Double Sideband (DSB) and Amplitude Modulation (AM)

Modules: Audio Oscillator, Wideband True RMS Meter, Multiplier, Adder, Utilities, Phase
Shifter, Tuneable LPF, Quadrature Utilities, Noise Generator, Speech, Headphones

0 Pre-Laboratory Reading

Double sideband (DSB) is one of the easiest modulation techniques to understand, so it is a good
starting point for the study of modulation. A type of DSB, called binary phase-shift keying, is
used for digital telemetry. Amplitude modulation (AM) is similar to DSB but has the advantage
of permitting a simpler demodulator, the envelope detector. AM is used for broadcast radio,
aviation radio, citizens’ band (CB) radio, and short-wave broadcasting.

0.1 Double Sideband (DSB) Modulation

A message signal 𝑥(𝑡) can be DSB modulated onto a carrier with a simple multiplication. The
modulated carrier 𝑦(𝑡) can be represented by

𝑦(𝑡) = 𝑥(𝑡) ∙ cos⁡(2𝜋𝑓𝑐 𝑡) (1)

where 𝑓𝑐 is the carrier frequency. The Fourier transform 𝑌(𝑓) of the modulator output is related
to the Fourier transform 𝑋(𝑓) of the message signal by

1 1
𝑌(𝑓) = 𝑋(𝑓 − 𝑓𝑐 ) + 𝑋(𝑓 + 𝑓𝑐 ) (2)
2 2
For a real 𝑥(𝑡), |𝑋(𝑓)| is symmetric about 𝑓 = 0 and |𝑋(𝑓 − 𝑓𝑐 )| is symmetric about 𝑓 = 𝑓𝑐 .
For the present discussion, it is only necessary to consider that part of 𝑌(𝑓) that lies in the
positive half of the frequency axis. The signal content that lies in the frequency domain below 𝑓𝑐
is the lower sideband. The signal content that lies in the frequency domain above 𝑓𝑐 is the upper
sideband. This is the origin of the term double sideband.

When studying and testing analog modulation schemes, it is convenient to use a sinusoid as the
message signal. This is a good choice for several reasons. First, when testing a system in the
laboratory, it is desirable to use a periodic signal since a stable oscilloscope display with
continuous signal capture is then possible. Second, the mathematics are usually simpler with a
sinusoidal message signal. Third, a sinusoid is easy to generate in the laboratory. After a
communication system goes into the field and becomes operational, the message signal would
not ordinarily be sinusoidal, of course.

If 𝑥(𝑡) is a sinusoid of frequency 𝑓𝑚 , the modulated carrier can be written as

𝑦(𝑡) = cos(2𝜋𝑓𝑚 𝑡) ∙ cos(2𝜋𝑓𝑐 𝑡)⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡

1
 1 1
= cos[2𝜋(𝑓𝑐 − 𝑓𝑚 )𝑡] + cos⁡[2𝜋(𝑓𝑐 + 𝑓𝑚 )𝑡] (3)
2 2
This last expression indicates that when a carrier is DSB modulated by a message sinusoid, the
modulated carrier is equivalent to the sum of two sinusoids: one having the difference frequency
𝑓𝑐 − 𝑓𝑚 and the other with the sum frequency 𝑓𝑐 + 𝑓𝑚 . In the frequency domain there is signal
content lying on both sides of the carrier frequency. When 𝑥(𝑡) is a sinusoid, each sideband is
one discrete spectral line (on the positive half of the frequency axis).

In the general case where 𝑥(𝑡) is real but not a sinusoid, each sideband is more complicated than
a single discrete spectral line. However, it is still true that there are two sidebands: a lower and
an upper.

In the time domain, a carrier that is DSB modulated by a message sinusoid looks like this:

DSB modulated carrier (solid curve) and message sinusoid (dashed curve)

A DSB modulated carrier is normally demodulated with a synchronous detector. This means
that the modulated carrier is multiplied by a local oscillator and the product is then sent to a low-
pass filter. With synchronous detection the frequency and phase of the local oscillator are
important. Its frequency must match that of the carrier. The local oscillator phase must
approximately match that of the carrier.

When the modulated carrier 𝑦(𝑡) of Eq. (1) is sent to a synchronous detector, the demodulator
(detector) output 𝑧(𝑡) can be written as

𝑧(𝑡) = ⁡𝒮{𝑦(𝑡) ∙ cos⁡(2𝜋𝑓𝑐 𝑡 + 𝜙)}⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡

= ⁡𝒮{𝑥(𝑡) ∙ cos⁡(2𝜋𝑓𝑐 𝑡) ∙ cos⁡(2𝜋𝑓𝑐 𝑡 + 𝜙)}⁡⁡⁡⁡⁡⁡⁡

1 1
= ⁡𝒮 { 𝑥(𝑡) cos(𝜙) + 𝑥(𝑡)cos⁡(4𝜋𝑓𝑐 𝑡 + 𝜙)} (4)
2 2

2
The local oscillator is modeled as cos⁡(2𝜋𝑓𝑐 𝑡 + 𝜙). It has a frequency that matches that of the
carrier. However, allowance is made for a phase that might not match that of the carrier. The
low-pass filtering is represented by 𝒮{∙}.

synchronous detector
modulator

message x x PF
LPF

carrier local
oscillator

DSB modulator (left) and synchronous detector as demodulator (right)

The message signal 𝑥(𝑡) is presumably low pass. The term 𝑥(𝑡)cos⁡(4𝜋𝑓𝑐 𝑡 + 𝜙) represents a
bandpass signal centered at 2𝑓𝑐 in the frequency domain. The bandwidth of the low-pass filter
should be large enough that 𝑥(𝑡) passes through the filter with little distortion. If the bandwidth
of 𝑥(𝑡) is small compared with 2𝑓𝑐 (as it will be in this experiment and as it normally is in
practice), then a bandwidth for the low-pass filter can easily be selected that passes 𝑥(𝑡) but
blocks 𝑥(𝑡)cos⁡(4𝜋𝑓𝑐 𝑡 + 𝜙). The demodulator output is, in this case,

𝐾LPF
𝑧(𝑡) = ⁡ cos(𝜙) 𝑥(𝑡) (5)
2
where 𝐾LPF is the lowpass gain of the filter. There will also be a time delay in going through the
filter; this is not indicated in Eq. (5) because it is not important for the present discussion. The
filter output is a scaled and delayed version of 𝑥(𝑡). A demodulation that produces a scaled and
delayed version of the original message is considered successful.

An important consideration is the phase difference 𝜙 between the local oscillator and the carrier.
The amplitude of the demodulator output is a function of 𝜙. If 𝜙 = ±90° , 𝑧(𝑡) is zero. It is
essential that 𝜙 remain close to 0° so that the message not suffer fading at the output of the
demodulator. Keeping 𝜙 close to 0° is normally accomplished within the receiver by a circuit
that performs carrier synchronization.

0.2 Amplitude Modulation (AM)

If a DC component is added to the message signal 𝑥(𝑡) before multiplication with a carrier, then
the modulation scheme is known as amplitude modulation (AM). The purpose of the DC
component is to permit the modulated carrier to be demodulated at the receiver by a means other
than synchronous detection. More will be said about this later.

3
 message + x
DC

carrier

Modulator for AM

The output of the modulator can be written:

[𝑐1 + 𝑐2 𝑥(𝑡)]cos⁡(2𝜋𝑓𝑐 𝑡)

Where 𝑐1 and 𝑐2 are constants.

For AM the maximum of the absolute value of the message signal 𝑥(𝑡) is of importance; this is
denoted here by 𝑝 (𝑝 ≥ 0).

𝑝 = max⁡|𝑥(𝑡)| (6)
It is useful to define a normalized message signal 𝑥𝑛 (𝑡) as follows:

𝑥𝑛 (𝑡) = 𝑥(𝑡)⁄𝑝 (7)

The maximum of the absolute value of 𝑥𝑛 (𝑡) is therefore

max⁡|𝑥𝑛 (𝑡)| = 1 (8)

This is what is meant by normalization. The modulator output can be rewritten as

[𝑐1 + 𝑐2 𝑝𝑥𝑛 (𝑡)]cos⁡(2𝜋𝑓𝑐 𝑡)

Factoring 𝑐1 out of the bracket and defining

𝜇 = 𝑐2 𝑝⁄𝑐1 (9)
as the modulation index for AM, the modulator output can then be written

𝑐1 [1 + 𝜇𝑥𝑛 (𝑡)]cos⁡(2𝜋𝑓𝑐 𝑡)

The modulation index of Eq. (9) is always non-negative (that is, positive or 0) since 𝑝 is positive
(by definition) and the constants 𝑐1 and 𝑐2 should always have the same sign.

4
The term 𝑐1 cos⁡(2𝜋𝑓𝑐 𝑡) is a residual carrier and the term 𝑐1 𝜇𝑥𝑛 (𝑡)cos⁡(2𝜋𝑓𝑐 𝑡) is a DSB-
modulated carrier. The ratio of the peak value of the DSB term to the peak value of the residual
carrier is
𝑐1 𝜇
| | = 𝜇⁡
𝑐1

In the above, we recognize that 𝜇 is positive. The ratio of the peak value of the DSB term to the
peak value of the residual carrier is the modulation index.

There are some AM calculations that involve the ratio of the DSB term to the residual carrier.
For this class of problem, it is acceptable to forget about the common factor 𝑐1. The important
problem of selecting an appropriate value for the modulation index 𝜇 is an example of this class
of problem. Therefore, in the analysis that appears below, we sometimes forget about 𝑐1 and use
the following definition:

𝑦(𝑡) = [1 + 𝜇𝑥𝑛 (𝑡)]cos⁡(2𝜋𝑓𝑐 𝑡) (10)

There are, of course, problems for which the constant 𝑐1 must be considered. One problem of
this type is when we must determine if any stage in the transmitter is overloaded (the voltage is
so large that the stage is unintentionally operated in a nonlinear regime).

As was done with DSB, the next step in this description of AM is to consider a sinusoidal
message signal.

𝑥𝑛 (𝑡) = cos⁡(2𝜋𝑓𝑚 𝑡) (11)

Eq. (11) satisfies Eq. (8). Three cases are of interest:

𝜇=1 100% modulation

𝜇<1 undermodulation

𝜇>1 overmodulation

The time-domain view of the modulated carrier is shown below for these three cases. It helps
when interpreting these figures to consider the quantity 1 + 𝜇𝑥𝑛 (𝑡). This quantity has a
minimum value of 0 for 𝜇 = 1. The minimum value is positive and non-zero for 𝜇 < 1. The
minimum value is negative for 𝜇 > 1. (It is assumed here that min 𝑥(𝑡) = −𝑝 and therefore that
min 𝑥𝑛 (𝑡) = −1.)

5
𝜇 = 1 (100% modulation): AM carrier (solid curve) and message sinusoid (dashed curve)

𝜇 < 1 (undermodulation): AM carrier (solid curve) and message sinusoid (dashed curve)

𝜇 > 1 (overmodulation): AM carrier (solid curve) and message sinusoid (dashed curve)

6
In the laboratory, the modulation index can be set with good accuracy when the message signal
is a sinusoid. The rms value of 𝑐1 𝜇𝑥𝑛 (𝑡) and the rms value of 𝑐1 are separately measured. For
the 𝑥𝑛 (𝑡) of Eq. (11), these rms values are |𝑐1 | 𝜇 ⁄√2 and |𝑐1 |, respectively. The ratio of these
two rms values is:

ratio⁡of⁡rms⁡message⁡component⁡to⁡rms⁡of⁡DC⁡component = 𝜇 ⁄√2 (sinusoid) (12)

A desired modulation index 𝜇 can be set by adjusting the gain on the (sinusoidal) message signal
until the ratio equals 𝜇 ⁄√2. It must be emphasized that Eq. (12) is valid only for a sinusoidal
message signal.

There is a technique for estimating the modulation index of an AM carrier that is quite general; it
does not require the message signal to be a sinusoid. This technique uses the XY view of an
oscilloscope. The multiplier input that is not the carrier is applied to Channel A. In the case of
DSB, this is 𝑥(𝑡). In the case of AM, this is 1 + 𝜇𝑥𝑛 (𝑡). To Channel B is applied the output of
the multiplier (that is, the modulated carrier). When this is done, one of the following figures is
displayed:

DSB AM: 𝜇 = 1 AM: 𝜇 < 1 AM: 𝜇 > 1

100% modulation undermodulation overmodulation

Consider first the DSB figure. The modulated carrier 𝑥(𝑡)cos⁡(2𝜋𝑓𝑐 𝑡) is plotted against 𝑥(𝑡). At
any instant when the message signal is 0, the modulated carrier is also zero. The center of the
“bowtie” represents this point; it is the origin. When 𝑥(𝑡) takes on the value 𝑥 ′ , the modulated
carrier can take on any value between −|𝑥 ′ | and |𝑥 ′ |. The larger |𝑥 ′ |, the wider the range of
values that the modulated carrier can assume.

For AM, the modulated carrier [1 + 𝜇𝑥𝑛 (𝑡)]cos⁡(2𝜋𝑓𝑐 𝑡) is plotted against 1 + 𝜇𝑥𝑛 (𝑡). If 𝜇 = 1,
then 1 + 𝜇𝑥𝑛 (𝑡) takes on only non-negative values, so only the right half of the “bowtie” is
present. If 𝜇 < 1, then the minimum value of 1 + 𝜇𝑥𝑛 (𝑡) is a positive and non-zero value. If
𝜇 > 1, then the minimum value of 1 + 𝜇𝑥𝑛 (𝑡) is negative.

The nice thing about using the XY view to estimate the modulation index is that this works for
any message signal, even an audio signal (from a microphone, for example). The problem of
calculating the modulation index for a general audio signal is that it can be difficult even to

7
identify 𝑝 and therefore 𝑥𝑛 (𝑡) for a signal that appears random. The XY view provides a quick
but approximate characterization of the AM.

It is worthwhile to consider how an AM carrier looks in the frequency domain. This is done here
for the case of a sinusoidal message signal. Eqs. (10) and (11) together can be rewritten as

𝑦(𝑡) = cos(2𝜋𝑓𝑐 𝑡) + 𝜇 cos(2𝜋𝑓𝑚 𝑡) cos(2𝜋𝑓𝑐 𝑡)⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡

𝜇
⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡⁡= cos(2𝜋𝑓𝑐 𝑡) + cos[2𝜋(𝑓𝑐 − 𝑓𝑚 )𝑡]
2
𝜇 (13)
+ cos⁡[2𝜋(𝑓𝑐 + 𝑓𝑚 )𝑡]
2
It is clear from Eq. (13) that the spectrum of an AM carrier with sinusoidal message signal
contains three discrete spectral lines (on the positive half of the frequency axis). The frequencies
are: 𝑓𝑐 , 𝑓𝑐 − 𝑓𝑚 , and 𝑓𝑐 + 𝑓𝑚 . This is similar to the case of DSB, except for the additional
discrete spectral line at 𝑓𝑐 , which is known as a residual carrier. Each of the sideband lines has a
line height that is different from that of the residual carrier by

sideband⁡line⁡height⁡relative⁡to⁡residual⁡carrier = 20 log(𝜇 ⁄2) ⁡⁡dBc (14)

The unit dBc is the number of decibels relative to the residual carrier. For 100% modulation
(𝜇 = 1), each sideband is −6 dBc.

Demodulation of an AM carrier is normally done with an envelope detector. An envelope

detector, implemented as a rectifier followed by a low-pass filter, extracts the envelope of the
modulated carrier. For the modulated carrier of Eq. (10), the envelope is |1 + 𝜇𝑥𝑛 (𝑡)|.

rectifier PF
LPF

Demodulator for AM

If 𝜇 ≤ 1, 1 + 𝜇𝑥𝑛 (𝑡) is always non-negative, so the envelope is just 1 + 𝜇𝑥𝑛 (𝑡). With AC
coupling (to remove the DC component) a scaled (and delayed) version of the original message
is recovered.

With overmodulation (𝜇 > 1), the envelope |1 + 𝜇𝑥𝑛 (𝑡)| does not equal 1 + 𝜇𝑥𝑛 (𝑡). In this
case, only a distorted version of the message can be recovered. However, if the overmodulation
is slight, then the distortion is slight.

8
Ideally, AM should be used with 𝜇 close to 1. If 𝜇 is significantly greater than 1, then distortion
occurs in envelope detection at the receiver. If 𝜇 is significantly less than 1, then the system is
inefficient because an unnecessarily large fraction of the transmitted power is devoted to the
residual carrier, which contains no information about the message. It is the sidebands that
contain the information about the message; the residual carrier is only present to make distortion-
free envelope detection possible at the receiver.

Envelope detection, unlike synchronous detection, requires no carrier synchronization circuit in

the receiver. This is why AM was selected over DSB for broadcast radio. The goal was to
permit the many receivers of broadcast radio to be simple (and therefore cheap) and reliable.

1 DSB Modulation and Synchronous Detection

You will build a DSB modulator. This will be paired with a synchronous detector.

Initially, you will use a 5-kHz sinusoid as the message signal. Adjust the frequency of the Audio
Oscillator to approximately 5 kHz and place this message signal at one of the inputs of the
Multiplier. It is necessary that the Multiplier be set for DC coupling. Connect a 100-kHz
sinusoid (Master Signals) to the second input of the Multiplier. The output of the Multiplier is
the modulated carrier.

Simultaneously observe the message signal on Channel A and the modulated carrier on Channel
B of the oscilloscope. Use the TTL output from the Audio Oscillator as an external trigger
source. This will stabilize the envelope, but you will see that the fast carrier oscillation inside
the envelope is not stabilized. This is a consequence of the fact that the message sinusoid and
the 100-kHz sinusoid are not coherently related. (The message sinusoid has a frequency that is
only approximately 5 kHz; it is not exactly 100 kHz divided by 20.)

Channel A: 5-kHz message sinusoid

Channel B: DSB carrier

Switch to an XY View (Views > X-Axis > A). Verify that this view gives the correct
picture for DSB. (In order to obtain the desired XY display, the message signal must be on
Channel A and the modulated carrier on Channel B.)

Channel A: 5-kHz message sinusoid

Channel B: DSB carrier

Observe the DSB carrier on the spectrum analyzer. Note the frequencies of the two tall spectral
lines.

Channel B: DSB carrier

9
We want to generate a local oscillator for the demodulator that has a variable phase relative to
the 100-kHz carrier. Place the 100-kHz sinusoid that you used for the DSB modulator on the
input of a Phase Shifter. Using the oscilloscope to view the Phase Shifter input and output,
adjust the Phase Shifter for a phase difference between input and output of 0° . (In other words,
the output will be in phase with the input.) This phase difference will be changed later.

Now build a synchronous detector. This will serve as the DSB demodulator. (A synchronous
detector is a type of demodulator. When discussing DSB, we will use the terms demodulator and
detector to mean the same thing.) You will need a multiplier for this demodulator; we will call
this the detector’s multiplier. You should use one of the multipliers in the Quadrature Utilities
module for this purpose. Connect the modulated carrier to one input of the detector’s multiplier.
Use the output of the Phase Shifter as the local oscillator, so this phase-shifted 100-kHz sinusoid
should be connected to the second input of the detector’s multiplier.

You will use a Tuneable LPF in the demodulator. Before placing it in the demodulator, adjust its
bandwidth to approximately 6 kHz. (The Tuneable LPF’s clock output has a frequency equal to
100 times the bandwidth.) Use the Noise Generator module to get a quick display of |𝐻(𝑓)|.

Channel A: Tuneable LPF output, showing |𝐻(𝑓)|

Connect the output of the detector’s multiplier to the Tuneable LPF.

Simultaneously observe the original 5-kHz message sinusoid (coming out of the Audio
Oscillator) and the Tuneable LPF output on the oscilloscope. Use the TTL output from the
Audio Oscillator as an external trigger source. It should be possible to stabilize simultaneously
both sinusoids in the display since they share a common frequency. You may want to increase
the gain of the Tuneable LPF in order to get a nice, strong signal out of the detector.

Channel A: 5-kHz message sinusoid from Audio Oscillator

Channel B: demodulator output (0° phase offset)

Notice that the detector has reproduced the original message, with some scaling and a delay. (Of
course there is delay, there is always delay in passing through a filter.) In judging whether a

10
demodulation is successful, we ask whether the shape of the message signal is reproduced on the
demodulator’s output. The amplitude of the demodulator’s output is not important here (as long
as it is not too small); in a practical receiver, the demodulator’s output can always be rescaled
with an amplifier.

You will now record the amplitude of the detector’s output for a 0° phase offset of the 100-kHz
local oscillator from the 100-kHz sinewave that is used in the modulator.

DSB

Now adjust the Phase Shifter so that the output lags the input by 90° . To do this, you will
simultaneously observe both the input and the output of the Phase Shifter on the oscilloscope,
adjusting the phase of the Phase Shifter until the phase difference is 90° . The XY view is helpful
for this. (When the phase offset is 90° , the ellipse axes will be horizontal and vertical.) You
should note, however, that the XY view does not tell you which sinusoid is leading and which is
lagging.

You should not change the gain of the Tuneable LPF before making this measurement. We want
the LPF gain to be the same for the two different phase offsets: 0° and 90° .

Simultaneously observe the original 5-kHz message sinusoid (coming out of the Audio
Oscillator) and the Tuneable LPF output on the oscilloscope. Use the TTL output from the
Audio Oscillator as an external trigger source.

Channel A: 5-kHz message sinusoid from Audio Oscillator

Channel B: demodulator output (90° phase offset)

Now with a 90° phase offset between local oscillator and carrier, record the amplitude of the
detector’s output.

As you can see, synchronous detection requires not only that the local oscillator match the carrier
in frequency but also (at least approximately) in phase. This is an important issue in receiver
design. For the purpose of this laboratory, the receiver has “stolen” a copy of the (unmodulated)
carrier from the transmitter. This is feasible as long as both transmitter and receiver are sitting in
the same laboratory. In the field, it is generally impossible for the receiver to steal a copy of the
(unmodulated) carrier from the transmitter. Instead, the receiver must perform carrier
synchronization.

11
2 AM and Envelope Detection

You will build an AM modulator. This will be paired with an envelope detector.

2.1 Sinusoidal Message

Build a modulator for AM. Use a 5-kHz sinusoid from the Audio Oscillator as the message
signal. Use a weighted adder to combine the message signal and a DC component. The DC
component will come from the Variable DC panel, and the knob on this panel should be set
clockwise from the vertical position, so that a positive DC value is supplied. Connect the output
of the Adder to a Buffer Amplifier and the output of this amplifier to a multiplier (set for DC
coupling). The other input to the multiplier will be a 100-kHz sinusoid (Master Signals).

In the Adder, the gain on each input is negative. The purpose of the Buffer Amplifier, with its
negative gain, is to cancel minus signs so that the net gain through Adder and amplifier is
positive for both the DC component and the message signal. Care should be taken that the
absolute value of the gain is not set too high. If it is too high, saturation (overloading) can result.

Set the modulation index 𝜇 to 1. According to Equation (12), a modulation index of 1 is

achieved by having the rms voltage of the message component equal to 0.707 times the rms
voltage of the DC component. You will measure rms voltages by connecting the Buffer
Amplifier output to the Wideband True RMS Meter. You will adjust the rms voltages using the
following procedure. First, disconnect the message signal from the Adder, so that only the DC
component is present. Second, adjust the gain applied by the Adder to the DC component until
the output of the Buffer Amplifier is 1.00 V rms. Third, reconnect the message signal to the
Adder input and disconnect the DC component from the Adder, so that only the message
component is present. Fourth, adjust the gain applied by the Adder to the message component
until the output of the Buffer Amplifier is 0.707 V rms. Fifth, reconnect the DC component to
the Adder. The ratio of the rms voltage of the message component to that of the DC component
should now be approximately 1⁄√2, so that 𝜇 = 1.

Connect the 5-kHz message sinusoid to Channel A and the AM modulator output to Channel B.
Use the TTL output from the Audio Oscillator as the trigger source. This will stabilize the
display of the envelope but not the fast carrier fluctuations inside the envelope.

12
 Channel A: 5-kHz message sinusoid
Channel B: AM carrier (𝜇 = 1)

Place the Buffer Amplifier output on Channel A and the AM modulator output on Channel B.
Switch to an XY view and observe the resulting figure. You should see the characteristic shape
for 100% modulation (𝜇 = 1).

Channel A: Buffer Amplifier output

Channel B: AM carrier (𝜇 = 1)

Observe the spectrum of the modulated carrier. You should see three strong spectral lines: the
residual carrier at 100 kHz, a lower sideband, and an upper sideband. The lower and upper
sidebands should have approximately the same level. You should use a logarithmic vertical
scale, having units of dBu. Make a note of the upper sideband line height in dBu and the
residual-carrier line height in dBu. Subtract the residual-carrier line height (dBu) from the upper
sideband line height (dBu). The difference is the upper sideband line height in dBc (decibels
relative to the carrier). This should be a negative number, since in this case the upper sideband
has a smaller level than the residual carrier.

Channel B: AM carrier (𝜇 = 1)

In addition to recording the measured upper sideband line height in dBc, also record the line
height as predicted by Eq. (14).

AM sinusoidal message

You will use a Tuneable LPF in the AM demodulator. Before placing it in the demodulator,
adjust its bandwidth to approximately 6 kHz. (If you use the same Tuneable LPF that you used
for DSB demodulation, it already has a bandwidth of 6 kHz.)

You will now build an envelope detector. This will serve as the demodulator for the AM signal.
Connect the modulated carrier to a rectifier (on the Utilities module) and the rectifier output to a
Tuneable LPF (with bandwidth 6 kHz). The combination of the rectifier and the low-pass filter
makes an envelope detector.

Connect the 5-kHz message sinusoid to Channel A and the rectifier output to Channel B. Use
the TTL output from the Audio Oscillator as the trigger source.

Channel A: 5-kHz message sinusoid

Channel B: rectifier output (𝜇 = 1)

13
On Channel B replace the rectifier output with the envelope detector output (that is, the
detector’s Tuneable LPF output). For the envelope detector output, set the coupling for AC.
(Now that the envelope detection is done, the DC component is no longer needed.)

Channel A: 5-kHz message sinusoid

Channel B: envelope detector output (𝜇 = 1)

You will have noticed that no local oscillator (and therefore no stolen carrier) is required for
envelope detection. For an AM receiver, carrier synchronization is not required; and this
simplifies the receiver design.

Set the modulation index to the new value of 0.5. You will use the same procedure that you
earlier used for setting the modulation index to 1, except that now the ratio of the rms voltage of
the message component to that of the DC component will be 0.35, in accord with Equation (12).
The table below summarizes one way (but not the only way) of achieving this ratio.

𝜇 message component (V rms) DC component (V rms) ratio

measured at Buffer Amplifier output measured at Buffer Amplifier output
1.0 0.71 1.00 0.71
0.5 0.35 1.00 0.35
1.5 1.06 1.00 1.06

With 𝜇 = 0.5, observe the AM carrier on the oscilloscope. Use the TTL output from the Audio
Oscillator as the trigger source.

Channel A: 5-kHz message sinusoid

Channel B: AM carrier (𝜇 = 0.5)

Place the Buffer Amplifier output on Channel A and the AM modulator output on Channel B.
Switch to an XY view and observe the resulting figure. This should look like undermodulation,
𝜇 < 1.

Channel A: Buffer Amplifier output

Channel B: AM carrier (𝜇 = 0.5)

Observe the spectrum of the modulated carrier and record the line height of the upper sideband
relative to that of the residual carrier, characterizing the difference in decibels as some number of
dBc. Make sure that you get the algebraic sign correct. (If the upper sideband is smaller than the
residual carrier, then the upper sideband line height is a negative number of dBc.)

Channel B: AM carrier (𝜇 = 0.5)

14
Place the 5-kHz message sinusoid from the Audio Oscillator on Channel A and the envelop
detector output on Channel B. For the envelope detector output, set the coupling for AC.

Channel A: 5-kHz message sinusoid

Channel B: envelope detector output (𝜇 = 0.5)

Set the modulation index to the new value of 1.5 with the help of Equation (12). In this case, the
5-kHz message signal will be distorted by the envelope detector. You should increase the
bandwidth of the Tuneable LPF (beyond the original bandwidth of 6 kHz) in order to get a clear
picture of this distortion.

With 𝜇 = 1.5, observe the AM carrier on the oscilloscope. Use the TTL output from the Audio
Oscillator as the trigger source.

Channel A: 5-kHz message sinusoid

Channel B: AM carrier (𝜇 = 1.5)

Place the Buffer Amplifier output on Channel A and the AM modulator output on Channel B.
Switch to an XY view and observe the resulting figure. This should look like overmodulation,
𝜇 > 1.

Channel A: Buffer Amplifier output

Channel B: AM carrier (𝜇 = 1.5)

Observe the spectrum of the modulated carrier and record the line height of the upper sideband
relative to that of the residual carrier, characterizing the difference in decibels as some number of
dBc. Make sure that you get the algebraic sign correct.

Channel B: AM carrier (𝜇 = 1.5)

Place the 5-kHz message sinusoid from the Audio Oscillator on Channel A and the envelop
detector output on Channel B. For the envelope detector output, set the coupling for AC.

Channel A: 5-kHz message sinusoid

Channel B: envelope detector output (𝜇 = 1.5)

15
2.2 Audio Message

Use an audio signal from the Speech module as the message signal. Otherwise, your
configuration will remain the same.

You will use the XY view to adjust the modulation index (instead of using rms voltage
measurements). Place the Buffer Amplifier output on Channel A and the AM modulator output
on Channel B. Switch to an XY view. You will notice that, given the random nature of the
audio message, the XY display changes with each frame. Adjust the relative gains of the
weighted adder in the modulator in order that the modulation index appears as approximately 1
(the right triangle is complete, and the left triangle is absent) in those frames where the message
signal is at its largest value. (For some frames, the display will look like undermodulation.) We
consider this to be an overall modulation index of 1.

Channel A: Buffer Amplifier output

Channel B: AM carrier (𝜇 ≅ 1)

Connect the AM modulator output to the envelope detector. Connect the detector output to the
headphones. Listen to the recovered audio message.

Repeat the above steps, but setting 𝜇 to a value less than 1 (that is, undermodulation). In this
case, the display should look like undermodulation in all frames.

Repeat the above steps, but setting 𝜇 to a value greater than 1 (that is, overmodulation). In this
case, the display should look like overmodulation in all frames.

Record your (subjective) judgment of audio quality, using adjectives like “good”, “okay” and
“poor”.

AM audio message

16