Maximum Power Transfer
Too small loses, too large loses
Slide the load resistance R_L and watch the power into the load. At zero there is no voltage across it; at infinity no current flows; either way the power is zero. Find the one point between them where the power is maximal, R_L = R_th.
You cannot have both voltage and current
Power is the product of the voltage across the load and the current through it. With a small load resistance the current is large but almost no voltage appears; with a large load resistance the voltage is high but almost no current flows. The two trade off against each other, so leaning either way shrinks the product. Power is greatest when they are in balance.
The balance is at R_L = R_th
Differentiate the power P = (V_th/(R_th+R_L))²·R_L with respect to R_L and set it to zero, and the maximum falls exactly at R_L = R_th. Setting the load equal to the source’s internal resistance is called matching. The voltage then splits evenly, putting V_th/2 across the load, and the maximum power is P_max = V_th²/(4R_th).
Maximum power is not maximum efficiency
At the match the load and internal resistances are equal, so the total power splits evenly between them. The load receives exactly as much as the internal resistance throws away as heat, so the efficiency is precisely 50%. Matching is the answer when you want to extract the most power, but where efficiency matters — like power transmission — the load is made deliberately large to cut losses. They are different goals.
Back to the first screen
As you slid the load resistance, the load power started at zero, climbed to a peak, and came back down, and that peak fell exactly at R_L = R_th, where the load equals the internal resistance. There the voltage split evenly and the efficiency was 50%. Once you fold a circuit into one source and one resistor with the Thévenin equivalent, what to attach for the most power becomes visible at a glance like this.