How does the use of a reflector affect the performance of a receive antenna?

[berkeman]

If you place a second antenna within the near field of the first, the antennas will couple and confuse the analysis. Standing waves will make it a challenge.

I wasn't proposing placing them in the near field. They can be far enough away not to distort each other's near field, say a few wavelengths away. The depth of the null would indicate what fraction of the incident energy was being reflected away (or scattered sideways I guess).

There will be some scattering or reflection, but it will not always be 50% of the incident energy.

That seems reasonable.

It occurs to me that such scattering must be accounted for in phased array receive antenna configurations. I"ll have a look around to see if the literature addresses inter-antenna scattering in phased array antenna systems...

 

[tech99]

Krause ends section 2:14 with the statement; “Although the above discussion of scattering aperture is applicable to a single dipole (λ/2 or shorter), it does not apply in general.”

Kraus does not support your original assertion that a receive antenna reflects half the incident energy.
Kraus does point out that half the energy can be reflected when a receive antenna is operated into a load having the same impedance as the antenna loss resistance. But that deliberate analogy with a generator having an internal resistance, makes the same assumption that the voltage is fixed, independent of load, and that unlimited power is available. Receive antennas have a fixed power available and maximise energy efficiency by minimising loss resistance relative to output impedance.Kraus does not confirm that a paraboloid will reflect half the incident energy.
A “paraboloid” is used as a mirror to direct axial energy onto a transducer at the focus. A parabolic reflector is made of conductive metal, so it reflects close to 100% of the incident energy. You must separate the analysis of the reflective surface from the analysis of the transducer at the focus.

While that is certainly an interesting idea, a conductive screen reflector could make the same shadow. When a signal is canceled by having the reverse phase signal added, that simple cancellation would appear to annihilate two sources of energy, so it must be impossible. There must be other directional effects that conserve energy.

It is unclear what Kraus means by saying it does not apply in general. He seems to be distinguishing between aperture antennas and dipoles, where the collecting area is greater than the physical area.

Antenna reflects half power. I have never said that unlimited power is available. I said unlimited power is not available. The max that can be extracted is half.

Paraboloid. If the energy arrives at the focus, then the feed unit will re-radiate half the power and it will create a beam going back towards the transmitter.

Conductive screen. It would reflect the power back to the transmitter, whereas a 377 Ohm sheet backed by a reflector does not do so.

May I mention that an omni directional receiving antenna will reduce the incoming EM wave all around it, not just as a shadow. This is because it radiates a cancelling wave itself.

Of course, as I mentioned previously, parasitic elements reflect all the incident power, and power can be extracted as desired by altering the terminating resistor, up to a max of half.

 

[Baluncore]

I wasn't proposing placing them in the near field. They can be far enough away not to distort each other's near field, say a few wavelengths away.

The near field is usually specified as being out to 60 wavelengths. Now I am confused.

It is unclear what Kraus means by saying it does not apply in general. He seems to be distinguishing between aperture antennas and dipoles, where the collecting area is greater than the physical area.

I believe Kraus is considering only isolated short dipoles and apertures in section 14:2, Kraus is being careful to make sure the analysis in that section is not applied to antennas in general.

Conductive screen. It would reflect the power back to the transmitter, whereas a 377 Ohm sheet backed by a reflector does not do so.

Correct. Now consider replacing the space cloth with an array of dipoles in front of the screen. (Ideally spaced λ/4 from the screen).

Paraboloid. If the energy arrives at the focus, then the feed unit will re-radiate half the power and it will create a beam going back towards the transmitter.

As above, the antenna at the focus of a paraboloid is also operated against a small screen. You must analyse the antenna at the focus with the screen, independently to the paraboloid. Then multiply "the pattern of the antenna at the focus" by "the array factor of the parabolic reflector aperture".

May I mention that an omni directional receiving antenna will reduce the incoming EM wave all around it, not just as a shadow. This is because it radiates a cancelling wave itself.

But while the direct energy arrives from one direction, the scattered energy is re-radiated in the radial pattern of the antenna. That forms standing waves in all directions, except directly behind the antenna in the close shadow.
So it does not reduce the field with destructive interference everywhere near the antenna. In many places, the waves will sum to increase the field by constructive interference.

 

[berkeman]

The near field is usually specified as being out to 60 wavelengths. Now I am confused.

Me too. Can you provide a reference link? The more normal definition is a couple of wavelengths. I will also look for a link.

 

[berkeman]

https://en.wikipedia.org/wiki/Near_and_far_field

resulting in relative lack of near-field effects within a few wavelengths of the radiator.

 

[berkeman]

From my Google searching... I guess I need to subscribe to the IEEE to get to this whole paper...

http://ieeexplore.ieee.org/document/1296172/

Abstract:
This paper discusses the amount of power, which is scattered and absorbed by a receiving antenna and in particular, whether an antenna can absorb the entire power incident upon it. The absorbed and scattered power from dipole arrays in either free space, or over ground plane is considered. By defining a suitable "aperture efficiency" for the receiving case, a dipole array without a ground plane can best absorb half of the incident power (scattering the rest), while an array over a ground plane can absorb all of the incident power. It is shown how aperture efficiency varies with load impedance, which is of practical interest for array designers.

 

[Baluncore]

I guess I need to subscribe to the IEEE to get to this whole paper...

The conclusion is reproduced, outside the paywall, at the end of the abstract page.
D. Pozar. “Scattered and absorbed powers in receiving antennas”. 2004.
“In closing, I think that Allan Love's intuition that it is possible for a receiving antenna to “capture” all available incident power, without re-radiating or scattering any of that incident power, is correct, and is demonstrated by these results. In fact, of course, practical antenna performance would be seriously hampered if this were not the case.”

 

[Baluncore]

Following the 2004 paper by David Pozar, “Scattered and Absorbed Powers in Receiving Antennas”, the refinements and discussion continued. It reaches some maturity by 2009 in an interesting paper by Do-Hoon Kwon and David M. Pozar. "Optimal Characteristics of an Arbitrary Receive Antenna". The fundamental conclusions of the paper appear to be that;
1. An isolated dipole or an array of dipoles will have an aperture efficiency of 50%.
2. Placing the dipole or array over a ground plane will increase the aperture efficiency to 100%.

For those of you visiting Tohoku University see; http://www.sawaya.ecei.tohoku.ac.jp/common/item/pdf/doctor/100506.pdf

 

[tech99]

Me too. Can you provide a reference link? The more normal definition is a couple of wavelengths. I will also look for a link.

We need to distinguish two types of near field:-

(a) The Radiation Near Field, which is the region in front of a directional antenna where the pattern is not fully formed. It may be considered as extending to a distance equal to the Rayleigh Distance, approximately (Diameter^2) / 2*Lambda. It is also called the Fresnel region. In this region, an aperture antenna has an essentially parallel beam.

(b) The Reactive Near Field, or Induction Field, located very close to the antenna, where the fields due to the voltages and currents on the antenna are predominant. Usually extends to Lambda/2*pi. These fields contain stored energy rather than radiated energy.

I think Berkman was intending (b).

 

[tech99]

Following the 2004 paper by David Pozar, “Scattered and Absorbed Powers in Receiving Antennas”, the refinements and discussion continued. It reaches some maturity by 2009 in an interesting paper by Do-Hoon Kwon and David M. Pozar. "Optimal Characteristics of an Arbitrary Receive Antenna". The fundamental conclusions of the paper appear to be that;
1. An isolated dipole or an array of dipoles will have an aperture efficiency of 50%.
2. Placing the dipole or array over a ground plane will increase the aperture efficiency to 100%.

For those of you visiting Tohoku University see; http://www.sawaya.ecei.tohoku.ac.jp/common/item/pdf/doctor/100506.pdf

On this topic, Kraus says that a flat metal sheet has an aperture that collects the energy over four times its area. When configured as a receiving antenna, such as a paraboloid with a resistor at its feedpoint, the antenna can then have a maximum aperture equal to its physical area.

 

[tech99]

As above, the antenna at the focus of a paraboloid is also operated against a small screen. You must analyse the antenna at the focus with the screen, independently to the paraboloid. Then multiply "the pattern of the antenna at the focus" by "the array factor of the parabolic reflector aperture".

On a point of clarification, pattern multiplication applies to the case of a broadside array of unit antennas, where the overall pattern is the product of the patterns for the complete aperture and that of individual radiators. In the case of the paraboloid, if we increase the gain of the feed, it causes it to have a narrower beam, and this reduces the illuminated area of the dish, lowering the overall gain of the system and broadening its pattern. It is not correct to multiply the patterns (or gains) of the feed and reflector.

 

[Baluncore]

On this topic, Kraus says that a flat metal sheet has an aperture that collects the energy over four times its area.

Where does Krause write that? The devil is in the detail. Kraus has several different definitions of aperture; A_physical, A_geometric, A_collecting, A_scattering, A_effective, A_maximum-_effective, A_receive-effective, and A_transmit-effective. It is important to use consistent definitions which makes mixing definitions from different references very risky. You can hide almost anything behind an aperture definition.

It is not correct to multiply the patterns (or gains) of the feed and reflector.

Do you think pattern multiplication only applies to discrete point arrays and not to continuous apertures? Do you have a reference that shows the product of an illuminated continuous aperture and the driven element gives incorrect results?

 

[sophiecentaur]

It is not correct to multiply the patterns (or gains) of the feed and reflector.

Because it is not an element factor times and array factor; it's all one bit radiator and not a 'separable variable' problem. Each element on the surface of the dish is fed differently and 'pointing in a different direction'.

 

[tech99]

Do you think pattern multiplication only applies to discrete point arrays and not to continuous apertures? Do you have a reference that shows the product of an illuminated continuous aperture and the driven element gives incorrect results?

It would be very difficult for me to find a paper describing something that is incorrect.

Just for clarification, the aperture of a dish antenna can be considered as an array of Huygens Sources.
The feed is just one means of creating, so far as possible, a uniform amplitude and phase across the aperture. But significant radiation from the feed does not directly reach the receiver, and so it does not form an array in conjunction with the aperture. Therefore, we cannot apply pattern multiplication because the two sources do not constitute an array.

 

[sophiecentaur]

Do you think pattern multiplication only applies to discrete point arrays and not to continuous apertures? Do you have a reference that shows the product of an illuminated continuous aperture and the driven element gives incorrect results?

Pattern multiplication can only be used where the individual sources have identical radiation patterns (i.e. main beam direction). When you are using a paraboloid, this is not the case. Fourier Optics tells us that a paraboloid (or a lens) produces the inverse Fourier transform at infinity of an object at the focus. Altering the directivity of the feed will actually have the inverse effect on the overall beam width which is not a 'multiplicative' effect. (The very opposite, in fact.)
There are situations where a continuous set of radiators can be analysed by multiplying but it's not a general thing.

 

[Baluncore]

The feed is just one means of creating, so far as possible, a uniform amplitude and phase across the aperture.

That is the last thing I would want. By tapering the illumination to the edge of the dish, the side lobes can be significantly reduced while only slightly broadening the beam.

The slope of a parabola varies in proportion to the radius, the angle is Atan( slope ). The illumination, or aperture distribution, can be calculated from the geometry of the dish and the polar radiation pattern of the focal element. Once the aperture distribution is known, the beam pattern of the combination is the 2D Fourier transform of the 2D aperture distribution. Any step in illumination due to the edge of the reflector will result in ringing in the frequency domain. That makes the big side lobes.

But none of that detracts from the fact that the energy scattered by a “paraboloid” receive antenna is determined only by the focal transducer configuration, and not by the presence of the reflector dish.
The direction of scattered energy will be determined by the presence of the dish, but not the total energy scattered, nor the efficiency of the receive system.

The energy reaching the receiver can be doubled by replacing a bare dipole at the focus with a more complex antenna such as a reflector backed dipole, a yagi or a horn. +3 dB makes a big difference near the noise floor. That explains why radio astronomy dishes do not have a simple dipole at the focus. The reciprocal, or transmit analogy, is that a reflector backed dipole as a driven element will double the energy illuminating the dish surface compared with a simple dipole.