Sun tracker piloted by GPS

Hello,
we try to design and build a sun tracker carrying a plate photvoltaîque, piloted by GPS,

GPS must locate the position of the sun and gives an order to move towards this position throughout the day,
but we really do not know where to start

all your suggestions and ideas will be welcome,

thank you in advance for your helps,
cordially

so your solar panel is mobile and tracking the sun ? wow that's ambitious, if perhaps unwieldy maybe?
If you have GPS coordinates, you can calculate the sun position.
you can find a lot by googling.
eg **broken link removed**

xaccto said:

so your solar panel is mobile and tracking the sun ? wow that's ambitious, if perhaps unwieldy maybe?

eg **broken link removed**

Click to expand...

HELLO,
yes exactly like this!
but we have a difficult.. the program that calculate the position in real time ??? and how can we transform these informations to control signal ?

thank you!

Actually the sun is always in the same place its the earth and yourself that move around. I feel that if this kit is placed on a vehicle then the speed of turning corners and weaving down roads will always leave the GPS coordinates behind, if the kit is on an ocean going ship then its speed is so slow that GPS would be unnecessary other then to initially get the PV array locked on the sun. What you really need are look up tables for the suns position at certain times and latitude, then correction for your own time and latitude is easy. A compass is a good tool, you can get a GPS location of where you are :- XX degreees N, YYY degrees W, but this will not tell you which way west or north is, until you move and some thing changes.
Frank

If you search via google, you'll be able to find source code to plot the positions of constellations and the sun and the moon. Pretty much all programs that do this are based on the math that were published in one book (I forrget which one) a long time ago.
Anyway, I once tried as an experiment to take the source code (which was intended for compiling into an x86 executable for a DOS package, and it was fairly easy to modify it to run on a 68k CPU with not much RAM and ROM. Unfortunately I don't have the code any more : (
It would be easier in your case, since you only need the algorithms for the sun portion. It would run pretty fast (not that you need much speed) on any processor running at even 8MHz or so.

Alternatively, you could have some array (like a fish-eye) of LDRs, and then you instantly know where approximately the sun is. I guess even if it is not directly facing the sun but at an angle, the solar cell should still work (I'm no expert on that).