How can I give a unique id for each transmitter ?

Thread Starter

Dohab

Joined Oct 17, 2009
9
It's been 6 months now since I started my project.

It's my first one, but since this is not my field of creativity so I could use your help in that.

Let's say that we have 50 transmitters and one receiver, all transmitters send on the same frequency about 30 signals/second.

1. how can each transmitter have and send a unique id of his own to the receiver?
2. how can the poor receiver know which transmitter is this ?

I attached an illustrator of the general idea.

If you can help, please treat me as a beginner in electronics, because I'll try to understand whatever you explaining.

so please attach any circuits diagrams + electronics parts + real explanation may I need.

I swear the god, I can't understand any guidelines :D

yet, anything you can do will be Appreciated.
 

Attachments

SgtWookie

Joined Jul 17, 2007
22,230
You'll have a big mess with 50 transmitters blasting away on the same frequency all at once. You'd be lucky to even get 1 signal received not garbled.

You need one transmitter to be a master, preferably on a different frequency, that sends out a request for data from a particular receiver/transmitter pair.
All of the other transmitters should be connected to individual receivers that listen to the master's frequency. They should only transmit when they have received a request from the master.

The master has a receiver tuned to the other 50 transmitters' frequency. Each transmitter should start with a block of data that identifies itself, and then the data that is to be sent. That will help if a transmitter gets "stuck" on.

The master receives the data, and once processed, sends a request for the next transmitters' data to be sent.

This is not really a project for a beginner.
 

flat5

Joined Nov 13, 2008
403
First thing they send is a unique digital code that IDs the sender. or analog dual tone encode/decode.

The 30/sec has me a bit confused.

Perhaps more info would be helpful.
What are you doing?
 

Thread Starter

Dohab

Joined Oct 17, 2009
9
You'll have a big mess with 50 transmitters blasting away on the same frequency all at once. You'd be lucky to even get 1 signal received not garbled.

You need one transmitter to be a master, preferably on a different frequency, that sends out a request for data from a particular receiver/transmitter pair.
All of the other transmitters should be connected to individual receivers that listen to the master's frequency. They should only transmit when they have received a request from the master.

The master has a receiver tuned to the other 50 transmitters' frequency. Each transmitter should start with a block of data that identifies itself, and then the data that is to be sent. That will help if a transmitter gets "stuck" on.

The master receives the data, and once processed, sends a request for the next transmitters' data to be sent.

This is not really a project for a beginner.
doing this, will kill the idea of getting the data at the same time which is the only purpose of this project, beside sending a 30 signal every second.

Also the receiver would be as big as a workstation mate :D , and I have plans to add more transmitters in the future, you can see how stuck I am.

the following link may match my thoughts, have a look you may find some other ways:
http://www.cdt21.com/parts/files/press/PR00204.pdf

I know it's not a beginner good start, but I can't sleep really, it's killing me.
that's the reason I asked your help, because you are not a beginners.
once I finished this project, I'll be professional.

you see, all technicians in my area are proud of building a flashing LEDs all the time , but they don't know what I am talking about :D. they're like : can't help you, it's illegal, Is this a huge radio station?
and i'm like : yes, yes, whatever.

SgtWookie, do you think of any another way to send signal at the same time? no waiting time between transmitters signals? without any action from the receiver except for getting signals and separate them?

First thing they send is a unique digital code that IDs the sender. or analog dual tone encode/decode.

The 30/sec has me a bit confused.

Perhaps more info would be helpful.
What are you doing?
exactly what I am asking my self all the time.
should I send a unique code or use encode/decode stuff?
And you know what? I don't know.
because I don't know how a couple of transistors and a capacitor soldiered to a 3cm of wire, can create a unique id, Rather than send it !! :D

what do suggest flat5? which way is the accurate and the best for sending about 1500 siganl from 50 transmitters at the same time to one receiver ?

If I tell you what I am building, you will be like: GET OUT OF HERE :D

but it's a lot of fun, oh yeah.

so, right now, let's just stick with: we need about 50 transmitters to send 30 signal each , every second to one receiver, you can use any technology you suggest.

once we finish, you will be like: Did we just made this for real ? we are geniuses :D I promise.

back to reality, can we make such thing becoming a real thing?
 

SgtWookie

Joined Jul 17, 2007
22,230
doing this, will kill the idea of getting the data at the same time which is the only purpose of this project, beside sending a 30 signal every second.
OK, so you'll just have to explain what you're trying to do then.

Also the receiver would be as big as a workstation mate, and I have plans to add more transmitters in the future, you can see how stuck I am.
It doesn't matter how "big" your receiver is. You will only be able to receive one transmission at a time.

If you feel that you need to receive multiple transmissions simultaneously, you will need to use transmitter/receiver pairs that are on different frequencies.
 

Thread Starter

Dohab

Joined Oct 17, 2009
9
Okey, as soon as I tell you what I am about to build, you will never help me like others.
but you know what ? I'm gonna tell you, coz I believe that you will help me.

Oh god,
I am building an RF ( wireless ) motion capture. TADDA!

that's why I need to receive multiple signals transmissions simultaneously in one frequency, which brought on surface the idea of using unique ids, after searching for a solution of course :D.

Is there any way?
 

SgtWookie

Joined Jul 17, 2007
22,230
Video will take up a heck of a lot of bandwidth.

One way to minimize the bandwidth is to only send the pixels that changed in between transmissions. That's basically how MPEG compression works.

If most of the time there is no movement to report on the individual cameras, minimal bandwidth usage will be needed. However, if there are multiple cameras with movement, you'll run into bandwidth problems.

Each camera will need a receiver, a pretty fast uC to process the frame data with lots of memory aboard, and the ability to recognize when the master transmitter requests the latest update.

A better way to do this would be via fiber optics; the bandwidth of a single strand of fiber optic is extraordinary. You don't have to worry about RF transmission interference, either.
 

Thread Starter

Dohab

Joined Oct 17, 2009
9
now we're talking,
but you have to excuse me, I am a little bit confused now,
Are you talking about capturing human body motion from a camera?
if so, then no sir, that is not what's in my mind. coz I captured my motion already a year ago using markers on a suite.
I'm not thinking of using cameras or any optical processing at all this time.

The general idea is to make three receivers, then calculate signal traveling time to each receiver and then positioning it in a virtual 3d world ( this will done using arduino and c++), it's more accurate than opticals, don't you think ?

If using fiver optics will be the best way considering not using any camera or an existing video, then, hell yeah we're gonna use your way.

can you please illustrate it in general?

or if it's easy to build you will be thankful if you attach any circuits we may need.
 

SgtWookie

Joined Jul 17, 2007
22,230
now we're talking,
but you have to excuse me, I am a little bit confused now,
Are you talking about capturing human body motion from a camera?
if so, then no sir, that is not what's in my mind. coz I captured my motion already a year ago using markers on a suite.
Do you mean, markers on a suit?

I'm not thinking of using cameras or any optical processing at all this time.
Then why are you interested in using transmitters?

The general idea is to make three receivers, then calculate signal traveling time to each receiver and then positioning it in a virtual 3d world ( this will done using arduino and c++), it's more accurate than opticals, don't you think ?
It will be rather difficult at short range.
RF travels one statue mile (5,280 feet) in roughly 5.376uS, or nearly 1 foot per nanosecond. I don't know if you realize how brief of an amount of time that is.

If using fiber optics will be the best way considering not using any camera or an existing video, then, hell yeah we're gonna use your way.

can you please illustrate it in general?
If your idea is to measure distance, then fiber optics really won't be much of a help.

I guess I'm still not sure what you're trying to do.
 

Thread Starter

Dohab

Joined Oct 17, 2009
9
yes sorry, I meant suit.

And the reason that makes me interest in transmitters is because I can use either ultrasonic, AM or FM signals, which can be captured even in dark environment.

You see optical tracking could be inaccurate enough for tracking processes.
some markers could be disappear when the actor move faster, beside, I may need at least 3 HD cameras to track'em. I didn't mention fingers movement yet.

RF transmitters can send there waves through human body, so even if the actor is moving fast or turn around him self or dancing, the signal will be sent, and the receiver will find it no matter what situation is, and we will get the tracking data at the same time without further processing.

If you check the link I posted, you will see that they are using one frequency which is 434MHz FM narrow band.
And maybe they are using what flat5 suggest, analog dual tone encode/decode.
 

kubeek

Joined Sep 20, 2005
5,794
So you are basically trying to make a little version of GPS, and send the locations wireless. As SgtWookie said, it seems you don´t realize the dificullty of such task.
How good resolution does the optical tracking with HD cameras make? My guess is about one inch, maybe even better.
The RF signal travels at 1 feet/ns, that is if you wanted to get 1/10 foot resolution, you would have to be able to distinguish at least 100ps intervals, that is comparable to sampling at 10GHz frequency. I really can´t imagine anyone breadboarding such thing.

I suggest you try to orinet on ultrasonic data transfer, because the sound travels at cca 1 foot per milisecond, which is 6 ordes slower than RF. That gives your processor enough time to calculate the delay, and also digital processing of the receiced signal will be much easier.

This way you can spread the spectrum of the transmitters across the receivers range, and then use FFT or similar techniques to decode the data and recognize each transmitter.

Each transimtter will also need to have an receiver (preferably RF), so that your master device sends a fire pulse to all the slaves, which all immidately send their code via ultrasonic signal. You sample the ultrasonic signal, and compute the time it took for each slave´s signal to reach your master. That way you can get a triangulation of the position of each point you want to observe.

When implementing this, you have to be aware that reflection of the signal will occur, so you have to use only the signal that comes first, and even then you might get into some strange functionality..
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
There was a female Navy Captain who was famous for her work in early computer programming and student lecturing style; her name escapes me at the moment.
[eta]
Rear Admiral Grace Hopper: http://en.wikipedia.org/wiki/Grace_Hopper
She was on a "60 Minutes" TV episode I watched in the early 1980's; at that time she was a Captain.

She would walk around the classroom and hand out nanoseconds; pieces of wire that were cut to about 1 foot in length.

Then she would show the students a microsecond; she only had one example. It was the same type of wire, but it was a roll that was 1,000 feet in length.

She would have liked to have shown the students a millisecond, but they didn't make spools of wire that were 1,000,000 feet in length, and it would take a big truck to deliver it.

Let's say that you have a microprocessor that's running with a clock frequency of 10MHz. During every clock cycle, RF would travel 100 feet.

I worked on a prototype ground imaging radar back in the early 1980's. One of the processor boards used (in those days) very high speed Motorola ECL (Emitter Coupled Logic) that operated at 117MHz. This gave the system a resolution of 10 meters, which would be woefully inadequate for your purpose.

Breadboarding is not practical once you get into the 10s of MHz range. When you're in the GHz range, even a very short piece of wire will act as an inductor, and a component anywhere near another component will have capacitive coupling. Even to an experienced person, RF in the GHz range seems to have near-magical properties.
 
Last edited:

MikeML

Joined Oct 2, 2009
5,444
My PhD dissertation adviser, Ivan Sutherland, build a 6DOF head tracking system using ultrasonics at MIT in 1963. Look on Wiki.
 

bull1894

Joined Oct 18, 2009
4
it may be that you need to reed if signals are present. if so try tone identity. it is music to your receiver. use prime numbers and mix and match them to get unique id's
 

bull1894

Joined Oct 18, 2009
4
I'm sorry. I thought you needed somthing else. if you need head tracking, it is already available. body tracking is currently done optic and I don't see it done better in the future. body maping can"t be done fast enough for real time with current cpu speeds.
 

Thread Starter

Dohab

Joined Oct 17, 2009
9
Do you mean, markers on a suit?

It will be rather difficult at short range.
RF travels one statue mile (5,280 feet) in roughly 5.376uS, or nearly 1 foot per nanosecond. I don't know if you realize how brief of an amount of time that is.
Sorry, I missed that part.
It's so important, how did I not see it?
I thought RF is a kind of sound waves, someway.


So you are basically trying to make a little version of GPS, and send the locations wireless. As SgtWookie said, it seems you don´t realize the dificullty of such task.
Exactly, its a pretty much a tiny GPS.

using optical tracking with HD camera is fun, "I confess ", but it's so annoying thing when a marker disappear, sometimes we had to re-film the scene. I've tested that experience before, When I came up with this wireless tracking idea :D

Using of ultrasonic may be the solution, in fact it was my first way I thought of using.
then, after searching for a circuit diagram, I couldn't found one that match multiple transmitters, but I found one and it consists of a HUGE number of parts on one big breadboard.
and now you are bringing back the hope of using Ultrasonic.
but this:
Each transimtter will also need to have an receiver
is what I'm trying to avoid.

Okey, I will use Ultrasonic, can you please give me circuits diagram and electronics part that I may need to assemble that in my home, and test the result?

I really,really need your help in this.

There was a female Navy Captain who was famous for her work in early computer programming and student lecturing style; her name escapes me at the moment.
[eta]
Rear Admiral Grace Hopper: http://en.wikipedia.org/wiki/Grace_Hopper
She was on a "60 Minutes" TV episode I watched in the early 1980's; at that time she was a Captain.

She would walk around the classroom and hand out nanoseconds; pieces of wire that were cut to about 1 foot in length.

Then she would show the students a microsecond; she only had one example. It was the same type of wire, but it was a roll that was 1,000 feet in length.

She would have liked to have shown the students a millisecond, but they didn't make spools of wire that were 1,000,000 feet in length, and it would take a big truck to deliver it.

Let's say that you have a microprocessor that's running with a clock frequency of 10MHz. During every clock cycle, RF would travel 100 feet.

I worked on a prototype ground imaging radar back in the early 1980's. One of the processor boards used (in those days) very high speed Motorola ECL (Emitter Coupled Logic) that operated at 117MHz. This gave the system a resolution of 10 meters, which would be woefully inadequate for your purpose.

Breadboarding is not practical once you get into the 10s of MHz range. When you're in the GHz range, even a very short piece of wire will act as an inductor, and a component anywhere near another component will have capacitive coupling. Even to an experienced person, RF in the GHz range seems to have near-magical properties.
Now I can see what your telling me.
I like the story.
I'm losing my focus sometimes, a lot of times actually.
Obviously, I'm gonna use ultrasonic now but I have no idea how .:confused:

My PhD dissertation adviser, Ivan Sutherland, build a 6DOF head tracking system using ultrasonics at MIT in 1963. Look on Wiki.
I'm not focusing on head tracking in particular, it may be a part of the project, but not the main of it. I think there is a huge difference.

I'm sorry. I thought you needed somthing else. if you need head tracking, it is already available. body tracking is currently done optic and I don't see it done better in the future. body maping can"t be done fast enough for real time with current cpu speeds.
I am trying to avoid any optic tracking.
I will use Optical tracking only to capture facial emotions.
The reason I want to build it my self ( with help of course ) is because it will be away cheaper than a commercial one.

=================
Addition:
You guys are giving me a hope and god knows how much I need your help, really.
So, thank you very much.
 
Last edited:

Thread Starter

Dohab

Joined Oct 17, 2009
9
Not always true and more often than not false.
Rright, if you are building a TV or flash memory. :D

But as you know, motion capture systems is a special equipment, no one gonna buy it with no good reason.

You can't find it even at eBay or Amazon too.

The cheapest one I found, costs 6k$. which I don't have, I can't have too.

and it comes with a strange suit which makes the performance guy looks freak even to the Joker, and it uses optical tracking :D
 

Thread Starter

Dohab

Joined Oct 17, 2009
9
After searching, I found the Ultrasonic is limited in distance, means I can't track transmitters farther than 5 Meters !
I can't use it, honestly..
 

kubeek

Joined Sep 20, 2005
5,794
I don´t know where you read that, but I don´t really think that´s true. The practical distance depends on the sensitivity of the receiver and reflections in the room.

You definitely need some kind of receciver in the transmitters, because you need to get some kind of synchronisation between them, to make them all send at the same time.
You could use infrared light instead of RF for the trigger, but that will again introduce risk of blocking the sensor and not firing when it should.


If you want to give this a try, make two ultrasound transmitters, each generating some kind of tone, and try to receive it with one receiver. To keep it simple, connect the transmitter by wires, place them at a typical distance from the receiver, and start trying to distinguish which one came first, and the exact times of arrival for each one.
 
Top