6 GHz parabolic dish materials 1

[blueandwhiteg3]

I'm exploring the idea of putting together my own parabolic dish(es) for special applications where off the shelf is not satisfactory. The transmitter would not put out more than 28 dBm around 6 GHz.

I would like some suggestions / ideas / input on the material selection and its response in terms of RF. One of the factors is my workability of the material.

Here are some of the ideas that have come to mind:

- Cold cast rf-neutral polymer resin filled with metal powder or particles. Magnesium might they might be a good choice for weight and settling issues. How effective? How much? What particle size?

- A thin layer of aluminum foil over my choice of rf-neutral non metal substrates. How thick for the foil? How effective?

- Aluminum sheet. Once again, how thick, how effective? Does the alloy matter?

- Copper foil or other metal foil - would something other than aluminum be more effective?

- Soft steel sheet - how does this compare to aluminum in terms of effectiveness? How thick?

- Other metal sheets - is there something else I should consider, which would be workable?

 

[VEBill]

In general, it's not critical. The reflector simply needs to be a good conductor. Thickness and exact alloy are almost non-issues (w.r.t. RF reflectivity).

But I'm not too sure about the 'metal powder or particles' where they're not in contact with each other. Sounds doubtful since it wouldn't be a conductor.

 

[blueandwhiteg3]

Alright, so soft steel or aluminum foil backed by something would do the trick.

If I used metal particles or similar, there would be some contact between them but not extensive.

I wonder about some kind of paint on, or spray on, or spread on coating. Any ideas?

If a paint based mix was not suitable, I could probably make or buy an alloy that melted say a bit over 100 degrees C. I could apply this to a resin parabolic and get a nice layer the thickness of a foil.

Does the conductivity of the metal matter a lot? I mean steel veresus aluminum versus tin versus gallium... Big conductivity differences.

 

[Higgler]

what gain are you looking for?
You might consider making a horn instead of a dish? It's much easier mechanically if you have the room and your gain requirements aren't that high.

You can buy an old reflector rotting in somone's yard if you want a 6 foot diameter reflector.

Optimum reflector designs count on the feed antenna pattern to match the reflector shape. If you don't have the proper feed antenna, your reflector gain won't be efficient.

kch

 

[blueandwhiteg3]

Hmm... I had not thought about feed horns seriously, but that is an interesting idea. I have not really worked enough with them to have a feel for the sizing I would need to be looking at for my performance goals.

Upper 20 dBi range to lower 30 dBi range gain is the performance goal, and narrow focus is useful. Many parabolic dishes out there achieve suitable gain with a 1 meter range sized dish, or less.

I am looking to keep the size modest, all those 6, 8, 9 foot dishes are just too big and really, they're too powerful anyhow.

Yes, I know about the feed issues, and I will need to address this carefully, but dealing with a small piece of material efficiently is not as challenging as finding a way to deal with the rest of the parabolic.

 

[Higgler]

For a 30 dB gain horn at 6 ghz, you'd need a 20"x27" aperture and 100" long.

A 20 dB horn is very small and I've attached a link, less then 12" in all dimensions.

Hence that extra 10 dB gain from 20 dB to 30 dB gain is a big factor.

The nice part about a horn like this is you buy the adapter (or you could make it too), then the rest is sheet metal parts.

kch
here is a 20 dB standard in your band, it's nicely small and easy to build, they give you all the dimensions.

http://www.pasternack.com/Pdf/PE9860-20.pdf

 

[blueandwhiteg3]

Yeah, this could be a really attractive option for situations where lower gain is required, and it looks dirt-easy to build. I also think that a 100" antenna would be pretty viable for a few situations where high gain is needed, but surface area is an issue.

If the end of the horn was covered with a radome material, I am guessing the 100" version would actually have lower peak wind loading than a similarly sized parabolic, at least a parabolic without a large radome. And the length would mean that mounting for wind resistance should be easier. Ice could still be an issue, but limiting the radome surface area (and making it flat) would probably make active de-icing a lot easier.

One thing I am unclear on... the PDF you linked to states 20 dB gain from 4.9 to 7.05 GHz. I normally expect antennas to have gain increases as frequency increases. Any reason why this is not the case here? Do you have links to other horn designs, perhaps larger? 25, 30 dB, etc. would be nice to see.

However, in many uses here, a true parabolic will still be necessary, particularly where I need high gain and I don't have a lot of depth to work with.

Any further comments on the material options? Would it matter if a parabolic or feedhorn was made out aluminum versus mild steel versus zinc versus tin versus magnesium? Would foil thickness (30 microns) be as effective as a thick sheet of metal?

Can anybody comment further on the metal-filled composite concept? The metal particles would not be touching 100%, but the metal to metal contact would still be significant, particularly if I used more stick-like filings of metal in the casting resin.

I have also been pondering casting a dish from resin, then applying a low-temperature solder to the surface, almost like a layer of paint. I have also heard electrically conductive paints exist, which might be usable in a similar method, sans the heat.

 

[Higgler]

The gain does increase with frequency normally. But when you make a physically short horn, there is phase error across the aperture.

All horns increase in gain with frequency to a point where this phase error makes the edges of your horn radiate out of phase with the center and the gain levels off, then above that frequency the gain drops making a dip in the boresight antenna gain pattern and actually forms two beams equally spaced out from the center, and these two beams move away from center as frequency increases.

Hence, if you make the antenna length a certain amount, and look at the quadradic phase error across it, i.e. draw circle with radius straight down the center of the horn, then see how much longer the edges are than the center, when this delta physical length equals 1/4 wavelength electrical length, that is the frequency where the horn gain is peak. Gain then gets lower above that frequency.

Most horns have a height to width ratio of 68/51 to equalize the beamwidths in both elevation and azimuth. The shorter length is in the direction of the connector, or the eplane direction. The shorter dimension is the one to set at the quarter wavelength delta from center to edge.

On materials, some reflectors are made from very thin resistive coatings, which are inherently lossy, but minor in the overall scheme of loss. Loss is based on currents, and a reflector spreads the currents out over a very wide surface, and these currents are tiny. Hence reflectors can even be made of dense, non metallic surfaces that interrupt the wave. Aluminum or sheetmetal is ok too. Only high frequency waveguide (20-100 GHz) usually requires polished surfaces with gold or copper plating to minimize loss.

kch

 

[blueandwhiteg3]

This is all very good information, and I appreciate it very much.

This suggests that a painted on solder would be potentially viable, as would a metal-filled casting resin of some type.

 

[blueandwhiteg3]

Higger, can you tell me more about large horn antennas??

Nobody seems to make anything over 24 dBi around 6 GHz. I could get something larger whipped up here, but I don't have any plans.

Ideas? Directions? Plans??

 

[Higgler]

seems like 25 dB gain is peak for most commercial horns.

http://www.nearfield.com/Sales/sghorns.htm

has some dimensions. Alot of other people sell them.

You'll need to double the aperture size from 25 dB to get to 31 dB, but the length has to grow 4x that amount based on this formula below.

example, pick a nearfield systems horn, 25 dB gain unit, i.e. the sg75 unit, to add 6 dB gain, it'll have to be 5.94x2 by 7.94x2 and length is 18.9x4, or 11.88"x15.88"x75.6 inches long. That's for 15 ghz. You can select your frequency and change the size of this 30 dB gain horn just by the ratio of these sizes. i.e. if you want a 6 gHz horn, 30 dB gain, it's 15/6x longer in all dimensions than this horn. That gets pretty big. 30"x40"x189" (15.75 feet long). But you can shorten it alot if you can add a lens inside the horn.


the gain for any antenna flared in the E plane peaks when the phase error is 90 degrees (H plane flare at 135 degrees phase error is peak gain before it starts rolling off in gain), or;
(aperture size E plane)^2/(8*wavelength*length) = 1/4, Hence if you arbitrarily pick a dimension to make your horn, you can see the frequency at which that horn will peak in gain from this simple formula.


here's a 4 foot dish with feed from Andrew with 35 dB gain. I think their price is probably $5K, (recently I was quoted a $7.6K 15 foot dish from them).

http://awapps.commscope.com/catalog/product_details.aspx?id=1983

hence, a 25 dB gain horn is easy, get up to 31 dB and it grows fast in length.
kch

 

[blueandwhiteg3]

Higger,

Can you clarify on the lens concept? Are you talking about essentially embedding a small, parabolic-type dish within the horn antenna? If there were any kind of plans or diagrams, it would be helpful.

My primary interest in horns is the possibility of achieving greater gain over a unit of frontal area versus a parabolic, or similar gain over the surface area but taking advantage of the square shape to pack the antennas more efficiently.

Does a lens allow a reduction to the frontal surface area?

The horn design you linked previously certainly exceeds what I can do with a most parabolics within the same amount of space. However, when I ran the numbers on a 30" x 40" 30 dBi horn, the gain per unit of surface area was appreciably lower than a traditional parabolic, or even really crude grid parabolics.

As for the Commscope.com product, it looks like a regular parabolic... are you just pointing out they use a small horn as the feed? I don't see this mentioned anywhere on the page.

 

[Higgler]

the lens only allows the length to be reduced by correcting phase error as the wave propagated down the horn length. It will effectively shorten the length of the horn.
There are some short horns 2-18 ghz with a fresnel lens in the front.

The lenses get heavy of course. You can make a dense foam lens, any lens helps straighten out the phase. More dielectric in the center, less on the edges to delay the propagation down the middle.

those are regular parabolics and come with a feed antenna, hence the designated bandwidth. Simple parabolic reflectors don't have a bandwidth spec. typically.

kch

 

[blueandwhiteg3]

Higgler... that's just fascinating. I knew this was possible, but I did not know of any really practical applications of microwave lenses.

Do you have any design examples of lenses, etc? I really can't begin to see precisely what I'd have to prepare here...

 

[Higgler]

here's a good link that's a production product using lenses just as we've been discussing.

http://www.flann.com/Products_Home/Antennas/Lens_Horn/lens_horn.html

here's their list of antennas.

http://www.flann.com/Products_Home/Antennas/Lens_Horn/FmiCat078085.pdf

they start at 7.9 ghz and go way up in frequency. 46.5 dBi gain max. at the highest freq., but only standard gain 22-25 dBi at the lower frequency and you need to get up to 18 ghz to break your 30 dBi gain barrier. I think the lenses get too heavy for lower frequency large horns, so mechanically they don't even bother trying to make them.

If you were to attempt this, I'd suggest dense foam as a dielectric to maybe shorten the length by ?x%? and the foam can be a structural support for your sidewalls. Could be a fun experiment.

kch

 

[blueandwhiteg3]

Higger, that's a great example of a commercial product. It is doing exactly what I wanted - being a superior solution versus a parabolic.

Do you have any more detailed diagrams or plans on the actual dielectric lens mechanism?

I have optical fresnel lenses made out of acrylic, so I get the concept, but I have no idea where to start as to the exact material or design pattern scale I'd be considering. If there was a diagram, a sketch, plans, etc. it would be really helpful.

I really don't know if this is at all viable or not for my uses, but you have me really curious!

 

[Higgler]

I only know the electrical shape will be teardrop.
You have to delay the energy in the center of the antenna more than at the edges to equalize the quadraditic phase error, thereby making a flat phase front coming out the antenna.

You could try to make the lens from one flat slab of dielectric that's cut to fit inside your horn shape, then drill holes to change the electrical length, no holes in center, lots of holes, or deeper holes out near the edge. It's all about the electrical delay, not the physical delay. That hole drilling will make the mechanical cutting easier and you could start with some holes, and add more holes later.

Energy does bounce back from an air/dielectric interface, so that's a concern from an efficiency standpoint. The energy that bounces back will hurt your VSWR. That website I referenced may have curved front faces on the lenses and teardrop shapes to optimize the VSWR too.


kch

 

[blueandwhiteg3]

This is so interesting, I wish I had at least some simple example designs I could reproduce, verify, and then modify.

What kind of materials would be of interest for use around 6 GHz?