Just finished testing a prototype unconventional link dipole in the back garden antenna range: covers 20/30/40m, with just 2 links rather than the usual 4 to cover 3 bands. Now that my fingers have thawed (it was 4˚C out there) I can write it up.
I discovered this approach when experimenting with a link dipole for 40m and 80m, and found that if I connected just one of the links it was resonant near 60m. Since I recently acquired a QRP CW radio that covered 40/30/20m, this seemed like something to try.
I started with a 20m dipole (5.1m on each side) and put a link on each end. On one side I added another 4m of wire, so the whole thing resonates on 30m, but fed off-center. On the other side I added 6.2m of wire, which resonated on 40m with both links closed.
With the center at 6m and the ends at 3m, the SWR was over 1.5 : 1 on the two higher frequency bands, but under 2 : 1 across the band. On 40m the SWR was 1.2 : 1 at the dip, and also under 2 : 1 across the band.
One downside is that common mode current is worse when an antenna is fed off-center. I noticed this particularly with the original 80/60/40m version. I’m using a pair of #31 ferrite tubes with ~4 turns of RG-178 through them for a balun. Haven’t tried this one without it yet, but I suspect that the SWR curve will shift around with the cable placement with no balun.
Specific article is here: unconventional link dipole for 20/30/40m – Practical Antennas
General article with sample dimensions: A Link Dipole – Practical Antennas