Building a Solar Parabolic

[MishraArtificer]

How much for a basic light meter? Or hardwire in a piezoelectric cell (I think that's what I'm thinking of...).

[JesseK]

For 1 degree of freedom tracking (1 axis that turns 'left or right' as the day progresses) you could have two light sensors with a slight angle offset (one pointed slightly east, the other slightly west). One will detect more light than the other. Simply mount the sensors on the solar array and repoint the solar array in the direction of more light until both cells detect equal amounts of light. This makes the assumption that the vector that symmetrically divides the angle between the two is perfectly orthogonal to the Earth's surface at the sun's peak.

-- edit -- it's probably best to add some sort of light-blocking barrier (weather resilient) in between them too so that one will be put in the 'shadow', guaranteeing that the other sees more light at some point, therefore moving the arrays. Should be a simple way to maximize output, if only by a few degrees worth of lumens over the whole array.

[ebarker]

About 50 years ago an engineer at TI invented a 4 quadrant sensor, used for guided missles.

Same concept



up/down difference = (sum A + B) - (sum C + D)

left/right difference = (sum A + C) - (sum B + D)

You need 4 light sensors

you could do the whole thing in analog electronics, op-amps and such,

OR

do it all in software using 4 A/D channels on a processor, like an old robot controller.

[MrForbes]

how sharp is the focus? If it's not too sensitive, you could use a non-feedback system. I'm assuming you have one axis of rotation. Just program what angle it needs to be at each time of day, have it move a bit when needed. The position will change over the year, so program that in as well.