On 4/20/2013 2:07 PM, Denny Esterline wrote:
> I don't know. I think trying to track the sun with PV panels based solely
> on time and date is a little suspect.
> As you've pointed out, you need a _very_ precise long term clock, but you
> also need angular position of the panel as well. How are you measuring
> that? What's it's repeatability and reliability and how well is it
> calibrated during install?
> I would think that some type of sensor that finds the brightest spot in t=
he
> sky would be a more consistent solution. Granted there would be additiona=
l
> controls (reset for morning, limit hunting during cloudy days, etc)
> Possibly some combination - time based algorithm with the clock disciplin=
ed
> by a sun detection. Use a common crystal (~5min/yr) and limit the
> discipline to something on the order of a second or two per day.
>
> -Denny
>
>
These have been sold and a few are being sold now, but the results are=20
poor. I have personally talked with two different owners of this kind of=20
system from Northern Mexico, and both told an identical story.

The system works pretty well on a clear-blue day, but almost no days are=20
like that. When a cloud drifts by, the system begins to wander all over=20
the sky, burning precious electricity in extraneous movement. Our system=20
KNOWS where the sun is, so even though it is momentarily blocked by a=20
cloud, it is always in the right position. The absolute minimum of=20
electricity is needed for motor movement. We are moving once a minute.

Angular position of the panel.. not sure what you mean.

Installation IS important. We mount a shadow target on the PV array and=20
watch the result as it moves. The azimuth (east-west) movement is pretty=20
accurate, its a huge load-bearing joint with very little whiplash. The=20
elevation movement is more error-prone as the positioning device (linear=20
actuator) is non-linear, especially at the ends of its travel. We had to=20
plot a few hundred points on its movement then create a table to=20
calibrate the actuator, but we only do this when we change actuator=20
brand or length. Each individual actuator does not need to be calibrated.

A lot of precision is needed when pouring the footings, and they are deeper=
..

--Bob A



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