Dual-band Feedhorn for the DSS Offset Dish 
5760 & 10368 MHz 

Paul Wade N1BWT

I recently completed a new transverter for 5760 MHz in a fairly small package ¾ it fits on top of my 10 GHz transverter next to the wedge which supports the RCA DSS offset dish. I designed a 5760 MHz feedhorn for the dish using my HDLANT21 computer program, built one, and modified the transverter slightly to allow for quick changing of the feedhorns with two wingnuts. Now I had a package, shown in Figure 1, for a compact two-band rover station.

I was wondering if it was possible to make a good dual-band feed when Dick, K2RIW, mentioned that WR-112 waveguide covers both 5760 MHz and 10368 MHz; even though the handbooks don't list it as usable for 5760, the cutoff frequency is slightly lower so it still works.

The next problem was designing a feedhorn to cover both bands with decent illumination for the dish. A few trial calculations showed that a 10 GHz horn providing -10 dB edge illumination taper would provide a -3 dB edge illumination at 5760 MHz -- most of the energy would miss the dish! On the other hand, a horn designed for 5760 would have a much narrower beam at 10 GHz, so the outer portions of the dish would receive very little illumination energy; only the performance of a much smaller dish would result. After some fiddling of the numbers, I found a compromise which might have the same loss of efficiency at both frequencies.

The final design, using the HDLANT21 template shown in Figure 2, has an illumination taper of roughly -16 dB at 10.368 GHz, so it is somewhat under-illuminated, and roughly -5 dB at 5760 MHz, somewhat over-illuminated. I adjusted the horn length to match the phase centers at 10.368 GHz, since it is most critical at the higher frequency.  Here is the template for the dual-band horn.

The next problem was getting a good VSWR at both frequencies. The surplus WR-112 waveguide-to-coax transitions I had weren't very good at 5760 MHz, so tuning was required. I put a small ball bearing inside the waveguide and moved it around with a magnet on the outside until I located a spot which improved the VSWR at 5760 MHz without making the 10368 MHz VSWR too much worse. Then I marked the spot, drilled and tapped the waveguide, and put in a tuning screw. Next I adjusted the screw for best VSWR at 5760 MHz, then put the BB back in and looked for a spot which improved both frequencies. A second screw was added here, then both screws adjusted for a compromise with reasonable VSWR at both frequencies. The final tuning had a VSWR under 1.6 at both 5760 MHz and 10368 MHz, but it is not a broadband match.

Does it work? YES!

I completed it just in time for sun noise measurements at the July 1997 N.E.W.S. meeting, and tested it there on 10368 MHz. The DSS dish with a single-band horn feed has an efficiency better than 60%, while the dual-band feed is around 50%; the gain difference works out to about 1.2 dB.

The next day, I set up a sun noise measurement at 5760 MHz, with similar results: the DSS dish with a single-band horn feed has an efficiency of about 60%, while the dual-band feed is around 50%; the gain difference works out to about 1 dB on this band.


An RCA DSS dish with this dual-band feedhorn provides two band performance only 1 dB down from a single band feedhorn on each band. I've never seen a multiband feed with performance this good. This compact antenna is ideal for rover operations.


Q - Is a tri-band feedhorn possible?
A - Not with ordinary waveguide, which cover a frequency range of less than 2 to 1 between cutoff and an upper frequency where other modes can propagate. Ridged waveguide can cover a wider range, but the horn design involves even more compromises.

Q - Is a dual-band horn possible for lower bands?
A - Yes, with a larger offset dish. A dish should be at least 10 wavelengths in diameter for good performance, so the 18 inch RCA dish isn't big enough below 5760 MHz.