A Brolly Dish for AO-40

Created 31 August 2001

Updated 2, 3, 7 September 2001 - Constructional detail and waterfall patterns

See also: Portable version

Here's an antenna I've been trying to make work for a while: a portable high gain antenna for AO-40. It's work in progress, but it turns up a lovely strong signal, even with  the linear polarization shown. Certainly better than the 60cm dish. This one's 1.2m.

Construction is on a composite (ie, non-conducting) frame umbrella. It's a stressed dish using the tension of the aluminium fly screen itself to maintain shape. But, before attaching the fly screen, you need to do a bit of practical geometry.

When selecting an umbrella, consider the following:

o    The support struts (main struts it don't matter) should be fibreglass or non-metallic. I used a Fulton Technoflex which has a fibreglass frame.

o    With the umbrella open, the main struts should come out at 90 from the boom

o    Choose an umbrella which isn't too deep. Take a ruler and a calculator with you and calculate the f/D with the umbrella unmodified. You want over 0.25 if you can. Smaller than 0.25 will probably be difficult to feed effectively.

f/D = D / (16d) where D is the Diameter and d is the depth.

Start by removing the original fabric from the umbrella. The stress of the original fabric will probably not be particularly close to a parabola. With an engineering trick we'll rectify that.

When defining the parabola, we already have two constraints: one point at the base of the boom where all the main struts emanate, and a second point where the support struts join the main struts.

You can make a parabola by temporarily using a stressed dowel (or similar bendable material) the same length as the strut as a template. The dowel has a single constraint at the boom base (at 90, like the existing struts). Stressing the far end of the dowel towards the boom, you can find the correct parabola by matching the second constraint, where the support strut connects to the main strut.

If you don't use the temporary dowel as a template, you'll never know whether you have a parabola because of the support strut constraint.

Making a parabola in this manner is based on the theory that for 'small' deviations, the stressed dowel will follow the line of a parabola. We may be taking the definition of 'small' to its limit here, but empirically the results are good.

When you have the correct stress on the dowel, measure R, the perpendicular distance from the boom to the far end of the dowel. Now you know R, you can discard the dowel template.

Now you know R, you can calculate the length between strut ends, L:

Take some nylon string (or similar non-stretching material) and mark out 9 marks starting about 10cm from the end, and then 8 marks separated by length L. Leave another 10cm or so at the end and cut the string - total length 8L+20cm. My umbrella had small holes at the strut ends, so with the umbrella closed, I threaded the string through the holes and joined up the start and end carefully in a knot which would give a circumference of precisely 8L in length.

Open up the umbrella (do this outside if you're superstitious) and match the strut ends to the eight marks made on the string. Now you have eight perfect (!) parabola sections.

Now make up a template using cardboard or paper for the eight umbrella segments, by marking the paper as it's dressed over the struts. Add at least 100mm to each of the measured dimensions - you can trim later. This should look like this:

Now cut up your fly screen segments (you can get this from pest control outlets). MIND YOUR FINGERS! That aluminium is very sharp. What's very important, in order to maintain stress, is that the mesh weave is kept parallel to each edge L. My mesh was supplied with a double seam along the edge and so I used that as edge L.

Attach the segments to the struts. In total I used a total of seven attachments to each strut. The more the better! I used solid copper wire as shown to make the attachments, but fishing line may be a better choice. Try to keep the attachments in line with the weave in order to maintain stress.

Try to make an even well stressed dish, but don't waste too much time trying to make it perfect.

Once the fly screen segments are all attached, remove the string stressing the dish.

I spent about five hours in total on the whole dish structure, but a lot of that was working out the logistics which I've documented here.

My dish's f/D was 0.27 - quite deep. depth=270mm, Diameter=1160m, f=310mm.

I also modified the TransSystem AIDC feed with a reflector made of PCB behind the dipole as shown. This made mounting a lot easier.

A note about closing the umbrella: I couldn't get the umbrella to close fully. Indeed, if you force it you'll permanently disfigure the surface, so be careful. 

Taken at squint 15, range 50,000km. S/N on S2 beacon 33dB. SSB 23dB S/N. As armchair as you're going to get.

Mail Howard, G6LVB

Home