Input Impedance of a Terminated Transmission Line

On a lossless line, Zin = Z0·(ZL + jZ0·tan βl)/(Z0 + jZL·tan βl), so the input impedance repeats every half wavelength while |Γ| and VSWR stay fixed. A 100 Ω load seen through λ/8 of 50 Ω line becomes 40 − j30 Ω, still at VSWR 2.

Engineer’s note: As the line in front of a load gets longer, its impedance circles around the Smith chart. At high frequency a few millimetres of trace can turn a perfect 50 ohm load strongly inductive or capacitive. That's why the matching parts in front of a chip belong right up against its pin.

For practitioners

A load does not appear at the input of a line as it is. Unless the load equals the characteristic impedance, the line transforms it, and the transformation depends on the electrical length. This page derives the lossless input-impedance formula, checks it against the cases you already know, and works one example by hand.

Setting up the problem

Take a lossless line with characteristic impedance Z0, terminated in a load ZL. Measure distance l back from the load toward the generator. The question is what impedance Zin a source sees at the input, at distance l.

On the line, the voltage and current are the sum of a forward wave and a reflected wave. At the load, the ratio of reflected to forward voltage is the load reflection coefficient:

ΓL=ZL−Z0ZL+Z0\Gamma_L = \frac{Z_L - Z_0}{Z_L + Z_0}

The phase constant is β = 2π/λ. Here λ is the wavelength on the line, not in free space.

Use the guided wavelength. On coax, microstrip or stripline the wave travels slower than in free space, so λ on the line is shorter than the free-space value at the same frequency. If you compute βl from the free-space wavelength, every phase in this article comes out wrong, and the error grows with line length.

Deriving the formula

Write the voltage and current at distance l from the load as forward and reflected waves:

V(l) = V⁺·(e^(jβl) + ΓL·e^(−jβl))

I(l) = (V⁺/Z0)·(e^(jβl) − ΓL·e^(−jβl))

Divide one by the other and factor out e^(jβl). This gives the impedance in terms of the local reflection coefficient:

Zin=Z01+Γ(l)1−Γ(l),Γ(l)=ΓLe−j2βlZ_{\text{in}} = Z_0 \frac{1 + \Gamma(l)}{1 - \Gamma(l)}, \quad \Gamma(l) = \Gamma_L e^{-j 2 \beta l}

This form is already useful. It shows that moving along the line only rotates Γ, and that is the basis of the Smith chart. To get a form in ZL directly, substitute ΓL = (ZL − Z0)/(ZL + Z0) into the first expression, expand e^(±jβl) = cos βl ± j·sin βl, and divide top and bottom by cos βl:

Zin=Z0ZL+jZ0tan⁡βlZ0+jZLtan⁡βlZ_{\text{in}} = Z_0 \frac{Z_L + j Z_0 \tan \beta l}{Z_0 + j Z_L \tan \beta l}

This is the lossless input-impedance equation. It applies to any load, real or complex, and any length.

Special cases worth checking

Before you trust a formula, check it against cases with known answers.

  • Matched load, ZL = Z0. The numerator and denominator become Z0·(1 + j·tan βl), so Zin = Z0 at every length. A matched line is transparent to its own length.
  • Zero length, l = 0. tan βl = 0, so Zin = ZL. The line has not done anything yet.
  • Half wavelength, l = λ/2. βl = 180°, tan βl = 0 again, so Zin = ZL. The impedance repeats every λ/2.
  • Quarter wavelength, l = λ/4. tan βl goes to infinity. Divide through by tan βl and the limit is Zin = Z0²/ZL. This is the quarter-wave transformer: a high impedance becomes low and a low one becomes high.
  • Short circuit, ZL = 0. Zin = jZ0·tan βl, a pure reactance. It is inductive for lengths below λ/4.
  • Open circuit, ZL → ∞. Zin = −jZ0·cot βl, also purely reactive, and capacitive for lengths below λ/4.

All of these follow from one equation. If your own algebra or code does not reproduce them, find the error before you go further.

Worked example: 100 Ω through λ/8 of 50 Ω

Let Z0 = 50 Ω, ZL = 100 Ω, l = λ/8. Then βl = (2π/λ)·(λ/8) = π/4 = 45°, and tan βl = 1.

Zin=50⋅100+j5050+j100=40−j30 ΩZ_{\text{in}} = 50 \cdot \frac{100 + j50}{50 + j100} = 40 - j30 \ \Omega

To check the division, multiply top and bottom by the conjugate (50 − j100). The numerator is (100 + j50)(50 − j100) = 5000 − j10000 + j2500 + 5000 = 10000 − j7500. The denominator is 50² + 100² = 12500. The ratio is 0.8 − j0.6, and 50 times that is 40 − j30 Ω.

A purely resistive 100 Ω load now looks like 40 Ω in series with a capacitive reactance of 30 Ω. Nothing was added to the circuit. The line alone did this.

Now check the mismatch. At the load:

|ΓL| = (100 − 50)/(100 + 50) = 1/3

At the input:

Γin = (40 − j30 − 50)/(40 − j30 + 50) = (−10 − j30)/(90 − j30)

|Γin| = 31.6/94.9 = 1/3

Working it through exactly, Γin = −j/3. That is ΓL = 1/3 rotated by −2βl = −90°, as Γ(l) = ΓL·e^(−j2βl) predicts. The magnitude is unchanged, so VSWR = (1 + 1/3)/(1 − 1/3) = 2 at both ends.

l = λ/8 (βl = 45°)Z0 = 50 ΩZL = 100 ΩZin =40 − j30 ΩinputloadVSWR = 2 at both ends|ΓL| = |Γin| = 1/3Input impedance vs line lengthLine length lβlZin00°100 Ωλ/845°40 − j30 Ωλ/490°25 Ω3λ/8135°40 + j30 Ωλ/2180°100 Ωλ = wavelength on the line
A 100 Ω load on a 50 Ω lossless line, showing how the input impedance changes in λ/8 steps while the VSWR stays at 2. Concept drawing.

The table in the figure continues the same calculation in λ/8 steps. At λ/4 the formula gives Z0²/ZL = 2500/100 = 25 Ω. At 3λ/8, tan βl = −1, and the result is the complex conjugate of the λ/8 value, 40 + j30 Ω. At λ/2 the impedance is back to 100 Ω. Every entry sits at VSWR 2.

What changes with length and what does not

On a lossless line, length changes the phase of Γ and nothing else. In practice this means:

  • The impedance moves on a circle. Every value of Zin for a given load lies on one constant-VSWR circle of the Smith chart, centred on Z0. Adding line rotates the point clockwise, toward the generator, by 2βl. One full turn takes λ/2.
  • The match does not improve or degrade. Adding lossless line cannot fix a mismatch or make it worse. Return loss and mismatch loss are set by |ΓL|.
  • What the line does is move the mismatch. A real load can come out complex, and complex can come out real. That is why line sections are used as matching elements: you rotate to a point where a simple series or shunt reactance brings you to the centre.
  • Frequency matters. βl is proportional to frequency for a fixed physical length. A section that is exactly λ/8 at one frequency is a different electrical length at another, so Zin spreads along the circle across a band.

When you measure a load through a cable or fixture, this is the effect that de-embedding or port extension removes. The VSWR you read is correct. The impedance is the transformed one, not the load.

In plain words · for students and beginners

Input impedance of a transmission line in plain words

An everyday analogy

Picture a long jump rope tied to a wall. You shake one end and a wave runs to the wall and comes back. What your hand feels depends on two things: how the rope is tied at the wall, and how far away the wall is. The return wave gets back to your hand late, and how late depends on the length of the rope.

A transmission line works the same way. The load is the wall, and the impedance is what the signal source "feels" at its end of the line. The reflected wave comes back with a delay, so the source sees the load mixed with that delay.

Where the analogy breaks down: a rope loses energy, and a real rope's wave changes shape. This article assumes a lossless line, where the reflected wave keeps its full strength and only its timing (its phase) changes.

The worked example in simple steps

  1. The line is 50 Ω. Its characteristic impedance Z0 is the impedance a wave sees while travelling along it. The load is a 100 Ω resistor.
  2. The line is one eighth of a wavelength long. The wavelength is measured on the line, which is shorter than in air.
  3. One eighth of a wavelength is a 45° phase shift, and tan 45° = 1.
  4. Put the numbers into the formula: 50 × (100 + j50)/(50 + j100) = 40 − j30 Ω. The "j" marks the reactive part, which stores energy instead of turning it into heat. A minus sign means it acts like a capacitor.
  5. Check the reflection. At the load it is 1/3 of the wave. At the input it is still 1/3, only shifted in phase. So the VSWR (standing wave ratio, a measure of how bad the mismatch is) is 2 at both ends.

On a Smith chart, the load and the input are two points on the same circle. Adding line length moves the point around that circle without changing its size:

r = 2r = 0.5Circle of VSWR 2|Γ| = 0.333 everywhere on itEvery load on the circlereflects 11.1 % of the power.Centre = perfect match,outer edge = total reflection.
Constant-VSWR circle (VSWR 2) on a Smith chart, computed from the reflection coefficient. Concept drawing.

Mistakes beginners make

  • Using the free-space wavelength. The wave is slower on a real cable or PCB trace, so the wavelength on the line is shorter.
  • Thinking a longer cable fixes a mismatch. On a lossless line, extra length only moves the point around the circle. The VSWR stays the same.
  • Assuming a resistor still looks like a resistor. Through λ/8 of line, a 100 Ω resistor looks like 40 Ω plus a capacitive part.
  • Forgetting the repeat. The impedance comes back to the load value every half wavelength, not every full wavelength.

Remember

  • Zin = Z0·(ZL + jZ0·tan βl)/(Z0 + jZL·tan βl), with λ measured on the line.
  • Line length changes the phase of the reflection, not its size, so VSWR is the same everywhere on a lossless line.
  • Special lengths: λ/2 gives the load back, λ/4 gives Z0²/ZL, and a matched load gives Z0 at any length.

Frequently asked questions

What is the formula for the input impedance of a lossless transmission line?

Zin = Z0·(ZL + jZ0·tan βl)/(Z0 + jZL·tan βl), where β = 2π/λ and λ is the wavelength on the line. It applies to any load ZL and any length l.

Does the VSWR change along a lossless transmission line?

No. The reflection coefficient at distance l from the load is Γ(l) = ΓL·e^(−j2βl), so only its phase changes. |Γ| and the VSWR are the same at every point on the line.

What is the input impedance of a quarter-wave line?

At l = λ/4, tan βl goes to infinity and the formula reduces to Zin = Z0²/ZL. A 100 Ω load seen through λ/4 of 50 Ω line looks like 25 Ω.

Why does the input impedance repeat every half wavelength?

Γ(l) rotates by 2βl. At l = λ/2 that rotation is 360°, so the reflection coefficient, and therefore the impedance, returns to its value at the load.

References and further reading