A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Bode magnitude plot uses a logarithmic scale (dB vs log ω). Asymptotic approximations simplify sketching: zeros and poles contribute slopes of ±20 dB/decade; corner frequencies mark transitions. Actual magnitude may exceed asymptotes near resonance. This technique enables rapid qualitative analysis without computation.
From your study of frequency response, you know that a transfer function G(jω) assigns a gain |G(jω)| and phase shift ∠G(jω) to every frequency ω. Plotting gain across a wide range of frequencies (say, from 0.01 rad/s to 10,000 rad/s) on a linear scale produces a curve that is essentially flat at low frequencies and nearly zero at high frequencies — the interesting variation is compressed into a tiny region. The Bode plot solves this by using a logarithmic frequency axis and expressing gain in decibels: 20 log₁₀|G(jω)|. Both transformations work together: the log frequency axis spreads out the interesting behavior, and the dB scale converts the multiplicative structure of |G(jω)| into a sum of individual contributions.
This sum structure is the key. A transfer function like G(s) = K(s+z)/[(s+p₁)(s+p₂)] has a magnitude that is a product of factors: |K| · |jω+z| / (|jω+p₁| · |jω+p₂|). In dB, this product becomes a sum: 20log|K| + 20log|jω+z| − 20log|jω+p₁| − 20log|jω+p₂|. Each term can be plotted separately and the results added graphically. The asymptotic approximation makes each term easy to draw: for a real zero at frequency z (the corner frequency ωz = z), the magnitude contribution is approximately 0 dB for ω ≪ z and rises at +20 dB/decade for ω ≫ z. A real pole at p contributes 0 dB for ω ≪ p and falls at −20 dB/decade for ω ≫ p. The transition happens at the corner frequency; the maximum asymptote error is 3 dB right at the corner.
To sketch a complete Bode magnitude plot: begin with the DC gain (set ω = 0 and compute 20log|G(0)|) as a horizontal starting line. Then process each pole and zero in order of increasing corner frequency. At each corner frequency, add ±20 dB/decade to the running slope — +20 for a zero, −20 for a pole. Integrators or differentiators (poles or zeros at the origin) set the initial slope rather than changing it: a pole at the origin means the curve starts with a slope of −20 dB/decade through the entire low-frequency range. Complex conjugate pole pairs introduce a −40 dB/decade slope change at their natural frequency ω_n, plus a resonant peak whose height depends on damping ratio ζ — the asymptote underestimates the actual gain near ω_n when ζ is low.
The practical payoff is design insight without computation. A flat low-frequency response followed by a −40 dB/decade rolloff identifies a second-order low-pass filter. A rising +20 dB/decade slope that levels off identifies a zero followed closely by a pole — the signature of a lead compensator. A slope that changes from −20 to −40 dB/decade indicates a system that will have poor phase margin at the second corner, alerting you to a potential stability problem before you've touched a calculator. Bode magnitude asymptotes are the vocabulary for reading and designing frequency-domain behavior by inspection.