Bode Plot, Gain Margin and Phase Margin (Plus Diagrams)

What is a Bode Plot?

A Bode plot is used in control system engineering to show the frequency response of a control system or component. The Bode magnitude plot shows 20 log10 of the magnitude, usually in decibels, while the Bode phase plot shows phase in degrees. Both use a logarithmic frequency axis.

Hendrik Wade Bode developed these logarithmic methods at Bell Labs in the 1930s. A Bode plot can represent transfer functions with right-half-plane poles or zeros. However, classical gain and phase margins alone do not establish closed-loop stability when the open loop has unstable poles, several crossovers or unusual encirclements. Use the Nyquist stability criterion or a closed-loop pole check for the full count.

Gain margins and phase margins estimate how much separate gain or phase variation a nominal negative-feedback loop can tolerate before losing stability. They are relative-robustness measures, not complete stability proofs.

Gain Margin

Gain Margin Impact: At a selected phase crossover, gain margin (GM) is the factor by which loop gain must change to reach unity magnitude. A positive dB margin usually describes allowable gain increase; a negative margin can describe stability loss when gain decreases. Always confirm that the nominal closed loop is stable.

Read gain margin at the phase crossover frequency, where loop phase is -180° modulo 360°. The magnitude plot’s signed distance from 0 dB gives the gain change needed to reach unity loop magnitude. If several phase crossings exist, evaluate every associated margin and use the limiting result.

Gain vs. Gain Margin: Loop gain is the plotted response. Gain margin is a distance to the critical unity-gain condition at a phase crossover. In dB, GM equals the negative of the loop magnitude in dB at that frequency.

Gain Margin Formula

The formula for Gain Margin in decibels can be expressed as:

   

Here G is the loop magnitude in dB at the selected phase crossover frequency. In absolute units, GM = 1/|L(jωpc)|.

In the diagram, the gain (G) at the shown phase crossing is +20 dB. Therefore GM = 0 – 20 dB = -20 dB. Under the standard single-loop assumptions, this crossing indicates that the nominal gain lies beyond the stable gain boundary. A full conclusion still requires the open-loop pole count and all crossovers.

Phase Margin

Phase Margin (PM) measures, in degrees, the additional phase lag needed at a unity-gain crossing to reach -180° under the standard negative-feedback convention. A larger positive value often means a less fragile nominal loop, but bandwidth, damping, delay and multiple crossovers also matter.

Reading Phase Margin: Measure PM at the gain crossover frequency, where the loop magnitude crosses 0 dB. It is the phase curve’s distance from -180°, not its distance from the phase plot’s x-axis. Evaluate all unity-gain crossings when more than one exists.

The phase lag and the Phase Margin are not the same things. Phase lag is the loop phase at the selected frequency; phase margin is the remaining angular distance to the critical -180° condition.

Phase Margin Formula

The formula for Phase Margin at a unity-gain crossing can be expressed as:

   

Here is the unwrapped loop phase in degrees at the gain crossover frequency. With the usual negative phase convention, PM = 180° + φ.

In the diagram, the phase at the shown gain crossing is -189°. Therefore PM = -189° – (-180°) = -9°. Under standard assumptions, the negative margin signals that the nominal loop has passed the phase boundary at that crossing.

If an amplifier’s loop gain crosses 0 dB where phase is -120°, PM = -120° – (-180°) = 60°. The result is a positive classical phase margin. Confirm closed-loop stability separately if the open loop has right-half-plane poles or other crossovers.

Bode Plot Stability

The following terms support Bode construction and classical margin analysis:

1. Gain Margin: The gain margin is the loop-gain change needed to reach unity magnitude at a phase crossover. Bode plots normally show it in dB.
2. Phase Margin: The phase margin, expressed in degrees, is the additional phase lag needed to reach -180° at a gain crossover.
3. Gain Crossover Frequency: A frequency where loop magnitude equals one, or 0 dB.
4. Phase Crossover Frequency: A frequency where loop phase equals -180° modulo 360°.
5. Corner Frequency: The break frequency where adjacent straight-line magnitude asymptotes meet for a pole or zero factor.
6. Resonant Frequency: The frequency of a magnitude peak when a suitable underdamped factor produces one. Not every system has a resonant peak.
7. Factors: Factor the loop transfer function L(s) = G(s)H(s) into constant gain, poles or zeros at the origin, first-order terms and second-order terms.
8. Slope: Each pole or zero changes the asymptotic magnitude slope by a defined number of dB per decade after its break frequency.
9. Angle: Add the phase contribution of each factor in degrees to obtain the total phase curve.

These standard factors give the basic asymptotic magnitude slopes and exact phase contributions:

• Constant term K: A positive constant shifts magnitude by 20 log10 K dB, adds zero slope and contributes 0° phase. A negative real K also contributes ±180° phase.
• Integral factor 1/(jω)ⁿ: For integer n, this factor has slope -20n dB per decade at all frequencies and phase -90n°. It has no finite corner frequency.
• First-order factor 1/(1+jωT): This pole has a 0 dB/decade low-frequency asymptote and a -20 dB/decade high-frequency asymptote after the corner 1/T rad/s. Its exact phase is -tan^-1(ωT).
• First-order factor (1+jωT): This zero has a 0 dB/decade low-frequency asymptote and a +20 dB/decade high-frequency asymptote after the corner 1/T rad/s. Its exact phase is tan^-1(ωT).
• Second-order pole pair: 1/[1 + 2ζ(jω/ωn) + (jω)²/ωn²]: The high-frequency asymptote is -40 dB per decade and the natural break frequency is ωn rad s^-1. Damping ratio ζ controls any resonant peak and the phase transition shown here.

  

How to Draw Bode Plot

For a rational SISO loop transfer function, use this procedure to construct asymptotic Bode plots by hand:

1. Write the open-loop transfer function as L(s) = G(s)H(s), then substitute s = jω.
2. Factor L(s), identify poles and zeros at the origin, and tabulate all positive corner frequencies.
3. Choose a logarithmic frequency range that starts below the lowest corner and ends above the highest. Use separate magnitude and phase axes; mark magnitude in dB and phase in degrees, including the -180^° reference where needed.
4. Calculate the constant-gain offset and the initial slope from poles or zeros at the origin.
5. At each corner, add the slope change from its pole or zero factor and then calculate the exact phase contributions.

For the asymptotic Bode magnitude plot:

• Mark every corner frequency on the logarithmic frequency axis.
• Tabulate the factors in a consistent sequence.
  1. Constant gain K.
  2. Poles or zeros at the origin.
     
  3. First-order pole factors.
     
  4. First-order zero factors such as (1+jωT).
  5. Second-order pole or zero factors:

     

• Start from a known magnitude or the constant-gain intercept. Change the cumulative slope at each corner, then add exact corrections near corners if the true curve is required.
• Locate every -180° phase crossing and read the corresponding gain margins from the exact response.

For the Bode phase plot:

1. Calculate each factor’s phase and add the contributions with consistent unwrapping.
2. Evaluate the total phase at enough logarithmically spaced frequencies to show each transition, then draw the phase curve.
3. Locate every 0 dB gain crossing and read the corresponding phase margins.

Bode Stability Criterion

The classical Bode margin test assesses a nominal SISO negative-feedback loop from its open-loop response. It is reliable only when the Nyquist assumptions and all crossover frequencies are handled correctly.

Use the following interpretation after confirming the nominal closed loop and the open-loop right-half-plane pole count:

1. For a Stable System: Positive limiting gain and phase margins support the existing closed-loop stability result. They estimate separate gain and phase tolerances; they do not prove stability by themselves.
2. For a Marginally Stable System: A zero limiting margin places the loop at a classical stability boundary, subject to the crossover and pole-count assumptions.
3. For an Unstable or Fragile System: A negative limiting phase margin or an adverse gain margin signals that the loop has crossed a classical boundary. Never compare gain margin in dB with phase margin in degrees.

Advantages of a Bode Plot

Bode plots provide these practical benefits:

1. Straight-line asymptotes give a quick hand estimate of logarithmic magnitude and phase.
2. Multiplication becomes addition and division becomes subtraction on the dB magnitude scale, so factor contributions combine cleanly.
3. Crossover frequencies and margins show useful relative-robustness information when classical assumptions hold.
4. Bode plots show relative stability measures through gain margin and phase margin, while Nyquist or closed-loop poles establish absolute stability.
5. The logarithmic frequency axis covers many decades in a compact plot.