Recipe

Size a solar panel and battery for a small node

Use when: a duty-cycled device with a known average current should run on solar power through the whole year

Goal: a panel rating and a battery capacity, each with the assumptions that produced them

Taught in Solar and energy harvesting

1. 1

   Multiply the average current by the circuit voltage and 24 hours, then divide by the path efficiency.

   Example

   3.3\ \text{V} \times 3.25\ \text{mA} \times 24\ \text{h} = 257 mWh; at 70% efficiency, 0.257 / 0.7 \approx 0.37 Wh/day.

   Why: panels and batteries are rated in energy, and losses are paid on every Wh the device uses.

2. 2

   Use solar data for the site, month by month, and take the lowest.

   Example

   Example site: 1.5 peak sun hours in December.

   Why: the device must keep running in the darkest month, not on average.

3. 3

   Panel watts \ge need / (h_{sun} \times k), then round up to a real panel with margin.

   Example

   0.37 / (1.5 \times 0.6) \approx 0.41 W, so choose a 1 W panel.

   Why: the derating covers losses you can't avoid; the margin covers the assumptions you might have got wrong.

4. 4

   Usable battery energy \ge need × days of autonomy; divide by the usable fraction and the cell voltage for mAh.

   Example

   5 days: 5 \times 0.37 = 1.85 Wh; usable 80%: 2.3 Wh; at 3.7 V, about 620 mAh, so choose 1000 mAh or more.

   Why: a run of cloudy days is when a solar node fails if the battery is small.

5. 5

   Pick a charger IC for the cell's chemistry whose input range covers the panel, and check the cell's charging temperature range against the site.

   • the site freezes — make sure charging stops below the cell's minimum, for example with a thermistor input.
   • it doesn't — still use the controller; a panel is never wired straight to a cell.

   Why: the panel's voltage swings with light, and lithium cells must not be charged outside their temperature range.