Professional Documents
Culture Documents
Bode Diagrams
We can write the frequency response function G(jω) in polar form as:
G ( jω ) = G ( jω ∠φ
where G ( jω ) is the gain of the system or, alternatively, the Amplitude Ratio, AR.
The amplitude ratio is a more informative term as it reminds us that the output can be
either amplified or attenuated i.e. the gain may be less than unity.
The argument φ = ∠G ( jω ) gives us the phase difference between the excitation
frequency and the output frequency. We are of course assuming that the system is
linear and consequently the response of the system to a sinusoidal input is a sinusoid
which has been amplified or attenuated and which has been phase shifted with respect
to the input.
Also, the AR and the phase difference between input and output are frequency
dependent. We often represent these relationships graphically as direct polar plots i.e.
AR vs. frequency and phase versus frequency. One form of this representation is the
so called Bode diagram which consists of two plots – AR vs. frequency and phase
angle vs. frequency.
An example of a Bode diagram is given in Figure 1. It shows the frequency response
of the following system:
s 2 + 0.1s + 7.5
G (s) = 4
s + 0.12 s 3 + 9 s 2
Note that the Bode diagram is plotted on log-linear graph paper with frequency
plotted logarithmically. Also, it is usual to plot the amplitude ratio AR in decibels
since it allows the diagram to represent a much greater dynamic range. Decibels are
defined:
dB = 20 log10 AR
While it is possible to plot the Bode diagram for a system using Matlab (e.g. using the
Control Systems Toolbox) another method is to use asymptotic approximations to plot
the AR. (Note that while asymptotic approximations are widely used to plot the AR
they are rarely used to plot the phase angle – this is normally done by tabulating the
phase angle and drawing a smooth curve through a selected number of points). The
method of asymptotic approximations is very convenient because the Bode plot for
each identifiable component in the overall transfer function can be plotted
independently and then the Bode diagram for the complete transfer function
determined by summing the components graphically.
To ease this process we will examine the frequency responses for the most common
functional blocks and plot their Bode diagrams.
We can determine asymptotic approximations to each of the above and use these
approximations to assist in sketching the Bode diagram for more complex transfer
functions. Note that to use these asymptotes we will have to transform the numerator
and denominator terms in the complex transfer function into the Bode forms given
above.
Determining Asymptotes
First let’s sketch the Bode plots for some of the above functional blocks and, in doing
so, determine the appropriate asymptotes (Note that not all are done here – you should
try to do the others yourself). We can also generalise from the case of a single block
to multiple similar blocks in series e.g. in a Type II system we will have two
integrators.
1 1
5. The simple lag . We find AR = and φ = − tan −1 ωτ . We will
1 + jωτ 1+ ω τ
2 2
need the low and high frequency asymptotes to sketch the amplitude response
of the simple lag. We divide the amplitude response into two distinct regions
delineated by the corner frequency, ωc which occurs at ωτ=1. For ωτ<<1,
AR≅1, so the low frequency asymptote is the 0 dB line on the Bode diagram.
1
For ωτ>>1, AR ≅ . In decibels, this is a line of slope –20 dB/ decade of
ωτ
frequency passing through the point ωτ=1, AR=1 (0dB). The phase lag is -45°
at the corner frequency and is bounded by -90° .
1 1 1
7. The PI Controller 1 + . We find AR = 1 + and φ = − tan −1 . At
jωτ (ωτ I ) 2
ωτ
1
low frequencies we can see that the term dominates the amplitude
(ωτ ) 2
1
response. We can make the approximation AR ≅ for ωτ << 1. When
ωτ
expressed in dB this is a line of slope –20dB which intersects the high
frequency asymptote at ωτ I = 1 .
1
8. A 2nd order term . Note that the amplitude and phase responses
τ s + 2ςτ s + 1
2 2
Figure 6 Bode diagrams for damping factors of 0.1, 0.25, 0.8, 1.2
Note the sharp transition in the phase plot at the natural frequency and the
magnification factor for under-damped 2nd order systems shown in the amplitude
response.