Wide Angle Multi-Radio Coverage via Simulsat / Offset Parabolic? 1

[blueandwhiteg3]

I'd like to cover a wide area with many independent, very high gain radio 'beams' so to speak. I am basically shooting towards the horizon in all directions, so the beams' "height" only needs to be a few degrees in most cases. The key is getting enough beams to shoot around part or all of the horizon.

One approach is to simply install many parabolic dishes in a ring, shooting towards the horizon. However, this seems like a wasteful use of space and resources, plus it is ugly, wind load is a concern, etc.

One solution I have been considering is the use (and perhaps design and production) of a simulsat-type shaped parabolic. With this and multiple offset feeds, it seems like I could use quite a few offset feeds and serve a wide angle. See what I mean here:

http://www.simulsat.com/

Looking at the simulsat site, it looks like they produce dishes which can serve up to 70° vertically and 152° horizontally. This is amazing, as 3 of these dishes could serve a full 360° radius around my site.

Take a look at this diagram:

http://www.simulsat.com/engineering/CA, San Jose Simulsat-5 C-Ku Configuration.jpg

I will be operating at the 5.25-5.825 GHz range for both RX and TX. Each radio will use a unique channel. My goal is to achieve 30-33 dBi gain range, possibly as much as 35-36 dBi, but such dishes begin to get quite large at that point

I'm trying to learn here, so I would really appreciate links, ideas, suggestions, conceptual introductions, etc. Once I get my feet wet properly, things tend to rapidly gel, but sometimes that first dip is very hard!

What I need to learn more about is:

- What is the practical size required for a design like a simulsat to achieve the gains I want? How does the horizontal and vertical coverage goals affect the size?

- What kind of math or design model is used to define the shape and size of a simulsat-type dish?

- With a design like this, could I simply alter the feed to trade-off gain for beam width? Some beams I might want a high gain, narrow beam, while others I might like a large beam, lower gain.

- How much can the shape of the beam be modified by the feed alone? For example, I might prefer a wider beam on some feeds, and a taller beam on others.

 

[VEBill]

Combining extremely high gain, multiple beams and feeds, covering all azimuths - yikes.

The basic problem with trying to invent a new-design single ring-like reflector to cover all directions (with ONE ring-like reflector) is that one dimension (H) would be essentially dedicated to the all-azimuth coverage leaving essentially only the other dimension (V) to try to achieve 30+ dB gain per feed. It's almost a contradiction.

Your instincts to build on the simulsat concept (using 3 or 4 dishes) is probably the right decision. It's more-or-less exactly what you want, just with 3 or 4 quadrants.

This stuff all scales with wavelength, so you can work out much of the conceptual design using just that.

Standby in case anyone else has input.

 

[blueandwhiteg3]

Thanks for the reply. I'd love to hear more specifics about the angular concepts involved with these kinds of designs, so I can wrap my head around the core ideas. Once I do that, the math and design can be worked out.

Simusat says their products simply are not suitable for transmit purposes, due the dish shape not properly focusing outgoing signals. So a design like they are using is just not going to do the trick.

Per their suggestion, I did find another company who's product may be more suitable for this application, but I am still waiting for more data on their product. The indications are that the dish is capable of transmit and receive due to the more open shape, but it's not clear the transmit performance.

Take a look at the product in question:

http://www.gdsatcom.com/Antennas/Data_Sheets/655-0037B_7mTorus.pdf

Now, this thing is way too large, and the gain values they are showing are also way more than I need, but the core design may be of interest.

If somebody could get me pointed in the right direction with modeling the angles, shapes and sizes here, it would be extremely useful!

 

[blueandwhiteg3]

Also, I would appreciate it if anybody could warm me up on the concepts behind a ring reflector. I can see how it might work, and I'd tend to need a narrow feed. I am not too concerned about excess signal sidelobes in this application.

I know it probably gets really messy, and may be wholly impossible or non-viable, but I'd like to at least understand why for myself.

 

[VEBill]

"...not properly focusing outgoing signals..."

Reciprocity of antennas (including reflectors) is considered by many to be a fundamental principle.


For the sort of basic design concept information that you're looking for, you might want to hit the USPTO website and start looking up patents. Some commercial products will even mention the patent numbers which makes finding the info trivially easy. Otherwise, you'll have to apply your Googly skills to the Advance Search option on the USPTO site. It sometimes helps to guess the Assignee name (often the company making and selling the product). The downside of patents is that they're often intentionally written in gibberish. The images are often more entertaining (you'll need a free TIFF viewer). It's not a one-day job.

 

[blueandwhiteg3]

VE1BLL:

It is rather hard to argue with somebody who claims to be an RF engineer with multiple degrees.

I am going to continue to look into patents and see what I can find. The problem is finding anything in the huge volumes of patents out there.

Moreover, the patents explain the theory, not the best implementation. I know all too well how this works, after having filed a few...

What I'd really like is a good implementation design I can copy or scale!

 

[blueandwhiteg3]

Take a look at this PDF:

http://ieeexplore.ieee.org/ielx5/5654/15201/00702071.pdf?arnumber=702071

Is this what I want?

 

[VEBill]

"Is this what I want?"

We're branching off into deeply philosophical questions...

;-)

 

[Higgler]

everything changes when gain increases 6 dB (going from 30 dBi gain to 36 dBi gain antennas).
Your system analysis and link budget better set an accurate gain number since this gain number changes your antenna type.

If 30 dBi works, why not 28 dBi?, now it's a standard horn and you can shape the beam size by adding length to these standard horns (maybe a lens inside too).

What about; cost, footprint area, antenna polarization, sidelobe levels, future variability needs, schedule?

You haven't mentioned cost and maximum footprint area.
k

 

[blueandwhiteg3]

VE1BLL: Hahaha... thank you for your sense of humor... and patience...

Higgler: Would 28 dBi work? Perhaps. We have the potential to use a variety of antennas.

Motivations are primarily to minimize the total outward facing area required for transmit and recieve for purposes of reducing total size of remote site. e.g. an army of many individual, large parabolics is not preferred.

Depth is not so much a concern, though 10 feet is getting impractical, hence my aversion to some of the horns we discussed. I find the lens concept interesting, but without some kind of a design layout starting point (and hopefully math showing that it works) it's rather tricky to begin to approach.

I want to be able to feed my choice of horizontally or vertically polarized signals, possibly both simultaneously in a MIMO configuration.

Sidelobes are not a huge concern. We have a ton of bandwidth and an environment where multipath induced by sidelobes won't be too bad.

As for cost, less is always better. I do not believe anything on the market exists which exactly scratches the itch I have here, so I've been wondering about designs for doing something custom.

 

[Higgler]

You're asking if 28 dBi gain would work?
It might, but there's alot of math (often combined with some measured data) that one has to perform before saying what works and what doesn't. Many times you'd have multiple people do the same work and compare results.

antenna gains are usually derived from an exhaustive systems analysis that includes alot of unknown and estimated multipath and fade margins.

k

 

[blueandwhiteg3]

I looked more closely into the issues here. The concern of the RF engineers is transmit sidelobes. Some of the two-way satellite services heavily re-use spectrum and they say you can easily interfere if your sidelobes are not well controlled.

 

[Higgler]

reflectors have inherently low sidelobes compared to many other antennas.
The energy in the center is always 10 to 12 dB stronger than the edge. That amplitude taper gives you a lowish sidelobe. Hence a friendliness of reflector antennas for satellite communication.

k