Antenna Impedance Matching
Theory and Physics
Overview — Why Matching is Necessary
Professor, how exactly do you perform antenna impedance matching? Why is it even necessary in the first place?
Simply put, it's the operation of canceling the reactance component at the feed point and bringing the resistive part close to $R=50\,\Omega$, because the antenna's input impedance changes with frequency. If this is misaligned, power is reflected, which can degrade transmission efficiency or even damage the transmitter.
What? It can break? That's scary...
Yes. For example, if a 50W PA (Power Amplifier) is used in a radio, when the VSWR exceeds 3, the reflected power exceeds 5W. This directly hits the transistor in the PA stage and can cause thermal breakdown. That's why transmitters have protection circuits that automatically reduce output when VSWR is high.
I see... What determines the antenna's impedance in the first place?
The antenna's input impedance $Z_{in}$ is composed of three parts: radiation resistance $R_{rad}$ (corresponding to energy radiated as radio waves), loss resistance $R_{loss}$ (conductor loss, dielectric loss), and input reactance $X_{in}$ (energy storage):
For example, a half-wave dipole has $Z_{in} \approx 73 + j42.5\,\Omega$ (free space), which results in a slight mismatch when directly connected to a 50Ω system. For a shortened dipole ($\ell < \lambda/2$), $R_{rad}$ becomes drastically smaller and $X_{in}$ becomes a large capacitive reactance, making it unusable without a matching circuit.
Reflection Coefficient and VSWR
VSWR, return loss... various numbers come up and I get confused...
They are all connected, so let's organize them. First, the reflection coefficient $\Gamma$ is:
$Z_0$ is the transmission line's characteristic impedance (usually 50Ω). $\Gamma$ is a complex number, and its magnitude $|\Gamma|$ indicates the degree of reflection. Perfect match is $\Gamma = 0$, perfect reflection is $|\Gamma| = 1$.
From this, all three metrics are derived:
What are the typical target values for each?
| VSWR | $|\Gamma|$ | RL (dB) | Reflected Power | Judgment |
|---|---|---|---|---|
| 1.0 | 0.00 | $\infty$ | 0% | Perfect Match |
| 1.2 | 0.09 | 20.8 | 0.8% | Excellent |
| 1.5 | 0.20 | 14.0 | 4.0% | Good |
| 2.0 | 0.33 | 9.5 | 11.1% | Acceptable Limit |
| 3.0 | 0.50 | 6.0 | 25.0% | Needs Improvement |
In practice, VSWR $\leq$ 2 (RL $\geq$ 10 dB) is a common design target, and mass-produced antennas are often required to have VSWR ≤ 1.5.
How to Read a Smith Chart
The Smith chart is that round diagram, right? I have no idea what's drawn on it...
The Smith chart is a mapping of complex impedance onto a "round map." It's a transformation that maps the normalized impedance $z = Z/Z_0 = r + jx$ onto the reflection coefficient plane:
With this transformation, lines of constant resistance become "circles shifted to the right," and lines of constant reactance become "circular arcs curving up and down." The chart's center is $z = 1$ (50Ω in a 50Ω system), the perfect match point. The rightmost point is open ($z = \infty$), the leftmost point is short ($z = 0$).
How does adding a matching circuit move the point on the chart?
This is the essence of the Smith chart:
- Add a series L → Move upward (inductive) along a constant resistance circle
- Add a series C → Move downward (capacitive) along a constant resistance circle
- Add a parallel L → Rotate downward along a constant conductance circle
- Add a parallel C → Rotate upward along a constant conductance circle
- Add a transmission line → Rotate clockwise around the center (corresponding to electrical length $\beta \ell$)
So matching design is a puzzle of "creating a path from a point on the chart to the center using L, C, and transmission lines."
Bandwidth and Q Factor Trade-off
Even if matched, doesn't it become unusable if the frequency shifts slightly?
Good point. There is a fundamental trade-off between the matching circuit's Q factor (Quality Factor) and bandwidth BW:
Here, $\text{BW}$ is the frequency range satisfying VSWR $\leq$ 2 (or RL $\geq$ 10 dB). A higher Q means a sharper resonance and narrower bandwidth.
There's a theoretical limit called the Bode-Fano limit, which states there is a physical upper bound to the achievable bandwidth with any matching circuit:
In other words, if the antenna itself has a high Q (e.g., small antennas), broadband matching is impossible no matter how hard you try. This is a "wall of physics" like Shannon's limit in information theory. For example, built-in antennas in smartphones have $Q \approx 10\text{-}30$, and the fractional bandwidth is limited to about 3–10%.
There's a physical limit... So how do designers work within that limit?
There are three approaches commonly used in the field:
- Multi-stage matching: When a single L-match doesn't provide enough bandwidth, use 2 or 3 stages. Increasing stages gets you closer to the Bode-Fano limit.
- Improving the antenna side: Lower the antenna's own Q. For example, adding slots, optimizing dielectric loading, etc.
- Dynamic matching (Aperture Tuner): Avoids the circuit Q constraint by switching matching conditions in real-time for each frequency band.
The Story of Philip Smith, Who "Drew" the Smith Chart by Hand
The "Smith Chart," which visualizes impedance matching, was conceived by Philip Smith of Bell Labs in 1939. At that time, there were no computers, and complex impedance transformations were done by manual calculation. Smith realized that "if you overlay constant resistance circles and constant reactance circles, the transformation becomes clear at a glance," and he created the chart by hand. Modern simulation tools draw Smith charts at the push of a button, but the essential beauty of the chart lies in the geometric transformation of equations. Trying a manual calculation on a paper Smith chart once will deepen your understanding immensely.
Mathematical Relationship Between Reflection Coefficient and Smith Chart
- Normalized impedance $z = r + jx$: A dimensionless quantity obtained by dividing $Z_{in}$ by the characteristic impedance $Z_0$. The Smith chart maps this $z$ onto the $\Gamma$ plane via a Möbius transformation.
- Constant resistance circles: The locus of $r = \text{const}$. Center is $(\frac{r}{r+1}, 0)$, radius is $\frac{1}{r+1}$. The circle for $r=0$ is the chart's outer circumference, $r=\infty$ degenerates to the origin.
- Constant reactance circular arcs: The locus of $x = \text{const}$. Center is $(1, \frac{1}{x})$, radius is $\frac{1}{|x|}$. The upper half-plane is inductive ($x > 0$), the lower half-plane is capacitive ($x < 0$).
- Constant VSWR circles: Circles of $|\Gamma| = \text{const}$. Center is the origin, radius equals $|\Gamma|$. The VSWR=2 circle corresponds to the $|\Gamma|=0.333$ circle.
Physical Meaning of the Bode-Fano Limit
- For an RC series load: $\int_0^{\infty} \ln\frac{1}{|\Gamma(\omega)|}\,d\omega \leq \frac{\pi}{RC}$
- For an RL parallel load: $\int_0^{\infty} \frac{1}{\omega^2}\ln\frac{1}{|\Gamma(\omega)|}\,d\omega \leq \frac{\pi L}{R}$
- This limit applies to passive matching circuits (lossless reactive elements only)
- Making $|\Gamma|$ smaller within the band narrows the bandwidth, widening the bandwidth makes $|\Gamma|$ larger — this trade-off cannot be fundamentally overcome even by increasing the number of matching stages
- Active matching (negative resistance or feedback circuits) can circumvent this limit, but introduces stability and noise issues
Matching Circuit Design Methods
L-Network Design
L-match is the most basic matching circuit, right? How do you design it?
The L-network is the simplest matching circuit, consisting of two reactive elements (a combination of L and C). When matching a load impedance $R_L$ to $Z_0$, for the case $R_L > Z_0$:
For example, to match $R_L = 200\,\Omega$ to $Z_0 = 50\,\Omega$: $Q = \sqrt{200/50 - 1} = \sqrt{3} \approx 1.73$, $X_s = 1.73 \times 50 = 86.6\,\Omega$, $B_p = 1.73 / 200 = 8.65 \times 10^{-3}\,\text{S}$. If we realize $X_s$ with an inductor and $B_p$ with a capacitor, then $L_s = X_s / \omega$, $C_p = B_p / \omega$.
Related Topics
なった
詳しく
報告