The "Cizirf-Special" Receiving Antenna

A 40/80m receiving antenna optimized for short and medium distances

Patrick Destrem F6IRF




In a recent article, I presented a receiving antenna concept for short distance communications on the 40 and 80m bands using 2 broadband dipoles placed at low height and fed in various phase relationships(*). This antenna is not a DX antenna, but an antenna whose purpose is to facilitate "domestic" communications, which in these times of minimum solar flux are often more difficult to achieve than DX contacts. This is especially true on the 40m band. This antenna may also be interesting for those who are affected by one or more sources of local noise that typically arrive at low angles such as urban neon signs, transformers or power lines.


The following is the complete description of this antenna with various available options. Although this antenna was initially designed with our “national contest” in mind, it may also be interesting for other applications.


(*)  (original in French only so far. However you may try an automatic translator or just look at the pictures!)


NB: If you have never used a separate receiving antenna, please note that a basic precaution consists in short-circuiting the input of the receiver (for example, by means of a coaxial relay) during transmission.  The RF "collected" by the receiving antenna can reach dangerous power levels for the receiver.  Even in the case of a transceiver having an " RX-ANT " input, it is recommended that you make sure from the schematic that this input is disconnected or shorted-circuited before the first stage of the receiver (this is not always the case...). Also, the receiving antenna should be located as far as possible from the transmitting antenna.


The broadband dipole

Figure 1 the broadband dipole. As you can read it in the text, it is better for the dipole to be non resonant.

Why use a broadband dipole of twice15m? The first reason is to maintain an acceptable SWR on the transmission line. Typically with this dipole, the SWR should hardly exceed 1.5:1 on 40 and 80m, whatever its height (from 1 to 10m).

W8JI, in his article explained the reasons why low-Q antennas should be preferred for this kind of receiving systems. Please refer to his article 


Figure 2  On top SWR versus wire-length (in meters) for the fig1 dipole. At the bottom Gain versus height (in meters) . In all the cases it is better to avoid resonance...

The minimum length of the elements (2x 15m) is imposed by the gain-limit of -20dBi, below which receiving antennas are likely to require preamplification. The length of wire is not very critical. However, it is better to avoid half-wave resonance, as the SWR may reach about 4:1, due to the low radiation impedance versus the loading resistance.  Because of its low-Q, this antenna should not disturb the other antennas located nearby (*). For short/medium distances, it is not desirable to make the dipole too long, because with length, the directivity increases, reaching a maximum at 2 x 5/8 λ, beyond which the secondary lobes become dominant. In short, 2 x 15m seems a good compromise for 80 and 40m bands. For more “broadband” applications, typically 3 to 10Mhz (i.e. SWL’s) I would recommend to use a terminated folded dipole (see W4RNL pages) which maintains a low SWR over a wider bandwidth.


(*) the opposite may not be true: As mentioned by W8JI in the above mentioned article, a transmitting antenna may affect greatly the performances of your receiving antenna system. I did experience this myself: I noticed quite a difference in levels and noise between my East antenna and my West antenna… until I detuned my transmitting vertical!


Figure 3: Vertical gain and SWR of the broadband dipole on 40 and 80m versus height in meters for an average ground (NEC2). For a receiving antenna the gain is not a determining factor, but a minimum is required.


The antenna

Figure 4 the antenna consists of 2 broadband dipoles, such as those of figure 1.

The antenna is made with  two  the broadband dipoles like the one pictured  in Figure 1. Optimum spacing depends on what is desired. If one wishes an extreme NVIS antenna, a half-wave is a good spacing. If one prefers a directional antenna for the intermediate vertical angles, a more reduced spacing will be preferable. For use on 40 and 80m it is a matter of compromise, but around fifteen meters of spacing between dipoles appears to be a good choice. However, this is not very critical.



Figure 5 vertical gain and gain at 45 degrees elevation, versus dipole spacing, for the two Figure 1 dipoles fed in phase at 7m AGL (NEC2).

Figure 6: Vertical pattern for spacings of  0.2, 0.5, 0.6 and 0.8 wavelengths (MMANA). A half-wave (in red) is good spacing for a NVIS antenna  without a low lobe on the horizon. However 0.6wl(in blue)  provides nearly 10dB more attenuation of the signals arriving at 40 degrees.  


Feeding the dipoles

For the simple NVIS version (Figure 6), the simplest solution is to mount the dipoles parallel to each other and use identical feeder lengths to a broadband matching device (25/50 ohms) such as a UNUN. For the “high angle” directional pattern option (Figures 7 and 8) and if the goal is a particular distance (1) or direction, it is possible to use dissimilar lengths of cables to the combining device. If the dipoles are installed in phase (2), it is the feed line of the dipole nearest the desired direction which should be made longer (see Figure 7). If the dipoles are mounted with reversed phase it is the feeding line of the dipole opposite to the desired direction which should be made longer (3).

If you want to reverse the desired direction, I have described on my blog a simple solution, using a single coaxial relay and a UNUN. The article is in French, but the diagram is quite obvious. This simple solution assumes that the 2 dipoles are absolutely identical, and that the SWR of each one is low (you may need to adjust the loading resistors and/or the wire lengths). With this simple solution the main problem remains the mutual coupling between the 2 dipoles fed in quadrature (or close to it). However due to the low Q of the dipoles, the degradation of the patterns compared to the simulation should be acceptable.

Ideally a hybrid coupler should be use to combine the 2 antennas when they are not fed in phase or 180 degrees out of phase, but such couplers are generally designed for one band (like the Comtek PVS-2). You may also try something like the DH-P or the BB2 power splitter, used as a combiner. Anyway for receiving, a more flexible solution is provided by phasing boxes such as such as the MFJ-1025 or the DX-engineering NCC-1 (see tests below). This said, even a very simple solution just using switched coaxial cable lengths  and a quickly made UNUN can sometimes provide very surprising results, as demonstrated in this 80m recording.  


(1) You may have a look at the relation vertical-angle/distance in this article (original in French but the diagrams are self-explanatory)

(2) The two dipoles, with equal feed line lengths may be fed either in phase or 180 degrees out of phase, simply by reversing their connection to the feed line - one dipole's left side to the center conductor, the other dipole's right side connected to it.

(3) For feeding out of phase by more than 90 degrees this solution requires less cable. Demonstration: starting with 2 dipoles A and B installed 180 degrees out of phase, adding 1/8wl to the feed line of the dipole A, will create a phase relation of 180+45 = 225 degrees (in other words -135 degrees), and the directivity will be A towards B (Our dipole A occupying the same situation as the reflector of a HB9CV). If the 2 dipoles are in phase, it will be necessary to add 3/8wl in antenna B feed line to achieve the same result.


Simulated performances on 40 and 80m

The diagrams below were generated with MMANA (and verified with NEC2) using a spacing of 15 meters and a height of 7m. As can be seen in Figure 3, there is some gain reduction by reducing the height.  However the absolute gain is not a determining factor for a receiving antenna and overall, the real performances should be close, even with a height as low as 1m.


Figure 7 : 40M patterns for various phase differences between the dipoles. All the horizontal patterns are generated for 45 degrees elevation. For the 2 dipoles in phase the maximum Gain given by NEC2 is minus 2.85dBi and the SWR 1.42. For the 2 dipoles fed in quadrature the max gain is minus 3.69dBi and the SWR for each dipole is 1.4 and 1.47. For 15m spacing, the maximum F/B ratio is obtained for a phase-relationship of about 110 degrees (green curve). If the low-lobes on the horizon are a problem, a phase relationship of 70 to 90 degrees will also provide excellent results


Figure 8 : The same configuration, but on the 80m band. As you can see, the gain decreases notably when the phase difference is increased. The maximum gain given by NEC2 is -10.28dBi (SWR 1.44) for the dipoles in phase and of -15.65dBi (SWR 1.47/1.45) for 140 degrees phase relationship. For 180 degrees, the maximum gain goes down to -20.83dBi. For 120 degrees the diagram is excellent with no secondary lobe near the horizon.


As can be seen in Figure 7, the gain variation is minimal on the 40m band. At 45 degrees elevation, it is the same for the phased-dipoles and the anti-phased dipoles. That is the ideal case!

On the 80m band the gain varies more. However the antenna remains usable with good performances.

Of course, it goes without saying that continuing to turn around the phase-circle rotates the diagram by 180 degrees in azimuth.

Note:: It is not as simple as you may imagine: if you consider the two dipoles fed in quadrature (90 degrees out of phase), adding 180 degrees will provide the " mirror " diagram...  it is not true any more if you start at 135 degrees (135+180 = 315), because in this case the phase difference between dipoles will only be 45 degrees.


Performances on adjacent bands

The performances on the adjacent bands is interesting, as it gives us an idea of what could be done to improve the design (with regard to spacing and element lengths).


Figure 9 Patterns on 30m. It is quite obvious that, if an NVIS antenna is the choice, a half-wave is a good spacing between dipoles ( 0.6wl is even slightly better). On the other hand as soon as the phase relation is changed, a back lobe appears rather low on the horizon. This goes against the desired goal. It can also be seen that the horizontal pattern (at 45 degrees elevation) becomes resolutely oval.  However the antenna remains usable for short-range operation.

Figure 10 Patterns on 160m. The antenna becomes difficult to use due to the insufficient gain 11.



I built the baluns on high-AL ferrite beads, initially intended for choking RG58 or similar cables (AL index around 1000 uH/100 turns). With three bifilar turns the SWR is in conformity with calculations (1.2 to 1.3 from 1Mhz to 30Mhz). The actual construction is far from being critical; even the “random choice" of resistors does not seem to affect the results at the considered frequencies.


Figure 11 The two baluns are virtually identical. The resistors used for the tests are two 510 Ohm 1/4W in parallel. It will replace by 3- 820 ohm 3W resistors when available because these are likely to burn if the antenna is in close proximity to the transmitting antenna.

Figure 12 Checking of the balun/resistor assembly. SWR 1.2 to 1.3 up to 30MHz... Note my high-class network analyzer!

Figure 13 Mounting of  the SO239 coax jack.

Figure 14  The assembly is made in a small electrical box. . Like the Balun, the dipole is constructed with a black and a white wire (of 0.5mm diameter- not critical ) to allow easy identification of the phase.

Figure 15 Here, the antenna is installed on a fiberglas mast  supported by a tripod. The dipole apex is approximately 4m AGL and the ends approximately 2m AGL.


Tests with one dipole

With reference to my vertical, the results with only one dipole are already interesting. With the 2 antennas levels balanced on the background noise, the gain in terms of SNR(*) is spectacular, frequently 10 to 15dB on the stations located at less than 500 kms. Only the most remote stations arrive stronger on the vertical. The improvement is such that certain stations, completely inaudible on the vertical, appear above the noise when switching on the low dipole. Admittedly it is necessary to mitigate this result by the fact that the comparison is made with a vertical. Compared to a relatively low dipole (< 0.5wl), it is likely that there will be no difference. On the other hand, on 40m, compared to high yagi or a dipole (>.0.5wl AGL) I think that the results will be similar to this 40m recording (the HB9 station is about 300 kms away, TM8P is around 400 kms away)

With this version using two 15m wire lengths, the level is sufficient on the 40 and 80m bands, not even requiring the use of the transceiver preamplifier. On 160m, it is necessary to use the preamp (for the IC756pro2, preamp position 2 which is approximately + 20dB gain). In spite of the very negative gain of the antenna on this band, the results are also still spectacular (the station on this audio-clip is located about150kms away). The local stations emerged from the noise as if by magic.


(*) Signal to Noise Ratio


Tests with the two phased dipoles

Making a point of testing a dedicated NVIS antenna, I initially installed the 2 dipoles (still 4m AGL) with 20m of spacing.  The two antennas were connected by equal cable lengths to the 2 ports of a "stackmatch" (to allow quick comparison between a single dipole and two phased ones). On 40m the first immediate result was a reduction of the background noise from a 380 kV power line by about 7 dB (compared to a single dipole). Of course on 80m the difference is tiny (approximately 2dB), because the 2 antennas are too close for this purpose. On 40m monitoring nearby stations (100 to 300 kms), I noted about 20 dB improvement of the SNR compared to the vertical, partly due to the reduction of the noise floor. The improvement of the SNR with the 2 phased dipoles is also very clear when compared to a single dipole.


Figure 16 One of the NVIS dipoles at 1m AGL. In the background is the vertical, used as reference antenna.

Interested in testing the limits of this antenna system, I then mounted the 2 antennas at only 1 meter AGL, with 25m spacing (0.6wl on 40m). Although this configuration resulted in having to use the transceiver preamp, it is clear that the results were still there. With the two phased dipoles, the results on 40 and 80m were surprising. Up to 30dB improvement of the SNR (with reference to my vertical) when monitoring local stations on the 80m band was achieved. As it is logical, the difference was reduced as the distance to the transmitting station increased. However on 80, as on 40m, the difference was still some 10 to 15 dB in favour of NVIS antenna for stations located between 400 and 600 kms away.

You may listen to the various recordings made with this configuration, but the most interesting is probably this one because it shows how it is possible to "sort" between 2 stations, located in the same direction but at different distances, when switching from one antenna to the other. Note, though, that the reference antenna is a vertical, antenna which is particularly unsuited for domestic traffic (on the scale of a relatively small country like France!).


Figure 17  Audiogram obtained on a local station (40kms) when switching from the NVIS antenna to the vertical. The receiver AGC was turned off and the two antennas balanced in level with respect to the background noise. As can be seen, the difference can reach up to thirty dB when switching from the vertical to the NVIS antenna.


Playing with the Phase

The idea of varying the phase between two close spaced dipoles is not new. In 1937, Dr. John Kraus, W8JK, described the principle of bidirectional antenna using 2 close-spaced dipoles fed with 180 degrees phasing. Antennas like the HB9CV, F8DR and ZL-special use a similar principle (2 close spaced dipoles - typically 1/8 wavelength), except that in this case the dipoles are not “anti-phased” anymore but fed with phasing of about 135 degrees, which makes the pattern more or less unidirectional, but limits the use to one frequency band.

If you imagine being able to vary the phase relationship between the 2 dipoles from 0 to 359 degrees, while controlling the amplitude of the currents in the 2 elements, it becomes possible to obtain a complete palette of patterns going from the phased-dipoles to the W8JK-array (Figure 7 and following). On the transmitting side, the control of the currents flowing into the elements is the most delicate part, which results in using complex feeding systems such as those used for the 4-square arrays (see " Low-band DXíng " from ON4UN). On the other hand, for receiving purposes, using resistively loaded dipoles tends to minimize the impedance variations (and therefore the current imbalance) to something more acceptable. Ultimately if it becomes possible to “fine-tune” the phase and amplitude of the 2 dipoles, all the diagrams envisaged by the models become achievable.

A simple way to vary the phase is by lengthening one of the feed-lines. By using switched sections of one feed line (see blog *) it is possible to obtain front-to-back ratios in the order of 15 dB (or more if you are lucky!). In practice if we are able to vary the phase from 0 to 180 degrees per 45 degrees steps, the missing 180 degrees sector can be obtained using a 180-degree broadband transformer. A single band system does not require more than 2 line sections. A section of a eighth of the wavelength results in a 45 degrees rotation and a quarter-wave section results in a 90 degrees phase rotation. With 3 line sections and a transformer we can easily build a system working on 40 and 80 meters (figure 18).


(*) automatic translation from French provided by Babelfish (don’t ask too much!)


Figure 18 Draft of a "rustic" 80m system, to be inserted in one of coaxial lines. It allows phase-shifts from 0 to 337.5 degrees in 22.5 degrees steps.  It is also usable on 40m, but the step becomes 45 degrees (on 40m SW2 becomes useless since it has the same effect as SW1). All the lines are cut on 80m, taking into account the cable velocity factor.


The limitation of the above “simple system" and one of the reason for reduced performance (compared to simulations), is that it is not possible to balance the amplitude of the signals arriving at the coupling device. Even if the antennas are virtually identical, it is possible to note rather important differences in the amplitude of the signals arriving from one antenna and the other. This difference can be caused by the terrain configuration, surrounding obstacles, other antennas present on the site or even by the losses introduced by the variable-phase system.  In short, to achieve the results envisaged by the simulations it would be necessary to add a step attenuator (or better a variable gain amplifier) in each line.

Electronics offers an alternative to us. Without going into the details, an "active" device allowing continuous phase and amplitude control should produce the anticipated results. To my knowledge there are currently on the “amateur” market " 2 magic boxes” allowing such a wonder: The MFJ-1025 (less than $200 in the USA) and the NCC1 from DX Engineering (a little less than $500 in the USA).

The MFJ-1025, which I tested personally, suffers from a certain number of weaknesses:

- It allows only a phase variation from 0 to 130 and 180 to 310 degrees, which is sometimes insufficient. It had to add an external "switchable" piece of line to cover the two missing 50 degree sections.

- It has negative gain (again, see W8JI page dedicated to this device)

- Its noise figure does not allow the use low output antennas (I had to raise my two broadband dipoles from 1 to 7m AGL, to get a sufficient level to cover the noise introduced by the box)

- Adjusting the phasing potentiometer requires precision, and over time repeating such precise adjustments becomes tiring as the potentiometers are not smooth enough.

This known… it is inexpensive and it does work, as attested by the videos and audio recordings published on the blog. Typically 20 to 30dB of rejection of an unwanted signal are made possible (of course the desired signal must come from a direction and/or a vertical angle different from the useful signal).



Figure 19: One of the "broad band" dipoles used for the tests of the MFJ-1025. It was necessary to raise the dipoles height from 1 to 7m to obtain a sufficient level to cover the noise generated by the device.  By the way, you can see in the background the power line passing approximately 200m away from my aerial which makes DXing on the low bands so difficult.

Figure 20 MFJ-1025 in test with F6IRF

The NCC-1, tested on the "Cizirf-special" antenna at F6KNB radio club, seems much easier to use than the MFJ-1025. Moreover, it includes two 20dB preamplifiers resistant to strong signals.  It covers 0 to 360 degrees without any hole and with an easy to use “phase button”. With this device it should be possible to use the broadband dipoles as low as 1m AGL.


Note: In a high power environment (i.e. a contest station) it is highly advisable to check if the protection is sufficient for your phase-box inputs..


Figure 21 The NCC-1 under test at F6KNB.  F6CIS, F5JZA and F6IRA are pictured. 


Summary and conclusions

My initial goal was to design a receiving antenna to improve the quality of short and medium communications on 40 and 80m: Mission accomplished!!


According to your personal requirements, you may choose from the options described above.


- If NVIS is your main requirement, 2 phased dipoles with 0.6wl spacing on the highest band (i.e. 25m on 40m), is the best option. Assuming your receiver has sufficient gain capability, the height of the 2 dipoles can be reduced as low as one meter. It is also the best option for those who live in a noisy environment (In this last case, reducing the spacing to a half-wave will eliminate the secondary lobes, low on the horizon). For this option a simple UNUN will do a perfect job.

- For those who want a longer range antenna offering some front to back ratio, the spacing can be reduced to fifteen meters and the dipole in the preferred direction, fed with a "fixed" phase delay ranging from 70 to 140 degrees (see Figures 7 and 8). For this option and though a simple UNUN may do the job, I would recommend a combining device providing some isolation, such as the BB2 . 

- If one wishes an antenna offering a more extended range of possibilities (choice of direction and vertical angle) and typically a ten to twenty dB's of front to back ratio, then I would recommend fifteen meters spacing and something similar to the simple phasing system of Figure 18 to be inserted in one of the feed line after a combining device.

- Ultimately, if the very best performance is desired (typically 20 to 30dB rejection of unwanted signals), you will have to use a box such as the MFJ-1025, or even better the NCC-1 (or design and build a phasing box ).


Finally your individual use should be considered if an adjustable device is used.

If you are a “ragchewer”, no problem.  You have all the time necessary to adjust the rejection of undesirable signals.

On the other hand, in a contest situation, the QSO will be already finished before you finally find the optimum adjustment. Personally, for contest use, I would appreciate having a few “presets” on the phasing-box to make it really useful.  This said, it may still be very useful over longer periods to reduce or eliminate annoying key-clicks or splatters from a station too close in frequency (our bands are not wide enough, especially during SSB contests!). However nothing yet replaces the individual operator’s skill to pull out weak signals from the noise and QRM.




73' S - Patrick



A summary of all the tests carried out with this antenna is available by following this link . Videos and audio recordings, are used to illustrate the topic.


Copyright F6IRF January 2008. Thank you for contacting me for any reproduction even partial of this document.



Main references

The W8JI pages: especially the receiving pages

The W4RNL pages: 

“Low band DX’ing” book by ON4UN

“Transmission line transformers” book by W2FMI  Rolling Your Own: building antenna splitters that perform better than most commercial units by John Bryant and Bill Bowers





4NEC2 by Arie Voors


Audacity  “a free digital audio editor”


Acknowledgments: I would like to warmly thank all who have actively contributed in a way or another to this publication, namely F4AJS, F6CIS, F6IRA, the F6KNB team, N4ZR, W7ZR and many others who demonstrated their interest or prefer to remain “anonymous”.