Connect with us

How to create 3D omnidirectional antenna set?

Discussion in 'Radio and Wireless' started by Farukh Khan, Sep 27, 2018.

  1. Farukh Khan

    Farukh Khan

    123
    2
    Jun 12, 2015
    Hello guys,

    I have been working on a AI based quadcopter recently. And I am using the 2.4GHz WiFi frequency for communication purposes. The 2.4GHz Wifi is built into an onboard SBC (something like raspberry pi zero w). Now I want to make an antenna set which will cover both TX and RX for the wifi on a 3D vector plane. So, basically I want this drone to catch signal from all the possible directions on a 3D plane. And I know that there is a tradeoff for distance with omni direction but what is the best possible way to achieve both transmission and reception on all possible 3D directions with optimal range?

    What antennas or antenna configuration is best for reaching all possible directions?

    On my ground station side I can cover up for any range loss by pointing multiple directional antennas where necessary. But on the drone I need an optimized version of omnidirection which will reach almost all possible directions, because the drone will be continuously moving in a 3D plane.

    BTW, I have only one IPEX connector on the SBC, so I have to squeeze every possible antenna configuration to work through this one IPEX.


    Any help would be appreciated guys. Thanks.
     
  2. Hopup

    Hopup

    210
    25
    Jul 5, 2015
    It is common to use omnidirectional and directional antenna in the receiver same time as diversity to benefit from both antenna types. In your quad you can have omnidirectional antenna.

    Since you are talking about control link, then I would advice using directional antenna instead of omnidirectional on both sides. If you want more distance you need UHF type system, might be illegal.
     
  3. Farukh Khan

    Farukh Khan

    123
    2
    Jun 12, 2015
    but if I use directional antenna on the quad itself, then won't it be hard to control it's rotation and backwards flying?
     
  4. Hopup

    Hopup

    210
    25
    Jul 5, 2015
    How far distances are we talking?
     
  5. Farukh Khan

    Farukh Khan

    123
    2
    Jun 12, 2015
    5km-10km. but I am ok to put loads of antennas on the ground station, but cannot put that many on the quad though.
     
  6. kellys_eye

    kellys_eye

    4,289
    1,143
    Jun 25, 2010
    Vertical dipole.
     
    hevans1944 likes this.
  7. Hopup

    Hopup

    210
    25
    Jul 5, 2015
    I would not expect to be able have reliable connection that far if you are doing some other than stable movement. You must have multiple receiver in your quad if you want to do that.
     
  8. hevans1944

    hevans1944 Hop - AC8NS

    3,908
    1,867
    Jun 21, 2012
    It is a provable fact that any bare antenna (no in-line receiver amplifiers allowed!) works equally well for transmission as well as reception. When antenna transmit/receive directional patterns are measured or calculated (from an antenna modeling program such as EZNEC) the directional gain is referred to an imaginary isotropic radiator, a point-source antenna, located in the center of a sphere, that radiates and receives equally well in any direction. Such antennas do not exist as anything physically realizable, but that would be what you need for
    Next best thing would be a vertical dipole, as @kellys_eye suggested in his post #6. These are omni-directional in a plane perpendicular to the vertical axis of the antenna.
     
  9. Farukh Khan

    Farukh Khan

    123
    2
    Jun 12, 2015
    What if I use the verticle dipole configuration and change it to a 3 axis or 3d configuration using basically three antennas? Will that cause interference or signal elimination problem?

    The configuration I am talking about is as follows:

    Three omni-directional antennas attached one horizonal x axis, one vertical y axis, and one in the z axis and all of the three antennas sharing single (IPEX 2 pole) wifi chip.
     
  10. kellys_eye

    kellys_eye

    4,289
    1,143
    Jun 25, 2010
    Why three antennas? The radiation pattern of a dipole is (figuratively) a 'donut' shape with 360o coverage. Whilst there are two distinct null points (top and bottom) these would only be an issue for reception directly under/over the antenna and, even then, 'leakage' would probably make it a non-event. Certainly the 'top' null point won't ever be a problem!
     
  11. duke37

    duke37

    5,103
    687
    Jan 9, 2011
    A vertical dipole will give a vertical polarised omnidirectional output.

    A pair of horizontal dipoles mounted in a horizontal cross at 90 degrees and fed with signal at 90 degrees will give a horizontal polarised omnidirectional output.

    Some FM aerials are made as a loop, called a halo, to give omnidirectional horizontal reception. These have a bad reputation.
     
  12. Farukh Khan

    Farukh Khan

    123
    2
    Jun 12, 2015
    But what if the signal is transmitting through a totally horizontal radiation antenna? As the drone will be moving in all the directions possible, the vertical dipole won't remain vertical all the time. It will rotate to various directions according to the drone's position on the air.
     
  13. duke37

    duke37

    5,103
    687
    Jan 9, 2011
    If the polarisation is not exact, then there will be a sine wave error. I.E. not much difference initially but going to zero at 90 degrees. Do you do loops or rolls?
     
  14. Farukh Khan

    Farukh Khan

    123
    2
    Jun 12, 2015
    what you mean by loops or rolls?
     
  15. duke37

    duke37

    5,103
    687
    Jan 9, 2011
    A loop is where the aircraft is nosed up and goes over the top.
    A roll is where the aircraft flies in a straight line but tips upside down and continues until it is fying level again.

    I have not seen a drone and do not know what aerodynamics they are capable of. Helicopters and even a Boeing 707 have been rolled.

    If the base of the drone tips by less than 20 degrees, then I would not think that polarisation would be a problem as long as the two aerial polarisations are the same.
     
  16. Farukh Khan

    Farukh Khan

    123
    2
    Jun 12, 2015
    I think quadcopters do both loop and roll. For your better understanding let me link a good video which shows some possible aerodynamics of a quadcopter drone.

     
  17. Hopup

    Hopup

    210
    25
    Jul 5, 2015
    There is no way to do that kind of stuff if you are multiple km's away without relying satellite or purely autopilot, at least I wouldn't attempt doing any of that stuff LR.

    *anything but stable movement.
     
  18. Farukh Khan

    Farukh Khan

    123
    2
    Jun 12, 2015
    What you mean by relying on satellite? Is there any open source projects which allows to share speed and channel from satellites?
     
  19. Hopup

    Hopup

    210
    25
    Jul 5, 2015
    I don't think there is any satellite systems for low prices. You could use 3G network for this if the area you are located has one, it would be much more reliable than using 2.4 transmitters, as far as I know.
     
  20. Farukh Khan

    Farukh Khan

    123
    2
    Jun 12, 2015
    Actually we do have 3G. Thanks for the info. I will try to work something out with this suggestion.
     
Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.