ECExplain

Troubleshooting

Common Errors in Bode Plot Sketching

Learn to avoid the most frequent mistakes in Bode plot sketching, including incorrect slope counting, wrong corner frequency placement, and phase plot errors.

Visual

Wrong ApproachCorrect ApproachvsSlope +20 dB/decPole adds gainWrong magnitudeSlope -20 dB/decPole reduces gainCorrect sketch

Topic overview

A Bode plot represents the frequency response of a linear system using two separate plots: magnitude in decibels versus log frequency, and phase in degrees versus log frequency. Errors in Bode sketching typically stem from misreading the transfer function form, calculating corner frequencies incorrectly, or applying the wrong slope increments. Since each pole and zero contributes independently, a single misidentified term corrupts the entire sketch.

Common mistakes

Mistake 1: Not converting the transfer function to standard Bode form before reading corner frequencies
Why it happens: Corner frequencies are read as the value of omega where a factor equals zero, but this only works if the transfer function is in the Bode standard form where each first-order factor is written as (1 + s/omega_c). Students read pole locations from the denominator roots directly (e.g., s = -10) and use 10 rad/s as a corner frequency without converting.
Correct approach: Write each first-order factor in the form (1 + s/omega_c). The corner frequency is omega_c. For a pole at s = -10, the denominator factor is (s + 10) = 10*(1 + s/10), so omega_c = 10 rad/s. The factor of 10 goes into the DC gain K.
Example: Wrong: H(s) = 1/((s+10)(s+100)). Student sets corner frequencies at s=10 and s=100 without normalizing. Uses K=1. Slope changes drawn at wrong magnitudes. Correct: H(s) = 1/(10*(1+s/10)*100*(1+s/100)) = 1/1000 * 1/((1+s/10)(1+s/100)). DC gain K = 1/1000 = -60 dB. Corner frequencies: 10 rad/s and 100 rad/s. Slope -20 dB/dec after 10 rad/s, -40 dB/dec after 100 rad/s.
Mistake 2: Adding +20 dB/decade slope for a pole instead of -20 dB/decade
Why it happens: A zero in the numerator adds +20 dB/dec. A pole in the denominator adds -20 dB/dec. Students flip the sign when they write the factor from the denominator and forget to negate the slope contribution.
Correct approach: For every pole at omega_c, the magnitude slope decreases by 20 dB/decade after omega_c. For every zero at omega_c, the slope increases by 20 dB/decade after omega_c. A double pole changes slope by -40 dB/dec.
Example: Wrong: H(s) has poles at omega=10 and omega=100. Student draws slope: 0 dB/dec below 10, +20 dB/dec from 10 to 100, +40 dB/dec above 100. Correct: 0 dB/dec below 10, -20 dB/dec from 10 to 100, -40 dB/dec above 100.
Mistake 3: Drawing the phase plot as a step function instead of a gradual transition centered at the corner frequency
Why it happens: The asymptotic Bode phase approximation uses a linear transition that starts one decade before the corner frequency and ends one decade after. Students draw an abrupt 90-degree step exactly at omega_c.
Correct approach: For a single pole, the phase starts at 0 degrees well below omega_c, transitions linearly from -45 deg/decade starting at 0.1*omega_c, passes through -45 degrees at omega_c, and levels off at -90 degrees at 10*omega_c.
Example: Wrong: Phase = 0 deg for omega < 10, phase = -90 deg for omega >= 10 (step at corner frequency). Correct: Phase = 0 deg for omega < 1 (one decade before omega_c=10). Linear ramp from 0 to -90 deg between omega=1 and omega=100. Phase = -90 deg for omega > 100.
Mistake 4: Ignoring the contribution of a pure integrator (pole at s=0) to the low-frequency magnitude
Why it happens: A factor of 1/s (pole at origin) does not have a corner frequency. It contributes a -20 dB/decade slope for all frequencies and -90 degrees of phase at all frequencies. Students start drawing from 0 dB at low frequencies and only add the integrator's effect after some corner frequency.
Correct approach: A pole at the origin contributes -20 dB/decade starting from the lowest frequency on the plot. The magnitude line at omega=1 rad/s passes through 20*log10(K) dB where K is the system gain excluding the integrator. Phase starts at -90 degrees and decreases further with additional poles.
Example: Wrong: H(s) = 10/(s*(1+s/100)). Student draws 0 dB flat line below omega=100. Correct: At omega=1, magnitude = 20*log10(10/1) = 20 dB. Slope = -20 dB/dec for all omega < 100. After omega=100 (pole at 100 rad/s), slope becomes -40 dB/dec. Phase = -90 deg for all omega < 10, then further transitions to -180 deg.

Debugging tips

  • Before sketching, list every pole and zero in a table: location, type (real/complex), and the slope change it introduces. Sum the slopes from left to right across the frequency axis.
  • Check the DC gain or low-frequency gain first. Substitute s=0 (or omega approaching 0) into the transfer function. The magnitude at very low frequency should match 20*log10(|H(0)|).
  • For the phase plot, mark the frequencies omega_c/10 and 10*omega_c for every corner frequency. The linear phase transition must start and end at these marks, not at omega_c itself.
  • Count total poles minus total zeros. The high-frequency slope in dB/decade is -20*(poles - zeros). Verify your sketch asymptotes to this slope at the far right.
  • If the transfer function has complex conjugate poles or zeros, the corner frequency is the natural frequency omega_n, not the pole location in the s-plane. The quality factor Q affects the peaking near omega_n, which is not shown in asymptotic Bode plots.

Exam warnings

  • Problems sometimes give H(s) = K*s^n / (denominator) where s^n in the numerator represents n zeros at the origin. Each zero at the origin adds +20 dB/dec to the initial slope and +90 degrees to the initial phase. A factor s^2 in the numerator means the magnitude starts with +40 dB/dec slope and phase starts at +180 degrees.
  • When a Bode plot is given and the question asks to find the transfer function, read the initial slope to determine poles or zeros at the origin, then read each slope change at corner frequencies to identify additional poles and zeros. A common trap is a -60 dB/dec high-frequency slope that students interpret as 3 poles when it could be 1 zero and 4 poles.
  • Phase margin is read from the phase plot at the gain crossover frequency, which is where the magnitude plot crosses 0 dB. Students sometimes read it at the phase crossover frequency instead. These are two different frequencies.
  • For a second-order system with low damping ratio, the magnitude Bode plot has a peak near omega_n. Asymptotic Bode approximation does not show this peak. If the question asks for the actual peak magnitude, use the formula M_p = 1/(2*zeta*sqrt(1-zeta^2)) and do not read from the asymptotic sketch.