land o plenty

Get them caps too Jeremy.

Kind of fun sometimes to strip down an old TV.  hehe.

let get the hot iron works....

use a torch and knock them all off at once! lol ~Russ

salvaged 10uf caps on a 16 pin breakout for temperature sensors. it fits into the analog slots on an arduino mega 2560 and is part of the new controller for my solar heating.

Haha. Oh snap son!  Nice!!!!!!~Russ

working upstairs

fixed....hopefully

i will try and remove parts as revisions are made....work in progress

the breakaway hinge on 'panelstand.stl' is a bit touchy....i had 1 side break clean and the other not. second print in progress before i revise it to beef up the material around the hinge.

the story thus far

Looking pretty good Jeremy.

What's going to be your control strategy for tracking?  Absolute based on time day, day of year, or will you add some feedback to actually optimize the panel output, say once per hour?

If you want to have some fun, use a little artificial intelligence to actually learn the solar pattern.  Then you can place these things anywhere, set'm and forget'm.  :)

im planning to use a system similar to daniel gendrons tracker....with 4 photo-diodes. i had thought of making 'blinds' over the diodes to focus the light....but the tubes work very well also  :)
we shall see.

https://www.youtube.com/watch?v=pHhqhxgKcnE

a little more meat around the hinges and standoffs to keep the panel out of the way.

Quote from jeremy gwilt on March 5th, 2017, 11:22 AM

im planning to use a system similar to daniel gendrons tracker....with 4 photo-diodes.

Interesting, though I have a comment about his system...  Not a real good idea to have this thing constantly tracking at that rate of feedback.  It could potentially waste a lot of electricity moving unnecessarily.  Better if it uses heuristics to predict the approximate sun location and then have slow feedback updates maybe every ten minutes for tight alignment.

It should be pretty easy to assess how much angle actually shows itself as a power increase or dip.  For full optimization, you do not want to make any moves where the power it costs you to make that move is more than you gain by making it.  Again, some AI here is helpful so the system learns what good and bad moves look like and remembers for the next time it encounters that situation.  Start by counting Joules in & out and calculate from there.

This does look to be a pretty neat project.  Lots of little tweak you can make.

"Not a real good idea to have this thing constantly tracking at that rate of feedback"

i dont see the sun bouncing around that much....but as i explained to my wife:
this is set-it and forget it. i really dont want to go through the hands-on hassle of calibrating the panel to match the season with no feedback but my own input. yikes.....im curious but not that committed.
yes, it is reactive and will have some extra jitter....but i dont expect it to be much. i mean...the sun isnt flying through the sky in some random arc or anything. automation is what you make it.
on the other hand.....PV reacts very well to diffused sunlight. if the sun is behind a thick cloud, but shining harder through some thinner clouds a couple degrees off....i want the panel to pick that up. waste not want not....or why track at all?

so theres my 2-cent rebuttal  :)   remembering that im no expert either.......i havent even seen a tracking system with my own eyes....which is why im building this small model to collect data. a stationary panel will act as control, and i can data log wattage throughout the day (used by motors as well) and get some REAL information about the viability. thats what this is all about right.....?

isnt it......?

Quote from jeremy gwilt on March 5th, 2017, 03:33 PM

...which is why im building this small model to collect data. a stationary panel will act as control, and i can data log wattage throughout the day (used by motors as well) and get some REAL information about the viability. thats what this is all about right.....?

Yes sir.  It sure is.

Once you have some of that data, you'll soon become the resident expert.  Or at least the person who knows what works and what doesn't.

...

...

3dprinted card