Shop OBEX P1 Docs P2 Docs Learn Events
Directional XBEE signals — Parallax Forums

Directional XBEE signals

garyggaryg Posts: 420
edited 2013-02-17 21:52 in Accessories
I've been looking at a real world position control of my scarecrow platform.
It appears that GPS control can't be done to within a few feet of real world coordinates.
It also would also appear that Infrared communications with lots of sunlight is not the ansower.
As I'm building a new differential wheel control platform, I still need to find a resolution to my real world location concerns.

The most recent thought concerning resolving my delima has to do with a very directional radio
signal.

Has anyone done any experiments involving setting up a very directional antenna, or restricting the
outgoing signal so that it will be received in only a very narrow band.

My thought is this:
Set up an XBee transmitter in my yard.
Set up another XBee transmitter/receiver on my platform mounted on a servo.
After running my platform for about 30 seconds, stop the platform and rotate the servo until the transmitter
is seen by my platform.

Possibly I could steer the platform towards its home position.

Sorry if i'm being a bit vague here, but I'm looking for the meaning of life.

Thanks
gg

Comments

  • W9GFOW9GFO Posts: 4,010
    edited 2013-02-14 02:19
    I'll start by saying that I don't know. You can poll the XBees for signal strength so you should be able to do better than testing for the presence or absence of a signal. In radio direction finding contests (fox hunts) often times the null is the best indicator. I would guess that a purpose built antenna can be fitted to an Xbee for exploiting the null zone.

    Precise outdoor navigation is very challenging. I'll be interested to see how this discussion goes.
  • JBWolfJBWolf Posts: 405
    edited 2013-02-16 02:05
    Many RC Airplane enthusiasts use a directionalized "yagi' type antenna on a pan/tilt mount to track the airplane for extreme ranges (miles).
    I have been wanting to get into this someday... maybe it can give you some ideas:
    http://www.fpvpilot.com/Pages/AntennaTrackingStations.aspx
    http://youtu.be/_lxUd96-1tw

    I believe the rest f the tracking equipment to make it follow is on their site
  • garyggaryg Posts: 420
    edited 2013-02-16 05:04
    Thanks for your comments
    I looked at information on fpvpilot.com, Yagi antenna info and did some searching for Fox Hunts.
    I've also done some searching in some of the older forum threads.
    The size of transmitter,receiver,antenna that I found were all pretty large in size, however
    I did see a comment in an older post that said that the XBEE chip antennas are somewhat omni-directional.
    It looks like I'll need to try some experiments.
  • JBWolfJBWolf Posts: 405
    edited 2013-02-16 18:59
    yes all the Xbee's i have seen have an omni antenna... but i believe there is a model with a pigtail for plugging in your own type of antenna.
    Excerpt from FPVpilot:

    Q. What does antenna tracking do?

    A. Antenna tracking can be used with some GPS/OSD systems like Eagle Tree Systems. Antenna trackers use the GPS data from the aircraft's down link and keeps the antenna (patch or high gain antenna) pointing at the airplane as it flies around giving you more range and better picture with less static. It uses tracking hardware to pan and tilt servos that the video RX antenna is attach to. The aircraft's telemetry is sent through the video down link where the antenna tracker decodes the info which in turn operates the tracking.Some systems use the audio channel which is not recommended by most FPV pilots since audio can cut out sooner then video making the tracker stop during flight.


    Thats a pretty good idea... if you have 2 gps modules, you could always make the ground station point at the received gps coord's
  • xanaduxanadu Posts: 3,347
    edited 2013-02-17 12:55
    I've played around with using XBees as beacons and also for navigation. I never got the same results twice. Maybe in perfect conditions, ie Faraday cage you could get a nice baseline to work from. I'm not all that great at programming in SPIN however I don't think it takes a good programmer to see how the environmental conditions effect the signals so I offer up that advice. The inconsistencies aren't huge, but they're there.

    The more directional you make the XBee signal the better signal you get and the less resolution you'll have for measuring it, so I'd start with the lowest power XBee if you're using RSSI as the variable.

    I was able to do something pretty simple here - http://forums.parallax.com/showthread.php/139684-XBee-Locator-Beacon
  • garyggaryg Posts: 420
    edited 2013-02-17 13:31
    Xanadu
    I've downloaded the XBEE Tutorial V1.0.1 .
    It seems like I'll need to purchase several XBEE components to make what you have done work for me in some manner.
    Just in checking the Parallax Store, It seems like to start with, I'll need something like the XBEE wireless Pack and
    possibly the separately purchased antennas.
    I'll need to do quite a bit of reading on this subject to get familiar with it.
  • xanaduxanadu Posts: 3,347
    edited 2013-02-17 13:55
    garyg wrote: »
    Xanadu
    I've downloaded the XBEE Tutorial V1.0.1 .
    It seems like I'll need to purchase several XBEE components to make what you have done work for me in some manner.
    Just in checking the Parallax Store, It seems like to start with, I'll need something like the XBEE wireless Pack and
    possibly the separately purchased antennas.
    I'll need to do quite a bit of reading on this subject to get familiar with it.

    I'm wondering if maybe you could use RFID and magnetic compass? How big is the area you need to canvas?
  • garyggaryg Posts: 420
    edited 2013-02-17 14:25
    I'm attempting to keep my platform within approx. 25ft x 50ft area.
    I've done some small experiments with the HMC5883L compass that Parallax sells.
    My 1st platform makes lots of electrical noise, which confuses the compass.
    My next platform, which I'm currently working on should be less noisy.
  • xanaduxanadu Posts: 3,347
    edited 2013-02-17 15:09
    I've seen some big buildings using RFID and line following, I was thinking maybe you could use RFID and a compass but didn't consider the compass error. 125 RFID tags at even $1.00 each would be a similar cost alternative and a lot less code. You'd just need to get the robot to navigate on it's own between them. A 10' gap doesn't seem that bad but I don't know what the surface is like. I always imagine everything on a smooth flat plane when reading about it hehe.

    The other method of triangulation I've been wanting to try is using a laser range finder and some basic odometry.

    There are robot lawn mowers that use ultrasonic signal measuring beacons and it works really well supposedly (5cm). No idea if this is feasible for DiY but worth a quick Google :)

    Another thing you can use is machine vision. RoboRealm is definitely worth every penny in its ease of use (think it's $50). You can teach it to recognize something, and based on where it is in the field of view move the robot. You don't need to put a computer or the camera on your robot either. You can use an IP cam (Dlink DCS-920 is perfect) to send the signal back to a house powered ground station, and two XBees to relay the commands back to the robot. With the help of some SPIN code from forums members here I had a tracking system up and running in minutes and it worked very well.

    Sorry to keep throwing out different options. There is really a lot to think about with accurate navigation outside in the elements and some good systems use a combination of different methods to get good results. Reading about how the Mars rover navigates is a really good primer on that aspect. I'd explore lots of options before choosing one taking into consideration the ground conditions.
  • garyggaryg Posts: 420
    edited 2013-02-17 21:52
    All of the robotic lawn mowers I've looked at (which has always been of interest to me) use a buried wire technique
    as a boundary.
    This similar to the buried wire dog fence techniques.
    While I like the concept of the buried wire boundary, I have not been able to figure out how to impliment such a boundary.
    It would appear to me that the buried wired boundary thing gets to be a very complex thing that may be outside of my current skills and knowledge.

    The RFID thing could hold some promise, but will take some time to figure out if I can impliment a solution using this.

    I'll be looking at information concerning the RoboRealm. I have not heard of this, so it will be worth some investigation.

    I fully realize the statement from W9GFO "Precise outdoor navigation is very challenging"
    Which I believe is why there is not a lot of information available about a very simple solution to a very complex concern.

    It will take me several weeks to complete my new basic platform, and at that time, I'll be able to try my HMC5883L compass again.

    I have yet to play with any type of basic odometry. Since I'm working on getting this thing to correctly track on a lumpy lawn surface,
    I believe I will have some inconsistencies on my still in construction platform.
    I'm also planning on using a Propeller controller, which may make this goal more achievable.


    I know there is a relatively simple solution to an invisible barrier and I'm determined to find one.

    Thanks for all of your thoughts.
Sign In or Register to comment.