Reverse engineering the Folded Skeleton Sleeve Antenna

Recently I saw someone online mention the Folded Skeleton Sleeve antenna. This was a design that I’d not previously encountered, so my curiosity was piqued.

The design includes two elements, the first (low band) is a partially folded dipole that is connected in the centre to the coax feedline, while the second (high band) is not electrically connected and instead it uses parasitic coupling.

The original designer of this antenna was Joel Hallas, W1ZR (SK). It was described in the May 2011 edition of QST magazine, and also in the ARRL book “Small Antennas for Small Spaces“.

This design provides several advantages:

1. A fair impression of a single wire dipole that covers two UNRELATED bands.
2. A reduced length relative to the halfwave size of the low band element.
3. A primary harmonic radiation pattern for both the low and high band elements (i.e. no lobes).
4. Modest gain on the high band element relative to a single band dipole.

One disadvantage is that, though there are some example dimensions for a range of band pairings (refer to the diagram), there is no calculation (that I could find anyway) to be able to derive the dimensions for novel band pairings.

So I decided to reverse engineer the design.

My first observation is that the total length (A) reduces as the ratio of the low/high bands increase. That is, the (A) length is less for the 80/10 pairing than for the 80/40 pairing.

I graphed those ratios against the (A) length as a percentage of the low band’s half-wave length and saw there was an obvious trend.

I started to dust off some neurons and manually derive the slope (y = mx +b) but rapidly came to my senses and used a spreadsheet instead. This gave me the formula to derive the (A) length for any given band pair ratio.

My next observation was that the total length of the low band element, being (A) plus the two folded ends, increases as the ratio of the low/high bands increase. That is, the (A)+Fold length is greater for the 80/10 pairing than for the 80/40 pairing.

I graphed those ratios against the (A)+Fold length as a percentage of the low band’s half-wave length and again saw there was an obvious trend.

Using the spreadsheet this gave me the formula to derive the (A)+Fold length for any given band pair ratio.

Finally, although the (B) length, being for the high band element, didn’t appear to vary with the band pair ratio, I graphed it anyway just to be sure. As expected this doesn’t vary and I can safely use an average instead.

This gave me the formula required to model new band pairings for a folded skeleton sleeve dipole antenna. I expect these are “good enough” to build from, but obviously haven’t all been tested. As with any antenna some adjustments will need to be done to allow for local environment – and for possible inaccuracies in my model.

Testing it against the “known good” examples reveals a reasonable match. There is some variation as expected since the originals weren’t a perfect fit to the trend line. I also made assumptions as to the target frequency within each band, very likely different from those of the original.

Here is the spreadsheet if you want to model your own band pairs.