Free simulation software for antenna and feed modelling

[sophiecentaur]

I want to be able to predict the performance of modest (Amateur) Radio Telescope Equipment.
Most of what I can find from a basic search will cost a lot of money. (Not surprisingly). I located a free, student, version of CST Suite from Dassault and it appears to be very powerful and can deal with many different EM problems. But the tutorials and descriptions are very weak - consisting mainly of reading out loud what the student can read from the screen and icon labels, without the basic principles of operation being described. Very boring and requiring a lot of material to be held in the head before you can achieve anything.
I may be trivialising this too much and people may well say that I should start with a general primer on the subject of simulation. Something free would be very useful to me.
Also, an alternative package that is free and approachable would also be useful.
Any ideas, chaps?

 

[Baluncore]

I want to be able to predict the performance of modest (Amateur) Radio Telescope Equipment.

The main problem with simulation of radio astronomy antennas and feeds, is that it tends to get in the way of actually assembling something and making real RA observations.

More important is how the structure will withstand snow, ice and wind loads. Then there is the problem of operating at different declinations, the gravity vector will distort the aperture and pull the focus out of position.

If you want to compute the beam pattern of an illuminated aperture, then you can solve that by taking the 2D autocorrelation of the aperture illumination distribution.

For a simple geometry you may now be able to do it using a math engine.

But for arbitrary systems, it will need to be numerically, by using FEM.
You can write that code yourself.

 

[sophiecentaur]

is that it tends to get in the way of actually assembling something and making real RA observations.

I have already found that out with optical astronomy. Making stuff can be at least as great a pleasure as using it. Time isn't an issue. Thing is also that making stuff can be done indoors in the warm. Obs'ing tends to be in the cold and dark. It would offend me to be paying good money for stuff that may in fact not be very special at all. At least with optical images, there's a pretty picture to look with and aberrations are visible. Also, accuracy of manufacture is not hard to achieve.

snow, ice and wind loads.

Not with the size of equipment I would be contemplating. With anything DIY there is a risk of bad performance that you are just not aware of without proper test equipment .

numerically, by using FEM.
You can write that code yourself.

Too hard, guv. ;-) I just couldn't be sure of any results I got; the results would be more dodgy than the results of making something up in metal.

I will stick with CST if there's no alternative and get over the style of presentation.

 

[nsaspook]

Take a look at https://openems.de/start/

A demo from a few year ago running on one of my Linux servers.
https://openems.de/index.php/Tutorial:_Parallel_Plate_Waveguide.html

 

[Baluncore]

Not with the size of equipment I would be contemplating. With anything DIY there is a risk of bad performance that you are just not aware of without proper test equipment .

Test equipment can be satellite TV equipment or SDR. There are target satellites out there that you can use to verify your system.

A small dish or array, say up to 5 m diameter, will detect the presence of astronomical radio sources, but that is the limit. HAM Earth-Moon-Earth is a better application for that type of equipment.

For an amateur to get more interesting results will require a wire mesh dish of about 12 m diameter, or an array of elements that can be operated as an interferometer. Unfortunately, you are observing from a latitude of +50°N, while the brightest pulsar (Vela) is at declination -45°, well below your southern horizon, but close to my overhead.

There is little point in an amateur duplicating or competing with the government funded traditional RA observatories.

Rather than looking for the peaks of energy sources at microwave frequencies, you may do better by measuring the radio noise floor, between the sources of interference in time and or frequency. The technique works well as a RA pre-processor when there are RADAR or TV horizontal sync pulses everywhere.

Here is an example of a possible high-tech low-cost amateur radio astronomy technique. At HF, transparent holes open in the ionosphere from time to time. During that period there is no HF skip, and the holes let the external HF universe in. With an inverted V antenna for HF, continuously digitise the RF with a 16 bit A-D, at say 120 MHz, using today's $100 technology. Every 100 msec, compute the FFT of the last 12 Mwords data with a desktop PC. Yes, it can now be easily done in real time, and beyond the 16 bit conversion, you will get a transform gain that drops your noise floor by √(12e6). Accumulate the minimum of the spectrum and let the computer wait for the view of an HF radio source through a hole in the ionosphere.

 

[sophiecentaur]

For an amateur to get more interesting results will require a wire mesh dish of about 12 m diameter, or an array of elements that can be operated as an interferometer.

You seem to be making assumptions about the course that I would take into RA.
As a first stab into RA, 12m would be ridiculous - I'm sure you would agree. I do not live in a field. If I were to want that sort of facility then it would have to be with an established group and way down the line.

According to what I have heard from friends, significant results can be obtained with a fraction of that arrangement. By significant, I mean results that show the presence of some identifiable sources and that can be obtained on very humble equipment.

But whatever equipment and what ever signal levels, it can't be a good thing to be relying on what you can get from a 'supplier'.

 

[Baluncore]

By significant, I mean results that show the presence of some identifiable sources and that can be obtained on very humble equipment.

But whatever equipment and what ever signal levels, it can't be a good thing to be relying on what you can get from a 'supplier'.

Why ever not?
As I wrote;

A small dish or array, say up to 5 m diameter, will detect the presence of astronomical radio sources, but that is the limit.

Your humble equipment only needs humble test equipment to verify it is up to specification. You do not have to rely on any suppliers, you can build it yourself if you want. But I see that homebrewing as an unnecessary duplication, it will take you significantly longer, and it will cost you more.

 

[sophiecentaur]

@Baluncore i wanted some help with locating simulation software. You haven’t helped in that direction. Instead, you are telling me how to run this thing. You suggest a crazy level of investment in a first stab at the subject. Why?

Also, to suggest that entry level off the shelf equipment can be relied on is a bit naive. Did you ever buy a TV aerial or a hi fi system? Never any problems ?

 

[Baluncore]

i wanted some help with locating simulation software. You haven’t helped in that direction.

I explained that you needed to do a 2D autocorrelation of the aperture illumination distribution to get the beam pattern. You could have done it by now.
That is really a prerequisite for radio astronomy, mainly because you will have to write an awful lot of accurate code to get anything out of RA.

You baulked at the first hurdle, because you thought it was too hard; "Too hard, guv. ;-) I just couldn't be sure of any results I got ". Maybe, since 1D FFTs are easier to comprehend, you need to start with the beam pattern of a 1D distribution. You can compare your simulation results with published data.

You suggest a crazy level of investment in a first stab at the subject. Why?

I do not suggest crazy investment. I suggest that entry level budget stuff borrowed from satellite TV, or off the HAM radio shelf, will indicate the presence of radio sources. To do any more than “indicate the presence” with a dish will require you build a team, and invest big time. A small dish is both a rite of passage, and a cul-de-sac.

Also, to suggest that entry level off the shelf equipment can be relied on is a bit naive.

Why buy mil spec? If it works for satellite TV, it will work for demonstration of RA. If you can't fix (or replace under warranty) faulty modules, then you will get nowhere in RA.

Did you ever buy a TV aerial or a hi fi system? Never any problems ?

The commercial TV antennas sold are designed for a band or three, not a specific transmit site or viewshed. That is why they are always sub-optimal, or fourth best. The manufacturers of my specific TV antenna designs, supply installers in those specific areas. They have always given me samples when needed, and they have always worked better than the best commercial TV antennas. Why would I buy an antenna when major governments and NGOs let me play with theirs.

I think I can say I have never purchased a new hi fi system. I have built a few from scratch, and repaired a few hundred. What I have is always third best because when an idiot throws out a faulty quality system as scrap, I fix it and use it. I do the same with RF test equipment.

 

[sophiecentaur]

I explained that you needed to do a 2D autocorrelation of the aperture

Is that enough, without details of the feed pattern?

If it works for satellite TV, it will work for demonstration of RA.

Not really. The performance of a sophisticated digital broadcast system is very different from the performance of a very basic measurement system, which will always be 'marginal' for simple equipment. The two environments are significantly different.

I have never purchased a new hi fi system.

Not relevant. For someone with your obvious experience, you will know about advertising specs and that the purchased products often fall way short.

I think you are trying to be helpful but you haven't come up with an answer to my actual question. Your replies have given me some useful insight but I still don't have my suitable simulation software. If I never, in fact build a system, a simulation would still be the source of information and interest for me.

The worlds of Amateur Radio and Amateur Radio Astronomy are a strange mixture of, on the one hand, great experience and knowledge and, on the other, a lot of myth and snake oil. I want to be able to distinguish between the two.

 

[Baluncore]

This was used to model the geometry of feed details and the dish surface.
https://en.wikipedia.org/wiki/Numerical_Electromagnetics_Code

 

[marcusl]

The only free EM SW I'm familiar with is NEC2. It's a method of moments (MOM) simulator with a clumsy interface that is a hold-over from the days of punched cards, but multiple vendors provide GUI interfaces, automatic meshing, etc., as well as upgrades to the later NEC4, for a fee that is quite modest. A bigger problem is that MOM is poorly suited to modeling solid surfaces (the walls of a horn and the reflector surface). I'll bet there are some SW put together by radio amateurs that you can find on the web, but if you can invest the time to learn how to use the CST studio, you'll have a top-notch professional tool at your disposal.

To augment your simulations, you might want to study material on the theory of reflector antennas. (I always like to have a general idea of the answer before simulating anything.) Yaya Rahmit-Samii has a nice overview without a lot of equations in ch. 8 here
Deep Space Comms
My favorite antenna book is Balanis, Antenna Theory, which provides more technical treatments of reflectors and horns. You can pick up a used copy of the older 2nd edition cheaply. Since reflector technology hasn't changed a lot over the years, another resource is Silver, Microwave Antenna Theory and Design, vol. 12 of the famous Rad Lab radar series. Written prior to the invention of computers, it emphasizes approximate equations and intuition. See the chapters on "Aperture Illumination and Antenna Patterns," horn feeds, and "Pencil Beam Antennas" (reflectors). The series passed into the public domain in the 1950s and has been reprinted numerous times, so used copies are cheap. Alternately, you can get it and the other 25 volumes in the Rad Lab series on a single DVD for less than $8--search online.

Have fun!

 

[Baluncore]

The MIT Radiation Laboratory Series is available for free.
https://archive.org/details/mit-rad-lab-series-version-2

 

[sophiecentaur]

This was used to model the geometry of feed details and the dish surface.
https://en.wikipedia.org/wiki/Numerical_Electromagnetics_Code

Ah yes. I remember 'we' used this in the 80s. No doubt its modern version is a bit more convenient to use.

The MIT Radiation Laboratory Series is available for free.

Definitely full of information but, being a textbook, basically, it gives its own examples, followed by the answers. It a good source of information but it isn't a 'convenient tool'. I do not apologise for wanting a convenient tool; we all converse on PF, using a very convenient tool and we don't have to know the ASCII code for the characters we're typing in.
If such a tool exists, I would be daft not to make use of it.

 

[Baluncore]

Ah yes. I remember 'we' used this in the 80s.

I think you probably used one of the early ham radio wire antenna modelling programs based on Mini-NEC that used MoM. If you had used NEC-2 you would know it, and you could use it again to model your feed and dish surface.

Definitely full of information but, being a textbook, basically, it gives its own examples, followed by the answers.

I always used the MIT series as a reference to the theory of the common practice at that time. I was quite unaware that it was a textbook, or that it included worked examples.

If such a tool exists, I would be daft not to make use of it.

I can't see why you would need such a tool, because you appear to have have skipped the design phase, and so have no design to model. If you specified and optimised your design parameters, you would have confidence in your design. If you went through the design process you would realize that simulation was not really necessary for such a small and well defined structure.

It can easily become procrastination by simulation. What were your design parameters, and what will you model ?

 

[sophiecentaur]

I think you probably used one of the early ham radio wire antenna modelling programs based on Mini-NEC that used MoM. If you had used NEC-2 you would know it, and you could use it again to model your feed and dish surface.I always used the MIT series as a reference to the theory of the common practice at that time. I was quite unaware that it was a textbook, or that it included worked examples.I can't see why you would need such a tool, because you appear to have have skipped the design phase, and so have no design to model. If you specified and optimised your design parameters, you would have confidence in your design. If you went through the design process you would realize that simulation was not really necessary for such a small and well defined structure.

It can easily become procrastination by simulation. What were your design parameters, and what will you model ?

I was only a ‘Ham’ for a short while. I worked for a major Broadcast organization. My design experience was from LF to UHF, mostly broadcast applications; no dishes.

I hear your advice but I do know what I want, thanks. Diffrent strokes for diffrent folks, perhaps.

PS
@Baluncore said "you appear to have have skipped the design phase, and so have no design to model." Where did you get that idea? Whatever design, I would still need to know its likely performance. And since when was "procrastination" a bad thing? I have saved myself thousands of GBP that way.

 

[Tom.G]

I can't really figure out if this is what you are after, but here is a MATLAB program for parabolic antennas:
https://www.mathworks.com/help/antenna/ref/reflectorparabolic.html

(above found with:
http://www.google.com/search?&q=matlab+program+for+parabolic+reflector+antenna
which was a Google suggestion for a search of:
'program for parabolic antenna')

 

[Baluncore]

Baars J. - The paraboloidal reflector antenna in radio astronomy and communication.. theory and practice-Springer (2007). (Astrophysics and Space Science Library 348)
Includes Mathematica Routines for design and modelling.

 

[sophiecentaur]

here is a MATLAB program for parabolic antennas:

Looks interesting. Thanks.