RFID Questions
John R.
Posts: 1,376
I have not yet puchased a STAMP set, but am leaning in that direction.
My specific project is to implement RFID on a model railroad.· The concept would be for each car and locomotive to have a small ID tag embedded somewhere (low to the track), and have readers at various points (hidden just beneath the tracks), reporting what cars went past (and when).
There are a few questions I have.
First, what STAMP starter set and additional components will I need to get a "prototype" up and running.··I have a "programming" background, and basic electronic knowledge and a fair amount of "DIY" experience.· I know that RFID will not be the first STAMP program I write , but want to get all the components I need from the start.
I would want to test with at least two RFID readers.
With respect to the RFID readers, I beleive I will have some challenges (opportunities for learning) with the following issues:
Reader proximity: I will have readers too close to each other, and will need to work out timeing, etc. so that only one is active in a give zone.· I believe I can work around this, in worse case, by cycling the readers (ultamately > 10) so only one is on at a time, or at least cycle enough of them off so that they don't interfere with each other.
ID Tag spacing:·It is my understanding that only one tag can be in the "read field" at a time.· On small scale rairoad cars (or short cars on narrow gauge railroads) the tags will very possibly be within a few inches (potentially < 3.00 inches [noparse][[/noparse]75mm]).
In order to minimize issues with both of the above, and especially the ID Tag spacing, is it possible to build some type of shielding to shape and/or minimze the RF Field of the reader?
Any thoughts or ideas?
·
My specific project is to implement RFID on a model railroad.· The concept would be for each car and locomotive to have a small ID tag embedded somewhere (low to the track), and have readers at various points (hidden just beneath the tracks), reporting what cars went past (and when).
There are a few questions I have.
First, what STAMP starter set and additional components will I need to get a "prototype" up and running.··I have a "programming" background, and basic electronic knowledge and a fair amount of "DIY" experience.· I know that RFID will not be the first STAMP program I write , but want to get all the components I need from the start.
I would want to test with at least two RFID readers.
With respect to the RFID readers, I beleive I will have some challenges (opportunities for learning) with the following issues:
Reader proximity: I will have readers too close to each other, and will need to work out timeing, etc. so that only one is active in a give zone.· I believe I can work around this, in worse case, by cycling the readers (ultamately > 10) so only one is on at a time, or at least cycle enough of them off so that they don't interfere with each other.
ID Tag spacing:·It is my understanding that only one tag can be in the "read field" at a time.· On small scale rairoad cars (or short cars on narrow gauge railroads) the tags will very possibly be within a few inches (potentially < 3.00 inches [noparse][[/noparse]75mm]).
In order to minimize issues with both of the above, and especially the ID Tag spacing, is it possible to build some type of shielding to shape and/or minimze the RF Field of the reader?
Any thoughts or ideas?
·
Comments
It would take care of the problem of multiple reads,as an optical approach would read only the car traveling in the beam.
Bob N9LVU
Photo interrupt or other mechanisms can (and will be used for other reasons) detect the presence of a car, but not (easily) identify specifically which car went by. Trying to use punch cards, etc. on the way by would create unsightly appendages on the otherwise finely crafted models
Bar code is a possibility. The downside is the visibility of the bar code label on the bottom of the car. (There are some very small ID tags that can be "embedded" and hidden in the structure/body of the cars.) I am looking at a "capsule" tag as one possibility, 1/2" long X 1/8" Dia. The thin flat tags can also be hidden in the floors of most cars.
I also havn't seen (but haven't looked too hard yet) a bar code reader as cheap as the RFID reader.
I'm open to other options, and bar code actually would be nice in that you could more easily control what "number" or information for a car was scanned (as opposed to a database mapping RFID tags to cars) by printing your own labeles.
RFID is attractive for its non-contact and "non-visible" methods.
Thank you for the thoughts, and if you know of any low cost bar code readers, I'd love to know about them.
John R.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jon Williams
Applications Engineer, Parallax
How long does the tag need to be "over" the reader? We're talking "model railroad", as opposed to "toy train", and are not trying to see how fast things can move. Other then when trying clean the track, I don't think I've hit "full throttle" in about 10 years.
Interesting complication that was not brought up at "another site" with a similar reader.
In some simple testing, I was able to read bar codes on moving cars with a "mid range" stationary laser reader (borrowed from work). Does this equate to the ability to read the RFID tags?
If distance affects scan acquisition speed, I can probably get the reader to within 0.500 inches, but I'd prefer about 1.00 inches (vertical).
In either case, I see some type of STAMP kit in my near future. As I've been poking around, I have become intrigued...
John R.
But, I'm just speculating -- theo only way to know for sure is to order a reader and a couple tags and give it a whirl.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jon Williams
Applications Engineer, Parallax
Back to one of the original questions I had:
What do I need to get:
I know I'll need the following:
One of the starter kits.
A reader
Some tags
Being new (brand new) to STAMPs, I don't know if I'll need anything else (other than wire, etc.) to interconnect the reader and the STAMP.
I guess it's time to start downloading the gazillion bytes of docs you guys have online...
Now I just have to convince my wife that this isn't ANOTHER hobby, but part of my model railroad.
John R.
·· Our most popular starter kit is the Discovery Kit.· It comes with our, "What's A Microcontroller?" kit, which helps you get started on learning the BASIC Stamp 2 programming and interfacing.· I have posted a link below for you.
http://www.parallax.com/detail.asp?product_id=27207-700-00065-28162
This is also available in a USB version.
http://www.parallax.com/detail.asp?product_id=27807-700-00065-28162
And of course the RFID Reader:
http://www.parallax.com/detail.asp?product_id=28140
And tags...
http://www.parallax.com/detail.asp?product_id=28142
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
Thanks for the help. I'll doing some reading, and probably jumping in with both feet next week sometime.
This fourm has been quite a different experience than the other one I was dealing with. Great job to all.
Could you use an Infra-red line follower module to read a custom code on the underside of the car? the code I was thinking of
would be a binary number of up to 8 bits with a narrow dark line for each 0 bit, and a wide line for a 1 bit, or even better, 2 line follower modules
mounted in a \ __ / fashion to read a double bar code that is monted on underside like \_/, the outside bits on each side would be BOTH dark, with the next
bit on one side equal to a 1, with corresponding bit on other side not marked, that way it doesn't matter what the direction the car is traveling in, with an mirror image that can be read in either direction, like so;
"SB---1---1---1---1---1
1--EB-EB-1
1---1---1---1---1---SB"
"==-==-==-==-==-==
==-==-==-==
==-==-==-==-==-=="
"==
==-==
==-==----==-==
=="
"SB
0---0
EB-EB----0---0
SB"
The above shows a binary number of 11111001 from either direction, which is 249 decimal.
You can use the 1st n bits to designate type of car, with the 8-n bits left for the amount of cars for each type...
sorta like setting up a network with subnets... and have a display show ( boxcar) # (25)
which could be 111111001 in binary , the 1st three bits are your ( car type) 111 xxxxx binary, or 224 decimal, the last 5 are the car (number) xxx11001 binary, or 25 decimal, with three bits for type, you can have 6 different types, with a max of 31 cars per type. I would see this a sure fire way to id cars, with no cross interfearance that you would get with rf, and using optical sensing this way you are not held to a speed constrain for the train.
If both side are marked its the start or end of data, any data after both start bits has to have 1 or 2 pulses per bit on one side or other, not both to equate to the binary logic.
As for a low cost bar-code reader, do a search for CueCat on Google, lots of info on hacking it.
Bob N9LVU
Interesting thoughts. I'd need to expand the bit count (this is targeted at larger layouts that may have over 100 of a given car), but this could work. When you say "Infra-red line follower module", is this a "pre made" module, or one I'd have to develop?
I could also just do like the real bar codes, and have a "start" and "stop" end. If I saw start first, I'd use the order read, if I saw stop first, I know I'd have to reverse the order.
I'll be ordering some stuff next week, and start some testing on both the RF and other options. I'll look up the CueCat, but as I remember it, it needed to be virtually in contact with the bar code to read it.
On the RF, I've got mixed feedback on how well this will work in terms of a moving train. The folks at Phidigits were talking about cycling the readers on a layout so each would only be on 1/20th of a second. (20 readers on a layout, and cycle through activating them once a second. I don't know if they really thought that would work, or were just "talking smart". If these things can power up and read that fast, then the moving train part shouldn't be an issue. If it can work, this also makes changes in speed (or stationary) less of a factor.
Most significantly, I've made a transition from a purely PC based solution to at least looking at a uProcessor approach. This has some meaningful implications on a number of other areas in layout automation.
Keep the ideas coming.
John R.
However, I would not recommend that you use the RFID reader.· The reader has many useful applications - I am presently using it for inventory control.· The smallest tags that come with the RFID reader are 2" diameter, which is larger than an HO car.· If you are using G scale it might work.· In addition, I can see all kinds of mechanical problems in trying to bury the reader underneath the track, and it addition, the reader needs to be "enabled" when you wish to scan a tag.· You'll be much better off using bar code or some derivative of it.
For what it's worth.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Sid Weaver
Do you have a Stamp Tester yet?
http://hometown.aol.com/newzed/index.html
·
A line follower module like;
www.parallax.com/detail.asp?product_id=555-27401
would work, and thier are several others, all you need is an IR pair (tran/rec) that are focused to read a flat surface... also to hide any bar coding, use a cover
of dark red plastic, to mask the white/black of the barcode... look at some remote controls, or recievers that use IR, alot use the same plastic to cover inside electronics..
With the cue cat, if you make the lines thicker ( larger barcode ect) you should be able to increase the distance that it would read, you could also roll your own code...
Bob N9LVU
Size of any ID marking/technology will be a significant factor. An "N" scale car is only about 3" (some less) long, and the track is a little under 3/8" between the rails. Aestetics will also be a significant factor, less for me, more for other potential users (the infamous "rivet counters").
There generally is some type of a "center sill" running down the center of the car on the bottom that I could probably hide something in/on.
John R.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
I've seen the smaller tags, including a "capsule" type (mentioned above). Your team here raised concerns about the "read time" required for a tag to be over the reader on a moving train.
My plan is to order some stuff to test with next week (need to negotiate with the budget committee) and see what happens with the RFID.
If the read is fast enough, then I need to find a way to deal with tags close together. I'm hoping to find a way to shield the reader and/or replace the antenna (naughty, I know...) with a less sensative one. (I've got to look into this more on the Grand Idea site.)
But one step at a time. Let's get into STAMPS, and test the RFID...
Is there a way to get the 2p40 Professional Starter set with all the other stuff found in the "discovery set" (parts for learning excercises)? (or swap in the 2p40 board and STAMP module to the discovery set)
John R.
·· That would entail buying the "What's A Microcontroller?" parts & text kit separately from the BS2p40 Starter Kit.· Mind you, the examples in the WAM Kit are all based on a BS2, and therefore many will not run on the BS2p40 without making changes to the code.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
I've got stuff on the wish list, we'll have to see what OMB (Office of Management and Budget, aka Lana, aka Wife) has to say...
Thanks for the help and direction.
John R.
(Do you guys ever take time off?)
·· What's time off?· I like your attitude about just buying one of each and having the whole kitten-kaboodle!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
That is kit-and-kaboodle.
Respectfuly, sir.
Sid
·· Thanks!· But you still didn't explain what a kaboodle is?!?·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
Sid
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Jon Williams
Applications Engineer, Parallax
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support
csavage@parallax.com
Yes, I have all the information I need for now. MBO is working nights this weekend, so we probably wont get appropriation until sometime next week. It just makes life so much easier to ask first.
I also enjoyed hearing the "down under" lineage of caboodle. Being from Wisconsin, the home of cheese and bratwurst, I always assumed it had a German heritage.
Thanks to all for your help, I'm sure you haven't heard the last from me (for better or worse).
John R.