The idea of using a Solar panel battery to charge various gadgets seemed to me very promising, but of course, 21W as a universal charging is not enough – you want to be able to charge not only in sunny weather, but for this you need a power reserve. Therefore, full-fledged solar panel were purchased and experiments with them began.
Solar panel
There are different options, but the main limitation on the balcony is the availability of free space. To understand the order of prices, a 50W battery costs about 5000 rubles and looks like this:
Balconies are different in size, based on the dimensions of the panels, it is quite possible to place 2 or 4 pieces without any problems, it will no longer fit. For the test, 2 panels of 50W were bought. Such a battery gives about 18V under load or 24V without it, which means that when using 2x batteries, you need to count on a total voltage of up to 50V (for example, many dc-dc converters work normally up to 30V). It is possible to connect the batteries in parallel, but then the losses due to the length of the wires will be slightly higher.
2. Controller
There are 2 options here:
– Solar panel + controller + battery
This is a classic design: the controller charges the battery when there is sun, the user uses this energy when he needs it.
This system has several advantages:
– energy can be used whenever, and not only when it is light,
– the ability to connect an inverter and receive 220V at the output,
– as a bonus, a backup source in the house in case of a power outage.
There is only one drawback: the use of a large-capacity battery fundamentally kills the environmental friendliness of the idea of this event. The number of charging / discharging cycles of batteries is limited, they do not like overdischarge, besides, batteries and controllers are quite expensive. The price of the controller ranges from 1000r for the cheapest PWM version, up to 10000-20000r for the more expensive (and efficient) version with MPPT support (you can read what MPPT ). The price of the battery is from 5000r for a regular gel battery of 40-50A * h, some use LiFePo4 batteries, of course they are more expensive.
– Grid-tie inverter
This technology is the most promising at the moment.
The bottom line is that the converter converts and delivers energy directly to the home power grid. At the same time, the energy consumed from the general network decreases, the house electric meter records lower readings.
Ideally, if solar panel provide enough energy for all consumers, the value on the electricity meter will not increase at all. And if the consumption of an apartment / house is less than the production of solar panel, then the meter will record the “export” of energy, which must be taken into account by the electricity supplier. In Russia, however, such a scheme does not work yet – moreover, most of the old electricity meters consider energy “modulo”, i.e. you will also have to pay for the energy supplied. It seems that in 2017, they promised to start solving microgeneration issues at the legal level. But however, for the panels on the balcony, all this is only of theoretical interest – their production is too small.
The price of a grid-tie inverter starts from $ 100, depending on the capacity. Separately, it is worth noting micro-illuminators – they are placed directly on the battery, and immediately give off the mains voltage, however, the recommended panel power is at least 200W. The inverter mounts directly on the back of the solar panel, which allows them to be connected like this:
But for the balcony, of course, this is irrelevant.
Testing
The first step was to find out what real power can be obtained from solar panel. For this, an ADS1115 ADC board for the Raspberry Pi was purchased for $ 15:
It is easy to use, the input voltage is divided by a divider and fed to the analog input, at the output we have digital values. Source codes for working with ADC can be taken here . An ACS712 current sensor was also bought, the voltage sensor was made from a bunch of resistors (only one nominal value was found at home). An Solar panel ordinary 100W light bulb was installed as a load. Of course, from 48 volts it did not burn (the light bulb is designed for 220V), but only barely glowed. The resistance of the spiral is 42 Ohm, which allows you to roughly estimate the power by voltage (although the resistance of an incandescent lamp is nonlinear, but for a rough estimate it will do).
The first test version looked like this:
Technofetishists don’t look!
The source was added so that the data and the current time were saved in CSV, and a web server was launched on the Raspberry Pi to download files over the local network.
The results for an ordinary, quite clear day with variable cloudiness look like this:
It can be seen that the voltage peak occurs in the early morning, which is a consequence of incorrect installation of the panels – ideally, they should not stand upright.
And this is what the “failure” looks like on the day when the clouds came and it started to rain:
Given the voltage of 44V and the resistance of the lamp filament of 42Ohm, we can roughly estimate (we ignore the nonlinearity of the lamp resistance) that, at best, the power received is P = U * U / R = 46W. Alas, the efficiency of a 100-watt Solar panel when installed vertically is not very good – the sun’s rays do not fall on the panel at right angles. In the worst case (cloudy, rain), the power drops even to 10W. In winter and summer, the total energy received will also differ.
The experiment with the transfer of energy directly to the network turned out to be unsuccessful: a 500-watt inverter from 45 watts simply did not work. In principle, this was expected, so the inverter was left for the future.
As a result, given the decision to abandon buffer batteries, the only viable option was to use dc-dc converters directly: for example, such a converter can charge any USB devices, its output already has a USB connector:
There are models that are slightly more expensive, they have a higher maximum current and a larger number of USB connectors:
There is also an idea to find a dc-dc converter for charging a laptop, their choice on eBay is quite large.
Conclusion
This system is experimental in nature, but in general we can say that it works. As can be seen from the graph, Solar panel from about 7 am to 5 pm, the power delivered by the panels is more than 30W, which, in principle, is not so bad. In very cloudy weather, the results are obviously worse.
Of course, there is no question of economic feasibility – with the generation of 40W * h for 7 hours, 2 kW * h will be generated in a week. Everyone can estimate the payback in prices of their region on their own. The question, of course, is not the price, but the gaining experience, which is always interesting.
But what to do with the energy, the question is still open. Using 40W to charge USB devices is overkill. On eBay there are 300W grid tie inverters with an operating voltage of 10.5-28V, but there are few reviews on them, and you don’t want to spend $ 100 on a test. If a Solar panel suitable solution is still not found, we can assume that one 50-watt panel is the optimum for a balcony – it can be used to charge different gadgets, redundancy in this case is minimal.
At least now all home digital devices (phones, tablets) are switched to green energy without much hassle. There is an idea to consider using a buffer LiFePo4 battery – but the question of choosing both the battery and the controller is still open.
In addition: as suggested in the comments, you can use a lead-acid battery, such as a car battery. Yes, this is a really Solar panel cheap and working option, with a 100-watt panel, something like this controller will be enough for only $ 10-20 on eBay:
