What's the best method to design a narrow band pyramidal horn antenna?

[drkirkby]

How do I design a narrow band pyramidal horn antenna?

I'm looking to try to design a pryamidal horn antenna for the 3 cm amateur band (10.0 to 10.5 GHz) to mount to WR90 (WG16) waveguide.

I'd like to look at antenna deisgns with gains in the range of 15 to 25 dBi. Anything less than 15 dBi is not really good enough, and anything greater than 25 dBi looks to be excessively long.

I have access to:

* Antenna Magus - no idea of the algorithm it uses
* HFSS
* HFSS antenna deisgn kit. I've no idea what algorithm that uses, and its a bit useless as you can't set the gain - only the centre frequency.
* IEEE papers
* Plenty of antenna books (Jasik, Krauss, Balanis).
* Some amateur radio microwave books.
* Engineering drawing from various manufacturers of horn antennas which are on the web.

Looking in the IEEE papers, it seems there to be an almost endless number of papers on the deisgn of horns, but as someone who has never looked at designing one before, I find it hard to see the wood from the trees. I'm just overwhelmed wth information. Virtually every author of a paper claims his/her method is a significant improvement over what others have done, which is my experience is not unusual in academic papers!

As far as I can determine, there are 3 main variables.

* Width W at front of horn
* Height N at frong of horn
* Lenght L of tapered section.

but how to chose what combination is best?

My main design goals are:

1) High gain.
2) Not too big - which conflicts with (1) of course.
3) Low sidelobes would be nice.
4) Preferably some design goals which don't requie a maths Ph.D. to undertand.
5) Something which avoids the complexity of neural networks, particle swarm optimisser, genetic algorithms etc.

Ideally something where I could stick a deisgn calculator on a web page for others in my radio club to use, so one puts in a gain figure and frequency, and it generates some decent dimensions.

Dave

 

[jiripolivka]

Re: How do I design a narrow band pyramidal horn antenna?

I'm looking to try to design a pryamidal horn antenna for the 3 cm amateur band (10.0 to 10.5 GHz) to mount to WR90 (WG16) waveguide.

I'd like to look at antenna deisgns with gains in the range of 15 to 25 dBi. Anything less than 15 dBi is not really good enough, and anything greater than 25 dBi looks to be excessively long.

I have access to:

* Antenna Magus - no idea of the algorithm it uses
* HFSS
* HFSS antenna deisgn kit. I've no idea what algorithm that uses, and its a bit useless as you can't set the gain - only the centre frequency.
* IEEE papers
* Plenty of antenna books (Jasik, Krauss, Balanis).
* Some amateur radio microwave books.
* Engineering drawing from various manufacturers of horn antennas which are on the web.

Looking in the IEEE papers, it seems there to be an almost endless number of papers on the deisgn of horns, but as someone who has never looked at designing one before, I find it hard to see the wood from the trees. I'm just overwhelmed wth information. Virtually every author of a paper claims his/her method is a significant improvement over what others have done, which is my experience is not unusual in academic papers!

As far as I can determine, there are 3 main variables.

* Width W at front of horn
* Height N at frong of horn
* Lenght L of tapered section.

but how to chose what combination is best?

My main design goals are:

1) High gain.
2) Not too big - which conflicts with (1) of course.
3) Low sidelobes would be nice.
4) Preferably some design goals which don't requie a maths Ph.D. to undertand.
5) Something which avoids the complexity of neural networks, particle swarm optimisser, genetic algorithms etc.

Ideally something where I could stick a deisgn calculator on a web page for others in my radio club to use, so one puts in a gain figure and frequency, and it generates some decent dimensions.

Dave

Your approach seems to me too complex. There is an easy way. Google a "standard-gain horn" manufacturers and note the dimensions they use on their horns. Make a similar-size horn from a tinned steel sheet, or use brass or copper sheet. Solder carefully with a 200W iron.
For you narrow-band application, almost any horn will work. The secret is to adjust a good match in the coaxial to waveguide transition. You can make one and adjust it for the best results.

You can study books and manuals but good horns were designed in 1940s. Since then they have been copied by many with a full success. I have made good X-band horns of hard paper and lined by aluminum foil. They all are as good as $450 models you can buy.

 

[drkirkby]

Re: How do I design a narrow band pyramidal horn antenna?

Your approach seems to me too complex. There is an easy way. Google a "standard-gain horn" manufacturers and note the dimensions they use on their horns. Make a similar-size horn from a tinned steel sheet, or use brass or copper sheet. Solder carefully with a 200W iron.
For you narrow-band application, almost any horn will work. The secret is to adjust a good match in the coaxial to waveguide transition. You can make one and adjust it for the best results.

You can study books and manuals but good horns were designed in 1940s. Since then they have been copied by many with a full success. I have made good X-band horns of hard paper and lined by aluminum foil. They all are as good as $450 models you can buy.

I'm quite sceptical of that approach, as I know what happens in the antenna industry - a lot of people just copy their competitors without actually evualating the design. I used to work for a company that made antennas, and I know fully well the specs in the catalogues bear no resemblemblence to what is actually achieved, but a remarkable resemblence to what their competitors cataloge says. This is not confined to the cheap antennas aimed at the consumer market, but fairly expensive antennas aimed at a "professional" markert.

I tried what you suggested and see a range of results. Here's what I find for 3 commerical antennas, all for X-band, all with 20 dB nominal gain. I've only considered here the width (A) and height (B) of the apeture.

Pasterneck A=5" B=3.7" A/B = 1.35
Fairview Microwave A=4.87" B=3.62" A/B = 1.35
Flann A=4.3" B=3.1" A/B= 1.39

So the Flann antenna, which uses a larger aspect ratio at the output flage, is quite a bit smaller than the others for the same gain. I believe Flann are the more professional of those 3 companies. Certainly they sell things of much greater complexity than the other two companies.

I can't help feeling that there is a better approach than just copying commerical manufacturers, who will quite often just copy each other.

I wont be using coax directly. I think my best approach is probably to just calibrate the VNA with a waveguide calibration kit. I've been thinking about how to make one of them, and think I have that sorted out.


Dave