The title may seem to be a contradiction but I assure you it's not. The design was spurred from wanting to operate as many bands as possible while keeping the number of antennas to a minimum, and everything had to fit into a plot of only 16.5 X 17 metres.
This antenna was fed from a FT101ZD and as my interest is only in emergency frequencies (e.g. 3695 and 7070 in South Africa) the need for wideband operation was not a prerequisite.
The theory of loaded dipoles is unfortunately far too involved to cover here but lets just state that loaded dipoles (using inductors short of the ends) electrically lengthens a short dipole so as to 'look like' a full dipole of one half wavelength.
These inductors are strategically placed and used as chokes thus making the higher band portion of the antenna 'end' at their location. On the lower frequency they become coils that resonate with the tips (much like a mobile antenna). Therefore the trap specified is not the conventional one (parallel LC) used for 'blocking' the higher frequency to the rest of the antenna but rather a series LC forcing current to flow in the whole of the antenna at the lower frequency when required. This has a two-fold advantage over conventional traps:
Construction is easy. Apart from the usual bits for a normal dipole i.e. 1:1 balun, wire for the elements and polyethylene rope to tie off the ends, all one needs is PVC electrical conduit (I used 32mm), some enamelled wire (preferably 0.8 to 1mm) and some wire to create the tips for the lower frequency.
Firstly, determine if the higher band dipole will stretch your property (if not already in place) and determine the length left over to play with. In my case I had approximately 6 feet at each end with 6 inches to spare for tying the ends down with rope. The trick is to first create coils that will resonate with the tips now added on to the dipole, mine turned out to be 115 turns close-wound on the 32mm conduit. Some experimentation may be necessary and good guidelines are found in many antenna handbooks under the 'mobile antenna' section.
After these have been constructed and installed and you have a match fairly close to the required lower frequency (aim for a bit lower than needed), have a look what has happened to the upper frequency matching. Depending on losses and ground effects this could have moved in either direction i.e. up or down in frequency, depending if the coil & end combination is capacitive or inductive at the upper frequency.
Once this has been determined start pruning/adding the main dipole till the match gets close to the required upper frequency. Look again at the lower frequency (while pruning the upper frequency, the lower frequency may rise as the ends are taken further from the ground). Slowly bring the two frequencies to the points required.
The 40m bandwidth (SWR < 2:1) is well beyond the band limits with 80m being approximately 50 kHz (allowing one to move to an alternative when other so-and-so's insist on using the emergency frequency for chit-chat and putting your life in danger!).
The narrow bandwidth is also suitable for receivers that cannot tolerate high levels from nearby commercial stations e.g. working the top-end of the 80m band near the commercial 75m one. With the antenna having a very narrow effective bandwidth will assist the broad input receiver by saving the front end from overload.
This idea works well on a half-size G5RV for 80m operation or to get you active on 160m using your 40m or 80m dipole.
I hope you have as much fun (and less frustration) as I did building, and then working this antenna.