1
$\begingroup$

I want to null out a particular direction so that i can't receive some vertically polarized interference on 10 m coming from that direction.

Which 10 m antenna will give the most rejection in one direction in the azimuth plane for received signals which are vertically polarized ?

I've already tried 3 and 4 element horizontal and vertical yagis and these don't provide enough rejection in the nulls in their patterns. I need 40 dB of rejection or more. Can i use parasitic elements to somehow cancel out in one direction only ?

$\endgroup$
  • 1
    $\begingroup$ Do you need to reject a single frequency, or a band of frequencies? $\endgroup$ – Brian K1LI Oct 15 '19 at 14:09
  • $\begingroup$ A band of frequencies about 1 MHz wide. $\endgroup$ – Andrew Oct 16 '19 at 2:39
2
$\begingroup$

There are several commercial units that work with a pair of antennas to reduce signals from a specific direction. These products are typically called "noise canceller", "signal enhancer" or "receive antenna phasing system".

The block diagram of a typical product is shown below:

NCC-2 block diagram

By adjusting the relative amplitudes and phases of received signals arriving from two antennas, unwanted signals arriving from particular directions may be attenuated. Note that one of the two input antennas may also be the station transmitting antenna and some units include a collapsible whip antenna, potentially obviating the need for additional structures.

The adjustable nature of these designs makes it possible to achieve substantial interference reduction over a wide range of frequencies - though probably only over a relatively narrow range of frequencies for any specific set of amplitude and phase adjustments. Product documentation includes antenna recommendations for signal enhancement, signal rejection and noise rejection. One product manual includes a complete schematic diagram of the unit.

The OP seeks to achieve a very substantial amount of interference reduction. The choice of receiving antenna(s), their spacing and height above ground may substantially affect rejection of signals arriving at specific elevation angles. It may be useful to model candidate antenna systems to evaluate performance before construction.

| improve this answer | |
$\endgroup$
  • $\begingroup$ that is an excellent answer and exactly the kind of idea i was thinking about, but i didn't know where to start - well done and thank you (sorry for saying thanks in a comment). : ) $\endgroup$ – Andrew Oct 18 '19 at 22:04
  • $\begingroup$ i am investigating this, can you give me specific product details ? $\endgroup$ – Andrew Oct 18 '19 at 22:06
  • $\begingroup$ one other question, i have a coupe of Yaesu FT2000s, do these have any special functions that would help with this idea ? $\endgroup$ – Andrew Oct 18 '19 at 22:08
  • $\begingroup$ An internet search using the terms in the answer reveals products from DX Engineering, MFJ Enterprises and Timewave Technology. There may be others. The manuals for several of these products include schematic diagrams. $\endgroup$ – Brian K1LI Oct 20 '19 at 12:09
  • $\begingroup$ In addition to an input for a separate receive-only antenna, the FT2000 includes so-called $\mu$-tune jacks to provide for addition of RF preselectors. It might be possible to introduce differential phase into one or both of these paths so that listening to the combined outputs of the main- and sub-receivers would provide the cancellation you seek. $\endgroup$ – Brian K1LI Oct 20 '19 at 12:24
4
$\begingroup$

Constructive and Destructive Interference, Directivity

This is an excellent time to point you towards antenna arrays!

So, maybe you remember the double-slit experiment from school. In case you never did that, or need a refresher:

Imagine having e.g. a laser, that is, something emitting a plane wavefront (from the left in the below picture). You shine that light on a plane with two small slits:

from wikipedia:

You've got two very small slits, emitting (half-) circular wavefronts that have the same phase (in the plane of the slits); when you look from a distance, you'll see that these wavefronts overlay constructively, whenever the run length difference of both waves is 0 or any other multiple of a wavelength. They interfere destructively, i.e. they cancel out, whenever the run length difference is an odd multiply of half the wavelength. These things line up in directions from the centre between the two slits, so that we can talk of directivity!

As you can imagine, adding more slots in the same spacing just makes the 0-difference direction "stronger" and all suppressed directions even more suppressed.

Light's just slightly higher-frequent radio

So, light is just electromagnetic waves, just as 10m radio waves. The latter are just slightly easier to handle – we don't need to build something that makes a plane wave front and then have some opaque material with slits in there. We can just put antennas with circular emission pattern at the point of the slits!

The whole idea of "we add delayed versions of the same RF wave up" is what makes multi-element antennas (and any aperture antenna) work: The elements of the Yagi are just sized and spaced that way, that they re-emit energy just in the right moment to lead to constructive interference of the radio wave in the main direction of the Yagi antenna. By the way, what works in TX works exactly the same in RX.

Due to the geometry of the antenna system defining that main direction this way, Yagis are fixed-direction (unless you rotate them). That's fine for many applications, but I think you need something with which you can find a good direction to receive from.

Enter: Antenna Arrays

Now, what would you have to change in the above double-slit experiment to change the direction(s) in which the main maximum occurs? You'd simply change the distance between the slots. Or, you could find a "magic" device that you install in one of these slots that delays the phase of the wave coming from that slot by some adjustable fraction of the full wave cycle, and with that you could steer the beam.

While that's nontrivial for photonics, it's relatively easy for the relatively narrowband HF signals: We call such a device a phase shifter:

                                                       v
                              +----------------------> |   Antenna 1 
Transmitter amplifier ---> splitter
                              |                        v
                              +-----> Phase shifter--> |   Antenna 2
                                         ^
                                         |
                                    Direction control

Tadah, by adjusting the phase shift, you adjust in which direction your transmitted radio waves constructively add up; that's your antenna system's main direction! And you get a gain in that direction. If you want better directivity, repeat the above scheme; from wikipedia:

Remember, what works in TX works in RX, too! Adjust the phase of the received wave of one antenna, add the electric signals up, and get a signal where the reception is good from one direction, and worse from the others.

How to Shift your Phases (All your Phase are belong to us)

For most of the previous century, using (mechanical!) analog phase shifters was state of the art for beam forming, and it's been extensively used, especially in radio sensing applications; and since that's something that military people like to do (for example, to have knowledge of where someone's transmitter is standing, or to build a radar with which you can look very far in a very specific direction), cost was …… less of an issue.

That changed around the turn of the millennium, when Software-Defined Radio (SDR) became a viable thing. Idea is simple: Take your RF signal, digitize it (like a sound card converts electrical signals from a microphone to digital numbers), and do the phase shifting and adding up in software. Done! Computers are cheap, and fast enough.

So, what you'd need is

  1. a set of omnidirectional antennas in the polarization of your choice
  2. a place to set them up in a line, in typically quarter- or half-wavelength distances
  3. a SDR receiver for each of them
  4. a way to "null" the phases of these receivers (otherwise they'll just be random)
  5. a computer (seems like you have one!)
  6. and a tiny bit of software to add up these streams

For 1., antennas: The easiest and – perpendicular to its conductors – omnidirectional antenna is the dipole. And in fact, many laaaaarge antenna arrays are simply made of dipoles, or similarly low-gain antennas. See EISCAT, for example, which I know hams that were involved with:

For 3. Receivers: For 30 MHz, basically all RTL-SDR dongles will do their job. Depending on where, in which quality and in which quantities you buy them, 6 to USD 40 a piece.

(from the OSMOcom wiki linked above:)

cracked open RTL-SDR dongle of an early generation

For 4., you'll just need something to calibrate your receivers. A transmitter in a known direction would work – you can calculate the phase shifts that the antennas should have with easy trigonometrics, and then just adjust their phases to have that, in doubt, manually. It'd be good idea to feed all receivers from the same reference oscillator source, so that you don't have continuously readjust (because the phases will drift away pretty quickly if they run from different reference oscillators), and that'll involve a bit of soldering (unsolder the oscillator from all but one dongle, add an amplifier and clock distributor to that one dongle, solder in cable to where the oscillator(s) used to be).

For 6., software: a tiny bit of GNU Radio, and pipe the result (e.g. via network) to the SDR receiver software of your choice (GQRX, SDR#, HDSDR, LinRad, I don't know a tenth of the candidates) to do the demodulation.

| improve this answer | |
$\endgroup$
  • $\begingroup$ excellent answer thank you, i wasn't a big fan of SDRs but now is the opportunity to change my mind : ) $\endgroup$ – Andrew Oct 18 '19 at 22:21
4
$\begingroup$

I might be missing something here, but one of the antennas in common use that has a very sharp null in one direction (and its opposite, so I guess two directions) is the one used by ARDF enthusiasts the world over - the magnetic loop.

You can design and use a magnetic loop at 28MHz, which would be surprisingly compact, and would do a reasonable job of receiving in all other directions - but be aware that it also has a very sharp pass-band, so you would want to make it tunable. This means if you were to mount one outside, you would have to come up with a remote tuning capability (a motor to drive the variable capacitor, for example).

As a transmit antenna, the magnetic loop is not particularly efficient, but it works fine as a receiver. In fact, if you open a regular domestic medium wave radio (so-called "AM" radio receiver), you will find one in there wound on a ferrite rod. You will find that you can turn the radio while listening to a strong station on the "AM" band and find a position where the station all but disappears. Similarly, if you build a magnetic loop antenna for 28MHz, you could turn it to reject the signal you are trying to avoid, and you simply won't hear it.

| improve this answer | |
$\endgroup$
  • $\begingroup$ Hmm ... how about a large 10 m version of the AM radio ferrite rod antenna without the ferrite rod ? $\endgroup$ – Andrew Oct 16 '19 at 2:42
  • $\begingroup$ @Andrew I don't know what you mean by that, but what Scott describes is a time-proven design. EDIT: It will have two sharp and very deep nulls at low angles 90° to the plane of the loop (broadside), 180° apart. $\endgroup$ – Mike Waters Oct 16 '19 at 3:17
  • 1
    $\begingroup$ Google for magnetic loop designs - you have to be aware of extremely high voltages if you transmit through one, but other than that it's just a loop of metal with a high-voltage air-spaced variable capacitor on it. Simple and relatively inexpensive to make or buy. $\endgroup$ – Scott Earle Oct 16 '19 at 5:29
  • 1
    $\begingroup$ The magnetic loop's pattern is identical to that of a small dipole. It's not special at all in having a sharp null, in fact by the hairy ball theorem every antenna must have at least one complete null. $\endgroup$ – Phil Frost - W8II Oct 17 '19 at 18:11
3
$\begingroup$

You might want to try not one, but 2 identical vertical dipoles spaced a half wavelength apart in line with the noise source and fed to an active noise cancelation box or a coherent dual-input SDR.

| improve this answer | |
$\endgroup$
  • $\begingroup$ ah didn't see your answer while writing mine: Yeah, the smallest possible antenna array :) $\endgroup$ – Marcus Müller Oct 15 '19 at 12:48
  • 1
    $\begingroup$ If the noise is vertically polarized, then horizontally polarized antennas would be even better... $\endgroup$ – Phil Frost - W8II Oct 15 '19 at 15:12

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.