How to Build Phased Verticals from Stuff at Home Depot

A lot of people on-the-air have asked about how to construct the verticals using materials available at your local Home Depot store. Although a lot of this is conveyed in the photographs, let's go through the construction of the final prototype, and add in a couple footnotes about how to beef it up mechanically.

Setting base supports in concrete is easy.

The two verticals are not guyed. They are supported by 4x4 posts set in concrete to which 2x4s are bolted. Two metal brackets are fastened into the top of the 2x4 which receive PVC rings acting as insulators for the aluminum tubes that pass through them. The 4x4 posts are 8' in length with about 2 feet set in the ground. 50 pound bags of Quickcrete, a fast-setting cement available at Home Depot, are poured around the posts after they have been set into the holes. Water is then drizzled over the cement over a period of about 10 minutes. In about an hour the cement is set. Before pouring the water, the 4x4 posts are made standing true with a carpenter's square. Only a few implements are needed to set base supports for verticals. 

Introduction

T-Connectors simplify Christman phasing.

Phased verticals exhibit directivity due to a difference in the phase of the signal exciting them. This holds true for transmit as well as on receive. It is caused by two things: (i) how far apart the verticals are from one another, and (ii) the lengths of the transmission lines connected to them. In the case of receive, the phase difference varies in accordance with the compass heading of the incoming signal. If it arrives at the front of the array, the phase difference adds the signal up, making it louder. If the signal arrives behind the array, the phase difference cancels the signal out, making it weaker. Gradations between the two extremes are distributed over the intervening compass headings. So how great are these two extremes? A signal arriving in front of the array is enhanced by a factor of two (3 dB), whereas a signal arriving behind the array is reduced by several orders of magnitude (20 - 30 dB). Thus two phased verticals do not exhibit much forward-gain since they only double the signal on transmit and receive. They do so over a wide swath (~120-degrees), after which radiation drops off precipitously towards the rear (180-degrees). So when you hear someone switching the directivity of a pair of phased verticals back and forth on the air, the great difference you hear is not due to the forward gain. It's due to the reward rejection of the antenna. A pair of phased verticals oriented East/West on the East coast of the USA enhances coverage of European, African and South American signals by about 3dB, while reducing stateside QRM and QRN by about 20 to 30 dB. Thus on a clear winter night, a pair of phased verticals can produce an astounding listening experience. 

Initial Summary

Phased verticals do not exhibit high forward-gain. On receive, the compass heading of the incoming signal determines the phase difference between the verticals. On transmit, the phase difference between the verticals determines the compass heading of the outgoing signal.

- INSERT FIGURE 1 -

Steering the Array: Line Delay & Space Delay

The phase difference can be varied by means of electrical wizardry. The Christman method, amongst others, accomplishes this feat by adding an extra piece of transmission lines to one of the vertical's feedlines. This compels the transmitted signal to take a little more time to reach one vertical compared to the other, and is called the "line-delay". It constitutes a part of the total phase difference between the two verticals. The other part is derived by how long it takes the signal to travel from one vertical to the other vertical, and is determined by the distance between the verticals. Both are expressed as degrees of the signal's oscillation. 

Since the purpose of the extra piece of feedline is to delay the signal, it is called a "delay line". It's length is measured by the number of degrees the signal advances in one of its oscillations during the time it takes to travel down the delay line. Interestingly, delay lines are measured in units of time, not distance. And these units of time, in turn, are not measured in seconds, but in degrees of a 360-degree oscillation. Since our array requires a 90-degree phase-shift, the length of our delay cable will be equal to how far the signal travels down this cable while it progresses 90-degrees, or 1/4, of an oscillation. As you might suspect, the 90-degree delay line ends up being 1/4 wavelength long. Equally uninspiring might be the realization that a "180-degree delay line" is 1/2 wavelength long, and that a "270-degree delay line" is 3/4 wavelength long. And, yes, we do have to take into account the velocity factor--but not until we are finished with all the theoretical calculations. In other words, don't even think about the velocity factor until you find yourself in the field with the wire cutters in your hand. 

How phased verticals work is a little more complicated than this. And we will get into these details in a moment. For now, let this explanation suffice on our initial attempt to understand how this fascinating aerial works.

Cutting Delay Lines: Some Practical Tips

It is important to get into the habit of thinking of distance in terms of time when discussing phased arrays. You will find this to be true whenever delay lines, feedlines and inter-element spacings are discussed amongst aficionados  These phased-array experts appear to think of space as time measured not in seconds, but in the number of degrees a signal advances in an oscillation as it moves through space. This type of thinking transposes one-wavelength into 360 degrees, 1/2 wavelength into 180 degrees, and a 1/4 wavelength into 90 degrees. I reiterate this point due to the following quandary: what happens if you need an 84-degree line? Or a 71-degree line? How many feet is that?

To answer this question, let's suppose you're assisting an RF engineer constructing a phased array in the field. And he or she asks you to cut an 84-degree line of coax. What do you do? One approach is to reformulate the request as meaning they want you to cut off part of a 1-wavelength cable, which you know is 360 degrees "long". And that the engineer wants 84 degrees of it--or "84 parts out of 360". So, you divide 84 by 360 (on your iPhone) and get .2334. And then you calculate how long 1-wavelength is in feet (936/f), and multiply this by .2334. This gives you the length of the "84-degree cable" the engineer wants you to cut. After measuring it out on the ground, you hover over it with your wire-cutters. But then you abruptly stop. Why? Because you remembered to figure in the velocity factor!

Actually, you don't have to do that if you have an antenna analyzer. You can go ahead and cut the coax because the velocity factor, which Einstein would agree is always less than 1, guarantees the final length will be less than the amount you've measured out. After snipping off this length of transmission line, you can trim it to the precise "84-degree length" using an antenna analyzer, as covered elsewhere in this tome. How much will you end up trimming off? If the velocity factor is .82, which it oftentimes is, about 1/5th of the coax will be trimmed away. For a velocity factor of .66, it'll be about 1/3. If you don't have an antenna analyzer, now's the time to apply the velocity factor and make the final cut, adding on a couple of extra inches in case you screw-up putting on the connectors. Done. Close enough for government work. 

Wiring the Remote Box

Remembering the shields are all tied together, let's examine how to wire up the remote relay box used to switch the directivity of the array, as provided in Figure 2, below.

Christman Phasing Method

One intriguing aspect of the Christman phasing method is that it uses a 71˚ delay line, rather than the 90˚ delay line one would expect to see used. The 71˚ delay line is accompanied by 84˚ runs of coax connecting the remote switching box to the verticals. This odd combination of line lengths achieves the phase shift required to produce the unidirectional pattern while dividing the power equally to both verticals. The Christman method gets away with this because it is not necessary to phase the two verticals 90˚ in order to produce the unidirectional pattern. This pattern emerges over a range of phase differentials with slight variations. The weird cable lengths associated with the Christman method merely make fortuitous use of the fact that these lengths divide power equally between the verticals at the phase difference needed to produce the unidirectional pattern. Incidentally, if you need longer runs of cable from your verticals to the switch box than 84˚, the Christman method can be re-worked to produce them. (Insert footnote providing links to technical articles).

required and equal current distribution between verticals required for this system to work is due to the fact that each vertical, when it radiates its signal, induces a current in the other vertical. The two play off one another through the same mutual coupling found at work in a parasitic array. So when one vertical radiates as energy the current fed to it by the transmission line, it induces a current in the other vertical, which then re-radiates this energy resulting in another current being induced in the first vertical. And vice-versa. It all gets quite complicated, requiring the kind of complicated mathematics thankfully consigned to theoretical physics. The bottom-line for our review is this: although equal currents are forced at the feedpoints of both verticals by Christman's odd cable combinations, the two verticals do not end up radiating equal amounts of energy. Some of the energy radiated by one vertical is coming from the other vertical. If it was not this way then feeding two verticals would only result in twice the effected radiated power produced by a single vertical. And this is, in fact, what we do see happening in the "forward" direction of two phased verticals--a 3 dB increase in the radiated energy. But this doubling of the radiation off the "forward" end of the array is not the result of all the complicated interactions between the two verticals due to mutual coupling and phase differences. These aspects of the array are what cause the tremendous reduction of radiation to its "rear". The complicated interactions of currents inducing other currents due to the mutual coupling between the verticals causes the huge rear null in the array's radiation pattern. This is the product of these complexities--not the 3 dB of forward gain. The 3 dB of forward gain is a side-product of the fact that the energy not being radiated by the rear of the array has to go somewhere; it can't disappear. It's spill-over. The physical analogy can be approximated by imagining sticking your finger (deeply) into a party balloon: one side becomes conically depressed as the other side slightly expands. 

And you can tell which is the primary result and which is a concomitant of all the complicated current interactions between two phased verticals by comparing the essence of the rearward null to the slight forward enhancement. The rearward null has the more pronounced essence in the form of reduced radiation several orders in magnitude dispersed over a well-defined  sharp angular displacement. Compare that to an almost imperceivable 3 dB doubling of radiation dispersed over a wide, 120-degree swath. Which is the result of the complicated current interactions dictated by phase differentials between two verticals, and which one is the result of the fact that radiation displaced by the former has to end up somewhere? One way to intuit an understanding is to imagine the complex current interactions between the verticals causing a doubling of radiation over 120 degrees on one side of the array, causing the displacement of several orders of magnitude of radiation over well-defined, sharp angles on the other side.

***

Use of T-connectors to link the 84, 71 and 84-degree cables to each other (in that order) simplifies the wiring of the remote switching box, to which the T-connectors are attached by means of SO-239 connectors. This arrangement enables the 71-degree delay line to be neatly coiled and hung off the remote relay box which, itself, is mounted between the two verticals.

Operation

When the remote switching relay connects the main feedline to the first T-connector, the first vertical gets fed through its 84-degree line while the second vertical gets fed through its 71 and 84-degree lines. This causes the signal radiated by the second vertical to lag behind the signal radiated by the first. When the relay connects the main feedline to the second T-connector, the second vertical is now being fed through its 84-degree line while the first vertical is being fed through its 71 and 84-degree lines. This causes the signal radiated by the first vertical to lag behind the second vertical. This reverses the directivity of the array--which beams towards the vertical that has the extra coax in its feedline. 

Construction Tip: Wire up your remote switching box so that the extra coax is added to the vertical pointing towards the DX region you work most. This will cause the phased array to point in this direction when there is no power sent to the remote relay box.

A second relay shorts out the 71-degree delay line, causing both verticals to be fed through 84-degree lines. This causes them to radiate in phase, which produces the omnidirectional pattern. 

Whether or Not to Isolate the Elevated Radials

At some point you need to decide whether or not to tie the elevated radial systems together wherever they intersect. You can wrap wires or solder them at their intersecting points, or electrically isolate them with insulated spacers. I opted to isolate them with 6" PVC spacers because a footnote I read in a technical paper indicated that doing so adds about .75 dB to the array's forward gain.

Keeping the elevated radial systems electrically isolated necessitates lifting the shields of the coaxial cables at the switching box. This, in turn, requires relays to switch the shields along with their respective center-conductors in accordance with the various configurations just described. Using a plastic box isolates the SO-239 connectors. However, if you use a plastic box in a system that ties the shields together, grounding the SO-239 connectors with a piece of wire introduces a reactive component which degrades the array's performance. The workaround is to either (i) ground the SO-239 connectors with a copper or aluminum strap, or (ii) use a metallic box for the remote relay enclosure.

Purists can employ plastic boxes in which relays switch coaxial shields along with center-conductors. If you are using low-power, a single DPDT relay can be tasked to switch the array's directivity East-West, and another DPDT used to short out the delay line as described. For higher power operations it might be advisable to double-up the contacts of a DPDT relay to increase it's power handling capabilities. If you opt for this approach, you will need two (2) DPDT relays with the shields all tied together, and four (4) DPDT relays for the ground-isolated configuration.