On 28/02/2011 07:40, IVP wrote:
>
> If you were to make a tracker that was totally self-initialising, it woul=
d
> have on board a level, a compass, GPS, or something like that, so it
> could be just plonked down in any reasonable orientation and find its
> own bearings. From that point on it would, by learning, determine what
> view it has of the sky and whether to use solar tracking based on
> calculation or light intensity or both. For example, pause before a known
> obstacle, like a chimney, until it's time to fast-forward past the obstac=
le
> to catch the sun again as per the calculation

Or you could have the panel in four sections and use differential=20
voltages to find "brightest" part of sky. Or  a photo-transistor on each=20
edge. Suitable algorithm would then rotate and change elevation=20
periodically. No need for a compass / gps / level. It can "learn" its=20
position, starting azimuth, slope and set its software solar clock to=20
precalculate positions. In a few days to a week it would figure out the=20
season also. So it can calculate local solar time, season/date,=20
latitude, how level it is, direction of north but not the Longitude=20
without external UTC clock setting. GPS etc is more power consumption.

Is a weak spring and escapement the best way to "rotate back" to start=20
before dawn or use the motor?

--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.