- Because high frequency ham radio antenna design is maybe not your “thing” it’s a good idea to read the Saturday column (Part 1) to grasp some of the underlying technical angles.
- There is one general shop tool note at the end of this article. If you’re not a ham, that may be more useful than extreme ham radio antenna designs!
It began with a Cage Dipole. Years ago, I lived on my 40-foot sailboat at Ballard Mill Marina in Seattle on the Ship Canal. My buddy Scott was doing mixed gas school over at Diver’s Institute (last I heard he was welding gold bars, or something, under oil rigs in the Gulf). To make ends meet in school, he was caretaker on the light ship that was awaiting restoration. The “Relief” I think it was. Decades long lightship off Astoria, Oregon.
Gave me a chance to climb around and see one of these mythical antennas close-up and personal.
While the antennas were tied together at the center insulator, and again at the far end by the support insulator, they were spaced a foot to 18-inches apart around circular insulators in-between.
That’s how I got started on Super Antennas (I thru IV).
Thank Antenna Modeling!
My first breakthrough was discovering that a simple off-center fed Dipole could have its performance radically enhanced by simply adding two additional wires on the long side of the antenna. Connected at the center, far ends 4″ apart and NOT connected out there.
Let’s take a 50-foot one side and 90-foot the other Off Center Fed Dipole (OCFD). When we look at it’s SWR curves, it’s very good, but unfortunately resonates down around 3.3 MHz which is outside the ham bands. But look at the return loss – again we’re in that >95% transmit efficiency range
Adding wires, the resonant frequency increased. What’s more, when we look at the total antenna gain, we find the additional wire generated a fraction of a decibel more total system gain on transmit, resulting in slightly less loss.
My First Breakthrough
Having modeled the Cage Dipole on the lightship, adding these additional wires on the long side of the antenna? Suddenly, gain appeared. A lot. Especially on harmonic bands.
A noticeable fraction more than a dipole. Meanwhile, the SWR was well-behaved and the return loss was extremely high. This meant this design was radiating like mad…
What’s more, the major lobe going skyward (ideal for 80-meters daytime as NVIS) began to pull in and was now several decibels better than the dipole.
To a point, the more wires I added on the cold (long) side – spaced down 4 inches at their far ends from the one above, and not connected there, the better things got.
Going too far, though, it became a mechanical engineering problem, so there were practical limits emerging.
Offset Near-Field Parasitics
In pushing antenna modeling to either its (or my – we’re not sure whose) limits, I finally arrived at what will be this year’s Seven Magic Wires build.
What I discovered along the way (and lay claim to herein) is that an antenna’s gain and impedance can be remarkably modified with a very closely set yet not connected antenna element or two. Key thing: Parasitic tuning is NOT connected to the antenna system.
This is vastly different than the classic Yagi-Uda parasitic reflectors and directors of the Yagi-style beams. Different than directors and reflectors of other parasitic antennas, such as quads, top. These were much wider-spaced. Fifth of a wavelength and greater. This novel use of near-field parasitic elements gained control of, bandwidth and standing wave ratio (and thereby the return loss) as well as gain.
Thus, when you inspect wire placement in this model, realize your construction practices not only need to conform closely to the model, but you will likely need a lot of “antenna corona dope” in order to reduce or minimize fire risk from corona effects when running high power, over a few hundred watts of output.
The Wire Plan
I’ve taken to buying #14 THHN stranded house wiring in 500-foot spools. A single antenna with a few inches each end for connections is about 520 feet worth. (510 plus whatever)
Here are the element lengths and placements:
Remember these measurements are in feet. S0 0.3 feet is 3.6 inches. (.3 X 12″) Parasitics are 0.1 inch offset from the driven elements, so 1.2 inches. And on the long side, the vertical spacings are 3.6 inches.
Remember when building, the wires with nothing in the Conn(ection) column are the two parasitics of 32 and 66 feet (plus a fraction of a foot overlap). They are 1.2 inches off either side of the top wires on the long side. The overlap (-0.3) is critical for proper gain adjustment. An additional parasitic element could be added:
-0.3, 0 50.1 Z on the left and 66, 0, 50.1 Z on the right, however this becomes a bit touchy on tuning and the incremental increase in gain is partially offset by losses.
Literally days of modeling has gone into this. There are versions with much higher gains (+26 dBref or 28.3 dBi) but each had certain shortcomings. Might not be acceptable on 40 meters, for example. Or, the SWR on 20 was not smooth-enough (and 20 is my “all purpose band”). This is a balancing act.
It is in electronics what Ben Seaborn went through designing the Thunderbird sailboat class. A delicate and well-performing balance.
Just as you can design higher performance sailboats, this is not the be-all, end-all antenna. Such things are always a balance. Trade-offs.
Antenna Performance by Band
80 Meters first.
Resonance: 3.825 and the SWR is under 2:1 from 3.675 up to 3.975 MHz. Although it will work well across the entire band, thanks to the open wire/ladder line use, it will encourage me to work on my classic AM “boat anchor gear.”
Overall gain is 4.89 db (compared with -0.69 db for the plain dipole in the same setting) while the maximum lobe for NVIS is more than 10 dBref (dBref is isotropic (dBi) minus 2.15 dB). Around 12.81 dBi if you want to think of it the dBi way.
When I get this built, I’m not sure what to expect on 80-meters on long winter nights. Typically, low take-off angle antennas are the cat’s meow. I’d expect this to perform OK. But daytime and early evenings the high gain NVIS should let it outrageously outperform less highly developed antennas.
I’ve mentioned (*I think it was in SuperAntenna IV) that highly evolved antennas almost have a kind of “traction” on the air. Quite remarkable to experience. Like an 80-foot high flat-top I had years ago that would consistently out-perform other USA signals in Longbranch, New Zealand late nights on the 75-meter phone band in the winter. 20 dB over s-9!.
40 Meters Begins to Scream
Right smack-dab in the middle of the 40-meter band you have resonance at 122 ohms and some inductive reactance. The return loss is 11.8 dB and that means 91-92 percent of the signal goes out on first bump. OK – I tried for better…
Before you go off whining about the SWR figures, hold on and look at the gain in the plot below. Average gain is? 5.45 dB. Almost a Six times power multiplier over a transmitter using a simple dipole! But wait until you look at the pattern of the 40-degree take-off angle!
The green max lobe pointer is just about on the Major’s heading from our place in the sticks. To put this into perspective, it potentially could multiply power by about 15-times. And receive by as much, too. Superior to any amplifier lash-up!
Remember, when dealing with power, 3 decibels doubles your effective power output. 100 watts out thus becomes 200 at 3 dB gain, then 400 watts at 6 dB gain, then 800 watts at 9 dB gain. and then 1600 watts at 12 dB gain. We miss that by only a bit.
Now Be Amazed at 20-Meters!
This is breathtaking. The worst SWR is better than many hams best SWR in life:
OK…about perfect. Centered exactly in the middle of the Extra Class DX phone portion of the 20 meter band. Yet great gain and return loss/low SWR to the edges. Then it gets even better:
Behold the gain figures. Over a dipole? Here’s how the overhead view of 20 degree take off angles looks:
This is as much as twice the “antenna power” as people with modest 3-4 element beams come up with. It’s an astounding little bump in physics to exploit. 100 watt transmitters ought to get out like 1,600 watts. Better? Super hot receiving, too, since antennas work on transmit as well as receive.
All that’s needed? Add additional wires on the cold side of any OCFD and then calculate and carefully model offset near-field parasitics to get things under control. Hint: Y Axis is critical!!!
Let’s say you have open wire line feeding this antenna and you wander up to 15-meters: 21.225 the return loss is only 4.7 and the SWR is about 3.75 to 1. (Remember though, SWR isn’t much consideration on open wire or ladder line!).
The gain and lower (15-degree) take-off angle is dandy here, as well! And the gain? OMG…more than 14 dBref which is more than 16 dBi.
Parasitic Elements: Gain and SWR
Since I don’t think anyone is using near-field parasitics as a gain, SWR/return loss, and resonance tool, a few words on the spacing matter are in order.
The first fact is that parasitics (like adding additional wires on the cold (long) site of an OCFD can increase – or decrease – both system gain overall, as well as maximum lobe power density.
In the process, it will also change (somewhat) the cloverleaf patterns on 20 and 15 meters. Essentially, you can eek out more gain (a dB or 5) but in the process the directional lobes become sharper. Also, the SWR changes.
Some deeper discussion of the parasitics is therefore to our advantage.
In the wire table, you will see two wires – one 32.3 feet in length and one 66.3 feet that are offset (insulated) by 0.1 foot (1.2 inches, but likely 1.25 inches will work fine).
Since 40-meter SWR can be particularly cantankerous with this type of antenna, we can focus on parasitic spacing on that band in order to “learn quickest and mostest!”
You can see in the 40-meter section above that 0.1 foot (1.2 inches) results in a relatively even, though somewhat inductive reactance of +j 106.7 ohms. (I’m skipping the Smith chart and pocket vector network analyzer class to prevent electrical malpractice suits!)
When we increase the parasitic spacing to 0.15 foot (1.8 inches) look what happens to the SWR and return loss! Down to 1.57 to 1 and the return loss increases to more than 13 dB. So, why not call it good and use that?
The short answer is because the overall system gain drops from 5.4 dB to 4.61 dB. I might be able to argue that’s OK, but such a small tweak really ripples throughout a complex antenna like this.
Because, what was providing maximum lobe gain 11.24 dBref (which is 13.39 dBi (isotropic) decreases to 10.45 dBref.
Before making the judgement call on this kind of thing, we next need to assess how this antenna will work on 20-meters. Because this is my “go to” band year-round that 41.1 dB of return loss and damn-near flat end-to-end SWR impacts negatively.
Before (from the azimuth plot for 20 meters) we see the maximum lobe is 13.23 dBref. After – moving the parasitics away drops gain to 12.8 dBref. There is also a fraction of a dB lost on 80 meters, as well. Thus, the version presented in the wire table is the highest gain commensurate with three band operations. It’s essentially taking either the highest average gains (adding) or the highest maximum lobes and adding those.
That’s the case of moving the parasitic out a bit. What about going the other way? Say one inch offsets for the parasitics? (0.083 foot)?
In theory, this will pick up some gain on 20 meters and 40 meters, but the SWR (and thus lower return loss figures) might result.
In this configuration, the best SWR/return loss is 56.3 dB – which almost sucks the RF into the sky! Unfortunately, though, proximity also effectively shortens the antenna so its best frequency is “pulled” up to 3.825 MHz.
Our problem child is 40 meters. Where the SWR has risen. Ever so slightly above 2.0 to 1 at band edges. On the other hand, average system gain comes up to 5.96 dB which is great. And maximum lobes become 11.81 dB. Tiny fraction of a gain. BUT our azimuth review says the tiny bit of gain out toward the Western Pacific took a fraction from European cloverleaf headings. Hmm…
Last, the gain on 20 meters increases ever so slightly (on the order of a quarter decibel or less). But with closer parasitic spacing my concern (with perhaps 5,000 volts spaced one inch, that such a design might be more prone to coronal discharge.
The easiest way to tweak this is in EZ-NEC and use of that software platform is highly recommended along with the FCC’s data set of real local ground conditions.
Planning and Execution
Obviously, it’s going to take some time to get this thing built and up in the air.
There are plenty of projects around here to get done – all competing with the ham radio fun stuff. The new deck is just one of them.
There’s also some 3D printer design work involved. I will be generating some special spacers because proper spacers must be used throughout. Dimensions are critical in this monster.
Also: A reader suggested I use a plastic called ASA instead of ABS plastic. This threatens to become another “Just George following his brain around and getting lost in minutia for weeks…” It’s how I’m wired, I guess.
Not sure how my 3D printers will like ASA high temps, so measurements will be necessary. Specs call for Nozzle: 250 C; Bed: 90 C; Speed: 50mm/s. Pretty slow and pretty damn hot…Odds are high I will push back learning ASA printing and just go with black ABS which is in house so no lag time or learning curve on that one.
All that remains is cooler weather, some time on https://tinkercad.com to draw up the .STL files and updating of Cura to the build of the month in order to slice and print the spacers…
More as developments warrant. Shop order is in. So this moves to WIP (work in progress). First deliverables are the stereo lithography files for the 3D ABS spacer prints. And then, a tower-mounted jig so one ea. 72-geezer and figure out (and build up) where 508-509 feet of wire goes… Final piece is machining the end pieces from another Dollar Store high-density polyethylene cutting board.
Yes…this is what passes for fun at this age…but beats the livin shit out of a nursing home, lol.
General Shop Notes
There’s a new table saw out there that has caught my eye. It’s called a Sirocco Dustless. Been eyeing one. But with a radial saw in restoration, a table saw, and a chop saw, I’m having a hard time coming up with a reason to buy one. Not a bad price, though, and very cool saw.
But then I read about this stuff called Zero Clearance Tape. Don’t know if you’re aware of it, or not. Take a look at FastCap Zero Clearance Tape, 5 Strips, 2″ x 16″, Made in USA. About $11 bucks on Amazon.
Lower your blade all the way, apply the tape and press firmly. Then (Clear!!!) turn on the saw and raise the blade while it’s on. I’ve set the front edge of my tape well-forward (by the lip of the table) and it doesn’t catch. The tape is thick enough, it doesn’t seem to have an appreciable impact on my work (which is somewhere between rough and slipshod, depending on how much time I’m taking…). Things Elaine sees tend to look a lot more finished. Tough audience.
OK, back to the Laboring weekend.
Write when you get rich,