How to make a solar panel follow the sun across the sky

[Shawnzyoo]

Does anyone have any suggestions of how to make a device follow the sun acrossed the sky?
It will be a solar panel mounted on a satellite dish acting as a large sunlight collector, but it needs to be aligned with the sun for optimal power.
Thanks for any input.

 

[BobG]

How about a simple digital sun sensor with no moving parts. You need a box with a slit cut into the top for the Sun to shine through. The bottom has two layers - a screen and your sensor layer.

The sensor layer just has several strips of light detecting material - eight, sixteen, or however many bits you need. Each hexadecimal value will indicate how far off you are from being lined up on the Sun. You might only need to account for 4 values - lined up, move left, move right, no sun.

The screen just has a Grey Code (easiest for cutting if you actually want to determine how far out of alignment you are) or binary code cut into it. It allows the Sun to shine through to different light detectors depending on the Sun angle. If you only needed 4 values, lined up might be both detectors illuminated (the openings in the screen overlap just enough that sunlight shining directly down through the slit shines through both openings), move left one detector illuminated, move right would have the other detector illuminated.

This only gives you single axis control, so you'd need another detector of the same design set up perpendicular to the first to control the second axis.

 

[russ_watters]

How about just gearing it to turn at the rate the sun moves across the sky? No actual tracking logic required.

 

[Pengwuino]

Does anyone have any suggestions of how to make a device follow the sun acrossed the sky?
It will be a solar panel mounted on a satellite dish acting as a large sunlight collector, but it needs to be aligned with the sun for optimal power.
Thanks for any input.

Follow the suggestions. I've already seen multiple examples of people building their own solar trackers for their solar panels! And they say its very simple even for those with little electrical experience (I mean one site literally ran off a list of radio-shack part ID #'s to get)

 

[BobG]

How about just gearing it to turn at the rate the sun moves across the sky? No actual tracking logic required.

Actually, there would be, since the Earth is tilted. You'd at least need some software that knew where the Sun should be for that day and time.

 

[pallidin]

It will be a solar panel mounted on a satellite dish acting as a large sunlight collector .

Be aware that most solar cells are not designed to have substantially additional full sun light. Mirrors have been used to provide additional sunlight to solar panels, but this is not recommended unless you know EXACTLY what you are doing.

The reason is substantially increased UV and IR hitting the panel(s) using mirrors. One can easily fry the solar panel if mirrors are used at high noon.

 

[Pengwuino]

Actually, there would be, since the Earth is tilted. You'd at least need some software that knew where the Sun should be for that day and time.

Yah, it also needs to be rotated based on the time of the year.

 

[russ_watters]

Actually, there would be, since the Earth is tilted. You'd at least need some software that knew where the Sun should be for that day and time.

Only if you're tracking in two dimensions. If you're tracking in one (and most do track in one), the sun is at precisely the same place every day (otherwise a sundial wouldn't work).

 

[hitssquad]

There are passive solar trackers. These might be the easiest to make.
http://www.e-marine-inc.com/products/mounts/tracker.html

--
Track Racks are engineered for unmatched reliability and do not rely on drive motors, gears, clutches, high-tolerance pistons, electronics, batteries, or sun sensors. The Track Rack derives its tracking motion with a simple, elegant design that uses the sun's heat to move liquid through a sealed system integrated into its structure. As liquid moves from one side of the rack to the other, gravity causes it to rotate and follow the sun.
--

 

[Kenneth Mann]

Your situation is similar to that of a spacecraft that I worked on several years ago. It had a flat solar array which was mounted on a shaft so that, as the shaft rotated its array would always face the sun. This was done by mounting a pair of single solar voltaic cells on the shaft itself slightly displaced from that line along the shaft that faced the sun. They were on both sides of the 'sun normal' line, so that if the shaft rotated one way, one of the cells faced more directly toward the sun and the other further away. If it rotated the other way, just the opposite. I don,t recall the circuit used in the spacecraft , but there are several ways to do it. As an example, the cell outputs can go to a difference amplifier. Hope this is of use.

KM

 

[Danger]

This isn't in one of my areas, but I'll chip in anyhow. Bob's solution is elegant, and no doubt very effective, but I've had a bit too much beer to follow it exactly. Kenneth's is much closer to how I would approach it myself. Three or four photovoltaic cells mounted equidistantly around the perimeter of the dish, and angled slightly outward from centre (to give a wider effective diameter), can be fed into summing junctions (comparitors) and command the servos to move the dish until all are producing equal voltage.

 

[Francis M]

I think there was an article in QST (Ham radio) magazine on how to build a sun tracker as well. It went the photovoltaic cell route, and tracked on one axis only. Unfortunately I can't remember which monthly issue it was in, to narrow down the search for you, sorry, but I hope this helps in at least giving another area to search for information. Good luck and I hope the application works out well!

 

[brewnog]

I encountered this problem in a control systems lab at university. The solution was pretty much that which Kenneth suggested.

 

[DuncanM]

The same stores that sell photovoltaic arrays usually also sell trackers.

Alternately, a search should turn up many sources, for example, (passive and active trackers):

http://www.realgoods.com/renew/shop/list.cfm/dp/1000/sd/1007

According to a few sources I have read, tracking makes minimal difference in the continental US. People are advised to simply re-point the array about every three months.

If you really want to do it yourself, here are a couple ideas:

i) write a small program with data input for the Sun's position in the sky for each day. The tracker would then be controlled by the program without real feedback (closed loop).

ii) simply use a timer to rotate the array at a fixed rate throughout the day. For example, you might know sunrise and sunset of a particular day, so set the timer accordingly. Once a month (say) adjust the timer to account for the lengthening or shortening of the days. Crude, but cheap and simple.

Regards,


Duncan

 

[russ_watters]

The reason I'd go completely passive (no sensors at all) is that clouds would interfere with such a device.

 

[Kenneth Mann]

The reason I'd go completely passive (no sensors at all) is that clouds would interfere with such a device.

I must disagree. First, a precisely geared (clock-like) drive is at least as complex in driving an array forward as is a simple sun-tracking drive. Second, comes the problem in driving the array back to its starting point during the night. Unless the array is to have expensive slip-rings, it must be driven back in reverse. The problem here occurs in the Summertime when the return must be faster than the tracking movement during the day. In this case, either it will be necessary to have a clutched, two-speed gearing arrangement, or there must be a way to run the motor at two speeds. Either way, this becomes a more complicated arrangement than the tracking drive. Finally, the tracking sensors work quite well on most cloudy days. If it is too cloudy for the tracking sensors, it's also too cloudy for the solar power array itself.

KM

P.S. Is the plan really to concentrate solar radiation on the solar cell array? This sounds a bit risky to me.