# Re: simplified heliostat program

Posted by nicksanspam on July 24, 2007, 8:58 am

Can a stamp run BASICA? Most simple BASICs seem to lack floating point,
sines, cosines, logs, and exponentials.

Nick

Posted by Morris Dovey on July 26, 2007, 4:18 am

David Williams wrote:
| -> I think you may be making the problem more difficult than really
| -> necessary. Only two axis motion is necessary and the range of
motion
| -> (for most locations) will be 45 degrees for elevation and 90
degrees
| -> for azimuth, with a time spread of at least six hours.
|
| Hmmm... My heliostat routinely moved through almost 90 degrees in
| elevation and 360 in azimuth. It was in the northern hemisphere, and
| required to reflect light almost northward at an angle of elevation
| that was intermediate between those of the sun in summer and
| winter. At various times and seasons, pretty well any azimuth was
| needed, and also elevation. This wasn't a problem, of course, but
| we shouldn't assume that ranges of motion will be smaller than they
| may have to be in practice.

Hmm. I'm missing something essential here. I'm thinking that when the
sun is on the horizon, elevation should be greater than zero (unless
the target is at ground level, of course) and at noon, elevation
shouldn't be more than half of (90 - latitude + 23.6). Azimuth change
from horizon to horizon is 180 degrees. What am I neglecting?

| -> There's no obvious need for mirrors to move "in step" - only that
each
| -> mirror satisfies the system-specific requirement for being
| -> "on-target".
|
| Quite so. Essentially, they can be independent little heliostats,
| each moving on its own.
|
| -> One of the more interesting thoughts that popped up was that it
might
| -> work well to use the nighttime "off-hours" to
pre-calculate/schedule
| -> all mirrors' movements for the following day...
|
| Let's not make more unjustified assumptions! Suppose the thing is in
| the arctic, or close to it. There may be little or no "nighttime off
| hours". But no matter. The earth turns quite slowly. Doing
| calculations in real time fast enough to keep pace with it is not
| much of a challenge.

I wouldn't think so - but then I probably wouldn't try to cram a
general solution (which included not only the arctic, but also the
entire 'other' hemisphere) into a minimal-cost micro-controller.

--
Morris Dovey
DeSoto Solar
DeSoto, Iowa USA
http://www.iedu.com/DeSoto/

Posted by Morris Dovey on July 26, 2007, 6:22 pm
David Williams wrote:
| -> Hmm. I'm missing something essential here. I'm thinking that
| when the
| -> sun is on the horizon, elevation should be greater than zero
| (unless
| -> the target is at ground level, of course) and at noon, elevation
| -> shouldn't be more than half of (90 - latitude + 23.6). Azimuth
| change
| -> from horizon to horizon is 180 degrees. What am I neglecting?
|
| We're talking about the aim-direction for a mirror, right? So
| suppose the "target" azimuth is diametrically opposite the azimuth
| of the sun, and the elevations of the sun and target are equal,
| then for the mirror to reflect sunlight at the target, it has to
| point straight upward, at an elevation of 90 degrees! This
| situation used to happen for my heliostat at about noon on a couple
| of dates per year.
|
| On a date just on the "winter" side of either of these two dates,
| the mirror would point at a very high elevation, but just less than
| 90 degrees, and at an azimuth roughly south as the sun passed to the
| south. A few days later (or earlier), just on the "summer" side of
| either of the two "vertical" dates, the mirror would also point
| almost vertically upward, but at an azimuth that was roughly
| *north*!
|
| On the dates when the mirror went most closely to vertical (it was
| never absolutely exact), its azimuth would rotate through 180
| degrees, from roughly east to roughly west, in a period of just a
| minute or so. Watching it was kinda fun. Of course, since the
| mirror was pointing almost straight up, the direction in which
| light was reflected was affected very little by this abrupt
| rotation in azimuth. Just on the "winter" side of vertical, the
| rotation would be clockwise. On the summer side, it was
| *anticlockwise*. In fact, during the "summer" part of the year, the
| azimuth drive would move anticlockwise for part of every day, but
| clockwise in the early morning and evening.
|
| The motions were a whole lot more complicated than I anticipated
| when I made the machine. Fortunately, the software figured it all
| the thing from moving as it should.

Hmm. I think I need to take a bit of time to digest what you wrote. I
think we're somehow using the same words to describe different
solution strategies...

| -> I wouldn't think so - but then I probably wouldn't try to cram a
| -> general solution (which included not only the arctic, but also
| the
| -> entire 'other' hemisphere) into a minimal-cost micro-controller.
|
| Why not?

Because I don't think it's a great idea to make people pay for what
they aren't going to use - in my mind, that's not cost minimization. I
am, however, perfectly willing to let arctic (and antarctic) dwellers
pay slightly more for a product for which (I suspect) there will be
few purchasers. Obviously, YMMV.

--
Morris Dovey
DeSoto Solar
DeSoto, Iowa USA
http://www.iedu.com/DeSoto/