Solar Tracker Project

This is my first Post

This is just the place I have been looking for ….
I am starting a project for a solar tracker using a stepper motor for the drive

The current design (not yet up and working) uses an ‘off the shelf’ lighting timer that ‘knows’ the sunrise and sunset times for each day of the year.

After this gets debugged I would like to start on a microprocessor version instead of the timer – to include an LCD interface and maybe some diagnostics and alarms for problems (problems ? what could go wrong)

My background is a field engineer for the last ~~30 yea ~~ well let’s just say a long time so I am pretty comfortable with electronics – programming not so much

I wonder which hardware package would be adequate ( I think the Mega 2560)

So let the flaming begin …

So let the flaming begin …

With the exception of a few sarcastic folks who will note that you did not ask a question, why would we flame you?

Oh, and welcome to the forum.

Back in the early 70's one of my teachers showed how to "follow the sun" (or stars) using a clock mechanism. After all, it's Earth-turn you're following. But then you have to get it set up with correct inclination and all.

Thank you

so is the mega 2560 gonna handle this ?

I see it has cos tan functions in the library

is there enough ram/eprom to handle this ?

I believe folks have used sunrise / sunset calculations in the much more modest Uno. I vaguely recall Jack Christensen recently mentioning the calculations and another user offering a simplified version. I suspect you will find a plethora of information searching the forum (I suggest using Google and initially including "Jack Christensen").

I probably should also answer the question...

so is the mega 2560 gonna handle this ?

Yes.

Thank you again CB

this is going to be exciting

Ultimately I would like this to be a kit or freeware so people can get more from their solar investment

Commercial 2 axis trackers are in the $4000 and up range -- I think i can get this to under a kilobuck or less

I found the Jack Christensen link -- it looks perfect

very nice not having to reinvent the wheel

The "down n dirty" way I use is with photo resisters. Normalize the output, build a jig that has them looking up with about 15 degrees between them . hook them up to analog 0 and 1. Read the output every so often( whatever you want) ant based on the delta V between them drive a stepper motor in the direction of the brightest light ( hopefully the sun).

You can program a min brightness so that the whole thing rests during the night. Since you are going to be tracking from East to West, you can either count the steps and when the brightness threshold drops below minimum, then step back East that number of steps and be ready for the next day. You can also just leave things as they are and the tracker will track back incrementally the next day.

This eliminates precise stepper calculations since the tracker stops when the photo resistors both bracket the sun.
I,ve done this, it works, but you have to get creative in lensing in the sun as it travels far to the North during the summer and passes overhead of the detector. I used an arched clear plastic lens with light baffles around the photo resisters so that a narrow strip of sky was presents with an arc of about 60 degrees N-S

Have Fun

As RPCoyle was saying

The "down n dirty" way I use is with photo resisters.

There is some updated code that isnt posted LCD And Limit switches

SolarPanels,
All you ever wanted to know about electronic trackers and sensors, http://www.redrok.com/electron.htm scroll down. I've built several versions and they all work but now I'm looking at the ones that interface with a microcontroller to prevent "hunting" and other problems with the simple ones.
Using LEDs as sensors really works better than I thought . The LEDcooker3 amazes everyone with its simplicity. 3parts.
Have fun.
TomJ

Why not use a GPS to give you an accurate time and location fix and then calculate the position of the sun? The algorithms for the location of the sun are very accurate and not computationally expensive so they would easily fit on your choice of MCU with a lot of headroom for the other code.

It has been a long time since I implemented solar location software so I do not have a link, but google is you friend.

wade

If I put a hole in a piece of opaque material then sunlight shining through will make a dot on a surface below. If I put a ring of sensors around that then when the sun moves the dot would strike a sensor. My accuracy goes up with the distance between the material with the hole and the surface below and the number of sensors in the ring.

For me, something like that might be used to tilt reflective panels to put more light on the collector. After all, they use solar panels in space where the light is far brighter than down on the ground -- especially at the latitudes most of us live.