Phasing Improves Kaz Antenna Nulls
Mark Connelly, WA1ION - 12 JUL 2001 (links updated 14 JUN 2013)

The "Kaz" antenna was introduced by John Bryant's article "Testing Two 'Kaz' Squashed Delta Antennas" (reference 1). Further test results were presented in my article "Pennant and Kaz Antenna Tests" (reference 2). The antenna has the form of a delta with the apex spaced above the center of the horizontal base at a distance of 1/4 to 1/3 the length of the base. Base height above ground can be as little as 0.3 m / 1 ft., though I find improved performance at 1.5 m or higher. The top wire (forming the two sloping sides) is one conductor and the base wire is a second conductor. Feeding and/or termination occurs at the pair of wires on each end of the base at the point of approach of the upper conductor.

This article describes how the nulling abilities of the Kaz antenna can be enhanced by having combination feed-and-termination boxes installed at each end of the antenna. The two coaxial feedlines are presented to inputs of a phasing unit.

Homebrew phasers such as DXP-2, DXP-3, Superphaser-1, and Superphaser-2 work well for this application. Articles on these units can be obtained via links on my RF Circuits page (reference 3). Commercially-available broadband phasing units such as the Quantum Phaser, modified MFJ-1026, and JPS ANC-4 could also be used.

Each feedline supplies a pickup that is somewhat cardioid in shape. If the Kaz antenna is set up on an east-west axis, the feedline coming from the box at the west end of the antenna has a pattern which, to some extent, nulls signals from the east (+/- 30 degrees typical). The feedline coming from the box at the east end of the antenna has a pattern that tends to null signals from the west. Depending on antenna layout, height, surrounding conductive objects, and termination resistance, the maximum null of each cardioid can vary from as little as about 6 dB to as much as 40 dB. In many cases Vactrol control of termination can improve null depth over what can be had with a fixed termination value (typically chosen to be about 1000 +/- 200 ohms). Even if the terminations are fixed and cardioid front-to-backs are only coming in around 10 dB each, very deep nulls can still be obtained by phasing the two opposing direction cardioids against each other.

In the case of wanting to null a signal from the west with the above set-up, the feedline from the west end of the antenna may be presented to phasing unit Channel 1 and the feedline from the east end to Channel 2. Channel 1 has a western station signal about 10 (and maybe more) dB stronger than that same signal on Channel 2. Also, desired eastern signals are on the order of 10 dB or more weaker on Channel 1 than on Channel 2. When equalizing Channel 1's western pickup to be about equal with that from Channel 2, the level pot adjustment introduces about 10 dB of loss on Channel 1. At this point there is at least 20 dB of strength difference between the two channels on eastern signals. When the phase adjustment to null the western signal is enacted, there is virtually no effect on eastern signals even if they too are made to be 180 degrees out of phase between the two channels. The system's overall front-to-back ratio can be pushed to better than 50 dB on groundwave and 25 dB on most skip with this arrangement: superior to a single feedline approach even with Vactrol control.

The simplest way to do a combination feed and terminate at each end of the antenna is to connect the antenna wires through a step-down transformer matching 950 ohms to 50 ohms. John Bryant's article "Fabricating Impedance Transformers for Receiving Antennas" (reference 4) recommends an FT114-75 (FT114-J) Amidon ferrite toroidal core with a 20-turn high-impedance primary winding (antenna) and a 5-turn low-impedance secondary winding (coaxial feed). He also mentions the alternative of an FT114-43 core with a 45-turn high-impedance primary and a 10-turn low-impedance secondary. A Mini-Circuits T16-6T-X65 transformer has also been used successfully in some Kaz and Pennant antenna installations. If the "shack end" cable terminations (e.g. at the two phasing unit inputs) are reasonably close to 50 ohms, the correct termination impedance for each side of the antenna will be passed through the respective transformer (acting as a step-up in that direction).

Because adjustment of a deep null is accomplished by phasing, there is less need for Vactrol remote termination than with a single-feedline system. Two opposing cardioids of fairly mediocre null depth can be combined to produce impressive front-to-back ratios. This is quite the same as noted here with opposing-pickup slopers even when each antenna is only good for about 8 - 10 dB of front-to-back on its own. Also, as noted with slopers, a bit of spatial separation helps too. Phase shift on desired signals is less likely to be in the 180 degree null range executed on opposite-direction signals when feedpoints are separated by 1/15 to 1/3 wavelength (as compared to being co-located or at some multiple of 1/2 wavelength). As compared to a small 12 m base version, the larger size Kaz antennas (base in the 20 - 40 m / 65 - 130 ft. range) will deliver heftier signals in the first place and will also have the added benefit of a lower likelihood of collateral nulling of desired-direction signals along with null-direction "pests" when inherent cardioid null depths are below the ideal of at least 12 dB each.

The transformer feed / terminate scheme without additional amplification is usually adequate for Kaz antennas having areas of 50 square meters or above (e.g. base 20 m, apex 5 m above base). Smaller antennas can benefit from amplification. Since coaxial cable has very little loss below 2 MHz, MW DXers and 160-m hams will usually have a 50-ohm input / output amplifier on each coaxial feedine at the "shack" end, out of the weather. The W7IUV design (reference 5) is a good choice. The Kiwa broadband amplifier and the Mini-Circuits model ZHL-6A are ready-made, though pricey, options.

There is one instance in which amplification located at the two antenna feed points may be desired. This is the circumstance when Vactrol control of termination resistance is desired. Because DC (typically 12 volts) to power an amplifier and a separate Vactrol DC voltage both must be presented to each of the two combination feed / terminate boxes, a single wire (carrying Vactrol DC) must accompany the coaxial feed that carries DC to and RF from the amplifier at a given end of the antenna. This separate wire should be broken up with several RF chokes to keep it from influencing the antenna. It can be physically attached to its accompanying coaxial line by means of nylon cable ties. The amplifier should be a high-impedance input to 50-ohm output buffer type. My BUF-E and BUF-F models (links via reference 3) work well here. The Vactrol's variable resistance can be placed from the input of the buffer amplifier card on one side and, on the other side, to a 220 ohm resistor to circuit ground. The two antenna leads go to the primary of a custom 1:1 high-impedance transformer (to be described); the secondary of this transformer goes to the buffer amplifier card input and to circuit ground. This arrangement gives about 15 dB of gain compared to no amplifier; it also enables simultaneous Vactrol control of the null observed when the phasing unit is set to the channel corresponding to the output of the amplifier on the opposite side of the antenna. The custom 1000 ohm 1:1 balun transformer consists either of 21 turns primary / 21 turns secondary on an FT114-75 (FT114-J) core or, alternately, 45 turns primary / 45 turns secondary on an FT114-43 core.

The Vactrol controller should be a dual version incorporating aspects of the model presented in Figure 3 of my article "Pennant Antenna with Remote Termination Control" (reference 6). Chokes, dropping resistors, and diode protection of the LED portion of the Vactrol (in each feed / terminate box) should be configured similarly to Figure 5 of the Pennant article.

Having the independent Vactrol controls and "field site" buffer amplifiers at each end of the antenna adds a good deal of complexity compared to the simple transformer-feed method, but it is seen as a valuable approach to take on smaller Kaz antennas like the 10 X 40 ft. / 3 X 12 m model that is becoming popular for temporary installations.

You could use the buffer amplifiers without the Vactrol if desired. In that case you'd install a 1K fixed resistor or, better yet, a 2K pot across the input of each buffer. During installation each end's pot could be tweaked to null a target station in the middle of the frequency range. You'd "listen" to the output of the amplifier opposite the one where you were adjusting the pot. An adjusted value of 800 ohms to 1.2K would be the typical result.

Phasing is achieved by observing the normal operating procedure for the unit being used. This generally consists of, first, equalizing the amplitudes of Channel 1 and Channel 2 on the signal to be nulled and, secondly, adjusting the phase shift control to produce a null. The procedure concludes with small interactive adjustments of one or both amplitude pots and the phase shifter.

If feedline pickup is a problem, one or more coaxial chokes (consisting of 17 turns of RG-174 on an FT140A-J core) may be inserted in series with the coaxial line.

Anyone contemplating the use of a Kaz, Pennant, Flag, or Delta terminated loop should look into two-feedlines-to-phaser schemes similar to those outlined above. The enhanced nulling performance makes the slightly greater system complexity well worth the effort.


(Note: Over time Web URL's may change. If this occurs, it may still be possible to retrieve the articles by going to known DXer Web sites or to search engines for links. Hard copies are likely to be available from the National Radio Club and International Radio Club of America reprints services.)

1. Testing Two 'Kaz' Squashed Delta Antennas, John Bryant, 2001

2. Pennant and Kaz Antenna Tests, Mark Connelly, 2001

3. RF Circuits page (links to construction articles)

4. Fabricating Impedance Transformers for Receiving Antennas, John Bryant, 2001

5. W7IUV Amplifier

6. Pennant Antenna with Remote Termination Control, Mark Connelly, 2000
"" and


APPENDIX: Some Reversible Kaz / Flag experiments by Andy Ikin
Originally sent to: [email protected]

Date: Tue, 8 May 2001 20:40:39 +0100
From: "Andy Ikin" <[email protected]>
Subject: Reversible KAZ-FLAG

Approximately two weeks ago I decided to give the KAZ Delta loop a try to see if there was any improvement over my existing K9AY.

The KAZ Delta loop E-W ( 10 foot by 40 foot ) was set up near to my K9AY E-W loop. The base of the KAZ was 1 foot off the ground, with one end connected to 20:1 z matching xmfr, the other end was connected with a Perkins/Elmer VTL5C4 Vactrol to provide remote controlled termination. Testing was conducted between 10 am and 3pm local time.

On LW, both Allouis 162 kHz and Europe No 1 183kHz yielded 17.5dB F/B, whilst RTL 234kHz and Kalundborg 243kHz provided 12.5 dB F/B.

Medium wave F/B varied from 17.5dB for Paris 864kHz and Belgium 621/540kHz to 30dB for Lille 1377kHz and Flevoland 1008kHz ( Flevoland 747kHz F/B was 20dB ). 13dB of pre-amplification at the receiver was used to raise the signal above the receiver noise floor so that the F/B could be measured. Without pre-amplification, the low gain degraded the reception quality of a significant number weaker stations. The average F/B for LW was 15dB and MW was 23dB. The use of a common mode feeder isolation choke next to the matching xmfr made no difference to the F/B.

Comparing the above without pre-amplification to Gary Breeds K9AY un-amplified ) with remote controlled termination. The K9AY provided about 3 dB higher F/B for RTL and Kalundborg and the same F/B for Allouis and Europe No 1. The K9AY gain was typically 10dB higher on LW and 15dB higher on MW. On MW, the K9AY the F/B varied from 20 to 40dB. The average F/B for LW was 15.8dB and MW was 30dB.

I also tried the K9AY using the KAZ size loop. On LW the F/B was the same as the K9AY. For MW the F/B was typically 7db lower. At 19:00 hours using the K9AY I was able to null Flevoland ( 400kW ) to provide useable reception of Cadiz ( 10kW ). The K9AY ( KAZ ) null was not deep enough to provide a useable signal from Cadiz. The gain difference between the two K9AYs was about +7dB in favour of the Gary Breed K9AY. However, I don't see the performance difference between the two K9AYs as a problem, but something to be expected from two loop shapes.

My real concern was the poor F/B performance of the KAZ on MW, so was I doing something wrong or was some other factor affecting the performance?

I did not compare the KAZ to another Pennant. So it would be unfair to say that the problem was just with the KAZ.

My first thoughts were just to dismiss this exercise as another failure. However, I decided to do some more experimentation; first was to increase the antenna gain by placing an amplifier directly to the loop. This would provide reverse isolation of the feeder to the antenna and so prevent any feeder induced signal degrading the F/B. The Amplifier I used was a DATONG AD270 Active Dipole Head Unit. This resulted in increasing the average LW F/B to 19dB and the MW F/B to 27dB. The down side to using the DATONG AD270 was too much gain on LW resulting in RX overload with Broadcast Stations.

Next, I had the idea of using an amplifier/Vactrol combination at each end of the antenna to provide a remote control of the antenna direction together with remote termination control and feeder isolation. Unfortunately, I didn't have another AD 270, so I built two Hi z input Amplifiers using VMOS FETS in a differential/push-pull configuration (antenna connects directly to the FET gates via coupling caps). Using Hi z input Amplifiers allows for the Amplifier to shunt the Vactrol with only a minor reduction in the F/B. Each Amplifier feeder and the Vactrol control line was brought back to a Control box next to the Rx. The antenna direction was achieved by simply switching one of the Amplifiers to the Rx and controlling the Vactrol at the opposite end of the antenna. This Reversible KAZ provided an average LW F/B of 18dB and the MW F/B of 28dB. Unfortunately, I didn't slug the HF gain of the VMOS amplifier, thus I did experience some intermod on MW. However, I considered that this antenna design was a success, as I had integrated four improvements to KAZ/Flag/Pennant; 20dB gain, Remote Reversibility, Remote Termination and Reverse Feeder Isolation.

Later, I abandoned the VMOS FET Antenna Amplifier in favour of a Bipolar design ( a 20dB gain version used in some Wellbrook K9AYs ). This Amplifier was used with a 20:1z input xmfr and a DPDT relay to switch the antenna to either the Vactrol or the amplifier. With no power ( 12volts via the feeder ) applied to the Amplifier, the Vactrol would terminate the antenna. Applying power the Amplifier, the relay is energised to isolated Vactrol and connect the Amplifier to the antenna.

The remote Vactrol termination is controlled via a separate single wire from the Control Box, the control voltage return path is via the feeder screen. Thus, by just selecting ( powering up ) the amplifier, remote beam reversal is achieved. This amplifier configuration provided the same performance to the VMOS version, with an average LW F/B of 18dB and the MW F/B of 28dB without any intermod problems.

This Antenna amplifier system could be expanded to several antennas with the Vactrol control line being daisy changed to the amplifiers.

A single pole multi-way switch in the Control Box is just required to select the require antenna amplifier. Alternatively, a relay box could be used at the 'Antenna Farm' to reduce the number of feeder cables coming back to the antenna control box.

An interesting feature of this antenna design is the possibility of its use with the EWE antenna and the Delta and Diamond variants of the Pennant/Flag.

To summarise;
20dB gain, Remote Reversibility, Remote Termination and Reverse feeder isolation is achievable in one design.

Although this antenna configuration is more complicated than the K9AY, I am pleased that it achieve the same performance.

Finally, I think that one issue with the Flag/Pennant antenna needs further investigation; Is reverse isolation of the feeder a key factor in achieving a good F/B ? ( a 20:1 z xfmr does not provide any reverse isolation, it only reduces capacitive coupling). Or is simply increasing the gain of the antenna before the feeder improving the F/B? If either is the case, then placing the amplifier next to or very near to the antenna may be necessary design feature!

Kind regards and good listening


Andy Ikin

e-mails: [email protected]; [email protected];
[email protected]