Solar powered light powered by Raspberry Pi

Support
1

Hi all, I’m working on my first electronics project and I’d appreciate a sanity check from people who know what they’re doing.

The goal is to build a lamp that runs entirely from a solar panel, controlled by a Raspberry Pi Pico 2WH. I’m aiming to keep everything no-solder if possible (as I don’t have tools or experience) but could find a maker space for this if you think it’s important.

The light itself will be a WS2812B LED strip (though I’ll probably only be driving a handful of pixels at a time to reduce power draw). The Pico will handle scheduling, colour control and brightness (my idea is to have the light follow the sunrise-sunset of another time zone). The solar panel will be outdoors on the roof, and the battery plus electronics will live indoors, so I figure I’ll daisy chain DC extension cables between the two.

Here’s the full parts list I’ve put together:

  • Raspberry Pi Pico 2WH

  • Adafruit USB/DC/Solar Li-ion charger v2

  • Voltaic P110 10 W 5 V solar panel

  • DC extension cable (5.5 mm × 2.1 mm)

  • Panel plug adapter (3.5 mm to 5.5 mm)

  • 2×18650 battery holder

  • Two 18650 Li-ion cells (3.7 V, 2600 mAh each)

  • 1 m WS2812B strip (144 LEDs/m)

  • 3-pin LED strip connector cable

  • 1000 uF electrolytic capacitor

  • Logic-level shifter (3.3 V to 5 V)

  • 300-point solderless breadboard

  • Jumper wire sets (M/M, M/F, F/F)

  • USB-A to micro-USB cable to power the Pico from the charger

I believe this collection of parts should all play nicely together, but since I’m new to this I’d love to know if there’s anything obviously missing, any easier alternatives, or any places where I’m making things too complicated.

One question I’m not fully confident about is the power side: is a 10 W panel feeding two 2600 mAh 18650s enough for running a Pico plus a small number of WS2812B LEDs for ~6 hours per day?

Any feedback or warnings would be very welcome.

1 Like
2

The following is just a high level to give an starting point. I would recommend getting the pico and led strip running as you want of a power supply, then when running as you want, measure the current/power draw and add a buffer. At this point select the parts for the power and charging…

The following should NOT be used as a recommendation, Its an example of how the math works. You will see that the more power you use the more power storage you will need and more solor charging you will need. So the get the best match, you need to know the actual power needed (rather then guess) i.e. build and measure, tweak and re-measure till you happy. Also you need to check the data sheets for the actual parts you chose and check if they can keep up.

Example Math… Lets others double check my math as I may have missed something…

From memory the WS2812b leds if all fully on are about 45mA per led (note different data sheets say different current from 35mA - 50mA)
Lets say, out of the 144 in your stipe, at any time you only have 10 on. Of those 10 lets say 1/2 are at fully on (which would be white) and 1/2 are at 50% on, some other dimmer color.
So …
5 x 45ma = 225 mA
4 x (45/2 ma) = 112 mA

The amount of power the Pico will use will depend on if WiFi us used how fast you run the clock etc… so lets just use 50mA (but this should be measured when its running as you need.

Pi Pico 50mA
Leds 337ma

Total about 387mA
Assuming a 24hr run time total power consumed (not including losses and other needs) = 387mA * 24hrs = 9288mA/hrs most of this pull is at 5V for the leds.
So total power => 5V * 9.288A = 46.44 watts… lets round up to 50watt/hrs

Next we take that 50w/hrs and see if it can fit into the power storage.
18650 Li-ion cells (3.7 V, 2600 mAh each) => 3.7 * 2.6 = 9.62 * 2 = 19.24 Watt/hrs

Based on my math in this example, we need 50Watt/hrs, but the power storage only has 19 Watt/Hrs. At this point we need to lower the power needs, e.g. less LEDs on at less power (which could still work for you, run some tests)
or we need more power storage. Going with this example
50W/h / 9 Watt/hrs per 18650 = 5.56 18650 so lets round up to 6 Batteries.

Now that we know the power storage, 6 x 9 Watt/hr 18650. = 54 Watt/hrs
How long to charge. (Need to get the actual charge rate for the batteries and charge controller as they can affect things)… but to keep this example running, lets assume the cells can charge at a rate of 0.5C we can pump in power at 2600 / 2 = 1300 mA (1.3A) per 18650 * 6 = 7.8 Amps charge rate. Assuming a charge voltage of 4.2V, then power need to charge would be 32.76 Watts for 2hrs.
Given a 10W panel, it will max out at 10 watts, so are charge rate then is less the 0.5c so longer to charge. 32.76/10 = 3,2 times 2hrs = 6.4hrs.

So if your happy with 7 hrs of 100% output from the panel per day then it should be ok, but I suspect you wont get that as peek output will only be for a few hours for the day… so 20watt might be better.

1 Like
3

Hi Streamline

Be a little bit careful here.
Do you want the cells serial or parallel. These holders come in both flavours.
Also make sure the cells will fit. If the 18650 has built in protection they could be almost 70mm long and are extremely hard to fit into some holders, and damn near impossible to remove.
I was advised by Core on a previous occasion that all the cells they market are 65mm long. I suppose that still applies. Core might like to confirm
Cheers Bob

1 Like
4

Thanks both - Michael that advice about getting it working on a power supply before introducing solar/battery is very sensible, especially with the panel/charger being the most expensive parts. And your maths worked example will be helpful for me to follow once I have real data.

Robert - thanks again, I have no idea about serial/parallel, that’s something for me to research :slight_smile:

1 Like
5

Hi Streamline
Quick hint.
Parallel will result in 3.7V 5200mAhr
Serial will result in 7.4V 2600mAhr
Cheers Bob

1 Like