Comparing Solar Energy Harvesters

There doesn’t have to be much more to setting up a simple solar panel installation than connecting the panel to a battery. Of course we would at least recommend the use of a battery management system or charge controller to avoid damaging the battery, although in a pinch it’s not always strictly necessary. But these simple systems leave a lot on the table, and most people with any sizable amount of solar panels tend to use a maximum power point tracking (MPPT) system to increase the yield of the panels. For a really tiny installation like [Salvatore] has, you’ll want to take a look at a similar system known as a solar energy harvester.

[Salvatore] is planning to use an energy harvester at his small weather station, which is currently powered by an LDO regulator and a small solar cell. While this is fairly energy efficient, the energy harvesters that he is testing with this build will go far beyond what an LDO is capable of. The circuit actually has two energy harvesters built onto it which allows him to test the capabilities of both before he makes a decision for his weather station. Every amount of energy is critical when using the cell he has on hand, which easily fits in the palm of one’s hand.

The testing of this module isn’t complete yet, but he does have two working prototypes to test in future videos to see which one truly performs the best. For a project of this size, this is a great way to get around the problem of supplying a small amount of power to something remote. For a larger solar panel installation, you’ll definitely want to build an MPPT system though.

11 thoughts on “Comparing Solar Energy Harvesters

      1. Would the supercap help with over curent protection providing time for a fuse to blow
        If say, pv cells being used indoors, producing something like 5mA were taken out into full sun causing current to hit 600+mA?

  1. I’d like to bring up Gleyzes’ project based around an ESP32 and a generic buck converter. It’ll do just as well as the dedicated commercial solutions, can be adapted to other MCUs (could even do one of those $0.04 MCUs from LCSC!) and bucks (or make your own buck with the MCU’s timers!), is typically way cheaper, and you can reliably find the parts. As a side benefit, you also have an MCU that can do other stuff between MPPT calc interrupts!

    1. Cool project, but wont work off the solar panel featured in this video as I believe it’ll use far too much power and just drain the coil cell battery to zero. What we’re ‘supposed’ to be talking about here is tiny solar cells and coil cell batteries. I bought one of Jasper Sikken’s solar harvester boards a while back ….. not had a chance too use it yet as my main PCBs are stuck in some dark rat infested place on route from Seeed PCB fission. Should have gone to PCBWay :(

      1. Good point! I’d have to look at the design again, but even if this exact layout isn’t viable a very similar one would be (really just depends on layout of MPPT gates and pullups/pulldowns). When voltage from the panel drops below usable, put the MCU in hibernation, where it draws 1 microamp. Wire the panel output to the wake pin, and it’ll come back to life when the sun comes back. With a CR2032 (battery size for this project) of 240mAh, that gives a battery life of 10,000 days! Even if you want to maintain RTC, that only quintuples power draw, so 2,000 days. And, after digging through the project this will power, he’s already using an ESP32 with a deep sleep cycle for the weather station!

        Now, there will be other sources of quiescent power draw on the battery, but on a well-designed board they will all be orders of magnitude less than it takes to run the power station itself.

  2. Ummmm ….. this looked quite interesting, but is it just me or did anybody see any meaningful results? I saw what looked like the profile of a flashing LED on the ‘scope, but not a lot else :( BTW, PCBWay are a good company to get PCBs made from, if a little pricey.

    1. They can be super cheap as long as you get a simple PCB, $5 (plus $25 shipping) for 5x 100x100mm 2-layers with 5mil/5mil tracks, an extra $15 for lead-free, an extra $20 to go up to 4-layer (pretty necessary if you want to do anything high-speed and generally convenient), and an extra $1.50 to keep them from adding their PN to the silkscreen. Some of these costs have overlap, so you can get 5x pretty decent boards for $55 including shipping, or $30 if you’re fine with lead and 2-layer.

        1. Yes, in order to keep the output from the solar cell near max power point. The issue is that there are a number of algorithms, and a lot of the cheaper implementations utilize the less efficient, but computationally cheaper, ones. Any comparison without actual efficiency/power output numbers isn’t very useful as performance can differ drastically between products.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.