Introduction: Why S Parameters Matter in Wireless Power Transfer

Wireless power transfer (WPT) systems have moved from laboratory curiosities to everyday technologies, enabling cordless smartphone charging, medical implants that never need battery replacement, and dynamic charging of electric vehicles. Behind the seamless user experience lies a demanding radio-frequency (RF) engineering challenge: transferring energy efficiently across an air gap, often with variable coil alignment and distance. Engineers solve this challenge using scattering parameters—commonly called S parameters—a formalism originally developed for microwave network analysis that provides a complete, measurable description of how RF signals behave at every port of a network. Far from a niche academic exercise, understanding and applying S parameters is the practical foundation for designing, testing, and optimizing modern WPT links.

This article explores the role of S parameters in WPT, from fundamental definitions through advanced measurement techniques and real-world applications. We will show how a single set of S-parameter measurements—easily obtained with a vector network analyzer (VNA)—can reveal impedance matching, coupling coefficient, coil quality factor, and overall link efficiency. By the end, you will understand why S parameters are indispensable for anyone serious about wireless power system design.

Fundamentals of S Parameters: A Refresher

S parameters, introduced in the 1960s, describe the linear relationship between incident and reflected voltage waves at the ports of an RF network. Unlike impedance (Z) or admittance (Y) parameters, which require open-circuit or short-circuit terminations—impractical at high frequencies—S parameters are measured with ports terminated in a reference impedance, typically 50 Ω. This makes them uniquely suited for broadband characterization of microwave circuits, antennas, and coupled resonator systems.

For a two-port network, the scattering matrix is:

[ b1 ]   [ S11  S12 ] [ a1 ]
[ b2 ] = [ S21  S22 ] [ a2 ]

Here, a1 and a2 are the incident power waves at ports 1 and 2, while b1 and b2 are the reflected waves. The four S parameters carry intuitive physical meanings:

  • S11: Input reflection coefficient – the fraction of power reflected from port 1 when port 2 is terminated in the reference impedance. Ideally low (e.g., ≤ –10 dB) for efficient power delivery.
  • S21: Forward transmission coefficient – the fraction of power transmitted from port 1 to port 2. High |S21| indicates low insertion loss and strong coupling.
  • S12: Reverse transmission coefficient – typically equal to S21 for reciprocal passive networks, making it redundant in many analyses.
  • S22: Output reflection coefficient – describes the impedance match at port 2, guiding receiver-side matching network design.

Each S parameter is a complex number varying with frequency, allowing engineers to visualize how the network behaves across the operating band. For a rigorous introduction, refer to the comprehensive Wikipedia entry on scattering parameters.

Wireless Power Transfer as a Two‑Port Network

A typical WPT system—whether inductive, resonant, or capacitive—can be abstracted as a linear two‑port network. The transmitter coil (plus its driving source) is port 1; the receiver coil (with its load) is port 2. The electromagnetic coupling between the coils, along with parasitic capacitances and losses, defines the scattering matrix. This abstraction holds as long as the system operates in the linear regime, which covers most practical WPT applications except for extreme magnetic saturation.

From S11 and S21, engineers directly assess two fundamental performance metrics:

  • Power transfer capability: |S21|2 represents the power delivered to a matched load relative to the available source power. A high |S21| means little insertion loss through the coupler.
  • Impedance match quality: |S11|2 indicates mismatch loss. A low S11 (in dB, typically below –10 dB) ensures that most source power enters the network rather than being reflected back toward the amplifier.

However, raw |S21| does not directly equal system power transfer efficiency (PTE), because the actual WPT load—often a rectifier with nonlinear impedance—differs from the VNA's 50 Ω reference. The true PTE, defined as DC output power divided by DC input power, requires modeling the entire chain including rectifier and amplifier efficiencies. Still, S parameters provide the essential RF link characterization upon which all later stages depend. Formal methods to derive link efficiency from measured S parameters have been published; see, for example, this IEEE paper on WPT measurement techniques.

Impedance Matching and Maximum Power Transfer

The maximum power transfer theorem states that maximum power is delivered when the load impedance is the complex conjugate of the source impedance. In a WPT system, the effective source impedance seen from the receiver side is transformed by the coupled coils, and the load directly influences the transmitter's reflection coefficient. This bilateral interaction is elegantly captured by S parameters.

Engineers use S11 to design a matching network at the transmitter that cancels reactive components and transforms the impedance to the desired value (often 50 Ω). A well‑designed match yields S11 near zero, meaning nearly all power enters the coupler. Similarly, S22 guides the receiver-side matching network to ensure the rectifier presents the optimal load to the resonant tank. Achieving simultaneous conjugate matching—where both S11 and S22 are minimized—maximizes transducer power gain, a quantity directly calculable from the S matrix and terminations.

In practice, fixed matching networks only work at one coil separation and alignment. S parameters measured over a range of positions reveal how S11 drifts, causing detuning and efficiency collapse. Adaptive matching circuits, using varactors or switched capacitor banks, employ real‑time S-parameter monitoring to maintain optimal match as conditions change. This is a hot research area for dynamic WPT in electric vehicles and robotics.

Key S‑Parameter‑Derived Metrics for WPT

Beyond the raw magnitudes, several derived metrics from S‑parameter data guide design decisions:

  • Coupling coefficient k extraction: From the two‑port S matrix, the impedance matrix and then the coupling coefficient can be computed. A typical method uses S21 phase and magnitude along with known coil inductances. High k (close to 1) yields high S21 but may cause frequency splitting, requiring careful operating point selection.
  • Operating frequency optimization: When coils are overcoupled, |S21| versus frequency shows two peaks instead of one. The frequencies of these peaks (odd‑mode and even‑mode resonances) shift with coupling. By measuring S21 across a band, engineers choose the peak offering the best trade‑off between bandwidth and sensitivity to misalignment.
  • Unloaded Q factor estimation: S11 of a single coil in isolation (with the other coil removed) can be used to extract the coil's Q. This is vital for assessing the intrinsic efficiency limit of the link, as maximum achievable efficiency is a function of k and the Q factors of both coils.
  • Available gain and maximum stable gain: Using the S matrix, engineers compute the maximum available gain (MAG) and maximum stable gain (MSG) to understand the theoretical upper bound of power transfer under matched conditions.

Thus, a single set of S‑parameter measurements, skillfully interpreted, reveals all the essential RF characteristics of a WPT coupler.

Accurate measurement of S parameters is the starting point for systematic WPT design. A vector network analyzer sweeps the excitation across the desired frequency range and measures reflected and transmitted waves with high dynamic range. The typical setup connects the transmitter coil to port 1 and the receiver coil to port 2 via coaxial cables. However, this straightforward arrangement presents several pitfalls that must be addressed.

Calibration and De‑embedding

Calibration sets the measurement reference plane. A standard SOLT (Short‑Open‑Load‑Thru) calibration places the reference at the VNA cable ends. Yet in WPT testing, the actual device under test—the coils—often connects through test fixtures, baluns, or pigtails. To obtain the true S parameters of the coupler alone, the effects of cables and connectors must be mathematically removed. This process, called de‑embedding, uses measured S parameters of the fixture and port extensions. For highest accuracy, custom TRL (Thru‑Reflect‑Line) calibration standards can be fabricated directly on the coil PCB, moving the reference plane to the coil terminals. Using TRL eliminates errors from connector transitions and PCB trace parasitics, giving results within 0.05 dB of the true coil behavior.

Measuring in Realistic Environments

WPT systems never operate in anechoic chambers. Metal objects, a user's hand holding a phone, or the steel frame of a vehicle can detune the coils. Engineers reproduce these conditions during VNA measurements by placing representative materials near the coils and recording the S‑parameter shift. Such experiments reveal the coil's sensitivity to extrinsic perturbations and inform the design of compensation mechanisms. For consumer electronics, a hand phantom made from tissue-equivalent dielectric material is used to mimic the loading effect of a human palm. Measured S11 shifts of 5–10 dB are common when a hand is present, necessitating adaptive tuning algorithms that rely on real-time S-parameter feedback.

For dynamic applications, real‑time S‑parameter monitoring is possible using compact VNA‑on‑a‑chip solutions or impedance analyzers integrated into the WPT controller. By continuously probing S11 at the transmitter, the system detects load changes and adjusts matching in microseconds. This closed‑loop approach, detailed in application notes from test‑equipment leaders such as Keysight's guidance on impedance measurements, ensures robust power delivery even as a robot moves across a charging pad.

Large‑Signal and Load‑Pull Considerations

A significant challenge is that S parameters are small‑signal linear parameters, yet the WPT load—a diode rectifier—is strongly nonlinear. The impedance presented by the rectifier varies with input power level, so the S parameters measured at low power (typical VNA drive levels) may not represent the actual operating point. To overcome this, engineers use load‑pull techniques: the VNA is replaced with a tuner that emulates the nonlinear rectifier impedance at the desired power level. This yields effective S parameters at the large‑signal operating point, though the setup is more complex and time‑consuming. For many practical designs, a reasonable approximation is to measure S parameters at a power level close to the intended operating point, then iterate with harmonic balance simulations. Active load-pull systems, which use a second signal source to inject power back into the DUT, provide the most accurate representation of the rectifier's nonlinear behavior during S-parameter extraction.

Common Measurement Pitfalls and How to Avoid Them

Even with careful calibration, several common errors can corrupt S-parameter measurements of WPT coils:

  • Cable movement during calibration: Any flexing of cables after calibration introduces phase errors. Use phase-stable cables and secure all connections with torque wrenches.
  • Improper port extension: When adding adapters or PCB traces, incorrect port extension values shift the reference plane. Verify extension lengths using a time-domain reflectometry (TDR) feature on the VNA.
  • Coil proximity to conductive surfaces: The VNA's metal chassis and nearby cables can couple to the coil, altering S21 readings. Place the coils on a foam stand at least 30 cm from any large metal object.
  • Insufficient averaging: Low signal levels at deep nulls in S21 can be buried in noise. Increase VNA averaging to at least 16 sweeps for reliable measurements near resonant nulls.

Design and Optimization Workflow Using S Parameters

The use of S parameters in WPT extends from early simulation to manufacturing. A typical workflow involves:

  1. Electromagnetic simulation: Full‑wave solvers (e.g., Ansys HFSS, CST Studio) export S‑parameter Touchstone files for the coil geometry. These files are imported into RF circuit simulators (e.g., Keysight ADS, AWR) where matching networks, rectifiers, and source amplifiers are co‑designed. The Touchstone format (.s2p for two-port networks) preserves both magnitude and phase information across the frequency range.
  2. Prototype measurement: VNA measurements validate the simulation and fine‑tune component values. Discrepancies often arise from PCB parasitics not fully modeled; measured S parameters feed back into the simulation for refinement. Typical discrepancies include 5–10% shifts in resonant frequency due to capacitor tolerances and PCB dielectric variations.
  3. Link budget analysis: Using measured S21, the required transmit power to meet a given rectifier input threshold is calculated. System efficiency is projected by cascading the matched S‑parameter block with measured power‑added efficiency (PAE) of the amplifier and efficiency of the rectifier. The cascaded efficiency is the product of each individual efficiency, so a 2 dB insertion loss in the coupler (63% efficiency) directly cuts the overall system efficiency by the same factor.
  4. Compliance and safety verification: Regulations (FCC, ETSI) set limits on radiated emissions. While S parameters cannot directly quantify radiation, they help ensure the coils operate in a well‑defined near‑field mode with minimal unintended radiation. Combined with E‑field and H‑field probes, S‑parameter data verify that the coupling mode is predominantly magnetic and that no significant common‑mode currents exist on the coaxial cables.

An example optimization might target |S21| ≥ –3 dB (50% power transfer through the coupler) and S11 ≤ –15 dB at 6.78 MHz for an AirFuel resonant system. If measurements show |S21| = –5 dB and S11 = –8 dB, the engineer iterates on matching component values or changes the coil design to improve coupling, then re‑measures. Typically, 3–5 design iterations are needed to converge on the target specifications.

Practical Applications Across Industries

The S‑parameter methodology scales from microwatt implants to kilowatt vehicle chargers. Each application presents its own frequency band, coil design, and regulatory constraints, but the underlying RF characterization remains universal.

Electric Vehicle Wireless Charging

Systems following the SAE J2954 standard operate at 85 kHz, delivering up to 11 kW or more. The ground assembly (transmitter) and vehicle assembly (receiver) are modeled as a two‑port network. S parameters measured over a range of parking misalignments—up to ±150 mm in horizontal offset—help define interoperability requirements. Automakers use S‑parameter data to design vehicle‑side matching networks that work with charging pads from different manufacturers. Adaptive tuning based on live S11 readings compensates for varying air gaps (100–250 mm) and load conditions, maintaining efficiency above 90% across the alignment envelope. Recent SAE J2954 test procedures mandate S-parameter measurements at 10 discrete misalignment positions to verify compliance.

Consumer Electronics

Qi (WPC) chargers typically operate at 87–205 kHz or up to 13.56 MHz for resonant extensions. S parameters of the transmitter coil with a reference receiver, and vice versa, are specified in compliance testing. Manufacturers use VNA measurements to verify that |S21| at the operating frequency exceeds a threshold while S11 stays low, ensuring interoperability. During development, engineers measure S parameters of the coil assembly inside the phone under various hand placements, mimicking the effect of the human hand on detuning and adjusting the matching network accordingly. A typical hand detuning shifts the resonant frequency by 5–10%, which must be accommodated by the system's tracking bandwidth.

Medical Implants

Implantable neurostimulators, pacemakers, and cochlear implants rely on transcutaneous WPT through skin. The secondary coil is often embedded in biocompatible encapsulation, and tissue properties (high permittivity, conductivity) dramatically affect coil impedance. S‑parameter measurements using tissue‑simulating phantoms reveal the optimal operating frequency—often in the low MHz range—where |S21| is acceptable and tissue absorption is minimized. For a cochlear implant, careful impedance matching based on measured S parameters ensures that the required power (tens of milliwatts) is reliably delivered over several millimeters, even as the patient moves. Recent studies continue to refine these techniques; see this ScienceDirect article on implantable WPT efficiency.

Industrial and Drone Applications

In industrial automation, WPT powers rotating equipment, robots, and autonomous guided vehicles. S parameters measured across rotational misalignment angles reveal that coupling can drop by 50% when coils are not coaxial. Engineers use these data to design oversized transmitter coils that maintain S21 within acceptable limits even with mechanical tolerances. For drone charging pads, wind-induced misalignment can cause rapid S11 shifts, and real-time S-parameter monitoring is used to trigger frequency hopping between available ISM bands (6.78 MHz, 13.56 MHz, or 27.12 MHz) to maintain power transfer during descent.

Challenges and Emerging Solutions

While S parameters are indispensable, their application to WPT is not without hurdles. The very act of connecting a VNA to the coils alters the network because the VNA's 50 Ω terminations differ from the actual source and load impedances. We already discussed load‑pull techniques as a partial solution. Another challenge is frequency splitting in overcoupled regimes: the |S21| curve bifurcates into two peaks, and the optimal operating frequency shifts with coupling. Using S‑parameter data, adaptive frequency tuning algorithms search for the peak that maximizes efficiency while staying within the allowed ISM band. Real‑time spectral analysis of S21 through embedded mini‑VNAs is an active area of innovation, with commercial devices like the NanoVNA enabling low-cost implementation.

Multi‑coil systems (MIMO WPT) extend the S‑parameter concept to N‑port networks. Here, crosstalk terms like S31, S23 become critical, and full N×N matrices must be measured and managed. Advanced transmitters use S‑parameter‑driven calculations to steer magnetic fields toward the receiver, analogous to beamforming in wireless communications. For a 4-coil array, the full 4×4 S matrix contains 16 complex parameters, which can be measured in a single VNA sweep using a multiplexed setup. This data enables real-time field optimization, but the computational load for matrix inversion increases as O(N³), limiting the number of coils that can be actively controlled.

Furthermore, the growth of simultaneous wireless information and power transfer (SWIPT) demands S‑parameter design for dual‑purpose couplers that preserve both high power transfer and wide modulation bandwidth. The scattering matrix formalism extends naturally to these composite systems, ensuring that the fundamental physics of reflection and transmission continues to illuminate the path forward. In SWIPT, the isolation between the power and data channels can be quantified by S21 at different frequencies, and engineers aim for >20 dB isolation to prevent the high-power carrier from saturating the data receiver.

Future Directions

As WPT moves into higher powers, dynamic environments, and higher frequencies (for smaller coils), S parameters will remain a core analytical tool. The integration of high‑speed, low‑cost impedance analyzers directly into power management ICs will enable real‑time S‑parameter tracking without external instrumentation. Machine learning models, trained on thousands of measured scattering matrices, will predict optimal matching network states and operating frequencies instantly, even in the presence of unknown foreign objects. Recurrent neural networks (RNNs) have already demonstrated the ability to predict S11 shifts within 2% accuracy based on past impedance trends, enabling proactive matching adjustments before efficiency drops.

Moreover, the convergence of WPT and data communication will push the boundaries of dual‑purpose coupler design. The scattering matrix formalism will be essential for optimizing trade‑offs between power transfer efficiency and signal integrity. With these developments, S parameters will continue to be the language in which the RF behavior of wireless power couplers is described, analyzed, and perfected—from the initial benchtop VNA trace to sophisticated adaptive control loops. Understanding and applying S parameters enables engineers to push the boundaries of efficiency, reliability, and user experience in wireless power transfer systems across every scale. For those seeking to deepen their knowledge, the IEEE Power Electronics Society offers a dedicated tutorial track on RF measurement techniques for WPT at their annual conference, providing hands-on training with VNA calibration and S-parameter interpretation.