How to build an inexpensive device?

Next we needed to try and reduce the cost of our device.

Table 7: Cost of the components of our device


Cost in €

Solar dish + shipping fees


High temperature black paint


Aluminum basket (artisans)


Flexible hose pipe and connection


Annealed copper pipe for cooling system


Stand to track the sun (artisans)




The purchased solar dish we were using cost €150 (including the delivery), an expense which did not fit with one of the criteria given by our partners.

Our inquiries about the satellite dishes in Marrakesh revealed that the most common sizes were 80cm and 140cm in diameter. It was obvious that a diameter of 80cm would be insufficient compared to a diameter of 140cm. The solar flux captured by the small dish would be 3 times less than the larger.

We had to choose the dish with as large a diameter as possible.

Recycling a satellite dish

Écrit par Rohan NOWAK et Waldan GIRARD le . Publié dans How to build an inexpensive device?

a) Covering the dish

There were several ways to cover the dish with reflective material:

    • Using aluminum foil and glue, the cheapest but probably the weakest solution. The reflecting surface would be irregular and in fact would diffuse too much light.
    • Using sticky mirror paper, a simple but more expensive solution that we chose because it is still relatively inexpensive.

Our first problem was that this kind of dish is not at all flat, which of course is why it functions. We looked for a way of covering it, minimizing the number of air bubbles.

Looking at the diagram in Fig. 10 below it is obvious that the distance SA is greater than the radius R. We decided to divide the surface of the dish into 16 sectors so as to reduce the formation of air bubbles that would result because the surface is not a flat disk.

In the diagram in Fig. 10 the sector of a circle with a radius R and an angle a is represented by dotted lines. Actually, the sector of a circle that we cut has to have the same length arc, but a radius SA>R. Therefore, this sector needs to have a smaller angle; angle b,  is less than angle a, as shown in the diagram.

Therefore, the arc length is: l =R.a=SA.b so  with a=2p/16 = p/8.




Fig. 10: Cross-sectional view of the dish and template of a sector of mirror paper to be cut out


When we went to Morocco in January 2016, our friends from the Dar Bellarj foundation had salvaged a satellite dish with a large diameter (140cm). We were surprised to find that the diameter of the dish was not regular, but varied from between 120 and 140cm. However we decided to cover it, choosing the greatest diameter, being 140cm.


b) Location of the support

A parabolic mirror enables the incident rays, which are parallel each other, to be concentrated in one focal point F, but only if the rays are parallel to the optical axis of the dish.


Fig. 11 Reflection of the rays a) parallel to and b) on a slight incline from the optical axis

If incident rays are not parallel to the optical axis, they don't converge in a single point anymore, but we can determine a minimal surface on which all these rays concentrate.

This is that case which had to be chosen because for our device, the support passes through the center of the dish.

For the purchased commercial solar dish, we wanted to determine the location of the focal point F with respect to its vertex Sto see if the manufacturer of the dish had set the support close enough to the focal point.


Fig. 12: Reflection of the rays on the pressure cooker



To determine the focal length f, we used a halogen floodlight we placed 20m from the dish so we could make the assumption that the rays of light arrive parallel to each other on the dish. We made the assumption that these rays were parallel to the optical axis of the dish.

Using a black plate (to minimize the reflection) that was small in size compared to the dish, we searched for the position in which the reflected rays formed the smallest possible spot of light (indeed, the rays did not all converge in a single point)

Results: we obtained a focal length f = (61±1)cm.

We have always supposed the solar dish was a paraboloid.



Fig. 14: Satellite dish covered with mirror paper and concentrating the sunlight under the pressure cooker

This device gave us satisfying results and an output equivalent to that from the purchased dish, being an output of 55%. It enabled us to distill, like the purchased dish did, but it led us to search for another solution because the focal length was much greater, making the support longer and the device unstable.

Constructing a geodesic dish

Écrit par Rohan NOWAK et Waldan GIRARD le . Publié dans How to build an inexpensive device?

Thanks to 3D design software called Solidworks, we planned to build a new dish using recycled printing offset plates made of aluminum. We chose the same measurements as the purchased solar dishes, being a diameter of 1.50m and a focal length of 61cm.


Fig. 15: SolidWorks modelling of a geodesic parabolic dish





Length in cm

Metal strips long enough for holes (cm)

Length of the sides of the triangles (cm)


























Fig. 16: Breakdown of the structure of the dish


We cut aluminum strips to build the framework, and also the corresponding triangles in accordance with the measurements given in Figure 16 above.

Before assembling all the elements, we were surprised to notice that the recycled aluminum plates did not reflect light in a single direction, but they diffracted it as can be seen in the photograph in Figure 17. After, we wanted to understand why this phenomenon happened and we that is why we observed the surface of the plates with an optical microscope using x40 magnification.

As the photograph in Figure 18 shows, the plates were made by brushing them all in the same way. So their surface acts as a diffracting grating, or like a glass rod. We realized we could no longer hope to concentrate the light under the pressure cooker with that anymore... However, we decided to keep the plates as a support because they were flexible (only 0.3mm thick) and we covered them with mirror paper, considering that the cost in Morocco is about €6 (≈US$6.3) per m².


Fig. 17: Laser diffraction by the offset plate Fig. 18: Observation of surface under microscope (x40)

      Fig. 19: Assembling the dish


The tests we did showed the device could boil 6L of water in 52min in May when the solar flux was 1080 W/m² with an angle of inclination of the rays on the dish of 14°. Calculating as described in part A, this gave us an output of only 45%.

The power output was weaker than for the purchased dish. That can be explained in part by the imperfect paraboloid shape taken by the dish, but also by the less efficient reflection of light on the aluminum strips, which are visible, than on the plates.

To conclude, this device was not at all simple to make and we spent a lot more time than we expected. At the moment, we are in contact with a company in the steel sector in Marrakesh to estimate the cost of manufacturing blades equivalent to those used to make the purchased solar dish.