Near Field of Horn Antenna

[waveboy]

I am trying to find the near_field-far_field boundary for a Horn antenna with a rectangular aperture of 10cm by 20cm. I am using it for frequencies 8 to 10 GHz.
I have read that this boundary is given by 2*D^2/lambda (where D is the largest dimension of the antenna). -However, am I correct in stating that the "D" in this equation is NOT 20cm for this horn? -Surely the aperture is just the end of the tapered-out waveguide, and the "antenna" is really the feed probe in the rectangular waveguide section of the horn.
Any help greatly appreciated.

 

[VEBill]

It gets more amusing...

Once you've got it all figured out for your horn, then consider what happens if the horn is then suspended hundreds of feet in the air and positioned at the focal point of the Arecibo dish.

What happens to the near-field / far-field boundary then?


Arecibo info:

http://en.wikipedia.org/wiki/Arecibo_Observatory

 

[RF0]

Hello all,

is horn antenna electric or magnetic or both (depending on configuration) in the near-field? I dont know why I was in impression that it can act as magnetic antenna?

sorry about the dumb question. It just came to my mind when I saw this thread.

 

[waveboy]

Thanks VE1BLL...that's interesting. For the Aricebo dish, I belive it is correct to say that the "D" (see equation above) of this antenna would not be the diameter of the dish...surely the dish is just a reflector [although since reflection is really re-radiation then do we say that the dish itself is actually an antenna in this system? (as well as the feed probe in the waveguide section of the feed horn of Arecibo)]
-Also, there is the "reactive" near field and "radiating" near field to be considered in this system.

Also, (generally) the bigger the dish, the more closely the waves coming out of it are plane_wave-like, and since the far field is (partly) characterised by being "plane_wave-like", then the big Arecibo dish would have a very "close" far-field (?)

 

[VEBill]

If a dish isn't an antenna, then is a horn? After all, a horn is simply a section of waveguide that transitions from waveguide to open space. The actual antenna element is often way down inside, perhaps a monopole.

I've always thought that the definitions of near-field and far-field were man-made, arbitrary and variable (several different definitions). As I've mentioned before, nothing significant happens at the boundary no matter where the committee decides to scratch the line.

 

[waveboy]

I am wondering how to test various antenna's for their antenna pattern. (I have no anechoic chamber but just a small field backing onto the Forest of Dean). My intended set-up is a source antenna (horn) which radiates toward the antenna_under_test (AUT). The AUT will be rotated 90 degrees either side of boresight. I must ensure that both source and test antenna's are out of each others significant reactive_near_field (else their radiation patterns will be affected as in a yagi array). Also, to diminish aperture_phase_errors, i must ensure that each antenna is out of the other's significant radiating_near_field. However, i also want the antennas to be as close as possible, so that the need for an RF amplifier on the source horn is obviated (since high gain, low noise, high P1dB, RF amplifiers at several GHz are very expensive).
Therefore, I would be most grateful if anybody knew what the equation for the near_field-far_field boundary distance is for all types of antennae. (e.g. Horns, Parabolics, Dipoles, Yagi's, Planar spiral antennae, arrays of patch antennae, log periodic dipole antennae, small patch antenna etc etc)

 

[VEBill]

You'd already mention the accepted formula for the near-field / far field 'boundary'.


Everyone has access to an anechoic chamber: it is outdoors in an empty field (look straight up). The main practical issue is that of suspending one end or the other without causing significant reflections. I guess bad weather might be another practical issue... This also assumes that you can find a quiet frequency and use a narrow band receiver.

Usually, you just need to keep reflections of the main beam out of the sidelobe measurements. That's where the ratio (xx dB down) of the reflections might affect the measurements. Therefore, (outdoors) you aim the source straight up and move everything else. This assumes that you have a directional standard for the source.

I'd think that this would apply in chambers too, where the source should be fixed aimed at the best area of the RAM absorbers. Aiming the source all over the place would ensure that your results would contain garbage from every possible reflection.

There's been a lot of papers published on making near-field measurements and then adjusting the results for the far field. Way over my head, but I've seen the references.

Once you have the 3D pattern (or enough slices), integrate it and make sure it averages out to 0 dBi. It's fairly easy in MS-Excel. Don't forget to weight down the over-represented data (typically near the poles of a Mercator data set). I've seen patterns where the antenna must contain a power source (ha ha) because the published pattern integrated to + several dBi.

Scale models are common so that the experiment fits into the available resources.

SW models are increasingly common.

 

[logbook]

I don't know as much about this subject as I would like to but if you bounce a horn antenna off a curved metal surface the D is the diameter of the mirror as far as I know. I work with microwaves and mirrors and the optics guys always use the final mirror diameter for the Rayleigh resolution limit. A flat plate would be a null surface but a curved mirror shapes the beam. A typical horn antenna spreads the beam too far. By using a huge dish you get a more directional beam so yes it does affect the near field point.

 

[IRstuff]

In optics, the curved reflector collimates the beam so that the emitted radiation is far-field. There is effectively, no near-field in that case

TTFN

 

[waveboy]

Hi, I'm really grateful for all these replies. Actually, concerning the Antenna Pattern Testing set-up. -I do not have the resources to suspend the Antenna_Under_Test up above the source antenna. -In fact, I just have to make do with having the souce and test antenna 2 metres above the ground. This means that there is a possibility of having a ground reflection which would interfere with the antenna pattern. My Antennas under test tend to be tested at increments of 1GHz between as much as 2 to 18GHz or more.

I was thinking that it would be best to get the Source_Antenna and Antenna_Under_Test as close as possible. -That way, the ground reflected ray would be well out of boresight and likely to be just a weak sidelobe ray. However, i must ensure that i do not get the antennas so close that they are in each others' significant near field. -That's why I would like to find some really solid information on the Near Field equation (e.g. what the "D" actually is for all different types of antennas)

To check that I have the source and test antennas in each other's far field, i was thinking of just doubling their separation and noting if the received power goes down by 6dB. (Though this would be time consuming and I would prefer to have a solid equation for the acceptable separation of source and test antennas.)

 

[waveboy]

Concerning RF0's question, -i think it depends on the type of feed used in the horn (loop or probe)