Lecture 6: Nonlinear Analysis

What this lecture covers

Earlier lectures restricted attention to linear circuits, where superposition and the Thevenin method apply. This lecture answers what happens once a circuit contains a genuinely nonlinear element, and shows that the same node method used for linear circuits still works, since nonlinear circuits are still lumped circuits under the lumped matter discipline.

Using a fictitious exponential device, the lecture demonstrates the analytical (node-equation) method and a graphical method for solving nonlinear circuits, then motivates why nonlinear behavior matters practically with a live demo: sending music over a light beam through a garage-door-opener style LED, and showing how the exponential relationship between current and voltage distorts the signal.

Key ideas

• Digital circuits are nonlinear in general but locally linear: a digital gate's transfer curve is nonlinear overall, but for a fixed set of switch (input) settings, the circuit reduces to a linear combination of voltage sources and resistors that can be analyzed with linear techniques.
• Nonlinear circuits remain lumped circuits: because they still obey the lumped matter discipline, KVL, KCL, and the node method all still apply directly to circuits containing nonlinear elements.
• Exponential device model: the lecture's fictitious nonlinear element follows ID = A * e^(B * VD), a stand-in for real nonlinear devices such as diodes and LEDs.
• Thevenin reduction around a nonlinear device: any linear network connected to a single nonlinear device can be replaced by its Thevenin equivalent (a voltage source in series with a resistor), reducing an arbitrarily complex circuit to a simple, recurring pattern.
• Analytical (node) method for nonlinear circuits: write the node equation using the nonlinear element's VI relationship in place of Ohm's law, then solve the resulting equation, by trial and error or substitution if no closed form is available.
• Graphical (load-line) method: plot the nonlinear device's VI curve and the linear relationship implied by the source and resistor (a straight line) on the same axes; their intersection is the circuit's operating point, and this straight line is called the load line.
• Nonlinear distortion in practice: passing a time-varying signal (such as music) through an exponential device produces an output where positive excursions are amplified sharply and negative excursions are compressed, distorting the waveform.
• Motivation for small-signal analysis: this distortion motivates the next lecture's method, incremental (small-signal) analysis, for handling nonlinear devices operating near a fixed point.

Walkthrough

Review: the linear and nonlinear playground (0:00)

The lecture reviews the lumped circuit abstraction and its linear subset, where superposition, Thevenin, and Norton methods apply, then introduces the nonlinear region of the same playground. It shows that digital gates are nonlinear devices overall (their transfer curve is not a straight line), but that for any fixed combination of switch inputs, the circuit becomes linear and can be analyzed with the tools already covered.

Introducing a nonlinear element and Thevenin reduction (10:22)

A fictitious nonlinear device, called the exponential device, is defined with the relationship ID = A * e^(B * VD). The lecture emphasizes a recurring circuit pattern: a voltage source and resistor connected to a nonlinear device. Because everything except the nonlinear device is linear, that portion of the circuit can always be reduced to a Thevenin equivalent, so most nonlinear circuits studied in the course reduce to this same simple template.

The analytical method (17:34)

Applying the node method to the voltage-source-resistor-nonlinear-device circuit, the lecture writes one node equation combining the resistor current with the exponential device's current, yielding one equation in the unknown node voltage VD. Since a closed-form solution is not always available, the lecture solves an example numerically by trial and error (with sample values V = 1V, R = 1 ohm, B = 1/V, A = 1/4 A, giving VD around 0.5V and ID around 0.4A), noting that the course does not require formal numerical methods for this.

The graphical method (22:46)

The same two equations, the device's exponential VI curve and the straight-line relationship from the source and resistor, are plotted on the same ID versus VD axes. Their intersection gives the same solution found analytically. The straight-line relationship from the source and resistor is named the load line, a term the lecture says will be explained further in a later lecture.

Motivating example: music over a light beam (27:51)

The lecture builds a circuit resembling a garage-door-opener light beam: a music signal drives an LED (treated as having the same exponential VI characteristic as the fictitious device), whose light intensity is proportional to its current; a receiving photoresistor converts received light back into a current that is amplified and played through a speaker. Because the LED's current-voltage relationship is exponential rather than linear, the lecture shows graphically that a sinusoidal input voltage produces a badly distorted output current: positive excursions are amplified into sharp peaks while negative excursions are compressed almost flat.

Live demonstration of distortion (38:04)

A physical version of the circuit plays music over the light beam, and the output waveform on the oscilloscope shows the predicted distortion, with negative excursions squashed. The lecture contrasts this with what an ideal linear light-emitting device would produce, an undistorted sinusoid, and poses the open question of how to handle nonlinear devices without this distortion, to be picked up in the next lecture's treatment of incremental (small-signal) analysis.

Before you watch

• Watch Lectures 1 through 5 first; this lecture assumes fluency with the node method, superposition, and the Thevenin method for linear circuits.
• Comfort with reading a device's VI (current-voltage) curve, introduced in earlier lectures for resistors and diodes, is needed to follow the graphical load-line method.

Check your understanding

1. Why does a digital gate's overall transfer curve look nonlinear, even though the circuit can be analyzed with linear techniques for any fixed set of switch settings?
2. How does the presence of a Thevenin-reducible linear network around a nonlinear device simplify nonlinear circuit analysis?
3. What is the load line in the graphical method, and how is the circuit's solution found from it?
4. Why does an exponential current-voltage relationship distort a sinusoidal input signal asymmetrically, amplifying positive excursions while compressing negative ones?