Another contraption: solarTracker

Showroom for MicroPython related hardware projects.
Target audience: Users wanting to show off their project!
Post Reply
viiveli
Posts: 5
Joined: Sat Jan 09, 2021 5:53 pm
Location: Finland
Contact:

Another contraption: solarTracker

Post by viiveli » Sat Mar 06, 2021 8:29 am

I guess when you're bored enough, you'll build your own toys.

https://github.com/viiveli/solarTracker

Had some solar cells and a supercapacitor module lying unused and since I have nothing but time to spend at this point of my life I decided to build something: Autonomous photovoltaic energy harvester running on ESP32 devkit board and micropython. Device uses solar panels to charge a supercapacitor module and once the voltage is high enough the "brain" will boot up and align the solar panels with the brightest source of light ...usually the sun. The device can rotate 360 degrees horizontally and roughly 110 degrees vertically. Average power production level can be monitored and axes controlled through a web UI.

A few design flaws, sketchy soldering, some glue and lots of coding anxiety but eventually a working and kinda useless contraption - fun and rewarding!

These days it seems like you can order pretty much everything by mail, including custom parts and PCB:s - and this is what I wanted to test with this project. Of course this way the price per part is kinda high and of course it would be nice to have your own 3d printer or a CNC router but it's not necessary if you want to build something. Those tools can be really expensive, and they require space and maintenance too. Sure it takes a bit more time this way and if you make a design mistake you can't just instantly make a replacement, but let's not be in hurry all the time

User avatar
scruss
Posts: 360
Joined: Sat Aug 12, 2017 2:27 pm
Location: Toronto, Canada
Contact:

Re: Another contraption: solarTracker

Post by scruss » Sat Mar 06, 2021 5:23 pm

Neat - but what are those shades for? If you shade even a tiny part of a module, you'll get near zero current.

Some commercial trackers use two cells about the size you're using, mounted at 90° to one another, and do some iteration to find the optimum sun value. They use to get really confused by snow, though

viiveli
Posts: 5
Joined: Sat Jan 09, 2021 5:53 pm
Location: Finland
Contact:

Re: Another contraption: solarTracker

Post by viiveli » Sat Mar 06, 2021 5:57 pm

scruss wrote:
Sat Mar 06, 2021 5:23 pm
Neat - but what are those shades for? If you shade even a tiny part of a module, you'll get near zero current.

Some commercial trackers use two cells about the size you're using, mounted at 90° to one another, and do some iteration to find the optimum sun value. They use to get really confused by snow, though
It seemed like the panels got "saturated" real easy and produced a near identical voltage even with around 45 degrees tilt angle in bright sunlight. Shades are there to exaggerate the differences - voltage drops on the opposite side as the sun shifts even a bit

The whole gizmo is a bit over-engineered if tracking the sunlight is the only goal. There are plenty of nice examples of really usable "dumb" trackers without any microcontrollers in YouTube, such as https://youtu.be/wL9PcGu_xrA

User avatar
pythoncoder
Posts: 5956
Joined: Fri Jul 18, 2014 8:01 am
Location: UK
Contact:

Re: Another contraption: solarTracker

Post by pythoncoder » Sun Mar 07, 2021 2:56 pm

That YouTube idea is very neat. It depends on how you define "simple". It's arguable that closed loop designs are overkill as you're tracking something whose position is known at any time. An open loop design will point in the right direction even if the sky is completely overcast.

Azimuth rotates exactly once every 24 hours*. Elevation can be computed as a function of time and date given latitude and longitude. A pair of stepper motors will deliver the correct position.

* unless you're at or near the equator, when you only need one degree of freedom.
Peter Hinch
Index to my micropython libraries.

viiveli
Posts: 5
Joined: Sat Jan 09, 2021 5:53 pm
Location: Finland
Contact:

Re: Another contraption: solarTracker

Post by viiveli » Sun Mar 07, 2021 4:57 pm

pythoncoder wrote:
Sun Mar 07, 2021 2:56 pm
That YouTube idea is very neat. It depends on how you define "simple". It's arguable that closed loop designs are overkill as you're tracking something whose position is known at any time. An open loop design will point in the right direction even if the sky is completely overcast.

Azimuth rotates exactly once every 24 hours*. Elevation can be computed as a function of time and date given latitude and longitude. A pair of stepper motors will deliver the correct position.

* unless you're at or near the equator, when you only need one degree of freedom.
Yeah. It's all math if you know where you are. Solutions that rely on light intensity (like mine) gets distracted with reflections or pretty much anything that blocks the light.

...GPS modules costs next to nothing these days right?

User avatar
scruss
Posts: 360
Joined: Sat Aug 12, 2017 2:27 pm
Location: Toronto, Canada
Contact:

Re: Another contraption: solarTracker

Post by scruss » Mon Mar 08, 2021 2:07 am

viiveli wrote:
Sat Mar 06, 2021 5:57 pm
It seemed like the panels got "saturated" real easy and produced a near identical voltage even with around 45 degrees tilt angle in bright sunlight. Shades are there to exaggerate the differences - voltage drops on the opposite side as the sun shifts even a bit
Or maybe you just discovered what most solar developers know: trackers are an almost complete waste of time. You can pay for a lot of fixed racking + extra modules for the price of trackers, and you avoid the maintenance aspect.
viiveli wrote:
Sun Mar 07, 2021 4:57 pm
...GPS modules costs next to nothing these days right?
Yup. Just make sure you put them above the modules, as solar panels are fantastic at blocking radio waves. They're already in use in solar installations on remote-ish parts of the grid (like this one in Northern Ontario) to ensure the inverters are supporting grid phase and not fighting against it. They're also the cheapest way of ensuring that aircraft warning lights all flash in sync on a wind farm.

I kind of miss my days designing huge renewable energy installations ...

viiveli
Posts: 5
Joined: Sat Jan 09, 2021 5:53 pm
Location: Finland
Contact:

Re: Another contraption: solarTracker

Post by viiveli » Mon Mar 08, 2021 6:39 am

scruss wrote:
Mon Mar 08, 2021 2:07 am
Or maybe you just discovered what most solar developers know: trackers are an almost complete waste of time. You can pay for a lot of fixed racking + extra modules for the price of trackers, and you avoid the maintenance aspect.
Now come on, hobby projects are never a waste of time :) ...but in all seriousness, for anything even remotely professional I too would think twice

Areba21
Posts: 3
Joined: Sat Nov 13, 2021 5:04 am

Re: Another contraption: solarTracker

Post by Areba21 » Tue Nov 16, 2021 4:12 am

Personally, I've never worked on or designed a dual-axis tracker. We've ruled them out due to their intricacy and land needs, but they do make sense for a residential application, especially if you're willing to conduct some upkeep yourself. I believe that a slanted single axis or a single vertical axis azimuth tracker is the next best option for yield and is also the most straightforward.
Active tracking is something I would avoid at all costs. On foggy days, I've seen them perform some amusing things.
I found some good tracker and bike accessories for my bike dirtbikeadvisor

Post Reply