Suntracker Optimizes Solar Panels While Visualizing Sun’s Path

If you have solar panels, you want soak up as much sunshine as you can to get your money’s worth. If you don’t have space for a lot of panels, the next best thing is repositioning the panels to catch the most rays. For his entry into the Hackaday Prize, [Frank] built a gorgeous solar tracker prototype to both validate his theories and to serve as a learning platform.

A solar tracker’s purpose is — you guessed it — tracking the Sun’s location to determine optimal positioning for solar panels and other sun-seeking payloads. In the latest revision, [Frank]’s tracker follows the Sun’s azimuth angle, aka its horizontal movement.

The Sun’s path is represented along a ring of 32 red/green LEDs. It moves around the ring as a green LED, according to a real-time clock and a set of pre-determined solar positions stored on an SD card.

Two red LEDs show the sunrise and sunset azimuth angles, and a third LED indicates North as detected with a magnetometer and adjusted for local magnetic declination. In the center of the ring, a stepper motor drives an arrow that always points at the Sun LED. As the tracker is moved around, all the LEDs shift around the ring to follow their targets.

Though it already shines, we think this ongoing project has a bright future. Be sure to check out the demo video after the break.

31 thoughts on “Suntracker Optimizes Solar Panels While Visualizing Sun’s Path

  1. So you can’t just go outside and see where the sun is?

    have the panels on an axis/rotor pointed at the north star that rotates every 23:56. Then shift them at an angle at that shaft that goes up or down dependig on the date – they will be parallel at the equinoxes, and 23 degrees at the solstices.
    Sun compasses (with elevation! which doesn’t appear to be in the build, or a GPS)

  2. useful or useless, whatever. This is so overlay complicated, you’ve got to love it.
    The 3-double indicator (LED, needle and OLED (actually two OLED displays), the lightshow before it starts. The layout is nice and tidy, the PCB is well protected by the plexiglass. Then there is the HUGE pcb, it must have cost a small fortune to make. Although everything in me screams that this is not the way to solve a “technical” problem… somehow I like it! I think I file this one under “art” and I can only conclude, nice project and well done!

    1. The solar panel can tell you when it is pointed at the sun by tracking the output, an ESP8266 can grab the exact time for the location then the sun position is calculated in the usual manner. This means that a sweep of the panel on two axis will set it correctly then the change in time will allow for the position to be continuously changed to match the sun position. You can get extra smart and work out the solar time just by the sun position, so the project can be both very simple and very “smart” if you design it well.

      1. I was contemplating that. I guess that might be better for larger setups. Maybe a half sphere with lots of gas cylinders painted in vanta black, with cowling around each one, each attached to it’s corresponding pneumatic cylinder so the combination of the moments keeps the panel pointed in the right direction.
        It would be nice to forego the tracking all together and have light pipes and prisms, so no matter what direction the light always comes in at 90° but generally the materials have too many losses.

        1. “generally the materials have too many losses. is not the problem. Materials themselves can have arbitrarily low and insignificant losses. The problem is simple geometry. You simply can’t arrange any sort of optical system to take light from a set of arbitrary directions and somehow direct them all into one direction into a finite aperture (like a solar panel). The general concept to search for is Etendue

          1. Perhaps some sort of fractal etched layering. A cross between a bell curve distribution peg board and a Fresnel lens. Increase the refractive index nonlinearly through the material. It should work with radio waves amd sound too.

        1. Probably not, these are usually for Ground Mount setups. The roof gets in the way for an east west setup in that you won’t be able to point the panels at the sun past a certain angle depending on the slope of your roof. The other thing is, lifting one side the panels off of the roof is probably not a good idea. They become a sail attached to what could be your new a-frame house boat when the roof comes flying off in a big wind. Just add a body of water !

    2. Or this jaw-droppingly-simple electromechanical solar tracker built by NighthawkInLight, inspired by the Parker Sun Probe. It uses not much more than four solar cells, two small DC motors and s shade baffle. I think it’s the most elegant design I have ever seen (Youtube video):

        1. Sure equatorial would be simpler and cheaper…
          And tracking system could also be reduced to… nothing… just a 24h motor cycle… unless the earth decides to change it’s daily rotation :-)

  3. I would be interested to see the curve on the graph of the power generated. On my fixed panels it is pretty much a semicircle with maximum power only available for a short time.

  4. Unless your solar power panel is on a mobile set up (e.g. RV), the electronic compass is a waste of time. The more high tech junk you have outside exposed to the elements, the more likely they will mess up in the long run.

  5. I don’t get it, where does the sun shine onto? Give us a simple clock/ calendar, an algorithm (including stowage), some thought out drive mechatronics, and I’m sold. – Fanciful play!

    1. The sun changes it directions with time and so to say with the sensor at the robotic arm of the riser the sensor detects the movement of the sun through the help of the solar optimised.

    2. Wan’t there some ancient dude, someone scientist with a really good reputation for inventions, working on a purely optical, passive tracking system? I remember seeing the release, stating that a very significant and revolutionary disclosure was imminent, but that was the last I heard?

  6. I’ll go along with the “art” judgement. These days for small solar projects, it’s cheaper and simpler to simply have several cells in an arc that matches the trajectory of the sun, with blocking diodes to shut off the current drain in ones that aren’t putting out much.

  7. >” you want soak up as much sunshine as you can to get your money’s worth”

    Depends on whether you want to sell it, or put it in a battery. That’s the Achille’s heel of solar power: what happens when everyone are “getting their money’s worth” and pushing power to the grid at exactly the same time? Well, of course the price plummets to zero.

    Except when the price can’t go to zero thanks to feed-in tariffs, right-of-way, and net metering laws. The price goes right past zero and into the negative – utilities are forced to take the solar power, pay for it, and then pay someone to take it off of them in turn.

    1. Or, from the other point of view: what happens if everyone’s not selling their power and instead using it to power their homes in favor of buying electricity. Well, the demand goes to zero, and then it goes back up a couple hours later. This on-off action on the power grid then demands the use of inefficient natural gas turbines to take up the slack both ways.

      From a systems point of view, it would be far better if half the people aimed their solar panels at the morning sun, and half the people at the evening sun, which evens out the mid-day peak and makes the output conform better to the actual demand – but this doesn’t produce nearly as much energy or money from subsidies.

      1. Or simply, plug you EV to charge while it’s sitting on the parking, and consume that energy when you need it by night. How come it’s not forced by law you can extract energy from a 36kWh battery you have sitting in your property yet ?

        1. That’s assuming you have an EV, and that it’s parked at home, and that it has capacity instead of being full all the time because you don’t drive all that much.

          And it assumes the battery is cheap enough that you don’t mind wearing it out prematurely by cycling your house power through it.

          For example, a battery that lasts 2,000 cycles and costs $150/kWh actually ends up at 7.5c/kWh on top of the price of your power. If you wanted to store solar power, there are cheaper alternatives than vehicle batteries.

  8. Why?

    Free software can show you were the sun will be through the year.
    Blender 2.8 and Sun Position.
    Model your buildings and trees, run Sun Poisition for lat & long, date & time.

    Or get a plugin that does all of the PV analysis of the site’s potential, so you know what tracking panels will generate, what fixed panels positions where will generate and where to place fixed for optimum.

    Or other software that does it without the 3D renderings.

    1. Aww, but what would the fun be in that?

      I’m still mystified why they took the trouble to precompute sun position and put a big table on a SD card for a small number of fixed locations. The micro would have no trouble computing it on the fly, for arbitrary locations, even real-time from GPS. It might even take less program space than the SD file I/O routines.

      The code on github even includes the location, hardcoded in the source. It’s here: https://goo.gl/maps/mkNmT8xSwUVMgiJZ9

Leave a Reply

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