The Scariest Chart in Electrical Engineering

The Smith Chart is famously intimidating to electrical engineering students, often described as 'terrifying' and 'black magic.' Despite this, it is widely used—millions of copies have been printed and it is embedded in modern software—because it solves a paradoxical problem in power transfer. The chart's ability to represent infinite impedance within a finite circle is one of its most puzzling features.

In 1928, Philip H. Smith began working at Bell Labs on a project to send radio signals across the Atlantic using a directional array of over 20 antennas connected by more than two kilometers of transmission line. The goal was to focus the beam to about 10 degrees, increasing power in that direction by 400 times. However, Smith discovered that reflections along the line caused significant power loss, preventing the signal from reaching the antenna efficiently.

Smith observed that when sending a signal from the source to the antenna, part of the wave bounced back due to impedance mismatches, reducing the power that reached the antenna. This reflection problem is specific to alternating current (AC), where voltage and current oscillate as sine waves. In contrast, direct current (DC) flows steadily without such reflections. Minimizing these reflections was critical for long-distance radio transmission.

Using a slinky analogy, the video explains that waves have a consistent wavelength (distance between peaks) and frequency (peaks per second). The speed of the wave equals wavelength times frequency. For electromagnetic waves in a vacuum, speed is constant at about 300 million meters per second, but in a transmission line it is slower. When the wavelength is long compared to the line length, reflections cause minimal distortion; but when the wavelength is short, reflections significantly alter the signal shape, leading to power loss.

Impedance is defined as the ratio of voltage to current in AC circuits, with magnitude indicating wave size and angle indicating phase shift. To match a load like an antenna (e.g., 10-j30 ohms) to a 50-ohm transmission line, you cannot simply add a resistor because it wastes power as heat. Instead, reflections cause impedance to vary along the line, so you can find a point where the real part matches 50 ohms and then cancel the imaginary part with a lossless inductor or capacitor.

In the 1930s, as political tensions rose, clear long-distance signal transmission became strategically important for coordinating ships, aircraft, and supply networks. Oliver Heaviside had earlier described transmission line behavior with equations, but using them to eliminate reflections was tedious with slide rules. Independently, Philip Smith in the US, Amiel Volpert in the Soviet Union, and Tōsaku Mizuhashi in Japan all sought a simple graphical system for quick impedance matching.

Philip Smith started designing a chart based on the complex plane but with a key modification to handle different characteristic impedances (e.g., 50 ohms vs. 75 ohms). If raw impedance values were plotted directly, each system would require its own chart with a different target, so Smith needed a normalized approach to make the chart universal.

The team successfully demonstrates impedance matching using a stub—a shorted or open-circuited length of transmission line. By trimming a coaxial cable stub to a precise length (77 mm or 92 mm), they cancel the reactive component of the load impedance, achieving optimal power transfer and eliminating reflections. The process involves starting at the open circuit position on the Smith Chart, walking around the constant reflection coefficient circle to the desired reactance, and converting the angular distance into a physical stub length.

The Smith Chart graphically represents the entire impedance matching problem, enabling radio engineers to solve it without complex calculations. It was independently developed by Phillip Smith (1937), Mizuhashi in Japan (1937), and Volpert in the Soviet Union (1939). The chart's outer circle corresponds to a reflection coefficient magnitude of one (open or short circuit), and moving along the line corresponds to traveling around the chart, with one full lap equaling half a wavelength.

By adding a short or open circuit stub of appropriate length, engineers can synthesize any reactance value, replacing discrete inductors or capacitors. The stub acts like a cove off a river: the wave reflects off the open end, and the stub's length sets the timing of the returning wave. This allows cancellation of the imaginary part of the load impedance, as demonstrated by trimming a 92 mm stub down to 77 mm to achieve a match.

The series stub configuration shown is educational but impractical because it requires cutting the main transmission line. In practice, parallel (shunt) stubs are preferred, but parallel impedances do not add linearly—admittance does. Engineers use the Admittance Smith Chart (a flipped version) to easily design parallel stubs. The team notes that the stub is made of the same cable it is fixing, and the process mirrors what RF engineers do daily.