Geodesic Parabolic Dish Antenna

[Yoshiyuki Takeyasu]

Hi,

Recently, it’s been a popular trend to repurpose inexpensive solar cookers into parabolic dish antennas. However, don't you find their weight makes them a bit difficult to handle?

I’ve just updated my design chart for a Geodesic Parabolic Dish Antenna that is both lightweight and easy to DIY. Please feel free to give it a try:
http://www.terra.dti.ne.jp/~takeyasu/GeoParaAnt_9.pdf

You can see the final version of frame structure in the photo below.



I’ve received a report from a German radio amateur who successfully built a 2-meter dish using this specific structure:
http://www.terra.dti.ne.jp/~takeyasu/1753204419672.jpg

I hope my design charts prove useful to your projects!

Regards.

Yoshi Takeyasu
JA6XKQ

[Alex P]

Hello Yoshi,

Very interesting .. 

Have you ever tried a deep dish ~ f/D 0.25 ?

How close is the dish to a Parabolic vs a Spherical surface ?

Regards,

Alex Pettit

[Yoshiyuki Takeyasu (JA6XKQ)]

Hi,

I have never tried f/D=0.25 because, in principle, this design is not suitable for deep dishes. This dish antenna is not a spherical, so-called "geodesic dome." Instead, it features a structure where the ribs are formed by geodesic lines—the shortest distance between two points on a parabolic surface. Please refer to F4BUC's numerical analysis regarding surface precision at the following URL: https://qsl.net/f4buc/precision_surface.htm

Regards.

Yoshi Takeyasu

JA6XKQ

2026年1月14日水曜日 0:15:05 UTC+9 Alex P:

[b alex pettit jr]

Thank You.

If you want some baseline against which to compare your system, 

here is a recent eval I performed with a 1m f/D 0.35 dish .

https://github.com/AP-HLine-3D/HLine3D/blob/main/21cm_Step1b_KrakenDiscoveryFeed_Eval_05.pdf

Y axis : dB vs Cold_Sky    peak ampl @  Declination + 40 dg RA 20:30 hrs

=========================================================

'numerical analysis' is useful as a basis for building the structure, 

BUT, the actual physical shape of the dish needs to be validated.

Regards,

Alex Pettit

[Henri NICOT]

Hello Yoshi,

Thank you for sharing your work on geodesic parabolic reflectors. I was very excited to read some work on this subject, as it happens to be one of my favorites. A few years ago, I posted some pictures, inviting interested members of SARA to get back to me but I had very few replies.

I remember reading your material and that of Mathieu Cabelic before 2020 when I started into geodesic reflectors. Let me state below the topics we seem to have in common and those which could be complementary.

1. Like you, I took a very mathematical approach, starting from scratches.
2. Are you aware that because the reflector is a paraboloid of revolution, each geodesic rod or rib is actually the intersection of that paraboloid with a plane crossing the axis of revolution? That helps in the mathematical calculations.
3. I also discovered that geodesic and triangular mesh do not agree very well. In order to minimize the number of nodes, it would have been desirable for the geodesics to intersect three by three rather than two by two. Unfortunately, according to my findings, this is only possible if one of them belongs to a plane including the vertex or one edge of the reflector. I am even surprised that you could have more actual geodesic ribs than those drawn in your "Camellia" document.
4. Also, for a given surface area, the perimeter of a triangular shape is longer than that of a square one, making the reflector with a triangular mesh 14% heavier than the squared mesh.
5. In theory, a square mesh retains one degree of freedom which could favour some scissoring deformation. In fact that can be fixed by just adding some cross bars which you can see in the corners of my 2 meter diameter prototype. Also the back support will tend to eliminate that weakness. It never was a problem in my two reflectors.
6. The design, the fabrication and the assembly of a pseudo-square mesh are easier than those required for a triangular mesh one.
7. For the above reasons, I decided on the pseudo square mesh, the quads.
8. Like you, I had some issues creating the circle around my first prototype 1.9 m. diameter. The circumferential segments and the long ribs interweave, creating unwanted thickness. The circumferential segments are not geodesic.
9. I came to realize, after the realization of the first prototype, 1.9 m. diameter that in fact a circular perimeter might not be necessary. Would the signal be deformed? The antenna retained diameter is that of the circle inscribed in the perimeter of the reflector. The ribs can then be prolonged until they meet two by two for a quad reflector and three by three for a triangular mesh reflector. Consequently, the reflector remains 100% geodesic.
10. Now the reflector should be considered as two (three) series of pseudo parallel ribs, oriented at an angle of 90° (60°) and layed on top of the previous one. For the quad reflector, the exact position of the reflector surface is precisely between the two rods. For the triangular mesh reflector, the exact position of the reflector surface is that of the second layer of ribs being squeezed between the first and the third. This is where lies the advantage of the geodesic nature of the reflector.
11. As you know, geodesic antennas can be very light. The 2 m. diameter one weighs 10 kg for the reflector and 20 kg including all (reflector, back support, source and surfacing). The 6 m. project weighs 50 kg for the reflector only. The reflectors are 6 mm thick !

I am now making tests, more related to radio astronomy, with the 2 m. diameter antenna . If the results, as I expect, are positive, I would have good justification to build a new geodesic paraboloid reflector, this time at least 6 m. diameter. I have already done the design of the paraboloid. I would still have to create the back support.

Here are some pictures of the two prototypes and a drawing of the future big one.

Let me know if you wish me to develop one of the above topics or if you have another question.

Henri Nicot

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/sara-list/0241c324-6ddd-87f7-34b1-5418021116a6%40terra.dti.ne.jp.

[b alex pettit jr]

Hello Henri,

Do you have any HLine data to post ?

Thanks,

Alex 

======================

On Tuesday, January 13, 2026 at 03:32:48 PM EST, Henri NICOT <henri.je...@gmail.com> wrote:

Hello Yoshi,

Thank you for sharing your work on geodesic parabolic reflectors. I was very excited to read some work on this subject, as it happens to be one of my favorites. A few years ago, I posted some pictures, inviting interested members of SARA to get back to me but I had very few replies.

I remember reading your material and that of Mathieu Cabelic before 2020 when I started into geodesic reflectors. Let me state below the topics we seem to have in common and those which could be complementary.

[Yoshiyuki Takeyasu (JA6XKQ)]

Hi,

I am the Triangle Lover.

Sesame Street #4144 - The Triangle Lover of the Day

https://www.youtube.com/watch?v=il8Po9AzuGY

In order for the bars to maintain its shape (or parabolic surface) when tension is applied as it bends, it needs a restraining member on the circumference. And it is this pre-tension that gives the Geodesic Parabolic Dish Antenna, a grid shell structure, its strength.

Regards.

Yoshi Takeyasu

JA6XKQ

2026年1月14日水曜日 6:03:26 UTC+9 b alex pettit jr:

[andrew....@googlemail.com]

Hi Yoshi (& others),

 

Can you tell me why the components of a geodesic structure form a natural parabola? I have been trying to find and explanation online but can’t find one.

 

Andy

--

--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/sara-list/b56b72e9-bfe3-443a-b724-209b001c61aen%40googlegroups.com.

[Yoshiyuki Takeyasu (JA6XKQ)]

Hi Andy,

> Can you tell me why the components of a geodesic structure form a natural parabola?

Here is the principle behind the design:

Reversing the perspective: First and foremost, the conceptual approach is reversed. It is not that "the components of a geodesic structure form a natural parabola," but rather that we "identify" a geodesic structure on the surface of a pre-defined parabola.

Approximation of the surface: The fundamental principle of the Geodesic Parabolic Dish Antenna is to approximate a parabolic surface using multiple planes. This is the same principle used in any antenna where mesh is attached to a frame.

Definition of a plane: A plane is uniquely defined by three points, not four. Using four points would result in two intersecting planes.

Triangular composition: Consequently, the parabolic surface is composed of a collection of triangles.

Structural efficiency: While the edges of these triangles form the structural ribs, it is highly advantageous if the shared edge of two adjacent triangles can be formed by a single rib. Please note that in a typical Geodesic Dome, the edges of adjacent triangles are often separate components.

Choice of material (Flat bars): To form a triangle, three ribs must intersect. Considering the connection and the thickness at these intersections, using flat bars is the optimal choice.

Constraint of geodesics: A flat bar, which can only bend in the direction of its thickness, must follow a "geodesic line" to conform to a parabolic surface.

The design process: My design begins by "identifying" geodesic lines on the parabolic surface, initially positioned at 120-degree intervals.

Engineering advantages: At the intersection of flat bars along these geodesic lines, the surfaces of the bars meet at a perpendicular orientation. This allows the faces of the flat bars to fit tightly together, which is a significant manufacturing advantage.

Unique determination of length: Once the vertices of the triangular segments (the intersection points of the flat bars) are identified on the parabolic surface, the lengths of the triangle's sides are uniquely determined.

Formation of the parabola: By joining the flat bars according to these uniquely determined triangular dimensions, the intended surface—the parabola—is naturally and uniquely formed.

References: For a better understanding of geodesic lines, I recommend this video: The Nature of Geodesics (https://www.youtube.com/watch?v=m6WY6VtPYrk).

Further reading: Additionally, I encourage you to read all the references listed in my Design Chart.

I hope this explanation helps your understanding.

Regards.

Yoshi Takeyasu

JA6XKQ

2026年1月14日水曜日 11:43:02 UTC+9 andrew....@googlemail.com:

[Henri NICOT]

Hello Andrew,

The rods tied together are not forming a natural paraboloid. As Yoshi mentioned, it works the other way around.

You start by stating the characteristics of your parabolic reflector: D and f. Then you try to lay down on its imaginary surface the curves which, tied together, will create the skeleton of the paraboloid. Among ellipses and parabolas, some curves have the ability to stick nicely to the surface because each point of those curves is tangent to the surface of the paraboloid. They are the geodesics of the paraboloid. They carry the ultimate ability to "wrap" the surface very smoothly. By choosing them judiciously, you can distribute the geodesic curves regularly over the surface, avoiding concentration and dispersing the stress all over the mesh. The mesh can be triangular or squad.

The parameters of the paraboloid targeted, the choice of distribution of geodesic curves over the surface and mathematical calculations which translate the wrapping ability of the geodesic curves will produce the coordinates of the points where they cross each other.

I hope this helps to answer your question. Please just ask if you need more.

Regards

Henri Nicot

[andrew....@googlemail.com]

Thanks Yoshi & Henri.

 

From: sara...@googlegroups.com <sara...@googlegroups.com> On Behalf Of Henri NICOT
Sent: 14 January 2026 19:37
To: sara...@googlegroups.com
Subject: Re: [SARA] Geodesic Parabolic Dish Antenna

 

Hello Andrew,

--

--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/sara-list/CALBFd%2BRY1DvVS2pTTw03GiD-eP3sk2_fP3g_fdTPYjuZWXL5yw%40mail.gmail.com.

[Stephen Arbogast]

Perhaps something like this...   ???

MatLab:    https://www.google.com/search?client=ubuntu-sn&channel=fs&q=matlab+coordinate+transformation+paraboloid+to+w-y+plane&udm=50&fbs=ADc_l-aN0CWEZBOHjofHoaMMDiKpaEWjvZ2Py1XXV8d8KvlI3p-ML-906rRL_m6h4jR-tdCeKIwp94h-QiJ4lJfObsqUPixp6KuAej6LdEw-ul8fudU82HJrKxYOXZophBT2wsL1A8SvxreDpxGmVqCpVkesfdCEYv4TQqqnLiFMWwhJ20FI1ggBlnQwmM6ujC1A3bTpl2yDZ6XL120IwEWHvfjJYiiIUg&ved=2ahUKEwi87Im5_ouSAxWtEDQIHdc9F7IQ0NsOegQIAxAB&aep=10&ntc=1&zx=1768427040679&no_sw_cr=1&mtid=Hg5oaf-gO4jz0PEP5uKekA4&mstk=AUtExfAJfIlbAZMp1xWtFmySH1J5lZe-OsTH2amEYJrPlIlexIyWechmEKTOiDnY1VkHL-x8KomJHVOXXepr3ZZGg-0NdOGV2vkWcqX-x0arTOiqdAjXpU9QqX2vuZ9RSvtR7Yg2P5juHWqWLv0W-fJuj4Aw1HFoo5bgaL-U84rhNlUgEZWU6eDi0tF4VYsqS5n0XG0aPjKmrUx3UMBKJU3qLghGOJyVi-DKoByGWFo3yBPdFbndH0urhDfRR0lRnzkvCoeQFWhDeWNyATZ_433qF4eQe22viTa_sdcuvtUAxHk7Uuc68xT7zgfqzTozrgf4abdWZ5j2n_b0rw&csuir=1

Python:  https://www.google.com/search?client=ubuntu-sn&channel=fs&q=python+coordinate+transformation+paraboloid+to+w-y+plane&udm=50&fbs=ADc_l-aN0CWEZBOHjofHoaMMDiKpaEWjvZ2Py1XXV8d8KvlI3p-ML-906rRL_m6h4jR-tdCeKIwp94h-QiJ4lJfObsqUPixp6KuAej6LdEw-ul8fudU82HJrKxYOXZophBT2wsL1A8SvxreDpxGmVqCpVkesfdCEYv4TQqqnLiFMWwhJ20FI1ggBlnQwmM6ujC1A3bTpl2yDZ6XL120IwEWHvfjJYiiIUg&ved=2ahUKEwiy8MXx_ouSAxUqPDQIHWDfKWsQ0NsOegQIAxAB&aep=10&ntc=1&mtid=qA5oacm8GaPb0PEP1LawuQE&mstk=AUtExfCye2vKjkb_mD4PnYM1l6dmATRYcPUqBByPyENqbYodR58TH8csPn9tj5ATs-EkGLohynj3vO_afpCjfElt_HWq8qfAPjUhh3tyyyifoLAUBd_nOiXdhk24ikYdB516jNUsfFW4IaCm_XhG9Y883x_-t5x8Ci9EWgR1uup-wJWlkWY38HmAM0GWbfbprk2giGli4v-rFxThU_K7E-yTv5gkVKdjL4GqPb4jv5k72mUBT9I2oB3L1Nl1sqGx-c1M23bjzvQNzHWgXvn52GRBFcEE93fM3bnFsFpk-b_TEX2WI-OszBD9ShzFSFQfQpv79IfYXtFAUdgADvqHaRbNqVH6MxrVZPH_dg&csuir=1

[Yoshiyuki Takeyasu (JA6XKQ)]

Hi Stephen,

I’ve tried the proposed MATLAB code. Your approach is correct if the objective is a transformation onto a plane parallel to the axis of revolution of the paraboloid. However, to accurately identify a geodesic line, the transformation must be performed onto "a plane relative to the surface normal" of the paraboloid.

I like MATLAB approach, thank you for your suggestions.

Regards.

Yoshi Takeyasu

JA6XKQ

2026年1月15日木曜日 6:48:23 UTC+9 Stephen Arbogast:

[Henri NICOT]

Hello Stephen,

The answer is "yes and no".

YES because indeed, projections like those described in the Matlab codes you sent, will create a mesh on the surface of the paraboloid, as on the drawing below:

NO because in the above described skeleton, the lines are not geodesic, except for the two passing by the vertex. They are parabolas, parallel to the vertical axis. In each of the drawings below, you can see a geodesic line. It is the intersection of the plane and the surface of the paraboloid. The plane must be positioned so that its normal vector is collinear with the vector tangent to the paraboloid, on the point of the intersection which has the highest vertical coordinate. Varying the point of tangency along the wall of the paraboloid and/or the angle of intersection with the plane, it is possible to create a mesh on the paraboloid.

I hope this has answered your question.

Regards

Henri Nicot

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/sara-list/ef5aa3a3-15ef-4d18-8cab-97131d3378ean%40googlegroups.com.

[b alex pettit jr]

Henri,

Do you have any H Line spectrum  plots obtained from a system to post ?

Thanks,

Alex

=============================

[Henri NICOT]

Alex,

Sorry for the delay. I have not forgotten your request. You may not remember an earlier exchange of emails during April 2025. I described my little observatory and the actions undertaken (design of geodesic parabolic antenna, calibration of the radio telescope, stabilization of the noise floor signal in full power). I am currently working on the last topic, still in full power. I have not yet worked on spectral analysis, I have not yet created a program to generate H-line spectrum plots using Fourier transformation.

If you would like to know the performance of the radio telescope, I can give you a Tsys of 125 °K and a Y of 3.1 for an ambient temperature of 25° C. I will try to send you some calibration curves sun/calibrated temperature/cold sky but in full power.

Regards

Henri Nicot

--

--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/sara-list/548154108.358688.1768484222520%40mail.yahoo.com.

[b alex pettit jr]

Hello Henri,

I exchange a Lot of Emails. I apologize, I do not recall  yours.

If you want to initially try some off-the-shelf software,  the set  SDR# > IFavg > HL3D  ( + Rinearn Graphics ) is a well tested combination  and easy to use

Otherwise, when you get Your software operational, post some Results !

Best Regards,

Alex Pettit

( software links here )

https://github.com/AP-HLine-3D/HLine3D

=======================================================

[b alex pettit jr]

It is difficult to see system performance with Total Power plots

Alex

=======================================================