772 � A!

"#$% F#&"%'(

15–1 B ASIC F ILTE R R E S PO N S E S

Filters are usually categorized by the manner in which the output voltage varies with
the frequency of the input voltage. The categories of active filters are low-pass, high-
pass, band-pass, and band-stop. Each of these general responses are examined.
After completing this section, you should be able to
� Describe and analyze the gain-versus-frequency responses of basic types of filters
� Describe low-pass filter response
� Define passband and critical frequency � Determine the bandwidth

� Define pole � Explain roll-off rate and define its unit � Calculate the critical

frequency
� Describe high-pass filter response
� Explain how the passband is limited � Calculate the critical frequency

� Describe band-pass filter response

� Determine the bandwidth � Determine the center frequency � Calculate the

quality factor (Q)

� Describe band-stop filter response
� Determine the bandwidth

Low-Pass Filter Response

A filter is a circuit that passes certain frequencies and attenuates or rejects all other fre-
quencies. The passband of a filter is the range of frequencies that are allowed to pass
through the filter with minimum attenuation (usually defined as less than -3 dB of attenu-
ation). The critical frequency, fc, (also called the cutoff frequency) defines the end of the
passband and is normally specified at the point where the response drops -3 dB (70.7%)
from the passband response. Following the passband is a region called the transition region
that leads into a region called the stopband. There is no precise point between the transi-
tion region and the stopband.
A low-pass filter is one that passes frequencies from dc to fc and significantly attenu-
ates all other frequencies. The passband of the ideal low-pass filter is shown in the
blue-shaded area of Figure 15–1(a); the response drops to zero at frequencies beyond the
passband. This ideal response is sometimes referred to as a “brick-wall” because nothing
gets through beyond the wall. The bandwidth of an ideal low-pass filter is equal to fc.
Equation 15–1 BW 5 fc

The ideal response shown in Figure 15–1(a) is not attainable by any practical filter.
Actual filter responses depend on the number of poles, a term used with filters to describe
the number of RC circuits contained in the filter. (The term pole has a number of meanings
in electrical work, but this definition is specifically for filters. In mathematics, it is a value
for which a function goes to infinity, such as when you divide by zero.) The most basic
low-pass filter is a simple RC circuit consisting of just one resistor and one capacitor; the
output is taken across the capacitor as shown in Figure 15–1(b). This basic RC filter has
a single pole, and it rolls off at -20 dB/decade beyond the critical frequency. The actual
response is indicated by the blue line in Figure 15–1(a). The response is plotted on a stand-
ard log plot that is used for filters to show details of the curve as the gain drops. Notice that
the gain drops off slowly until the frequency is at the critical frequency; after this, the gain
drops rapidly.
The -20 dB/decade roll-off rate for the gain of a basic RC filter means that at a fre-
quency of 10fc, the output will be -20 dB (10%) of the input. This roll-off rate is not a par-
ticularly good filter characteristic because too much of the unwanted frequencies (beyond
the passband) are allowed through the filter.

, ,

M15_FLOY2998_10_GE_C15.indd 772 03/08/17 5:46 PM

 B)(#! F#&"%' R%(*+,(%( � 773

Gain (normalized to 1)

0 dB
–3 dB Actual response of a
single-pole RC filter
Passband
–20 dB
Transition R Vout
region
–40 dB –2
0
BW dB Vs
/d C
Stopband ec
ad
region e
–60 dB f
0.01 fc 0.1 fc fc 10 fc 100 fc 1000 fc
(a) Comparison of an ideal low-pass filter response (blue area) with actual response. (b) Basic low-pass circuit

Gain (normalized to 1)

0 dB

–20 dB
–2
0
dB
–4

/d
– 60

ec
0d

ad

–40 dB
B

e
dB/d
/de
ca
ecad

de
e

–60 dB f
0.01 fc 0.1 fc fc 10 fc 100 fc 1000 fc
(c) Idealized low-pass filter responses
� F IGU R E 1 5 – 1
Low-pass filter responses.

The critical frequency of a low-pass RC filter occurs when XC = R, where

1
fc =
2pRC

Recall from your basic dc/ac studies that the output at the critical frequency is 70.7% of the
input. This response is equivalent to an attenuation of -3 dB.
Figure 15–1(c) illustrates three idealized low-pass response curves including the basic
one-pole response (-20 dB/decade). The approximations show a flat response to the cut-
off frequency and a roll-off at a constant rate after the cutoff frequency. Actual filters do
not have a perfectly flat response up to the cutoff frequency but drop to -3 dB at this point
as described previously.
The simple one-pole RC filter in Figure 15–1 is a passive filter because it is composed
only of passive components. More poles can be added to increase the steepness of the
transition region, but the downside is that the accuracy of the filter is less due to loading
effects. A better way to produce a filter that has a steeper transition region is to add active
circuitry (an amplifier) to the basic filter. Responses that are steeper than -20 dB/decade
in the transition region cannot be obtained by simply cascading identical RC stages (due to
loading effects). However, by combining an op-amp with frequency-selective feedback
circuits, filters can be designed with roll-off rates of -40, -60, or more dB/decade. Filters
that include one or more op-amps in the design are called active filters. These filters can
optimize the roll-off rate or other attribute (such as phase response) with a particular filter
design. In general, the more poles the filter uses, the steeper its transition region will be.
The exact response depends on the type of filter and the number of poles.

, ,

M15_FLOY2998_10_GE_C15.indd 773 03/08/17 5:46 PM

 774 � A!"#$% F#&"%'(

High-Pass Filter Response

A high-pass filter is one that significantly attenuates or rejects all frequencies below fc
and passes all frequencies above fc. The critical frequency is, again, the frequency at which
the output is 70.7% of the input (or -3 dB) as shown in Figure 15–2(a). The ideal response,
indicated by the blue-shaded area, has an instantaneous drop at fc,which, of course, is not
achievable. Ideally, the passband of a high-pass filter is all frequencies above the critical
frequency. The high-frequency response of practical circuits is limited by the finite band-
width of active components and unwanted stray capacitance in components that make up
the filter.

Gain (normalized to 1)

0 dB
–3 dB
Actual response
Passband
of a single-pole
– 20 dB RC filter
C
e Vout
c ad
e
– 40 dB /d
dB
0 Vs
–2 R

– 60 dB f
0.001 fc 0.01 fc 0.1 fc fc 10 fc 100 fc
(a) Comparison of an ideal high-pass filter response (blue area) with actual response (b) Basic high-pass circuit

Gain (normalized to 1)

0 dB

– 20 dB
e
ad
ec
/d

de
dB

e
eca
ecad
0
–2

– 40 dB
B/d
dB/d
0d
–4
– 60

– 60 dB f
0.001 fc 0.01 fc 0.1fc fc 10 fc 100 fc
(c) Idealized high-pass filter responses
� FIG U R E 15– 2
High-pass filter responses.

A simple RC circuit consisting of a single resistor and capacitor can be configured as a

high-pass filter by taking the output across the resistor as shown in Figure 15–2(b). As in
the case of the low-pass filter, the basic RC circuit has a roll-off rate of -20 dB/decade, as
indicated by the blue line in Figure 15–2(a). Also, the critical frequency for the basic high-
pass filter occurs when XC = R, where
1
fc =
2pRC

Figure 15–2(c) illustrates three idealized high-pass response curves, including the basic
one-pole response (-20 dB/decade) for a high-pass RC circuit. As in the case of the low-
pass filter, the approximations show a flat response to the cutoff frequency and a roll-off at

, ,

M15_FLOY2998_10_GE_C15.indd 774 03/08/17 5:46 PM

 B)(#! F#&"%' R%(*+,(%( � 775

a constant rate after the cutoff frequency. Actual high-pass filters do not have the perfectly
flat response indicated or the precise roll-off rate shown. Responses that are steeper than
-20 dB/decade in the transition region are also possible with both passive and active high-
pass filters; the particular response depends on the type of filter and the number of poles.
In general, active filters suffer from fewer loading effects.

Band-Pass Filter Response

A band-pass filter passes all signals lying within a band between a lower-frequency limit
and an upper-frequency limit and essentially rejects all other frequencies that are outside
this specified band. A generalized band-pass response curve is shown in Figure 15–3. The
bandwidth (BW) is defined as the difference between the upper critical frequency ( fc2) and
the lower critical frequency ( fc1).
BW 5 fc2 2 fc1 Equation 15–2
The critical frequencies are, of course, the points at which the response curve is 70.7% of
its maximum. Recall from Chapter 12 that these critical frequencies are also called 3 dB
frequencies. The frequency about which the passband is centered is called the center fre-
quency, f0, defined as the geometric mean of the critical frequencies.

f0 5 2fc1 fc2 Equation 15–3

Vout (normalized to 1)

0.707

BW

f
fc1 f0 fc2

� F IGUR E 1 5 – 3
General band-pass response curve.

Quality Factor The quality factor (Q) of a band-pass filter is the ratio of the center
frequency to the bandwidth.
f0
Q5 Equation 15–4
BW

The value of Q is an indication of the selectivity of a band-pass filter. The higher the
value of Q, the narrower the bandwidth and the better the selectivity for a given value
of f0. Band-pass filters are sometimes classified as narrow-band (Q 7 10) or wide-band
(Q 6 10). The quality factor (Q) can also be expressed in terms of the damping factor
(DF) of the filter as
1
Q=
DF

You will study the damping factor in Section 15–2.

, ,

M15_FLOY2998_10_GE_C15.indd 775 03/08/17 5:46 PM

 776 � A!"#$% F#&"%'(

EXAMPLE 15–1 A certain band-pass filter has a center frequency of 15 kHz and a bandwidth of 1 kHz.
Determine Q and classify the filter as narrow-band or wide-band.

f0 15 kHz
Solution Q= = = 15
BW 1 kHz
Because Q 7 10, this is a narrow-band filter.

Related Problem* If the quality factor of the filter is doubled, what will the bandwidth be?

*Answers can be found at [Link]/Floyd.

Band-Stop Filter Response

Another category of active filter is the band-stop filter, also known as notch, band-reject,
or band-elimination filter. You can think of the operation as opposite to that of the band-
pass filter because frequencies within a certain bandwidth are rejected, and frequencies
outside the bandwidth are passed. A general response curve for a band-stop filter is shown
in Figure 15–4. Notice that the bandwidth is the band of frequencies between the 3 dB
points, just as in the case of the band-pass filter response.
� FIG U R E 15– 4 Gain (dB)
General band-stop filter response.
0
–3

f
fc1 f0 fc2

BW

SECTION 15–1 1. What determines the bandwidth of a low-pass filter?

CHECKUP 2. What limits the passband of an active high-pass filter?
Answers can be found at www
.[Link]/Floyd.
3. How are the Q and the bandwidth of a band-pass filter related? Explain how the selec-
tivity is affected by the Q of a filter.

15–2 F I LTE R R E S PO N S E C HA R AC T E RI S T I C S
Each type of filter response (low-pass, high-pass, band-pass, or band-stop) can be tai-
lored by circuit component values to have either a Butterworth, Chebyshev, or Bessel
characteristic. Each of these characteristics is identified by the shape of the response
curve, and each has an advantage in certain applications.

, ,

M15_FLOY2998_10_GE_C15.indd 776 03/08/17 5:46 PM

 F#&"%' R%(*+,(% C-)')!"%'#("#!( � 777

After completing this section, you should be able to

� Describe three types of filter response characteristics and other parameters
� Discuss the Butterworth characteristic
� Describe the Chebyshev characteristic
� Discuss the Bessel characteristic
� Define damping factor
� Calculate the damping factor � Show the block diagram of an active filter

� Analyze a filter for critical frequency and roll-off rate

� Explain how to obtain multi-order filters � Describe the effects of cascading on

roll-off rate

Butterworth, Chebyshev, or Bessel response characteristics can be realized with most

active filter circuit configurations by proper selection of certain component values. A gen-
eral comparison of the three response characteristics for a low-pass filter response curve
is shown in Figure 15–5. High-pass and band-pass filters can also be designed to have any
one of the characteristics.

Av # F I G U R E 15– 5
Comparative plots of three types of
filter response characteristics.

Butterworth
Bessel
Chebyshev

The Butterworth Characteristic The Butterworth characteristic provides a very flat

amplitude response in the passband and a roll-off rate of -20 dB/decade/pole. The phase
response is not linear, however, and the phase shift (thus, time delay) of signals passing
through the filter varies nonlinearly with frequency. Therefore, a pulse applied to a filter
with a Butterworth response will cause overshoots on the output because each frequency
component of the pulse’s rising and falling edges experiences a different time delay. Filters
with the Butterworth response are normally used when all frequencies in the passband
must have the same gain. The Butterworth response is often referred to as a maximally flat
response.
The Chebyshev Characteristic Filters with the Chebyshev response characteristic are
useful when a rapid roll-off is required because it provides a roll-off rate greater than
-20 dB/decade/pole. This is a greater rate than that of the Butterworth, so filters can be
implemented with the Chebyshev response with fewer poles and less complex circuitry for
a given roll-off rate. This type of filter response is characterized by overshoot or ripples in
the passband (depending on the number of poles) and an even less linear phase response
than the Butterworth.

, ,

M15_FLOY2998_10_GE_C15.indd 777 03/08/17 5:46 PM

 778 � A!"#$% F#&"%'(

The Bessel Characteristic The Bessel response exhibits a linear phase characteristic,
meaning that the phase shift increases linearly with frequency. The result is almost no over-
shoot on the output with a pulse input. For this reason, filters with the Bessel response are
used for filtering pulse waveforms without distorting the shape of the waveform.

The Damping Factor

As mentioned, an active filter can be designed to have either a Butterworth, Chebyshev, or
Bessel response characteristic regardless of whether it is a low-pass, high-pass, band-pass,
or band-stop type. The damping factor (DF) of an active filter circuit determines which
response characteristic the filter exhibits. To explain the basic concept, a generalized active
filter is shown in Figure 15–6. It includes an amplifier, a negative feedback circuit, and a
filter section. The amplifier and feedback are connected in a noninverting configuration.
The damping factor is determined by the negative feedback circuit and is defined by the
following equation:
R1
Equation 15–5 DF 5 2 2
R2

� FIG U R E 15– 6
General diagram of an active filter. Frequency-
Vin selective +
RC circuit
Vout
–
Amplifie r R1

R2

Negative feedback circuit

Basically, the damping factor affects the filter response by negative feedback action.
Any attempted increase or decrease in the output voltage is offset by the opposing effect
of the negative feedback. This tends to make the response curve flat in the passband of the
filter if the value for the damping factor is precisely set. By advanced mathematics, which
we will not cover, values for the damping factor have been derived for various orders of
filters to achieve the maximally flat response of the Butterworth characteristic.
The value of the damping factor required to produce a desired response characteristic
depends on the order (number of poles) of the filter. In mathematics, a pole is a point at
which a function approaches infinity. For a filter, poles are determined by the resistors and
capacitors present; for example a filter circuit with one resistor and one capacitor used to
affect the frequency response is a one pole filter. The more poles a filter has, the faster its
roll-off rate is. To achieve a second-order Butterworth response, for example, the damping
factor must be 1.414. To implement this damping factor, the feedback resistor ratio must be
R1
= 2 - DF = 2 - 1.414 = 0.586
R2

This ratio gives the closed-loop gain of the noninverting amplifier portion of the filter,
Acl(NI), a value of 1.586, derived as follows:
1 1 R1 + R2 R1
Acl(NI) = = = = + 1 = 0.586 + 1 = 1.586
B R2 >(R1 + R2) R2 R2

, ,

M15_FLOY2998_10_GE_C15.indd 778 03/08/17 5:46 PM

 F#&"%' R%(*+,(% C-)')!"%'#("#!( � 779

EXAMPLE 15–2 If resistor R2 in the feedback circuit of an active single-pole filter of the type in
Figure 15–6 is 10 kV, what value must R1 be to obtain a maximally flat Butterworth
response?

R1
Solution = 0.586
R2
R1 = 0.586R2 = 0.586(10 kV) = 5.86 kV

Using the nearest standard 5% value of 5.6 kV will get very close to the ideal
Butterworth response.

Related Problem What is the damping factor for R2 = 10 kV and R1 = 5.6 kV?

Critical Frequency and Roll-Off Rate

The critical frequency is determined by the values of the resistors and capacitors in the
frequency-selective RC circuit shown in Figure 15–6. For a single-pole (first-order) filter,
as shown in Figure 15–7, the critical frequency is
1
fc =
2pRC

Although we show a low-pass configuration, the same formula is used for the fc of a
single-pole high-pass filter. The number of poles determines the roll-off rate of the filter.
A Butterworth response produces -20 dB/decade/pole. So, a first-order (one-pole) fil-
ter has a roll-off of -20 dB/decade; a second-order (two-pole) filter has a roll-off rate of
-40 dB/decade; a third-order (three-pole) filter has a roll-off rate of -60 dB/decade; and
so on.

Single-pole # F I G U R E 15– 7
low-pass circuit
First-order (one-pole) low-pass filter.
R
Vin +

C Vout
– R1

R2

Generally, to obtain a filter with three poles or more, one-pole or two-pole filters are
cascaded, as shown in Figure 15–8. To obtain a third-order filter, for example, cascade a
second-order and a first-order filter; to obtain a fourth-order filter, cascade two second-
order filters; and so on. Each filter in a cascaded arrangement is called a stage or section.
Because of its maximally flat response, the Butterworth characteristic is the most
widely used. Therefore, we will limit our coverage to the Butterworth response to
illustrate basic filter concepts. Table 15–1 lists the roll-off rates, damping factors, and
feedback resistor ratios for up to sixth-order Butterworth filters. Resistor designations
correspond to the gain-setting resistors in Figure 15–8 and may be different on other
circuit diagrams.

, ,

M15_FLOY2998_10_GE_C15.indd 779 03/08/17 5:46 PM

 780 � A!"#$% F#&"%'(

RC RC RC
Vin + + +
circuit circuit circuit
Vout
– R1 – R3 – R5

R2 R4 R6

� FIG U R E 15– 8
The number of filter poles can be increased by cascading.

� TA BLE 15– 1
Values for the Butterworth response.

1ST STAGE 2ND STAGE 3RD STAGE

ROLL-OFF
ORDER DB/DECADE POLES DF R 1 /R 2 POLES DF R 3 /R 4 POLES DF R 5 /R 6
1 –20 1 Optional
2 –40 2 1.414 0.586
3 –60 2 1.00 1 1 1.00 1
4 –80 2 1.848 0.152 2 0.765 1.235
5 –100 2 1.00 1 2 1.618 0.382 1 0.618 1.382
6 –120 2 1.932 0.068 2 1.414 0.586 2 0.518 1.482

SECTION 15–2 1. Explain how Butterworth, Chebyshev, and Bessel responses differ.
CHECKUP 2. What determines the response characteristic of a filter?
3. Name the basic parts of an active filter.

15–3 A CTI VE L OW -P A SS F I LT E RS
Filters that use op-amps as the active element provide several advantages over pas-
sive filters (R, L, and C elements only). The op-amp provides gain, so the signal is
not attenuated as it passes through the filter. The high input impedance of the op-amp
prevents excessive loading of the driving source, and the low output impedance of
the op-amp prevents the filter from being affected by the load that it is driving. Active
filters are also easy to adjust over a wide frequency range without altering the desired
response.
After completing this section, you should be able to
� Identify and analyze active low-pass filters
� Identify a single-pole low-pass filter circuit
� Determine the closed-loop voltage gain � Determine the critical frequency

� Identify a Sallen-Key low-pass filter circuit

� Describe the filter operation � Calculate the critical frequency

� Analyze cascaded low-pass filters

� Explain how the roll-off rate is affected

, ,

M15_FLOY2998_10_GE_C15.indd 780 03/08/17 5:47 PM