A Simple DIY Solar Tracker

solar panel

The sun is a great source of energy, however, efficiently collecting this energy can be hard to do.  One thing that can improve the results of solar use is to actually track the sun’s movement. [fanman1981] hooked up his own homebrew solar tracker using some pretty clever techniques.

For this hack he used two Harbor freight 45 Watt solar kits, some struts on the bottom, and a Dish Network satellite dish bracket hooked up to an “old school” antenna motor.  Although one might think this device tracks the sun using some sort of sensor, it’s actually just a matter of hooking up the device to rotate at timed interval with a remote control.  This interval is figured out with the benefit of some charts on livingonsolar.com.

To see it in use, check out the video after the break.  He gives a good explanation about how everything was put together, but if you just want to see it move, fast forward to around 5:26, really quite impressive.

[via hackedgadgets]

40 thoughts on “A Simple DIY Solar Tracker

  1. Quite clever with the antenna rotor.

    For maximum efficiency you would also want to rotate the panels on a second axis to have them adjusted 90 degrees towards the sun both horizontally and vertically.

  2. Very cool, just like the ones we were buinding in the 80’s out of old 10meter Sattelite mounts and jack arms.

    IF you wanted even simpler, the sun is easy to track, so you change out the motor so that it sweeps from left to right at a slow pace that matches the sun. ZERO computer needed. when you hit the far right limit switch you make it wait until the solar voltage drops low enough to release a relay that makes it go back the other direction to the other switch. when the sun comes up and the solar array makes enough to trip the relay, it goes back to tracking.

    Works great, no arduino needed, no PC, no sensors other than limit switches that cut power.

    1. To fartface:

      Im using 2 photocells from radio shack and a 2-1K cermet potentiometer as the light sensor to build a suntracker. Its hooked to a 12V automotive relay from radio shack. Cannot get the output from the photocells to operate the relay. Trying to figure out where to ground the photocell circuit to the relay? Any suggestion?

      – Thanks

  3. Yup. Shadows on cells like that are a major no no. Some panel assemblies handle this better than others. But on most arrays, a shadow like that across multiple cells is creating a major speed bump in efficiency.

  4. Oh, and we forgot to mention, that the motor to drive that tracker is eating the other 40% That array is 90 watts peak. – the wrong wire sizes, – the tracker motor, – the shadow on the panels, Should leave him with enough juice to charge the batteries for the tracker remote.

    1. I am working on a open source solar tracker with polarmount / actuator and controlled with a arduino. Arduino is made as a webserver and connected to internet.
      The solar tracker data can be viewed on a smartphone, iphone, ipad and any other items with internet connection. I will make a schematic, a prototype printed circuit board and arduino sketch Free for all. Language on my site i danish, use google translate on site or i might translate t english if you find this solartracker useful. First hardware i a Lego model and a gear motor simulating a polarmount.

  5. @Pete – You’re correct, “tracker” is probably a bit of a misnomer… Maybe it should be sun “predictor.” :)

    @fartface – Interesting, any way we could get a link to that info?

  6. if that bottom flange is all that’s holding it, the thing won’t last that long (just 100lb at a panel 2 feet from the axis would be pulling with about 2400lb at one of the screws)

    and chris is right, if you want to go with only one motor, do a polar mount: have the panels rotate around a pipe that is parallel to the earth axis. you avoid a second motor and can even secure the axis on both ends. (since it isn’t sticking up vertically but at about 40°)
    then, if you want to be super good, tilt the panels down a few degrees in spring and up in fall. for a rough timing: solar noon is south and each hour is 15°

  7. It is a little known fact that solar panel capacity factors are generally quite poor.

    I remember reading statistics from Germany where it turns out all the solar panels installed on government subsidies are producing roughly 10-15% of their nominal capacity, precisely because of no tracking, shadowing and weather, and the fact that half of the 24 hours is night.

    I think the practical potential for solar power is greatly exaggerated. At least you need a lot of batteries to give you any sort of stability.

  8. Its quite a smart idea not to use real tracking. Just is prone to errors. Predicting is better, it is absolutely clear where the sun will be when.
    However, I see the stability aspect of that construction critical. What happens during a thunderstorm? This thing is not going to withstand much force.

  9. Good Lord :( Tracking as used in this context is NOT a misnomer, if it points the array towards the sun during the day it’s tracking the sun, that’s as far I’ll venture into that peeing contest.

    I have to wonder about the long, and term durability of that antenna rotator. Most likely seasonal adjustments(as another mentioned) of the elevation would out way the cost, and complexity of adding continuous elevation adjustment. Could be better off putting that cost towards an additional panel array. Today is the first I read anything on the order that something that shades less than 40% of a solar panel will create a 40% loss in power output.

    There are are many fact’s, figures coming out of Germany, enough one can cite them to support what they desire to support. Many in the USA for whatever reasons choose to live off the grid, and photo Voltaic make up a good portion of their renewable energy power sources. Not that I’m saying that mean PV can do the same for grid power. At some point in time society is going to have to place a value on conserving finite fuels, and not polluting the environment.

    As always this is one persons ends to a means, we may have done it differently, but no one is required to duplicate anther’s project.

    1. If the power output on one cell drops the entire panel’s output drops proportionally. (unless the panel has bypass diodes). Also, the whole string on an array will also suffer the same drop in power output if a panel in that string suffers a drop in output due to shading. Hence the reason why less than 40% shading is capable of output losses greater than the shaded percentage.

  10. @D_

    Thanks – this is certainly tracking. I think a lot of people are getting confused about open loop and closed loop.

    @Steve C.
    I’ve never seen a UV rated white zip tie (so I generally agree). I have, however, seen black zipties that do not have any UV rating. (full disclosure, this was in China not in the States).

  11. “Today is the first I read anything on the order that something that shades less than 40% of a solar panel will create a 40% loss in power output.”

    Two reasons.

    If a part of a single cell of photovoltaics isn’t recieving sunlight, the voltage at that point will dip, and the charge from the surrounding areas will leak into it. Almost as if you had batteries in parallel, and the one with the weakest voltage will draw current from all the others, except in this case the charges don’t balance out, so shading one bit is equal to shading the whole cell.

    The individual cells in the panel are in series because the voltage of a cell is rather small. If one cell isn’t producing power, it’s simply sitting there acting more or less like a resistor and wasting power.

    And the cost of solar panels is a pretty good indicator of how much -other- energy they require to make, because by far solar panels are not made with solar energy. If you can’t produce enough energy to offset the cost, chances are you’re still indirectly using more limited fossil fuel resources in building/infrastructure/maintenance than what you gain from the solar panel.

    Just because part of that cost is invisible (cheap coal in china used to refine minerals) doesn’t mean it isn’t there.

  12. I made a solar tracker once, for a mobile application. Two axis and simple, analog electronics. Four photocells, four comparators and some resistors driving the switches to the axis servos. Didn’t implement it full-size, however, ran out of time. I guess a fifth photocell with buffer would be required for good stability in a wider range of lighting situations.

  13. The best “tracker” is a clock.

    Once you understand that the energy output of an array is proportional to the *Cosine* of the sun angle, “power hours”, and wind loading, you understand why trackers are pointless.

    Been there, wasted my time on that.

  14. … Get a black plastic film canister put an LDR in the center of the bottom pointing up and a small hole on the centre of the lid (to track when the sun is over head) maybe putting nine LDR’s in a grid to get more accuracy or maybe just a few canisters at different angles. Then put an LDR in a plastic film canister that is slightly translucent (to compare against the direction to find the ambient light level).
    Use the ambient light level to stop it trying to find the suns direction at night or on a cloudy day. And mount the directional one to the panel its self (at ninety degrees). Have the panel rotate and gradually chance angle once per revolution. Cycling up and down. Stopping when the ambient light level drops below a set point or the optimal directional brightness is found. Just a few simple resistors and relays to control it. Switching in and out a set of battery’s from being charged by the solar panel also used to change it’s orientation any exess power is sent off to be used.
    Maybe it’s current state could be reported back to a pc. Or my favorite just a few LEDs on a board on the wall. With the option to manually move it for cleaning and repairs. Or lock it’s current position.

  15. The power gain with a physical tracker is roughly 10% at best.

    A switch-mode “maximiser” or electronic Maximum Power Point “tracker” that matches the panel source resistance to the load will give more than this with securely fixed panels, and effectively zero power drain.

    Unless the array has an output of 1kW or more physical trackers simply aren’t worth the trouble, and continuous trackers may indeed use more power than they gain. Physically tracking an array of 1kW or more is an interesting proposition since it will have an area (and wind loading) of a garage door.

    Most stand-alone solar electric systems have undersize batteries for the array capacity and require regulators to avoid damaging the battery set, typically come up to charge by noon, thus wasting the afternoon power hours.

    It is better to spend the money on a bigger battery set and throwing the regulator away; on a windmill to provide power input diversity; and to use productive load-shed such as refrigeration or water pumping for overcharge regulation.

    After some consideration you realise that a clock is superior to an optical tracker, and after some more consideration that physical tracking is a dud.

  16. @aussietech
    yes and no:
    i don’t see your power argument: a well-built tracker with a gear system, counter weights, and a well balanced panel would need very little power to move the panel a degree every four minutes. on the other hand a solar panel that is two hours out of alignment already loses 30%.

    on the other hand you’re right: for the price of the tracker setup, the extra hardware to make it wind resistant (and the continuous maintenance!) one could just get a second panel. -unless you can do a lot of the work yourself.

  17. 30%! Where did you get that from?

    360/24=15 deg/hr.

    2 hrs = 30 deg.

    Cos(30)= 0.866, that is 13% loss at the *extremes*.

    The approximate *overall* loss (10am-2pm solar) is something less than 10%.

    The couple of hours before and after zenith are called the “power hours” because of the loss effect of the slant angle through the atmosphere. For the same reason it is normal practice to mount arrays perpendicular to the winter elevation, the summer takes care of itself.

    If you really want to move your array it only needs to adopt three positions, morning, noon and afternoon. *Continuous* trackers are a joke, optical moreso, and most of the time there is a wind load the tracker must grind away against. Nothing like having done it professionally to know the hidden pitfalls.

    1. i took a 56 volt panel, using a kiec dc/dc regulator(reducing to 12.8 volt), a 5 castor drawing stool, two 7.2 v nicad rechargeable hand drills(used one for jack screw,(elevation), a gearbox off a dead grinder (for azimuth), LM317’s, 555 timers for differential amplifier and 4 CS photo resistors to control azimuth and elevation, with a dish network throw-away sat dish( using as radio telescope with freeware Skypipe). Just a fun project, to charge a couple of marine deep cycles, for Ham radio demo to Boy Scouts and Cub Scouts. It’s fairly portable. You are right on about ‘power hours’ and rate of return for investment. The az/el operates on the drill batteries and of course the torque from the drills. The Skypipe program is driven by my laptop sound card, which has an 800 hertz tone oscillator driving it, the signal being down converted via a channel master satellite dish signal strength meter. So the end result is that I can get a general graphing output of the sun, as the dish will track both horizontal and vertical (through Zenith)planes, following the sun, and the 56 volt panel is doing the same . I have it apart right now, re- working the physical location of the water pump PV panel(i.e 56 volt) for better balance in wind.


      1. I envision this “system” with a solar cooker in place of the ‘Dish” in economically poor areas, being portable enough for well pump use… it might cut down on the ravagement of the environment,(i.e trees,etc.) and provide some evening lighting arrangements for
        small medical clinics and homes. Of course, it might be scaled larger.
        The marine batteries can be used as ballast for this small array.

    2. Your forgot what a lot of people forgot… The perceived surface of the panel is smaller at an angle. You only calculated the decreased performance of the panel itself… Not the fact that it was “covering” a smaller surface seen frol the sun.

    3. At 45° the perceived surface is halfed, eve considering a superpanel that remain with the same performance whatever the angle… At 45 it’s already half the power you got.

  18. GREAT idea!!!
    A quick suggestion.. rather than use zip ties, use a few stainless steel hose clamps to tie the two pipes together, because the nylon will become brittle being exposed to the sun.

  19. Well said, but we all need to realize that adding Solar in their property is an asset that should boost the long term valuation of their property if / when they choose to sell. With the environment the way it is going we are unable to disregard any system that gives zero cost electricity at no cost to both the consumer and more importantly the earth!

  20. What was the specific model # & make of that programmable remote control unit you used to control the array’s rotation time through the channel master…and what exactly we’re your programming steps or code you used?

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.