Post by thread:Sun Tracking

I have some friends' graduation report (A sun tracker, used with a solar cell, ... only a hardcopy)
major observations:
* Continuous tracking not necessary (very small gain in power), they put the 16c71 to sleep and woke it up every 30minutes (using a 555, AFAIR)
* They used a PIC (an improvemnt over an older PC-Based project)

Here's a few ASCII Drawings

+--------------+ |*1ssssssssss2*| | ssssssssss | | ssssssssss | | ssssssssss | |*3ssssssssss4*| +--------------+ s=solar cell *=sensor sensor amp +---+ *-------|>----| 1 | +------+ *-------|>----| 6 |----+Motor |----MOTOR#1 *-------|>----| c | | | *-------|>----| 7 |----+Driver|----MOTOR#2 | 1 | +------+ +---+They used a simple algorithm:
ON WAKE UP:
turned power on for everything
if Sensor1+Sensor2>Sensor3+Sensor4 then move up else move down
if Sensor1+Sensor3>Sensor2+Sensor4 then move left else move right
repeat until Differences very small (optimal position)
turned power off from most of the board
SLEEP for 30 minutes

At 02:12 25/12/98 , you wrote:
>Hi All,
>
>I am trying to design a project to track the sun.� Has anyone ever tried to
>do this.� Any help would be greatful.
>
>Brian
>

_____________________________________________________

Osama ALASSIRY osama@qatar.net.qa osama@alassiry.com

http://www.alassiry.com

Brian, Yes I have done a sun tracker for PV arrays What do you have in mind. Will there be a PIC controller? Basically one can make a quadrature encoder (a plus sign with some depth for shading) using low sensitivity phototransisters of CDS cells with shading over them. There is also a need to have the home location of the tracker - like overnight have it return for catching morning sun. The method I used was to use two steppers motors and a PC running on a Dig Analog I/O card with Borland C. It would be possible to push this into a PIC like a 16C76 type. Keep in touch... Eric

Merry Christmas all! I have an interest in this as well ... for a solar panel tracker. James Cameron is on this list and has interest and ideas on this subject. I would like to drive the panel mount with a 12vdc trailer jack as they are robust, available and reasonably priced. This means an 'H' drive circuit and a 12c508 with a couple of photo transistors would make the decisions. I am told the best setup is to have both sensors shaded when the panel is aimed and maybe a startup/low light sensor would be a hot idea. If this is anywhere close to what you want to do, let me know, I'll work along with you. Wil Reeder spam_OUTteachtechTakeThisOuTbc.sympatico.ca Vancouver,Canada ---------- > From: Brian Gracia <.....brian_graciaKILLspam@spam@HOTMAIL.COM> > To: PICLISTKILLspamMITVMA.MIT.EDU > Subject: Sun Tracking > Date: Thursday, December 24, 1998 6:12 PM > > Hi All, > > I am trying to design a project to track the sun. Has anyone ever tried to > do this. Any help would be greatful. > > Brian

At 20:12 1998-12-24 -0600, you wrote: >Hi All, > >I am trying to design a project to track the sun. Has anyone ever tried to >do this. Any help would be greatful. > >Brian Hi Brian and all I am thinking to build a solar heat collector the coming year, dual-axis rotating against sun. As the position of the sun is *very* predictable... I have decided to use just time for positioning, to avoid any problems with contamination of sensor, like fies shit, autumn leaves, partial snow cover, or degradation of the sensor (the collectors and environmental casing is designed for 30 yrs, so i don«t want a lousy sensor to shit it up) The position feedback sensors are probably much cheaper than a suitable optical sensor arrangement, and also make more use, like make it stand correct at sunrise or after a half day dark clouds, without any additional logic. For the turn-around it is just to follow the clock: n revolutions of motor per minute, and an index at 12 a.m to calibrate zero twice a day. For the height using a sine with frequency of the day, amplitude modulated with a constant + sine with frequency of the year. Time is fetched from logging system. Update every hour might give unmeasurably less power tna continuous, but i think i will settle for every five minutes or so, keeping a more constant temperature in the motors, making them always slightly warmer than outside, thus keeping them dry iven if small leak in casing, as there always will be in 30 yrs time... Good luck /Morgan Morgan Olsson ph +46(0)414 70741 MORGANS REGLERTEKNIK fax +46(0)414 70331 H€LLEKS (in A-Z letters: "HALLEKAS") SE-277 35 KIVIK, SWEDEN .....mrtKILLspam.....iname.com ___________________________________________________________

I'm also interest in this idea. The prospect of fixing the sensors so they are both DARK is one I haven't thought of. Does anyone use somethind besides CDS cells, say phototransistors? I would think a small phototransistor would be easy to aim in a drilled hole. -- Lawrence Lile => Median Filter Source Code => AutoCad blocks for electrical drafting => Brownout tester plans => Amateurish pictures of my family at: http://home1.gte.net/llile/index.htm

Morgan Olsson sez > As the position of the sun is *very* predictable... > > I have decided to use just time for positioning, to avoid any > problems with contamination of sensor, That's a good idea. I would suspect that a very accurate timekeeper, not just a crystal, would be needed. Either a watch crystal or one of those timekeeping devices from Dallas. If you had more that a few minute's error per month you would accumulate large position errors. How far from the equator are all you solar tracker makers? At higher lattitudes the winter to summer variation in dawn times is signifigant. Perhaps, though, you are only interested in heat in the winter. In that case calibrating it for winter Dawn times is enough. -- Lawrence Lile => Median Filter Source Code => AutoCad blocks for electrical drafting => Brownout tester plans => Amateurish pictures of my family at: http://home1.gte.net/llile/index.htm

> > As the position of the sun is *very* predictable... > That's a good idea. I would suspect that a very accurate timekeeper, > not just a crystal, would be needed. Either a watch crystal or one > of those timekeeping devices from Dallas. If you had more that a few > minute's error per month you would accumulate large position errors. You're possibly overlooking the obvious. Why try and maintain a long term clock for a short term requirement ? A single light sensor will let you detect when the world changes from dark to light and back to dark again. You'd only need half a dozen bytes to make a long counter and measure the length of the day. Since today is going to be very close to the length of yesterday and you can see the sun rise, you have all the information you need to predict where the sun is. You could even use a median filter. :-) Steve. ====================================================== Steve Baldwin Electronic Product Design TLA Microsystems Ltd Microcontroller Specialists PO Box 15-680, New Lynn http://www.tla.co.nz Auckland, New Zealand ph +64 9 820-2221 email: EraseMEstevebspam_OUTTakeThisOuTtla.co.nz fax +64 9 820-1929 ======================================================

WIL REEDER wrote: > I have an interest in this as well ... for a solar panel tracker. > James Cameron is on this list and has interest and ideas on this > subject. True, but I haven't actually got a current need for it yet; perhaps in another four or five months when our road is in place. On the other hand, with as much solar panel area as I will need (for 7kw/h per day) pointing them is not trivial. I may just buy some more. I like the idea of sleeping for thirty minutes between checks to cut down the power consumption of the controller. -- James Cameron (cameronspam_OUTstl.dec.com) OpenVMS, Linux, Firewalls, Software Engineering, CGI, HTTP, X, C, FORTH, COBOL, BASIC, DCL, csh, bash, ksh, sh, Electronics, Microcontrollers, Disability Engineering, Netrek, Bicycles, Pedant, Farming, Home Control, Remote Area Power, Greek Scholar, Tenor Vocalist, Church Sound, Husband. "Specialisation is for insects." -- Robert Heinlein.

James Cameron wrote: > > WIL REEDER wrote: > > I have an interest in this as well ... for a solar panel tracker. > > James Cameron is on this list and has interest and ideas on this > > subject. > > True, but I haven't actually got a current need for it yet; perhaps in > another four or five months when our road is in place. > > On the other hand, with as much solar panel area as I will need (for > 7kw/h per day) pointing them is not trivial. I may just buy some more. > > I like the idea of sleeping for thirty minutes between checks to cut > down the power consumption of the controller. You don't need a controller at all. Use two photosensitive devices seperated by a sheet of opaque material. Use their outputs to drive two comparators. The comparators are used to keep the panel in a 'window' where the motors will not be driven. You can drive the H bridge directly from the comparators. If you use FET type opamps, there is no way you can do better (current consumption wise) with a controller. -- Friendly Regards /"\ \ / Tjaart van der Walt X ASCII RIBBON CAMPAIGN @spam@tjaartKILLspamwasp.co.za / \ AGAINST HTML MAIL |--------------------------------------------------| | WASP International | |R&D Engineer : GSM peripheral services development| |--------------------------------------------------| |SMS KILLspamtjaartKILLspamsms.wasp.co.za (160 chars max)| | http://www.wasp.co.za/~tjaart/index.html | |Voice: +27-(0)11-622-8686 Fax: +27-(0)11-622-8973| | WGS-84 : 26¡10.52'S 28¡06.19'E | |--------------------------------------------------|

part 0 1038 bytes ive seen a sun tracker for a solar panel in a book i have it was made with fischer technic a german plastic construction toy frequently used toy at universitys for prototyping the program was made in basic it was controlled by a schneider, sinclair zxspectrum, commodore 6*/8* acorn or msx way back in 1985 later fischer technik developed an ibm interface the have made more interesting things like robotic arms plotter/scanner/digitiser etc. i think if you do a web search on fischer technic you wil find pages about this subject maybe there is even a news group fischer technic the booklet i have is dutch robot besturing met uw homecomputer (translated robot control with your home computer) publisher Elektuur ISBN90-70160 (first press date 1986) hope this helped a bit attached image suntrack.gif greeting's luberth http://home.wxs.nl/~luberth home build qbasic plotter engraver vinyl-cutter pcb-driller http://home.wxs.nl/~luberth/plotter/plotter.htm Attachment converted: wonderland:SUNTRACK.GIF (GIFf/JVWR) (00023476)

On Tue, 29 Dec 1998, James Cameron wrote: > WIL REEDER wrote: > > I have an interest in this as well ... for a solar panel tracker. > > James Cameron is on this list and has interest and ideas on this > > subject. > > True, but I haven't actually got a current need for it yet; perhaps in > another four or five months when our road is in place. > > On the other hand, with as much solar panel area as I will need (for > 7kw/h per day) pointing them is not trivial. I may just buy some more. > > I like the idea of sleeping for thirty minutes between checks to cut > down the power consumption of the controller. You mean, sleeping a PIC which draws some 10 mA with all sense amps etc, on a 7kW installation ? ;) Are you far away from civilized places ? I.e. in the wilderness ? If so, you might want to know that there is a PICless panel pointing system that relies on the differential pressure in two freon-filled chambers exposed to the sun. The chambers are shaded equally when pointing into the sun and deliver pressure for action cylinders to move the panel(s). This also works in az/el (4 vessels, 4 actuators on a swashplate or gymbals). By choosing the pressures right when there is not enough sun/heat, the panel 'parks' face down and rolls into the expected next sunrise by gravity (the base plate is inclined towards that side). The face down position also reduces rain etc mudding the mirrors (actually they get washed because slanted). There is no need for maintenance, someone simply walks around with a freon bottle and fills the two (four) circuits for each panel up from time to time. By differential filling small alignment errors can be fixed. I've read about this system a long time ago, in a solar/environment magazine that predates the 'greens' ;) hope this helps, although OT, Peter