Be generous with PV watts
Commodity rigid panels can be remarkably inexpensive per watt. The entry system aims to use enough panel area that charging does not depend on squeezing every last percent through expensive electronics.
Early-stage engineering project
Useful electricity first.
Explore a radically simple household solar system built around inexpensive photovoltaic watts, rechargeable appliances, field-serviceable parts and a path to larger systems as a household’s needs grow.
This is a development site, not a product offer. Costs, specifications and financing shown here are working targets under test.
~150 Wworking entry-panel target under evaluation
US$55–60working hardware-cost target
No inverterat the entry level
No central batterystorage lives in appliances
A different starting point.
Conventional solar home systems concentrate cost in a battery, inverter and controller. This project asks whether the first useful rung can be much simpler: a commodity solar panel feeding efficient DC converters and appliances that already contain their own small rechargeable batteries.
Put storage where it is used
Lights, fans, radios, phones and small televisions can carry their own batteries. That avoids a single central battery becoming the cost and failure point of the whole household system.
Keep the first system DC
USB and low-voltage DC cover a surprising range of useful loads. An inverter becomes an upgrade rather than a requirement on day one.
Design for a screwdriver
Connectors, replaceable modules and straightforward wiring matter. A field technician should be able to isolate a failed branch and replace a low-cost part without specialized equipment.
Make upgrades additive
A second panel, larger storage, 24 V loads or a future inverter should add capability without discarding the useful appliances and PV already purchased.
Test ugly conditions
Clouds, low sun and repeated brownout/recovery matter more than a perfect bench supply. The development rig is being built to replay measured solar profiles and expose bad charger behavior early.
One solar lead-in, one household charging hub.
~150 W panelWorking entry target · outside in useful sun
PV lead-in cableOne low-voltage run into the house
Distributor V1Independent 5 V and 12 V DC channels
Charging shelfLights · fan · radio · phone · power bank
The distributor can mount immediately above a simple shelf so appliances stay organized, off the floor and close to short charging leads.
The long-range objective.
This is deliberately ambitious: explore a ten-year program capable of bringing 200 million African households to at least a useful Tier-2 level of electrical service, with voluntary, self-funded progression toward much larger systems. What do the tiers mean?
200 million households
At least Tier-2-scale useful electricity: lighting, communications, fans and other low-power household services.
50 million to Tier 3
Additional PV, storage and power capability as household demand and ability to pay increase.
12.5 million to Tier 5
Full household electrical service for a first large cohort. The broader aspiration remains a path to Tier 5 for everyone.
Prototype now
From idea to hardware.
The first breadboarded distribution unit uses several inexpensive USB buck modules. A rechargeable COB light is bright while charging, and an old cell phone successfully returned to service after being connected to the prototype.
✓
Multiple independent 5 V branches.
One misbehaving load need not collapse every charging outlet.
One misbehaving load need not collapse every charging outlet.
✓
12 V branch under development.
For small TVs, larger fans and other useful DC appliances.
For small TVs, larger fans and other useful DC appliances.
→
Next test: qualify the Distributor with real PV and realistic mixes of partially discharged appliances; current prototypes use ~100 W-class panels while the working deployment target is moving toward 150 W.
Distributor V1
Five charging channels on one 100 × 100 mm board.
The first integrated distributor PCB has now been laid out and ordered for prototype fabrication. It combines three independent 5 V buck-converter channels feeding six USB-A sockets with two independent 12 V channels feeding four barrel-jack outputs.
3
5 V channels.
Each channel feeds one stacked dual USB-A connector.
Each channel feeds one stacked dual USB-A connector.
2
12 V channels.
Each channel feeds a pair of center-positive 5.5 × 2.1 mm outputs.
Each channel feeds a pair of center-positive 5.5 × 2.1 mm outputs.
PV
Intended source.
A nominal-12 V panel in the ~150 W class is now the working deployment target, typically around 18–20 V at maximum power and roughly 22–25 V open circuit. Current ~100 W-class prototypes are being used to qualify the electronics first.
A nominal-12 V panel in the ~150 W class is now the working deployment target, typically around 18–20 V at maximum power and roughly 22–25 V open circuit. Current ~100 W-class prototypes are being used to qualify the electronics first.
↔
Replaceable connector leads.
The field-use concept is to leave short, inexpensive pigtails plugged into the board so routine wear occurs at a replaceable cable end rather than at the PCB-mounted socket.
The field-use concept is to leave short, inexpensive pigtails plugged into the board so routine wear occurs at a replaceable cable end rather than at the PCB-mounted socket.
V1
Prototype mechanics.
Several generic connector and protection-component footprints still need to be checked against the delivered parts.
Several generic connector and protection-component footprints still need to be checked against the delivered parts.
Solar that behaves like solar.
A programmable bench source is being built around an ESP32-S2 Mini and a buck power stage. It will replay measured Nova Scotia solar-current profiles so charger recovery can be tested against passing clouds, poor mornings and repeated power collapse—not just a laboratory supply.
Why brownout recovery matters. A directly connected solar panel is not a stiff power supply. When cloud cover suddenly reduces available current, panel voltage can collapse under load; inexpensive converters and charging devices may then cycle, stall or recover poorly when sunlight returns. The emulator exists to stress-test clean automatic recovery under those ugly real-world transitions.
PAYGo without making the electronics expensive.
A working financing hypothesis is deliberately simple: make the upfront hurdle small, let the household own the system after a fixed period, and use repayment history as a bridge to larger solar rather than adding complicated lockout electronics to every entry system.
$10working down-payment target
Then $3 per month for 24 months
That produces US$82 in nominal receipts. The present working cost model is roughly US$55–60 of hardware plus field installation and infrastructure. Real financing cost, payment fees, defaults, warranty and local operating costs still have to be demonstrated.
$10down payment
24 × $3monthly payments
$82nominal total
Electricity can become productive capital.
A useful reason to design for upgrades is that household solar can support income as well as lighting and communications. A 2019 GOGLA/Altai Consulting follow-up study tracked PAYGo solar-home-system customers in five East African countries and revisited them about 15 months after purchase.
A longer follow-up
The study revisited customers well after the first novelty period rather than measuring only the first weeks of ownership.
Reported additional income
Twenty-eight percent of households reported additional income associated with economic activity using their solar home system.
Average reported gain
For households reporting additional income, the reported increase averaged about US$46 per month.
What this does—and does not—show: the research covered 1,419 solar-home-system owners in Kenya, Mozambique, Rwanda, Tanzania and Uganda. It was observational rather than a randomized trial, so the result should not be read as a promise that solar will raise every household's income—or as a number that can simply be multiplied by universal access. The report found that 21% of systems were being used in a business or income-generating activity after 15 months. Among those businesses, paid phone charging was common: 39% named it as the primary activity and 48% offered charging as a service. The report explicitly notes that this customer base shrinks as neighbors acquire their own solar systems. Those early-adopter scarcity rents should therefore be separated from more durable benefits such as longer working hours, a reliably charged phone, avoided charging and kerosene trips, and electricity used as an input to other businesses. GOGLA, Powering Opportunity in East Africa (2019) ↗
Useful add-ons, not a locked bundle.
The entry system should support things a household actually wants to add. The most interesting options either use daylight directly or bring their own storage.
Representative appliance types, not a locked bundle or vendor recommendation. Exact products, battery capacities and charging behavior still need field testing.
Portable storage: charge by day, use anywhere.
Small batteries move with the appliance instead of being concentrated in one expensive central pack.
USB power bank10–40 Wh of portable nighttime energy.
12 V televisionPreferably with an internal battery.
Larger rechargeable fanUseful comfort load without an inverter.
Laptop / tool batteriesPortable stored energy that can also support productive work.
Daytime-direct loads: use surplus sunlight without storage.
When the sun is strong, some work is cheaper to do immediately than to route through a battery.
Water pumpRun when solar input is high and store the water rather than the electricity.
Immersion heaterTurn strong midday sun directly into stored hot water.
Food warmer / slow cookerInsulation can carry midday thermal energy into the evening.
Second 150 W panelAdd it in parallel (1S2P) for a 300 W batteryless system. Later, the same two panels can be rewired in series (2S1P) when a 24 V battery and solar charger are added.
A deliberately additive growth path.
The current design direction is to make the first panel the beginning of a household power system, not a disposable starter kit.
Start batteryless
One nominal-12 V panel feeds the Distributor directly. Storage stays inside the lights, radio, fans, phone and power banks that need it.
Add the same panel in parallel
A second 150 W panel in 1S2P roughly doubles daytime energy without replacing the original panel or requiring a central battery. Add-on Distributor modules can absorb the extra useful power.
Add shared storage when refrigeration earns it
Rewire the same two panels to 2S1P, add a low-cost 24 V solar charger and a 25.6 V / 50 Ah LiFePO₄ battery. The battery bus can feed the original Distributor and an efficient DC refrigerator.
Why refrigeration matters. After lighting and air movement, refrigeration may be the next transformative household service. It is also a load large and continuous enough to justify moving from “battery in the appliance” to a shared household battery.
Where the project stands.
This is intentionally public before it is polished. The point of the site is to document what is being built, what has actually worked, and what remains only an attractive hypothesis.
Already demonstrated
- Low-cost USB DC-DC modules operating rechargeable appliance loads.
- A prototype charging an old cell phone back into service.
- ESP32-S2 MicroPython toolchain and 10 kHz PWM generation.
- Solar emulator schematic and PCB design completed; prototype PCBs ordered.
- Integrated 100 × 100 mm Solar Distributor V1 designed, electrically checked and sent for prototype fabrication.
Still to demonstrate
- Reliable automatic recovery of each appliance after repeated cloud-induced voltage collapse.
- Real landed BOM cost at meaningful production volume.
- Actual battery capacity and lifetime of very inexpensive rechargeable appliances.
- PAYGo repayment, service and warranty economics in real communities.
- A field installation model that remains simple at very large scale.
Important: This site documents research and prototype development. It is not currently soliciting customers, investments or donations, and no prices or performance figures should be treated as a commercial offer.