Part 1 established the $k \cdot Q$ rule; Part 2 covered distance and frequency. This part is about circuit architecture: how the compensation network shapes what the link can deliver, and why system efficiency is a chain that no single stage can rescue.

5. Compensation topologies: the four basic configurations

TopologyTX sideRX sideOutput characterBest for
SSSeriesSeriesCurrent-source typeConstant-current drive, LEDs
SPSeriesParallelVoltage-source typeGeneral-purpose power
PSParallelSeriesHigh input impedanceHigh voltage, low power
PPParallelParallelHigh gain, self-bootstrapping at light loadWeak coupling, long range

In long-range, weakly coupled scenarios ($k < 0.1$), a parallel-compensated receiver uses resonance to amplify the induced voltage by $Q$ — making PP the default choice for cross-material and cross-distance applications. Paired with a regulator stage that has enough voltage headroom, it tames both extreme operating points at once: no-load overvoltage and full-load droop.

6. The end-to-end efficiency chain

System efficiency is a chained product — no single stage can compensate for the rest:

$$\eta_{total} = \eta_{drv} \times \eta_{tx} \times \eta_{space} \times \eta_{rx} \times \eta_{rect} \times \eta_{reg}$$

StageTypicalDominant loss mechanism
Driver $\eta_{drv}$90–96%Switching loss, gate drive
TX resonant cavity $\eta_{tx}$92–97%Coil ESR, capacitor ESR
Space link $\eta_{space}$30–85%Set by $k \cdot Q$ (Part 1)
RX resonant cavity $\eta_{rx}$92–97%Same as TX
Rectifier $\eta_{rect}$90–95%Schottky forward drop
Regulator $\eta_{reg}$85–95%LDO dropout / DC-DC

The space link is almost always the weakest stage — which is why, across this series, coil geometry and $Q$ get more engineering attention than any circuit trick.

The Wireless Power Transfer Engineering Series — Part 1 · Part 2 · Part 3 · Part 4