need help to build a dual axis solar tracker.

[nilanjana]

I am a final year student of electrical engineering,and am currently
engaged with the development of a microcontroller controlled dual axis
parabolic solar tracker for the final year project. But I cannot
understand how to proceed, what to use and mainly how to assemble the
required components to build one.

I shall be grateful if you would help me out by providing me with
basic know how of how to build a tracker.

 

[MikeMl]

Come back when you have read the "solar seeker/tracker" threads on the Renewable Energy forum. There are hundreds of postings there, some of which have my name on them (see post 9)..

 

[alec_t]

Mods: Perhaps this could be moved to the Renewable Energy forum?

 

[dr pepper]

I have built a single axis version of the same thing, never used it just built it.
Its a pic16f88 microcontroller with 2 norp12 light sensistive resistors and a rc servo.
The micro positions the servo by adjusting a pwm o/p so that the light falling on the 2 cells produces an equal voltage within a bandgap.
Theres a couple of niceties, when the overall light level fails, the unit positions the solar panel facing east ready for sun up and goes into sleep, waking up every few secs to check the light level.
Shouldnt be that hard to increase to 2 axis.

I was going to build a r/c compatible servo from a car wiper motor, but never did for reasons below.

The problem with this method is clouds, the panel constantly positions itself in cloudy weather wasting precious power, I think systems that use gps and calculated solar position are superior.

 

[Mr RB]

The micro positions the servo by adjusting a pwm o/p so that the light falling on the 2 cells produces an equal voltage within a bandgap.
...
The problem with this method is clouds, the panel constantly positions itself in cloudy weather wasting precious power, ...

That's easy enough to fix in software, by making sure the motor can only move forward (will totally stop wind-back caused by clouds) and fixing a max forward rate which is slow, say 1.5 times the speed of the sun's movement means that wind-forward will be fairly minimal.

 

[alec_t]

If the tracker is going to be used at a fixed latitude/longitude you could do the whole thing by calculation of sun position. No need for light sensors.

 

[dr pepper]

Roman thats a good idea, I did implement slowing down the response as you mentioned, but I never thought of only driving forwards.

Building a servo controller into a pic is also a bit of a hurdle.

 

[MikeMl]

That's easy enough to fix in software, by making sure the motor can only move forward (will totally stop wind-back caused by clouds) and fixing a max forward rate which is slow, say 1.5 times the speed of the sun's movement means that wind-forward will be fairly minimal.

So how do you reset it for the next day? Timer and limit switches?

 

[dr pepper]

You could use what I mentioned earlier pick up the fact that the overall light has gone, ie sunset then position the cell east and wait for sunrise, resetting the position.

 

[Joe G]

I build one w/ duel tracking also, the first one I used the a2d module in a 16f866, that did not fair to well. next one I used a circuit similar to Mike led tracker, I used avoltage inverter instead. two old hp plotter motors moved the two axises. I have to wait till it gets warmer out to finish debugging it I used a ldr for day detection. when it goes dark, it waits a bit then moves east to a limit switch, and sleeps. when it receives first light it waits for a bit then rechecks and moves.......I was building it just for
fun also.

 

[dr pepper]

One of the simplest tracker controllers I've seen was just a rc servo, an rc servo tester and a couple of photocells.

The author just connected the servo tester to the servo, positioned it in the centre of its motion, disconnected the feedback pot in the servo and connected the 2 photocells in its place, so the photocells produced an error signal for the servo board and the servo positioned itself to balance the difference out again, only doing it through sunlight over the 'cells rather than the pot wiper.

Brilliant idea, only it doesnt sort out the cloud problem.

 

[Mr RB]

So how do you reset it for the next day? Timer and limit switches?

Yep, as DrPepper said you just wait for dark and then wind it back to the limit switch, that can be nice and slow as it has all night to do that.

The whole setup can be pretty energy efficient as it only needs to move very slowly, you can use a tiny gearmotor with high gearing.