Horizontally mounted small active receiving antennas
Chavdar Levkov , LZ1AQ, email@example.com
Last revision 1.10; 28 Sept 2015 ©LZ1AQ
We can mount a small loop or a small dipole horizontally as shown on Fig.1. It is interesting to compare the small horizontally mounted antennas with the identical “standard” vertical mount. The only difference is that the antenna polarization is changed. In the next text, to save space, we will denote vertically and horizontally mounted small loops and dipoles respectively H-loop, V-loop, H-dipole and V-dipole. Vertical and horizontal polarizations will be denoted as VP and HP. There are numerous textbooks and articles on antenna polarization subject so the reader is advised to get acquainted with fundamentals.
Fig.1 Horizontally mounted small loop antenna at 7 m height. It consists of two crossed coplanar loops [1, 2] each with 1 m diameter. A vertical loop is seen on the picture mounted at 2.5 m height above the ground.
1. The benefits to use both VP and HP antennas
The polarization of the incoming wave must match the polarization of the antenna, otherwise the signal will be attenuated. The polarization is just as important as the directivity. This is well known in VHF but on HF bands due to the random change of polarization usually this is not a scope of interest. Change of the polarization of the waves reflected from the ionosphere is random – it can be slow as well as very fast. The polarization plane might wobble slowly or it can rotate with low frequency. It is known that the substantial part of the fading comes from polarization changes. Fig.2. shows a case where the fading of the signals from two HP and VP antennas placed on the same mast (Fig.7) do not have the same phase. If we combine in some way these two signals (diversity reception [9, 10]) there will be improvement in reception. Very often one of the polarizations is prevailing and switching to the matching antenna will improve the reception substantially Fig.3. Sometimes even 1 dB improvement of S/N ratio is the difference between successive and failed contact.
Another benefit is that the noise sensitivity to local sources will be different for HP and VP antennas and it might be possible to eliminate effectively the noise with one of the antennas .
Fig.2 Fading with different phases form HP and VP antennas. The SDR Linrad S-meter (RMS value) window shows the relative strength of the signal as a function of time. Two small active identical loops (Fig.7) horizontally and vertically mounted are periodically switched alternatively with period of approximately 2 sec. Linrad SDR software  is used with Perseus DDC receiver . A digital transmission station is received at 16.475 MHz. The moments of antenna switching are clearly visible. H-loop levels are marked with red color and V-loop with green. At the beginning H-loop is stronger but after 20 sec the polarization is changed. The differences between the signals reach almost 18 dB at time. If a diversity reception is used the fading variance will be only 2 to 3 dB. The V-loop is not pure omni-directional VP sensor but its VP is predominant if pointed to the direction of the incoming wave (Fig.9 ). The receiver is set in AM mode, with wide bandwidth and digital transmission station is chosen to avoid the effects of selective fading.
Fig.3. The case of prevailing VP which reaches 28 dB difference at times for digital signal at 10.57 MHz. Somewhere in the middle of the window both polarization give equal levels and then VP again becomes stronger. The peaks are from the switching process between antennas. If we imagine that the RX noise floor is at 60 dB level, the H-loop will not hear the signal at all while the V-loop will have decent signal. Here the diversity reception will not have any advantage compared to V-loop used alone. The setup is the same as in previous picture.
Note that in both pictures the V-loop and H-loop signals (the fading curve) are negatively correlated (a peak in one of them coincides with a valley in the other) which proves in some sense that the sensor polarizations are orthogonal. Perfect negative correlation is not possible since the fading is due not only to polarization rotation but also to other random factors e.g. multipath interference etc.
2. HP antenna needs height to be efficient
If a small antenna has good sensitivity to electric field vector E that is vertical to ground surface it can work at very low height. But horizontally mounted small loop will be very inefficient receiving antenna if it is placed near the ground since for vertical electric vector it has almost null sensitivity.
The reason is the ground reflections. All our antennas work in the region near (up to several wavelength distance) to the ground. In this region a complex interference picture is formed in the process of the interference between the incident and the reflected from the ground waves.
For the simplest case of a pure HP incident wave, the reflected wave is also HP but has 180 deg. phase shift. Both waves are superimposed forming an interference pattern as a standing wave in vertical (z) direction. It is a similar process to the case of a reflection from a short circuited transmission line. The amplitude of this horizontally polarized standing wave at the ground boundary is zero and increases with height. The height of the standing wave maximum depends from the arrival elevation angle (take-off angle) of the incident wave . Fig. 4 and Fig.4a show the dependence of the standing wave amplitude of HP wave as a function of the height (z coordinate) to wavelength ratio and parameter the arrival angle. The less is the arrival angle the higher must be the antenna height to compensate the attenuation. In this case the resultant E vector does not have any z component at any height e.g. it is always parallel to horizontal plane.
Fig.4 Field intensity as a function of relative height above the ground surface .This is the intensity of the standing wave which results from the interference of the direct incident HP wave and the wave reflected from the ground surface. The standing wave maximum depends also from the arrival angle (elevation) of the incoming HP wave. The curves are computed for the case of ideal reflection without any losses. The dB scale represents the relative amplitude ratio between resultant field and the field of the incident wave in free space. The maximum gain of 6 dB is at a height point where the phases of the direct and the reflected waves are the same.
Fig. 4a Reflection interference at a point 8 m above the ground of three HP waves coming from different arrival angles for 7 MHz band (40, 15 and 5 deg.) . The resultant field intensities W40, W15 and W5 are taken from the plots on Fig.4. It is obvious that the important reflections are at points in immediate vicinity of the antenna. The distance of the reflection point is increased when the arrival angle becomes lower. The interference maximum is +6dB ( two times) and will be at different heights above the ground for each wave. For W40 it is at 16 m height but for 5 deg. wave it will be somewhere above 80 m.
There is a very spectacular java animation applet at [7, 8] which demonstrates visually (by solving Maxwell equations) the process of reflection of a horizontally polarized plane wave. The color plot of the Poynting vector (flow of energy) is given on Fig.5 in the plane of incidence of incoming wave. The reflection plane is at 45 deg. and normal to the plane of incidence. It can be seen that there is a traveling wave which is parallel to the reflecting surface. The energy concentration in the standing waves is seen in direction normal to the surface,.
Fig.5 “Computer simulation of wave reflection from a surface. The direction in which energy is traveling at any instant is proportional to the cross product of the electric and magnetic fields, known as the Poynting vector. The arrows show the direction of the Poynting vector averaged over the simulation, while the colors show its strength. The electric field vector is parallel to the reflecting surface (out of the plane of the page).” From http://www.met.reading.ac.uk/clouds/maxwell/total_internal_reflection.html  The simulation is for the case of HP wave.
The reflection of VP waves is more complex (HP and VP are explained in details in the Appendix). Now there is a maximum very near to the ground surface. There, the resultant E vector has only z component (it is vertical), but its intensity depends very much from the real properties of the reflection surface and also from the arrival angle. In [4, 5] there is comprehensive description of the reflection process and plots of the filed intensity where the reflection factor of the real ground is also taken into account.
As a result of reflection of an arbitrary polarized plane wave, a complex interference picture is created near the ground surface. First of all the flow of energy is parallel to the ground surface, so we can assume that a traveling wave with direction parallel to ground exists. The resultant E vector of this wave changes its amplitude and direction with height forming the interference picture but E always lie in a plane normal to the ground surface. Near the ground this vector is almost vertical with negligible horizontal component. This is the field that is captured by H-loop or V-dipole placed at low height. There is an interference maximum (standing wave maximum) for the horizontal component of this vector at certain height. We must place the HP antenna near this standing wave maximum in order to capture maximal power from the horizontal E field component. The specific interference picture depends from the arrival angle and the polarization of the incident wave. Obviously any receiving antenna, is “immersed” in an interference field created by the superposition of the direct and reflected waves. The idea that we capture the incoming wave directly with our antennas is simply not true. This explains the fact that there is not much reason to tilt our short wave antennas to the sky from where the waves come. Our antennas must capture the electromagnetic energy that flows parallel to the ground surface.
What must be the minimal height for HP antenna in order to have sufficient sensitivity? Probably 0.1 of wavelength (wl) is useable. The ground is not an ideal reflector and not a pure flat surface. The real phase shift after reflection is also not 180 or 0 degrees  and the total effect of the “standing wave” probably will be not so strong in real ground case. Also the surrounding land shaft (near hills and valleys, large buildings) can play a significant role creating specific interference picture. This is true for urban environment where unexpected high levels of HP waves might occur at low heights. The user is encouraged to try also lower heights since for some propagation conditions the effect of matched polarization can compensate the standing wave effect. Higher heights will benefit the small HP antenna as is the case with any other full sized HP antenna. An interesting conclusion can be drawn: if the ground reflection is weak (lossy or absorbing media) the HP antennas can work at lower heights. Is this true ?
3. Diversity reception with dual polarization
Diversity reception is a method of combining two signals from different antennas to improve the S/N ratio . Most often the diversity reception is performed with two antennas (with the same or different types) placed at some distance between each other. The fading of the signals from these two antennas usually do not have the same phase and if we combine in some way these two signals there will be improvement in reception. Two phased locked receivers are needed for this purpose . This technique is usually associated with big antennas and large spaces, but it can be applied anywhere, even on the balcony of an apartment (Fig.7) using small active antennas with different polarizations and very simple additional hardware . Two small active antennas must be placed on the same mast each with different polarization. The effect will be maximal if the antennas have mutually orthogonal polarization. More about polarization diversity reception can be found in another publication of the author “Diversity Reception with a Modified SDR Receiver and Small Active Antennas “ . The best pair is V-dipole and crossed coplanar (CC) H-loop [1, 2]. They are pure VP and HP antennas even above real ground. Pairs V-loop & H-loop , V-loop & H-dipole or V-dipole & H-dipole will also work but only the first pair is omni-directional and orthogonal for all azimuth and elevation angles. In the same publication I present several audio records with stereo diversity reception with VP and HP antennas  . There are evidences that sometimes the incoming wave physically consists from two eliptically polarized waves with right and left rotation [20, 21]. These waves have different ionospheric path and can be detected with suitable equipment. Two orthogonally polarized antennas are needed and V-H pair can be used for this purpose.
On Fig.6 is given the experimental setup with single AAA-1 amplifier . With this setup we have 3 modes which can be switched in order to compare different antennas. The horizontal CC loop is connected to A terminals and a separate small vertical dipole is connected to V terminals. J1a and J1b = OFF. The loop is “pure” horizontally polarized and the dipole is a “pure” vertically polarized both with ideal circular radiation pattern. A third vertically mounted CC loop is connected to B terminals. This is the most versatile antenna setup where we have three antennas with vertical, horizontal and “mixed” polarization.
Fig. 6 Multi mode setup with AAA-1. Here we have pure HP with horizontal loops, pure VP with vertical dipole and mixed polarization with vertically placed loop.
On Fig.1 is shown CC circular HP loop , horizontally mounted at 7 m height. The loop diameter is 1m with 14 mm PE coated tubes. The location is in a small village with relatively low manmade noise. It was compared to vertical mounted CC loop with the same size which was mounted at 3 m height. AAA-1 amplifiers are used. On 1.8 MHz the HP loop noise floor was limited by the amplifier noise, since the signals were relatively weaker due the low height. But this loop was very quiet – I have a steady strong noise from high voltage line passing 100 m from the house. This noise can not be rejected with the vertical loop even if the loop is turned to optimal noise reduction position. But with HP loop this noise was not present at all. On higher bands the HP loop was a very good performer and having both antennas will give the best receive possibilities for particular situations.
In order to increase sensitivity for HP mode I build another dual polarization antenna with larger size (Fig.17 in Appendix) placed at the same location. The HP and VP antennas are with the same size placed on the same mast with the same AAA-1 amplifiers. The horizontal loop is at 7 m height. Very interestingly down to 1 MHz the noise floor in all modes was determined by the external atmospheric and manmade noise so there is no reason to build larger antennas. On Fig.6a are shown the spectrograms in the most problematic LW and MW bands. As it can be seen HP antennas have lower S/N ratio compared to HP antennas especially for weaker stations which are supposed to come from lower arrival angles. But for stronger closely located stations the S/N ratio is almost the same or even better probably due to decreased manmade noise in HP mode. For higher bands the performance of the antennas is as expected - which is the best antenna depends on the specific polarization of the incoming signal.
Fig. 6a SDR spectrograms of the band from 0.1 to 1.7 MHz made with 4 different antennas, almost at the same local time 22:30 ( the difference is not more than 5 – 6 minutes) on 11 July 2015. Note the lower noise level of HP antennas. For loop antennas the difference is almost 10 dB and for dipoles is almost 20 dB! The best S/N ratio is with VP-loop. The worst is with HP loop which has pure HP. H-dipole has better performance since it has VP for waves coming from the dipole axis direction. Irrespective of the low height the horizontally placed antennas have decent performance. Both dipole antennas have strong electric field noise (in 200 to 400 KHz band) probably from high voltage line passing 100 m away from the antennas.
I installed two H and V loops at my apartment balcony in Sofia city. Both CC loops (Fig.7) are with identical construction 0.7 m diam. which is described on Fig.3.6 of the Antenna part of the AAA-1 manual . This is a concrete 6-store building and the antenna was mounted on the second floor approximately 5 meters above the ground level.
I made careful measurements to compare the horizontal and vertical loops performance. As expected, the low frequency performance of the HP loop was poor. On MW even the strong BC stations were only few dB above the noise level. I measured up to 30 dB difference between HP and VP modes on BC station carrier on 900 KHz , on 1323 KHz the difference dropped to 18 dB. But above 10 MHz the H-loop was an excellent performer! First of all, in HP mode the band noise dropped down with 1 to 3 dB. The signal level increased with 5 to 7 dB giving the overall S/N improvement up to 10 dB compared to V-loop mode. At 21 and 28 MHz the H- loop S/N ratio was 6 to 10 dB better than the V- loop! On 7 MHz the signal level with V- loop was higher but with some stations H-loop gives better S/N. It might be possible that HP wave component in urban environment has less attenuation compared to VP component but this is just a guess not experimentally proven fact. The vertical concrete walls of the buildings act as reflection surfaces and the interference patterns might be quite strange.
The good news is that the small active HP antennas are very good performers for users that live at high urban buildings and if they occupy higher floors they will probably have very good sensitivity even on MW band.
My urban location is electromagnetically very polluted. I have a lot of interference from different noise sources. The VP and HP loop have different sensitivity to local noise sources as expected but I can not say that the one is better than another at my specific location. Some sources are strong with HP but very weak with VP and vice versa. I have a neighbor with legal limit PA who is quite strong – but in HP mode his signal strength dropped down with 10 to 15 dB which means that signal is predominantly with VP. As a results of these experiments I now keep this dual polarization loop setup as permanent. Before I used a pair of vertically orthogonal loops.
Fig.7 Dual polarization setup with a single AAA-1 amplifier. Both CC loops are identical, only the placement is mutually orthogonal. Horizontal loop has pure HP, the vertical loop has mixed polarization but pure VP in the direction normal to the building’s face.
Horizontally placed small active receiving antennas give very good results if they are placed at least 0.1 wl above the ground. It is possible to place two very small antennas on the same mast and to have a possibility to choose the optimal polarization. The best pair for dual polarization reception is a horizontal crossed coplanar (CC) loop and a vertical dipole. These two antennas have pure H and V polarizations even above real ground and are omnidirectional with a zero toward the zenith. Another good pair are V-loop and H-loop especially for urban environment since the vertical dipole is very noisy there.
A1. Antenna polarization coordinate system
We are using plots of VP and HP components and it must be explained what is the meaning of these plots to avoid any misinterpretations. The propagating electromagnetic wave is characterized by the direction vector and the polarization plane - there are a lot of textbooks and publications about that. In polarization plane lay both vectors - the electric field vector and the wave direction vector. Bear in mind that the electric filed vector E is always normal to the wave direction vector (electric and magnetic vectors are transversal to the direction of wave travel). The simplest case is the plane wave, where the polarization plane is fixed.
The terms Horizontal (HP) and Vertical (VP) polarizations are connected to the coordinate system locked to the ground surface (horizontal plane) and with an observer located at its center point. I will explain them exactly as they are calculated in antenna CAD software (usually denoted as H and V components) since their names are somewhat misleading especially the VP wave. Let us have xyz coordinate system where xy axes define the horizontal plane. The plane of incidence of a wave is defined as a vertical plane (normal to horizontal plane) where the wave direction vector lies. This plane is passing through the center of the coordinate system (Fig.8).
We have VP if the electric field vector E lies only in a vertical plane (in the plane of incidence). Do not mix the xyz coordinates of the E vector with wave polarization! The polarization plane of VP waves is always normal to horizontal plane but that not means that the E vector is normal to the horizontal plane. An example: Let us have a VP wave coming from 80 deg elevation angle. E is normal to wave direction vector and lies in the plane of incident thus it has an angle of only 10 deg. to horizontal plane. Fig.8
If E is normal to the plane of incidence we have HP. For HP waves E vector is always parallel to horizontal plane and it does not have z component in xyz coordinates.
An arbitrary plane wave has tilted polarization. Any tilted plane wave can be represented as it consist of two independent HP (Eh) and VP (Ev) waves with the same wave direction vector (same plane of incidence). The vector sum of Ev and Eh gives the actual E vector. Most of the antennas used in HF spectrum radiate plane waves. These waves after the reflection from the ionosphere continuously change their polarization usually in random way (even a full rotation might occur) but this change is slow and we will assume them plane waves. We will not assume them circularly (or elliptically) polarized – this term will be used only for waves where the frequency of rotation of the polarization plane is equal to the frequency of the signal. (But there are cases where after the reflection of plane wave the polarization can be changed to elliptical [20, 21]) .
Fig.8 The antenna coordinate system. The polarization of a wave is determined always to the reference plane of incident. For VP E lies in the plane of incident and for HP E is normal to the plane of incident. Any other polarization can be represented as it consist of two independent HP and VP waves with Eh and Ev electric field vectors with the same wave direction. E vector is also always normal to wave direction vector.
The concept of wave polarization can be applied to the antennas. As with the waves, we can assume that antenna radiates two independent orthogonal HP and VP components of E with the same wave direction vector. They are named H and V in CAD antenna software and it is possible to plot these two components separately. Any antenna radiates in all directions but H and V plots give the exact polarization of the radiated wave for each direction vector defined by azimuth and elevation angles. The azimuth angle defines the plane of incident and the elevation angle defines the wave direction vector in this plane. The vector sum of H and V components gives the actual E vector. The usual “total” radiation pattern which is the module of the vector sum of H and V components does not give any information about the polarization but just for the magnitude of E. Most of the HF antennas have tilted polarization patterns and wave polarization is different for each direction - the tilt of the polarization plane depends from the azimuth and elevation direction.
If an antenna radiates only HP waves then as receiving antenna it is sensitive only to HP components of the incoming wave and we said that this is HP antenna. The same holds true for the VP antennas. This is very important - looking at the H and V plots we can estimate what is the sensitivity of the receiving antenna to the wave polarization coming from different directions.
A2. Small antenna polarization patterns above real ground
The next models are computed with NEC2 engine (Nec2 for MMANA  ) or Mininec (MMANA) antenna CAD software. For all models the HP antennas are placed at 12 m above the ground level with Somerfield/Norton ground model (when NEC2 is used) , Eps.=13, sigma=5 . The center of VP antennas is placed 3 m above the ground. The VP components are always marked with red color. Different loop shapes are used in models because I used pictures from older investigations, but the loop shape does not matter at all.
Vertically mounted loop
The vertically mounted small loops above real ground have mixed polarization as shown on Fig.9 The polarization plane varies with the azimuth and elevation directions. It has stronger VP component but the HP can not be neglected. For waves coming form x direction this antenna has VP. For the waves coming from y directions it is HP. This explains why there is not much sense to rotate the small loop for the cases where the polarization is not defined e.g. for the signals reflected from the ionosphere. The small loop has figure 8 pattern only for pure VP or HP waves. Note that the loop has very good sensitivity to VP waves which are coming at low angle elevations. That means that this loop will work near the ground since there are only VP waves traveling almost parallel to the ground surface.
Fig.9 H+V patterns of vertically mounted small crossed coplanar loop 12m above the ground. The model is for 3.5 MHz .
Horizontally mounted small dipole
The H-dipole is very similar to V-loop just the H and V components have changed their places. For some azimuth directions these two antennas have the same polarization so they are not very suitable for dual polarization reception.
Fig.10 Horizontal dipole with 1 m arms. This antenna has mixed polarization which varies with azimuth direction and its radiation maximum is at the zenith because at 3.5 MHz 12 m height is too small. Note that, at 30 deg. azimuth, the vertical and horizontal components are equal thus we have tilted to 45 deg. polarization at this direction. For the 90 deg. azimuth direction we have pure HP and for 0 deg. – pure VP.
The next antennas have pure HP and VP above real ground in very wide frequency range. They are ideal pair for dual polarization reception since their mutual polarization is orthogonal for any azimuth and elevation angles. Both have circular azimuth radiation pattern which makes them universal receiving couple.
Horizontally mounted small loop - pure omni-directional HP sensor
Horizontally mounted small receiving loop is insensitive to VP wave. The magnetic vector H of VP wave is always horizontal irrespective of the azimuth and elevation of the incoming wave and thus, according to Faraday low, the induced voltage in loop is zero. It is omnidirectional in azimuth for HP waves but has a zero for HP waves coming from the zenith since their H-vector becomes parallel to the loop plane. As expected the radiation (transmitting antenna) patterns have the same polarization properties. Fig.11
Horizontally mounted single small loop has almost pure HP if the loop perimeter is smaller than 0.1 wavelengths. For larger perimeters VP component appears. A modification of the small loop is the so called crossed coplanar (CC) loop - a parallel connection of two or more loops where the loops are coplanar (lay in one plane). The horizontally mounted CC loops have almost pure HP in much wider frequency range (Fig.11, Fig.12 ). This configuration is preferred in wideband loop designs .
Fig.11 Two horizontal crossed coplanar rhomb shaped loops with 1 m2 total area (MMANA). CC type loop is the “perfect” horizontally polarized antenna since the vertical component is eliminated. This holds true for very wideband frequency range – see the next figures.
H= 0.075 wl H= 0.15 wl H= 0.3 wl
H= 0.6 wl H= 0.9 wl H= 1.2 wl
Fig.12 H+V patterns of horizontal CC loop 12 m above the ground for different frequencies (NEC2). The height in terms of wavelength is given also. For all frequencies the azimuth pattern is circular and is not given to save space. There is no vertical component for these frequencies.
Vertical small dipole - pure omni-directional VP sensor
The vertical small receiving dipole has pure VP in wide frequency range as shown on Fig.13. It is insensitive to HP wave. The E- vector of a HP wave is always horizontal irrespective of the azimuth and elevation of the incoming wave and thus it does not induce voltage at the dipole terminals. It is omnidirectional in azimuth for VP waves but has a zero for VP waves coming from the zenith since their E-vector becomes normal to the dipole.
Fig.13 H+V patterns of vertical dipole 2 x 2.5 m, 12 m above the ground for different frequencies. For all frequencies the azimuth pattern is circular and is not given to save space. There is no horizontal component.
Phased arrays with HP small loop antennas
Phase arrays with VP elements are usually used in low-bands (1.8 and 3.5 MHz). But we can build also a phased array with HP elements. There is no difference in phasing method and time delay settings. The radiation pattern of a phased array with two horizontal CC loops is plotted on Fig.14 The distance between elements is 12.8 m. This is a subtractive (endfire) array  and the time delay is set to optimal = 43 ns . The vertical component is negligible in wide frequency range (Fig.15 ). It will be interesting to build such an array in order to test the HP directive antenna especially on low bands. To be useful, the height of the elements should be at least 0.1 wl and the distance between elements should be above 0.2 wl in order to increase the signal levels. I do not know whether the HP will lead to improvement in reception in certain conditions but it might be quite possible. [16, 17]
Fig.14 H+V pattern of 2- element subtractive phased array with horizontal CC small loops. The height is 12 m above real ground at 3.5 MHz.
Fig.15 H+V patterns of 2- element phased array with horizontal CC loops at different frequencies 12 m above the ground. The delay between elements is optimal . Above 10 MHz the unidirectional pattern is degraded and is not given. There is no vertical component.
Here is an example of 4-square with horizontal loops with pure HP (Fig.16). Its radiation pattern is similar to a 4-square with vertical dipoles which is pure VP antenna.
Fig.16 A 4-square array with horizontally mounted small CC loops. The size of the square diagonal is 12.8 m. The height is 12 m at 3.5 MHz. The delay between elements is optimal . The azimuth plane is at 20 deg elevation. The vertical component is negligible.
A3. Universal dual polarization antenna
Universal dual polarization antenna was build in order to test all possible polarization placements of small dipole and loop antennas. The antenna consists of H-loop and V-loop in CC connection placed on the same mast Fig.17 . The loops act also as arms of small dipole in dipole mode of AAA-1 amplifier.
Fig. 17 Fig.18
Fig.17 Universal dual polarization antenna on a single mast. It is better to turn the horizontal part at 90 degrees than that shown on the picture in order to have the same direction of H –dipole and V-loop antennas pair (some other reasons due to location restriction are taken into account for this specific construction).
To increase the dipole sensitivity small whips with 80 cm length from 3mm diam. steel material are electrically connected to the external parts of the loops Fig.19. These flexible metal whips do not influence the loop performance but increase the dipoles length. The length of each dipole arm is therefore equal to 2.07 m. The loops were made from 16 mm PE coated aluminum tubes (for heating installations). The loop diameter is 1.27 m ( 4 m perimeter). The mast is made from wood and the center of the H-loop is at 7 m height. HP and VP antennas are connected to two AAA-1B amplifiers in order to have the possibility for diversity reception.
The protective earth of the amplifier is connected to the FTP shield inside the amplifier box Fig. 20 . Common mode chokes  are used for the FTP cable of each amplifier Fig. 18. The FTP shield between the choke and the antenna is grounded to copper rod inserted into the soil.
 Levkov.Ch. LZ1AQ, Wideband Active Small Magnetic Loop Antenna, 2011, http://www.lz1aq.signacor.com/docs/wsml/wideband-active-sm-loop-antenna.htm
 Levkov.Ch. LZ1AQ, Experimental Comparison of Small Wideband Magnetic Loops, 2013, http://www.lz1aq.signacor.com/docs/experimental-comparison-v10.pdf
 Leif Asbrink, SM5BSZ, Linrad , http://www.sm5bsz.com/linuxdsp/linrad.htm
 Kazimierz Siwiak, KE4PT An Optimum Height for an Elevated HF Antenna, 2011 http://www.arrl.org/files/file/QEX_Next_Issue/May-Jun_2011/QEX_5_11_Siwiak.pdf
 Kazimierz Siwiak, KE4PT, Pitching and Catching Radio Waves - The Last Bounce , The QRP Quarterly, 2013, pp 31 -32 ,
 Java applets, http://www.met.reading.ac.uk/clouds/maxwell/offset_dish.html
 Levkov.Ch. LZ1AQ, Diversity Reception with a Modified SDR Receiver and Small Active Antennas, 2015, http://www.lz1aq.signacor.com/
 Antenna diversity, http://en.wikipedia.org/wiki/Antenna_diversity
 NEC2 for MMANA, http://www.qsl.net/ua3avr/
 MMANA-GAL http://hamsoft.ca/pages/mmana-gal.php
 Levkov.Ch. LZ1AQ, Receiving Phased Array with Small Electric or Magnetic Active Wideband Elements. Experimental Performance Evaluation, 2014, http://www.lz1aq.signacor.com/docs/phased-array/2-ele_phased_array11.pdf
 Variable Delay Line Kit for Active Antenna Phased Arrays. Model VDL-1, http://active-antenna.eu/variable-delay-line-kit-vdl-1a/technical-documentation/
 Jose Carlos, N4IS, High performance RX antenna for a small lot, 2014, http://www.qsl.net/k4fk/presentations/WF-receiver-antennas-SFDXAs.pdf
 Carl Luetzelschwab, K9LA, Polarization Observations on 160-Meters, 2011, http://k9la.us/Polarization_Observations_on_160-Meters.pdf
 Tom Rauch, W8JI, http://www.w8ji.com/polarization_and_diversity.htm
[19 ] Forum discussion digest, http://www.dxmaps.com/discuss/polarization.html
 Levkov Ch. LZ1AQ, Reducing the Noise in Dipole Mode with Common Mode Filter , 2013, http://www.active-antenna.eu/tech-docs/comm-filter-ftp-10.pdf
 Eric Nichols KL7AJ , Gimme an X, Gimme an O, QST December 2010
 Jan Simons PA0SIM, Circular and elliptical polarisation and propagation on shortwave, http://www.pa0sim.nl/XOpropagation.htm
1.0 24 July 2015 Initial publication.
1.1 28 Sep 2015 Some additional description for circularly polarized ionospheric waves and additional links.