EE-410L Communication Systems Lab · Lab 7 · read before the labAll labsGo to the lab →

Send one sideband, get the whole message.

Ten minutes of reading. Move the sliders, then every waveform on the two trainers will make sense.

01

Why single sideband?

2 copies

An AM signal carries the message twice: once in the upper sideband (USB) and once in the lower sideband (LSB). One copy is enough.

0 info

The carrier holds no message at all, yet in AM it takes at least two thirds of the transmitted power.

fm

SSB needs a bandwidth of only fm instead of 2 fm, so a crowded band holds about twice as many SSB stations as AM stations.

DSB-SCs(t) = m(t) cos(2π fc t)balanced modulator output
One tonecos A · cos B = ½ cos(A − B) + ½ cos(A + B)LSB at fc − fm, USB at fc + fm
SSB (LSB)s(t) = ½A cos 2π(fc − fm)ta single line, a flat envelope
02

Power in the sidebands

With a modulation depth m (section 7.3.1 of the manual calls it the depth of modulation), the AM power splits as

Pt = Pc (1 + m²/2), each sideband = Pc m²/4

The carrier power never changes with the message. Only the sidebands grow with m, and even at m = 1 they hold just one third of the total.

LSB16.7%Carrier66.7%USB16.7%
Pt / Pc1.500
Useful (both sidebands)33.3 %
One sideband (SSB-SC sends only this)16.7 %
03

Making SSB: the filter method

The IT-4101 makes SSB in two steps. The balanced modulator multiplies the audio with the 455 kHz carrier; the carrier terms cancel and DSB-SC leaves. The ceramic band pass filter, centred near 450 kHz and only a few kHz wide, then passes the lower sideband and pushes the upper one down. Move fm and watch where the two sidebands fall on the filter curve:

440k445k450k455k460k465k470kcarrier 455k (suppressed)LSB 452.0kUSB 458.0k
LSB through the filter-0.2 dB
USB through the filter-34.1 dB
Sideband suppression33.9 dB
Why a ceramic filter and not an RC filter? The two sidebands are only 2 fm apart (6 kHz for a 3 kHz tone) at 455 kHz. Only a resonator with very steep skirts (ceramic or crystal) can keep one and reject the other. For very low audio tones both sidebands sit on the same skirt, which is why voice SSB filters start at about 300 Hz.
04

Getting the message back: the product detector

An SSB signal has a flat envelope, so an envelope detector cannot recover it. The IT-4102 uses a product detector: it multiplies the received sideband by a local 455 kHz carrier.

cos 2π(fc − fm)t · cos 2π fct = ½ cos 2π fmt + ½ cos 2π(2fc − fm)t

The first term is the message. The second sits near 910 kHz. Some of the 455 kHz carrier also leaks through. That is why the manual calls the detector output noisy.

The local carrier must match. If the receiver oscillator is off by Δf, every audio frequency moves by the same Δf. A few hundred hertz is enough to make a voice sound like a cartoon character. Try it below; the simulator itself keeps both oscillators at exactly 455 kHz.
Received LSB452.00 kHz
Receiver oscillator455.000 kHz
Tone after the detector3.000 kHz
Sum term (removed by the LPF)907.0 kHz
05

The low pass filter

tone 3kpasses455k leak907kLPF 8 kHzblocked

The IT-4102 low pass filter passes the audio band and blocks everything above it. After it, the 455 kHz leak and the 910 kHz term are gone and a clean tone is left.

  • The audio amplifier then raises the clean tone enough to drive the speaker or the headphone.
  • In the simulator the filter is taken as 8 kHz, 4th order, and the amplifier gain as 3. These are model values, not datasheet values.
  • On the analyzer the 455 kHz line is visible at the detector output and has vanished at LPF O/P.
06

The ground cord between the trainers

The IT-4101 and the IT-4102 each have their own supply. A signal cord alone has no return path, so the receiver sees the transmitter signal against a floating reference. Figure 4 therefore adds one cord from a GND socket of the IT-4101 to a GND socket of the IT-4102.

Forget it and you hear a buzz. In the simulator, without the ground cord the received tone gets weaker and 50 Hz mains hum appears at the product detector output. It passes the low pass filter and the speaker buzzes. The extra experiment "The ground cord between the modules" lets you see it on the scope.
07

The two trainers: what connects to what

7.4.1 SSB-SC modulation (Figure 1)
Audio Generator 3 kHz→Balanced Modulator→Ceramic Bandpass Filter→CH2 and analyzer
455 kHz Local Oscillator → Carrier I/P of the Balanced Modulator
7.4.2 Transmitter (IT-4101)
Audio O/P→Balanced Modulator→BPF I/P → BPF O/P
455 kHz O/P → Carrier I/P
7.4.2 Receiver (IT-4102)
SSB I/P→Product Detector→Low Pass Filter→Audio Amplifier→Speaker
RX 455 kHz O/P → Carrier I/P · ground cord between the modules

Before each part: Audio input select switch on Audio Generator Output, speaker OFF, antenna OFF, then power ON and set the audio to about 3 kHz and 700 mV p-p on CH1. In 7.4.2 both modules are powered.

08

Using the instruments

Oscilloscope

  • Clip the ground clip on a GND socket of the same module as the probe tip. Without a ground clip you see 50 Hz hum.
  • Trigger on CH1 (the audio). On BM O/P the envelope follows the audio; on BPF O/P the envelope is flat.
  • On the product detector output the tone looks thick and fuzzy: that is the high frequency residue.
  • Autoset picks the scales; Measure shows Vpp and frequency.

Spectrum analyzer

  • 455 kHz preset on BM O/P: two lines, 452 and 458 kHz, no carrier.
  • Same preset on BPF O/P: 452 kHz strong, 458 kHz far down.
  • Audio preset on LPF O/P: one line at the tone frequency.

Multimeter and counter

  • The meter reads RMS of audio signals: 700 mV p-p is 0.247 V.
  • The counter on LPF O/P reads the received tone; compare it with Audio O/P.
09

What you should see

WhereExpected (3 kHz, 700 mV p-p audio)
Audio O/P (CH1)3.00 kHz, 700 mV p-p
Balanced Modulator O/PDSB-SC, about 490 mV p-p; lines at 452 and 458 kHz, no carrier line
Ceramic filter BPF O/PSSB, about 220 mV p-p with a flat envelope; 452 kHz strong, 458 kHz about 34 dB lower
Product Detector Audio O/P (IT-4102)3 kHz tone with 455 kHz and about 907 kHz riding on it (noisy), about 1.2 V p-p in total
Low Pass Filter O/Pclean 3.00 kHz, about 430 mV p-p
Audio Amplifier O/Pclean 3.00 kHz, about 1.3 V p-p; the speaker plays the tone

Levels are those of the simulator model; the real trainers give values of the same order.

10

Check yourself

Q1The audio tone is 2 kHz and the carrier 455 kHz. Which frequency leaves the ceramic filter?

The lower sideband: 455 − 2 = 453 kHz.

Q2Why is the envelope of a single tone SSB signal flat on the scope?

SSB of one tone is one sine wave at fc − fm. A single sine has a constant amplitude, so the envelope does not move.

Q3The receiver oscillator runs at 455.2 kHz. Which tone comes out for a 3 kHz audio tone (LSB)?

|455.2 − 452| = 3.2 kHz. The pitch moves by the same 200 Hz error.

Q4Which lines does the analyzer show at the product detector output, before the low pass filter?

The tone fm, the leaked 455 kHz carrier and the sum term near 2fc − fm (about 907 kHz for 3 kHz).

Q5You forgot the ground cord between the trainers. What do you notice?

A weaker tone with 50 Hz hum on it; the speaker buzzes.

I am ready, go to the lab →