6
$\begingroup$

What is the best omni-directional antenna design for any RC flying objects? Many times I've lost signal and my copter falls due to lost signal. (Using default antenna). Since the receiver and transmitter position will be in different altitude, I can not imagine what is the best antenna design for this.

I am thinking about the receiver would use their default antenna, and the transmitter will use DIY omni antenna, is by this way it can control better?

$\endgroup$
  • $\begingroup$ best antenna is an amplifier.. $\endgroup$ – captcha Sep 23 '14 at 23:54
  • $\begingroup$ @captcha: That is not correct! You can't amplify signals if they are not there, so the first concern should be to capture the signal as best as you can! $\endgroup$ – jcoppens Oct 5 '14 at 17:38
  • $\begingroup$ @jcoppens Not if amplifier is in the copter! :-) Sorry, been contesting on hf too much.. $\endgroup$ – captcha Oct 6 '14 at 6:11
3
$\begingroup$

Wow... The simplest antenna (not entirely that simple :) for your use is the quadrifilar helicoidal antenna (abbreviated QFH or QHA). This antenna has a few characteristics which are important to you:

  • It radiates in a very wide angle (over 180 degrees)
  • It has circular polarization (which makes the position of the RC plane irrelevant)
  • It doesn't have a 'dead' spot above it, like the groundplane

Note that both the transmitter and the receiver have to be fitted with the same antenna type (and the 'twist' should be the same, i.e. both should be left- or right-hand circularity).

Have a look at: http://jcoppens.com/ant/qfh/index.en.php There's a calculator to design one for you. There are also a couple of references for more information. I just noticed that the link to the photos is missing - I'll update the page in a moment. Here's a link to pictures of a Vaisala GPS QFH antenna for 1.6 GHz: http://jcoppens.com/ant/qfh/index.en.php

$\endgroup$
  • $\begingroup$ Do you have any reference to a simulation or empirical measurement of such an antenna? I can't find much more than a picture of the radiation pattern. $\endgroup$ – Phil Frost - W8II Oct 6 '14 at 12:04
  • $\begingroup$ @Phil, look around in the website. There are simulations with several height/width ratios, and with and without a ground plane nearby. (jcoppens.com/ant/qfh/sim.en.php) $\endgroup$ – jcoppens Oct 6 '14 at 16:09
  • $\begingroup$ @Phil... Are you the same Phil that wrote the article below? You mention the Isotropic antenna yourself. Anyway. The Isotropic antenna is just a mathematical construction to obtain a really ($4\pi$) omnidirectional antenna. As it is infinitely small, you cannot determine its polarization (of the Isotropic antenna, that is). $\endgroup$ – jcoppens Oct 6 '14 at 16:13
  • $\begingroup$ Looks like your website is down. It would probably be good to include some key points in the answer. $\endgroup$ – Phil Frost - W8II Oct 13 '14 at 16:42
  • $\begingroup$ Website is back up. Was sown over the weekend because the hosting site changed operating systems. $\endgroup$ – jcoppens Oct 15 '14 at 4:52
2
$\begingroup$

A ground plane antenna might be what you want here; according to this post (http://www.rcgroups.com/forums/showthread.php?t=1333382), which also includes a detailed tutorial with pictures, the multiple grounded elements change the propagation pattern of the antenna and send the signal upward.

This sort of design is good for both the transmitter and the receiver, and can be cut for 2.4 GHz without much trouble (the elements will be about an inch each). I'll paraphrase and adapt the tutorial's building instructions for the substrate-type antenna (built on a small piece of protoboard) here:

  1. Calculate the quarter-wave length of each element as (2808/f) inches or (71323/f) millimeters, where f is the frequency in MHz you want to operate on.
  2. Cut several pieces of wire at this length to use as your elements. You can use any gauge wire you have available (low gauge is typically better as it is sturdier - 12 AWG is my recommendation). Cut any number of pieces you want, between 4 and 9. 5 is a good number because the construction is easy while still having a good radiation pattern.
  3. Solder all but one of your quarter-wave wires to a common point in the center of a small piece of protoboard (onto the copper side). The wires should be spaced an equal angle apart (e.g. for four wires, they should be joined to make an 'X').
  4. Drill a small hole in the substrate as close to the solder point as possible, large enough to pass the coaxial cable's center conductor through.
  5. Strip the insulation off your coaxial cable, pass the center conductor through the bottom of the board, and solder the last quarter-wave element to the center conductor so the wire rests directly above the solder point. Then solder the cable's shield to the solder point on the substrate.
  6. Measure all the elements to be sure they are exactly the length you calculated above, as measured from the center of the antenna, and trim if necessary (especially check the center one since you have joined it to a second wire).

The author's notes indicate that this antenna performs well when the four center conductors are supporting the antenna at a 120-degree angle from the ground plane (this is why you use low-gauge wire here, for strength).

$\endgroup$
1
$\begingroup$

The "best" antenna depends strongly on the relative orientation of the transmitter and receiver. Since any RC craft is constantly and rapidly changing orientation, there is no antenna that will work well for the entire flight. A dipole, ground plane antenna, loop, or any other basic antenna will all perform about the same.

There are two reasons one antenna can not work in all orientations:

Firstly, a truly isotropic antenna can not exist. This means that any antenna you can build will have at least one null, a direction in which it does not radiate. When your craft is in the transmitter's null (or the transmitter is in the receiver's null), you may lose the signal.

Secondly, transverse waves such as EM radiation used in radio all have a polarization. When these polarizations are orthogonal, no power transfer occurs between transmitter and receiver. Again, signal is lost.

The solution is to have multiple antennas, in different orientations. This is called antenna diversity. If the receiver or transmitter have a number of antennas available, each in a different orientation, thus radiating in different directions, with different polarizations, then it's possible to select the one that is best at each instant.

Of course, this means your communications system has some way to determine what is "best". This can be done, for example, if periodically the transmitter sends a known synchronization sequence. The receiver can then analyze signals from each antenna, and using a method like regression analysis, calculate coefficients for each antenna that result in the best fit to the expected synchronization signal. The received signals from each antenna are then multiplied by these coefficients and added, and the result is used to recover the message. In the context of antenna arrays, this is called minimum mean square error weighting.

Modern wireless communications protocols, such as 802.11n or LTE Advanced do this. Usually it goes by the name of MIMO, which is, in essence, what I described. MIMO also includes some more advanced techniques like spatial multiplexing for increased data rate.

Unfortunately, no commodity RC system does this to my knowledge. As such you may be left to implement it yourself.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.