Portable Phased Dipoles for 20m

Bob Rose, KC1DSQ & Bob Glorioso, W1IS


Phasing waves, acoustic, radio, radar, etc. is a way of directing energy where you want it.  We have all had the challenge of “phasing” our stereo speakers to get a correct “Stereo Effect.”  We do this by reversing the connection on one speaker, until positioned in the middle of the two speakers, the sound appears to come from the middle.  Similarly, we can adjust the phase of two antennas to put most of our signal where we want it.

Yagi antennas use passive radiators (reflectors and directors) in the antenna field to concentrate and direct energy in the direction the antenna is pointed1,2.  Here we create an RF Beam just like we phase speakers by driving both antennas and changing the phase of one of two antennas so the waves emitted by both antennas send their energy in the same direction at the same time.  This is usually done with vertical antennas with two or even four as in the 4-Square3,4 and more rarely done with horizontal dipoles.

This antenna, constructed with two horizontal 20m wire dipoles spaced ten feet apart using phasing to feed the antennas, has the ability to quickly switch directivity 180 degrees through a simple relay box containing the relay, matching circuits, phasing line, and Balun. 

Instead of using multiple towers to support our beam, we designed our antenna to be portable and hung from ropes in only two trees.  It is easily transported in a mini-van or on the roof rack of an SUV because the longest items are ten-foot dowels.  This makes it perfect for Field Day, Parks on the air, or even for DX-peditions where the spreaders can be obtained locally leaving only wires, cable and a 4”x4”x2” plastic box that need to be transported.


Basic Theory of Operation of Phased Dipoles 3,4

Figure 1 shows two dipoles viewed from the end separated by one-quarter wave, a distance equivalent to 90 degrees of phase shift.  At the design frequency 14.175 MHz, this distance is 69.38 ft / 4 = 17.345 ft.  An observer placed right of and facing Dipole 2 sees the forward direction. An observer placed to the left of Dipole 1 sees the backwards direction.


      < -- Backwards            Dipole 1           Dipole 2           Forward -- >

Figure 1. End View of Two Dipoles


Figure 2 shows the waveforms driving Dipoles 1 and 2.  Dipole 1 is driven by a sine wave while Dipole 2 is driven by the same sine wave delayed by a 90-degree phase shift.

Figure 2. Delay line causes the feed to Dipole 2 to lag the feed to Dipole 1 by 90-degrees.

An observer placed in the forward direction to the right of Dipole 2 sees the waveform from Dipole 1 arrive after a delay of 90 degrees due to the 90-degree spacing between the dipoles.  The waveform driving Dipole 2 is also delayed by 90 degrees.  Thus, Dipole 1 and Dipole 2 waveforms are in phase and the far field is the sum of the two, Figure 3. 

Figure 3. The delay in signal propagation plus the delay line causes radiation from antennas to add in the forward direction.


An observer placed in the reverse direction to the left of Dipole 1 sees the waveform from Dipole 2 arrive after a delay of 90 degrees due to the spacing between the dipoles. Dipole 2 already has a delay of 90 degrees due to the phasing line. Thus, it arrives at dipole 1 with a total phase shift of 180 degrees causing the two waveforms to cancel for the far field and the amplitude in that direction is zero, Figure 4, creating a directional radiation pattern with strong signal forward and a much weaker signal to the rear. 

This is normally implemented with equal length feedlines and a quarter wavelength phasing line inserted in the feed to the dipole for the forward direction. The forward direction can be switched by moving the phasing line to the other dipole via a relay.

Figure 4. Cancellation of the signals from the two antennas in the reverse direction minimizes the radiation from the rear of the array.

These are the fundamentals of a phased dipole array.  However, in practice we need to deal with the complexities caused by the interaction of the two closely spaced resonant dipoles. 


Shortening the Separation Between Dipoles

To achieve portability, we reduced the separation from 17.345 ft to 10 ft. This reduces the separation delay to less than a quarter wave, 90 degrees, to 51.9 degrees.  Given that the gain of an antenna is proportional to its directivity, we chose to optimize the design by adjusting the phase for cancellation in the reverse direction allowing most of the radiation to be forward.

 We achieve backwards cancellation by increasing the length of the phasing line to make up for the shorter separation.  Therefore, the total delay from dipole 1 to dipole 2 needs to be 180 degrees and the length of the phasing line must be 180 – 59.1 = 128.1 degrees. 

Simulation and testing proved that directivity is equivalent to 90-degree dipole separation with similar forward gain and front to back ratio.


Design of the Phasing Network

Because of our requirement for portability, our phasing network design departs from the techniques commonly used for other arrays.  To start, we chose to use feedlines from the dipoles to reach the center, at least five feet.  Next, through simulation, we adjusted the length to transform the impedance of the dipoles at 40 feet to 50-ohms.  For RG-59 with a VF of 0.66, the length is 5’ 9”.  This spans the 10 ft dipole spacing to feed to a common point between the dipoles leaving slack in the cables. 

With the dipole’s length at 32 ft at a height of 40 ft, simulation indicates 11+ dBi forward gain at a take-off angle of 25 degrees and 16 to 25 dB front-to-back ratio across the 20m band, Figures 5a and 5b.  A dipole’s gain at 40 ft with a 25-degree takeoff angle is about 4 dBi. Thus, the expected gain of the phased dipoles over a dipole at 40 feet is at least 6 dB or 4X power gain. 

Figure 5c shows the SWR across the 20 m band.  While the take-off angle changes from 40 degrees at 20 feet to 10 degrees at 80 feet, the SWR curves at these heights are similar.

The impedance of the combined antennas is about 9 Ohms, which defines the matching network needed to get to a 50-ohm feedline.


Figure 5a.  Elevation Pattern
Figure 5b. Azimuth Pattern       
Figure 5c. SWR with 66 feet of RG-8X

Figure 5.  Simulated performance of the 20m phased dipoles array.

Later we will describe the results of our field tests to verify Performance, SWR and front-to-back ratio on the air.  


Building the Phased Dipoles

The layout of the antenna, Figure 6, defines the major components of the array.  The dipoles are simple 32-foot 20m dipoles each fed directly with 5.9 feet of 75-ohm coax.  The antennas are connected to a combination switch box, containing a coax delay line, impedance matcher, the relay, a Bias Tee to power the relay, and 1:1 Balun/Choke, the “Combo Box.”

 Figure 6. Top View of 20m Phased Dipoles


This separation between the two dipoles is provided by two 10’x1.25” wooden dowels available at most hardware and big box stores.  We use parachute cord for the rope harness and between the centers of the dowels to support the combo box and coax to the rig. This prevents the weight of the Combo Box and coax from pulling the dipoles towards the center.

The dipole dimensions are typical and, as with all antennas, the lengths should be cut 3” or 4” longer for tuning.  The wire recommended is #14 THHN from your electrical or hardware store or #14 FlexWeave™ from Davis RF.

The array depends on the Combo Box, Figure 7, to provide line isolation and impedance matching from the 9 ohms presented by the array to 50 ohms for the feed-line while the relay allows the direction of the array to be switched 180 degrees.  For example, if you are in the middle of the country for Field Day, you can select to focus on either the east or west coast.  On the other hand, we in the Boston area test the array aimed NE-SW allowing us to work Europe in one direction and along the east coast out to Texas in the other direction.

The relay is a single coil latching relay. A positive pulse latches it in one direction and a negative pulse latches it in the other direction. The relay specified is a DPDT relay. The two sets of 8.0 Amp contacts are wired in parallel creating an SPDT relay with twice the current capacity. The relay is mounted on a prototype PC board with pre-drilled, plated-through holes on 100 mil centers. The control signal for the relay is sent via a bias tee circuit described below.



Figure 7.  Diagram of Contents of the Como Box


The Combo box uses a Carlon A-273, 4”x4”x2” weather proof box available at local hardware and electrical stores, Figure 8.  The relay, matching network and balun are attached to the box with Scotch Exterior Mounting Tape. 

 Figure 8a. Combo Box Layout
Figure 8b. Completed Combo Box

The balun5 consists of 16 turns of RG-316 on a 140-43 core using a crossover winding, Figure 8a. 

The matching board has a low Q pi-network mounted on a single sided PC board.  Use a Dremel tool or hobby knife to isolate the lands.  The schematic, and PC board dimensions are given in Figure 9.  Attach the Balun input to the SO-239 and the Balun output to the 50-ohm side of the network.  Use one-inch leads from the 9-ohm side of the network to connect to the relay contacts.  Use short runs of RG-316 coax to attach the relay contacts to the SO-239s for the dipole feedlines.  Provide a short bus of bare wire on the small board near the relay to solder the shields of the coax and the ground side of the matching network. Good RF layout practice is important.  The 14-foot phasing line of RG-316 is rolled up and soldered across the two SO-239s for the dipole feedlines.  The Combo box can handle up to 500 watts.

 Figure 9a. PC Board
Figure 9b. Schematic
Figure 9. PC board layout and schematic of the matching network.


Bias Tee

The Bias Tee, Figure 10, delivers the DC control voltage for the relay over the coax.  The capacitor in series with the center conductor keeps DC from entering the rig.  Since there are no components at the antenna that can affect or be affected by DC, a capacitor is not needed in the Combo box.  A choke carries the DC from the control box and isolates RF from the DC control signal on both ends.  The control box housed in a metal mini-box uses two 9- volt batteries to charge a capacitor that is discharged down the line to switch the latching relay.  The double pole-double throw switch changes the polarity of the pulse sent to the relay by pushing the button.   


Figure 10. Bias Tee Schematic – Control box & Bias Tee Receiver in Combo Box


Tuning and Performance

Before launching the antenna make sure that the antenna wires connected to the center conductors of the RG-59 coax are attached to the same spreader and the wires connected to the shield are both attached to the other spreader.  Then raise the antenna to 35 or 40 feet, measure SWR, drop the array as needed to get low SWR in the middle of the band by adjusting the lengths of both legs of both dipoles.  All four antenna legs must be the same length.

The SWR of the antennas we have built and tested using 66 ft of RG-8X is represented in Figure 11.

 Figure 11. SWR 20m Phased Dipoles


We test all our antenna designs in W1IS’s back yard near Boston using a KX3 at 5 watts on SSB.  During the first tests of this design with the array pointed at Europe we worked six countries and noted that we didn’t hear many stateside signals.  Subsequent tests, this time with a working relay, with several local stations from 3 to 15 miles away, as well as DX stations we found that the Front-to-Back ratio was between 3 and 5 S-units, a bit better than we expected.  Even for fairly local stations, our S-5 signal became unreadable when the antenna phasing was reversed. 



Our thanks to our spouses, Dee, W1MGA, and Barbara Rose who tolerated our constant phone calls and time spent developing and testing this antenna as well as several fellow members of our local club, Police Amateur Radio Team of Westford, MA, including Steve, W1KBE, Colin, W1DJR, Bonnie, AC1IY, Frank, KB1HSC, Allison, KB1GMX, and Greg, N1DAM for assisting with the Front-to-Back measurements and proof reading the text.



  1. Yagi Uda Antenna, https://en.wikipedia.org/wiki/Yagi–Uda_antenna
  2. The ARRL Antenna Book, 24th Edition, Chapter 11
  3. The ARRL Antenna Book, 24th Edition, Chapter 6
  4. Low Band DXing, John Devoldere, ON4UN, ARRL
  5. Understanding, Building and Using Baluns and Ununs, Jerry Sevick, W2FMI, CQ, https://store.cq-amateur-radio.com/product-category/books/projects-and-construction/page/2/


Materials List


  • 75 Feet #14 THHN wire available at hardware stores or #14 FlexWeave™ from Davis RF.
  • 6 Insulators
  • 2 - 1”x10’ wooden dowels available at hardware and lumber stores.
  • 100 Feet #550 parachute cord
  • 15 Feet RG-59U 75-ohm coax
  • 20 Feet RG-316 coax
  • 1 - FT140-43 ferrite toroid for the Balun
  • 1 – T37-2 ferrite toroid for matching network inductor
  • 1 – 430 pF, 500 v, Mica Capacitor, CD15FD431J0F, Mouser P/N 598-CD15FD431J03F
  • 1- 330pF. 500 v, Mica Capacitor, CDV16FF331J03F, Mouser P/N 598-CDV16FF331JO3F
  • 4”x4”x2” Waterproof Box, Carlon A-273, available at Electrical and Hardware stores.
  • Latching Relay, Potter and Brumfield RT424A12, Allied 70288718
  • Prototype PC board with plated through holes on 100 mil centers
  • Single Sided Copper Clad Board for matching network
  • Scotch Exterior Mounting Tape
  • 2 – SO-239, Coax Sockets, Connect to Dipoles
  • 1- SO-239, Coax Socet, Cable to Radio, J3

Bias Tee

  • Metal Controls Box, Bud Cu234, 4.7”x 3.7”x 2.2” Allied 70148694
  • 2- .01 uF 100 V Disk Ceramic Capacitor, C2, C4
  • 3- .01uF 1KV Disk Ceramic Capacitors in parallel, C3
  • 1- 500 uF 25 V Electrolytic, C1
  • 1- 2200 Ohm 1/4W resistor, R1
  • 1- SPST Switch, S2
  • 1- DPDT Switch, S1
  • 1- N.O. Push-Button Switch, PB1
  • 3- SO-239 Coax Sockets, J1, J2, J3

Leave a comment

Please note, comments must be approved before they are published