Terminations, Reflections & Standing Waves

01 / 10

Why “50 Ω” still fails

If the feed is called “50 ohms,” why can the sensor node still lose accepted power and show a large voltage peak?

Characteristic impedance describes the voltage-to-current ratio of one traveling wave on a uniform line. It is not a promise that every device attached to the line is 50 ohms. In the recurring node, let a 50 ohm antenna feed end in $Z_{\mathrm L}=30-j20\,\Omega$ at the product feed plane R2. The incident wave reaches a boundary whose required voltage/current ratio differs from Z0. A reverse wave supplies the boundary correction.

That same return is not equally large everywhere. At R2 it is Gamma L. At the source-side input plane R1, it has made a round trip through phase and attenuation. “The return loss” is therefore incomplete unless the reference plane is named.

Think about itIf the 50 ohm feed is cut to a different length but the load stays 30 − j20 ohms, does Gamma L change?

Answer

No. Gamma L belongs to the load boundary. The reflection observed at the new input plane rotates with twice the electrical-length change and shrinks if the added line is lossy; the input impedance can change dramatically even though ZL and Gamma L do not.

Common misconceptionIf both ports are labeled 50 ohms, they are automatically matched.

A 50 ohm connector, cable, or instrument setting declares a reference impedance. The connected device can present another complex impedance because of frequency, bias, packaging, geometry, or calibration-plane placement. Verify the boundary quantity at the same frequency and plane.

02 / 10

Load boundary and Γ

Which reverse voltage makes total voltage and current satisfy V/I = ZL at the load?

Use phasors with e^+jωt and +z from source to load. The forward voltage is V⁺e^−γz; the reverse voltage is V⁻e^+γz. Reverse-wave current carries a minus sign because positive current is defined toward the load. At the load plane, the boundary condition gives the voltage reflection coefficient.

$\Gamma_{\mathrm L}=\frac{V^-}{V^+}=\frac{Z_{\mathrm L}-Z_0}{Z_{\mathrm L}+Z_0}$ Gamma L is dimensionless and complex; ZL and real positive Z0 are in ohms, evaluated at the same frequency, mode, normalization, and load plane.

Derived · recurring 50 ohm case

30 − j20 ohms reflects with a −120.964 degree phase.

$\Gamma_{\mathrm L}=\frac{(30-j20)-50}{(30-j20)+50}$ Substitute the boundary impedance and line impedance.
$\Gamma_{\mathrm L}=-0.1764706-j0.2941176$ Rectangular form preserves the signs needed for later transforms.
$\Gamma_{\mathrm L}=0.3429972\angle-120.964^\circ$ Polar form separates reflected amplitude ratio from phase.
$Z_{\mathrm L}=Z_0\frac{1+\Gamma_{\mathrm L}}{1-\Gamma_{\mathrm L}}$ The bilinear inverse recovers the same 30 − j20 ohm boundary.

Common misconceptionReflections happen because RF is fast.

Propagation delay makes a reflection spatially observable, but the load discontinuity creates it. A matched line has no load reflection at any frequency inside the model; a mismatched boundary can reflect at low frequency once the interconnect is long enough for the distributed description to matter.

03 / 10

Canonical termination limits

What must the equations do at a match, open, short, and purely reactive load?

MatchedZL = Z0 · Gamma = 0No reverse load wave; all incident power is accepted.
OpenZL → infinite · Gamma = +1Voltage doubles at the boundary; current cancels to zero.
ShortZL = 0 · Gamma = −1Voltage cancels at the boundary; current doubles in magnitude.
Pure reactanceRe{ZL} = 0 · |Gamma| = 1Phase depends on reactance; average accepted power is zero.

Open and short are not “large” and “small” versions of the same voltage event. Their reflection phases differ by 180 degrees. A finite reactance fills the rest of the unit circle. A passive load with positive resistance lies inside that circle because some incident power can be accepted.

Think about itAt a short circuit, does a reflected voltage of −V plus imply zero current?

Answer

No. Reverse-wave current has the opposite propagation sign. With Gamma = −1, voltages cancel but forward and reverse current contributions add at the short. At an open, Gamma = +1 makes voltage add while the currents cancel.

Common misconceptionA reflected voltage wave changes the source frequency.

A linear time-invariant termination returns the same sinusoidal frequency. It changes complex amplitude. Frequency translation requires time variation or nonlinearity, neither of which is in this model.

04 / 10

Mismatch metrics

Which number reports reflected power, which reports a wave-amplitude ratio, and which one has lost phase?

$\frac{P_{\mathrm{ref}}}{P_{\mathrm{inc}}} = |\Gamma|^{2}$ For real positive Z0, the power-wave fraction reflected at the same plane.
$\mathrm{RL}=-20\log_{10}|\Gamma|$ Return loss is positive here: larger is a smaller reflection. A perfect match tends to infinite dB.
$\mathrm{VSWR} = \frac{1 + |\Gamma|}{1 - |\Gamma|}$ Lossless-line maximum-to-minimum voltage ratio; it keeps magnitude and discards phase.
$\mathrm{ML} = -10 \log_{10}(1 - |\Gamma|^{2})$ Mismatch loss under an available-power convention; it is not line attenuation or total insertion loss.

Derived · default load plane

Four metrics, one complex reflection, no interchangeable labels

|Gamma L| squared = 0.1176471, return loss = 9.29419 dB, VSWR = 2.04413:1, and mismatch loss = 0.54358 dB. The accepted fraction at the load plane is 0.8823529.

Common misconceptionReturn loss and insertion loss are the same dB quantity.

Return loss compares reflected with incident power at one port plane. Insertion loss compares transmission through a two-port arrangement, often mixing dissipation, mismatch, and the chosen normalization. They can be related inside a complete network model, but they are not synonyms.

05 / 10

Standing-wave pattern

Where do the voltage maxima move when only the reflection phase changes?

At a distance d from the load toward the source on a lossless line, the local reflection is Gamma(d) = Gamma L e^−j2βd. The factor two is the round-trip spatial phase. Total voltage is the phasor sum of the two waves; total current uses their difference. Constructive voltage addition produces a current minimum, and destructive voltage addition produces a current maximum.

$\begin{aligned}|V|_{\max}&=|V^+|(1+|\Gamma|)\\|V|_{\min}&=|V^+|(1-|\Gamma|)\end{aligned}$ These exact constant envelopes assume a lossless uniform line. With attenuation, the forward and reverse amplitudes vary along the line, so local extrema no longer repeat with identical height.

Class 1 · embedded teaching model

Reflection & Plane Walker

Change the boundary, then walk the observation plane. The load sets Gamma L; distance and loss transform what returns to the input plane.

reflection-plane-walker/1.0.0

Calculated · load plane to input plane

Gamma L = -0.1765 - j0.2941; Gamma in = 0.2689 + j0.2058

The load mismatch is unchanged. At the input, the reflection has accumulated a round-trip phase shift and 0.112 dB of round-trip loss.

|Gamma L|: 0.342997
Reflected power: 11.765%
Return loss: 9.294 dB
VSWR: 2.0441:1
Mismatch loss: 0.544 dB
Input impedance: 76.729 + j35.677 ohm

|V| Z0|I|Forward voltage is normalized to 1 at the input. Loss makes this a standing-wave pattern with unequal traveling-wave envelopes, not a perfectly stationary sinusoid.

Indigo is Gamma L at the load; cyan is Gamma in at the input. The plotted circle is the passive |Gamma| = 1 boundary, not a Smith chart.

1 W single-pass power ledger · Gamma S excluded
Plane or element	Toward load	Toward source / accepted
Input plane	1.000000 W incident	0.114652 W returned
Forward line pass	0.987188 W reaches load	0.012812 W dissipated
Load plane	0.116140 W reflected	0.871049 W accepted
Return line pass	—	0.001488 W dissipated
Conservation	1 W = 0.114652 W returned + 0.871049 W accepted + 0.014300 W line loss

Cause trace

The load boundary 30 - j20 ohm against 50 ohm sets Gamma_L.
Walking 0.28 lambda toward the source multiplies Gamma_L by exp(-2 gamma l): phase rotates by -201.6 degrees before wrapping.
The 0.056 dB one-way loss attenuates the reflected wave twice before it reaches the input plane.
Z_in follows from Z0(1 + Gamma_in)/(1 - Gamma_in); changing line length changes the observed plane, not the physical load.
Gamma_S = 0: the returning wave is absorbed at the source plane, so the single-pass 1 W ledger is complete.

Canonical boundary checks at the current Z0
Case	Load	Gamma L	Accepted	Interpretation
Matched load	50 + j0 ohm	0.000 + j0.000 · \|Gamma\| 0.000	100.00%	No load reflection; the input remains Z0 at every length.
Open circuit	infinite impedance	1.000 + j0.000 · \|Gamma\| 1.000	0.00%	Voltage reflects in phase: Gamma = +1; no real power is accepted.
Short circuit	0 + j0 ohm	-1.000 + j0.000 · \|Gamma\| 1.000	0.00%	Voltage reflects with 180 degree reversal: Gamma = -1; no real power is accepted.
Purely reactive	0 + j50 ohm	0.000 + j1.000 · \|Gamma\| 1.000	0.00%	Its reflection phase is neither the open nor short limit, but its magnitude is one.
Current load on lossy line	30 - j20 ohm	-0.176 - j0.294 · \|Gamma\| 0.343	88.24%	The load sets Gamma_L; the lossy walk changes its magnitude and phase at the input plane.

Source re-reflection and echo sequence

Gamma S is constrained to a real, nonnegative value at 0 degrees at the input plane. The pulsed ledger treats successive echoes as time-separated. For a continuous wave, coherent phasors settle by the geometric factor 1/(1 - Gamma S Gamma in).

Round-trip field ratio: 0.00000 + j0.00000
Round-trip power ratio: 0.000000
CW forward-amplitude factor: 1.00000 + j0.00000
CW load incident power: 0.987188 W

Gamma S = 0: only the initial pass remains
Pass	Input incident	Load incident	Load accepted	Return at source	Source absorbed	Next launch
1	1.000000 W	0.987188 W	0.871049 W	0.114652 W	0.114652 W	0 W

Numerical envelope samples

Every sixth model point is listed; the chart uses all 49 deterministic samples.

From input	From load	\|V plus\|	\|V minus\|	\|V total\|	Z0 \|I total\|
0.0000 lambda	0.2800 lambda	1.00000	0.33860	1.28545	0.75956
0.0350 lambda	0.2450 lambda	0.99919	0.33888	1.19352	0.89552
0.0700 lambda	0.2100 lambda	0.99839	0.33915	1.06648	1.04223
0.1050 lambda	0.1750 lambda	0.99759	0.33942	0.91945	1.17277
0.1400 lambda	0.1400 lambda	0.99678	0.33970	0.77707	1.27047
0.1750 lambda	0.1050 lambda	0.99598	0.33997	0.67709	1.32539
0.2100 lambda	0.0700 lambda	0.99518	0.34024	0.66077	1.33254
0.2450 lambda	0.0350 lambda	0.99437	0.34052	0.73636	1.29122
0.2800 lambda	0.0000 lambda	0.99357	0.34079	0.86885	1.20488

Model contract: uniform single-mode line, real positive Z0, passive finite load, frequency-local electrical length, and uniform attenuation. Phasors use exp(+j omega t), +z points from input to load, and the +z voltage wave varies as exp(-gamma z). Radiation, dispersion, connector modes, calibration, and nonlinear or time-varying loads are outside this model.

Server-rendered default · printable / no-script record

30 − j20 ohms at 0.28 lambda on a 50 ohm lossy line

Model reflection-plane-walker/1.0.0. One-way attenuation is 0.2 dB/lambda, so the selected line contributes 0.056 dB per pass. Gamma S = 0 at 0 degrees at the input plane.

Gamma L: -0.1764706 − j0.2941176
Gamma L polar: 0.3429972 at -120.964 degrees
Input impedance: 76.729 + j35.677 ohm
Return loss: 9.29419 dB
VSWR: 2.04413:1
Mismatch loss: 0.54358 dB

Calculated envelope samples · normalized forward voltage = 1 at input
From input	From load	\|V plus\|	\|V minus\|	\|V total\|	Z0 \|I total\|
0.000 lambda	0.280 lambda	1.0000	0.3386	1.2854	0.7596
0.070 lambda	0.210 lambda	0.9984	0.3391	1.0665	1.0422
0.140 lambda	0.140 lambda	0.9968	0.3397	0.7771	1.2705
0.210 lambda	0.070 lambda	0.9952	0.3402	0.6608	1.3325
0.280 lambda	0.000 lambda	0.9936	0.3408	0.8689	1.2049

Common misconceptionA standing wave means no net forward power flows.

Only |Gamma| = 1 on a lossless line forces equal forward and reverse powers and zero net delivery. For the default |Gamma| = 0.343, most incident power still crosses the load plane and is accepted, even though voltage and current show spatial maxima and minima.

06 / 10

Why VSWR is incomplete

Can a 2.044:1 VSWR tell whether the load is 30 − j20 ohms or locate the nearest voltage minimum?

No. VSWR maps every reflection with the same magnitude to one scalar. It cannot retain the sign of reactance, distinguish resistance/reactance combinations, locate extrema, or specify the reference plane. On a lossless uniform line, |Gamma| is unchanged with position, so VSWR remains fixed while the complex Gamma and Zin rotate continuously.

On a lossy line, |Gamma| decreases as the observation plane moves away from the load. A VSWR inferred at the source plane can therefore look better than the mismatch at the load. That is attenuation masking, not improved acceptance at R2.

Load plane R2

Gamma L = -0.1765 − j0.2941

The antenna-feed boundary sets the physical reflected fraction.

Input plane R1

Gamma in = 0.2689 + j0.2058

Round-trip loss shortens it and 0.56 wavelengths of phase path rotates it.

Common misconceptionVSWR uniquely identifies the load impedance.

VSWR gives only |Gamma|. Recovering ZL requires the full complex Gamma at the load plane and Z0. Even complex Gamma at another plane must first be transformed to the boundary through a declared line model.

07 / 10

Input-impedance transformation

How can the source see 76.729 + j35.677 ohms when the physical load remains 30 − j20 ohms?

Input impedance is the total V/I ratio at the selected input plane. Propagating both traveling waves from the load through a uniform line gives the lossy-line transform below. Here gamma = alpha + j beta and l is the source-to-load plane separation.

$Z_{\mathrm{in}}=Z_0\frac{Z_{\mathrm L}+Z_0\tanh(\gamma\ell)}{Z_0+Z_{\mathrm L}\tanh(\gamma\ell)}$ Equivalent plane-walk form: Gamma in = Gamma L exp(−2 gamma l), then Zin = Z0(1 + Gamma in)/(1 − Gamma in). The model uses the latter for stable canonical limits.

Derived · lossless checkpoints

Quarter-wave inversion and half-wave repetition catch sign mistakes.

$\ell=\frac\lambda4:\quad Z_{\mathrm{in}}=\frac{Z_0^2}{Z_{\mathrm L}}$ 57.6923 + j38.4615 ohm for the recurring load.
$\ell=\frac\lambda2:\quad Z_{\mathrm{in}}=Z_{\mathrm L}$ 30.0000 − j20.0000 ohm; the complex boundary repeats.
$l = 0.28 \lambda , \mathrm{loss} = \frac{0.2 \mathrm{dB}}{\lambda }$ 76.729 + j35.677 ohm at the input plane.

Go deeperWhy the factor of two appears in Gamma but not in the forward-wave delay

A forward phasor accumulates one traversal from input to load. The ratio V−/V+ at a new plane compares a reverse wave that traversed back with a forward reference that would traverse forward; their relative phase changes by twice the plane displacement. This is why moving by lambda/4 rotates Gamma by 180 degrees and inverts the normalized impedance.

Common misconceptionChanging line length changes the physical load impedance.

The load boundary remains 30 − j20 ohms. Length changes the impedance presented at another plane. That distinction is the basis of distributed matching, but component stress and accepted load power still have to be evaluated where they physically occur.

08 / 10

Source re-reflection and echoes

What happens when the wave returning to R1 encounters Gamma S instead of an absorbing source?

Part of the return is absorbed by the source termination and part is re-reflected toward the load. The next load arrival reflects again, producing a deterministic sequence. With Gamma S declared at the input plane, the complex field multiplier from one forward launch to the next is q = Gamma S Gamma L e^{−2 gamma l} = Gamma S Gamma in.

The interaction constrains Gamma S to a real nonnegative value at 0 degrees so one control cannot quietly hide source phase. For separated pulses, each echo power is the previous one times |q| squared. For a continuous sinusoid after transients settle, the forward phasors form the geometric series 1 + q + q squared + … = 1/(1 − q). These are different observations and should not be mixed.

Go deeperWhy this passive control range always settles

The load is passive, so |Gamma L| is at most one. Line attenuation cannot increase the round-trip magnitude, and this model limits |Gamma S| to 0.8. Therefore |q| is below one and both the echo power sequence and coherent phasor series converge. Active or negative-resistance boundaries require a stability analysis outside this lesson.

Common misconceptionAll power returning to a source is automatically absorbed.

That is true only for Gamma S = 0 at the declared source plane. A mismatched passive source absorbs a fraction and re-reflects |Gamma S| squared of each separated returning pulse; its phase also matters for continuous-wave settling.

09 / 10

Guided-wave reflection versus general field reflection

When is one scalar Gamma enough, and when must the field problem become larger?

The transmission-line Gamma in this module relates one forward and one reverse wave of the same guided mode at a reference plane. It assumes the line remains uniform away from a localized load. A plane wave meeting a material boundary can require polarization, angle of incidence, wave impedance for TE or TM polarization, transmitted fields, and possibly anisotropic or lossy media. A discontinuity in a real guide can also excite higher-order, radiating, or common modes.

Scope boundary · guide versus field problem

Use scalar Gamma only after the modal contract is credible.

Keep this model when one propagating mode dominates and port voltage/current or power waves are well-defined. Escalate to a multimode network or field solver when launches, gaps, bends, radiation, polarization conversion, or nonuniform cross-sections carry decision-relevant energy. The next module compares practical structures and their field assumptions.

This is also why a decorative “wave bounce” drawing cannot serve as measurement evidence. A VNA result depends on calibration, port normalization, fixtures, and the chosen reference plane; TDR adds transform, bandwidth, and windowing choices. Those workflows belong to the measurement path, not this analytic lesson.

Common misconceptionEvery electromagnetic reflection is the same scalar transmission-line Gamma.

Scalar Gamma is a single-mode boundary relation. Field reflection can be polarization- and angle-dependent, while structural discontinuities can redistribute energy among modes or radiation. The mathematical family resemblance does not erase those extra degrees of freedom.

10 / 10

Node power ledger

Starting with 1 W incident at R1, can every milliwatt be assigned without counting a loss twice?

Yes. The 0.28 lambda line has 0.056 dB one-way loss. It delivers 0.987188 W incident to R2. The load accepts 0.871049 W and reflects 0.116140 W. The returning pass leaves 0.114652 W at R1. Total forward-plus-return line dissipation is 0.014300 W.

Derived · 1 W incident at input plane R1 · Gamma S = 0
Quantity	Plane / region	Power	Accounting role
Incident toward load	R1	1.000000 W	Ledger input
Incident on load	R2	0.987188 W	After forward line loss
Accepted by load	R2 boundary	0.871049 W	Pinc,R2(1 − \|Gamma L\| squared)
Reflected by load	R2	0.116140 W	Pinc,R2 \|Gamma L\| squared
Returned to source	R1	0.114652 W	After return line loss
Line dissipation	R1 ↔ R2	0.014300 W	Forward plus return passes

Path 03 capstone artifact · part 2

Add termination and power evidence to the node reference-plane map.

Record Z0, ZL, frequency, mode, R1 and R2, +z, Gamma L, Gamma at R1, line length/loss, Zin at R1, and the full power ledger. State whether each value is defined, derived, simulated, or measured; do not blend values from different planes into one unlabeled number.

Verify Gamma by both rectangular arithmetic and a canonical-limit check.
Separate mismatch loss from physical line dissipation and any two-port insertion result.
Close 1 W = returned at R1 + accepted at R2 + total line loss.
If Gamma S is nonzero, append the declared source plane, phase, and settling model.
Escalate to multimode or measured evidence when the single-mode boundary contract fails.

Common misconceptionIf the load accepts 0.871 W, the missing 0.129 W is all reflected at the load.

The load reflects about 0.116 W at R2, not 0.129 W. The rest is line dissipation accumulated on the forward and return passes. At R1 only about 0.115 W returns because those plane-separated quantities include different portions of the line loss.

Ungraded review

Check your understanding

Answer each question in your own words, then reveal the model answer.

01What creates Gamma L: the line length, signal speed, or load boundary?
Model answer
The load boundary creates Gamma L = (ZL − Z0)/(ZL + Z0) at the load plane. Length and propagation transform that same reflection to another reference plane through Gamma(z) = Gamma L exp(−2 gamma d); they do not change the physical termination.
02For the default 30 − j20 ohm load on a 50 ohm line, what does |Gamma L| squared mean?
Model answer
It is 0.117647, the fraction of incident traveling-wave power reflected at the load plane under the real-positive-Z0 power-wave convention. The remaining 0.882353 is accepted by the passive load at that boundary.
03Why can two loads with the same VSWR produce different voltage patterns and input impedances?
Model answer
VSWR preserves only |Gamma|. It discards reflection phase, so it cannot distinguish different points on the same constant-|Gamma| circle. That phase positions the maxima and minima and enters the complex input-impedance transformation.
04What does a lossless quarter-wave line do to a finite load impedance?
Model answer
It inverts the normalized impedance: Zin = Z0 squared divided by ZL. For 50 ohms and 30 − j20 ohms, Zin is 57.692 + j38.462 ohms. A half-wave line repeats ZL.
05Why is returned power at the input smaller than reflected power at the load in the default case?
Model answer
The incident wave loses 0.056 dB before reaching the load, and the reflected wave loses another 0.056 dB on its return. Reflected-at-load power and returned-at-input power belong to different planes, separated by one lossy pass.
06What extra information is required before adding repeated source/load bounces?
Model answer
You need a source reflection coefficient with a declared phase and reference plane, plus a settling interpretation. This module constrains Gamma S to a real nonnegative value at 0 degrees at the input plane and separates time-spaced echo powers from coherent continuous-wave phasor settling.

Sources and further study

Accessed 5 September 2026. Equations, values, tables, and SVG diagrams are independently derived for the stated model. Measurement sources are orientation only; no calibration or de-embedding procedure is claimed here.

Theory and worked limits

David M. Pozar, Microwave Engineering, 4th ed., Chapter 2. Stable textbook treatment of terminated lines, reflection, standing waves, and impedance transformation.
University of Kansas EECS 723, Jim Stiles, Transmission Line Input Impedance. Official university notes with half-wave, quarter-wave, matched, reactive, and electrically short limits.
University of Kansas EECS 723, Power Flow and Return Loss and VSWR. Official teaching notes supporting the same-plane power and mismatch definitions used here.

Measurement and reference-plane orientation

IEEE, IEEE 370-2020, active; published 8 January 2021, with errata dated 21 January 2022. Normative context for high-frequency interconnect characterization to 50 GHz; detailed practice is deferred to Path 08.
Keysight, Transmission Lines and Reflected Signals. Vendor application note used for bench-oriented transmission-line and reflected-signal context, not as an independent standard.
Rohde & Schwarz, VSWR and return loss primer. Vendor orientation on the relationship among reflection coefficient, VSWR, and return loss.

Review the prerequisiteWhen an Interconnect Becomes a Transmission Line

Next · plannedReal Transmission-Line StructuresCompare field patterns, loss, dispersion, return paths, and manufacturability.

PracticePreserve the ledgerChange R, X, length, and loss; explain every output from boundary → walk → power.

Measurement boundaryDefer calibration detailsCarry named planes forward to Path 08 rather than guessing a fixture correction.

Back to Transmission Lines & Matching