Re: ATM Automatic Solar tracker

[Del Stanton <sdl20@pacificnet.net>]

Dear Valerie & Bob Goff

I have worked with several solar trackers.  Because of the energy available
it is high it is easy to make a sensor that will track the sun within an arc
second
or so.  The basic sensor is two solar cells side by side with a slit mask
several 
inches in front of them (the cells are in a box, the mask is a slit in the top
surface (solar facing) of the box).  The slit is above and parallel to the 
dividing line between the two cells.  The two solar cells are connected 
"series-opposing" so that when they are equally illuminated the current output
is zero.  As the Sun moves away from the sensor's axis the output current will 
become "positive" or "negative".  Connect this output to a simple 741 
operational amplifier wired in the "trans-impedance mode" and the resulting 
output voltage is the error signal for your tracking electronics.  (See the
Forrest 
Simms III book "Optotronics" at your local Radio Shack for the trans-impedance 
connection details.) Use separate sensors for the RA and Dec drives.  This 
device is the "fine-pointing" sensor and can acquire the Sun when it is several 
degrees off-axis.  It will work fine if used to make fine corrections to a
solar rate 
drive in RA.  In DEC it will make the small corrections required each day as
the 
Sun's track varies during the seasons.  This type of sensor is suitable for 
systems where you are forming a solar image and want accurate pointing.  Given 
sufficient bandwidth in your tracking system it should be capable of following 
some of the "motion" of the Sun's image due to "seeing".

For solar energy collectors that do not require such a degree of pointing 
accuracy the four-sided pyramid sensor is adequate and has the advantage of 
being able to acquire the Sun over wide error angles, approaching 140 degrees.  
The solar cells are mounted on the surfaces of a square-based pyramid, the 
included angle of the peak is perhaps 60 degrees.  When the axis of the 
pyramid is pointed at the Sun the current output of all the cells is equal.
Again 
the cells are connected "series-opposed" and the RA pair and the DEC pair 
will produce the required error signals for your tracking system.  The sensor 
pyramid should be mounted on a tall stalk, well clear of possible reflecting 
surfaces.  It wide angle of acceptance makes it vulnerable to spurious 
inputs, for example, sunlight reflected from a window pane that might be 
hundreds of feet away.  The "on-axis" sensitivity of the pyramid sensor can
be increased by having a square "shadow plate" mounted on axis, several 
inches above the solar cells.  Without this plate the sensor has a "cosine"
response, with relatively small changes in output while pointed close to the
Sun.  The "shadow plate" can increase the error signal when the Sun is 
near the sensor's axis by an order of magnitude or more.

I hope this helps.

Del Stanton  (My sig follows the quoted text.)

Del Stanton  - Burbank, California, USA (Near Los Angeles)

"sdl20@pacificnet.net"  (Lower case "SDL" followed by numeric "20")