A Simple Sun Tracker With Very Few Parts

There are a huge number of ways to track the sun if you have some reason to do so. You can use time-based algorithms, or feed in coordinates from the Internet, or you could do it with minimal parts and no electronic processing at all. The latter is how this project from [3D Printer Academy] works. 

One key thing about this project is that you shouldn’t be fooled by the solar panels. They’re not here to generate power for external use. Instead, they’re wired up in opposing polarities to a DC gear motor. The motor turns the panel assembly. As one panel is hit by the sun, it turns the assembly to bring the other panel into the sun as well simply by applying a DC voltage to the motor. The other panel is wired up the opposite way, so if it is in the sun, it brings the other panel into alignment as well.

This serves as a very simple planar solar tracker. If you want to track the sun with minimal parts, this is a very easy way to do it. You’ll just need to put whatever you want to actually aim at the sun on top of the assembly. if that happens to be a larger solar panel, it may be cumbersome and another more complex design may be more suitable.

It’s an ingenious and easy way of tracking the sun, even if it’s not immediately apparent how the device would be useful in its current form. If you’ve got an idea how you would use such a mechanism, let us know in the comments.

We’ve seen other solar tracker projects before, too. Video after the break.

58 thoughts on “A Simple Sun Tracker With Very Few Parts

  1. Ah a new take on it, thanks for post.
    I recall seeing an air/water ballast system that automatically moved the who panel arrays for time of day And season too some decades ago, part of the ballast shadowed by the panel from sunlight vs part not, nicely elegant fully sealed closed setup – other than storms fully reliable and maintenance free other than grease one or two bushes each 3 years or so, which I noticed could have been covered up with UV stable silastic boots.
    Very hard to beat that, though might refine the angles a tad to suit Earth’s orbital wobbles across a decade or so ;-)

  2. very clever, I always love elegant, simple designs that invoke a “I coulda/shoulda thought of that”
    >put whatever you want to actually aim at the sun on top… larger solar panel… may be cumbersome and another more complex design may be more suitable
    how about replacing the 10rpm motor with a geared, faster variety and stand? cumbersome isn’t really much of a problem if the thing sits on a roof for years after assembly.
    >immediate use
    i plan to attach to the electric power steering of a saturn minivan to allow automated vehicular wandering alongside maximized sungazing/ eyeball photosynthesis

    1. I’m afraid I have to agree, despite the cleverness of the design. The whole point in solar tracking is to keep the incident solar rays perpendicular to the solar panel, to maximize output of electricity. This solution guarantees that the solar rays will never be perpendicular to either panel.

    2. It’s a differential system intended to track the sun. It is not intended for power production by the solar panels, any more than the steering on a car is intended to propel the car.

  3. A more pertinent approach would be to use photodiodes attached next to the panel, and 4 of them separated by a cross beam of nothing more than a few centimeter will make for easy 3D tracking due to the proportional shadow of the cross beam on each phtodiode… direct aim at the sun = no shadow, easy to track

    There was such a tracker in Elektor in the 90’s I think

    1. Phot diodes are basically solar cells themselves, they only produce such a small current though, you would still need an amplifier circuit and a power supply to power the motor. At least that’s the way I am seeing it. I have seen where people have used LED’s to construct a tracking system and they need an amplifier also.

  4. It’s ingenious, but stupidly wasteful, as solar panels are completely useless for anything else. The better solution would be to use them to power two op-amp tracker that uses LDRs or photodiodes and H-bridge to operate the motor. Then most of the energy made by panels can be used to power something else, like a power bank…

    1. Cheap, simple and easy to integrate a positional disk/sensor to control a set of main panels. No need for over-engineering. Parts for a diode based solution is actually harder to get by in this developed country. Small solar panels can be found in every medium sized food shop as garden lights.

      1. Wasting two panels is “cheap”? The cost of a photodiode/amplifier solution is likely a fraction the cost of the two steering panels, and far more capable. While those two steering panels can drive a tiny motor, they’re not up to the task to drive a motor big enough to turn a big array.

        This is clever, but from a design standpoint, wasteful and inefficient.

  5. I think the solar panels are short circuiting each other this way. I.e. after the movement, when both panels have same amount of light shining in and produce the same amunt of power, they are just shorting each other out. All power is converted to heat. This will kill the panels over time.

    You have just contributed to global warming by creating waste electronics! Perfect.

    1. As if we haven’t had much more wasteful hacks here.

      Also, the panels are not shorting each other, they are balancing each other. No current flows when the motor is not moving.

      1. If I take two batteries, connect plus to minus for both pairs, you claim there will be no current flow?

        Why would it be different with two solar cells? If you look at the solar cells they are connected the same way. plus of cell A to minus of cell B, plus of cell B to minus of cell A. That is a serial circuit with a short in the end. The motor does not change that.

        1. I would have to agree the panels are connected plus to minus which is shorting the panels out.

          Though the energy produced is rather low when the device is pointing to the sun as each panel is on a significant angle while this works on a small scale I would suspect not would not work on a larger scale.

          And when one pannel is in the sun the other is in the shade so producing negligible current.
          So it’s not quiet like two batteries connected the same way.

          If you wanted smallest possible component count you can’t do much better but it’s not an optimum solution.

        2. At least the angle between the panels is shallow so neither panel should be making much power or heat in equilibrium. You could use smaller panels and put a flat plastic nose across the front to shade the panels when they’re in equilibrium.

    2. Take a hood look at your browser habits insted lmao. The energy used by these panels are allready here. You could claim there is a waste in making the panels. You can make the same claim about everything humans make. So how do you kive your life?

    3. Geez- lighten up.

      Every “art” project depicted on Hackaday is “wasteful.” Most serve no useful purpose other than eye candy. Every o’scope or meta clock is “wasteful” because of the needless complexity or power consumption of the displays. Every Halloween hack is “wasteful.” There is no “need” to animate plastic skeletons. Every sentry gun, balancing robot, and Twittering toilet seat is “wasteful,” because there is no material benefit to humanity to offset their consumption of materials and electricity. So what is your point?

      Is this project a practical application of the pieces and parts it uses? Most likely not. But it is clever, it demonstrates how differential currents can create what amounts to a closed-loop controller with no op amp or computer, and it might, indeed have an end use.

      A structure like this, fitted with a pot, could be used to provide sun-position information to the controller of a much larger array. If nothing else, you could attach a semi-circular scale with numbers at the base of this device, and the apex of the triangle formed by the two panels will point to the time of day–an active sundial, if you will.

      There is joy in experimentation, tinkering, building, and exploring. Also, toys like these often inspire others. The cult of the man-bear-pig might do well to remember this.

    4. Build it and see for yourself if it works or not ! Don’t just make a statement about something without being able to back up your opinion. The motor has impedance which makes it a current limiter. If there is no potential in either polarity than no current will flow. As far as the panels being a short circuit, NOT, I had an array of 400 watts at 12 volts connected to a battery bank with a total of around 1000 amps with no charge controller or blocking diode and there was no issue aside from a slow drain over night that pulled the bank down below useful voltage to run an inverter. If one panel is facing the sun it produces current and the other acts as a small battery or load, but the load is so small because the cells are diodes themselves or they wouldn’t produce DC. They all have forward and reverse currents, forward is higher than reverse. That’s what makes them charge. If they were a short, the wires wouldn’t be able to be as small as they are, they would melt. There were no fried cables or melted batteries, they all came back up by the end of the day. Of course I installed a control system eventually, at first all I used was a heavy diode that would handle 100 amps until I could afford a PWM control system. Do what you must with as much as you can afford at the time. The motor is acting as a differential circuit, it senses the difference the same as an isolated field regulator. The alternator in your car does the same thing, if the battery voltage is different than what the circuit is set to operate at it feeds a small difference to the rotor which is magnetized either stronger or less to vary the voltage/amperage produced by the stator output. Same principle just reversed, it consumes a little power instead of producing it.

    5. The panels are maybe 1-2W each and maybe 5V? Let’s assume 5V is at their peak power, so that would be maybe 300mA of current. IV curves are roughly rectangular, so we can assume their short circuit current is around there; let’s say 400mA. When cells are shorted, the voltage is determined by series resistance. Maybe that’s 1 ohm? So .4*.4*1 = 160mW per panel worst case. At an angle, though, there’s cosine losses, so we’ll call it 100mW per panel, or at least on that order roughly. That doesn’t seem likely to cause too much degradation. It also seems roughly around the total power dissipation for a more active circuit.

      I once had to build a sun tracker for a solar concentrator system mounted on a moving vehicle. I tried the differential photodiode approach, but it just wasn’t precise enough (I needed to be within 2-3 degrees or so). I ended up using an analog (PAL) video camera and decoding the signal with voltage comparaters and MCU interrupts. I used a few layers of overdeveloped film negative as an IR-pass filter, such that the sun was a white dot on an otherwise black frame. That gave me about 250 lines of resolution in one axis with maybe 90 degrees FOV, or ~0.3 degree resolution, at about 50Hz.

      1. Wow, I’m impressed. I assume also that it got easier to decode the signal because you were looking for hot lines?

        Not hot “pixels” (in the X-direction). That twist alone is very clever, I’m in awe.

        1. Yep, the signal was relatively clean. The sync pulses dipped well below (if I recall correctly) 0.7V, and the white pixels spiked well above 0.7. I had two trim pots to adjust the comparator levels. A better implementation could have dynamically detected the thresholds, but I had to hobble it together last minute from parts from Dick Smiths down in Australia. It was an invention of necessity, and I had to make it work with the 8-bit PIC MCUs I had with me.

          It was very easy to just count line syncs to get vertical position, and that’s the axis I used for control. I did approximate the horizontal position too just by timing the difference between line sync interrupts and hot pixel interrupts, but I didn’t use the information for anything other than telemetry.

    6. No, the panels should be connected plus to plus, and the minus terminals to either side of the motor. Or swap plus and minus,and switch the motor connections. It’s a way to get the *difference* in voltage (and thus illumination) between the two panels.

      1. As others pointed out, protect the panels against reverse current with a diode in parallel (I would think Schottky for low forward voltage); anode to the panel ‘-‘ terminal and cathode (banded end for discrete packages) to the panel ‘+’ terminal.

  6. Simple solution, however you basically shortcircuit the two panels, as there is current flowing. There’s no voltage between the two wires and current must be flowing.
    This could work with changing the circuit by connecting both solar panels in series, but with both + wires as the connecttion between the two panels. The black wires will then provide a voltage and current to the motor.
    Connect two diodes to the panels, to make sure that the ‘reverse current’ can flow to the motor.

  7. There is no need to TRACK Sun. It is there, every day. And if you know your position, you know where it was, where it is, and will be.

    Just like a missile ;)

    1. So you are saying that a stationary array of panels will produce the same amount of power if they are not facing directly at the sun as they would if they were 45 degrees from the sun? I think not ! If fact I know they won’t, I’ve been living with solar power for over 18 years now for electricity and all of that time they were stationary and I wish I had thought of this simple design 17 years and a few months ago. I have struggled with so many different circuits and designs hoping to make something work with less expense to do it.

        1. Well hrrm, you do definitively need; azimuth and compass bearing as the key data on each day presumably gleaned from accurate astronomical historical orbital data with maybe a comparatively tiny orbital wobble factor correction of some small % within those bounds – with proportional value to the size of the array overall as each fraction of a % affects overall throughput, which from what I’ve seen could be many 100’s of KWatts.

          IOW. Sure you can use as a base historical data re azimuth and compass bearing in relation to season & GPS and hopefully this is enough however, I wouldn’t discount perturbation factors such as various sources of noise – the way to get around (most of) that is power flow Not facile voltage output – otherwise some primate numbnut borderline amphibious grex could upset your control system with a torch at start or end of daily cycle…

          1. Not azimuth, you’d align it with the earth’s axis and run it at a constant 15 degrees per hour. Though that does leave out small seasonal adjustment which would be an error of a little less than 12 degrees in declination and about 4 degrees in right ascension over a year.

    2. Maybe just like SOME missiles, the ones that do not correct for course deviation.

      Sure. However that involves some sort of predetermined tracking behavior based on the location, and the design is stated as intended to be very simple with no electronics. Unless you are allowing continuous rotation 24 hours, something more than a motor is required.

  8. What a great demonstration of a clever concept. Gets you thinking and looking at making a better and more useful version.

    And for the usual naysaying trolls that have made their way out from under their bridges an Oddball quote.

    “Why don’t you knock it off with them negative waves? Why don’t you dig how beautiful it is out here? Why don’t you say something righteous and hopeful for a change?”

  9. Hm, I didn’t think this through I see, I think you may be right.

    And people discussed this before:



    My uneducated guess is that the current induced may not be enough to damage the panels, how could they otherwise handle a full load use-case? Combine that with when they are in equilibirum, neither panel is pointing directly into the Sun, thus not pushing as much current as they are capable of.

    I am humbled to have been confused by such a simple circuit!

    Another idea came to mind – what if there were 2 motors, each “fighting” other mechanically. No electric connection between the panels, just one panel hooked up to one motor each. This would give much the same result, but the motors would heat up more than the panels.

    Could the original design be improved with diodes somehow?

    1. “Could the original design be improved with diodes somehow?”

      Hello! Yes, I think that could be possible.
      Two diodes, say 1N4007, could be installed on each solar panel. The diodes then would prevent reverse polarity. However, arranging them in the right way would be tricky, I suppose. Also, they’d cause a voltage drop (ca 0.7v each). Germanium or BAT types have a lower drop, 0.3v each. Not sure if they can handle the current of a solar motor, though.

      If two diodes with different orientation met (anode mets to cathode), voltage flows into one of the panels.
      Polarity would be correct, however.
      –> o –>
      — o —

      If the diodes do not met on a single pin, but are connected in a crossed way, the result would be similar.
      –> o —
      — o <–

      Energy may "run in a circle" then, if both cells are into the sunlight. It's like a series circuit were the end connects to the beginning. So just wiring the panels "in series" (connect anodes to the motor contacts, connect cathodes together) may give a similar end result?!
      –+ o o +–

      Alternatively, what about using LEDs instead?
      They can better absorb energy, so maybe they can be used somehow in a reverse polarity situation? Again, not sure about the current that flows. There are also duo-leds, with common anodes/cathodes. Maybe they could be taken advantage of?

      These are just some ramdom thoughts, of course. 😅 In my childhood, I tried to use lamps and LEDs to solve all sorts of issues. A little lamp in series often helped preventing shorts of all sorts. If a motors was stuck, the lamp in series would glow brighter, for example. Likewise, a 12v lamp in series assisted in charging a 12v lead-gel battery. If the light became dimm, the battery was charged. Imagine, a whole battery charger, consisting merely of a lamp. And it worked in a more gentle way than commercial types do, even. And acted as a fuse, even.

      That's why I'd like to say "kudos" for the idea of this project! Designing "simple" circuits like this is more difficult sometimes than, say, programming an AVR that performs a similar task. Because, it requires a flexible mind that dares to think out of pre-defined conventions.

      PS: Maybe there are some errors in my thinking, I'm quite sleepy, too. 😴 I just thought that you deserved some feedback. So long! 🙂👍

  10. This tracker suggests an improvement. Set the two panels side by side, in the same plane. Put a shade between them, at right angles. When the pair directly faces the sun, both panels have the same voltage. When the pair are tipped, one or the other panels is partially shaded, and so generates less voltage.

    Wire the two panels in series. Their output can be used normally, to power anything you like.

    Connect two equal-value resistors in series across the panels. The voltage between them is half the total panel voltage. These resistors, and the two panels form a “bridge” circuit. The voltage difference between them tells you which way to run your motor. This voltage can be sensed with an opamp, comparator, a pair of transistors, or even a sensitive relay. When it turns “on”, it powers the motor with the full output from both panels.

    1. First useful comment amongst a plethora of naysayers.

      However, the use of active components (transistors/opamps/etc) negates the simplicity of the design. I’d bet that something with a perpendicular shade could be done without any active components…

      Personally, I’m happy with fixed panels. Even though they aren’t as efficient, they are certainly more robust. And unless you are severely space constrained, it is cheaper (and more reliable/robust) to install more panels.

  11. This is an updated (3d prints?) version of a BEAMbot described in the 1990s. The thing missing from the original design is that a 3rd panel for generation was mounted at an angle (by experimentation) such that IT faced the sun directly when the rotator panels overshoot.
    Also this is the base concept for one which used a capacitor and two motors to make a sun seeking bug.
    All the whinypoopets need to chill. BEAM is an incredibly cool robotics challenge. We need to see a lot more of it in HAD.

  12. Bad on many levels; it’s essentially a continuous short circuit. If the solar panels don’t contain a series diode they will be damaged when connected that way. Also, a lot of energy is being wasted when the panels are lit but there’s not enough energy to turn the motor. The correct way to make a solar follower is unfortunately more complicated and needs light sensors to read the sun’s position (panels can do that but under load the perceived read position could be altered), then the motor should be driven using PWM so that it can still turn very slowly when the panels are slightly off axis with the sun.

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.