Building a home photovoltaic solar panel using transparent polycarbonate

Building a home photovoltaic solar panel using transparent polycarbonate (updated)

In this entry we will see how to build a panel photoelectric home-made in the cheapest way possible, but at the same time be safe and durable (it has to endure 25 years). As we will see, the difficulty of manufacturing is very high, so I recommend it only for the more handyman ;). In addition the process is long, but very satisfying if you like the "do it yourself" and they fascinate you photovoltaic panels.
I've seen many examples online of little lasting home panels, some preparations with wooden frame, which I doubt that they last more than ten years, or in sealed box sealed with silicone (so that moisture enters, but then condenses inside not to leave quickly, corroding the cells), so the potential savings to make the panel of a home goes down the short duration it efficiently.
As we discussed in the previous post, my intention was to use polycarbonate, three times more economical than glass, in addition to much safer and lightweight, whose only problem is that it dilates 3 times more than the glass, and has one somewhat lesser light output than glass (and even more down to whiten with the Sun), but stays in a reasonable range for the first 30 yearswhich is what interests us.

To attach the solar cells, I searched the way more simple and economical; without support or back cover, but by attaching them directly to the polycarbonate, and thus kill several birds with one stone:
  • Reduces costs of materials,
  • You alleviate the panel,
  • You avoid possible accumulation of moisture or water, since polycarbonate is not completely (no plastic is waterproof it, unless you have a barrier of metal), so it is convenient that there are spaces where moist air may accumulate.
  • It allows some light to pass between cells, so the bottom is partially illuminated.
  • Cells give a performance more how much "cooler" are; for every ° C above 25 ° C, the panel lost 0.5% efficiency. If we leave the back "on the air" only with a layer of protection (paint, epoxy or silicone), they either better, last longer and give more electricity.
Also to be glued to the plastic cells, there are no air bags just and you eliminate potential corrosion problems in the cells by condensation.
Note: That was my intention, but as we will see in the end, if I want to join in this adventure dear reader, you'll see that cells join polycarbonate is not the best option, but that it is better to leave them separated, so the plastic dilate freely and heat of the cells do not pass directly to the same. Everything you learn. However, I will detail the procedure if you want to use glass, since this technical solution that would serve if you use glass, since the same expansion than the cells themselves (to the end and Cape are nothing more than very thin wafers of crystalline silicon).
The only problem that this configuration can have not put back cover and use polycarbonate, is that it is less strong in the gusts of wind (although I've used 5 mm colorless transparent polycarbonate plate to strengthen it), by what you must install it in such a way that the wind can not influence him violently; for example putting it totally stuck to the smooth surface of the roof, or wall, no hollow back by where the wind can gatecrash and "straining" pulling, since being anchored the polycarbonate in few points, wind stress > 50 kms/nr can be excessive to coming from the area of the anchor. We can also avoid this problem by placing it in an area protected from the wind, or a tilt and position so the more usual gusts of wind in the area are not front directly, and the panel the "cuts" in some way.

The great discovery was the neutral silicone, as we saw in the previous post; It allows us to much play thanks to its resistance to UV, high transparency, good adhesion on all types of materials, economic and easy to use.
It would have been ideal to use transparent liquid silicone to attach them (although it is no longer necessary as we will see), but I haven't found it anywhere, I have found liquid type Rubson to waterproof roofs but are coloured. In dense and transparent only in cartridge format.

The silicone layer should be enough so that the tension of the polycarbonate to the stretch does not directly affect the cells, but Dim in the silicone. I've calculated to eye that it might still be a layer of 2-3 mm, and hoped that it would serve and cells could withstand the surface tension which is longitudinally, but as we will see at the end of the input is not enough, and this separation, in case of joining them, should be between 5 and 20 mm, making it unfeasible this solution by its cost and reduction of efficiency.

Buying polycarbonate sheet

Plates of 5 mm. PC (Plasinco for example) stores have them in 3050 x 2050 mm. plate format that you can then supply cut in the format that you want.
Note: I opted for 5 mm. thick to give greater rigidity, but 4 mm. There big difference, moreover, as we will see later, we have more constructive options; We could have opted for a plate of 3 mm. and another 2 mm as back cover in the middle of which are the cells 'floating' United with points of the subsequent iron also joined the front plate with multiple points of silicone and silicone; currently I think it is more durable and resistant option, and thus I will do in the next panel, for within a month.
A whole plate cost 270 euros, and I to take advantage of materials, have tried to make the panel the greatest possible seeking balance between size and total weight, so I have chosen to buy half plate (1525 x 2050) cut in two to make two panels of 1525 x 1025 mm.
With that size, I have to put 5 rows x 9 columns of cells 156 x 156 mm; 45 cells in series of 0, 56V that produces each give me 25 volts maximum totals each panel, which conectare in parallel using a diode schottky double with positive common.
Note: can make configuration of rows x columns that we better come to our project, provided that we can leave at least 5 mm. between cell and cell to make connections, and trying all the panels that go to the same micro - inverter have similar voltage (albeit of varying power). Also it could have put 6 rows x 8 columns, but there was place on the extreme sides for the planned expansions of the polycarbonate.
As the Chinese investor I bought accepts up to 28V input, there is no problem:

Preparing the casting

For the fixation of the plate of polycarbonate we need to prepare a framework or frame of the required size, in some material that can withstand 30 years outdoors without deteriorating significantly. Ideally, be of the same size or 1 cm. smaller, polycarbonate protrudes and protect it from the rain.
I have recycled a few square tubes, leaving a gap between the sensor and the upper berths to remove a bit of the wall, as you can see in the drawing.

Drawing frame and panel with torn for dilation and mooring in wall
Necessary equipment: welding machine electrogena, little grinder with discs cut from 125 x 1 x 22 mm, and 125 x 3 x 22 screen auto-darkening, rule, square, tube of 50 x 50 x 1, 5 mm, electrodes of steel 2 mm, etc. We prepare two tubes equal to far and two others for the width taking into account the extent of the panel, as discussed.


With the help of a square and a large smooth surface present slats, we hold them and dot them with the size of the plate. You can use wood, but I do not recommend it because it deteriorates quickly in the open.

We finished welds and us should be something like this. Then apply you paint style forge welded areas and where were more exposed to corrosion, this painting carries particles of metal and is very durable, very good corrosion with just lend a hand.
In the Center lack of place a booster, who fails for the end to facilitate the placement of photovoltaic cells.

As indicated in the previous sketch, to attach the iron to support allowing that this can resize freely we prepare holes torn in the sense of the expansion. In those areas where there is much space we will strengthen the fastening flanges, being plastic allow some movement and are very durable (10 years without significant deterioration).
I have prepared a template for tearing holes, which I will use to mark their position and then finish them with the drill bit and drill:

Template with tearing of 8 mm. x 2.5 cms long
For this I have used large 25 mm in diameter, with rubber washers. Under the same I will place pieces of 1.5 mm polycarbonate so that it will slide (the rubber would be engaged, and will also deteriorate).

The screw thread plate I used are 8 mm, just go through the washer. A slightly smaller can also be used.

Once dry paint put the panel in your website presentation subject with a few jaws.

And we proceed to the marking of tearing, and with an 8 mm HSS drill bit We drill polycarbonate and the top of the metal, without crossing it. If the chosen rosca-chapa screw out smaller, would have to drill with a drill 1 mm. smaller than the inside diameter of the screw thread.

It is very important to make holes in the part more internal effort to leave enough material of the plate at the sides; 2. cms is more than enough, so that it is strong and not to break over easily.

To finish cutting the torn will use a well-sharpened bit that has good sidecut, and holding the template, help drill to continue tearing,

We rasgaremos also slightly outwards, that panel also endure contractions of possible frosts, and does not break where the screw when you reach the top. It is a delicate operation that must be done with skill and patience.

In my case, as the screws are very long, tie on the other side, so end of cross tube with a 6 mm drill bit. and then put the screws with their washers and polycarbonate separator, but loosely much, leaving half mm SAG approximately to allow free movement of the plate.
To fix them, as they are loose, I've given them a spot weld on the other side and I cut them.

Placement of photovoltaic cells on the plate

For that we first remove the protective film from the inside plate; the panel thanks to the same has been preserved without any Gall (remember that it is very easy to grate):

It is time to mark lines that we will serve as a guide to place the cells. Using a plastic of the same size as a guide we check the distances; We have to keep in mind to leave at least 2 cm. on each end for the expansions of the panel, and between cell and cell at least 4 mm. to facilitate the connections.

We make a few marks with the necessary divisions, and improvising a ruler mark with the pen or marker that is easily eliminated a dashed line.

By placing the cells

Once we have the photoelectric cells prepared with the track welded on its negative side, with enough length to join them after next positive in his union in series with each other, let them go putting on polycarbonate.
I will not explain here how are soldered, as there are many videos on the internet, and we only need a bottle of flux, tinned copper tracks, and a soldering iron at least 60W with a tip with an area of the track width, and much mana and patience.
Attention: the cells are extremely fragile glass wafers, have to handle them with care. Do not worry, because although at first you hard, you immediately acostumbraréis to handle them as if they were extremely fragile CD disks. Did I say fragile? Non-fragile, the following ;).

156 x 156 mm multicrystalline cells, 14.4% of efficiency and 3, 5W power acquired on Ebay
Place cells to presentation to see what space we leave between them, I will be placing one to one on a mattress of neutral silicone.

We started the process on the side where they will go loose tracks so that they do not hinder us according to go by placing cells. First we take a line of silicone of the size of the cell.

We smooth it with a spatula to leave the same thickness in all the box:

And then we put carefully the cell, pressing them against the silicone. To do this we have to use a smooth surface of the size of the cell, with a thin rubber on the side that will play the cell so that it adapts, and tighten carefully to reduce air pockets. (However This is not necessary if you are using another method more correct than this and we will see in the next post). 
If you hear a crunch, is that the cell has cracked or broken. This is not a serious problem if not is any piece of the cell separate from the rest. And although so does not reduce the capacity of the other cells. The problem can come if break affects the area of the slopes. In short: If you break a cell and crack can be seen with the naked eye, it is better to change it.
The fact of separating the polycarbonate with silicone cell will allow us to then solder them easily once the silicone is dry.
As we shall see in the following entry, it is not appropriate to make a layer of silicone, but it is better to apply a fine layer of protection directly to the cell on the side in contact with polycarbonate, in order to protect, and once dry put it on your website, with a line of silicone only by the outline to secure it and that is floating on the polycarbonate (before soldering them using a square of the size of the cell to isolate the polycarbonate glass); I'll explain you better in the next inning with a video. In this way we will avoid the problems associated with the expansion, we will improve the transfer of light and will spend less material, facilitating at the same time possible change of a damaged cell.
But you choose the materials you choose, the process of construction of a panel is similar; we are placing the cells with the tracks of each row to a different side and the following to the opposite side to be able to then join them easily in series (normally 36 cells, I put 45):

In the pictures below you can see how they are.

When I finished I was left a quarter of a silicone cartridge so I applied it on the sides of two rows.

You can see that the cells are placed making zig-zag
Not now we most to wait a week to allow the silicone to dry well. It takes one week because Crystal wafers is opaque to the passage of the air molecules, so silica can only harden is by the action of air molecules that penetrate the polycarbonate front and narrow lateral space, above all.

Welding cells tracks

The next step is to Finish welding cells together. So we need a powerful 60W soldering iron, and a few 50 meters from roll of track for 100 cells 156 x 156 mm. tinned copper 2 mm. width of good quality.

Verify that tracks going right on your website. The grey side is the positive cells, while the side that is facing the Sun is the semiconductor (negative) n.


As you can see in the picture below, due to dilatation of the polycarbonate is very important to leave a fold up on the tracks between cell and cell so that can stretch and shrink freely about 5 mm.
Once prepared, we can apply flux liquid of brush or pen (without welding is difficult, since it is necessary to remove the thin layer of conductive oxide layer):

We passed the welder holding the track in place slowly leaving that he melts the Tin to the cell. A pot of flux 15 ml. you will only come to prepare 10 solar panels like this. I bought it on Ebay for €5 + postage.

If you want to avoid having to remove any of the cells after welding them, it is highly recommended that you check them with the tester before joining them; in my case one of them, I don't know why, it gave me a very low, only 5 Ohm resistance, which indicated a clear short (perhaps damaged to solder track on the other side) while that is normal among 200 ohm and 1 K, depending on the light that affects them (the more litlower its resistance).

Appearance of the hollow after booting with the cutter the damaged cell.
Finally we have to join the ranks. Normally use a wider cable but it is enough to use two tracks together to allow a major step of electricity. As you can see, I have placed flat wires inwards to leave space at the ends and extricating the frame when the plate dilate. Very important to isolate the side positive, I have used two layers of duct tape. To fold the cables, try not touching the corner of the cell to avoid a possible short.
The thickness and width of the tracks used is sufficient to withstand 300W lossless, so it should not worry whenever welds are well made.

Detail of the union between two rows in series with double welded wire

Checking the performance of the panel

And it's time the truth; we will check the performance of the panel before you finish sealing the back of cells with silicone applied with brush.
As you can see in the image the appearance is very good, and although they have been air bags, they do not affect his performance as we shall see.

Panel to the Sun on February 18, 2014; Note the flanges that I have placed to improve the fixation
I have tried to put the panel as possible, Sun-facing and a maximum power of 3, 5W per cell, i.e. a total of 157W theoretical, I have obtained 132W peaks, so in principle seemed to be a success (taking into account losses of 10% of the inverter, the small error of placement that may be), tells us that we are losing approximately a 5% performance by polycarbonate and siliconewhich is not bad, and it confirms to us that these cells do not take advantage of the range of UV light (because this iron Lexan reflects this;) both sides are protected against UV).
I think that direct sunlight in winter and summer is practically the same, the only thing that changes is the angle on which affect rays (indeed in winter sun is a bit closer). Therefore the panel production conforms to the theoretical.

An electronic meter (measures the passage of electricity in both senses) helps us (source)
Note: If only one of the cells stays in the shade, the panel production will be 0 Watts. And if a cell shaded half the production will be half. In this situation of shadow panel does not produce, and the shaded area behaves like a resistance, preventing to pass electricity and consuming production from the rest. Why blocking diodes are important, as we will see in the later post when we try to tilt and protect the panels.
However, an hour in the Sun, the problems started. As seen in the following image, 3 cells that were already weakened by micro cracks by placing them, not have been able to endure stress and have been terminated of crack (illuminated cracks you can see):

Looking closely at these cells, we can see the gravity of the matter. When removing the panel polycarbonate temperature was 35 ° C, and some 45 cells. If this has happened to these winter temperatures I can not imagine what will happen in summer.

Cracks in the cells: when cooled just as reflected they

Not only that, but that when the negative side stuck to the polycarbonate, the same was loaded electrically, attracting the particles present in the air pollen that had been loaded statically by the Sun. in an hour the presence of pollen was quite clear. Possible solution to this unexpected? Perhaps since he was isolated from soil (over a wood), and when you place it tied to the wall this burden deviates from naturally.

To the relative failure of the experiment (I should have tried with a different sample forms drop cells before anything), I touch remove from all cells that are intact with a sharp spatula:(, y colocar_las con mayor separaci√≥n.) I will tell you about going, in case it is not recoverable, I lost €45 in cells and a few hours of work... on the whole process of learning committed errors ;)
Today 02/24/14 back to test panel and continues 130W full sun.Considering that it is unfeasible to release the cells without breaking them, I'll leave them as they are, and we'll see what happens with the rest in summer; Since the 156 mm expands cells measuring only 1 mm., it is possible that they have just broken that were already weakened. I install it and I will continue with the other, by placing them otherwise on polycarbonate.
Next will see how to position them separate PC floating, with a thin layer of silicone to protect cells first, and once dry, put them and attach with a bead of silicone around only for the outline to hold them in place and best performance of the sensor, in addition to reducing the work.

Update 05/05/2014

Finally as you can see in this entry, paste the cells into the polycarbonate has been bad idea. Not only the expansion has broken half of cells, but that silica reflects the light where it has taken off, reflecting light and hardly passes light, reducing the maximum power to a few theorists 160W 100W, and Moreover the silicone (cheap neutral silicone) has reacted chemically with cells "oxidizing them" (one of them hit it just about replacing another broken and still perfect(indicated in red, are the rest of silicone poorly removed from the previous).
In addition, when heats much the sensor, the delivered power is reduced to 20W spreading some of the cells into several pieces. Continues to operate, but for a short time delivers 100W (if the temperature is low).

To make better use of it, I can only remove all cells with a blade to clean electric cooktops and return to secure them in another way that does not put them at risk, placing them directly against the polycarbonate and then sticking them with a bead of silicone around. Or better put some support and over the polycarbonate...). We will continue investigating... but it seems the best thing to do is to use glass and then a coat of epoxy as it does this handyman:

Project of photovoltaic of polycarbonate panel (Apparently I have not been the first ;)
Encapsulation of photovoltaic home
Photocell Spanish Wikipedia