Hardware Store Goods And An Mbed Combine Help Solar Panels Track The Sun


If you have the space, and can build a tracking rig cheaply you’ll be able to get a lot more out of your solar panels. That’s because they work best when the sun’s rays are hitting them perpendicular to the surface and not at an angle. [Michael Davis] hit both of those stipulations with this mbed powered solar tracker.

At a garage sale he picked up an antenna motor for just $15. The thing was very old, but still wrapped in the original plastic. It’s beefy enough to move his panels, but he first needed a way to mount everything. After checking his angles he built a base out of wood and used galvanized water pipe as an axle. Cable clamps mate his aluminum angle bracket frame to the pipe. This frame holds the panels securely.

To track the sun he used two smaller cells which aren’t easy to pick out in this image. They are monitored by the mbed microcontroller which measures their output in order to point the assembly in the direction which has the most intense light. A couple of limit switches are included to stop the assembly when it reaches either side.

This technique of using small solar cells as the tracking sensors seems to work well. Here’s another project that took that approach.

23 thoughts on “Hardware Store Goods And An Mbed Combine Help Solar Panels Track The Sun

  1. Just curious, have you done any checks on system efficiency? I’ve looked at projects that aim solar panels before, and they universally use more energy to aim the panels than they gain from aiming the panels…

    1. A well known ballpark rule for estimating panel output is that panels which are fixed at an “average” sun angle will produce about half the power of panels that follow the sun. The kind of trackers I’ve seen are stingy with power consumption and don’t come anywhere near using as much power as you describe. You might want to revisit the question of how much power an efficient tracking system needs.

      1. I’ll admit, the other projects I was looking at were a little dated, so newer technology for aiming them may be more efficient, but the article says it was an older motor.

        1. I would have thought that with a balanced system, it would take very little power (using say a screw drive so that loading from wind/snow doesn’t mess things up) to actually move the thing around.

  2. Do you guys have any data to support your assertions?

    Here’s one report that says you’re probably wrong:

    A single axis tracker generates about 30% more power, and a dual-axis one about 80% more.

    Now we have some figures then if we know the power consumption of the motor(s), or the cost of the motor(s) we can say if they do or do not generate more power than is used, or cost more to implement than adding another panel.

    Especially the first poster, since you have looked at so many projects like this could you please send a link to just one of them that demonstrates your assertion?

    1. no, the light sensor thing is easily fooled by bright clouds, and hence these kind of trackers usually end up all looking into the most funny locations, but not the sun.
      Just calculating the theoretical sun position with the use of a realtime chip is superior usually. Also, its easier.

  3. As I read this I wondering if this is the same dude whose wind turbine project I recall, he is. I have seen where in other projects where unions where used the builder drilled, tapped at least one hole into the nut to used a set screw to lock it in place. For my location I’d build the legs of the ground mount longer with pads or trays for concrete block at the ends of them. No doubt that would require beefier wood joinery so they don’t become a weak ink. All in all a great build log of a good design, that one can modify as they need. I’d like to find that rotator as new old stock for only $15 myself, I need one for it’s intended purpose.

    1. Well, yeah, if that simple motor was connected to a very complex set of gears to account for the tilt of the earth, time of year, local latitude, local time preferably in solar h:m:s instead of sidereal, and you were willing to rebuild the whole thing every so often to maintain accuracy.

  4. As someone who did homestead and farm solar applications design and installation professionally I’ll just pose some questions to ponder;

    – why use optical tracking when a clock/calendar gives more reliable sun position (without wasteful array “hunting”)?
    – what happens to the required array footprint with more than a single tracking panel?
    – what are the implications of wind gust loading on large axially mounted PV arrays?

    Keeping in mind that insolation (array current) is proportional to the sine of the sun angle, in the paper quoted above;

    – why do the fixed (supposedly reference) panel values never match the tracked panels even at solar noon?
    – why does right ascension tracking give 32.17% improvement while declination tracking (a much smaller angle) supposedly gives an implausible 81.68%?
    – why do panel manufacturers specifically warn against any concentration by “placing a mirror or concave lens on top of the panel”?

    Overall, in practice why does Switch Mode Maximum Power Point Tracking give better results than mechanical tracking (and integrated “load shed” use of excess capacity beat both)?

    1. “Switch Mode Maximum Power Point Tracking” and “mechanical tracking” are two completely different things though – MPPT is about modifying the current draw from the panels to keep them at their optimum voltage, for a particular input light level. (Drawing too much current from the panels causes their voltage to fall *and this affects the efficiency* electronically).

      Mechanical tracking maximises the cross section of the sun’s rays that strike the panel, increasing the input solar energy.

      Array current is an electronic measure; insolation is an optical measure of how much sun you’re getting.

    2. Concentration is warned against because efficiency drops rapidly with temperature. Non-uniform illumination of a panel, as you’d get with a concentrator not properly aligned with the sun at all times, also causes a surprisingly large efficiency loss. Or so I’ve read.

      I do wonder however, if it’s possible to come out ahead with a concentrator/tracker with standard solar cells (not small triple-junction cells) and some sort of heatsinking, or even water cooling.

  5. There are a couple of other reasons that you might not need/want to track:
    1) If Off-Grid, then the real gain would be made during the summer when if fact you have ‘buckets’ of power. When harvesting power you need somewhere to store it.
    2) In a windy enviroment you are putting your panels at more risk of damage. A sturdy fixed frame would be significantly stronger than a tracker.

  6. When considering solar panel output, there are quite a few variables to be considered…. There are differences between the gains that you will see in a sun tracking setup at different latitudes ….. There is a difference in the gain from a “MPPT” system in hotter climates due to the voltage/current output curve changing in different temperatures, as well as at the overall photoelectric efficiency changing depending on the intensity of light hitting the cells….. This isn’t a simple matter. I have seen overall system output gains of only 15% in a dual axis sun tracking setup in the past which was right next to an otherwise identical setup which did not track at all but was fixed at the “ideal” angle. It is no doubt that sun tracking setups were dramatically more economical and desirable in the past when the solar panels themselves were much more expensive than they are today…… If you could boost the output of a 100,000$ system by 20%, then as long as the tracking system cost less than 20,000$ to implement, then it was worthwhile versus just installing additional panels. But when an identical system may cost a tenth of what it did 40 years ago????? The downside of “MPPT” regulators versus linear regulators in systems which require tight voltage regulation is the dramatically increased cost of the regulator, in addition to the much increased complexity of it’s system. Put simply:There is a lot more that can go wrong.

    All in all, I think that a much more productive and helpful ‘hack’ would be the development of an efficient and robust open source MPPT system which can be adapted for different system voltages etc….. The commercial offerings available leave much to be desired in terms of cost and reliability in addition to their lack of flexibility.

  7. Additionally, often times the efficiency of the panel can be increased by double digit percentages by simply applying an antireflective coating to the “topside” of the glass. In these homemade panels there is also room for improvement where there are losses due to reflection from the cells back up to the glass.

  8. this could be true in the past but not nowadays because solar cells have a much lower price.
    there’s no way a solar tracker is going to be worth it over just buying more solar cells!!!!
    this is pure time waste!

    1. Thinking and building things by yourself is never time waste. Playing WoW, CoD or Angry Birds IS time waste (and fun).
      This guy may or may not have the money back in his setup, but he had a theory, designed and built a working circuit, and this it-works!-feeling is amazing. I always have theories, some times I design a circuit, sometimes they work. Nobody cares, but I like when they work.
      Lots of people don’t even get any ideas and shout “is a waste of time”, “useless” and so. It’s like a nine grader that don’t write any poems telling a first grader that his poem is useless and boring.
      Building something is useless, right? Then go to any shop and buy everything, as people already thought for you. Thinking is hard, buying is easier. Let the pros do the thinking, and pay them for it.

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.