Smith Chart as an Engineering Map

Assumed

You can use complex arithmetic and the reflection/line transforms established in Module 03.2, plus the two-port reference-plane contract from Module 03.4. This lesson still restates every convention it needs at the point of use.

01 / 10

Why a memorized rotation fails

A dot moved clockwise and then upward. Did the network add a line, an inductor, a capacitor—or merely change coordinate view?

The shape alone cannot answer. A Smith trace is reviewable only when it declares the reference impedance, phasor convention, plane direction, whether the chart is showing impedance or admittance, and which physical topology caused the move. Otherwise a correct sketch can still encode the wrong network.

Normalization: Z₀ is real, positive, and stated in ohms.
Convention: e^(+jωt); +z points from source to load.
Plane walk: d is measured from the load toward the generator.
Coordinate: z = Z/Z₀ or y = Y/Y₀ = 1/z.
Topology: Series preserves r; shunt preserves g.
Verification: Transform the complex number after every move.

Think about itCan two engineers plot the same physical load at different chart positions and both be internally correct?

Answer

Yes. Different real reference impedances produce different normalized loads and Γ values. An impedance and admittance view also place the same physical state 180° apart. Their records become comparable only after Z₀, plane, and coordinate view are aligned.

Common misconceptionClockwise always means adding an inductor.

Clockwise is the plane-walk direction for a lossless line under this lesson’s convention. A series inductor follows a constant-r reactance arc. Direction on the page is not a component identity without the invariant and coordinate view.

02 / 10

The Smith chart is a bilinear map

What algebra turns every passive normalized impedance into a point on or inside one circle?

Normalize: $z=\frac Z{Z_0}=r+jx$
Map: $\Gamma = \frac{z - 1}{z + 1}$
Invert: $z = \frac{1 + \Gamma}{1 - \Gamma}$
Admittance: $y=\frac Y{Y_0}=\frac1z=g+jb$

These equations are the chart. The printed circles are coordinate lines created by the map. For a passive load r ≥ 0, |Γ| ≤ 1. The center Γ = 0 means z = 1, so Z = Z₀ and no wave is reflected at that plane.

Derived · checked default

Map 30 − j20 Ω on a 50 Ω system.

z = (30 − j20)/50 = 0.6 − j0.4. Substitution gives Γ = −0.1764706 − j0.2941176, |Γ| = 0.3429972, phase = −120.964°, and return loss = 9.294 dB. Inverting Γ returns 0.6 − j0.4, then multiplying by 50 Ω returns the physical load.

Common misconceptionThe Smith chart plots impedance in ohms.

It plots normalized impedance or admittance through Γ. The same 30 − j20 Ω load lands elsewhere when Z₀ changes. Always denormalize before reporting a physical result.

03 / 10

Orient center, rim, open, short, and sign

Before reading a curved grid, can you reconstruct its five non-negotiable landmarks?

Left rimshort · z = 0Γ = −1; |Γ| = 1.
Centermatch · z = 1Γ = 0; no reflected wave.
Right rimopen · z → ∞Γ = +1; |Γ| = 1.

Impedance chart

Upper half is +x and therefore inductive; lower half is −x and therefore capacitive under e^(+jωt).

Admittance chart

Upper half is +b and therefore capacitive; lower half is −b and therefore inductive under the same convention.

The entire rim is the |Γ| = 1 boundary. Purely reactive finite loads lie on it, as do the short and open limits. Moving outside the rim implies negative resistance under this normalization, so a passive-load calculation outside the rim is an immediate audit flag.

Think about itWhere does z = 1 + j1 appear relative to the center in the impedance view?

Answer

In the upper half because x is positive, on the r = 1 circle. It is not directly above the center: the bilinear map bends constant-r and constant-x coordinates into circles.

04 / 10

Constant resistance and reactance are algebra

Why do the printed curves intersect as an orthogonal coordinate grid rather than as decoration?

Write Γ = u + jv and substitute z = r + jx into the bilinear transform. Holding r or x constant produces two circle families. Their geometry is a consequence of the transform, so a plotted coordinate can always be checked with an equation.

Constant r: $(u - \frac{r}{r + 1})^{2} + v^{2} = [\frac{1}{r + 1}]^{2}$
Constant x: $(u - 1)^{2} + (v - \frac{1}{x})^{2} = (\frac{1}{x})^{2}$

Every finite z = r + jx is the intersection of one r circle and one signed x arc. The real axis is x = 0. Increasing r contracts the resistance circle toward the open point; changing the sign of x mirrors the reactance locus across the real axis.

Go deeperWhy the map preserves local angle

Away from its pole, a bilinear transform is conformal: it preserves the angle at which smooth coordinate curves intersect. That is why constant-r and constant-x loci remain an orthogonal coordinate system even after straight lines in the z-plane become circles in Γ.

05 / 10

Admittance rotates the coordinate view

Must you use a second chart to add a shunt element correctly?

No. For y = 1/z, the reflection coordinate formed from admittance is Γᵧ = (y − 1)/(y + 1) = −Γz. The same physical state therefore appears 180° away when the chart is relabeled as conductance g and susceptance b.

Derived · same default load

Reciprocal first, then read g and b.

From z = 0.6 − j0.4, y = 1/z = 1.153846 + j0.769231. The admittance point is Γᵧ = +0.1764706 + j0.2941176: exactly the negative of the impedance-view Γ. Its upper half now means positive capacitive susceptance, not inductive reactance.

Common misconceptionAn admittance chart changes the network state.

It changes coordinates, not hardware. The reciprocal and the 180° chart rotation are two descriptions of one plane. A shunt element changes the state only when its susceptance is actually added to y.

06 / 10

Moving the plane rotates Γ

What stays fixed when you observe the same mismatched load through an ideal lossless line?

$\Gamma (d) = \Gamma _{\mathrm{L}} e^{-\frac{j 4\pi d}{\lambda }}$ e^(+jωt); +z source to load; d measured from the load toward the generator.

Magnitude stays fixed and phase decreases, so the point moves clockwise around a constant-|Γ| circle. A quarter wavelength contributes −180° and gives z_in = 1/z_L. A half wavelength contributes −360° and returns to the load impedance.

Derived · checked line transforms

The default load at three planes

Distance toward generator	Z at plane	Γ	What stayed fixed
0 λ	30 − j20 Ω	-0.1765 − j0.2941	\|Γ\| = 0.3429972
0.25 λ	57.6923 + j38.4615 Ω	0.1765 + j0.2941	\|Γ\| = 0.3429972
0.50 λ	30 − j20 Ω	-0.1765 − j0.2941	Returns to load plane value

Common misconceptionA quarter-wave line always matches the load.

A uniform quarter-wave section transforms its normalized impedance to the reciprocal; it does not generally move Γ to zero. A quarter-wave transformer matches only when its own characteristic impedance and the terminal resistances satisfy the required relation.

07 / 10

Series elements preserve resistance

Which coordinate survives when an ideal reactance is inserted in series at the current plane?

Series L: $z_{2} = z_{1} + \frac{j\omega L}{Z_{0}} \qquad \Delta x > 0$
Series C: $z_2=z_1-\frac j{\omega CZ_0}\qquad\Delta x<0$

Only x changes; r is invariant. The chart path must therefore remain on the starting constant-r circle. This is the fastest topology audit: if a supposed series-only move crosses resistance circles, the construction or coordinate view is wrong.

Derived · ideal component at 2.45 GHz

After a 0.05 λ line, add 2.12250 nH in series.

The line produces z = 0.504508 − j0.153487. The inductor adds Δx = +0.653466, giving z = 0.504508 + j0.499980. The normalized resistance remains 0.504508 exactly within the ideal arithmetic; only the reactance changes sign and magnitude.

Think about itCan a single series reactance move any starting impedance directly to the center?

Answer

Only if its normalized resistance is already r = 1. A series element cannot change r, so another topology or a line-plane move is required when the starting resistance circle does not pass through the chart center.

08 / 10

Shunt elements preserve conductance

Why is reciprocal algebra safer than guessing a shunt path on an impedance-labeled chart?

Shunt C: $y_2=y_1+j\omega CZ_0\qquad\Delta b>0$
Shunt L: $y_{2} = y_{1} - \frac{j Z_{0}}{\omega L} \qquad \Delta b < 0$

Convert z to y, add the susceptance, and convert back. Only b changes; g is invariant. In the admittance view, a capacitor moves toward +b in the upper half and an inductor toward −b in the lower half.

Derived · third move of the pinned fixture

Use 1.28756 pF to close the remaining admittance error.

The previous z = 0.504508 + j0.499980 corresponds to y = 1 − j0.991025. An ideal shunt capacitor adds Δb = +0.991025 at 2.45 GHz, so y = 1 + j0, z = 1 + j0, and Z = 50 Ω. Conductance stays g = 1 through that shunt move.

Common misconceptionCapacitors always move downward on a Smith chart.

A series capacitor adds negative reactance and moves toward −x in the impedance view. A shunt capacitor adds positive susceptance and moves toward +b in the admittance view. Component type alone does not define a chart direction; topology and coordinate do.

09 / 10

Real matches are trajectories, not points

What happens to the three-move center point when frequency, loss, and component behavior are allowed to matter?

A frequency sweep is an ordered curve of plane states. A physical line’s electrical length scales with frequency, an ideal inductor’s reactance rises with frequency, and an ideal capacitor’s susceptance rises with frequency. These mechanisms move at different rates, so a center-frequency coincidence is not a broadband guarantee.

Illustrative · deterministic three-frequency check

Hold the load and physical parts fixed around 2.45 GHz.

Assumptions: Z_L remains 30 − j20 Ω, the physical line is 0.05 λ at 2.45 GHz and nondispersive, and L/C are ideal. This is a sensitivity demonstration, not an optimized or measured bandwidth claim.

Frequency	Final Z	Γ	Return loss	Local target
2.4 GHz	48.214 + j0.7969 Ω	-0.01812 + j0.00826	34.018 dB	inside
2.45 GHz	50 + j0 Ω	0 + j0	∞ dB	inside
2.5 GHz	51.8265 − j1.0036 Ω	0.01803 − j0.00968	33.779 dB	inside

On a uniform lossy line, round-trip attenuation reduces |Γ| as the observation plane moves away from the load, so the lossless circle becomes an inward spiral. Do not generalize that picture to every lossy network: discontinuities, dispersive parameters, and resonant components can create more complicated trajectories.

Go deeperWhat the chart still does not prove

A plotted match does not establish available gain, noise performance, stability, harmonic behavior, current or voltage stress, component self-resonance, layout parasitics, or tolerance yield. Those constraints belong in the matching objective and validation plan.

10 / 10

Propose and verify the node’s next moves

Can you name the next invariant before the chart draws it—and recover the final physical impedance afterward?

Use the workbench as an engineering notebook. Commit a plane, select a line or component topology, predict its locus and direction, then apply. Numbered markers, equation rows, and the coordinate table all come from model smith-map/1.0.0.

Derived · interactive

Smith Map with Algebra Trace

Predict the invariant and direction, apply one physical move, then audit the same state as Z, Y, z, y, and Γ.

smith-map/1.0.0

Choose a topology-aware move

One line section or one ideal lumped element per step.

Move family Lossless line Lumped element

CandidateIdeal component value nH

Current teaching range: 0.1–20 nH. It is recomputed only after the plane or candidate changes; rounded display values do not replace the SI calculation.

Predict before plotting

Name the invariant and sign before the chart confirms it.

Expected move

0 of 6 moves used.

Chart coordinates Impedance z Admittance y

Outside local 10 dB target

Marker 0: in the lower half, to the left of center; constant-r locus r = 0.6; −x (capacitive). |Γ| = 0.342997; target radius = 0.3162278.

Audit every coordinate

The chart is a view of this table, not a substitute for it.

Physical impedance Z	30 − j20 Ω
Physical admittance Y	0.023077 + j0.015385 S
Normalized impedance z	0.6 − j0.4
Normalized admittance y	1.1538 + j0.7692
Impedance-view Γ	-0.176471 − j0.294118
Admittance-view Γᵧ	0.176471 + j0.294118
\|Γ\| ∠ phase	0.342997 ∠ -120.964°
Return loss / VSWR	9.294 dB / 2.0441

Operation-by-operation algebra

Each row states the update, invariant, and physical cause.

Marker	Move	Equation	Invariant	Result
0	Load planeNormalize the physical load before reading any Smith-chart geometry.	$z_{\mathrm L}=\frac{Z_{\mathrm L}}{Z_0}=0.6 - j0.4$	Starting state	$z=0.6 - j0.4$ Γ = -0.1765 − j0.2941 · RL 9.29 dB

Model boundary: line moves are lossless and components are ideal at one frequency. A real line can shrink |Γ| through round-trip attenuation, and real components add loss and parasitics. The local pass circle is a teaching criterion, not a standard.

Workbench ready at the default load plane.

Derived · server-rendered fallback · smith-map-node/1.0.0

The default node and its pinned three-move verification

If scripting is unavailable, this chart and table preserve the complete default result. Start at 50 Ω, 30 − j20 Ω, and 2.45 GHz; walk 0.05 λ toward the generator; add an ideal 2.1224979 nH series inductor; then add an ideal 1.2875633 pF shunt capacitor.

Plane	Normalized z	Physical Z	Γ	\|Γ\| / phase
Default load · 0 λ	0.6 − j0.4	30 − j20 Ω	-0.176471 − j0.294118	0.3429972 / -120.964°
Quarter wave · 0.25 λ	1.1538 + j0.7692	57.6923 + j38.4615 Ω	0.176471 + j0.294118	0.3429972 / 59.036°
Half wave · 0.50 λ	0.6 − j0.4	30 − j20 Ω	-0.176471 − j0.294118	0.3429972 / -120.964°

Marker	Move	Equation	z	Γ	RL / target
0	Load plane	$z_{\mathrm L}=\frac{Z_{\mathrm L}}{Z_0}=0.6 - j0.4$	`0.6 − j0.4`	`-0.1765 − j0.2941`	9.294 dB · outside
1	Line 0.05λ toward generator	$\Gamma_2=\Gamma_1e^{-j4\pi\times0.05}$	`0.5045 − j0.1535`	`-0.3156 − j0.1342`	9.294 dB · outside
2	Series inductor	$z_2=z_1+j0.6535$	`0.5045 + j0.5`	`-0.1971 + j0.3978`	7.052 dB · outside
3	Shunt capacitor	$\begin{aligned}y_2=y_1+j0.991\\y_1=1 - j0.991\end{aligned}$	`1 + j0`	`0 + j0`	∞ dB · inside

Final algebra: Z = 50 + j0 Ω and |Γ| = 0. The local teaching criterion is return loss ≥ 10 dB, equivalently |Γ| ≤ 0.3162278. It is intentionally not presented as an industry standard.

Next handoff

Continue to Module 03.6, Matching Networks That Survive Reality. Carry forward the named planes, Z₀, frequency trajectory, topology, exact component values, idealizations, and target definition; the next module adds realizability, bandwidth, loss, parasitics, stress, and tolerance.

Ungraded review

Check your understanding

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

01Why must Z be normalized before it is plotted on a Smith chart?
Model answer
The chart is the bilinear map of z = Z/Z₀, not of impedance in ohms. Without the declared real, positive Z₀, a plotted point has no unique physical impedance and cannot be audited.
02Where do an open, short, and matched load appear in the impedance view?
Model answer
An open is Γ = +1 at the right rim, a short is Γ = −1 at the left rim, and a match is Γ = 0 at the center. The entire rim has |Γ| = 1.
03What does rotating an impedance point by 180° reveal?
Model answer
It reveals the corresponding normalized admittance y = 1/z because Γᵧ = −Γz. The physical state is unchanged; only the coordinate view has rotated.
04Under this lesson’s convention, which way does Γ rotate when the reference plane moves toward the generator?
Model answer
Clockwise. With e^(+jωt), +z from source to load, and distance d measured toward the generator, Γ(d) = ΓL e^(−j4πd/λ). A quarter wavelength rotates 180° and a half wavelength rotates 360°.
05Which chart coordinate is invariant for a series element, and which for a shunt element?
Model answer
A series reactance changes x at constant normalized resistance r. A shunt susceptance changes b at constant normalized conductance g. This topology check should be made before plotting.
06Does a center-frequency point inside the lesson’s 10 dB circle prove a practical broadband match?
Model answer
No. The 10 dB circle is a local teaching criterion, not a standard. A practical decision also needs the frequency trajectory, loss, component parasitics and Q, tolerance, power stress, reference planes, and system requirement.

Sources and model boundary

Sources were checked 5 September 2026. Equations here use e^(+jωt), a real positive Z₀, +z from source to load, and clockwise motion when walking toward the generator. The worked network is deterministic instructional data, not measured hardware.

Core microwave reference: D. M. Pozar, Microwave Engineering, 4th ed., transmission-line and Smith-chart treatment.
Smith geometry and moves: M. Steer, Smith Chart and Transmission Lines and Smith Chart.
Instrument context: Rohde & Schwarz, Understanding the Smith chart; Keysight, Network Analyzer Basics.
Reference-plane context: IEEE Std 370-2020 and its 21 January 2022 errata. Calibration, de-embedding, and uncertainty are outside this module.

Smith Chart as an Engineering Map

Why a memorized rotation fails

The Smith chart is a bilinear map

Map 30 − j20 Ω on a 50 Ω system.

Orient center, rim, open, short, and sign

Impedance chart

Admittance chart

Constant resistance and reactance are algebra

Admittance rotates the coordinate view

Reciprocal first, then read g and b.

Moving the plane rotates Γ

The default load at three planes

Series elements preserve resistance

After a 0.05 λ line, add 2.12250 nH in series.

Shunt elements preserve conductance

Use 1.28756 pF to close the remaining admittance error.

Real matches are trajectories, not points

Hold the load and physical parts fixed around 2.45 GHz.

Propose and verify the node’s next moves

Smith Map with Algebra Trace

Commit the starting plane

Choose a topology-aware move

Predict before plotting

Audit every coordinate

Operation-by-operation algebra

The default node and its pinned three-move verification

Check your understanding

Sources and model boundary