I'm trying to build a structure onto which I can mount a solar panel where the structure will be able to turn and track the sun - to maximize solar energy input.
Can anyone suggest a photosensor that can be used to accomplish this? I looked online on some sites and they advertise photodetectors, however I'm not sure how good they are for this kind of an application (whether or not they'll just be dumb and happy with ambient light).
If you want to dinamically adapt the position of the panel with the position fo the sun, one good sensor is the panel itself :)
Doing MPPT (Maximum Power Point Tracking), you can also measure the output power and do small movements of the panel to detect the changes in power, and move it accordingly. It will be a bit tricky (and probably not worth) to tilt and rotate it, probably it's more convenient just to rotate.
You may also find this DIY light direction sensor interesting.
I think that the easiest solution is to use pre-programmed settings, because you can predict where the sun will be at a certain time, and set the orientation accordingly. Note that you can do it with relative settings and a compass (to make it aware of its heading), so you'll have flexibility about the position.
Usually the recommended angle for a panel corresponds to the latitude, but it will obviously vary during the day and with seasons.
Often though, the energy spent for following the sun doesn't justify the additional expense for motors, and the energy necessary to operate them.
I agree with clabacchio's answer! :-) The "pre-programmed settings" one, that is.
A solar follower will get confused when the sun is behind clouds, and parts where the sky is clear are brighter than clouded parts.
You need formulas for azimuth and elevation at your location as a function of day of year and hour of day. You don't have to compensate for analemma, that's only a few degrees error, which will give you only 0.05% less sunlight. Same for equation of time, which is less than 15 minutes.
edit
If instead of a solar panel you would want to control solar furnace mirrors you would have to incalculate analemma and equation of time, since a few degrees error would make you miss the furnace.
But for a solar panel even a 5° error is just 0.4% less power.
Using a photosensor guided tracking system is not usually the best way to achieve solar tracking in relatively simple amateur tracking systems. A simple tracker may use more power than it saves by tracking, may point the panel to the wrong location and is more complex and expensive than technically better alternatives.
A simple mechanical gearbox and motor that tracks where the sun should be in a daily arc, and which is adjusted every few weeks for seasonal change, will be likely to be more cost effective overall than any more complex tracker.
A completely manually adjusted panel that is repointed to local noon every few weeks and which is kept clean (and ideally also cool) will outperform many optical trackers.
An optical 1 axis tracker with two sensors and a shield that allows both sensors to be equally illuminated when the panel is pointing at the sun is liable to be more than good enough.
Complex multi axis optical trackers will often cause complex multi axis problems.
For all except the very most exacting or purposes a 1 axis tracker is adequate that follows the arc of the sun across the sky. Adjustment for seasonal azimuth change can be made occasionally manually. Even as little as a few times per season may be good enough and once every week or two will get you about within the general noise margin due to other factors in your system.
It is usually good enough to drive your 1 axis tracker with a dumb drive algorithm that points the panels where the sun should be and returns it at day's end to wait the new dawning. Any day that the sun is not EXACTLY where it should be you have much bigger things to worry about than solar tracking!
You can get some minor gains by optical solar tracking with a good system. You can get substantial losses with a poor solar tracker. A good solar tracker allows the panels to point at the brightest area of sky when the sun is partially occluded but with enough output that the panel still makes some output. A good optical tracker will be able to decide which of a number of areas is brightest and ideally will have some ability to estimate how long the low sun condition will last so that it will not be waving the panels to and fro and using up energy and putting load on the drive unnecessarily. A good MPPT tracker will not get stuck on a false peak, will not hop from peak to peak for little or no gain and its sensors will match PV panel response so it is not pointing at a spot that is optimum for it but not for the panel.
A bad optical tracker may hunt unnecessarily, may use more energy in tracking to and for that it gains, may be a net maintenance liability, may get stuck on false MPPT peaks and may go hiving to an for chasing flitting bright spots when the sun level overall is too low to be worth the effort.
Which sort of optical tracker do you think the average amateur is most liable to build? ;-).
A well maintained and cleaned panel - and cooled if you can manage, that is pointed to local noon and suitable seasonal azumith angle and altered manually maybe weekly and otherwise not steered at all will give a good steered array a good fight and will not need a motor, gearbox, drive etc to be built.
IF you can cool a PV panel you'll get 10%+ more out on a hot day - more than you'll gain from most trackers.
Here is an interesting project utilizing Light Dependant Resistor:
http://www.pc-control.co.uk/howto_tracksun.htm
Though it is important to do the power calculations and size your panel accordingly so that you get a net positive power generation. You have to take into account the amount of power utilized to calculate and determine the position of the sun and also the power utilized to move the panel.
Have you ever considered using a rounded top concave glazed surface that would redirect the sun no matter what angle if would be in by chance ? I'm doing a final year project the the moment and that's my concept ..
