Light contains energy. When light hits a conductor (or semiconductor) some of the energy is translated into moving electrons, creating current. We can harness the current using solar cells (aka photovoltaic cells). When the sun shines on a solar cell, the current output is mostly constant, which is known as direct current, DC. DC is easy to use and great for charging batteries and powering microcontrollers, but using DC from solar cells can be tricky.
Solar panels come in many varieties. When shopping for solar panels for your system, there are a few specifications you need to follow. The specs below will refer to the Solar Cell Huge, shown above.
Output voltage range: The panel above will output a voltage between 0V and 10V depending on the intensity of light and what you have attached to the solar panel (i.e. the load). You will need to make sure whatever you attach to the solar panel will be able to handle at most, 10V. This is called the open circuit voltage.
Power: 5.2 Watts is the total power this solar panel can deliver at its maximum efficiency. The solar panel is most efficient at a point called the maximum power point, which is surprisingly not the maximum voltage, 10V, but a little less, 8V. When the panel provides its peak power, the current draw is around 650mA. This is called the short circuit current. Multiply 8V by 0.650A and you get 5.2W.
Maximum power point: With no load (nothing connected to the solar cell), in full CO sunlight, the solar panel will output around 10V (Voc). When you start to draw more current (increasing load), the voltage starts to decrease, and the efficiency of the solar panel hits a sweet spot, then tapers off.
Thanks to pveducation.org for the graph.
The sweet spot for operation is called the maximum power point (top of blue curve) and is around 8V @ 650mA for this specific panel. Sophisticated solar power systems use MPPT controllers to monitor and balance the system to maintain the max power point. To check if your system is at this point, simply measure the voltage on the solar cell with your load attached to see if you are close to 8V. It isn't too big of a deal if you are not exactly at the maximum power point, the panel will still work, but you will be losing efficiency. Notice, in our testing, we saw 550mA short circuit (Isc), which is outside of the maximum power point.
If you wanted to use this solar panel to power your RedBoard, make sure the voltage regulator can handle at most 10V on the input (which it does). It is almost this simple. Realistically, the solar panel will not be in direct sunlight most of the time and the panel could provide as little as 100mA or less on shady days. The RedBoard running the blink sketch (blinking the on board LED) draws a peak current of 60mA. This setup will not allow much room for additional hardware, due to the inconsistent and limited power draw from the panel.
Another design choice for solar power is to use the Energy Harverster board.
The Energy Harvester board is simply an efficient voltage regulator with an input window that allows charge to accumulate on a battery or supercap, up to a defined voltage, before turning on and powering your system.
In this setup, supercaps are attached to the input to give more control over when and how much power is delivered to the output at 3.3V. Supercaps act in similar ways to batteries, however they don't hold charge as well, but they do last through millions of charge and recharge cycles. The above design uses 4, 2.5V supercaps in series to equal a 10V rating. The Schottky diode is in there to prevent the caps from sending current back to the panel once charged.
As sun shines on the panel, the supercaps begin to charge. If you measure the voltage on the caps, you should see the voltage slowly increase from 0V as more sunlight hits the panel. When the supercap is charged over a certain voltage (between 4V and 5V, defined by the rising UVLO threshold, see page 3 in datasheet), the board will turn on, the caps will discharge through the board and output a regulated 3.3V to power your system. This is nice, because it allows a defined amount of charge to accumulate on the input caps and when the voltage hits a certain level, they are discharged through the board to power your system. The more capacitance you have on the input the longer the board will run, but the caps will take longer to charge. You can think of this system as supplying bursts of power, when it becomes available.
An simple application of this setup is a simple solar power robot using the Ardubot (no microcontroller required).
When the input capacitors charge up, the motors turn and propel the beast forward, until the caps deplete and everything is turned off. Once the caps charge again, the motors turn on and the cycle repeats. FOREVER. Well, maybe not forever, but much longer than the life of any battery! Whisker switches can be attached to each motor control to steer.
Here are some things to watch out for when designing for low power:
Everything draws power: wires (use thick ones), any resistor, capacitors also have an ESR, which contributes to power loss, LEDs, regulators, etc. All of these items will draw power from your system. Deciding what can stay and what can go is a first step to minimize excess power consumption. Check out the tutorial on How to Power a Project for more information.
What are you trying to do? For example, running motors takes much more current (10-100 times as much) than running an I2C sensor. If you are drawing large amounts of current, your storage device (battery or capacitor) will need a large capacity. The Energy Harvest board will also allow for supercaps to be attached to the output to support larger loads.
Aerogel supercapacitors are a great choice for power storage if you are really needing to save power. They hold charge almost as well as a battery! Regular supercaps are also okay if you need a lot of storage (more capacitance = more storage), but will drain quickly.
The uses for solar power are many and diverse. Let us know how you used solar in your project!
I give this five stars for the article And I hope you guys do more on this subject. .
I would also like to point out there is a cheap supply of 12V 5W Solar Panels on EBay. The car companies use them to keep car batteries charged while they sit waiting to be sold at car yards (before they get to the dealership- most of the time they recycle the panels for the next set of cars but sometimes they wind up on EBay.) There was an article about this in N&V a few years ago. I understand the excitement about using low voltage to supply your Arduino, but truth you can’t really use a 10V 3W solar panel to start your car, or really run a 12V bot either. Add to that; NASA is not going to run an USB cable from Earth to Mars to run their bots or to charge their Li Battery.
So, It would really be nice if we could get more information on how to use 12V solar panels- meaning: How To build MPPT charge controller for your 12V Panel, How to do DC to DC conversion AKA: how to turn that 18+V .5A (from your 12V solar panel) into more current so you could charge something like a 12V lithium iron phosphate (LiFePO4) battery pack or even a Li-ion battery pack like what the Crazy Aussie from EEVblog shows on his YouTube blog “LiPo Battery Discharge Testing” 5000mAh 20C battery. Plus, charge in a way that does not fry or kill your Li batteries. Heck, turning that 18V .55A into 13V .7615A would also be better for those folks who are using Lead Acid batteries too. You don’t get the Lead Acid battery trying to turn that 18+V into waste heat while charging which leads to shorter battery life.
Maybe I am wrong here, but I thought one of the ideas floating around was to aim for higher voltages when playing with motors because higher currents not only play havoc with your battery’s life but also put wear and tear on your motor (as in heat) and circuit’s life too, not to mention that you have to have bigger wires to deliver that current. Also I thought it was a lot easier to not dip below the bottom 3.xx Volt threshold for each cell in a series Li battery pack, then a parallel Li Battery pack. It would be nice to get real information on this subject.
Won't increasing our use of photovoltaics deplete the sun's hydrogen fuel faster?
no, it will not : )
As an indirect effect, these PVCs or Photon Vacuuming Cells - don't let the article convince you otherwise, do contribute to the depletion of our most precious non-renewable energy source. On the up side, the increase in Helium produced would be most fortunate if we could manage to bring some of it back to Earth.
Same way putting up more windmills will stop that wind from flowing around the world.
Actually, technically it does work this way. A finite amount of energy or power causes the wind to flow. If there is sufficient drain on this the wind will slow. I don't know about the relative magnitudes but wind turbines do subtract from the power of the wind.
There's a great image demonstrating part of this effect. Actually it shows the 'wake' with all its turbulence, meant to demonstrate that placing windmills in a regular grid could have a detrimental effect on efficiency. But the turbulence naturally decreases the energy more rapidly than a steady stream.
Great article - tons of good content, and way more accessible than the videos. I hope you do more like this.
The best way I have found to power an arduino project with solar is a pv panel charging a lithium ion battery and a dc to dc converter producing 5v. Reliable day in and day out for years now.
I wish things were as simple as shown in the above diagrams. But they are not - far from it.
If you connect any kind of arduino directly to a solar panel, you will have a system that may boot one time but fail to another time. If conditions are just a bit less than perfect, the panel will not provide enough current for the arduino, and as a result the system will end up in a malfunctioning state that it cannot recover from.
So put a capacitor across the panels, you say. No, that won't work. The capacitor will carge up to the point where the arduino can just begin to boot, and it will end up in the same malfunctioning state.
So use a SparkFun Energy Harvester board with that capacitor, you say. This MIGHT work. The caps will charge up to when the Energy Harvester board's UVLO disengages, and provide just enough for the output to run while the cap discharges from 4-5v down to 3.3v, which is very quickly unless you use a very large supercap. Also, the Energy Harvester output is rated at 100ma which I haven't ever been able to draw from it.
There seems to be no really good solution to powering electronics with solar panels and supercaps. Ideally you want to run your panel at its Vmp, and indeed there are buck converters available that feature this sort of functionality, but they are all without exception intended for battery charging. Which means the unit will try to "trickle charge" your Arduino at the configured Vout. Even with a cap on the output it won't work, because as the cap charges, it will once again reach that state where the Arduino can just enter that malfunctioning state that it cannot recover from.
I've been at this for a long, long time now and I still haven't found a good way to reliably power an Arduino with a solar panel. We need something that will charge a cap at the panels Vmp, and when the cap has reached a specific voltage, and ONLY THEN, it will provide power from that cap to your device - and keep doing this until the cap voltage drops below another specific voltage.
It's amazing how difficult this "simple" problem really is.
Do you think could I do the same with a 3v panel 2.7v 25F supercap to charge a 3.2v lifepo4 battery? Or do I need 2 supercaps? Thanks in advance for your help
If the cap only has a 2.7V rating then you would need two in series to use with a 3-3.2V bus.
In the figure where you show the capacitors hanging off the Vin pin, and you had to use a diode to prevent backflow into the solar panel, you could instead put the capacitors coming off the Vcc pin (Vout on the actual chip) and like Fig 11 on p. 16 of the 3588's datasheet you would NOT need a diode. The LTC3588-1 is able to have huge capacitance on its output without the need for a blocking diode...apparently. Haven't tried it myself.
Am I wrong?
I would like to see a topic on energy harvesting from a magnet moving through a stationary coil. similar to the flashligths that you shake. I have done some tests that show I cna make somethign work but I know that you all could make something much more refined and robust.
Also, I have seen application notes on inductive recharging. This would be an exciting area for spakfun ot have some modules in. Given the portable nature of so many of the project i see being built, I would be awesome to have inductive rechargine start showing up in the area.
Don't worry. We're working on something good for harvesting energy from moving magnets :)
Hi, I have replicated the solar panel example above, and connected a little motor between 3.3V and GND terminals, at the left of the graphic. I can see that the caps are getting charged, however no matter how long I wait, I can't make the harvester to start operating. I know that the series of caps are full charged because the motor works when I connect it to the positive and negative terminals of the series. I can't figure what's wrong. Do I need to connect something else? Thank you for your help...
Light doesn't contains energy, LIGHT IS KIND OF ENERGY!!!!!
??? E=hv
It's great to see Sparkfun posting renewable energy topics from time to time: keep it up!
Another super minor goof - the current provided at the MPP is less than Isc, the same way the voltage is lower than Voc. Granted, it'll be much closer to Isc than V will be to Voc. The point being, the MPPT is located somewhere JUST south of Isc and a little bit more west of Voc.
Anyone ever see those cool mendocino motors they make with solar cells? Pretty sweet.
Hey great we need some video explanation how to use SUPER CAP. I think that will be the great importance in future. Please make some experimental with SUPER CAPS
Shouldn't the diagram show voltage over current rather than current over voltage?
Is the PowerCell board a reasonable option for charging a Li-Po battery from a small solar panel?
Might work - wouldn't necessarily call it reasonable, though. There's some discussion in the PowerCell product page that you may wish to read.
To be honest, though, an off-the-shelf solution may be better / cheaper. Be that a Home&Garden type store with garden lights (make sure they don't charge a supercap / NiMH battery), or even the Solar Charger and Battery Pack - 3500mA that SFE carries. It gives you the charging (solar and USB), powering (only 5V), a solar cell and a battery for only $5 more than the PowerCell by itself, and looks like it's easy to take apart and repurpose piece-wise fo whatever you have in mind. Perhaps not as satisfactory as trying to puzzle something together yourself, though :)
One thing you didn't mention was the rated power is at standard test conditions, which may not be realistic for your application. For a project I was derating my panels to less than half the "nameplate" power. I admit it was an extreme case; 100% up time, safety application, northern climate (snow, low insolation); but I would not expect to get what the spec sheet says.
Solar panels don't draw power, they provide it. You goofed twice on that one :-)
Ah, I see where I goofed it. Thanks for the catch. Fixed.
Unfortunately, simply series-connecting the supercaps to achieve the higher voltage rating isn't enough to protect them long-term. You need some sort of charge balancing circuitry in there, at the very least a high-impedance resistor connected across the terminals of each super cap.
Useful Instructable reference: http://www.instructables.com/id/Lets-learn-about-Super-Capacitors-A-Practical-G/step7/Balancing-Your-Series-Banks/
Very good point, I definitely should have mentioned something about balancing series caps.
Although, FWIW, in this configuration, the highest voltage the series caps will ever see is 5V (barring any failure). The peak surge voltage is also 3V per cap. What would a failure look like in this configuration?
From various app notes, the failure mode is usually increased ESR/decreased capacitance, and there's also a temperature dependent factor. You could probably get away with simply putting 2.5V or lower zener diodes across the leads to clamp the voltage within tolerances.
App notes with all the gory details for the Cooper Supercaps can be found here: http://www.cooperindustries.com/content/public/en/bussmann/electronics/resources/library/apps_notes_technicalinfo.html
You have a dead link there on "How to Power a Project". Or "learn.sparkfun.com" is offline right now.
Should be back in a second. You caught us in the middle of a slightly problematic deploy.
Cool. It's up. Thanks.