How Do Solar Panels Actually Work?

The photovoltaic effect

A solar cell is a wafer of silicon with a deliberate imbalance built into it. The top layer is doped with phosphorus, giving it spare electrons and a negative character. The bottom layer is doped with boron, leaving it short of electrons and positive. Where the two layers meet, a small permanent electric field forms across the boundary — the p-n junction.

That field is the whole trick. When a photon of sufficient energy strikes the silicon, it knocks an electron loose from its atom. Without the junction, that electron would wander briefly and settle back. With the junction present, the field sweeps it in one consistent direction. Connect a wire between the two faces of the cell and those displaced electrons have somewhere to go. That directed flow is direct current.

Two consequences follow from this, and both matter more than most buying guides admit. First, the cell responds to light, not heat — which is why a panel in cold bright weather outperforms the same panel in hot hazy weather. Second, only photons above a threshold energy can free an electron at all. Silicon's threshold sits at about 1.1 electronvolts, so a large slice of the incoming spectrum, particularly infrared, passes through doing nothing useful. That single physical limit is the main reason commercial panels sit near 20% efficiency rather than 80%.

From one cell to a panel

A single silicon cell produces about 0.5 to 0.6 volts regardless of its physical size. Size determines current, not voltage. Half a volt powers nothing useful, so cells are wired in series to add their voltages together — sixty cells in series gives roughly 30 to 40 volts, which is a workable figure.

Those series-wired cells are laminated between a sheet of tempered low-iron glass on top and a polymer backsheet beneath, then framed in aluminium. The visible thin silver lines on a cell's face are busbars and fingers, the printed conductors that collect freed electrons and carry them to the cell's edge. They are kept deliberately narrow because every square millimetre of metal is a square millimetre not collecting light.

The series wiring has an important weakness. Because the cells form a chain, current through the whole string is limited by its weakest member. A single cell in shadow does not reduce output by its own share alone — it throttles the entire string. This is why partial shading is disproportionately damaging, and why bypass diodes are built into the junction box to route current around a shaded section. It is also the central argument for microinverters on complex roofs.

Turning DC into household AC

Panels produce direct current. Household wiring and appliances run on alternating current, at 230 V and 50 Hz in most of the world or 120 V and 60 Hz in North America. Bridging that gap is the inverter's job, and it does two things at once.

The obvious task is switching the DC into a stepped waveform that approximates a sine wave closely enough to be safe for appliances and acceptable to the grid. The less obvious task matters more for your yield: maximum power point tracking. A panel's voltage and current trade off against each other along a curve, and there is exactly one point on that curve where their product — the power — is highest. That point drifts continuously as light and temperature change. An MPPT controller hunts for it many times per second. A cheaper PWM controller does not, and simply loses the difference.

The three loss stages

Follow 1,000 watts of sunlight landing on one square metre of roof and you can see where it ends up.

Conversion is the big one. A 20%-efficient panel turns that 1,000 W into about 200 W of DC. The other 800 W is not being wasted through poor engineering; most of it is spectrum the silicon physically cannot use, and it leaves as heat. No commercial silicon panel escapes this, which is why efficiency arguments between a 20% and a 22% panel are arguments about the small remainder.

Temperature is the one people underestimate. Panels are rated at 25 °C cell temperature, but a dark panel in still air on a sunny roof commonly reaches 45 to 65 °C. At −0.35%/°C and a cell temperature of 55 °C, you lose about 10.5% against the rating. Mounting panels with an air gap beneath them, rather than flush to the roof surface, recovers a meaningful part of this.

System losses accumulate quietly. Cable resistance, connector losses, inverter conversion, dust, and mismatch between panels that are nominally identical each take a small percentage. Together they typically account for 8 to 15%.

Multiply those stages and a 5 kWp array on a hot clear day peaks somewhere near 4.0 to 4.2 kW rather than 5.0 kW. That is normal and expected, not a fault.

Why a 400 W panel rarely makes 400 W

The wattage printed on a panel is measured at Standard Test Conditions, and STC is a laboratory specification: 1,000 W/m² irradiance, 25 °C cell temperature, and an air mass of 1.5. Its purpose is comparability between manufacturers, not prediction of field output. A roof almost never satisfies all three conditions simultaneously — when irradiance is high enough to hit 1,000 W/m², the cell is usually far hotter than 25 °C.

This is why the more useful number on a datasheet is NOCT — Nominal Operating Cell Temperature, measured at 800 W/m² and 20 °C ambient with wind. NOCT output typically runs 20 to 25% below the STC figure and sits much closer to what you will actually see. When comparing two panels, compare their NOCT ratings and their temperature coefficients, not just the headline watts.

Rated wattage tells you what a panel did in a laboratory. Temperature coefficient and NOCT tell you what it will do on your roof in July.

Common questions

Do solar panels work on cloudy days?

Yes, at reduced output. Panels use diffuse light as well as direct beam, so an overcast day typically yields 10 to 30% of clear-sky output depending on cloud thickness. They do not stop entirely.

Do solar panels work in cold weather?

Cold helps. Output tracks available light, and lower cell temperature raises efficiency, so a cold bright winter day can outproduce a hot hazy summer one. Snow covering the glass stops production, but panels shed snow reasonably well once any part of the surface clears and warms.

How long do solar panels last?

Panels degrade rather than fail outright. Typical warranties guarantee 80 to 87% of rated output at year 25, implying roughly 0.4 to 0.6% loss per year. Inverters are the shorter-lived component, usually needing replacement somewhere between years 10 and 15 — a cost worth including in any payback calculation.

Can I run my house directly from panels without batteries?

During daylight, largely yes, through a grid-tied inverter that blends solar with grid supply. Without batteries you export surplus during the day and import at night. Note that standard grid-tied inverters shut down during an outage for lineworker safety, so solar alone does not provide backup power — that needs a hybrid inverter and storage.