Single Sideband: USB and LSB

Single sideband modulation — universally abbreviated SSB — is the dominant voice mode on the HF amateur bands. If you tune across 20 meters (14.0–14.35 MHz) during the day, virtually every voice signal you hear is SSB. The reason is straightforward: SSB is dramatically more efficient than AM, concentrating all transmitter power into the sideband where the information actually lives, while eliminating the carrier and the duplicate sideband that AM wastes power on. Understanding SSB completely — how it is generated, how USB and LSB differ, how a receiver extracts the audio, and what happens if the receiver is tuned slightly off frequency — is essential knowledge for every HF operator.

Why SSB Was Developed

The lesson on amplitude modulation revealed AM's fundamental inefficiency: at 100% modulation, two thirds of the transmitter's total output power is wasted in the carrier — which carries no information — and both sidebands, which carry identical information, each receive only one sixth of the total power. An operator running 100 W AM delivers only about 17 W to each information-bearing sideband.

Engineers and amateur radio experimenters began working on this problem in the 1930s, and by the late 1940s and 1950s SSB was well established. The key insight was that both sidebands contain the same information, so one of them is redundant. And the carrier itself was identified as pure waste — it exists only to make demodulation simpler. If the receiver could reinsert its own carrier locally (at exactly the right frequency and phase), the transmitter carrier could be eliminated entirely. Combining these two steps — remove the carrier, remove one sideband — gives SSB.

The resulting signal concentrates all of the transmitter's output power into the remaining sideband. That single sideband contains all the information that the full AM signal contained (since the two sidebands are mirrors of each other). The transmitter output power required for the same intelligibility at the receiver therefore drops enormously, or equivalently, the same transmitter power produces a dramatically stronger signal at the receiving end. This is why SSB displaced AM on HF almost entirely during the 1950s and 1960s, and why it remains the standard for HF voice today.

Upper Sideband and Lower Sideband

When you modulate a carrier at frequency f_c with audio ranging from f_low to f_high, you create two sidebands:

• Upper Sideband (USB): occupies frequencies from f_c + f_low to f_c + f_high. The low audio frequencies map to the bottom of the USB, and the high audio frequencies map to the top. Higher audio tones produce higher RF frequencies in USB.
• Lower Sideband (LSB): occupies frequencies from f_c − f_high to f_c − f_low. The high audio frequencies map to the lowest RF frequencies in the LSB, and the low audio frequencies map to the highest. The frequency order is reversed: higher audio tones produce lower RF frequencies in LSB.

The fact that LSB is a frequency-inverted version of USB is important for receiver design: when you demodulate an SSB signal, you must insert a local carrier at exactly the right frequency and use the correct sideband mode (USB or LSB) to restore natural audio pitch. If you attempt to receive USB with your receiver set to LSB, the audio comes out pitch-inverted — speech becomes unintelligible (it sounds like a chipmunk or a slow-talking monster, depending on the offset). This is not dangerous — just a mistuned or wrong-mode receive setting.

In terms of spectral content, USB and LSB are mirror images of each other reflected around the carrier frequency. They contain identical information — just with the audio spectrum flipped in the frequency domain. The choice of which sideband to use is entirely conventional; both work equally well technically. The conventions that have been adopted by the amateur community exist purely for compatibility — so that transmitters and receivers using the same band all select the same sideband, ensuring they can communicate.

SSB transmitter using the filter method. The balanced modulator produces DSB-SC (both sidebands, no carrier). The sideband filter (crystal or mechanical) removes the unwanted sideband, leaving only USB or LSB. The signal is then converted to the final transmit frequency and amplified. The spectrum diagrams show the signal at each stage.

View Larger

USB and LSB Conventions by Band

Amateur radio operators follow a convention established over decades that determines which sideband to use on which band. The convention is not a legal requirement but is universally observed because using the wrong sideband would mean your signal could not be received by other stations without them changing their mode setting:

The simple rule most operators memorize is: LSB below 10 MHz, USB above 10 MHz. The 60-meter exception (where FCC rules mandate USB) is worth knowing, and the VHF/UHF SSB operating convention (USB throughout) is important for operators active on weak-signal VHF/UHF. Most modern HF transceivers automatically select the correct sideband when you change bands, though you can always override this manually.

Generating SSB: The Filter Method

The most common method for generating SSB in practical transceivers is the filter method, which uses a bandpass filter sharp enough to pass one sideband and reject the other. The filter method works in three stages:

Stage 1: The Balanced Modulator

The audio signal and a carrier oscillator signal are fed into a balanced modulator — a circuit designed to mix the two signals together while canceling the carrier itself. The output of a balanced modulator is DSB-SC (Double Sideband, Suppressed Carrier) — it contains the upper sideband and the lower sideband, but no carrier. The carrier is suppressed by the circuit's balanced (symmetrical) design, typically to better than 40 dB below the sidebands. A common balanced modulator circuit is the diode ring modulator (four diodes in a ring or bridge configuration), which achieves good carrier suppression without requiring precision active components.

Stage 2: The Sideband Filter

The DSB-SC output from the balanced modulator is applied to a very sharp bandpass filter — typically a crystal filter or mechanical filter — that passes only one sideband and rejects the other. The carrier oscillator frequency is chosen so that the two sidebands are separated by twice the lowest audio frequency. For a voice carrier oscillator at 9 MHz with audio from 300 Hz to 3,000 Hz, the USB occupies 9,000.3 to 9,003 kHz and the LSB occupies 8,997 to 8,999.7 kHz. The filter must reject one of these bands by at least 60–80 dB to ensure the unwanted sideband does not interfere with the desired one. Crystal filters and mechanical filters have the very steep skirts needed for this — they achieve 60+ dB of rejection within just a few hundred hertz of the passband edge.

To switch between USB and LSB, the transceiver changes the carrier oscillator frequency slightly — moving the carrier to the opposite side of the filter passband — so that the other sideband falls within the filter passband and is passed instead. For example, a 9 MHz filter with a 2.4 kHz passband from 9.0003 MHz to 9.0027 MHz passes USB when the carrier is at 9.000 MHz (USB occupies 9.0003 to 9.003 MHz) and passes LSB when the carrier is shifted to 9.003 MHz (LSB now occupies 9.0 to 9.0027 MHz).

Stage 3: Frequency Conversion to the Transmit Frequency

The SSB signal emerging from the filter is at the intermediate frequency (IF) — typically 9 MHz, 10.7 MHz, or another standard IF frequency chosen for the filter design. A second mixer converts this to the desired transmit frequency by adding or subtracting the output of a variable frequency oscillator (VFO) or synthesizer. The dial frequency the operator sees corresponds to this final conversion. If the SSB signal is at 9 MHz and the operator selects 14.225 MHz on the dial, the VFO is set to 14.225 − 9.000 = 5.225 MHz (for high-side conversion the VFO would be at 14.225 + 9.000 = 23.225 MHz). The linear power amplifier then raises the SSB signal to the final transmit power.

Generating SSB: The Phasing Method

An alternative approach to generating SSB — used in some modern SDR-based transceivers and in many classic designs — is the phasing method (also called the Hartley method or the image-reject method). Instead of filtering out the unwanted sideband, the phasing method generates the two sidebands in such a way that they cancel in one combination and add in the other.

The phasing method requires two audio signal paths with a 90-degree phase difference between them, and two RF carrier paths also with a 90-degree phase difference. Mixing the shifted audio with the shifted carrier and summing the outputs cancels one sideband while reinforcing the other. The degree of sideband suppression depends on how precisely the 90-degree phase shifts are maintained across the full audio bandwidth — achieving good suppression (40+ dB) across 300–3,000 Hz requires excellent phase accuracy, which is why the filter method has historically been more popular in hardware radio design. However, with software-defined radios (SDRs), the phase shifts can be applied digitally with great precision, making the phasing method very practical in modern DSP-based transceivers.

SSB Bandwidth

One of SSB's most important advantages over AM is bandwidth. In AM, the signal occupies twice the highest audio frequency (above and below the carrier). In SSB, only one sideband is transmitted — so the bandwidth equals the width of that sideband, which is simply the span of the audio frequencies being transmitted:

In practice, the effective SSB bandwidth for voice is usually stated as approximately 2.4–2.8 kHz, reflecting the audio passband of the microphone, audio amplifier, and sideband filter combined. High-quality SSB transceivers often use sideband filters with a 2.4 kHz or 2.7 kHz passband. Some DX-oriented operators use narrower filters (2.1 kHz or even 1.8 kHz) to improve selectivity on crowded bands at the cost of slightly reduced audio quality. Wide-filter options (3.0 kHz or more) are sometimes used for better audio fidelity on clear bands.

The bandwidth of SSB has a direct practical consequence: on the 40-meter phone band (7.175–7.300 MHz in the US), a bandwidth of 125 kHz accommodates approximately 125 kHz / 2.7 kHz ≈ 46 simultaneous SSB contacts. On FM, each contact would occupy 10–16 kHz, allowing only 8–12 contacts in the same spectrum — three to six times fewer. On heavily used bands during contest weekends, this difference is the reason you can still find clear frequency even when hundreds of stations are active.

Calculator: SSB Bandwidth

The Product Detector: Demodulating SSB

An envelope detector — the simple diode circuit that works perfectly for AM — cannot demodulate SSB. The reason is fundamental: an SSB signal has no carrier to serve as the envelope reference. The envelope of an SSB signal does not correspond to the audio waveform at all; it is a complex shape determined by all the audio frequencies present simultaneously. If you apply an SSB signal to an envelope detector, you get distorted noise, not intelligible speech.

SSB demodulation requires a product detector — a circuit that multiplies the incoming SSB signal by a locally generated carrier signal. This multiplication (also called mixing or heterodyning) performs the mathematical equivalent of reinserting the carrier that was suppressed at the transmitter, and it recovers the original audio frequencies from the sideband.

The mathematics is straightforward: multiplying the SSB signal (which is a sideband at frequency f_c + f_audio for USB) by a local carrier at f_c produces sum and difference products:

The low-pass filter after the product detector removes the high-frequency sum component (at twice the carrier frequency), leaving only the difference — the original audio frequency. This process is linear and distortion-free as long as the local carrier (called the Beat Frequency Oscillator or BFO) is at exactly the right frequency.

The SSB product detector multiplies the incoming SSB signal by a locally generated carrier (the BFO). The multiplication produces the original audio frequency and a high-frequency image at 2f_c. A low-pass filter removes the image, leaving the recovered audio. The BFO must be at exactly the correct frequency or voices sound pitched up or down.

View Larger

The BFO and Carrier Reinsertion

The Beat Frequency Oscillator (BFO) in an SSB receiver generates the local carrier that the product detector uses for demodulation. The BFO must be set to the same frequency as the suppressed carrier — the carrier that was used at the transmitter but was not transmitted. If the BFO is off frequency, the audio comes out at the wrong pitch.

For USB, if the transmitter carrier is at 14.225 MHz, the SSB signal occupies 14.2253 to 14.2280 MHz (assuming 300 Hz to 3,000 Hz audio). The BFO in the receiver must be set to 14.225 MHz to recover natural-sounding audio. If the BFO is 100 Hz too high (at 14.2251 MHz), all audio frequencies shift down by 100 Hz — voices sound slightly lower-pitched. If it is 100 Hz too low (at 14.2249 MHz), voices shift up 100 Hz. At 500 Hz off frequency, voices become noticeably unnatural. At 1,000 Hz or more, speech becomes unintelligible — it sounds like a warbling alien. This sensitivity to BFO tuning is one of the characteristics of SSB that beginners find challenging, but modern transceivers solve this problem with their VFO controlling the IF conversion chain such that the BFO tracks automatically as you turn the dial.

The practical effect is that tuning an SSB signal requires finding the frequency at which the voice sounds natural. For USB, turn the dial until the voice sounds natural (not too high-pitched, not too low-pitched). For LSB, the same approach applies but the pitch shift direction reverses as you tune. Once you have a natural-sounding contact, you are on frequency. Any drift in the VFO of either the transmitter or receiver will cause the pitch to shift during the contact — which is why SSB requires a stable oscillator and why crystal-controlled and synthesized transceivers replaced VFO-only rigs for SSB operation.

The Power Advantage of SSB

SSB's power advantage over AM is one of the most significant practical benefits in amateur radio. Let's work through the numbers carefully to understand exactly where the 9 dB advantage comes from.

SSB transmitters are rated in PEP — Peak Envelope Power — rather than average power. This is because an SSB signal's amplitude varies with the voice, reaching a peak when the audio is at maximum and dropping toward zero during pauses. At full speech audio, the PEP equals the instantaneous power during a voice peak. In a correctly set-up SSB transceiver, the PEP is the transmitter's rated output, and the average power during a voice contact is typically 25–50% of PEP depending on the operator's voice characteristics and speech processing settings.

FCC Part 97 limits amateur stations on HF to 1,500 W PEP output, which corresponds to 1,500 W of SSB PEP. The legal limit is always in PEP for SSB because instantaneous power during peaks is what matters for interference and adjacent channel performance.

Operating SSB in Practice

Understanding SSB theory translates directly into practical operating skills. Here are the key things every SSB operator needs to know:

Setting Audio Levels

The microphone gain and speech processing must be set so that the ALC (Automatic Level Control) meter shows appropriate action — deflecting on voice peaks without pegging at maximum. Excessive microphone gain causes the ALC to clip the audio heavily, introducing distortion similar to AM overmodulation. Proper SSB audio level adjustment: speak normally into the microphone and set the microphone gain so that voice peaks drive the ALC meter to about 80% of full scale. Never drive the ALC into constant or heavy limiting.

Clarifier and RIT Controls

Most HF transceivers have a clarifier (also called RIT, Receiver Incremental Tuning) that allows you to adjust the receive frequency without changing the transmit frequency. If a station you are working has a VFO that is slightly off frequency, you will notice their voice sounds pitched or unnatural. Use the clarifier to shift your receive frequency until their voice sounds natural — this corrects for their VFO offset without changing your own transmit frequency. Clarifier range is typically ±10 kHz in 10 Hz steps.

Working Split

During DX pile-ups, the rare DX station often listens on a different frequency from where they transmit — this is called working split. They might transmit on 14.195 MHz but listen on 14.200–14.210 MHz, spreading the pile-up callers across 10 kHz of spectrum and making it easier to hear individual callers. Modern transceivers handle split operation with separate A/B VFO memories. Always listen to find where the DX station is listening before transmitting.

Calling Frequencies

Certain frequencies are designated as calling frequencies — where you call "CQ" (a general call to any station) and where you listen for others doing the same. Common HF SSB calling frequencies in the US include 14.225 MHz (20m USB, though the calling frequency varies by region), 7.200 MHz (40m LSB), 21.300 MHz (15m USB), and 28.400 MHz (10m USB). After making initial contact, it is good practice to move to a clear working frequency so the calling frequency remains available for others.

Spectrum comparison of AM vs SSB. AM uses 6 kHz bandwidth with power split between carrier and two sidebands. USB and LSB each occupy 2.7 kHz — less than half the bandwidth — with all power concentrated in one sideband. Note the suppressed carrier position (dashed line) in the SSB displays: no RF energy at that frequency, yet the receiver can still demodulate by reinserting a local carrier (BFO).

View Larger