I've always liked the OCFD, they're a simple multi band antenna for HF. One balun and some wire, and easy to tune gets you 5 or 6 bands.However conventional OCFDs present a set of trade offs. 80m and 40m resonance is low, and 20m and 10m resonance is high placing the low SWR points at the band edges. This is due to the harmonic relationship of the bands and the detuning effect the proximity of the ground has at lower frequencies, lowering the resonant frequency.
One can't quite get away from needing an antenna tuner of some description. We can make do with the internal "touch up" tuners in most rigs. I want something that's nicely matched so I don't have to fuss with tuners, or load the AL-80B into a high SWR. Tuner or not high SWR is still present at the balun which can stress the core heating it up, heat is lost power.
One of the more well known solutions is ON4AA's "CL-OCFD" design which moves the resonance on 80m (and optionally adds 30m too). This one locates a capacitor at the electrical center of the antenna, which only affects 80m resonance moving it up into the band.
K5GP's "A broadband 80/160 meter dipole" is another example of using a center loading network.
A July 2020 article in CQ magazine Multiband Off-Center-Fed Dipoles for 160 and 80 Meters uses a 20% offset (improves 15m band coverage), two capacitors in the 160m version, the 80m version uses one in the center same as CL-OCFD.
Focusing on 80m, the CL-OCFD fixes 80m but doesn't solve the rest, 40m is still low and 20m and up are still high.
I have modeled two possible solutions - model files here.
The first solution I call FPL (Feed Point Loading), making the antenna longer to bring 10m and 20m down, and adding 175 pF series capacitance at the feed-point (e.g. insert it between the long leg and the balun terminal) to pull up the lower bands by tuning out the inductive reactance.
The cap has a greater effect lower in frequency where it's needed more, and a diminished effect on the upper bands. This moves the resonances into the phone segments of 80, 40, 20, 10m, and 12m as a bonus. Unfortunately 17m and 6m resonance is still low in the model, can't win 'em all!
Per the model:
- 41.7m wire.
- 33% offset.
- 175pF cap at balun / feed point.
- 16m / 55 ft height above ground.
- 13 AWG, 0.5 mm PE insulated wire (The Wireman 531 material).
- 150 ohm feedpoint, ok with 4:1 current balun.
After a lot of searching, I managed to turn up a few references where a capacitor is placed at the balun:
- http://www.dj0ip.de/off-center-fed-dipole/80m-ocf/move-it-move-it/.
- http://www.andreadrian.de/sdr/Kurzwellenantenne.html#mozTocId272180.
- https://pa0fri.home.xs4all.nl/Ant/FD4/fd4eng.htm.
The second solution uses a compensation coil. I got the idea for this compensation coil from WA7ARK's End Fed Multi Band Antenna slides. By optimizing the antenna for 40m this shifts the upper bands further up and out of band. These can be corrected with a 1.7uH compensation coil placed in the end of the long leg, 6% in of the overall length. Getting this positioned right may result in some trial and error with out a good antenna analyzer. 80m resonance can optionally be moved by using ON4AA's OCFD-CL method of placing a capacitor in the electrical center.
Per the model:
- 40.7m wire.
- 34% offset.
- 300pF cap in the center (optional to move 80m resonance).
- 1.7uH compensation coil located in the end of long leg, 6% in overall total antenna length.
- 13 AWG, 0.5 mm PE insulated wire (The Wireman 531 material).
- 135 ohm feedpoint, use 2.5:1 current balun.
That's what the models say at least :-)
Balun Designs sell both 4:1 and 2.5:1 baluns designed for OCF antennas. I've bought several of their baluns over the years and they are very well made. However they can lack choking performance below 40m, adding a second CM choke below the balun is recommended.
The capacitor needs to be RF transmitting types or capable of handing the RF current passing through them. The cap should have a 1 to 5 Meg ohm 5 watt metal film resistor across it to protect it from static buildup.
Nov 2020, I finally got a compensation coil installed with mixed results which merits further experimentation..
- 10m came down around 500kHz to 28.8, so still 400kHz too high. I tried less turns, more turns, moved it between 6 and 8ft from the end, couldn't get it to come down further.
- 12m resonance below the band.
- 17m resonance below the band.
- 20m moved down about right, but SWR 1.5:1 across the band.
- 40m is about right.
- 80m resonance for some reason at my location has always been further up the band than expected and from previous experience with OCFDs.
I'm using my existing 4:1 current balun with this.
Oct 2021, Noticed SWR rising when running the amp on 40m (tuner off/bypassed). While troubleshooting I removed the LDG -RT-600 remote tuner I had inline about 60ft from feed-point. Put the analyzer on the antenna again, and found the resonances and low SWR points on 40, 20 and 10m had moved a lot! Long story short the RT-600 must have an impedance bump inside it? Which was why I didn't get the expected results from the compensation coil. Now it should be fairly easy to get the coil to do the trick.
However I decided I want agility across all ten bands from 160 - 10m, and be able to run the amp when needed. After research, the solution I have settled on is a 130 ft doublet fed with 600 ohm open line into a new 1 kW remote ATU via 1:1 balun. More about that at Doublet / T Antenna.
The OCFD has been retired, we had a good run, it is a nice simple antenna where the most expensive part is a good 4:1 current balun, the rest is wire and coax, gets you on several bands easily.
An analysis of the original Windom - Notes of Mr. Windom's "Ethereal Adornments" (by L.B. Cebik, W4RNL). An interesting read since what we might call modern coax fed OCFD and the more recent EFHW designs using 49:1 transformers evolved from the Windom.
I also found the original Windom QST article from September 1929 in the ARRL QST archives online, another interesting read.
Models are good starting point, and a way to investigate and better understand antenna systems. These tools can also help guide us to and validate the final result, if a good correlation is observed in the real world then we can have confidence the patterns and other information are accurate.
The models I have created and made available may contain errors, or overlook something someone more experienced can see.