Spectrum Analyzer

Overview

Click Start Microphone and the Spectrum Analyzer turns your live input into a Fast Fourier Transform (FFT) display — adjust the Sensitivity slider and switch between Frequency Bars and Waveform mode to watch your room's dominant frequency and peak level update as you go. The ambient sound level decibel meter is a fast way to see a live number instead of guessing by ear. The test your vocal range gives you a clear answer instead of guessing.

When you need to understand exactly what frequencies are present in a spectrum analyzer reading and how much power each component carries, your results tell a story that time-domain tools simply cannot narrate. Whether you are chasing unwanted emissions across a crowded ISM band, verifying the harmonic content of an amplifier output, or characterizing a radar transmitter's spectral purity, the frequency-domain view delivered by this tool gives you the clarity to make confident technical decisions. From the bench to the site, signal and spectrum analyzer tools have become indispensable across electronics, communications, acoustics, optics, and mechanical oscillation — anywhere that understanding power vs. spectral distribution is the key to solving a problem.

What Is a Spectrum Analyzer and What Does It Measure?

A spectrum analyzer is a test instrument that performs signal measurement of the magnitude of an input signal versus spectral position across the full input spectral coverage of the instrument. Where an oscilloscope maps signal level against time, this tool maps level against spectral position — giving you an immediate, visual representation of the power spectrum of known and unknown inputs. The Y-axis displays power (typically in dBm or dBµV), while the X-axis represents the spectral axis on either a linear or logarithmic scale, depending on the task. Generate exactly what you need with the create your own tone online, no install or sign-up required.

Beyond raw power levels, powerful analyzers expose a rich set of characteristics:

• Spectral components and their relative levels
• Harmonic content — second, third, and higher-order overtones
• Sensitivity floor and signal-to-noise ratio
• Occupied channel width and channel capacity
• Modulation quality and distortion characteristics
• Spurious emissions and cross-device coupling
• Carrier drift over short and long intervals

The primary use domains span radio frequency (RF) and microwave technical work, audio-frequency acoustics, optical spectrum analysis, and mechanical oscillation analysis. In communications and broadcasting, reliable signal measurement of RF inputs ensures regulatory adherence.

In EMC evaluation and pre-compliance work, these instruments verify that products stay within mandated emission limits. In audio production, they reveal overtone distortion in amplifiers and music signals. The sheer breadth of applications — from 9 kHz through 7.5 GHz and beyond into the microwave band — reflects why the spectrum analyzer has become a cornerstone of modern instrumentation.

Definition: A spectrum analyzer performs signal measurement of input signal energy plotted as spectral position vs. level, revealing the power spectrum of both known and unknown inputs across the electromagnetic spectrum.

How a Spectrum Analyzer Works: Theory of Operation and Signal Analysis

Understanding how this tool processes an input is essential for selecting the right settings and interpreting your results correctly. At the heart of every spectrum analyzer — whether swept or FFT-based — is the task of decomposing a time-domain input into its constituent spectral components so that you can see each component's power independently. The fundamental operation involves either sweeping a local oscillator across the measurement window or digitizing a wideband capture and applying a fast Fourier transform (FFT) for band analysis. Run the free absolute polarity blind test for a few minutes a day to build real listening skill.

Center Frequency, Span, and Signal Analysis Setup

Every measurement begins with two foundational parameters. The center frequency is the midpoint of the displayed spectral range — the spectral position at the center of the screen.

The sweep width defines the total spectral range shown, stretching equally above and below the center point. For example, setting a center of 2.4 GHz with a sweep width of 100 MHz displays the range 2.35 GHz to 2.45 GHz, covering the entire 2.4 GHz ISM band used by Wi-Fi and Bluetooth.

Selecting the right sweep width is a balance between coverage and spectral resolution: a wide window shows the full picture but reduces the ability to separate closely spaced inputs, while a narrow window zooms in for fine band analysis at the cost of missing energy outside the window. Most modern signal analyzer tools allow you to set the sweep width from a few kHz up to the full spectral coverage of the instrument in a single command.

Resolution Bandwidth, Filter, and Spectral Resolution

The resolution bandwidth (RBW) is the passband width of the bandpass filter — the narrowing filter — applied at the intermediate-frequency (IF) stage. It determines the minimum separation between two inputs that the instrument can distinguish as separate peaks. A narrower RBW provides finer spectral resolution and lower displayed noise, but demands a longer sweep time because the filter must settle at each point:

In practice, if you halve the RBW, sweep time increases by approximately a factor of four. For the superheterodyne architecture, the instrument downconverts the input to an intermediate frequency, passes it through the narrowing filter, and then assesses the power at each spectral point in the selected window.

The detector level at each step is recorded to build the power vs. spectral trace. Choosing RBW requires careful judgment: too wide and closely spaced inputs merge; too narrow and sweep time becomes impractical for dynamic inputs.

Post-Detection Filter and Spectral Trace Smoothing

After the detector, a video bandwidth (VBW) filter averages the detected power over time, smoothing the displayed trace to reduce fluctuations. Reducing VBW below RBW smooths the trace and makes weak inputs easier to see against the sensitivity floor, but it slows the effective update rate.

Setting VBW equal to or greater than RBW preserves brief signal events. The relationship is:

In spectral analysis of noisy inputs — such as emission floors or satellite link characterization — reducing VBW to one-tenth of RBW (or less) reveals signals masked by noise that would otherwise be invisible.

Detector Modes: Peak, Sample, and Average Detection

The detector mode controls how the instrument condenses multiple input samples within each RBW cell into a single display point. The three primary modes are:

Peak detection

Captures the highest level encountered during the sweep through an RBW cell. Ideal for identifying spurious emissions and ensuring no input peak is missed — the default mode for EMC evaluation and adherence work.

Sample detection

Records a single instantaneous sample from each RBW cell. This preserves the noise statistics accurately, making it the preferred mode for sensitivity floor characterization and DANL verification.

Average detection

Averages multiple samples within each cell, reducing variance and improving the visibility of low-level, continuous inputs. Equivalent to applying additional time averaging to the power reading.

Displayed Average Noise Level (DANL) and the Sensitivity Floor

The displayed average noise level (DANL) is the internal sensitivity floor of the instrument itself — the lowest power it can detect above its own thermal and electronic background. Any input below the DANL is indistinguishable from the instrument's own background. DANL is typically specified in dBm/Hz and improves (gets lower) as you reduce RBW, because the narrower filter admits less thermal background:

High-performance benchtop instruments can achieve DANL values below −170 dBm/Hz, enabling the isolation of weak inputs and signals masked by noise in sensitive receiver work. For site work, handheld instruments trade some DANL performance for portability, but modern handheld devices still achieve DANL levels suitable for most interference detection tasks.

Main Types of Spectrum Analyzers: Swept-Tuned, FFT, RSA, VSA, and Hybrid Architectures

Modern spectrum analyzers come in several distinct architectures, each optimized for different band analysis needs. Choosing the right type depends on whether your inputs are stable or dynamic, whether you need to capture brief events, and whether your application demands decoding and vector readings alongside spectral views. Below is a structured comparison of every major architecture, including a worked scenario showing where each approach succeeds or fails.

Swept-Tuned Superheterodyne Analyzers (SA) — Traditional Architecture

The swept-tuned superheterodyne instrument is the traditional workhorse of RF signal measurement. Its architecture works by downconverting the input using a voltage-controlled local oscillator, mixing it to an intermediate frequency (IF), passing the result through the narrowing filter, detecting the power, and then stepping the local oscillator to the next spectral point.

The result is a power vs. spectral map built one step at a time. This approach delivers excellent dynamic range and reading precision for stable, unchanging inputs.

The fundamental limitation is that this approach is effectively blind to events occurring while it dwells at other spectral points. A frequency-hopping input that dwells on each channel for only 1 ms will almost certainly be missed by an instrument whose sweep time exceeds several seconds across a wide window.

For controlled, static inputs — such as a CW carrier, a stable oscillator, or a transmitter spectrum emission mask verification on a continuous input — this remains an excellent, cost-effective choice. It covers spectral windows from as low as a few kHz up to the full input spectral coverage of the instrument, with RBW values as narrow as 1 Hz for high-resolution spectral work.

FFT-Based Spectrum Analyzers — Fast Fourier Transform Processing

An FFT-based instrument digitizes the time-domain input with a high-speed ADC, then applies a Fourier transform — a computationally efficient implementation of the discrete Fourier transform — to convert the time-domain waveform into a spectral representation. Because all components within the capture window are computed simultaneously from a single digitized block, these tools are dramatically faster than swept instruments at low spectral positions and for narrowband captures.

The Fourier transform relationship is foundational:

In practice, the discrete Fourier transform (DFT) operates on sampled data, and the window function applied before the FFT determines the trade-off between spectral resolution and leakage. This approach is ideal for audio-frequency work and low-rate oscillation analysis, where the narrow spectral range can be captured in a single ADC sample block with exceptional speed. At higher RF positions, a hybrid front-end is typically required.

Real-Time Spectrum Analyzers (RSA/RTSA) — 100% Probability of Intercept

The real-time spectrum analyzer (RSA or RTSA) represents the most advanced architecture for characterizing dynamic inputs. Unlike swept or conventional FFT instruments, an RSA performs real-time FFT processing with overlapping DFT frames, ensuring that every sample captured is included in at least one FFT computation. This gives the RSA a 100% probability of intercept — it processes inputs indefinitely without gaps between capture cycles.

The RSA uses an RF downconverter to shift the input to a wideband intermediate-frequency section. An ADC digitizes the IF output at a rate that exceeds the Nyquist criteria for the full capture window.

Real-time DSP then transforms overlapping time-domain waveform blocks into spectral outputs at a DFT transform rate fast enough to ensure no analysis interval is missed. The result is a gapless stream of spectrograms — often displayed as a spectrogram graph — that reveals rare short-duration events, intermittent transmissions, glitches, switching transients, and inputs that shift spectrally over time, including those that would be completely invisible to a swept instrument.

The RSA can trigger on events in the spectral, time, encoding, statistical, and code domains, seamlessly capturing triggered segments into memory for post-acquisition processing and multiple-domain examination.

Vector Signal Analyzers (VSA) — Magnitude and Phase for Encoded Inputs

A vector signal analyzer (VSA) goes beyond power-only readings to capture both magnitude and phase of the input, enabling full decoding and modulation analysis. The VSA digitizes all RF power within the passband using a wideband ADC, stores the digitized waveform in memory, and then applies DSP algorithms — including processing for decoding — to extract I/Q data. This captured waveform contains the complete characterization needed for EVM (error vector magnitude) work, 256QAM linearity evaluation, and phase-accurate readings across multiple antennas and synchronized channels.

The VSA's batch processing architecture means it stores a block of data, then examines it — making it well-suited for analyzing complex encoded inputs like DVB-S, DVB-S2, DVB-T, 802.11ac, and other encoding schemes. Its limitation is the gap between capture cycles: single or infrequent brief events can slip through undetected, and external triggering may be required for reliable capture of rare events. A vector network analyzer (VNA or network analyzer) extends VSA concepts to include transmission and reflection readings for component characterization, though the VNA is a distinct instrument category.

Hybrid Superheterodyne-FFT Analyzers — Best of Both Worlds

The hybrid superheterodyne-FFT architecture combines a swept superheterodyne front-end — which handles wide spectral coverage and downconverting the input to a manageable intermediate frequency — with an FFT-based back-end that provides real-time signal bandwidth and speed. This approach allows the instrument to cover a wide input spectral range (from kHz to tens of GHz) while still offering significant real-time signal bandwidth for band analysis. Many modern benchtop instruments use this architecture, digitizing a wideband IF with a fast ADC and then processing overlapping DFT frames to provide near-real-time performance across a selectable capture window.

Form Factors: Benchtop, Handheld, Portable, and Networked Analyzers

Spectrum analyzers are available in four primary form factors, each targeting different band analysis needs:

• Benchtop instrument: The highest-performance form factor, offering the widest spectral coverage, lowest DANL, best phase noise, and most comprehensive reading functions. Benchtop performance is optimized for laboratory use — in the workshop, these instruments support the most demanding verification, waveform examination, and EMC evaluation tasks. A benchtop instrument like the RSA5100B exemplifies this category, offering residual EVM specifications adequate for 256QAM work and real-time signal bandwidth of 165 MHz for 160 MHz WLAN channels.
• Portable instrument: Mid-range form factor that balances performance with transportability. Suitable for site deployment when benchtop performance is not required but a handheld device lacks the necessary reading precision.
• Handheld analyzer: Compact, battery-powered device designed for site use — interference hunting, spectrum management, and site surveys. Modern handheld instruments offer spectral coverage from 9 kHz to 7.5 GHz and above, with sufficient dynamic range for most radio communications and cellular radio troubleshooting tasks. USB-powered portability is a feature of some compact modular designs.
• Networked device: Remotely accessible instrument that transmits reading data over a network, enabling distributed monitoring across multiple sites simultaneously. Networked instruments support continuous, unattended assessment for spectrum management and interference detection across wide geographic areas.

Applications Across Frequency Domains: From RF to Audio, Optical, and Vibration Spectrum Analyzers

The true power of spectrum analyzers becomes apparent when you consider how widely the spectral domain view applies across science and technical disciplines. Whether you are analyzing RF circuitry in a workshop, evaluating the acoustic spectrum of a music performance, examining optical wavelength distribution in a fiber-optic system, or diagnosing mechanical oscillation in rotating machinery, the core principle is the same: decompose a complex analog signal into its spectral components and perform frequency measurement of each component's power independently.

Radio-Frequency and Wireless Applications: Interference Hunting and Signal Characterization

Radio-frequency uses dominate the spectrum analyzer market. In communications infrastructure, RF professionals use these instruments for transmitter characterization — verifying that power output, occupied channel width, and encoding quality meet specification.

In cellular radio design, they confirm that voice and data transmission inputs (including 4G LTE and 5G NR) operate within their allocated channels without producing spurious out-of-band emissions. Radar transmitter verification analyzes pulse shape, spectral purity, and adherence to spectrum management requirements. EMC evaluation and pre-compliance work use peak detection to capture worst-case emission levels across the electromagnetic spectrum for regulatory submission.

For Wi-Fi work and wireless network testing, spectrum analyzers have become essential tools. Consider the following worked example for interference hunting in the 2.4 GHz band:

1. Set center frequency to 2.4 GHz: This places the IEEE 802.11 unlicensed band (ISM band) at the center of the display. For 5 GHz 802.11ac or 802.11n wireless networks, shift the center to the higher band.
2. Set sweep width to 100 MHz: This reveals all channels from 2.35 GHz to 2.45 GHz, covering the full Wi-Fi band including 802.11g and 802.11n channels.
3. Adjust RBW to 100 kHz and VBW to 30 kHz: This provides sufficient spectral resolution to separate individual 20 MHz Wi-Fi channels while smoothing the trace enough to reveal weak interferers beneath the sensitivity floor. The narrower VBW suppresses fluctuations, making it easier to detect inputs that are signals masked by noise or a rogue access point transmitting between legitimate channels.
4. Switch to peak detection: This ensures that intermittent transmissions from competing access points — which may appear only momentarily — are captured at their true peak power level.
5. Observe the spectral display: Legitimate 802.11g and 802.11n channels appear as flat-topped humps approximately 20 MHz wide. Channel overlap, a microwave oven operating in the ISM band, or a Bluetooth piconet appear as additional spectral energy between or overlapping the Wi-Fi channels.
6. Identify the rogue input: A narrowband interferer shows as a sharp spike at a specific spectral position; wideband coupling from a microwave oven appears as broadband raised energy across a wide portion of the band.

For work with the newer 802.11ac and 802.11n standards using 80 MHz channels on wireless networks, the instrument should have a real-time signal bandwidth of at least 120 MHz — specifically 120 MHz — to capture the full 80 MHz channel within its capture window and assess transmitter emission mask adherence. For 160 MHz WLAN channels, instruments such as the RSA5100B with 165 MHz real-time capability and residual EVM of −49 dBm provide sufficient linearity and sensitivity specifications to verify 256QAM encoding performance. Assessing multiple spatial streams across multiple antennas requires up to 8 independent capture channels that are synchronized for phase-accurate readings — a capability found in high-end RSA platforms.

Beyond Wi-Fi, RF uses include satellite input characterization (DVB-S and DVB-S2 forward link analysis, DVB-T terrestrial broadcast verification), video broadcast monitoring, and spectrum management for regulatory agencies tracking unlicensed band activity.

Audio-Frequency Measurements: Harmonic Distortion and the Audio Spectrum Analyzer

An audio spectrum analyzer operates in the spectral range from approximately 20 Hz to 20 kHz — the range of human hearing — and is an essential tool for sound professionals, musicians, and acousticians. Unlike RF work where inputs travel through coaxial cables, audio-frequency measurements deal with acoustic pressure waves converted to electrical inputs, or directly with outputs from amplifiers, mixers, and sound processors.

Music and other acoustic sources produce complex waveforms that are rich in harmonic content. Even a nominally pure sine wave from a generator will exhibit some additional overtone content at integer multiples of the base tone when passed through a real amplifier.

The audio spectrum analyzer makes these overtones visible. Consider measuring harmonic distortion on an amplifier used in music production:

1. Apply a test tone at the base tone: Connect a low-distortion sine wave generator at, for example, 1 kHz. Verify that the generator itself has distortion well below your target.
2. Set center frequency and sweep width: Set center to 5 kHz with a window covering 20 Hz to 20 kHz to display the base tone and at least the first nine overtones on screen simultaneously.
3. Select peak detection mode: Use peak detection to capture the true level of each overtone. Read the base tone at 1 kHz and the second overtone at 2 kHz, third at 3 kHz, and so on.
4. Calculate Total Harmonic Distortion (THD): THD is the ratio of the RMS sum of all overtone levels to the base tone level: THD = fracsqrtV₂² + V₃² + V₄² + ·sV₁ × 100% where V₁ is the base tone level and V₂, V₃, … are the overtone levels.
5. Interpret results: A high-quality amplifier might show THD below 0.01%, meaning overtone peaks are more than 80 dB below the base tone. A rich overtone structure with additional peaks at higher spectral positions indicates poor linearity or clipping.

Audio-frequency uses extend to acoustic spectrum assessments of musical instruments — a violin recording of music reveals a complex comb of overtones that define its characteristic timbre and pitch; a human whistle produces a near-pure sine wave with minimal overtone content, appearing as a single sharp spike. Even underwater sounds like whale clicks and orca recordings, or the complex patterns of bird songs and bird calls, yield fascinating spectrogram graphs when analyzed.

Input from preset recordings, uploaded sound files through a player interface, or real-time microphone capture can all be processed. In education and science demonstrations, the spectrogram graph shows students how different sounds — police sirens, dolphin clicks, 56k dial-up modem tones — produce distinctive spectral patterns that are immediately recognizable once you learn to read the display. A virtual oscilloscope can show the same acoustic input as a time-domain trace (displacement vs. time), revealing the waveform shape, while this tool displays spectral intensity — showing where the energy is concentrated across the acoustic range.

For recordings and sound files, the spectrogram view adds a third dimension: time. Color-coded spectral intensity at each moment reveals how the overtone structure evolves — the attack transient of a musical instrument shows broad high-frequency energy that settles into a steady-state overtone pattern, and sound visualization makes this immediately apparent. Whether you use the play and pause controls of a software player or feed live acoustic input, the analytical insight is the same: the acoustic spectrum is far richer and more informative than a simple level vs. time trace.

In recording production, audio-frequency spectrum work also supports sensitivity floor characterization — verifying that the noise characteristics of a recording chain (microphone, preamp, ADC) stay below audible thresholds, and that inputs at both high and low intensities remain undistorted and accurately captured.

Optical Spectrum Analysis: Wavelength and Power in the Optical Domain

The optical spectrum analyzer (OSA) extends the spectral-domain concept into the optical domain — from the near-infrared through the visible light range. Rather than assessing power vs. spectral position in Hz, an OSA measures power vs. optical wavelength in nanometres (nm), though the underlying science is governed by the same relationship (since wavelength and optical frequency are inversely related: f = c/lambda). OSAs use diffractive optics or tunable Fabry-Pérot filters to perform the equivalent of a sweep across optical wavelengths, and are essential tools in fiber-optic communications and laser characterization.

In optics and fiber communications, OSAs characterize the wavelength and power of each channel in a dense wavelength-division multiplexed (DWDM) system — verifying that each carrier is within its allocated slot, assessing optical signal-to-noise ratio, and detecting channel crosstalk. Laser diode work with this type of instrument reveals mode structure, linearity, and side-mode suppression ratio. The optical domain typically spans from 600 nm to 1700 nm, corresponding to frequencies from roughly 175 THz to 500 THz — far beyond the spectral coverage of any electrical test instrument.

Vibration Spectrum Analysis: Mechanical Oscillation and Predictive Maintenance

A vibration spectrum analyzer applies FFT-based spectral analysis to mechanical oscillation inputs, typically measured by accelerometers attached to rotating machinery, structural components, or embedded systems. The spectral domain view of oscillation levels reveals fault signatures that are completely invisible in a simple time-domain trace: a bearing defect creates a characteristic set of component frequencies related to the bearing geometry and rotation speed; gear mesh frequency appears as a peak at the product of tooth count and shaft speed; imbalance manifests as a dominant peak at the base rotation rate (1×).

In predictive maintenance work, oscillation spectrum analysis enables condition monitoring of pumps, motors, compressors, and turbines without shutdown. A shift in oscillation levels at specific component frequencies — or the appearance of sidebands around a gear mesh peak — signals developing faults weeks or months before catastrophic failure.

The oscillation spectrum is typically displayed on a logarithmic scale for level (to accommodate the wide amplitude spread from very low to high intensities) and a linear spectral axis spanning from 0 Hz to several kHz or tens of kHz, depending on the rotational speed and fault frequencies of interest. Oscillation analysis also underpins structural resonance work in civil, aerospace, and automotive disciplines — identifying resonance positions where structures respond with amplified oscillation to applied forces, a critical concern for mechanical systems near the natural frequencies of bridges, aircraft, or vehicle bodies.

Spectrum Analyzer vs. Oscilloscope: Frequency Domain vs. Time Domain Signal Analysis

Engineers frequently debate whether a spectrum analyzer or an oscilloscope is the right instrument for a given task. The honest answer is that the two instruments are complementary, not interchangeable. Each reveals a fundamentally different aspect of the same physical input — the oscilloscope shows you how the level changes over time (time-domain view), while this tool shows you what spectral components are present and how much power each carries (frequency-domain view).

An oscilloscope is the right choice for waveform examination — looking at rise times, pulse widths, overshoot, and timing relationships between multiple inputs. It excels at capturing rare brief events when triggered correctly and displaying the actual waveform shape.

A dedicated spectrum analyzer, by contrast, offers superior dynamic range (often exceeding 100 dB), far better spectral resolution (down to 1 Hz RBW), and specialized detection modes optimized for the spectral domain. Modern real-time oscilloscopes can apply an FFT to compute an approximate spectrum, making them a combined scope-and-analyzer tool of a sort — but dedicated instruments deliver superior sensitivity, phase noise performance, and calibration traceable to national standards for rigorous RF signal measurement work.

In a well-equipped laboratory, both instruments are standard tools. The oscilloscope answers questions about when and how an input behaves; the spectrum analyzer answers questions about what spectral components are present and at what power level.

Examination domains also extend beyond these two: the VSA adds the encoding domain, the RSA covers the statistical and code domains, and specialized instruments like the vector network analyzer address spectral response and impedance — each revealing a different facet of the input's behavior. Combining multiple domains and instrument types gives you the most complete characterization possible, which is why application notes from major instrumentation manufacturers consistently recommend using signal analyzers alongside oscilloscopes rather than treating them as substitutes.

Support resources, verification services, and product announcements from leading test equipment manufacturers continue to expand the reading functions available to professionals working with signal and spectrum analyzers. Whether your band analysis needs center on RF adherence, acoustic quality, optical communications, or mechanical oscillation, the right instrument — chosen for its spectral coverage, form factor, real-time capability, and reading precision — gives you the frequency measurement and signal bandwidth insight that drives confident, reliable technical decisions.