Re: Switching between RFID antennas - one reader 100 antennas - how?

John Barrett wrote:

"John Barrett" <ke5crp1@xxxxxxxxxxx> wrote in message news:GgnMh.9962$8B1.4620@xxxxxxxxxxx

<yvonne.vanderbilt@xxxxxxxxxxxxx> wrote in message news:1174535138.538737.197700@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

On 22 Mrz., 07:06, "John Barrett" <ke5c...@xxxxxxxxxxx> wrote:

<yvonne.vanderb...@xxxxxxxxxxxxx> wrote in message

news:1174503030.828602.213300@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx


Hi,
I would like to connect multiple antennas (more than 100) to a RFID
reader and switch between these antennas using a microcontroller. Can
anyone point me to the right direction what type of switch can be used
for this?

Thanks
Yve

-----------------------------------------------------------------
Yvonne Vanderbild
mail: yvonne.vanderbilt@@@hotmail.co.uk
http://www.flickr.com/photos/7432253@N03/

100 antennas for one reader is getting a little outa hand in my opinion --
you might consider independent readers with 4-8 antennas + a microcontroller
(which you need anyway to handle antenna switching), and interconnect with
ethernet or multi-drop serial. i'm assuming the antennas will be at least
somewhat seperated, and you are creating a huge headache backhauling 100+
coax feeds, not to mention coax losses in the longer runs.

there are schematics and code online for building MCU based RFID readers for
just a few dollars in parts, and adding an 8 channel multiplexor to the
front end of one of those should be simple enough. RS-485 is easy enough to
cope with for the multidrop connection, or get an MCU with on chip ethernet.

Personally -- I would consider some of those 900mhz data radio modules for
the backhaul link -- thats the setup we are doing for a warehouse RFID
system -- 4 antennas per reader with 900mhz RF backhaul -- 50 readers for a
total of 200 antennas covering 20 dock doors, 5 fork lifts, and 25 key
transit points throughout the warehouse. We had some real fun setting up so
that the readers from one dock door didnt pick up traffic from adjacent
doors :)

Thanks guys for you suggestions, a bit more background on what I am
trying to build: I try to identify chess pieces on a 8x8 chess board.
The pieces have an implanted RFID glass-tag which I want to read from
below the chess board. The board is just a few millimeter thick and
made of plastic. One solutions of course is to move the antenna on an
simple xy slegde but I want to aviod moving/mechanical parts. I also
want to use RFID and not any of the alternative techniques like Reed-
contacts, etc.
So, the 64 antennas are on a regular 8x8 matrix, one below each field
which is 5x5cm. What ever the antennas pick up will be processed by a
microcontroller this gives the option to resolve ambiguities because
the MC knowns at least most of the time where the pieces stand and
only the piece that has been moved needs to be found. However, the
main thing I am worry about at the moment is how the MC switches the
antennas. One of the few things I konw about RFID is that the antenna
when in resonance carries up to 100V which is too much for most analog
switches and multiplexer.

Yve

-----------------------------------------------------------------
Yvonne Vanderbilt
mail: yvonne.vanderbilt@@@hotmail.co.uk
http://www.flickr.com/photos/7432253@N03/



thats too close -- the antenna in any square will likely pick up the chips in ALL the pieces !! not a bad idea in general concept -- but off the shelf RFID hardware has too much range for this particular application -- you need something much more tightly coupled and lower power -- it can still be based on RFID technology, but scaled down for the millimeter ranges that you are talking about -- I would SERIOUSLY look at the MCU based rfid reader schematics out there -- at least you have the circuit so you can tune the power such that you only activate the chip in the correct square. The antenna coils need to be scaled back -- less turns, such they create less field to activate adjacent chips, and only activate the closely coupled chip in a given square.

Since you are passing AC signals to the antennas, you could use 16 triacs to create the 8x8 matrix switch -- 8 for each row, 8 for each column -- the power should be low enough that the partially connected antennas will not activate their chips.

Here is a website with an MCU based RFID reader including source code for the MCU.

http://instruct1.cit.cornell.edu/courses/ee476/FinalProjects/s2006/cjr37/Website/index.htm





Just an additional thought -- use ferrite cores for the antennas to direct the field more vertically -- like a stick antenna -- short piece of ferrite with a few turns of wire around it.


I'd try it with pot core halves. Done that in the past, worked great and it's cheap.

Yve: At such short distances I'd operate it at lower signal levels so you don't run into lots of volts. Also, you can limit that with diodes if needed.

--
Regards, Joerg

http://www.analogconsultants.com
.

Relevant Pages