Disclaimer: This post was written with the assistance of Claude Sonnet

Solar-Powered Tool Shed: My Setup Guide

This is a writeup of the small solar charging system I built on top of my tiny tool shed. It’s designed to keep my Ryobi power tool batteries charged from sunlight, with enough overhead to also run a low-power Raspberry Pi as a static web server.

The Components

Component	Role
SOLPERK 50W 24V Solar Panel	Generates power from sunlight
10A PWM Solar Charge Controller (included with panel kit)	Regulates charging of the battery
ECO-WORTHY 12V 20Ah LiFePO4 Battery	Stores energy for use when the sun isn’t out
24V→12V Buck Converter (10A / 120W)	Steps down the controller’s load output for 12V devices
CharIend One+ Ryobi Charger	Charges Ryobi 12V/14.4V/18V tool batteries from 12V input
Raspberry Pi 400	Lightweight static web server, powered via USB

How It’s Wired

The system has two distinct paths: a charging path and a load path.

Charging Path

Solar Panel (24V)
       ↓
PWM Charge Controller
       ↓
LiFePO4 Battery (12V)

The charge controller is connected directly to the battery. It auto-detects the 12V battery and applies an appropriate charge profile, while the panel feeds it on the input side.

Load Path

PWM Controller (load output)
          ↓
Buck Converter (24V→12V)
          ↓
     Ryobi Charger
     ↙         ↘
Tool Batteries  USB Port
                   ↓
             Raspberry Pi

The buck converter steps the load output down to a stable 12V, which the Ryobi charger expects as a vehicle-style input. The charger then handles the battery chemistry-specific charging for the tool packs, and offers two USB-A ports as a bonus for charging phones or running a small device.

A Note on the Voltage Mismatch

The panel is a 24V nominal panel and the battery is 12V. With a PWM controller in the middle, this means the system only captures roughly half of the panel’s rated wattage — a 50W panel behaves more like a 25W panel in this configuration.

This is because PWM controllers don’t convert excess voltage into extra current; they essentially just connect the panel to the battery and modulate the connection. The “lost” voltage is dissipated rather than harvested.

Why I left it this way: the system meets my actual demand. My tool batteries get charged, the battery stays topped up, and I have headroom for the Pi. If I ever find myself power-limited, the upgrade path is to swap the PWM controller for a small MPPT controller (something like the Victron SmartSolar 75/10 or an EPEVER Tracer 1210AN). An MPPT controller would convert the excess panel voltage into usable current and roughly double the harvest from the same panel.

Powering a Raspberry Pi from the Ryobi Charger’s USB Port

The Ryobi charger has two USB-A ports in addition to its tool battery port. For a Raspberry Pi 400 running nothing but a static site, this works well.

A static site server is an extremely light workload — the Pi will idle around 2-3 watts, well within what a basic 5V USB-A port can supply. The main things that matter:

• Use a good USB-A to USB-C cable. Cheap, thin cables cause voltage drop, which causes the Pi’s undervoltage warning. A short, thick cable rated for 3A solves this.
• Make the SD card resilient to power loss. Since power depends on battery state, enable the Pi’s read-only overlay filesystem, or use log2ram to keep writes off the SD card. This prevents filesystem corruption on sudden shutdowns.
• Pick a lightweight server. Caddy is the easiest choice for static hosting (and it handles HTTPS automatically). nginx works too. Either will idle near zero CPU.

Power Budget

A back-of-envelope calculation for the Pi’s draw on the battery:

• Pi at ~3W continuous = 72 Wh/day
• LiFePO4 battery capacity: 12V × 20Ah = 240 Wh
• One full day with zero solar input: battery drops to ~70%

Even mediocre solar days should easily make up for that, especially since the tool battery charging is intermittent rather than continuous.

Future Upgrades I’m Considering

• MPPT controller swap — to actually capture the panel’s full output.
• Direct 12V-to-USB-C buck converter for the Pi — bypassing the Ryobi charger entirely would be slightly more efficient and give the Pi a cleaner 5V supply.
• A second panel — if I ever add more loads, doubling up panels is cheaper than upgrading the controller alone.

Lessons Learned

The biggest one: system design matters more than component specs. A 50W panel that’s only delivering 25W of usable power isn’t a “broken” system — it’s a system that works, just not optimally. Knowing where the inefficiencies are means I can target upgrades when (and only when) the system stops meeting my needs.

The other lesson: LiFePO4 is wonderful for this use case. It tolerates partial-state-of-charge cycling far better than lead-acid, has a built-in BMS for protection, and is light enough to mount easily inside a small shed. Worth the price premium over a sealed lead-acid battery for any solar project that gets daily use.