capture done by white. The same applies
> when the piece set down is not placed in the place of any of the pieces
> that where picked. It was an "en passant" capture.
What if the player decides not to make that move? Many OTB players feel
that you can decide not to make the move so long as you have not set
down the captured piece.
>
>> What if two pieces are picked up and both set down? I think these
>> scenarios pose a problem.
>
>
> Like in the castling? Again, the logic of the program can deal with it
> with ease. Suppose that the rook at h1 and the king at e1 where picked,
> and two pieces where placed on f1 and g1. It was the king-side castling.
> But it's necessary that the chessboard "knows" the rules of chess, and
> can detect what is happening.
In OTB games, which I see this thing as being the most useful for,
people often ignore rules about touching pieces, or picking up pieces,
and such things. I think the board would have to do more than rely on
where pieces are to get its information about what they are or it can
fail in many general use applications.
And like "Doom and Gloom Dave" mentions, pawn promotions cause problems
without ignoring tournament rules.
Cesar A. K. Grossmann
>
> What if the player decides not to make that move? Many OTB players feel
> that you can decide not to make the move so long as you have not set
> down the captured piece.
The chess rules tell that a pieced touched must be played, and so on.
*But* you can make the chessboard consider the lance incomplete until
you hit an electronic clock attached to the chessboard, or a switch that
tells the chessboard that you have completed the move.
> In OTB games, which I see this thing as being the most useful for,
> people often ignore rules about touching pieces, or picking up pieces,
> and such things. I think the board would have to do more than rely on
> where pieces are to get its information about what they are or it can
> fail in many general use applications.
I think that most of the cases can be handled if you sinalize when the
move was completed, so the chessboard will compare the new position with
the old position and make the necessary assumptions. Like a blind chess
game, where you don't see the pieces, only the moves.
> And like "Doom and Gloom Dave" mentions, pawn promotions cause problems
> without ignoring tournament rules.
Pawn promotion is the worst case until now. You must have a way to tell
the chess board what piece you choose. You can put 4 switches in a side
of the board, and when you promote a pawn, you also touches one of the
switches to tell the chessboard what piece you choose. It's the problem
with the RFID solution too, you must have some spare pieces to deal with
the promotion. AFAIK you can have only 2 queens of the same color at a
time in the board with DGT chessboards.
Any other problem?
[]s
P.S.: Sorry the "engrish"
--
.O. Cesar A. K. Grossmann ICQ UIN: 35659423
..O http://www.LinuxByGrossmann.cjb.net/
OOO Quidquid Latine dictum sit, altum viditur
http://www.clickfome.com.br/ - Clique e doe!
Soja Transgênica: http://tinyurl.com/2pfou
Noah Roberts
> Any other problem?
Many. 3 pieces are picked up: a white rook, a black pawn, and a white
bishop. 2 pieces are then set down, one where the white rook was, and
another where there was no previous piece. How are your magnet switches
going to be able to account for this?
Rearanging the board is common place in OTB. Especially if you are
studying. Older electronic boards had ways to indicate positional
setup, but I think what we have here is more sophisticated than those
older models.
I think your implementation makes too many assumptions, and something
that makes too many assumptions is bound to fail.
Finally, I don't see a clock or any switches to say that a move was
finalized or what piece has been put in what square at the link the OP
provided. There is an optional clock attatchment but it does not appear
to be a standard option.
You might be able to use magnets to represent piece color and type, but
such a configuration would be much more complicated than what you have
suggested. I suppose you could use field polarity and direction to
signify what piece is being moved, but such configurations would have to
account for radial movement. It is probably possible.
I think, though, that some sort of low class RFID would be much easier
and is most likely what was used.
Phil Innes
This is not how sensory boards works at all.
Remember that Disney chess set that Grandma gave you one Christmas? You
could use those, or any set of pieces on a sensory board! The software is
set to understand that the piece that starts on f8 is the Black King's
Bishop, and remembers the subsequent action of this Bishop and its location
throughout the game. No need for special antennae in the piece : )
Cordially Phil
ps: check out www.ruschess.com at the top left of the screnn it says
'distance chess' where there is much more information on how these systems
work. If you are interested in the technical basis of the system, write a
question to the letters column at the same site.
"David Richerby" <dav...@chiark.greenend.org.uk> wrote in message
news:W7v*m1...@news.chiark.greenend.org.uk...
Dr. David Kirkby
>
> I'm curious to know how the DGT Chessboard works
> (http://www.dgtprojects.com/eboard.htm). Can someone explain this to me?
> In great details, please.
>
> []s
I don't know, but here is my guess.
Each piece has a tuned circuit in it, consisting of a capactor and
inductor. The tuned circuit would be 'resonate' at one frequency, and
therefore absorb energy at that frequency, and not at others too
distant. (f = sqrt ( L/C) where L is the inductance in Henrys, C the
capacitance in Farads, and f the resonate freqency in Hz).
The board emits a range of radio freqencies and determines which
frequency the piece aborbs at. So white pawns abosrb at says 10 MHz,
black ones at 15, white rooks at 20 MHz ... or similar.
There is a device called a 'grid dip meter' which is used to determine
the resonate frequency of a tuned circuit. I have one here, and it
covers (with a few plug in coils), somewhere between about 1 and 200
MHz. I've used that to tune the resonate frequency of amplifier
components.
The idea of one poster of different ariels is not really practical, as
the resonate frequency of ariels is of the order of their size (i.e.
the wavelength would be of the order of mm), so the frequency of
resonance is likely to be around 100 GHz. That is just not practical
on a cheap board. Wheras making a tuned circuit with an inductor and
capacitor resonate at much lower freqencies would be very practical.
As I say, I'm only guessing. But that's my considered view as an
electronics engineer, who specialises in radio freqency engineering.
Dr. David Kirkby PhD CEng MIEE.
David Richerby
> This is not how sensory boards works at all.
>
> Remember that Disney chess set that Grandma gave you one Christmas? You
> could use those, or any set of pieces on a sensory board! The software
> is set to understand that the piece that starts on f8 is the Black
> King's Bishop, and remembers the subsequent action of this Bishop and
> its location throughout the game. No need for special antennae in the
> piece : )
Really? How does that work?
Dave.
--
David Richerby Confusing Love Apple (TM): it's like
www.chiark.greenend.org.uk/~davidr/ a tasty fruit that you can share
with someone special but you can't
understand it!
David Richerby
> The idea of one poster of different ariels is not really practical, as
> the resonate frequency of ariels is of the order of their size (i.e.
> the wavelength would be of the order of mm), so the frequency of
> resonance is likely to be around 100 GHz. That is just not practical
> on a cheap board. Wheras making a tuned circuit with an inductor and
> capacitor resonate at much lower freqencies would be very practical.
Yes. When I said aerial, I was using the term extremely loosely and had
some manner of coil (=inductor) in mind. I forgot that you'd need a
capacitor, too. Your post makes this sound much more plausible. :-)
Dave.
--
David Richerby Expensive Carnivorous Pants (TM):
www.chiark.greenend.org.uk/~davidr/ it's like a well-tailored pair of
trousers but it eats flesh and it'll
break the bank!
Cesar A. K. Grossmann
>
> Many. 3 pieces are picked up: a white rook, a black pawn, and a white
> bishop. 2 pieces are then set down, one where the white rook was, and
> another where there was no previous piece. How are your magnet switches
> going to be able to account for this?
It was one move? If it was one move, it is a weird one... BUT, if you
have a chess computer attached to the board, it must have to find the
only legal move that goes from position BEFORE to position AFTER.
> Rearanging the board is common place in OTB. Especially if you are
> studying. Older electronic boards had ways to indicate positional
> setup, but I think what we have here is more sophisticated than those
> older models.
Well, if you are studying, then the simplest form of the electronic
board I was thinking is of no help. There's the need to inform it that
you are going back to some old setup, and this indicates the need for
some computerized interface (maybe a interface to a chess program).
> I think your implementation makes too many assumptions, and something
> that makes too many assumptions is bound to fail.
I'm not makink any assumptions, or so I think. It's an evolving idea,
you feedback (and the feedback of the others) is giving me an idea os
what kinds of problems DGT people has to come around, and how complicate
things can be.
I don't know if this will end in something that can hurt my feet or if
it will be only an idea.
> Finally, I don't see a clock or any switches to say that a move was
> finalized or what piece has been put in what square at the link the OP
> provided. There is an optional clock attatchment but it does not appear
> to be a standard option.
To the DGT Chessboard? I don't know how it works. I have an idea:
probably it uses RFID tags to identify every piece, but I don't know
anything more than that.
> You might be able to use magnets to represent piece color and type, but
> such a configuration would be much more complicated than what you have
> suggested. I suppose you could use field polarity and direction to
> signify what piece is being moved, but such configurations would have to
> account for radial movement. It is probably possible.
If you use a swith that consist in a magnet between two plates, then you
can think this way: if the basis of the piece is N, then the magnet in
the switch is repelled, and touchs the plate that is under it. if the
basis of the piece is S, then the magnet is atracted, and it touches the
plate that is above it:
-------------< (fixed, meaning 'white piece')
+---+
| N | ^
----+ | moves up an down
| S | v
+---+
-------------< (fixed, meaning 'black piece')
But I think it is clumsy and unnecessary (in the making you must assure
that every black piece has the north pole pointing down, and vice-versa,
and that the north pole is "pointing up", for all the 64 cells of the
chessboard). You can use a switch that consist in two plates that are
atracted when immersed in a magnetic field (some home alarms use this,
it's very simple), and doesn't have to bother with the orientation of
the magnets in the bases of the pieces. BUT you don't have a way to know
if the piece is black or white, you only have the information that
there's a piece on that cell. You MUST have to incorporate some chess
computer in it if you want something more usefull. BUT you has to do
that anyway, so...
> I think, though, that some sort of low class RFID would be much easier
> and is most likely what was used.
I think you are right, but what is the cost for that? You will need 64
readers, and 32 RFID tags, all of them marked (like 'black pawn', 'black
knight', etc.). The electronic doesn't need to figure out (via
backtracking) what piece is where, but it needs to accept a more complex
signal (a signal that says "I'm cell e4, and a white pawn just landed
here").
And the design has more questions:
1. when a piece is 'picked up' or 'lands' on a cell, the signal is
feeded immediately or there's some delay, to prevent that a "j'adoube"
is interpreted as the starting of a move?
2. How does you inform that the move is completed? A clock must be attached?
3. When studying, how you tells the chessboard that you are not making
moves, but that you are "undoing" moves? I think that the simplest way
is to tell the board to "look to the other side" when you are placing
the pieces and when you are ready you tell it to read the position.
[]s
Phil Innes
Sorry for late response to your question.
You can talk directly with the inventor of the product at www.ruschess.com -
where you can ask your question in English (top right, 'Feedback') and
receive a response in more or less the same language :)
To inform folks here and at rec.games.chess.computer, can you please write
what you understand to be the reply?
Additionally, at the top left of the screen at ruschess are some articles on
the actual practice of play of DC (Distance Chess) using sensory boards
where you will see my name listed along with the high and mighty :).
Cordially, Phil Innes
"David Richerby" <dav...@chiark.greenend.org.uk> wrote in message
news:cJd*kA...@news.chiark.greenend.org.uk...
David Richerby
> Sorry for late response to your question.
No problem!
> You can talk directly with the inventor of the product at
> www.ruschess.com - where you can ask your question in English (top
> right, 'Feedback') and receive a response in more or less the same
> language :)
Aha. I'll do that when I get the chance and post the results here.
Dave.
--
David Richerby Accelerated Robot (TM): it's like
www.chiark.greenend.org.uk/~davidr/ a high-tech robot but it's twice
as fast!
Jet Jaguar
Excalibur electronic boards work. When you move a piece, you tap the
square it starts from and then the square you're moving to. It knows
where the pieces are because it knows the starting position and keeps
track of where they moved. Your idea of magnetic sensors would
eliminate the need to tap the board since the computer would know when
a piece has been picked up and put down.
---
Jet Jaguar
I have a spam blocking address. Replying to me is like pulling teeth.
Cesar A. K. Grossmann
> Your basic idea would work. In fact, this is pretty much how the
> Excalibur electronic boards work.
What model? One of this http://www.chess4less.com/2-computers.htm ?
Jet Jaguar
<cesa...@bol.com.br> said:
>Jet Jaguar wrote:
>
>> Your basic idea would work. In fact, this is pretty much how the
>> Excalibur electronic boards work.
>
>What model? One of this http://www.chess4less.com/2-computers.htm ?
>
>[]s
AFAIK, all of them. I know for sure the Ivan 2 and the Chess Station
work this way.
Ben Bulsink DGT Projects
fundamentals.
This is how it works:
Study patent documents US5129654 and 5,188,368 and (NL 1009574 or US
09/345593 or DE 19929931A1 (being identical).
They describe some backgrounds.
Some on the insight of the board:
Under the squares, a polyester foil is mounted, with silver-ink
printed traces. The traces form 16 loops: one loop under each column
and one under each row. The loops are connected to some electronics at
the side of the board, with overall PCB size about 4 cm x 13 cm.
The electronics contain a microprocessor with RS232 interface, and
various logic ports.
The column loops go through a 8 to 1 multiplexer analog switch, which
is connected to an analog amplifier input. The output of the amplifier
goes throuh a 1 to 8 multiplexer analog switch, which connect to the 8
row loops.
The pieces contain a LC tuned resonator on a ferrite rod, in the
frequency range of 90-350 KHz.
By selecting one loop and one row, the ampifier is a feedback system
that oscillates when a piece is on the selected square. determine the
frequency (takes 3 mS), do it over all squares, communicate, ready.
Hope this clarifies a lot.
Please send any further questions directly to my email address
b...@dgtprojects.com, while I am too busy to follow all kinds of
threads.
Ben Bulsink
"Cesar A. K. Grossmann" <cesa...@bol.com.br> wrote in message news:<c0qvn8$19td5k$1...@ID-210972.news.uni-berlin.de>...
Cesar A. K. Grossmann
>
> Study patent documents US5129654 and 5,188,368 and (NL 1009574 or US
> 09/345593 or DE 19929931A1 (being identical).
Do you know is there's some sensory chessboard that is not covered by
patents?
> The pieces contain a LC tuned resonator on a ferrite rod, in the
> frequency range of 90-350 KHz.
So, no RFID/SmartTag...
Tugy tuk
Quadibloc
> The pieces contain a LC tuned resonator on a ferrite rod, in the
> frequency range of 90-350 KHz.
That avoids reliance on such high-tech as RFID, but it still allows individual pieces to
be differentiated, making the board more convenient than old-style "sensory" boards
which still required players to take certain artificial steps when playing.
Frank Camaratta, the man behind House of Staunton, holds a patent on a method of
making heavily-weighted chess pieces that can still be used with DGT, patent US20160059115A1.
John Savard
