simple trick to build a better POTY

  • during my work on the 3-helix-feed I made a lot of measurements to compare it to other feeds, including the POTY.

    I have build many Potys until now (for DH5RAE, DO1RX, DG6RCH, DG6FAX, DL3MX, DL1EV and myself), all worked OK, but had some drop in TX performance compared with a simple helix antenna. Something between -2 to -3dB compared with a helix (when monitoring my own signal via QO-100 in a spectrum monitor).

    Also I talked to many people on QO-100 who had similar experiences, mainly because it was very hard to tune the "two" virtual antennas to the correct frequencies and to see the two resonances in a VNA. This results in a poor circular (or even no circular) polarisation which explains the -2 to -3 dB drop.

    I have read the manual (written by G0MJW, PA3FYM, M0EYT) hundred times, but this week I focused on this remark on page-4 about the distance between the to copper plates:

    "Aim for 3.0mm, not about 3mm”.

    When building a POTY I use 3mm aluminium spacers while soldering the copper plates. But this always resulted in a distance of the finished (and cooled down) antenne of 3.1 to 3.2mm.

    This week I did it different. I sanded my 3mm spacers to exactly 2.9mm and used it to build the Poty. The distance was now exactly 3.0mm.

    What a difference !!!

    When connecting the antenna to my VNA, it was almost perfect without any adjustment. The two dips are almost exactly as shown on page 7 "modelled return loss". It tooks only a few minutes to squeeze the copper plates to a perfect match.

    Also I found that the way how the connector is mounted on the reflector has hugh influence on the tuning of the Poty. I do not use screws any more, instead I solder the connector directly to the reflector. Also the hole for the central wire must be small. If its too big, the few mm² of missing copper will disturb tuning.

    Today I mounted it to my Gilbertini offset dish (f/D 0.67) and the signal was almost identical to the 4-turn helix.

    I also did a few RX tests with different dielectric lenses because I found STL files on But none of these lenses worked for me. The only lense working perfectly is the cap of the Red Eagle Black Buster.

    Thanks to G0MJW, PA3FYM, M0EYT for this excellent POTY!

  • Hi Kurt,

    thank you very much for your interesting information.

    When I moved from my previous patch feed, which was based on the design of DJ7GP, to the POTY (after tuning it carefully) I measured about 2 to 2.5dB better signal on QO-100. Having measured the polarization of the DJ7GP feed to be rather elliptical, I assumed that the improvement from the POTY is because it is producing a proper circular polarization.

    Kind regards