// RF Theory › Impedance Matching

Impedance Matching

A practical, example-driven guide to impedance matching — from the fundamentals of reflection to quarter-wave transformers, single stub and double stub matching — with fully verified calculations and Smith chart walkthroughs at every step.

// 01 — Fundamentals

Why Match Impedances?

In any RF system a signal travels from a source through a transmission line to a load. When the load impedance ZL differs from the characteristic impedance Z₀ of the line, part of the incident wave reflects back toward the source. That reflected energy is wasted, and in high-power systems it can destroy amplifiers.

Impedance matching eliminates the reflection so that all available power is delivered to the load — maximising efficiency, minimising VSWR, and ensuring the source operates into its design impedance.

Key insight: Maximum power transfer occurs when ZL = ZS* (complex conjugate of the source). For a purely real source ZS = Z₀ this simplifies to ZL = Z₀.

Where matching matters most

LNA input: Maximises power from antenna and minimises noise figure.
PA output: Delivers maximum power to antenna and protects the PA.
Filter terminations: Filter response is valid only with correct source and load impedance.
Between stages: Cascaded stages need matched interfaces to achieve predicted gain.

VSWR / Return Loss Calculator → Γ Reflection Calculator →
// 02 — Reflection Basics

Reflection Coefficient & VSWR

Reflection Coefficient
Γ = (ZL−Z₀)/(ZL+Z₀)  ·  Return Loss = −20·log₁₀|Γ| dB
VSWR = (1+|Γ|)/(1−|Γ|)  ·  Mismatch Loss = −10·log₁₀(1−|Γ|²) dB
Quick Reference Table — verified
|Γ|Return LossVSWRMismatch Loss
0.000∞ dB1.000 : 10.000 dB
0.10020.00 dB1.222 : 10.044 dB
0.20013.98 dB1.500 : 10.177 dB
0.31610.00 dB1.925 : 10.458 dB
0.5006.021 dB3.000 : 11.249 dB

|Γ|=0.1 → RL=20.00 dB · VSWR=1.222 · ML=0.044 dB ✓  ·  |Γ|=0.316 → RL=10.00 dB · VSWR=1.924 · ML=0.458 dB ✓

Compute VSWR & Return Loss → Plot S11 from Touchstone file →
// 03 — Smith Chart

Smith Chart Basics

The Smith chart maps the complex impedance plane onto the complex Γ plane. Every point corresponds to a unique normalised impedance z = Z/Z₀ = r + jx.

Constant-r circles — r=0 is outer rim, r=1 passes through centre, r=∞ is rightmost (OC).
Constant-x arcs — upper half = inductive (+jx), lower half = capacitive (−jx).
Centre — z=1+j0, Γ=0, perfect match. Leftmost — z=0 (SC).

Moving on the Smith chart: A lossless line of length ℓ rotates the load point clockwise by 2βℓ radians around a constant-|Γ| circle. One full revolution = λ/2 — not λ. Adding a shunt susceptance moves along a constant-g circle.
Interactive Smith Chart Tool →
// 04 — Quarter-Wave Transformer

Quarter-Wave Transformer

A λ/4 section of transmission line with characteristic impedance ZT inserted between Z₀ and a real load RL. The input impedance of a λ/4 line terminated by RL is ZT²/RL. Setting this equal to Z₀ gives:

Quarter-Wave Transformer
ZT = √(Z₀·RL)  ·  Zin = ZT²/RL = Z₀ (at design frequency)
Valid only for purely real loads. For complex loads: use a line or stub to rotate load to real first.

Worked Example — 100 Ω to 50 Ω at 2.4 GHz on FR4

Example 1 — QWT on FR4 microstrip
1
ZT = √(50×100) = 70.71 Ω
2
Verify: Zin = ZT²/RL = 5000/100 = 50.0 Ω ✓
3
Microstrip synthesis: W/h≈1.14 → W≈1.82 mm, εeff≈3.06. λ = 300/(2.4×√3.06) = 71.44 mm. λ/4 = 17.86 mm
4
Smith chart: zL=2+j0 (r=2 circle). λ/4 rotation → Zin=ZT²/RL=50 Ω — load moves to chart centre ✓
✓ 70.71 Ω, 17.86 mm microstrip section matches 100 Ω load to 50 Ω at 2.4 GHz.
Microstrip Width & Length → Stripline Calculator → Verify on Smith Chart →
// 05 — Single Stub Matching

Single Stub Matching

A single shunt stub connected in parallel with the main line at distance d from the load. Works entirely in admittance (Y=1/Z). The stub provides a susceptance that cancels the imaginary part of the line admittance at the junction.

Single Stub — Shunt Procedure
1. Normalise: yL = Z₀/ZL
2. Find d: y(d) = 1+jb (g=1 circle on Smith chart)
3. Stub: SC: βℓ = arccot(−b)   OC: βℓ = arctan(−b)
4. Total: (1+jb) + (−jb) = 1+j0 → matched ✓

Worked Example — Match 25−j30 Ω to 50 Ω at 1.0 GHz

Example 2 — Single shunt stub (Vf=0.66, λ=198 mm)
1
zL = (25−j30)/50 = 0.500−j0.600
2
|Γ|: |num|²=0.610, |den|²=2.610, |Γ|=√(0.610/2.610) = 0.4834
3
yL = (0.500+j0.600)/0.610 = 0.8197+j0.9836
4
Rotate to g=1: y = 1.000+j1.248 at d = 0.089λ = 17.6 mm
5
SC stub: cot(βℓ)=1.248 → βℓ=arctan(0.8013)=38.74° = 0.1076λ = 21.3 mm
6
Verify: ytotal = (1+j1.248)+(0−j1.248) = 1+j0 → Z=50 Ω ✓
✓ d=17.6 mm, SC stub ℓ=21.3 mm. Solution 2: d=81.4 mm, ℓ=77.7 mm.

Smith Chart Walkthrough

A
Plot zL=0.500−j0.600: r=0.5 circle, below real axis. |Γ|=0.4834 — your constant-|Γ| circle.
B
Flip to admittance: Rotate 180° → yL=0.820+j0.984. Now working in admittance coordinates.
C
Rotate to g=1 circle: Clockwise along |Γ|=0.4834 circle until g=1. First hit: y=1+j1.248 at d=0.089λ.
D
SC stub length: From SC point (y=∞), rotate clockwise to Im=−j1.248. Read ℓ=0.1076λ=21.3 mm.
E
Verify: (1+j1.248)+(0−j1.248) = 1+j0 — chart centre. Match complete.
F
Two solutions: g=1 intersected twice. Solution 2: d=0.411λ, ℓ=0.392λ. Shorter d preferred — lower loss.
Interactive Smith Chart → Microstrip Line Length → Check VSWR after matching →
// 06 — Double Stub Matching

Double Stub Matching

Two shunt stubs at fixed spacing d (typically λ/8), with adjustable stub lengths. Ideal for tunable bench tuners. The cost is a forbidden load region: loads with gL > 1/sin²(βd) cannot be matched.

Double Stub — Key Conditions
Forbidden region: gL > 1/sin²(βd)  ·  d=λ/8 → forbidden if gL > 2.0
yB = (yA+j·tan(βd))/(1+j·yA·tan(βd)) must have Re{yB}=1

Worked Example — Match 25−j50 Ω at 2.0 GHz, d=λ/8

Example 3 — Double stub tuner, Vf=1.0 (air), λ=150 mm
1
yL=(0.500+j1.000)/1.250 = 0.400+j0.800. gL=0.400 < 2.0 → matchable ✓
2
Set Re{yB}=1 for d=λ/8 (tan=1): 0.800/[(1−BA)²+0.16]=1 → (1−BA)²=0.640 → BA=0.200 or 1.800
3
Sol 1: b1=0.200−0.800=−0.600 → SC stub βℓ=arctan(1.667)=59.04°=0.164λ=24.6 mm
4
yA=0.400+j0.200. yB=(0.400+j1.200)/(0.800+j0.400)=(0.800+j0.800)/0.800=1.000+j1.000 ✓
5
Stub 2: cancel b=+1.000. SC: βℓ=45°=0.125λ=18.75 mm
✓ Stub 1=24.6 mm, Stub 2=18.75 mm, spacing=18.75 mm. ytotal=(1+j1)+(0−j1)=1+j0 ✓

Smith Chart Walkthrough — Double Stub

A
Plot yL=0.400+j0.800 on admittance chart (g=0.4 circle, upper half).
B
Draw rotated g=1 circle: Rotate entire g=1 circle counter-clockwise by λ/8 (90°) — the locus of yA values that arrive on g=1 after λ/8.
C
Intersect at g=0.4: Vertical line at g=0.4 intersects rotated circle at yA1=0.4+j0.2 and yA2=0.4+j1.8.
D
Stub 1: b1=Im{yA}−Im{yL}=0.2−0.8=−0.6 → SC ℓ=0.164λ.
E
Rotate yA forward λ/8: yA1=0.4+j0.2 rotated 90° CW → yB=1.0+j1.0 confirmed on g=1 ✓
F
Stub 2 cancels b=+1.0: SC ℓ=0.125λ. Total: 1+j1+0−j1=1+j0 ✓
Forbidden region fix: If gL > 1/sin²(βd), add a short line section before Stub 1 to move the load out of the forbidden region, or change stub spacing d.
Smith Chart Tool → Microstrip Calculator → CPW Calculator →
// 07 — Summary

Method Comparison

MethodLoad TypeBandwidthTunable?Main Constraint
Quarter-Wave TransformerReal onlyModerateNoNeeds real load — pre-transform complex loads
Single Stub — ShuntAny complexNarrow (~10–20%)PartiallyStub position d set at design time
Single Stub — SeriesAny complexNarrowNoRequires line break — harder on microstrip
Double StubAny (except forbidden)NarrowYesForbidden region gL > 1/sin²(βd)
L-Network (LC)Any complexNarrowWith variable CComponent Q degrades at mmWave
PCB best practice: Use a single OC shunt stub (no via needed). Calculate d and ℓ using the admittance Smith chart, convert to physical dimensions with the Microstrip Calculator, verify S11 with the S-Param Plotter.
// 08 — Try the Tools

Put This Theory Into Practice

Use these RFLab tools to implement and verify the techniques covered on this page — from computing the stub length to checking S11 after matching.

Matching & Transmission Line Calculators
📡
S-PARAMS
Smith Chart Plotter
The graphical tool for all matching on this page. Enter zL=0.5−j0.6 from Example 2, flip to admittance, rotate 0.089λ to g=1 circle, add stub susceptance and verify y=1+j0 at centre. Upload a .s2p file to plot real S11 traces.
 ⚡
CALCULATOR
Microstrip Calculator
Convert stub electrical lengths to physical dimensions. Example 2 gives d=0.089λ and ℓ=0.107λ — enter your f, Z₀ and substrate to get physical mm. Also finds W for the 70.71 Ω QWT section in Example 1 (gives W≈1.82 mm on FR4).
 ⚡
CALCULATOR
Stripline Calculator
For multilayer PCB matching networks where the stub is a buried stripline. εeff=εr for stripline — shorter physical length than microstrip for the same electrical length. Useful for inner-layer matching in dense PCB designs.
 ⚡
CALCULATOR
CPW Calculator
CPW stubs don't require vias — the ground plane is on the same surface. Use for mmWave matching networks where via inductance (~0.66 nH per via) would degrade stub accuracy. Lower dispersion than microstrip above 20 GHz.
 ⚡
CALCULATOR
Γ Reflection Calculator
Enter ZL=25−j30 Ω, Z₀=50 Ω → get |Γ|=0.484, ∠Γ=−148°. This Γ value is the starting point on the Smith chart for Example 2. Instantly shows return loss and mismatch loss before you start the matching design.
 ⚡
CALCULATOR
VSWR / Return Loss
After matching, S11 should be <−20 dB (VSWR <1.22). Enter your measured S11 in dB to confirm the stub match worked. Target for a well-matched LNA input: S11 <−15 dB across the band.
S-Parameter Verification Tools
📡
S-PARAMS
S-Parameter Plotter
Export a .s2p of your matching network from simulation and upload here. Plot S11 vs frequency — a good match shows S11 <−20 dB at the design frequency with bandwidth set by the stub length. The ZIF in the stub response tells you where re-reflection occurs.
 📡
S-PARAMS
Cascade S-Parameters
Cascade the matching network .s2p with your LNA .s2p to see the matched LNA response. A well-designed input matching network flattens S11 of the LNA across the band — verify this is actually achieved, not just at one frequency.
 📡
S-PARAMS
Amplifier Stability
Check LNA stability before and after adding a matching network. A reactive stub at the input changes the impedance the LNA sees at out-of-band frequencies — which can trigger oscillation even if the design frequency is correctly matched.
 📡
S-PARAMS
De-embedding Tool
Remove test fixture parasitics from measurements of your matching network. SMA connectors and PCB launches add excess S11 ripple — de-embedding isolates the true matching stub performance from the measurement setup.
Related Theory Pages
📘
THEORY
Transmission Line Theory
The theoretical foundation for everything on this page. Zin(ℓ) formula, SC/OC stub reactance, and Smith chart rotation all come from transmission line theory. Read this first if the stub formulas here need derivation.
 📘
THEORY
Noise in RF Systems
The goal of LNA input matching is Γopt (minimum noise match) — not the conjugate match ΓMS for maximum gain. These two points differ on the Smith chart. The Noise page explains the trade-off and how to find the optimum noise operating point.
 📘
THEORY
RF Filter Theory
Filters only work correctly when terminated in their design impedance. When a filter must interface with a non-50 Ω source or load, a QWT or L-network matching network provides the interface — the complete design flow is on the Filter Theory page.
// Suggested workflow — match a 25−j30 Ω antenna to 50 Ω LNA at 1.0 GHz
1
Γ Calculator → ZL=25−j30 Ω, Z₀=50 Ω → |Γ|=0.484, ∠Γ=−148° — your Smith chart starting point
2
Smith Chart → plot zL=0.5−j0.6, flip to yL=0.82+j0.98, rotate 0.089λ to y=1+j1.248 on g=1 circle
3
Microstrip Calculator → at 1 GHz on FR4 (εeff=3.22, λ=69.6 mm) → d=6.2 mm, SC stub ℓ=7.5 mm
4
Simulate layout, export .s2p → S-Param Plotter → confirm S11 <−20 dB at 1.0 GHz
5
VSWR Calculator → enter S11 in dB → confirm VSWR <1.22 — match verified ✓