Stacked Dipoles for 2m?

(1/4) > >>

Jared Sherman:
Is anyone familiar with stacked dipoles? I've been looking at a half*ssed article in an old AARL book which shows a typical stack of four on one vertical pole. I say 'half' because they conveniently omitted some dimensions, and the logic behind them.

I'm thinking about a modified split stack for my own problem installation. Basically I have access only to whatever can be hung outside of a triple window about 5' tall and 7' wide in the brick wall of a building.

So I'm thinking, stacked dipole for the +6db gain, but split the stack into two halves, and put each one to one side of the window. I'd have two dipoles "hiding" as a pole on one side of the window, two on the other side, and the spread between them would take up most of the cable length that the original design shows.

Of course I've got no idea if any of the cabling lengths are critical...It looks like the antennas are kept 1/2 wavelength apart and the cables may be cut to provide a wavelength of cable (?) between each pair, and again between the two sets?

Working with the space and shape limits I've got, it looks like a reasonable way to get good gain with low cost.

Steve Katz:
Stacked dipoles cannot provide 6 dB gain; perfectly stacked and phased dipoles can provide 3 dB gain (over a single dipole), and for vertically polarized use must be stacked one above another and spaced one wavelength apart, center-to-center.  This means one set of (2) stacked vertical dipoles occupy exactly 1.5 wavelengths of vertical space, or 120 inches, which is ten feet.  If you stack them in any other fashion, you don't achieve gain.  

The classic "4-pole" or "4-stack" of vertical dipoles is twenty feet tall (top to bottom), using four vertical dipoles.  It's a very common, and popular antenna for repeater use and many home stations use them as well.  But you need 20 feet of vertical space to make one work.

The phasing harness which interconnects the vertical dipoles is made from odd quarter waves of 75 Ohm coaxial cable per branch.  Since each cable needs to be physically one wavelength long, and one wavelenght in coaxial cable is only 0.66 wavelengths long (due to the velocity factor of propagation of coaxial cable), usually three or five quarter wavelengths of coax are used per branch of the phasing harness.  For two stacked dipoles, two such identical lengths are required.  They combine signals in a coaxial "T" adapter, to the third port of which any length of 50 Ohm coaxial cable is attached for connection to the station equipment.

One quarter wavelength at 146.00 MHz in standard coaxial cable (any impedance cable, provided it's solid polyethylene dielectric) is 13.34".  Three quarter waves in coax, then, is 40", or just enough to reach from one dipole feedpoint to a point midway between the two dipoles, assuming there's no slack in the cable.


Jared Sherman:
Thanks, Steve. There are some major discrepancies between what you say and what the old ARRL book said. First being that they are no using center-fed dipoles, they are using what I would call loops, i.e. a rectangular "loop" which is not closed and is fed slightly asymetrically at those ends.

For some reason ASCII art isn't being supported here, draw a mental picture of a skiny rectangle the size of a cigarette carton. Make the loop around it the long way, cut the loop slightly off center on one side, that's the feed points. Shield to one side, center to other.

And, the loop is one wave long, but wrapped that way it only is 1/2 wave "long" in the space it fills. They also space them 60" center to center, which seems to be another 1/2 wave apart end-to-end, so they are getting the dipoles packed in half the space that you're thinking of.

This is from the 1974 AARL Antenna Book, which unlike the ridiculous 2004 Antenna "Phone Book" was pocket sized.

They're also specific about getting 6db gain from four dipoles stacked this way although they don't say with reference to what zero point.

I could scan the pages & upload them or email them to you if you wanted to see them.

With their much shorter length, they'd work very well for me. And as far as I can figure, the overall gain comes from the multiple antennas--if I break the "stack" into two sets of two, I should still see gain from each pair being combined with the same impedancesin the system. (Or, I've still got lots more to learn.<G>)

In theory if one antenna grabs "this much" of a signal, two antennas working in phase would grab 2x as much...and so on. I know, once I break the "strip" I'm breaking the phasing...I'd have to sneak a tape meaure outside to see if I could get away with all four, 210" according to their diagram, 17.5 feet. Maybe.

Any ideas for better alternatives?

Steve Katz:
MMI: The antenna you describe is a standard 4-pole array of dipoles, but each dipole is a *folded* dipole rather than a linear one.  The folded dipole is still 1/2-wavelength long (you don't count the perimeter of the loop formed as the length, you count the end-to-end length from one loop to the opposite one), and the advantage of using folded dipoles is they don't need any balun, gamma match or other matching device at the feedpoint of each dipole.  The 4-pole antenna using folded dipoles is ubiquitious and the most common configuration sold commercially in the 2-way field -- there's millions of them out there.

The "gain" isn't from "grabbing" more signal; the gain from stacked dipoles, or any gain antenna, is achieved by losing signal pattern in some directions, thus peaking it in others.  In the case of stacked vertical dipoles, they must be one above another, above another, above another, in order to create omnidirectional gain.  Stacking them in any other fashion or orientation might create some gain in one or two directions, but not omnidirectionally, which is the whole advantage of stacked dipoles.  (If you want to create gain in just one direction, a small Yagi or Quad, or even a corner reflector, is a much more efficient use of space.)

The omnidirectional gain is achieved by pattern compression, moving energy from high angles to lower angles.  If you can take advantage of a very low radiation angle antenna (which generally means your antenna needs to be quite high above ground, otherwise the lower angle pattern can work against you, rather than benefit you), a stack of dipoles is a great choice, and the most common "repeater" antenna out there.  You'll note most repeater antennas, made of dipole stacks, are located atop tall buildings or towers.

6 dB gain can be achieved, as I originally posted, with *four* dipoles in a stacked array; it cannot be achieved with *two.*  3 dB gain can be achieved with two.  If you stack two dipoles vertically (one above another) and then stack another two dipoles vertically (same way) but install the second set of dipoles to the side of the first set, rather than above or below it in the same plane, you do *not* achieve 6 dB omnidirectional gain.  It's impossible.

BTW, the contributing author to some of the antenna writeups in the Handbook, regarding VHF-UHF antenna array construction, was me.  


Jared Sherman:
<The "gain" isn't from "grabbing" more signal; the gain from stacked dipoles, or any gain antenna, is achieved by losing signal pattern in some directions, thus peaking it in others>
 I confess, I don't understand. I haven't found any explanation in simple physics terms for how stacking dipoles "lose" signal in some direction (vertical along their track) in order to gain it for the array.
 The obvious guess is that there is interaction between the ends of the elements...but is there? And how?

 And, how critical are the dimesions on the folded dipoles? IOW, I can probably hand an 18' long bamboo pole or inch wide plexi sheet outside the window. That would let me run the four-tall array, if I keep the folded dipoles squat enough to fit on it, i.e. a half way long but only an inch tall, basically making the folded dipoles thinner than the coax.

Can't just use alarm tape and "paint" them on the outside of the coax, huh?<G>

Did you work on the 1972 "FM" book, that's the one I was getting the dipole array article from. (Yes, a code rule can keep people out of things for a long time.)


[0] Message Index

[#] Next page