Sun-powered Stirling Engine With Automatic Tracking

Check out this solar-powered Stirling engine (translated). The build is part of a high school class and they packed in some really nice features. The first is the parabolic mirror which focuses the sun’s rays on the chamber of the engine. The heat is what makes it go, and the video after the breaks shows it doing just that.

But the concept behind the mirror makes for an interesting challenge. The light energy is focused at a narrow point. When the sun moves in the sky that point will no longer be at an efficient position to power the engine. This issue is solved by a pair of stepper motors which can reposition the dish. It’s done automatically by an Arduino Uno which makes readings from four LDR (photoresistors) in that cardboard tube mounted at the top of the dish. If the light intensity is the same for all four, then the tube is pointed at the sun. If not, the motors are tweaked to get the best angle possible.

[Thanks Lionel]

32 thoughts on “Sun-powered Stirling Engine With Automatic Tracking

  1. there is a tad bit of a simpler way to do this… no need for an aurduino.
    The book junkbots bugbots and bots on wheels has a thing like it. Then all you would have to do would be a small solar panel attached to it, like the cardboard tube.
    If there is no aurdiono, and no stepper controllers, then it should take less power.
    Also, the design in the book allows for use of tiny solar cells instead of photoresistors. Even more power.

      1. I suggest you get the books written by Forrest Mims. They are a great introduction to basic electronics. and he covers this in one of the very first circuits.

      2. Barefoot its simple. Not sure what implementation roboman was referring to, but one way is to have DC motors hooked up to solar panels, with baffles so that the solar panels don’t get any sunlight if the rig is aligned with the sun. But then as soon as the sun moves, the baffles are arranged so that the solar panel attached to the right motor gets sunlight and the system moves towards the sun.

        There’s a LOT of stuff you can do without microcontrollers. We flew to the moon with less computing power than an Arduino.

    1. The great thing about the parabolic dish and sterling engine setup is that if it’s not energy positive, relocate the engine, and add more dishes chained together. Same motors, negligible increase in power usage, double, triple, quadruple, whatever, the amount of heat input to the engine. Gets slightly more complicated aiming them all correctly at the engine, but that’s what the horsepower of the arduino is for eh?

    1. He can also use the micro to ‘Home’ the dish at dusk so it catches the early sunlight. The micro could also be used for data logging. You can’t do that with basic electronics.

  2. I see a lot of comments with people talking about using analog stuff over the arduino.

    well, before we had matches we used flints. I don’t see many people using flints anymore.

    These things are cheap and easy to use. If he wants, he can add new functions to the arduino as he thinks of them. Logging power output for example. As far as learning a new skill, Coding a microcontroller is more relevant.

    1. Its definitely a bit overkill to use sensors to track the sun, since its position is always predictable. The nice thing though is being able to throw the unit down and always have it in sync with things without having to do any alignment. In fact aligning the thing with the earth’s axis and properly setting up your timer circuit could actually be more complicated than just using sensors.

      1. I always wondered about this. I thought the movements of the sun were pretty well understood at this point, and wondered why systems would use sensors over a clock and some math.

  3. The op-amp circuit would cost what, less than a dollar? and probably have an equal component count to the external stuff they had to glob onto the arduino…

  4. He could have used a slave to direct the plate to the sun. This technology was available for more than 3000 years and still works! Why do we need an Arduino for a job that a cheap slave can do?

    1. Please give me the link to the slave store you’re using, as mine are very expensive.
      Also, I guess mine are not top-notch because after a bit of time, they get exhausted and start to show a lot of drift.

  5. I would love to do a fully mechanical solution. Two or four small areas around the engine’s hot-plate could be used as mechanical ‘sensors’. These ‘sensors’ would simply be the end of a tube filled with a liquid that turns to gas at low-ish (180°F) temps. This tube would connect to a cylinder that adjusts the dish in the corresponding direction.

    If the focal point is low, the low sensor heats up. This low sensor is tied to a cylinder/piston that raises the dish moving the engine and sensor up. Since the entire thing would be balanced, the pressure in the lower sensor/actuator does not need to be maintained.

    Simple, reliable cheap and bulletproof?

  6. There’s no details on the sterling engine itself? I’ve had a very difficult time finding any commercially available ones that could do anything useful, it seems they are all desktop display trinkets.
    And the DIY ones I’ve seen are a joke, you aren’t going to spin an alternator for long with some cut up beer cans.

  7. This thing is on a alt-azz mount. It needs to adjust around and up-down. It should be equatorial mount then only one constant slew is all that is needed to track the sun or stars. Like a telescope. Then the second adjustment is seasonal. With manual or auto reset to east only one motor sense is needed instead of four. This can be as simple as small PV runs motor till shadowed by the slew.

    1. Alt-Az is much simpler than an equatorial mount, however. Have you noticed that almost all the latest GO-TO telescopes use an Alt-Az mount? If you’re not doing photography, then field rotation is not an issue and Alt-Az is much cheaper. You say the number of sensors is reduced from 4 to 1, but sensors are dirt cheap, and machining an equatorial mount is not.

    2. If I’m understanding your terminology right, that won’t work.

      An equatorial mount can track something like satellites in equatorial orbit, the path of which does not change with the seasons.

      The sun’s path across the sky is quite different from winter (low, in the northern hemisphere) to summer (high), so wouldn’t you end up with a mount that only traced the right path twice a year?

Leave a Reply to meCancel 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.