Elsewhere on the internet...

The League of Reason has some social media accounts! You can find us on Facebook or on Twitter for some interesting links and things.

Design ideas for a Sun finding sensor?

Post new topic Reply to topic  Page 1 of 1
 [ 12 posts ] 
Design ideas for a Sun finding sensor?
Author Message
SparkyUser avatarPosts: 148Joined: Wed Mar 18, 2009 2:17 amLocation: New Zealand Gender: Male

Post Design ideas for a Sun finding sensor?

I'm just starting a project at University in which I am to design an electrical/electromechanical system that can detect the Sun's position in the sky and output signals to a control system that will rotate a solar panel array to the correct position.

The requirements are:
1) Sensors to detect the Sun's position
2) A search (so that the sun can be found if the system loses power for a period of time, etc) and tracking system of some kind
3) A digital or analogue output that indicates the way the solar panel needs to be moved to maximize power output

At present, my best idea is just to take the output power from the solar panel and to have a program run on a microprocessor that rotates the solar panel until it finds the maximum power. It will poll the program so continual adjustments will be made to ensure that the solar panel is always getting the maximum power possible.

I am posting here to see if anyone else can think of a better system. Keep in mind that the system can be electromechanical. One suggestion the Dr setting the project made was that the declination angle (related to the Earth's magnetic field) could be used in some way.
Believe those who are seeking the truth. Doubt those who find it.
~Andre Gide
Fri Jul 30, 2010 4:10 pm
NelsonUser avatarPosts: 183Joined: Tue Mar 10, 2009 6:52 pmLocation: British Columbia Gender: Male

Post Re: Design ideas for a Sun finding sensor?

Do you have a CCD or some sort of device that can detect flux at different positions on the chip in real time? You could build a primitive eye type device. Take some CCD. Place a opaque cover some height above it with a small hole in the middle (maybe the size of a pixel or so). The distance between the cover and the chip itself should be determined depending on the range of angles you need to measure. If the device is level with the ground, which pixels get illuminated should tell you the position of the sun on the sky.

Honestly though, I think your solution is better since I'm guessing these solar panels already give some energy output value and have a mechanical means to be adjusted. It should be fairly straight forward to write an algorithm to make adjustments toward the maximum output.

You could also just do this with a clock and calendar. It isn't hard to calculate the position of the sun in the sky with your latitude, date, and time of day. You could have a schedule where each day the panel tracks across the sky in a predetermined arc. This could all be calculated beforehand rather than on the fly.
I have approximate answers, and possible beliefs, and different degrees of certainty about different things, but I’m not absolutely sure of anything, and there are many things I don’t know anything about. - Richard Feynman -
Fri Jul 30, 2010 5:53 pm
NetheralianUser avatarPosts: 357Joined: Wed Feb 25, 2009 3:13 pmLocation: Underwater world Gender: Cake

Post Re: Design ideas for a Sun finding sensor?

Edit: Shit - you might want to ignore this - For some reason I thought you were building a satellite... Anyway, the solution is still kind of valid. Although the suns position is easy to calculate based on time and date from a ground position

What is your pointing accuracy requirement? I would imagine based on the purpose of the device that +/- 5 degrees or so would be sufficient (less than 1% loss on your solar array). Also, what are you volume/mass requirements?

There are plenty of Course Earth/Sun sensors available - but buying one wouldn't help you pass with your project.

Thermal based sensors could be pretty simple - especially if you use materials that have a sufficient electrical resistance sensitivity to thermal changes - then you can use the material itself as a thermistor (or just use thermistors on some aluminium plates). Then you just need to make a half sphere and pick the hotspot (although you have to thermally decouple the facets of the sphere to minimise cross talk between sensors). You can also use this to find the earth. Or you could combine this with the idea Nelson gave - just replace the CCD with thermal sensors. The benefit of this will be cost - its really cheap takes virtually no power and little processing is required. It can also be light - although it will probably be fiddly.

IR sensors could also be used - make a little porcupine of cylinders with an IR sensor at the base.

You solution is of course fine, but you could waste a lot of power hunting for the signal. Also, I doubt your attitude control guys would be overly happy with you using the solar array in this way. Come to think of it, the power guys may not either - even though it is ultimately helping them. You probably want something that already knows where the sun is - especially on orbit acquisition.

Keep in mind that if you use the earth's magnetic field, as your orbit decays (unless you have the propellant to keep your orbit fixed, this is a likely outcome), the direction of the magnetic field lines will also change.
Last edited by Netheralian on Fri Jul 30, 2010 6:43 pm, edited 1 time in total.
Fri Jul 30, 2010 6:41 pm
simonecuttlefishUser avatarPosts: 114Joined: Tue Jul 06, 2010 12:24 am Gender: Pinecone

Post Re: Design ideas for a Sun finding sensor?

OK - some idiot suggestions from the utterly clueless, but it's Friday night and there is a bottle of rum AND a bottle of bourbon on the table so lets rip into it.

1) An array of photo-resistors, or photo-transistors, or cheap solar mini panel thingers, with a conical shield surrounding them to make them more directionally sensitive (not too much), possibly with a semi opaque shielding over them to soften the effect of localised reflections from nearby objetcs, that can "add up" their values based on incoming light levels? Actually, thinking about it (clearly not too hard atm :) ) swapping the detectors to something that would more accurately detect incoming solar radiation, rather than reflected/visible light might be smarter.

If these were arranged in a static/fixed physical array, then the varying incoming values from each receptor should be able to allow a calculation of which direction to "point" the servo motors at. This might help the collector stay at a relatively well focussed position, while something else does the "searching" - moving the collector around searching for relatively "bright" spots might lower the overall efficiency, as it would have to point at dull/dead spots before it could "decide" that was relatively less efficient, and even worse, spin off in the wrong direction trying to see if it got brighter on the "other side". I assume you get brownie points for keeping efficiency high, so making the collector the directional host for the "brightness" indicators might be 'inefficiencyerising' ? :)

2) Get a bunch of those "dippy bird thingers" you used to see in Chinese restaurants, and arrange them "bum in the middle, circular pattern", with there hollow glass necks sticking out of a shaded (and well ventilated/heat insulated and preferably cooled) box. The glass of water they dip their beaks in would be exposed to the sun, and probably best dyed black to convert solar radiation into heat. Then, you could have mercury tilt switch thingers at the "hip" of the birds, to detect how many were "sitting up" and how many were "drinking" to easily switch on the solar collectors drive servos. I have no idea how they work so you might need to put them "bum out" instead, but whatever, you might get "HEY! I remember those in the Chinese restaurant when I was a kid!" points from your tutor/professor. Or slapped in the face.

3) More idiot suggestions coming as soon as I get come more cola.

PS - If you want me to stop just say :) But I'm thinking broad beans in a dark box with holes in it, or phototrophic bacteria or crustaceans and a scanner, and/or cockroach density (they hide from light) detectors!

Seriously though - good luck!

I have no requirement for the supernatural or magic to explain anything, finding purpose in life other than trying to enjoy it, fantasies to explain wonder, fear that reason might trivialise me, or demand of wonders greater than those that are evident.
Fri Jul 30, 2010 6:41 pm
simonecuttlefishUser avatarPosts: 114Joined: Tue Jul 06, 2010 12:24 am Gender: Pinecone

Post Re: Design ideas for a Sun finding sensor?

OK - How about this. Dippy bird position and cockroach density is tricky to set up in 3D without using an odd "pointy bit up" conical structure to mount then on/in, so that was a bit dumb, because you don't just need round and round position, but up and down position to keep your collector pointed in an optimal direction. SO ....

Arbitrary number time! YAY! I'll say ... 5

Decide if you are wanting to detect visible light levels, or (preferably I think) whatever solar radiation type best penetrates clouds (in case it's cloudy DUHHH!) and set them up in this structure.

1) Build a 3D thinger that has 5 outpointing rays from a substance that is SLIGHTLY opaque to the type of light you are wanting to detect. Like from celluloid sheets that intersect in the middle, but are still vertically high, like they are balanced on their "edge", not on their flat surface. Do NOT make them very tall.

2) Put 5 detectors of choice near their central apex. You then need to put a "second story" of this structure on top, but with a flat alligned floor of the same semi opaque material "on top", like the bit you stand on in the 2nd story of a building.

3) Put another round of detectors on top of the "floor" of the 2nd story. The 5 "walls" of this story are (a littler) higher.

Now you can detect the brightness levels in a "round and round" fashion from the relative light levels of the top story readings, and use the collective "opaqueness" filters of the walls effect on the lower detectors, when compared to their vertically aligned "2nd story" versions, to detect "up and downinessness" levels of the sun.

I didn't do so well at school as you may have guessed, so on that note I'll stop.

P.S. This smart arse http://en.wikipedia.org/wiki/Eratosthenes could probably have done it with a lighthouse and heaps of string, but no one takes geometry seriously these days.
I have no requirement for the supernatural or magic to explain anything, finding purpose in life other than trying to enjoy it, fantasies to explain wonder, fear that reason might trivialise me, or demand of wonders greater than those that are evident.
Fri Jul 30, 2010 7:15 pm
FaithlessThinkerUser avatarPosts: 618Joined: Wed Apr 07, 2010 10:41 am Gender: Cake

Post Re: Design ideas for a Sun finding sensor?

Use a camera, perhaps a wide angle one aligned east to west. The sun's place in the sky is quite predictable, as it moves across the sky in a traceable path at an uniform speed. So I dont see why you should search the sky for the sun especially if you mean to scan the whole sky looking for it.

Unless you live in a place where the sun is invisible (covered by clouds, for example), this system should work. Even during the occasional cover-up by clouds, a simple computer program can trace the path of the sun (using data already collected using the camera) and determine the current position of the sun. This can also be done for shot power outages, as computers can easily calculate how much time had past since the power was lost.
It was my honest attempt to find a more pure form of God that made me realise that there was none.”
— Master_Ghost_Knight
Previously known as anon1986sing. You can call me FT. :)
Fri Jul 30, 2010 7:36 pm
WWW
borrofburiModeratorPosts: 3527Joined: Sun Jul 26, 2009 5:27 pm Gender: Male

Post Re: Design ideas for a Sun finding sensor?

I'd just make a "wand" with a sphere at the end, that is decorated in sensors around the sphere, and then based on which sensors are reporting the most sunlight, you can get sun's direction from that and move all the other panels based on this one calibrating sensor device.
Fri Jul 30, 2010 8:28 pm
Master_Ghost_KnightContributorUser avatarPosts: 2750Joined: Sat Feb 21, 2009 11:57 pmLocation: Netherlands Gender: Male

Post Re: Design ideas for a Sun finding sensor?

A passive control of calculating the position of the sun in the sky isn't a very good one because that isn't control, the system will eventualy drift for several different reasons. The objective is to know where it is, not to think where it is.

Also I do not know in which setting is the control supoused to be aplyed, is it for something as trivial a a ground fixed solar panel? Or are they solar panels of something movable, like a satelite, or mounted on a truck. What degrees of freedom does it have?

Now trial samplig (the one you propoused) is bad in any of the cases simply for the reason that you will have to constantly adjust back and forward the solar panel and that consumes allot to operate it (in some situation more then what you would get if you justlet it stand still).

But for the sake of exemplification let's assume that it is a ground fixed system with yaw and pitch capabilities. With that it tells you that the sun can only be observed over the horizon and also that the sun can only be traced on a specific band in the sky (so it saves allot on sensors), and it can only move at a certain rate.
And what you can do is to have 3 sensor plates placed perpendicularly (after calibrated) and with that depending of the light intensity tha eachone recieves you can calculate (or estimate) the relative angles of the plates (or sensores) in respect to the sun. (or some sort of a eye as Nelson sugested to indicate the direction, plates would requier to be small to minimize disturbance caused for instance by clouds).
Now if you are thinking on runing a program on a micro processor you have gone to far, a small, cheap, litle energy consuming, digital system can do the work of comparing the inputs of the sensor and giving an input (that when aplyed the apropriate gain) can give a signal to adjust the panels (the system itself can even be powered by the solar input of the sensor itself if you make it the right size). If you find out that making a small digital system or even a smart electric circut isn't for you, you can have a microcontroler that switches on from time to time (to make small corrections) while it can even completly shut down if the sensor input goes bellow a certain treshold only to go back online when the sensor input comes back up (when the sun rises again next morning).
"I have an irrefutable argument for the existence of...." NO, STOP! You are already wrong!
Fri Jul 30, 2010 10:15 pm
SparkyUser avatarPosts: 148Joined: Wed Mar 18, 2009 2:17 amLocation: New Zealand Gender: Male

Post Re: Design ideas for a Sun finding sensor?

I like the ideas of an "eye" or wand that have been suggested. They will allow the Sun's position to be found more quickly than the method I proposed.

I also noticed people mentioning CCDs. On that topic, we have been told that the sensors will probably need to have a very large dynamic range due to the variability in the intensity of the Sun's light received. Will CCDs work well for this? Does anyone have any other particular sensor suggestions that would work well?
Believe those who are seeking the truth. Doubt those who find it.
~Andre Gide
Sat Jul 31, 2010 12:47 am
simonecuttlefishUser avatarPosts: 114Joined: Tue Jul 06, 2010 12:24 am Gender: Pinecone

Post Re: Design ideas for a Sun finding sensor?

simonecuttlefish wrote:1) An array of photo-resistors, or photo-transistors, or cheap solar mini panel thingers, with a conical shield surrounding them to make them more directionally sensitive (not too much), possibly with a semi opaque shielding over them to soften the effect of localised reflections from nearby objetcs, that can "add up" their values based on incoming light levels?


Mini compound eye perhaps?
A device that looks a bit like 1/2 a disco ball, flat side down, but the 'mirror tiles' are light sensors, and they are sunken in to the "ball" a little, to make them more easily fall into shade, to reduce the number of detectors needed. Like a very simple compound eye.

EDITED IN
Or a sphere with hollow tubes running from the outside edge inside a short way, around the top half. With a camera or other detector inside the sphere, the position of the "bright spot" on the spheres internal lower half surface should give you the angles needed. Sort of like the stick and shadow method. But that might be epic fail if it's cloudy and you were trying to detect visible light levels I suppose. Ultraviolet perhaps?
I have no requirement for the supernatural or magic to explain anything, finding purpose in life other than trying to enjoy it, fantasies to explain wonder, fear that reason might trivialise me, or demand of wonders greater than those that are evident.
Sat Jul 31, 2010 4:23 am
borrofburiModeratorPosts: 3527Joined: Sun Jul 26, 2009 5:27 pm Gender: Male

Post Re: Design ideas for a Sun finding sensor?

simonecuttlefish wrote:Mini compound eye perhaps?
A device that looks a bit like 1/2 a disco ball, flat side down, but the 'mirror tiles' are light sensors, and they are sunken in to the "ball" a little, to make them more easily fall into shade, to reduce the number of detectors needed. Like a very simple compound eye.

Hmm, this was what I was trying to describe, though I wouldn't do half a disco ball, more like 3 quarters at least. Only I don't know that I would make it quite so large; if it were as large as a disco ball it would be a bit unweildy.
Sat Jul 31, 2010 2:25 pm
scalyblueUser avatarPosts: 1417Joined: Sun Aug 30, 2009 3:02 am

Post Re: Design ideas for a Sun finding sensor?

You are aware that a nintendo wiimote with a wii motion plus is an IR sensor with a positional gyroscope and tri-directional accelerometers. I've seen people calculate ocean height to the fraction of a millimeter by pointing a wiimote at an IR source floating on the surface. Given that the sun is a large IR source, you can make something that can identify what direction the sun is in and what angle it is at. May be useful.
_________________
悪夢の王の一片よ
空のいましめ解き放たれし
凍れる黒き虚無の刃よ
我が力 我が身となりて
共に滅びの道を歩まん
神々の魂すらも打ち砕き
Sat Jul 31, 2010 4:14 pm
Post new topic Reply to topic  Page 1 of 1
 [ 12 posts ] 
Return to Technology/Engineering

Who is online

Users browsing this forum: No registered users and 4 guests