RFID tag reader?

Tar.Aldarion

Anybody here ever try to make one?
I am going to, what was it like? Run in to many problems?

KTRIC

I was looking into an off-the-shelf test kit for work last week and it was in the region of $3500. The initial outlay isn't worth it at the moment to be honest. I wouldn't even know where to start with a DIY job !!

Tar.Aldarion

This is a college thing, I have to make an RFID tag reader/wireless etc for my project, as a beginning.

KTRIC

Tar.Aldarion wrote:

This is a college thing, I have to make an RFID tag reader/wireless etc for my project, as a beginning.

Ah right, one of the biggest problems I had was finding a supplier. Have you seen anything online regarding sourcing parts etc ?

Tar.Aldarion

I have seen a couple of evaluation kits and some form of guide on how to build a lot of circuits. I am now looking at an RFID forum(there is a tag supplier guy there).
I didn't even know what it was until today, this will be fun.
I imagine I have to make a lot of the parts..

KTRIC

Tar.Aldarion wrote:

I imagine I have to make a lot of the parts..

Oh yeah I think you're in for a shock.

Tar.Aldarion

Like this?
http://www.eng.tau.ac.il/~yash/kw-usenix06/index.html

eeeeeeeeeeeeeek.
Building a reader is just a 'pass' grade.

copacetic

could you use a front end module like that Tar, or do you have to build the whole thing from scratch?

Tar.Aldarion

What front end module? Did they get the reader and build stuff from there?
I may have to build it all, not sure.

copacetic

Tar.Aldarion wrote:

What front end module? Did they get the reader and build stuff from there?
I may have to build it all, not sure.

this one:

RFID reader module we used was the Texas Instrument (TI) S4100 Multi-Function reader module, [TI03].

From the picture of the cct board with this on top there is a fair bit in this. I doubt they will expect you to do this yourself as you would need to have RF front end and a baseband digitial cct to go with it. which would be a trifle difficult. Is it a long term project or a few weeks?

Boston

Front end typically refers to the antenna and the usually the AD convertors. You should not be buliding the front end, when off the shell componenets will do.

Tar.Aldarion

4th year college project, so months. I would have no clue how to make it.
The first part of my project is 'Development of a standalone automated tag reader', hmm.
I imagine I would have to make the antenna etc. The reader itself maybe?
In an email he suggested I might have to make readers, or we may be able to buy them.

easyontheeye

can i ask are you designing an application that will use this reader or are you just building a reader? I used an off the shelf reader for an indoor positioning system...building an app to talk to the reader was hard enough! lol

Boston

Well nailed him down. Building the front end would be a final year project in an of itself, let alone an dynamic routing application on top of that.

easyontheeye

Boston wrote:

Well nailed him down. Building the front end would be a final year project in an of itself, let alone an dynamic routing application on top of that.

exactly thats what mines was...i wouldnt see the interest in building a reader, bit like reinventing the wheel unless your trying to achieve bigger transmission distances etc...you doing electronic eng i assume.

Boston

He is indeed. Final year projects are meant to be unique to a degree, or a new take on something old. The software routing part is the only unique aspect i can see. Wheres the point in building something you can buy off the shelf? I'd personally question the amount of work being asked of you here.

Tar.Aldarion

easyontheeye wrote:

can i ask are you designing an application that will use this reader or are you just building a reader? I used an off the shelf reader for an indoor positioning system...building an app to talk to the reader was hard enough! lol

Apparantly both.

Project milestones

Milestones:

Pass:

Development of a standalone automated tag reader

Average:

Provision of wireless interface onto the tag reader and development of basic system to control a network of such readers

Good:

Development of a more complex software system to manage the location of users and their devices and to provide an external software interface to provide remote access to the system to third party communication system.


Very Good:
Demonstrate a system with capability to re-direct incoming call attempts based on simple user configured rules.

Excellent:
Support for more complex subscriber configured rules in addition to non-voice call communication paradigms.

I'd be lucky if I got to average.
I'm In elec+comp eng yeah.

Well nailed him down. Building the front end would be a final year project in an of itself, let alone an dynamic routing application on top of that.

Thanks, I'll have a chat with him tomorrow, so i should suggest that I do the same as those at http://www.eng.tau.ac.il/~yash/kw-usenix06/index.html?
Or should I get all of that stuff off the shelf?

easyontheeye

Tar.Aldarion wrote:

Apparantly both.

Project milestones

I'd be lucky if I got to average.
I'm In elec+comp eng yeah.


Thanks, I'll have a chat with him tomorrow, so i should suggest that I do the same as those at http://www.eng.tau.ac.il/~yash/kw-usenix06/index.html?
Or should I get all of that stuff off the shelf?

eh that sounds very similar to my college project except i was utilising Bluetooth and RFID together...you would be mental to try and build the reader..think about it, you have to design the user interface, the backend database for storing position coordinates, users, info etc, design the coordinate logic map that to a graphical map of the building, design the software that will interface with the reader.... all that in itself is still a massive project for 2 people , btw i used Java and mySQL

Boston

Well just get him to sign off on the level. If he wants RTL then I think part one will be very hard. If he allows Integrated circuits like the TI S4100 then it shouldn't be a problem. At the very least get him to say you don't have to build you own antenna. The theory behind that goes into electromagnetic realms which I personally found quiet difficult. Remember you'll have to be able to do up a report on all your work.

Boston

easyontheeye wrote:

eh that sounds very similar to my college project except i was utilising Bluetooth and RFID together...you would be mental to try and build the reader..think about it, you have to design the user interface, the backend database for storing position coordinates, users, info etc, design the coordinate logic map that to a graphical map of the building, design the software that will interface with the reader.... all that in itself is still a massive project for 2 people , btw i used Java and mySQL

He wouldn't necessarily have to provide that level of functionallity. Something else also strikes me, you'll need to build three readers to demonstrate everything ( 3 points to locate an object in a plane) but it only mentions one. Are the other phases simulation only? Find that out as well. A matlab simulation of some C++ code would be very hardy.

copacetic

Boston wrote:

Front end typically refers to the antenna and the usually the AD convertors. You should not be buliding the front end, when off the shell componenets will do.

Not usually in RF, the front end is after the antenna, but may or may not included the baseband digital.

Tar.Aldarion

easyontheeye wrote:

eh that sounds very similar to my college project except i was utilising Bluetooth and RFID together...you would be mental to try and build the reader..think about it, you have to design the user interface, the backend database for storing position coordinates, users, info etc, design the coordinate logic map that to a graphical map of the building, design the software that will interface with the reader.... all that in itself is still a massive project for 2 people , btw i used Java and mySQL

Thanks for that, how did you fare with your project? Any advice? :-) I have no experience for mysql, what did you use both languages to do?

You might want to split this off to a new thread.

anyway, The reader itself will have to implimented on some hardware I persume, thats why i asked about fpga's or asic's. Alternatively you might be allowed to solely develope a software model.

Make life easy on yourself.

Give yourselve
1) 3 antanneas
2) Power RFID tags for better range
3) send very basic data using simple modulation such a PAM (simple to implement.
4) Go slap who ever assigned you that project and tell them if you actually implimented an UWB system it would not be legal (as of 3 months ago).
5) If I was doing that project, I'd use embedded erlang for my routing of an add hoc network.
6) Limit it to one user at a time.

The first part will be easy enough, the rest will be difficult.

Thanks for that, why was UWB made illegal recently?
I don't know what embedded erlang is, so maybe I have a lot to do...
I have a basic knowledge of vhdl, just starting to learn it, did simple compnants, used matlab a couple of times so a novice.

Boston

We'll have to agree to disagree on that point copacetic.

copacetic

if at all possible I'd avoid vhdl if you can, if you haven't got much experience of it. Getting a fpga demo board up and running can really suck the time. If they have a demo system up and running already with shell code you can add to then it won't be too bad.

It sounds like you need a lot of work on specing it asap otherwise you will find yourself knee deep in it in 6 months.

Tar.Aldarion

Boston wrote:

Well just get him to sign off on the level. If he wants RTL then I think part one will be very hard. If he allows Integrated circuits like the TI S4100 then it shouldn't be a problem. At the very least get him to say you don't have to build you own antenna. The theory behind that goes into electromagnetic realms which I personally found quiet difficult. Remember you'll have to be able to do up a report on all your work.

The theory is ok(The only bit I feel at all confident about), but I haven't built much circuits, I'll try and get him to specify what he wants tomorrow.

He wouldn't necessarily have to provide that level of functionallity. Something else also strikes me, you'll need to build three readers to demonstrate everything ( 3 points to locate an object in a plane) but it only mentions one. Are the other phases simulation only? Find that out as well. A matlab simulation of some C++ code would be very hardy.

Are you saying it would be more difficult if it was simulation?

Three readers hmm, I may have to buy some. An evaluation kit maybe too?
Man this is fecking hard what with exams and other smaller projects to do.

if at all possible I'd avoid vhdl if you can, if you haven't got much experience of it. Getting a fpga demo board up and running can really suck the time. If they have a demo system up and running already with shell code you can add to then it won't be too bad.

It sounds like you need a lot of work on specing it asap otherwise you will find yourself knee deep in it in 6 months.

With my lecturer in VHDL, I imagine I'll have to be ace at it in a couple of weeks. :-)

copacetic

Boston wrote:

We'll have to agree to disagree on that point copacetic.

fairy nuff, it's a minor point, I did this for a living for 8 years though!

Boston

Tar.Aldarion wrote:

Thanks for that, how did you fare with your project? Any advice? :-) I have no experience for mysql, what did you use both languages to do?



Thanks for that, why was UWB made illegal recently?
I don't know what embedded erlang is, so maybe I have a lot to do...
I have a basic knowledge of vhdl, just starting to learn it, did simple compnants, used matlab a couple of times so a novice.

Mysql is a database management langauge.

Erlang is a functional language developed by Ericsson's specifically for routing vocie calls to nodes.

VHDL is handly if you decide to implement on an FPGA (thought that would probably be extremely difficult).

Matlab is great for software simulations of communications systems.

UWB as used in the states uses a spectrum policy of underlay which is currently (Unless something changed recently) not allowed in Ireland (and afaik the rest of europe).

Boston

copacetic wrote:

if at all possible I'd avoid vhdl if you can, if you haven't got much experience of it. Getting a fpga demo board up and running can really suck the time. If they have a demo system up and running already with shell code you can add to then it won't be too bad.

The man is right, took me three months to get up an running on an FPGA demo bord for my final year project.

Tar.Aldarion

Thanks for those answers. I too know the pains of VHDL, it will be my hardest exam most like, what with all he wants us to know and be able to do in a few months.
Oh well, I'll take notes on these replies and talk to him tomorrow, thanks!

nellyshark

The Institute of technology Tralee have an RFID department *I Think*, maybe you could get in contact with them.

Tar.Aldarion

Thank you, I'll look in to it.

Do I really need three readers to locate the tag? I think I can do it with one? (RFID-Radar?) ?
Although, do I really need to know it's position and not just that it is within the range of the reader. I just need to know that the tag is in a no-go area and calls are to be diverted, so maybe I don't need to know the exact location?

Boston

Ok it would seem that Comreg has started issuing licenses for testing UWB equipment. If you're department has one you should be good to go in that direction, but i wouldn't advise it. RFID has been around a long time and theres a wealth of data out there.

Boston

Tar.Aldarion wrote:

Thank you, I'll look in to it.

Do I really need three readers to locate the tag? I think I can do it with one? (RFID-Radar?) ?
Although, do I really need to know it's position and not just that it is within the range of the reader. I just need to know that the tag is in a no-go area and calls are to be diverted, so maybe I don't need to know the exact location?

Well you need three values to determine position. You might get away with two readers and using some other property to determine the third variable, others would have more information on that. I hoesntly don't see how you can get anything bar distance from node using a single reader.

Tar.Aldarion

Hey thanks, I discovered you can d it with one reader, using signal strength, angles etc. i will have to see how hard, and if there is software expense, for that.

Boston

Interesting. You must post up how that works if you get it working. Personally I'd push for the software simulation of the routing rather then building complete position extrapolation hardware.

Tar.Aldarion

Yeah, I'll post it up if I do it that way. If software is easier I woul hope to do it that way. That would be all matlab I take it.

copacetic

not sure if they will let you stick with all mathlab, maybe to simulate, but if you are actually going to attempt to get things to work you'll have to code it up I assume.

As much as possible I'd get the thing spec'd and code written up front. You will get bogged down in getting things working and tracking down HW and SW bugs later but if you have things planned well up front you won't be losing valuable time for design when you are debgging.

It's all in the specification tbh, if you get that right and work out all your issues at a higher level before you get bogged down in the nitty gritty you are laughing.

Perhaps you could reduce the complexity straight off by agreeing location is simply within x of the rx'er showing the strongest signal strength. Getting this kind of thing specified up front could save you weeks and getting it in early is the way to go.

Boston

Copacetic is bang on. Also limit it to one user up front. so you don't have the hassle of having to id different Tags.

Tar.Aldarion

Ok, I'm going to buy a kit. I just need to find the best one.
I'll want several readers and several active uhf tags.
Time to find the best!
Code up front, eek, dunno where to begin even.

copacetic

Tar.Aldarion wrote:

Ok, I'm going to buy a kit. I just need to find the best one.
I'll want several readers and several active uhf tags.
Time to find the best!
Code up front, eek, dunno where to begin even.

well even a spec, some uml and a project plan would be a great start.
all that stuff is going to be gold to you in the real world and should
greatly impress the review board at the end of the project too if
you show a professional approach from the start.

Tar.Aldarion

Does anybody know if I can get a few active tags on their own?
The reader/active tags eval kit is far too expensive and I am looking up
feasibility/cost of developing your own reader by integrating ready made modules e.g. micro-processor, RFID read module, WLAN/Bluetooth module to form your own reader.
I don't know how much that might cost or how hard it would be.

barneysplash

Do you mean you want actual RFID tags?

I can get you 2 types they are both used in film cartridges.

There is a Texas Instruments one and another unmarked one that we use.

pm me with a postal address and I'll send 'em out.