Let’s face it — solar panels still aren’t that efficient. So why not pump as much juice out of them as possible? Building a 2-axis solar tracking unit can increase daily power output by around 30%!
[Jay Doscher] had his power go out back in 2011, and even though it was only for 12 hours, they realized how ill-prepared they were to deal with a power outage. Food was spoiled, flashlights were dead, candles were sparse… they needed to be prepared better for the next time this happened. This spawned one of [Jay’s] longest running projects on his blog Polyideas.
His goal was to build a fully automated solar tracking unit that could be setup anywhere, and automatically track the sun to ensure optimum ray catching. It makes use of a 12V gear reduced motor to provide panning, and a linear actuator with positional tracking to control the tilt. To track the sun he’s got a digital compass and an Adafruit Ultimate GPS breakout board. To control it all he’s using is an Arduino UNO, but he has been through multiple iterations including his first with a BeagleBone. It’s a very slick and well engineered system and [Jay’s] hoping to spread it around the world — the entire thing is open source. What a guy!
It’s not quite complete yet, but he’s got an amazing build log and a GitHub repository filled with info — plus the following video showing it off in its current state!
For a cheaper solar tracker you can build for fun, check out this baby-sized solar tracker that also uses an Arduino! Or how about this one built out of parts from the hardware store?
21 thoughts on “2-Axis Solar Tracker Always Gets A Tan”
Lots of great building, but it seems it might be easier to sense brightest part of the sky directly, and point at that.
Would it have been cheaper to buy 30% more solar cells?
It would depend on the scale of your system. If it was large enough then I would imagine that’d definetly is worth it.
We were doing this in 1976 with very simple analog parts (tracking the brightest part of the sky). Are you sure that is 30%? You should be getting a factor of 3 more power from tracking. It isn’t just the cosine of the angle. You get total reflection at low angles and some reflection all the time. So, tracking is a major improvement and great for a small installation. In solar farms there is shadowing of neighboring panels and it gets stupid pretty quickly.
It wouldn’t be getting a tan here today. I haven’t seen the Sun all day long. So far we’ve gotten almost an inch of rain though.
A couple of toughts:
1. Sky tracking is OK only if you have direct sunlight. With diffused light (=cloudy days), solar tracking is not very effective.
2. Is not zombie-apocalypse-proof. GPS would decay within a week without human intervention. 0/10. :D
If you want juice, buying (building!) a solar panel with more cells would be the best option. Unless you live in Sahara desert.
Thanks for all the comments! You can certainly use a larger panel, and the benefit is greater then as well. You don’t have to use a GPS unit- you can hard code the latitude and longitude into the sketch, but you’ll have to add a RTC since the Arduino doesn’t have one my default.
Rather than try to calculate the position of the sun, use two photodiodes and BEAM robotics to just adjust it until it’s right on.
Be interesting to see if the power gained by optimal tracking EXCEEDS the power cost of moving the panels.
Yes- even in winter here in California where the sun makes a sweep along the horizon fairly quickly, the Arduino sketch defaults to moving only every 10 minutes- so there’s not much power used- the Arduino idle uses very little.
Nice, and a good exercise. But WAY overkill when a couple of sensors and a couple of opamps would do the job.
Like I said, nice from a learning standpoint, and I’ll never say that’s a bad thing. But as far as building something like this for the job it does….. way overkill.
As a general rule it is more effective to track knowing your lat/lon and time. Unless you are careful, on overcast days you might just burn more energy tracking the sun than you gain by proper aiming
I sorta want to replace the solar panel with a satellite Internet dish and use it as an auto-aiming sat so I don’t have to muck with my mobile Starband system trying to aim it…
see the redrock link I posted below – you can drive the actuators to keep it in position
gps is overkill!
Why is GPS overkill? The object is to have a mobile system…without the GPS you would have to enter the lat/lon manually every time you you were in a (sufficiently different) location. The GPS and the compass would mean that you could plunk it down anywhere and not have to fuss with it.
For the problems it solves with a portable solar tracker, the GPS unit is fairly cheap- the $40 unit is one of the most important parts of the project. As I posted earlier, you can hard code your latitude and longitude, but you’d still have to add a $20 RTC because the Arduino doesn’t come with a clock.
If an enterprising individual wanted to create a single PCB with all the components onboard, you’d probably find this to be a relatively affordable solution compared to GoalZero stuff. You could even add dials or switches to manually enter your latitude and longitude and set the time, but you’d actually be adding significant cost because of that hardware.
Redrock makes a $20 kit for solar tracking – detailed schematics and plans on their website with lots of other solar tracking info – great resource http://www.redrok.com/led3xassm.htm#led3xc3
Why not use a polar mounted single axis tracker ? It would be simpler, and only requires some trivial setup after moving to a new location.
This project doesn’t require any other electronics to run- so no computation is needed by a smartphone or other device. If you use a 1-axis configuration, then you’d have to deal with the following drawbacks: A.) You would have to compute the fixed axis that your panel would roll on. B.) Your panel would usually be off-axis when operating during the winter or far from the equator, where the sun travels low across the sky.
Additionally, the rotation motor and compass are fairly cheap components- so removing them would add complexity for the user and not much cost. If cost is a concern and complexity is not, a fully manual tracker may be ideal for you.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)