SECTION 2:

FIRST-ORDER FILTERS
ENGR 202 – Electrical Fundamentals II
2 Introduction

K. Webb ENGR 202

 Filters
3

 We are all familiar with water and air filters

 Basis for operation is size selectivity
 Small particles (e.g. air or water molecules) are allowed to pass
 Larger particles (e.g. dust, sediment) are not

 Unwanted components are filtered out of the flow.

 Electrical filters are similar

 Basis for operation is frequency selectivity
 Signal components in certain frequency ranges are filtered out
 Signal components at other frequencies are allowed to pass

K. Webb ENGR 202

 Noise
4

 All real-world electrical signals are noisy

 You’ve seen this in the lab
 Zoom in closely on a low-amplitude sinusoid with the
scope (even one supplied directly from the function
generator) – it won’t look like a perfectly clean sinusoid

K. Webb ENGR 202

 Noise
5

 We will use the term noise to mean any electrical signal

that interferes with or corrupts a signal we are trying to
measure.
 Noise has many sources:
 Measurement instruments themselves
 60Hz power line interference
 Electrical components – resistors, transistors, etc.
 Wireless LAN, fluorescent lights, computers, etc.

 We’d like to be able to remove, or filter out, this noise

 Improve the accuracy of measurements
 Often possible, if we know the frequency characteristics of
the signal and the noise

K. Webb ENGR 202

 Filtering Noise
6

 We’ll learn how to design filters to remove noise

Noisy Signal Filtered Signal

Filter

 First, we must introduce two important concepts:

 Frequency-domain representation of electrical signals
 What is meant by “frequency characteristics” of an electrical signal?
 Frequency response of linear systems
 How does a linear system (e.g. a filter) behave as a function of
frequency?
K. Webb ENGR 202
7 Frequency Spectrum

K. Webb ENGR 202

 Frequency Domain
8

 We are accustomed to looking at electrical signals in

the time domain
 Amplitude plotted as function of time

 Can also be represented in the frequency domain

 Amplitude plotted as a function of frequency
 Frequency spectrum
 Describes the frequency content of a signal
 Can think of signals as a sum of different frequency
sinusoids
 What frequencies (sinusoids) are present

K. Webb ENGR 202

 Frequency Spectrum
9

 Frequency spectrum
 An amplitude vs. frequency plot
 X-axis is frequency – not time

 Y-axis is amplitude

 Amplitude units may be in decibels (dB)

 Shows the relative amount of energy at each

frequency
 Time-domain plot and frequency spectrum are
alternate representations of the same signal

K. Webb ENGR 202

 Frequency Spectra – Examples
10

 Single sinusoid: 𝑣𝑣 𝑡𝑡 = 1𝑉𝑉 cos 2𝜋𝜋 ⋅ 800𝐻𝐻𝐻𝐻 ⋅ 𝑡𝑡

Time-domain Frequency-domain

 Sum of three sinusoids:

𝑣𝑣 𝑡𝑡 = 1𝑉𝑉[cos 2𝜋𝜋 ⋅ 800𝐻𝐻𝐻𝐻 ⋅ 𝑡𝑡 + cos 2𝜋𝜋 ⋅ 1200𝐻𝐻𝐻𝐻 ⋅ 𝑡𝑡 + cos 2𝜋𝜋 ⋅ 2000𝐻𝐻𝐻𝐻 ⋅ 𝑡𝑡

Time-domain Frequency-domain

K. Webb ENGR 202

 Frequency Spectra – Examples
11

 White noise:
Time-domain Frequency-domain

 Band-limited (colored) noise:

Time-domain Frequency-domain

K. Webb ENGR 202

 Frequency Spectra - Examples
12

 Consider the following scenario  Same signal in the frequency

 Measuring a sensor output in domain:
the lab
 Know the signal is roughly
sinusoidal
 Suspected frequency: ~1 kHz

 Three interfering tones

 All near 100 kHz
 Can now design a filter to
remove the noise:
 Measured signal corrupted by
noise/interference
 Difficult to identify the
interfering signal from the
time-domain plot
K. Webb ENGR 202
 Fourier Transform
13

 Fourier transform
 Transforms a time-domain representation to a frequency spectrum
∞
𝑉𝑉 𝜔𝜔 = � 𝑣𝑣 𝑡𝑡 𝑒𝑒 −𝑗𝑗𝑗𝑗𝑗𝑗 𝑑𝑑𝑑𝑑
−∞

 Inverse Fourier transform

 Transforms from the frequency domain to the time domain

1 ∞
𝑣𝑣 𝑡𝑡 = � 𝑉𝑉 𝜔𝜔 𝑒𝑒 𝑗𝑗𝑗𝑗𝑗𝑗 𝑑𝑑𝑑𝑑
2𝜋𝜋 −∞

 A mathematical transform
 Two different ways of looking at the same signal
 A change in perspective not a change of the signal itself

K. Webb ENGR 202

14 Frequency Response of Linear Systems

K. Webb ENGR 202

 Frequency Response Function
15

 Linear systems (electrical, mechanical, etc.) can be

described by their frequency responses
 Frequency response
 Ratio of the system output phasor to the system input phasor
 In general, a complex function of frequency
𝐘𝐘 𝐘𝐘 𝜔𝜔
𝐻𝐻 𝜔𝜔 = =
𝐗𝐗 𝐗𝐗 𝜔𝜔

 Complex-valued – has both magnitude and phase

 Magnitude: ratio of output to input magnitudes
 Gain of the system
 Phase: difference in phase between output and input
 Phase shift through the system
K. Webb ENGR 202
 Frequency Response – Bode Plots
16

 Frequency response
 Description of system behavior as a function of
frequency
 Gain and phase

 Represented graphically – formatted as a Bode plot

 Magnitude plot on top, phase plot below
 Logarithmic frequency axes

 Magnitude usually has units of decibels (dB)

 Phase has units of degrees

K. Webb ENGR 202

 Bode Plots
17

Units of
magnitude
are dB
Magnitude
plot on top

Logarithmic frequency axes

Units of Phase plot

phase are below
degrees

K. Webb ENGR 202

 Decibels - dB
18

 Frequency response gain most often expressed and

plotted with units of decibels (dB)
 A logarithmic scale
 Provides detail of very large and very small values on the
same plot
 Commonly used for ratios of powers or amplitudes

 Conversion from a linear scale to dB:

𝐻𝐻 𝜔𝜔 𝑑𝑑𝑑𝑑 = 20 ⋅ log10 𝐻𝐻 𝜔𝜔

 Conversion from dB to a linear scale:

𝐻𝐻 𝜔𝜔 𝑑𝑑𝑑𝑑
𝐻𝐻 𝜔𝜔 = 10 20

K. Webb ENGR 202

 Decibels – dB
19

 Multiplying two gain values corresponds to adding their

values in dB
 E.g., the overall gain of cascaded systems
𝐻𝐻1 𝜔𝜔 ⋅ 𝐻𝐻2 𝜔𝜔 𝑑𝑑𝑑𝑑 = 𝐻𝐻1 𝜔𝜔 𝑑𝑑𝑑𝑑 + 𝐻𝐻2 𝜔𝜔 𝑑𝑑𝑑𝑑

 Negative dB values corresponds to sub-unity gain

 Positive dB values are gains greater than one
dB Linear dB Linear
60 1000 6 2
40 100 -3 1/√2 = 0.707
20 10 -6 0.5
0 1 -20 0.1

K. Webb ENGR 202

 Value of Logarithmic Axes - dB
20

 Gain axis is linear in dB

 A logarithmic scale
 Allows for displaying detail at very large and very small levels on the same plot

 Gain plotted in dB
 Two resonant peaks
clearly visible

 Linear gain scale

 Smaller peak has
disappeared

K. Webb ENGR 202

 Value of Logarithmic Axes - Frequency
21

 Frequency axis is logarithmic

 Allows for displaying detail at very low and very high frequencies on the
same plot

 Log frequency axis

 Can resolve
frequency of both
resonant peaks

 Linear frequency
axis
 Lower resonant
frequency is unclear

K. Webb ENGR 202

 Interpreting Bode Plots
22

Bode plots tell you the gain and phase shift at all frequencies:
choose a frequency, read gain and phase values from the plot

For a 10MHz
For a 10KHz sinusoidal
sinusoidal input, the
input, the gain is -32dB
gain is 0dB (1) (0.025), and
and the phase the phase
shift is 0°. shift is -176°.

K. Webb ENGR 202

 Interpreting Bode Plots
23

K. Webb ENGR 202

24 Example Problems

K. Webb ENGR 202

 A measured signal has the
frequency spectrum shown
here. Assuming the larger
signal component has an
amplitude of 500 mV, and
that both signal components
are in phase, write a time-
domain expression for the
measured signal.

K. Webb ENGR 202

 Determine the frequency
response function, 𝐻𝐻 𝜔𝜔 ,
for the following circuit.
What are the circuit’s gain
and phase at 200 kHz?

K. Webb ENGR 202

K. Webb ENGR 202
 The input to a circuit with
the following Bode plot is
𝑣𝑣𝑖𝑖 𝑡𝑡 = 1.2𝑉𝑉 ⋅ cos 2𝜋𝜋 ⋅ 10𝑘𝑘𝑘𝑘𝑘𝑘 ⋅ 𝑡𝑡

What is the output, 𝑣𝑣𝑜𝑜 𝑡𝑡 ?

K. Webb ENGR 202

29 Types of Filters
Filters are classified by the ranges of
frequencies they pass and those they filter out

K. Webb ENGR 202

 Filter Operation
30

 Frequency spectrum describes frequency content of

electrical signals
 Frequency response describes system (circuit) gain and phase
at different frequencies

 Can design circuits (i.e. filters) to have high gain at desirable

frequencies and low gain at undesirable frequencies
 Want to filter out high frequencies?
 Design a filter with low gain at high frequencies and high gain at low
frequencies.

 Want to filter out all signals between 1 MHz and 10 MHz?

 Design a filter with low gain in this range and high gain everywhere else.

K. Webb ENGR 202

 Types of Filters
31

 Filters are classified according to the ranges of

frequencies they pass and those they filter out
 Low pass filters: pass low frequencies, filter out high
frequencies
 High pass filters: pass high frequencies, filter out low
frequencies
 Band pass filters: pass only a range of frequencies,
filter out everything else
 Band stop (notch) filters: filter out only a certain range
of frequencies, pass all others

K. Webb ENGR 202

 Ideal Filters
32

 Ideal filter gain characteristics:

 Unity gain in the pass band
 Range of frequencies to be passed
 Zero gain in the stop band
 Range of frequencies to be filtered out
 Abrupt transition between pass band and stop band

 Signals with frequency components in the pass

band pass through the filter unaltered
 Signals with frequency components in the stop
band are completely filtered out – removed from
the signal
K. Webb ENGR 202
 Ideal Filters – Magnitude Response
33

Ideal Low Pass Filter Ideal High Pass Filter

pass band pass band

stop band stop band

Ideal Band Pass Filter Ideal Band Stop Filter

pass stop
band band
stop band stop band pass band pass band

 Note the use of a linear gain scale here

 Stop band gain of zero translates to −∞ dB
 Ideal filters often referred to as brick wall filters
K. Webb ENGR 202
 Real Filters – Magnitude Response
34

Magnitude response for a real low pass filter:

Pass band edge is freq. at which gain is down by 3 dB – the -3 dB frequency.
This is the filter’s bandwidth.

pass band
stop band

Roll-off rate between pass band and stop band depends

on the type of filter – particularly, the order of the filter.

K. Webb ENGR 202

35 First-Order Passive Filters
First-order – only one energy-storage element
Passive – contain only resistors and capacitors
or inductors – no opamps or transistors

K. Webb ENGR 202

 Filters as Voltage Dividers
36

 Already familiar with resistive voltage dividers:

𝑅𝑅2
𝑣𝑣𝑜𝑜 = 𝑣𝑣𝑠𝑠
𝑅𝑅1 + 𝑅𝑅2
 Frequency response function:
𝐕𝐕𝑜𝑜 𝑅𝑅2
𝐻𝐻 𝜔𝜔 = =
𝐕𝐕𝑠𝑠 𝑅𝑅1 + 𝑅𝑅2
 A real constant – independent of frequency
 Same gain at all frequencies
 No phase shift at any frequency

 Now consider a circuit whose resistances have been replaced with impedances :
𝐕𝐕𝑜𝑜 𝑍𝑍2
𝐻𝐻 𝜔𝜔 = =
𝐕𝐕𝑠𝑠 𝑍𝑍1 + 𝑍𝑍2
 Frequency response is now a complex function of
frequency
 Gain and phase vary as a function of frequency
 Basis for the design of first-order filters
K. Webb ENGR 202
37 RC Low Pass Filter

K. Webb ENGR 202

 RC Low Pass Filter
38

 Now, let 𝑍𝑍1 be resistive and 𝑍𝑍2 be capacitive

 Frequency response:
1�
𝑉𝑉𝑜𝑜 𝑍𝑍2 𝑗𝑗𝑗𝑗𝑗𝑗
𝐻𝐻 𝜔𝜔 = = =
𝑉𝑉𝑠𝑠 𝑍𝑍1 + 𝑍𝑍2 𝑅𝑅 + 1�
𝑗𝑗𝑗𝑗𝑗𝑗

1
𝐻𝐻 𝜔𝜔 =
1 + 𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
 Recall from ENGR 201 that the transient response of this same circuit is
characterized by its time constant, 𝜏𝜏 = 𝑅𝑅𝑅𝑅
 In the frequency domain, this is the corner frequency or break frequency

1 1 1
𝜔𝜔𝑐𝑐 = = and 𝑓𝑓𝑐𝑐 =
𝜏𝜏 𝑅𝑅𝑅𝑅 2𝜋𝜋𝜋𝜋𝜋𝜋

 The frequency at which gain is down by 3 dB

 The -3 dB frequency
 Frequency at which the magnitude of R and C impedances are equal
K. Webb ENGR 202
 RC Low Pass Filter
39

 To gain insight into the behavior of this filter circuit,

consider two limiting cases
 As 𝑓𝑓 → 0,
 Capacitor→ open circuit
 𝑖𝑖 𝑡𝑡 → 0
 𝑣𝑣𝑜𝑜 → 𝑣𝑣𝑠𝑠
 Gain → unity

 As 𝑓𝑓 → ∞
 Capacitor → short circuit
 𝑣𝑣𝑜𝑜 shorted to ground
 Gain → zero
K. Webb ENGR 202
 RC Low Pass Filter – Bode Plot
40

K. Webb ENGR 202

 RC Low Pass Filter – Magnitude Response
41

Low-frequency
asymptote
at 0dB
Response rolls off at
-20dB/decade, or
Gain is -3 dB at 𝑓𝑓𝑐𝑐 , the -6dB/octave above 𝑓𝑓𝑐𝑐
bandwidth of the
filter
𝑓𝑓𝑐𝑐 = 200 𝑘𝑘𝑘𝑘𝑘𝑘

K. Webb ENGR 202

 RC Low Pass Filter – Phase Response
42

Low-frequency
asymptote One decade below 𝑓𝑓𝑐𝑐
at 0°

𝑓𝑓𝑐𝑐 = 200 𝑘𝑘𝑘𝑘𝑘𝑘

Phase approximation rolls
off at -45°/decade around 𝑓𝑓𝑐𝑐

Phase is -45° at 𝑓𝑓𝑐𝑐

High-frequency
asymptote
One decade above 𝑓𝑓𝑐𝑐 at -90°

K. Webb ENGR 202

 RC Low Pass Filter – Magnitude Response
43

 Known slope can be used to relate

gains at different frequencies

𝑑𝑑𝑑𝑑 𝐻𝐻 𝑓𝑓2 𝑑𝑑𝑑𝑑 − 𝐻𝐻 𝑓𝑓1 𝑑𝑑𝑑𝑑

𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 =
𝑑𝑑𝑑𝑑𝑑𝑑 log10 𝑓𝑓2 − log10 𝑓𝑓1
-20dB/dec
𝐻𝐻 1𝑀𝑀𝑀𝑀𝑀𝑀 = −14𝑑𝑑𝑑𝑑  For example:

𝑑𝑑𝑑𝑑 𝐻𝐻 7𝑀𝑀𝑀𝑀𝑀𝑀 − 𝐻𝐻 1𝑀𝑀𝑀𝑀𝑀𝑀 𝑑𝑑𝑑𝑑

−20 =
𝐻𝐻 7𝑀𝑀𝑀𝑀𝑀𝑀 = ?? 𝑑𝑑𝑑𝑑𝑑𝑑 log10 7𝑀𝑀𝑀𝑀𝑀𝑀 − log10 1𝑀𝑀𝑀𝑀𝑀𝑀

𝑑𝑑𝑑𝑑 𝐻𝐻 7𝑀𝑀𝑀𝑀𝑀𝑀 − −14𝑑𝑑𝑑𝑑

−20 =
𝑑𝑑𝑑𝑑𝑑𝑑 6.845 − 6

𝐻𝐻 7𝑀𝑀𝑀𝑀𝑀𝑀 = −30.9𝑑𝑑𝑑𝑑

K. Webb ENGR 202

 RC LP Filter – Application Example
44

 Simple first-order RC low pass filters provide a quick

and easy way to remove noise from electrical
signals

 Consider for example a dual-tone multi-frequency

(DTMF) signal
 Touch-tone telephone signal (key 5 in this example)
 Tone is the sum of two sinusoids (key 5 = 1336Hz and
770Hz)
 Pressing the “5” key generates the DTMF signal
 Noise on the DTMF signal makes decoding impossible
 Filter noise to enabling decoding
K. Webb ENGR 202
 RC LP Filter – Application Example
45

 Key number 5 is pressed

 DTMF signal generated
 Sum of 770 Hz and 1336 Hz sinusoids

 Decoder at the receiving end decodes the DTMF signal and

determines that a 5 was pressed
K. Webb ENGR 202
 RC LP Filter – Application Example
46

 Consider a more realistic scenario

 DTMF signal corrupted by a significant amount of noise

 The decoder is no longer able to determine that a 5 was pressed

K. Webb ENGR 202

 RC LP Filter – Application Example
47

 The goal is to filter the received signal so that the decoder

can accurately interpret the DTMF signal

Filter

 Designing the low pass filter

 White noise
 Flat frequency spectrum
 Equal power at all frequencies
 DTMF frequency range: 697 Hz – 1633 Hz
 Want to attenuate as much noise as possible
 Want to attenuate DTMF signals as little as possible
 RC LPF with corner frequency at 10 kHz will limit DTMF-band
attenuation to < 0.2 dB
K. Webb ENGR 202
 RC LP Filter – Application Example
48

 RC LPF design
 Need to select R and C to set the
corner frequency
1
𝑓𝑓𝑐𝑐 = = 10 𝑘𝑘𝑘𝑘𝑘𝑘
2𝜋𝜋𝜋𝜋𝜋𝜋

 Say we have a 0.1 𝜇𝜇F capacitor available

 Solve for R
1
𝑅𝑅 =
2𝜋𝜋𝑓𝑓𝐶𝐶 𝐶𝐶
1
𝑅𝑅 = = 159 Ω
2𝜋𝜋 ⋅ 10 𝑘𝑘𝑘𝑘𝑘𝑘 ⋅ 0.1𝜇𝜇𝜇𝜇

𝑅𝑅 = 159 Ω, 𝐶𝐶 = 0.1 𝜇𝜇𝜇𝜇

K. Webb ENGR 202
 RC LP Filter – Application Example
49

Bode plot of the resulting filter:

DTMF signals lie in noise attenuated

this range – passed
through the filter
with little
attenuation

K. Webb ENGR 202

 RC LP Filter – Application Example
50

 Filter allows DTMF signal to pass mostly unaltered

 Noise below 10 kHz is mostly passed through
 Noise above 10 kHz is attenuated, but not removed
 Received signal is not noiseless, but clean enough to be decoded

K. Webb ENGR 202

51 Example Problems

K. Webb ENGR 202

 Design a filter to pass the
desired 500 Hz signal and to
attenuate the unwanted 100
kHz by 40 dB.
What is the signal-to-noise
ratio (SNR) at the output of
the filter?

K. Webb ENGR 202

K. Webb ENGR 202
K. Webb ENGR 202
55 RC High Pass Filter

K. Webb ENGR 202

 RC High Pass Filter
56

 Now, swap the locations of the resistor and capacitor

 Frequency response:
𝑉𝑉𝑜𝑜 𝑍𝑍2 𝑅𝑅
𝐻𝐻 𝜔𝜔 = = =
𝑉𝑉𝑠𝑠 𝑍𝑍1 + 𝑍𝑍2 𝑅𝑅 + 1�
𝑗𝑗𝑗𝑗𝑗𝑗
𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
𝐻𝐻 𝜔𝜔 =
1 + 𝑗𝑗𝑗𝑗𝑗𝑗𝑗𝑗
 Corner frequency is the same as for the low pass filter
1 1 1
𝜔𝜔𝑐𝑐 = = and 𝑓𝑓𝑐𝑐 =
𝜏𝜏 𝑅𝑅𝑅𝑅 2𝜋𝜋𝜋𝜋𝜋𝜋
 The frequency at which gain is down by 3 dB
 Frequency at which the capacitor impedance magnitude is equal to the
resistor impedance magnitude
 Now, gain is constant above 𝑓𝑓𝑐𝑐 and rolls off below 𝑓𝑓𝑐𝑐

K. Webb ENGR 202

 RC High Pass Filter
57

 To gain insight into the behavior of this filter circuit,

consider two limiting cases
 As 𝑓𝑓 → 0,
 Capacitor→ open circuit
 𝑖𝑖 𝑡𝑡 → 0
 𝑣𝑣𝑜𝑜 → 0
 Gain → zero

 As 𝑓𝑓 → ∞
 Capacitor → short circuit
 𝑣𝑣𝑜𝑜 shorted to 𝑣𝑣𝑠𝑠
 Gain → unity
K. Webb ENGR 202
 RC High Pass Filter – Bode Plot
58

K. Webb ENGR 202

 RC High Pass Filter – Magnitude Response
59

High-frequency
asymptote
at 0 dB
Gain is -3dB at 𝑓𝑓𝑐𝑐

𝑓𝑓𝑐𝑐 = 1 𝑘𝑘𝑘𝑘𝑘𝑘

Response rolls off at

20 dB/decade, or
6 dB/octave below 𝑓𝑓𝑐𝑐

K. Webb ENGR 202

 RC High Pass Filter – Phase Response
60

Low-frequency One decade below 𝑓𝑓𝑐𝑐

asymptote
at +90°
Phase approximation rolls off
at -45°/decade around 𝑓𝑓𝑐𝑐

Phase is +45° at 𝑓𝑓𝑐𝑐

High-frequency
𝑓𝑓𝑐𝑐 = 1 𝑘𝑘𝑘𝑘𝑘𝑘
asymptote
at 0°

One decade above 𝑓𝑓𝑐𝑐

K. Webb ENGR 202

 RC HP Filter – Application Example
61

 High pass filters are useful for removing low-frequency

content, including DC, from electrical signals.

 For example, consider the following scenario:

 Instrumented a flow loop in the lab
 Pumps, temperature sensors, pressure sensors, and flow meters
 Flow meter output seems to be erroneous every ~1 msec
 Suspected cause: coupled through the +12V power supply from one of
the pumps
 Want to measure the flow meter’s +12V power supply with a channel on
our data acquisition system (DAQ)
 Dynamic range of DAQ input: ±5 V
 Use a high-pass filter to remove the +12V DC component from the
power supply voltage
K. Webb ENGR 202
 RC HP Filter – Application Example
62

 Want to a +12 V supply with a ±5 V DAQ input

 High pass filter will remove the DC component of the supply voltage

 High pass filter used to remove DC signal components

 Couples only AC signal components to the DAQ input
 AC coupling
 Similar to the AC coupling setting on the scopes in the lab

K. Webb ENGR 202

 RC HP Filter – Application Example
63

 High pass filter design

 Want to remove DC
 Low corner frequency
 High RC time constant
 Large R and C
 Arbitrarily set 𝑓𝑓𝑐𝑐 = 10 𝐻𝐻𝐻𝐻

 DAQ system
 Datasheet says 𝑅𝑅𝑖𝑖𝑖𝑖 = 10 𝑀𝑀Ω
 Let 𝑅𝑅𝑖𝑖𝑖𝑖 be the filter resistance
 Calculate C to get desired 𝑓𝑓𝑐𝑐
1 1
𝐶𝐶 = =
2𝜋𝜋𝑓𝑓𝑐𝑐 𝑅𝑅 2𝜋𝜋 ⋅ 10𝐻𝐻𝐻𝐻 ⋅ 10𝑀𝑀Ω
𝐶𝐶 = 15.9 𝑛𝑛𝑛𝑛
 Or anything in that neighborhood
 Not critical – just want to block DC
K. Webb ENGR 202
 RC HP Filter – Application Example
64

RC high pass filter: High pass filter Bode plot:

The +12V DC component of the power

supply voltage is completely removed.

K. Webb ENGR 202

 RC HP Filter – Application Example
65

The noisy +12V power supply at the malfunctioning flow meter:

 DC value of signal is +12 V

 Outside ±5 V DAQ input
dynamic range

High pass filter output – AC coupled power supply voltage:

 DC value of signal is now 0 V

 Within ±5 V DAQ input
range
 Glitches clearly measured
with the DAQ

K. Webb ENGR 202

 Oscilloscopes – AC Coupling
66

 Scope inputs allow you to select between DC and AC coupling

 Usually under the channel menu
 DC coupling: input signal is terminated in 1MΩ and connected directly
to the preamp and ADC in the scope
 AC coupling: input signal is switched through a capacitor that forms a
high pass filter with the 1MΩ input resistor
 𝑓𝑓𝑐𝑐 ≈ 3.5 𝐻𝐻𝐻𝐻 – removes DC
 Useful for looking at power supply ripple, etc.

K. Webb ENGR 202

 Oscilloscopes – AC Coupling
67

High-impedance
scope front-end:

Configured for
DC coupling:

Configured for
AC coupling:

K. Webb ENGR 202

68 RL Filters

K. Webb ENGR 202

 First-order RL filters
69

 Can also use inductors to make RL low pass and high pass filters
 Capacitors are usually preferable for simple first-order filters
 Smaller
 Cheaper
 Draw no DC current

RL low pass filter: RL high pass filter:

𝑅𝑅 𝑅𝑅
Corner frequency: 𝑓𝑓𝑐𝑐 = Corner frequency: 𝑓𝑓𝑐𝑐 =
2𝜋𝜋𝜋𝜋 2𝜋𝜋𝜋𝜋

K. Webb ENGR 202

 RL Low Pass Filter
70

 RL low pass filter

 Frequency response:
𝑉𝑉𝑜𝑜 𝑍𝑍2 𝑅𝑅
𝐻𝐻 𝜔𝜔 = = =
𝑉𝑉𝑠𝑠 𝑍𝑍1 + 𝑍𝑍2 𝑅𝑅 + 𝑗𝑗𝑗𝑗𝑗𝑗
𝑅𝑅
𝐻𝐻 𝜔𝜔 =
𝑅𝑅 + 𝑗𝑗𝑗𝑗𝐿𝐿
 Corner frequency is one over the time constant
1 𝑅𝑅 𝑅𝑅
𝜔𝜔𝑐𝑐 = = and 𝑓𝑓𝑐𝑐 =
𝜏𝜏 𝐿𝐿 2𝜋𝜋𝜋𝜋
 The frequency at which gain is down by 3 dB
 Frequency at which the inductor impedance magnitude is equal to the
resistor impedance magnitude
 Bode plot identical to that of the RC low pass filter
 As it is for all first-order low pass systems
K. Webb ENGR 202
 RL Low Pass Filter
71

 Again consider the filter’s behavior for two limiting

cases
 As 𝑓𝑓 → 0,
 Inductor → short circuit
 𝑣𝑣𝑜𝑜 shorted to 𝑣𝑣𝑠𝑠
 Gain → unity

 As 𝑓𝑓 → ∞
 Inductor → open circuit
 𝑖𝑖 𝑡𝑡 → 0
 𝑣𝑣𝑜𝑜 → 0
 Gain → zero

K. Webb ENGR 202

 RL High Pass Filter
72

 Now, swap the locations of the resistor and inductor

 Frequency response:
𝑉𝑉𝑜𝑜 𝑍𝑍2 𝑗𝑗𝑗𝑗𝑗𝑗
𝐻𝐻 𝜔𝜔 = = =
𝑉𝑉𝑠𝑠 𝑍𝑍1 + 𝑍𝑍2 𝑅𝑅 + 𝑗𝑗𝑗𝑗𝑗𝑗

𝑗𝑗𝑗𝑗𝑗𝑗
𝐻𝐻 𝜔𝜔 =
𝑅𝑅 + 𝑗𝑗𝑗𝑗𝐿𝐿

 Corner frequency is the same as for the low pass filter

1 𝑅𝑅 𝑅𝑅
𝜔𝜔𝑐𝑐 = = and 𝑓𝑓𝑐𝑐 =
𝜏𝜏 𝐿𝐿 2𝜋𝜋𝜋𝜋

 Bode plot is identical to that of the RC high pass filter

 Gain is constant above 𝑓𝑓𝑐𝑐 and rolls off below 𝑓𝑓𝑐𝑐
K. Webb ENGR 202
 RL High Pass Filter
73

 Again, consider the two limiting frequency cases

 As 𝑓𝑓 → 0,
 Inductor → short circuit
 𝑣𝑣𝑜𝑜 shorted to ground
 Gain → zero

 As 𝑓𝑓 → ∞
 Inductor → open circuit
 𝑖𝑖 𝑡𝑡 → 0
 𝑣𝑣𝑜𝑜 → 𝑣𝑣𝑠𝑠
 Gain → unity
K. Webb ENGR 202
74 Audio Filter Demo

K. Webb ENGR 202

 Analog Discovery Instrument
75

 2-chan. Scope
 14-bit, 100MSa/s
 5MHz bandwidth
 2-chan. function generator
 14-bit, 100MSa/s
 5MHz bandwidth
 2-chan. spectrum analyzer
 Network analyzer
 Voltmeter
 ±5V power supplies
 16-chan. logic analyzer
 16-chan. digital pattern
generator
 USB connectivity

K. Webb ENGR 202

 Analog Discovery – Audio Demo
76

 Demo board plugs

in to Analog
Discovery module
 Summation of
multiple tones
 Optional filtering
of audio signal
 3.5 mm audio
output jack

K. Webb ENGR 202

 Analog Discovery – Audio Demo
77

K. Webb ENGR 202