Math & Units

Conversion (Z to S)

Q: What is the matrix formula to convert a Z matrix to an S matrix?

For an N-port network, S = (Z - Z0*I)(Z + Z0*I)^-1 when all ports share the same real reference impedance Z0 (for example 50 ohms) and I is the identity matrix. More generally, with a diagonal reference-impedance matrix and its square root, S = F (Z - Z_ref*) (Z + Z_ref)^-1 F^-1, where F normalizes each port by the square root of its reference resistance. For a single-port load this collapses to the familiar reflection coefficient Gamma = (Z - Z0)/(Z + Z0). The conversion requires that (Z + Z0*I) be invertible, which holds for any physically realizable passive network at a given frequency.

Q: Why convert Z-parameters to S-parameters at microwave frequencies?

Z-parameters are defined using open-circuit terminations, which are difficult to realize at RF and microwave frequencies because an open port has parasitic capacitance and radiates, making the measurement non-ideal above roughly 100 MHz. S-parameters are instead defined with matched terminations equal to the reference impedance, which a vector network analyzer can present accurately to over 110 GHz. Converting a Z matrix from a circuit simulator or a low-frequency measurement into S-parameters lets engineers compare it directly against VNA data, cascade it in a system budget, and check stability and matching on a Smith chart.

Q: Does the conversion result change if I pick a different reference impedance?

Yes. The Z matrix itself is independent of reference impedance because it describes the network in terms of port voltages and currents, but the resulting S matrix is defined relative to a chosen Z0. The same physical two-port renormalized from 50 ohms to 75 ohms produces a different S11 magnitude and phase even though the network is unchanged. Always state the reference impedance alongside any S-parameter set; mixing 50-ohm and 75-ohm S-parameters in a cascade without renormalization is a common source of error in system-level analysis.

/kuhn-VUR-zhuhn zee too ess/

Transforming a network's impedance matrix into S-parameters recasts open-circuit voltage and current descriptions into the traveling-wave reflection and transmission coefficients that vector network analyzers actually measure. For an N-port the relationship is S = (Z − Z₀I)(Z + Z₀I)⁻¹ when every port is normalized to a common real reference impedance (usually 50 Ω). Because the impedance description uses ideal open terminations that are impractical above roughly 100 MHz, this conversion is the standard bridge between a circuit simulator's Z matrix and laboratory S-parameter data, and it underlies cascading, matching, and stability analysis up to millimeter-wave bands.

Category: Math & Units

Reference Z₀: 50 Ω typical

Validity: Any passive linear N-port

From Open-Circuit Impedances to Traveling Waves

An impedance matrix Z relates the total voltages at a network's ports to the currents flowing into them, with each element Z_ij defined while every port except port j is left open. That open-circuit definition is convenient for lumped circuit theory but breaks down at microwave frequencies, where a true open is impossible to realize: fringing capacitance, lead inductance, and radiation all corrupt the termination. S-parameters sidestep the problem by defining each port relative to a matched reference impedance and measuring incident and reflected traveling waves. The Z-to-S conversion is the deterministic algebra that links these two viewpoints, letting a model expressed in impedance terms be checked against measured scattering data.

The conversion is purely a change of basis on the same physical network, so it loses no information and is exactly reversible through the companion S-to-Z transform. Reciprocity (Z₁₂ = Z₂₁) carries over to S₁₂ = S₂₁, and a lossless network keeps its unitary S matrix. The one practical caveat is reference impedance: the Z matrix is independent of Z₀, but the resulting S matrix is defined relative to whatever value you choose, so a 50 Ω result differs numerically from a 75 Ω result for the identical hardware.

The Conversion Equations

General N-port (common real Z₀):
S = (Z − Z₀I) × (Z + Z₀I)⁻¹

Two-port closed form (denominator Δ):
Δ = (Z₁₁ + Z₀)(Z₂₂ + Z₀) − Z₁₂Z₂₁
S₁₁ = [(Z₁₁ − Z₀)(Z₂₂ + Z₀) − Z₁₂Z₂₁] / Δ
S₂₁ = (2 × Z₂₁ × Z₀) / Δ
S₁₂ = (2 × Z₁₂ × Z₀) / Δ
S₂₂ = [(Z₁₁ + Z₀)(Z₂₂ − Z₀) − Z₁₂Z₂₁] / Δ

Single-port limit: Γ = (Z − Z₀) / (Z + Z₀)

Z₀ = reference impedance, I = identity matrix, Δ = two-port determinant term. Example: a single shunt 50 Ω resistor across the line has Z₁₁=Z₁₂=Z₂₁=Z₂₂=50 Ω, which the formula maps to Δ=(100)(100)−2500=7500, S₁₁=(50×100−2500)/7500=−0.333 and S₂₁=(2×50×50)/7500=0.667 (−3.5 dB). A pure series element has no finite Z matrix, so its S-parameters must be built directly rather than converted.

Worked 50 Ω Reference Cases

Network	Z-parameter inputs (Ω)	Resulting S₁₁	Resulting S₂₁	Note
Shunt 50 Ω resistor	Z₁₁=Z₂₂=Z₁₂=Z₂₁=50	−0.333 (−9.5 dB)	0.667 (−3.5 dB)	Lossy, partially reflecting
Shunt 25 Ω resistor	Z₁₁=Z₂₂=Z₁₂=Z₂₁=25	−0.5 (−6 dB)	0.5 (−6 dB)	Heavier shunt loading
Matched 6 dB T-pad	Z₁₁=Z₂₂=83.5, Z₁₂=Z₂₁=66.9	≈0 (matched)	0.501 (−6 dB)	Symmetric attenuator
Matched 10 dB T-pad	Z₁₁=Z₂₂=61.1, Z₁₂=Z₂₁=35.1	≈0 (matched)	0.316 (−10 dB)	Symmetric attenuator
Reactive (lossless)	Pure imaginary Z terms	\|S₁₁\|≤1	\|S₁₁\|²+\|S₂₁\|²=1	Unitary, power-conserving S matrix

Common Questions

Frequently Asked Questions

What is the matrix formula to convert a Z matrix to an S matrix?

For an N-port with a common real reference impedance Z₀, S = (Z − Z₀I)(Z + Z₀I)⁻¹, where I is the identity matrix. With unequal port impedances you normalize each port by the square root of its reference resistance before applying the formula. The single-port limit reduces to the reflection coefficient Γ = (Z − Z₀)/(Z + Z₀). The conversion is valid whenever (Z + Z₀I) is invertible, which holds for any realizable passive network at a given frequency.

Why convert Z-parameters to S-parameters at microwave frequencies?

Z-parameters rely on open-circuit terminations that are hard to realize above roughly 100 MHz because an open port has fringing capacitance and radiates. S-parameters use matched terminations a vector network analyzer can present accurately past 110 GHz. Converting a simulator's Z matrix into S-parameters lets you compare against measured VNA data, cascade blocks in a link budget, and assess matching and stability on a Smith chart.

Does the conversion result change if I pick a different reference impedance?

Yes. The Z matrix is independent of Z₀, but the S matrix is defined relative to the chosen reference. The same two-port renormalized from 50 Ω to 75 Ω yields a different S₁₁ magnitude and phase even though the hardware is identical. Always record the reference impedance with any S-parameter set; mixing 50 Ω and 75 Ω data in a cascade without renormalizing is a frequent system-analysis error.

Related Terms

← Conversion (S to Z) Converter EMI →