Powering your Raspberry Pi Zero W projects without electricity readily available can be a challenge. On the other hand, the sun provides energy, literally out of thin air, if we have the means to harvest it.
Is it possible to run a Raspberry Pi Zero W from solar energy? The short answer is yes, but let’s examine a few aspects of what it takes to do it. We’ll assume that the Pi is used in some sort of IoT implementation, where it’s largely idle.
Solar-Powered Pi Zero: Power Consumption
The first thing we’ll need to know is how much power the Rasbperry Pi Zero W needs. It runs at 5V, or preferably a tad higher, and the official Pi Mini-USB power supply can provide up to 2.5 amps. Certainly, however, the average current that it draws when idling is much lower.
A little research points to the Zero W consuming 120mA, which I confirmed experimentally. First, I used an inexpensive inline meter to measure voltage levels, which bounces around the 50 and 150mA range, agreeing with the ballpark of the first figure, if not the exact spec. A multimeter hooked up inline to a broken-out USB circuit gave an average of 132mA over a few minutes.
If we peg current draw at 150mA as a round number, we come up with .150A x 5V = .75W of power usage. Note that incandescent light bulbs are commonly rated at 100W, meaning they suck up roughly 130 times the power input of these little boards, when idle. For the Pi, this means only 6.57 kilowatt-hours of electricity are used per year, if one runs 24 hours per day. At an average US electricity cost of around 10 cents per kilowatt-hour, a year’s worth of Pi monitoring costs roughly $0.70. If you have access to the power grid, there’s little economic gain in going solar on such a small scale.
However, if you need to position your Pi in a remote area, solar begins to make a lot more sense. No one wants to run cable or intermittently change out batteries.
Raspberry Pi Solar Panel Capacity
You now need to figure out how much actual light is available. If you’re expecting full sun power for 12 hours a day, you may be disappointed.
According to Google’s Project Sunroof, my relatively bright location on the west coast of Florida (the Sunshine State) gets 1,768 hours of usable sunlight per year. When divided by 8,760 hours total in the year, this means that the sun is only shining well enough to provide solar power 20% of the time. Based on that, you’ll need a panel capable of providing five times the needs of your single-board computer, plus a battery backup for the other 4/5ths of the time.
This works out to be .75W x 5 = 3.75W panel capacity, which is easy to find in inexpensive fold-out portable packs. For a bit of informal testing, I obtained a solar charger/battery pack rated at 1.5W—enough to theoretically sustain itself during bright sunlight—with a 20,000mAh battery. I successfully ran the Pi on this for a day, allowing me to log on via SSH and check that it was working from my computer.
Solar Power Bank: Final Results
After running overnight, the charge lights were all blinking, noting that it was nearly out of batteries. If the Pi was running for 24 hours, putting a draw of 150mAh on the system, that works out to only 3600mAh of use. When the Pi was plugged in again, it died after a short time, leading me to think that this solar power bank — not purchased from Arrow — wasn’t quite as advertised. I’m also skeptical about its 1.5W charging ability. While it’s definitely possible to implement solar in a mission-critical system, you’ll want to buy quality components, test everything beforehand, and implement the proper redundancies.
In the next post in this series, we’ll go over how to cycle the Pi on intermittently with an ATtiny85, in order to potentially save a massive amount of power.