Tuesday, 6 September 2011

The W5GI “Mystery Antenna”

W5GI's “Mystery Antenna” gets a lot of attention. Possibly because of its daft name.

It looks like a G5RV, but W5GI (now SK) used coax stubs to give what he called “a coaxial colinear array on 20m”. It consists of a half wave length (at 20m) of 300 Ohm ribbon that feeds a dipole centre.

Either side of the dipole centre are two pieces of wire 16ft 6in long.These are then each connected to the inner of a 16ft 6in piece of RG58 coax. The outer is left unconnected. The far end of this is then shorted and connected to a further 16ft 6in of wire, giving an antenna with a total span of 99ft.

I've now been using one for about three months, so find out what I think of it, is it really a "mystery" and is it worth making one!

Download the report.

Update: I have been doing some more modelling with the antenna as an inverted V and if mounted at about 10m with the ends at 3m you'll find that the lobes vanish and it becomes pretty omnidirectional at higher frequencies. On 20m you may find it radiates best broadside to the wire and less off the ends - this seems to match what I have found. Plots for 20m and 10m Inverted V have been added to the report.

Further update: I have now included the original CQ magazine constructional info in the PDF. I can also confirm that the 50 Ohm resistor method I suggested to determine the exact half-wave electrical wavelength on 20m works a treat as I have just built another one.


  1. One would save a lot of time and headaches just putting up an 88' doublet fed with window line or even 300 ohm ribbon, especially on 20 meters.


  2. Very true - people seem to want to shy away from doublets and open wire feeder. I think it is because of the increase in automatic tuners with unbalanced output.

    Steve G0KYA

  3. People who are used to coax have a lot of problem understanding open feeder. It took me a while to work it out but once I got it, everything clicked.

    I set up a doublet at the local club which turned out to be poorly located and really not long enough and unfortunately the perception was that the feeder was rubbish. I heard comments such as "Why are we using this crap?" and "We should take this sh*t off and feed it with coax". I will add that we were using a true balanced feed tuner.

    The whole balanced/unbalanced thing can be quite difficult to understand. The natural way to look at open feeder is that "it must be rubbish because it's not screened, coax is screened so therefore it's better" but of course that's not true.

    Try getting that across though, it's a real uphil struggle. I'd explained it fully before I put up the doublet but still got those comments above.


  4. I built and tested the W5GI antenna. It soon became problematical. Coax would corrode, the length of the flat lead to get it to work seem to be a mystery. It sagged more. Wound up trashing it. Now I use a simple 110 ft center fed Zepp fed with 450 flat lead along with with an antenna tuner. This works much better overall and no maintenance problems.

  5. Where is the PDF of the original CQ article you said you have included? Link?

    1. Just download the report linked to in the story above - it is towards the end.