Es’Hail-2 Dual Band feed with patch tuning screws

I’ve made an attempt at building a dual-band feed for a dish (instructions & examples on this link). I had some thin copper sheet available. I just hand cut the sheet with metal shears. I used my step drill bit to drill the holes in the centres of the plates. I pushed the cleaned base plate and patch onto the piece of 22mm copper pipe and added a bit of liquid flux at the joins & some solder. I fitted an SMA socket, held in place by brass screws, tapped into the base plate and sawn short so that the ends of the screws were flush. Then I baked the whole thing on an electric hot plate & dabbed in a bit more solder. I let it all cool down.

The first attempt resonated around 2475 MHz (with a second resonance about 50MHz up on that), so somewhat high.

In an attempt to bring the resonances down to the Es’Hail-2 uplink band, I laid a couple of little fragments of copper to a couple of corners on the patch. This seemed to work, judging from my VNA. I soldered them on with the aid of a blowtorch. This gave the patch antenna a proper “bodged” look, unlike the pretty pictures of professional-looking antennas I have seen. Of course, one fragment floated & moved slightly out of position and then refused to budge under the blowtorch, so the result was still a bit high in frequency. With a bit more blow-torching, I further added a bit of brass (from a connector block), to hold an adjustment screw.


The solder blob on the right edge of the patch in the left-hand picture is the feed point from the USB socket.

The upper resonance seemed about right, but the lower one was still a bit high at about 2405 MHz. Feeding the signal generator into the patch, I set up the SG-Labs yagi a short distance away as a receiver & fed this into the spectrum analyser. Rotating the patch showed that the difference between the maximum and the minimum signal was about 5dB. So the polarisation seemed to be somewhat elliptical.

With some more blow-torching I added a second screw block and adjustment screw. That gave me some adjustment of each resonance.

With screw-twiddling I found I could get a good return loss or good polarisation, but not both, so I opted for a compromise. The elliptical distortion came down to about 2dB; RL (return loss) about -15dB, or better, across the satellite bands. Gluing on the red lens from a Rocket LNB for 3cm improved the return loss a little further. I don’t know why!

Whether the patch’s polarisation is right-handed or left-handed, I have no idea. Today, Barry G8AGN helped me by trying it out on his 1.1m dish, feeding it from an SG-Labs transverter. We could see a carrier coming back on the Goonhilly SDR about 10dB over the noise. I hope that means the polarisation is the right way round.

Update 4-April-2019: here is the patch with both adjustment screws. I have also bent back the ground plane plate in order to fit it in a nice box I bought from Toolstation. The box has no metal parts & it passed the microwave oven test, so I am optimistic the box is transparent to microwaves. The patch looks even more bodged now, but this is hidden in the plastic box.


The lid clamps down nicely with integral plastic screws. I used a step drill bit to make the hole in the lid for the red lens and in the base for a tight fit for the copper pipe waveguide. I added an N-type socket with an SMA tail attached to the 13cm antenna.


I sawed off the horn from the Octagon LNB. This left a short stub of waveguide. I filed the diecast case of the LNB, so the stub stood as proud as possible. I cut a short ring of 22mm pipe to act as a spacer in the 22mm compression coupler. I found the coupler clamped on to the LNB’s waveguide OK.

Another, smaller, Toolstation box gave protection to the Octagon LNB.

As a result of the edge bends on the ground plane plate and encasing the patch antenna in a plastic box, the operating frequency has dropped. There is now a very good match (RL < -25dB) for the whole of the 13cm band, 2300 MHz to 2400 MHz, rising to -15dB at 2408 MHz. I have not yet re-checked polarisation. So maybe I could have skipped adding the the corner fragments and adjustment screws & just bent the ground plane edges and put it in the box?

I’ve put some bleed-holes in the boxes so that any condensation can escape. I shall apply some sealant in order to keep the rain out.