smith chart
A Smith chart is a graphical calculator that turns a complex impedance into a point on a circular plot, so that an engineer can read an impedance, a reflection coefficient, a standing wave ratio, and the exact reactive element needed to cancel a mismatch all from one drawing. Before the chart was introduced in the 1930s, radio-frequency engineers had to juggle multiple equations and look up tables every time a load changed, and a single tuning iteration could take an afternoon of arithmetic. The chart collapses that arithmetic into geometry: every possible passive impedance appears somewhere inside a circle, the matched condition sits at the exact center, and the distance from the center to any point is the magnitude of the reflection coefficient, which is the quantity that decides how much forward power bounces back toward the source. In a foundry or a test laboratory the same drawing is used to plot a measured antenna feed, a plasma chamber load, or a filter network, and it remains the single most reproduced diagram in microwave engineering more than ninety years after it was published.
**The Smith chart is the reflection coefficient drawn as a map, not an impedance grid.** The chart uses a conformal transformation that takes the reflection coefficient, a complex number defined as the ratio of the reflected to the forward voltage wave, and draws it on a plane where circles of constant resistance and arcs of constant reactance curve across the interior. The center of the chart is the perfectly matched load, where the reflection coefficient is exactly zero and all power is absorbed, while the outer rim is the unit circle where the reflection coefficient magnitude is one and everything reflects back. A purely resistive load that is higher than the reference appears as a point on the right half of the horizontal real axis, a load that is lower appears on the left half, and a load with reactance moves off the axis into the upper or lower half of the chart.
**A point on the chart is worth two numbers, an impedance and a standing wave ratio.** The distance from the center of the chart to any plotted point, expressed as a fraction of the chart radius, is the magnitude of the reflection coefficient, and that single distance maps directly to a standing wave ratio through a simple formula. A point exactly at the center reads as a standing wave ratio of one to one, while a point at the midpoint of the radius reads as a reflection coefficient magnitude of 0.5 and a standing wave ratio of three to one. Because the circles of constant standing wave ratio are concentric around the center, an engineer can spin a compass across the chart and see in one motion how many points on a transmission line share the same mismatch, which is exactly the question a tuning session is trying to answer.
**The horizontal real axis is the home of purely resistive loads.** Any load whose reactance is zero lands on the real axis, and its position tells an engineer at a glance whether the load is above or below the reference impedance. A 100 ohm load on a 50 ohm line normalizes to 2.0 and lands to the right of center with a reflection coefficient of 0.333, while a 25 ohm load normalizes to 0.5 and lands to the left with a reflection coefficient of negative 0.333. Both points sit on the same constant standing wave ratio circle, because a load of 100 ohm is exactly as mismatched as a load of 25 ohm, and both reflect 11.1 percent of the incident power even though one is too high and the other is too low.
**A reactive load climbs off the axis and becomes a complex point.** When a load carries an inductor or a capacitor, its impedance has both a real and an imaginary part, and the point moves above the real axis for an inductive load or below it for a capacitive one. A load of 50 ohms in series with 50 ohms of reactance normalizes to 1 plus j1 and produces a reflection coefficient with a magnitude of 0.447, a standing wave ratio of 2.62, and a reflected power of 20.0 percent. Reading that point on the chart tells an engineer the direction to move to cancel the reactance, which is the entire reason the chart outlives its century: it shows not only where a load is but which direction a matching element must push it to reach the center.
**The chart is a map that a vector network analyzer turns into a picture.** Modern instruments that sweep a device across frequency produce the same circles that Phillip H. Smith drew by hand, but they plot thousands of points per second and overlay them with a calibrated reference circle. A network analyzer from Keysight or Rohde & Schwarz can mark the center of the chart as a 50 ohm reference, plot a filter's impedance as a curve that loops across the interior as frequency rises, and read the standing wave ratio and return loss at any point with a cursor. Field instruments from Anritsu and Bird bring the same chart to a mast or a production line, where a technician probes a feed line and watches the plotted point drift as a connector is torqued or a cable is moved.
```flowchart
flowchart TD
A[Measure S11: forward and reflected waves at the port] --> B[Normalize the load: z = Z / Z0]
B --> C[Plot the point on the Smith chart]
C --> D[Read |Γ| from distance to center, VSWR from the concentric circle]
D --> E{Is the point at the center of the chart?}
E -- yes --> F[Accept: 50 ohm reference matched, near zero reflection]
E -- no --> G[Read the direction toward center from the chart geometry]
G --> H[Add series / shunt reactance or a quarter-wave section]
H --> C
```
The table below translates the impedance points plotted on the chart into the reflection coefficient and the standing wave ratio, all normalized to a 50 ohm reference line. It is the arithmetic the chart hides behind its circles, and it is what an engineer recovers by reading the distance and direction of a point.
| Load z | Impedance (50 Ω) | Reflection Γ | VSWR | Reflected power |
|---|---|---|---|---|
| 0.5 | 25 Ω | −0.333 | 2.0 to 1 | 11.1% |
| 1.0 | 50 Ω | 0.000 | 1.0 to 1 | 0.0% |
| 1.5 | 75 Ω | 0.200 | 1.5 to 1 | 4.0% |
| 2.0 | 100 Ω | 0.333 | 2.0 to 1 | 11.1% |
| 1 + j1 | 50 + j50 Ω | 0.20 + j0.40 | 2.62 to 1 | 20.0% |
The transformation that builds the chart is compact enough to write down, and it is the reason every point on the drawing can be trusted. The reflection coefficient is computed from the normalized impedance by a single fractional expression, and the impedance can be recovered from the reflection coefficient by inverting it.
$$\Gamma = \frac{z - 1}{z + 1}, \quad z = \frac{Z}{Z_0}$$
The standing wave ratio follows from the magnitude of that coefficient, and the reflected power is its square, so the distance a point sits from the center of the chart is directly tied to how many watts come back.
$$|\Gamma| = \frac{\text{VSWR} - 1}{\text{VSWR} + 1}, \quad P_{refl} = |\Gamma|^2 \times 100\%$$
A quarter-wave transformer is the classic way to move a resistive load to the center of the chart without adding reactance. A length of line one quarter wavelength long, with a characteristic impedance equal to the geometric mean of the source and load impedances, transforms the load so it appears matched at the far end. For a 100 ohm load on a 50 ohm line the needed section is the square root of their product, which is 70.7 ohms, and for a 150 ohm load on the same line it is 86.6 ohms. On the chart the transformer rotates a resistive point along a constant standing wave ratio circle until it lands on the real axis at the center, which is why a quarter-wave section is drawn as a half-turn of a circle between the load and the source.
The instruments that draw the Smith chart are the same ones that measure a standing wave ratio, and their shared vocabulary makes the chart the meeting ground between a measurement and a design. Keysight and Rohde & Schwarz vector network analyzers plot the chart directly and overlay constant standing wave ratio circles, while Anritsu and Bird handheld units carry a simplified chart onto the field. In a wafer fabrication facility, the same chart is used to match the impedance of a plasma deposition chamber to a 13.56 MHz generator, where the load drifts as the plasma ignites and the matching network must steer the plotted point back toward the center in milliseconds. SMA and N-type connectors from Belden and Times Microwave are the ports these measurements are made through, and the reference impedance they all assume is the near-universal 50 ohm line.
The numbers that make the chart concrete are easy to remember once they are tied to hardware. On a 50 ohm line, a 100 W transmitter driving a 100 ohm load sees a reflection coefficient of 0.333 and reflects 11.1 W back, while a 25 ohm load reflects the same 11.1 W even though its point sits on the opposite side of the chart. A 75 ohm load reflects only 4.0 W out of that 100 W because its reflection coefficient is 0.20, and a perfectly matched 50 ohm load reflects nothing at all. When reactance is present, a 50 ohm series reactance on a 50 ohm base reflects 20.0 W out of 100 W, and a quarter-wave section of 70.7 ohm cable placed in front of a 100 ohm load brings the reflected power to nearly zero at the design frequency, within a band that broadens as the match improves. At 13.56 MHz a quarter-wave section in a coaxial line is about 2.7 m of cable, while the same section at 915 MHz is only about 40 mm, which is why quarter-wave matching at industrial plasma frequencies is practical to build from a short length of transmission line.
Read Smith chart through a *reflection-coefficient* lens rather than a *control-chart* lens: the drawing is not a plot of a process metric over time but a map of complex impedance, and every circle, arc, and point on it is a reflection coefficient in disguise. An engineer who reads the chart as a picture of where a load is, and then uses the geometry to steer that point toward the center, is doing in one glance what used to take a page of arithmetic. The professional habit is to read the distance from the center as the mismatch, to read the direction toward the center as the required match, and to know that the 11.1 percent reflected at 100 ohm, the 4.0 percent at 75 ohm, and the 20.0 percent at 50 plus j50 ohm are all just the same reflection coefficient written once as impedance, once as a standing wave ratio, and once as watts lost.