9 band HF Fan Dipole Antenna System (modification of EAntenna dipoles)

There is a long story behind this project, which took 2 months non-stop to complete. It all started when I decided to install an EAntenna Fan dipole for main HF bands, and later to connect it next to an EAntenna for WARC bands, using only a single coaxial line from the shack.

Start of parentheses.

As a background information, from the very beginning, the tuning of the EAntenna (Spanish design) dipoles was problematic – at least for my conditions. I haven’t found any information on the internet about these antennas or whether they work as they promise, but for my case I was on the unlucky side. From day 1, these EAntenna dipoles couldn’t work, no matter how well I was following the instructions by the manufacturer. Trimming and adjusting the dipoles to resonate on their pre-determined band, couldn’t be tougher than anything else. The manufacturer’s suggestion that the dipoles should be trimmed no more than 5 cm maximum on each, side failed completely. I had to shorten them as much as 35cm on EACH side for some bands to work. Then on the other hand, the classic λ/2 equations didn’t work either – write numbers on the paper, trim the dipoles, then measure: Resonance was showing completely OFF the calculated band. As another practise, adjusting length in kHz per length e.g. 50kHz per cm, didn’t work either! No matter what someone was going to try, there was no way to identify the ‘magic’ length that would make the each dipole resonate on its band! I’ve made dozens of dipoles in the past, but I’ve never came across to something like this.

Well well…it turned out that the BALUN designed for this antenna system was useless. Yes, the balun! How come?! It all started when I observed (to my horror) that the coaxial line (high quality RG-213) was actually influencing the SWR and resonant frequency on any band as I was moving the coaxial close or away from the mast and the antenna! How come this could happen when it is supposed that the presence of the BALUN was actually responsible for avoiding this problem? At that point it was obvious that the balun wasn’t cancelling the unbalanced currents from the coaxial into the balanced currents antenna and vice-versa. The purpose of the balun is exactly to do this: To isolate and minimise the effects of the unbalanced currents of the coaxial cable from the antenna. The fact that the location, length and orientation of the coaxial cable near the entry point of the dipoles was influencing the SWR and frequency of the dipoles severely, was a clear evidence that the balun wasn’t doing the job it was supposed to do.

As a side note, surprisingly, the antenna behaved the same bad when I completely removed the balun included with the EAntenna dipoles. It was obvious that the balun was not effective. So, I got rid of that balun. See on this article its interior.

The problem solved when I later constructed my own ‘ugly’ balun. The SWR across all bands wasn’t changing when the coaxial length or orientation was changing – what a relief! I later ‘upgraded’ the ‘ugly’ balun into a home-made 1:1 voltage balun (article on this to appear soon). So, I settled down using these parallel dipoles by EAntenna, but using my voltage balun. Both solutions (use of ‘ugly’ balun or voltage balun at the feed point) worked well for me, the tuning of the dipoles became more predictable. And I’ve applied them for my EAntenna 8040201510 bands (main) and the EAntenna 3017126DX bands (WARC).

End of parentheses.

So, now, I can start saying a few things about the purpose of this article. Having dealt with the balun problem described above, I was experimenting with ideas of how to connect the two EAntenna fan dipoles systems (one for HF main bands and one for the WARC bands) with my shack using one coaxial cable. My HF Transceiver had 2 UHF connectors available, so it wasn’t necessary for me to start this project, but for the shake of curiosity I went ahead: How could I (perhaps) save 1 out of 2 of the coaxial lines? Was it possible?

My idea was the following: If a multi-band fan dipole could be constructed for 5 bands, why not for 9 bands as well?! So, I started connecting the feed points of the 2 fan dipole antenna systems in parallel in various ways to experiment with. Firstly, each with its own voltage balun at the input after a simple T-split from the main coaxial cable. This didn’t work, I couldn’t tune the dipoles properly, although some of them were working on their bands. The problem solved further when I did this: A single voltage balun feeding ALL dipoles at a single point. And that worked (almost!). What I also did was to reduce ‘interference’ (i.e. interactions) between bands, was to position one set of dipoles away from the other in the form of a cross (+) as you view the antenna from the sky downwards. This approach solved nearly all problems. I was now able to tune each dipole on a predictable manner.

Except for the 80m band*, where a loading coil was used in order to cut-down wire length (I don’t have space on my roof for a 80m full-length dipole), all other bands were finely centered into their corresponding bands with the worst SWR at 1:1.75 for one side of the band for the 20m. No need for an antenna tuner! So, all in all, these were the the bands of operation: 80-40-30-20-17-15-12-10-6m (9 bands, 8 dipoles with 1 trap inductor for 80m shared with the 40m dipole wire).

*For operation on the 80m band, an antenna tuner was necessary, due to the shortening by the coil trap. The bandwidth of the 80m dipole was too narrowband to cover the entire 80m band, with worst SWR in the 3.5:1 level. The best I’ve managed with this band was to make the antenna resonate at ~3.570MHz with SWR=1.2:1, but moving away from this frequency the SWR deteriorates rapidly.

Photos of final functional 9 band HF fan dipole system is shown below:

 

Alt text goes here
Caption: EAntenna 8040201510m for ham radio bands next to a 3017126m dipole from EAntenna. (Both antenna systems have 1:1 current balun at the entry point but these were later became redundant – see text above for details. The photo is shown for illustration purposes only, to show readers the