Spectrum Analyzers
1. Definition and Purpose of Spectrum Analyzers
Definition and Purpose of Spectrum Analyzers
A spectrum analyzer is an instrument designed to measure and display the power spectral density of an input signal as a function of frequency. Unlike an oscilloscope, which reveals temporal characteristics of a signal, a spectrum analyzer provides critical insights into its frequency-domain behavior. This capability is indispensable for analyzing modulated signals, identifying interference sources, and characterizing noise performance in RF and microwave systems.
Fundamental Operating Principle
The core functionality relies on the Fourier transform relationship between time and frequency domains. For a continuous signal x(t), the power spectral density Sxx(f) is obtained through:
Modern implementations typically employ either:
- Superheterodyne architectures with tunable local oscillators for high dynamic range
- Fast Fourier Transform (FFT) processors for real-time analysis of baseband signals
Key Performance Metrics
The analyzer's capabilities are quantified through several critical specifications:
| Parameter | Typical Range | Impact on Measurements |
|---|---|---|
| Frequency Resolution | 1 Hz - 1 MHz | Determines minimum distinguishable frequency separation |
| Dynamic Range | 70-160 dB | Governs simultaneous measurement of strong/weak signals |
| Phase Noise | -80 to -160 dBc/Hz | Affects precision in narrowband applications |
Advanced Applications
Beyond basic spectral analysis, modern instruments enable:
- Adjacent channel power ratio (ACPR) measurements for wireless standards compliance
- Spurious emission detection in radar systems with sensitivity below -150 dBm
- Real-time spectrum analysis for transient signal capture using FFT rates exceeding 1 MSa/s
The choice between swept-tuned and real-time analyzers depends on the application's required probability of intercept versus frequency coverage. For pulsed signals with nanosecond-scale durations, real-time analyzers with persistent displays provide critical visibility into signal dynamics that would otherwise be missed by conventional sweep methods.

1.2 Key Components and Their Functions
Input Attenuator
The input attenuator is the first stage in a spectrum analyzer, designed to protect the sensitive mixer from overloading due to high-power input signals. It typically offers adjustable attenuation in steps of 5 dB or 10 dB, with a range of 0 dB to 70 dB. The attenuation level is dynamically adjusted based on the input signal strength to maintain optimal signal-to-noise ratio (SNR). Over-attenuation reduces sensitivity, while under-attenuation risks mixer compression or damage.
RF Mixer
The RF mixer performs frequency translation by multiplying the input signal with a local oscillator (LO) signal, producing sum and difference frequencies. For a mixer with input signal fin and LO frequency fLO, the output contains:
Nonlinearities in the mixer generate spurious responses, quantified by third-order intercept point (TOI) and conversion loss. High-performance mixers use Schottky diodes or FET-based designs for reduced distortion.
Intermediate Frequency (IF) Filter
The IF filter determines the resolution bandwidth (RBW) of the analyzer. A narrower RBW improves frequency selectivity but increases sweep time. The relationship between RBW (B), noise floor (N), and sweep time (T) is given by:
Crystal or digital FIR filters are common, with modern analyzers offering RBW settings from 1 Hz to 10 MHz. The filter shape factor (typically 15:1 for 3 dB:60 dB ratio) defines out-of-band rejection.
Local Oscillator (LO) and Phase-Locked Loop (PLL)
The LO generates the tunable reference frequency for the mixer. Advanced analyzers use fractional-N PLL synthesizers with phase noise below -110 dBc/Hz at 10 kHz offset. The LO stability directly impacts frequency accuracy, with high-end instruments achieving ±0.1 ppm aging per year.
Detector and Video Filter
The detector converts IF signals to baseband using envelope, peak, or RMS detection. Video filtering (VBW) averages noise while preserving signal peaks, with the time constant τ related to VBW by:
Modern analyzers employ logarithmic amplifiers for 70+ dB dynamic range, with detector modes (positive peak, sample, normal) optimized for different signal types.
Display and Processing Unit
The display subsystem renders amplitude-vs-frequency data with configurable scales (log/linear). Digital IF architectures use FFT processing for real-time analysis, with metrics like adjacent channel power ratio (ACPR) and modulation analysis computed via DSP algorithms. High-resolution touchscreens now provide interactive marker functions and demodulation capabilities.
Reference Source and Calibration
An internal 10 MHz OCXO or rubidium reference ensures frequency stability. Automatic calibration routines adjust gain flatness (±0.5 dB typical) and correct for mixer conversion loss variations across frequency bands. Temperature-compensated alignment data is stored in non-volatile memory.

1.3 Types of Spectrum Analyzers
Spectrum analyzers are classified based on their underlying operational principles, frequency range, and application-specific optimizations. The primary categories include swept-tuned, real-time, vector, and Fourier analyzers, each offering distinct advantages in resolution, speed, and dynamic range.
Swept-Tuned Spectrum Analyzers
These analyzers employ a superheterodyne receiver architecture, where a voltage-controlled oscillator (VCO) sweeps across the frequency range of interest. The input signal mixes with the local oscillator (LO) frequency, producing an intermediate frequency (IF) that is filtered and detected. The key performance metrics are governed by:
where Δf is the resolution bandwidth (RBW), f₀ is the center frequency, and Q is the quality factor of the IF filter. Swept-tuned analyzers excel in high-frequency measurements (up to THz) but suffer from slower sweep times due to mechanical tuning limitations.
Real-Time Spectrum Analyzers (RTSA)
RTSAs capture and process time-domain data continuously using high-speed analog-to-digital converters (ADCs) and FPGA-based processing. They compute the power spectral density (PSD) via parallel Fast Fourier Transform (FFT) engines, enabling:
- Instantaneous bandwidths exceeding 1 GHz
- Detection of transient signals with nanosecond durations
- Overlap processing to eliminate blind times
The minimum detectable signal is constrained by the ADC's effective number of bits (ENOB) and the system noise floor:
Vector Signal Analyzers (VSA)
VSAs combine the functionality of a spectrum analyzer with demodulation capabilities. They capture both magnitude and phase information using I/Q sampling, enabling:
- Modulation analysis (e.g., EVM for 5G NR, Wi-Fi 6)
- Time-correlated multi-domain measurements
- Digital predistortion (DPD) characterization
The complex envelope representation of a modulated signal is given by:
FFT-Based Analyzers
These instruments directly compute the discrete Fourier transform (DFT) of time-domain samples. The frequency resolution is determined by:
where fₛ is the sampling rate and N is the number of points in the FFT. Window functions (e.g., Hann, Blackman-Harris) are applied to mitigate spectral leakage, with the equivalent noise bandwidth (ENBW) defined as:
Hybrid Architectures
Modern analyzers often integrate multiple techniques, such as swept-FFT designs that combine the wide frequency range of superheterodyne systems with the RBW flexibility of digital processing. Applications include:
- Radar cross-section measurements using pulsed-RF analysis
- Phase noise characterization with cross-correlation
- MIMO OTA testing with spatial scanning

2. Frequency Domain Analysis
2.1 Frequency Domain Analysis
Frequency domain analysis is a fundamental technique in signal processing, enabling the decomposition of signals into their constituent sinusoidal components. Unlike time-domain representations, which describe signal amplitude versus time, frequency-domain representations reveal spectral content—essential for identifying interference, harmonics, and noise.
Fourier Transform and Spectral Decomposition
The Fourier transform (FT) is the mathematical foundation of frequency domain analysis, converting a time-domain signal x(t) into its frequency-domain representation X(f):
For discrete signals, the Discrete Fourier Transform (DFT) is used, implemented computationally via the Fast Fourier Transform (FFT) algorithm. The power spectral density (PSD), derived from the squared magnitude of the FT, quantifies signal power distribution across frequencies:
Spectrum Analyzer Operation
Modern spectrum analyzers employ heterodyne receivers or FFT-based processing to measure PSD. Key operational modes include:
- Swept-tuned analyzers: Use a voltage-controlled oscillator (VCO) and mixer to downconvert a narrow frequency band at a time.
- Real-time FFT analyzers: Digitize the input signal and compute the FFT, suitable for transient or non-repetitive signals.
The resolution bandwidth (RBW) determines the analyzer's ability to distinguish closely spaced frequencies. A narrower RBW improves frequency resolution but increases sweep time:
Applications and Practical Considerations
Frequency domain analysis is critical in:
- RF communications: Measuring channel occupancy, modulation quality, and spurious emissions.
- Vibration analysis: Identifying resonant frequencies in mechanical systems.
- Noise reduction: Isolating and filtering unwanted spectral components.
Dynamic range and phase noise are key performance metrics. For example, a spectrum analyzer with a third-order intercept point (TOI) of +30 dBm can accurately measure intermodulation products in high-power RF amplifiers.
Advanced Techniques
For non-stationary signals, the Short-Time Fourier Transform (STFT) provides time-frequency localization:
where w(t) is a window function (e.g., Hamming, Blackman-Harris). Wavelet transforms offer multi-resolution analysis for signals with varying spectral characteristics over time.

2.2 Swept-Tuned vs. FFT-Based Analyzers
Spectrum analyzers fall into two primary architectures: swept-tuned and FFT-based. Their operational principles, performance trade-offs, and suitability for different applications stem from fundamental differences in signal processing methodology.
Swept-Tuned Spectrum Analyzers
Swept-tuned analyzers operate by sequentially tuning a narrowband filter across the frequency range of interest. The core components include:
- A voltage-controlled oscillator (VCO) that sweeps through frequencies
- A mixer to downconvert the input signal
- A resolution bandwidth (RBW) filter for frequency selection
- A detector (typically envelope or peak) for amplitude measurement
The instantaneous bandwidth is determined by the RBW filter, with narrower filters providing better frequency resolution at the cost of slower sweep times. The relationship between sweep time (Tsweep), span (Δf), and RBW is given by:
where k is a constant dependent on filter shape factor. For a 3 dB bandwidth Gaussian filter, k ≈ 2.
FFT-Based Spectrum Analyzers
FFT-based analyzers digitize a time-domain signal and compute the frequency spectrum via the Fast Fourier Transform. Key aspects include:
- An anti-aliasing filter to bandlimit the input
- An ADC with sufficient sampling rate and dynamic range
- A window function (e.g., Hanning, Flat Top) to control spectral leakage
- Parallel processing of all frequency bins simultaneously
The frequency resolution (ΔfFFT) is determined by the sampling rate (fs) and FFT size (N):
Unlike swept analyzers, FFT-based systems capture the entire span in a single acquisition, enabling real-time analysis of transient signals.
Comparative Performance
| Parameter | Swept-Tuned | FFT-Based |
|---|---|---|
| Frequency Range | Up to THz (with harmonic mixing) | Limited by ADC technology (typically < 100 GHz) |
| Dynamic Range | Superior close-in phase noise | Better spur-free range |
| Speed | Slower for narrow RBW | Faster for wide spans |
| Transient Capture | Misses brief events | Real-time capability |
Phase Noise Considerations
Swept analyzers exhibit phase noise determined by the VCO's phase-locked loop (PLL), typically following:
where F is the noise figure, f0 is the carrier frequency, and foffset is the offset from carrier. FFT-based systems inherit phase noise from the clock source but can achieve better close-in performance with ultra-low jitter oscillators.
Modern Hybrid Architectures
Contemporary high-performance analyzers often combine both techniques, using FFT processing for wide spans and swept methods for narrowband high-resolution measurements. Digital downconversion (DDC) extends FFT capabilities by:
- Digitally mixing the signal to baseband
- Applying decimation filters to reduce data rate
- Enabling zoom FFTs for detailed analysis

2.3 Resolution Bandwidth and Its Importance
Definition and Fundamental Role
The resolution bandwidth (RBW) of a spectrum analyzer defines the smallest frequency separation between two sinusoidal signals that can be distinguished. Mathematically, RBW is the 3-dB bandwidth of the intermediate frequency (IF) filter used in the analyzer. A narrower RBW improves frequency resolution but increases sweep time due to the filter's settling time.
Trade-offs Between RBW and Measurement Parameters
Selecting an appropriate RBW involves balancing three key factors:
- Frequency Resolution: Narrow RBW (e.g., 1 kHz) resolves closely spaced signals but requires longer measurement time.
- Noise Floor: RBW directly affects displayed average noise level (DANL). Halving RBW lowers the noise floor by 3 dB.
- Dynamic Range: Wide RBW (e.g., 1 MHz) captures fast transient signals but may obscure weak signals near stronger ones.
Mathematical Relationship to Noise Power
The noise power measured by a spectrum analyzer scales linearly with RBW:
where k is Boltzmann's constant, T is temperature in Kelvin, and B is the noise bandwidth (typically 1.05–1.2 × RBW for practical filters).
Practical Implementation Considerations
Modern spectrum analyzers implement RBW through digital IF filters with shape factors ranging from 3:1 to 5:1 (ratio of 60-dB to 3-dB bandwidths). For pulsed signals, the RBW should be at least 1/τ where τ is the pulse width to avoid amplitude inaccuracies.
Advanced Applications
In 5G NR measurements, RBW selection becomes critical for accurate adjacent channel leakage ratio (ACLR) tests. The 3GPP specification mandates RBW values between 1% and 3% of the channel bandwidth. For a 100 MHz channel, this translates to 1–3 MHz RBW depending on the emission mask requirement.
RBW Optimization Techniques
Automated RBW selection algorithms in modern instruments use signal detection heuristics:
- Peak search routines adjust RBW to maintain constant measurement uncertainty
- Density-based methods increase RBW for broadband signals while preserving narrowband features
- Adaptive filtering techniques minimize sweep time while meeting resolution requirements
For phase noise measurements, the RBW must be significantly narrower than the offset frequency being characterized—typically at least 10× smaller than the measurement offset to avoid contamination from the carrier.

3. RF and Microwave Signal Analysis
RF and Microwave Signal Analysis
Fundamentals of RF Signal Analysis
Spectrum analyzers measure the power spectral density of RF and microwave signals, providing critical insights into signal integrity, modulation characteristics, and noise behavior. The core principle involves heterodyne reception, where the input signal is mixed with a local oscillator (LO) to downconvert the frequency to an intermediate frequency (IF) for processing. The relationship between input frequency fin and LO frequency fLO is given by:
Modern analyzers employ fast Fourier transform (FFT) algorithms to compute the power spectrum S(f) from the time-domain signal x(t):
Key Performance Parameters
The dynamic range of a spectrum analyzer is determined by its noise floor and maximum input power. The displayed average noise level (DANL) is calculated as:
where k is Boltzmann's constant, T is temperature, B is resolution bandwidth, and NF is the noise figure. For a typical analyzer with 10 dB NF at 1 GHz:
Advanced Measurement Techniques
Phase noise characterization requires specialized techniques due to the close-in spectral components. The single-sideband (SSB) phase noise L(f) is measured as:
Modern analyzers implement cross-correlation methods to achieve phase noise measurements below -170 dBc/Hz at 1 kHz offset. For modulated signals, error vector magnitude (EVM) analysis decomposes the signal into in-phase (I) and quadrature (Q) components:
Microwave Measurement Considerations
Above 20 GHz, waveguide interfaces and harmonic mixing become necessary. The cutoff frequency for WR-90 waveguide is:
where a is the waveguide width (22.86 mm for WR-90). Millimeter-wave measurements require correction for atmospheric attenuation, which peaks at 60 GHz (15 dB/km) due to oxygen absorption.
Practical Applications
- Radar systems: Pulse repetition interval (PRI) analysis using zero-span mode
- 5G NR: Channel power measurements with integrated bandwidths up to 400 MHz
- Satellite communications: Adjacent channel leakage ratio (ACLR) testing
Advanced triggering capabilities allow capture of transient events as short as 10 ns, while real-time spectrum analyzers can detect signals with dwell times under 1 μs. The frequency mask trigger (FMT) function enables automatic detection of spectral violations in crowded environments.

3.2 Troubleshooting Electronic Circuits
Identifying Signal Anomalies
Spectrum analyzers excel at detecting deviations from expected signal behavior. When troubleshooting, the first step is to compare the measured spectrum against the theoretical or reference spectrum. Common anomalies include:
- Spurious emissions — Unintended frequency components caused by nonlinearities, mixing, or parasitic oscillations.
- Harmonic distortion — Integer multiples of the fundamental frequency due to amplifier saturation or clipping.
- Phase noise — Broadening of the spectral peak caused by oscillator instability.
For quantitative analysis, the signal-to-noise ratio (SNR) and total harmonic distortion (THD) can be computed directly from the spectrum. The THD is given by:
where V1 is the fundamental amplitude and V2 to Vn are harmonic amplitudes.
Localizing Noise Sources
Noise manifests as an elevated noise floor in the spectrum. To isolate its origin:
- Broadband noise — Uniformly distributed noise suggests thermal or shot noise sources.
- Narrowband interference — Discrete peaks indicate coupling from clocks, switching regulators, or RF sources.
Narrowing the resolution bandwidth (RBW) improves sensitivity to low-level signals. The noise floor reduction is:
For example, reducing RBW from 10 kHz to 1 kHz lowers the noise floor by 10 dB.
Diagnosing Modulation Issues
Modulated signals exhibit characteristic sidebands and envelopes. Common problems include:
- Asymmetric sidebands — Indicates I/Q imbalance in quadrature modulators.
- Excessive spectral regrowth — Caused by amplifier nonlinearity, visible as non-ideal ACPR (Adjacent Channel Power Ratio).
The modulation error vector magnitude (EVM) can be inferred from the constellation diagram derived from the spectrum:
Practical Debugging Workflow
- Set the analyzer to span 5× the signal bandwidth.
- Use max-hold to capture intermittent anomalies.
- Enable marker noise to measure SNR.
- Apply time-gating for pulsed or burst signals.
For intermittent faults, the persistence mode helps visualize rare events by color-coding frequency occupancy over time.

3.3 EMI/EMC Testing
Fundamentals of EMI/EMC Measurements
Electromagnetic interference (EMI) and electromagnetic compatibility (EMC) testing evaluate the unintentional generation, propagation, and reception of electromagnetic energy. A spectrum analyzer measures radiated and conducted emissions across frequency bands, comparing them against regulatory limits such as CISPR, FCC, or MIL-STD-461. The key parameters include:
- Radiated emissions: Measured using antennas in anechoic chambers, quantifying field strength (dBμV/m) over 30 MHz–6 GHz.
- Conducted emissions: Assessed via line impedance stabilization networks (LISNs) for frequencies below 30 MHz.
Spectrum Analyzer Configuration
For EMI testing, the analyzer must operate with:
where RBW (resolution bandwidth) is critical for distinguishing adjacent signals. The detector mode is set to peak for initial scans and quasi-peak or average for compliance testing. A preamplifier (e.g., 20 dB gain) and external attenuators mitigate noise floor limitations.
Measurement Uncertainty and Calibration
Total measurement uncertainty (TMU) combines:
where \(U_{ant}\) is antenna factor uncertainty, \(U_{cable}\) accounts for transmission line losses, and \(U_{SA}\) derives from the analyzer’s amplitude accuracy. Regular calibration using traceable RF sources (e.g., NIST) ensures ≤1.5 dB uncertainty.
Time-Domain Analysis for Transient EMI
Fast Fourier Transform (FFT)-based spectrum analyzers capture transient events like switch-mode power supply noise. The relationship between time-domain sampling and frequency resolution is:
where \(T_{acq}\) is the acquisition time. Overlap processing (≥50%) minimizes spectral leakage during repetitive sweeps.
Advanced Techniques: Real-Time Spectrum Analysis
Real-time spectrum analyzers (RTSAs) use parallel processing to detect intermittent EMI with 100% probability of intercept (POI). Key metrics include:
- Minimum detectable signal: \(MDS = -174 \text{ dBm/Hz} + NF + 10\log(RBW)\)
- Processing gain: \(G_p = 10\log(\frac{f_s}{2RBW})\) for digital IF stages
Case Study: Automotive EMC Testing
ISO 11452-2 requires radiated immunity testing at 200 V/m from 1–18 GHz. A spectrum analyzer with a tracking generator and directional coupler verifies field uniformity by comparing injected vs. measured power:
Deviations exceeding ±3 dB indicate chamber reflections or antenna positioning errors.

4. Real-Time Spectrum Analysis
4.1 Real-Time Spectrum Analysis
Real-time spectrum analysis (RTSA) is a critical technique for capturing and analyzing transient or rapidly varying signals that traditional swept-tuned or FFT-based spectrum analyzers may miss. Unlike conventional methods, RTSA processes the entire signal bandwidth continuously, ensuring no gaps in data acquisition.
Fundamentals of Real-Time Processing
The core principle of RTSA relies on high-speed analog-to-digital conversion (ADC) followed by real-time digital signal processing (DSP). The Nyquist criterion must be satisfied to avoid aliasing:
where fs is the sampling rate and B is the instantaneous bandwidth. Modern RTSA systems employ undersampling techniques and digital down-conversion (DDC) to extend their effective bandwidth beyond the ADC's Nyquist limit.
Key Performance Metrics
The performance of an RTSA system is characterized by three primary metrics:
- Real-time bandwidth (RTBW): The maximum frequency span that can be processed without gaps.
- Probability of intercept (POI): The likelihood of capturing a transient event.
- Processing latency: The time delay between signal reception and display.
For a system with a sampling rate fs and FFT size N, the time resolution Δt is given by:
Overlap Processing and Time Resolution
To improve time resolution, RTSA systems use overlap processing, where successive FFTs are computed with overlapping time windows. The overlap percentage P is defined as:
where Tupdate is the time between FFT updates and Twindow is the FFT window duration. Higher overlap improves POI but increases computational load.
Applications in Modern Systems
RTSA is indispensable in:
- Radar signal analysis: Detecting and characterizing pulsed RF signals with nanosecond durations.
- 5G/6G development: Analyzing dynamic spectrum sharing and beamforming signals.
- Electronic warfare: Identifying and classifying hostile emissions in real time.
Modern implementations leverage field-programmable gate arrays (FPGAs) for parallel processing, achieving real-time bandwidths exceeding 1 GHz with microsecond-level latency.
Mathematical Derivation of Minimum Detectable Duration
The shortest detectable transient duration τmin is determined by the system's frequency resolution Δf and windowing function. For a Hann window:
This relationship shows the fundamental trade-off between frequency resolution and time resolution in real-time analysis.

4.2 Tracking Generators and Their Use
A tracking generator is an essential auxiliary module in modern spectrum analyzers, enabling swept-frequency network analysis by synchronizing its output signal with the analyzer's local oscillator (LO). This allows for precise measurement of device-under-test (DUT) frequency response, including insertion loss, gain, and return loss.
Operating Principle
The tracking generator produces a sinusoidal signal whose frequency is locked to the spectrum analyzer's instantaneous tuned frequency. Mathematically, the output frequency fTG is given by:
where fLO is the local oscillator frequency and fIF is the intermediate frequency. The generated signal is injected into the DUT, and the spectrum analyzer measures the transmitted or reflected signal power at each frequency point.
Key Applications
- Filter Characterization: Measures amplitude response, bandwidth, and roll-off of RF/microwave filters.
- Amplifier Testing: Evaluates gain flatness, compression points, and harmonic distortion.
- Cable and Antenna Analysis: Determines insertion loss and VSWR across frequency bands.
Calibration and Error Correction
To minimize systematic errors, a two-port calibration is performed using known standards (open, short, load, thru). The corrected measurement S21,corr is derived from raw data S21,meas and error terms:
where EDF is directivity error, ERF is reflection tracking, and ETF is transmission tracking.
Practical Limitations
Phase information is not preserved in scalar measurements, restricting use to magnitude-only analysis. Dynamic range is constrained by the tracking generator's output power (typically +10 to -30 dBm) and the analyzer's noise floor. For wide sweeps (>1 GHz), power flatness corrections may be necessary.

4.3 Phase Noise Measurements
Phase noise quantifies the short-term frequency instability of an oscillator, appearing as sidebands around the carrier signal in the frequency domain. It is a critical parameter in RF and microwave systems, affecting communication link quality, radar resolution, and clock synchronization precision.
Fundamentals of Phase Noise
Phase noise arises from random fluctuations in the phase of an oscillator's output signal, typically caused by thermal noise, flicker noise, and vibration-induced jitter. The single-sideband (SSB) phase noise L(f) is defined as:
where fc is the carrier frequency, f is the offset frequency, and Psideband is the power in a 1 Hz bandwidth at offset f from the carrier.
Measurement Techniques
Direct Spectrum Analyzer Method
The most straightforward approach uses a spectrum analyzer to measure the power spectral density:
- Center the analyzer on the carrier frequency with sufficient resolution bandwidth (RBW)
- Measure the carrier power Pcarrier
- Measure the noise power at various offset frequencies f
- Correct for the analyzer's noise floor and RBW effects
The measurement accuracy is limited by the analyzer's phase noise floor, typically -140 to -170 dBc/Hz for high-end instruments.
Phase Detector Method
For improved sensitivity, a reference oscillator and phase detector can be used:
where Kd is the phase detector constant, Δφ(t) is the phase difference, and n(t) represents system noise. The power spectral density of Vout relates directly to the device under test's phase noise.
Advanced Measurement Considerations
When making precision phase noise measurements:
- Residual noise: The measurement system's own phase noise must be significantly lower than the DUT
- Harmonic mixing: Can extend measurement range but introduces conversion loss
- Delay line discrimination: Used for ultra-low phase noise measurements by correlating two identical signals
The modified Allan variance provides additional insight for long-term stability analysis:
Practical Measurement Challenges
Real-world phase noise measurements must account for:
- Amplitude modulation components that can corrupt phase noise readings
- Microphonic effects in crystal oscillators
- Power supply noise coupling into the oscillator circuit
- Temperature fluctuations during measurement
Modern phase noise analyzers often incorporate cross-correlation techniques between multiple measurement channels to suppress instrument noise and improve sensitivity by 10-15 dB.
5. Importance of Regular Calibration
5.1 Importance of Regular Calibration
The accuracy of a spectrum analyzer degrades over time due to component aging, thermal drift, and environmental stress. Calibration ensures that the instrument adheres to its specified performance metrics, including amplitude accuracy, frequency response, and phase noise. Without periodic calibration, measurement errors compound, leading to unreliable data in critical applications such as RF design, signal intelligence, and compliance testing.
Sources of Measurement Drift
Key components contributing to calibration drift include:
- Local Oscillator (LO) stability: Frequency synthesizers experience phase noise degradation over time, affecting resolution bandwidth accuracy.
- Mixer nonlinearity: Diode-based mixers develop compression point variations, distorting harmonic and intermodulation measurements.
- IF filter shaping: Component tolerances in intermediate frequency filters alter the effective noise bandwidth.
- Reference clock accuracy: Crystal oscillators drift with temperature cycles, introducing frequency errors.
Quantifying Calibration Errors
The total measurement uncertainty Utotal combines systematic errors from calibration drift (Ucal) and random noise (Unoise):
For a typical spectrum analyzer, the dominant terms in Ucal include:
Where ΔG/G is gain variation, Δf/f is frequency error, and ΔP/P is power measurement deviation.
Calibration Standards and Traceability
Modern spectrum analyzers use NIST-traceable calibration with these reference standards:
- Amplitude: Calibrated signal generators with ±0.1 dB uncertainty at 1 GHz
- Frequency: Rubidium atomic clocks (1×10-11 stability)
- Noise floor: Thermal noise sources with ENR certified to ±0.05 dB
The calibration process typically follows ANSI/NCSL Z540-1 or ISO/IEC 17025 protocols, documenting measurement uncertainty at each test point across the frequency range.
Recommended Calibration Intervals
Industry-standard calibration intervals balance operational needs with measurement integrity:
| Application | Interval | Tolerances |
|---|---|---|
| Research labs | 6 months | ±0.3 dB amplitude, ±1 ppm frequency |
| Production testing | 12 months | ±0.5 dB amplitude, ±5 ppm frequency |
| Field measurements | 24 months | ±1.0 dB amplitude, ±10 ppm frequency |
High-precision applications like satellite communications often require monthly verifications using in-situ calibration techniques.
Automated Calibration Systems
Modern analyzers implement self-calibration routines for critical parameters:
- Internal power calibrators: Reference tones at -30 dBm and -50 dBm for amplitude verification
- Phase-locked loop diagnostics: Automated testing of LO phase noise and spurious emissions
- Digital correction algorithms: Stored error correction tables compensate for nonlinearities
These systems reduce downtime while maintaining measurement traceability between full calibrations.
5.2 Common Calibration Procedures
Spectrum analyzer calibration ensures measurement accuracy by compensating for systematic errors in the instrument. Advanced calibration procedures involve both hardware adjustments and software corrections, often requiring traceable reference standards.
Frequency Response Calibration
The frequency response calibration corrects amplitude variations across the analyzer's frequency range. A known flat signal source, such as a calibrated noise generator or comb generator, is used as a reference. The procedure involves:
- Connecting the reference source to the analyzer input
- Sweeping across the full frequency range
- Recording amplitude deviations at discrete frequency points
- Creating a correction table stored in non-volatile memory
where C(f) is the frequency-dependent correction factor, Vref(f) is the known reference amplitude, and Vmeas(f) is the measured amplitude.
Amplitude Accuracy Calibration
This procedure verifies and corrects absolute amplitude measurement accuracy using precision signal sources:
- RF power meter with known calibration factor serves as reference
- Multiple power levels are tested across dynamic range
- Correction factors are applied to mixer compression characteristics
- Temperature-dependent variations are characterized
Modern analyzers use vector error correction techniques that account for both magnitude and phase errors in the signal path.
Phase Noise Calibration
Critical for modulation analysis and low-level signal detection, phase noise calibration involves:
- Comparison against a low-noise reference oscillator
- Characterization of the local oscillator phase noise profile
- Measurement of close-in and far-out phase noise
- Creation of noise floor correction tables
Harmonic Distortion Calibration
To accurately measure harmonic content, the analyzer must first characterize its own distortion products:
- Pure tone input signal at known power levels
- Measurement of second and third harmonic generation
- Nonlinearity modeling of mixer and amplifier stages
- Correction algorithm implementation
where THD is total harmonic distortion, Pn is the power of the nth harmonic, and P1 is the fundamental power.
Temperature Compensation
High-performance analyzers implement temperature-dependent calibration:
- Internal temperature sensors monitor critical components
- Pre-characterized thermal drift coefficients are applied
- Real-time compensation of frequency and amplitude parameters
- Periodic recalibration triggers based on thermal history
For metrology-grade measurements, calibration procedures must follow documented standards such as ISO/IEC 17025, with traceability to national measurement institutes.
5.3 Troubleshooting Common Issues
Noise Floor Anomalies
Unexpected rises in the noise floor often stem from internal or external interference. Internally, degraded mixer performance or excessive local oscillator (LO) phase noise can elevate the displayed noise floor. Externally, ambient RF noise or ground loops may introduce spurious signals. Verify the noise floor by:
- Disconnecting all inputs and terminating the analyzer with a 50Ω load.
- Comparing the displayed noise floor against the theoretical thermal noise limit:
$$ P_{noise} = -174 \text{dBm/Hz} + 10 \log_{10}(RBW) + NF $$where RBW is the resolution bandwidth and NF is the analyzer's noise figure.
Frequency Drift and Instability
Frequency drift in swept-tuned analyzers typically arises from aging oven-controlled crystal oscillators (OCXOs) or thermal stress on reference clock circuits. For modern vector signal analyzers (VSAs), phase-locked loop (PLL) settling errors or software-based frequency correction faults may manifest as apparent drift. Mitigation strategies include:
- Preheating the analyzer for 30+ minutes before critical measurements.
- Using an external atomic reference (e.g., rubidium standard) for long-duration tests.
- Monitoring the 10 MHz reference output with a frequency counter to isolate hardware vs. software issues.
Spurious Signals and Intermodulation
Non-harmonic spurs often indicate mixer nonlinearity or power supply contamination. To distinguish internal artifacts from DUT-generated signals:
- Reduce input attenuation by 10 dB—internal spurs will decrease by ≤10 dB while external signals drop proportionally to the attenuation change.
- Inject a clean CW tone and observe whether spur amplitudes follow the expected nth-order intercept point (IPn) behavior:
$$ P_{spur} = nP_{in} - (n-1)IPn $$
Amplitude Accuracy Errors
Deviations exceeding the specified ±0.5 dB typically stem from:
- Calibration drift: Verify using a calibrated power sensor at multiple frequencies.
- Input stage compression: Test with –30 dBm and –10 dBm signals—readings should scale linearly.
- IF gain errors: Most analyzers provide a built-in calibration signal for IF path verification.
Display Artifacts in FFT Mode
When operating in FFT mode, spectral leakage and scalloping loss become prominent with improper windowing. For a sequence of N samples, the worst-case scalloping loss for a rectangular window is:
Mitigate this by selecting window functions (Hanning, Flat Top) matched to the signal characteristics, trading off frequency resolution against amplitude accuracy.
Phase Noise Measurement Pitfalls
When measuring phase noise below –150 dBc/Hz, analyzer self-noise dominates. The measurement sensitivity limit is given by:
For accurate measurements below this floor, employ cross-correlation techniques using dual reference receivers or external phase noise test sets.
Reference Level Saturation
Input stages may exhibit nonlinear behavior when operated near the maximum specified input power. The compression point P1dB defines where gain deviates by 1 dB from linear:
Always maintain input power at least 10 dB below P1dB for distortion-free measurements.
6. Recommended Books and Publications
6.1 Recommended Books and Publications
- Portable spectroscopy and spectrometry. 1, Technologies ... — Stanford Libraries' official online search tool for books, media, journals ... Spectrum analysis. Spectromètres. Bibliographic information. Publication date 2021 Title variation Technologies, instrumentation and applications ISBN 9781119636489 (electronic bk. ; oBook) 1119636485 (electronic bk. ; oBook) 9781119636366 1119636361 DOI 10.1002 ...
- PDF Electrical Spectrum and Network Analyzers: A Practical Approach — Spectra and Spectrum Analysis 1.1 Waveforms: A Review of Basics 1 1.2 The Fourier Transform 2 1.3 The Fourier Series 3 1.4 Measuring Spectra 8 1.5 Standards for Spectrum Analyzers 8 2 Methods of Spectrum Analysis 2.1 Overview 14 2.2 The Parallel Filter Analyzer 15 2.3 The Superheterodyne Analyzer 17 2.3.1 Determining the Internal Frequencies 20
- PDF Electronic and Photoelectron Spectroscopy - Cambridge University Press ... — 16.2 Assignment and analysis of the rotational structure 141 16.3 Band head formation 143 References 143 17 Photoionization spectrum of diphenylamine: an unusual illustration of the Franck-Condon principle 144 References 149 18 Vibrational structure in the electronic spectrum of 1,4-benzodioxan: assignment of low frequency modes 150
- PDF Modern RF and Microwave Measurement Techniques — 4 Real-time spectrum analysis and time-correlated measurements applied to nonlinear system characterization 64 Marcus Da Silva 4.1 Introduction 64 4.1.1 Types of spectrum analyzers 65 4.2 Spectrum analysis in real-time 68 4.2.1 Real-time criteria 69 4.2.2 Theoretical background 69 4.3 Spectrum analysis using discrete Fourier transforms 70
- PDF Introduction to Electromagnetic Compatibility — Output Spectrum of a Linear System 140 3.3 Spectrum Analyzers 142 3.3.1 Basic Principles 142 3.3.2 Peak versus Quasi-Peak versus Average 146 3.4 Representation of Nonperiodic Waveforms 148 3.4.1 The Fourier Transform 148 3.4.2 Response of Linear Systems to Nonperiodic Inputs 151 3.5 Representation of Random (Data) Signals 151
- Electronic Measurements and Instrumentation[Book] - O'Reilly Media — 1.7 - Statistical Analysis. 1.7.1 - Probability of Errors and Gaussian Curve; 1.8 - Measurement Standards ... 3.10 - Spectrum Analyser. 3.10.1 - Characteristics of a Spectrum Analyser; 3.10.2 - Applications of a Spectrum Analyser ... book. Encyclopedia of Electronic Components Volume 2. by Charles Platt, Fredrik Jansson ...
- SDR Books & Publications - BelmY/wikiSDoR GitHub Wiki — Recommended background: ECE 3311, MA 2621, familiarity with Simulink, familiarity with general programming. ECE 502. Analysis of Probabilistic Signals and Systems Applications of probability theory and its engineering applications. Random variables, distribution and density functions. Functions of random variables, moments and characteristic ...
- Top RF Engineering Tools & Resources | Your Guide to Success — 3 Books and Publications for RF Engineers. 3.1 Recommended Reading Lists; ... enabling efficient troubleshooting during the electronic design automation process. ... Spectrum analyzers and network analyzers stand out as essential components in the realm of RF engineering, contributing significantly to the evaluation and optimization of various ...
- PDF Fundamentals of Spectrum Analysis - UVa — Spectrum Analyzer 17 3.1 Fourier analyzer (FFT analyzer) 17 3.2 Analyzers operating in accordance with the heterodyne principle 27 3.3 Main setting parameters 30 4 Practical Realization of an Analyzer Operating on the Heterodyne Principle 32 4.1 RF input section (frontend) 32 4.2 IF signal processing 44
- PDF Fundamentals of Spectrum Analysis - TU Delft — Fundamentals of Spectrum Analysis 14 Fig. 2-4 Magnitude spectrum of approximated rectangular signal shown in Fig. 2-3 Fig. 2-5 shows some further examples of periodic signals in the time and frequency domain Non-periodic signals Signals with a non-periodic characteristic in the time domain cannot be described by a Fourier series.
6.2 Online Resources and Tutorials
- PDF Fundamentals of Spectrum Analysis - UVa — 6.2.5.1 Detectors, time constants 186 6.2.5.2 Measurement bandwidths 190 6.3 Channel and adjacent-channel power measurement 190 6.3.1 Introduction 190 6.3.2 Key parameters for adjacent-channel power measurement 193 6.3.3 Dynamic range in adjacent-channel power measurements 194 6.3.4 Methods for adjacent-channel power measurement
- PDF How to use spectrum analyzers for EMC testing - Tekbox — How to use spectrum analyzers for EMC testing_V1_3_20220409.docx V1.3 How to correctly use spectrum analyzers for EMC pre-compliance tests Author: Mayerhofer 2 of 18 TekBox Digital Solutions 9-Apr-22 1 Introduction A spectrum analyzer is a key instrument for conducting EMC testing. Analyzers with EMC-specific features
- Review: Tektronix RSA306 spectrum analyzer (part 1) - EDN — General Signal Analysis: Description: Spectrum analyzer: Spans from 100 Hz to 6.2 GHz, 3 traces + math and spectrogram trace, 5 markers with power, relative power, integrated power, power density and dBc/Hz functions. DPX Spectrum/Spectrogram: Real time display of spectrum with 100% probability of intercept of 100 usec signals in up to 40 MHz span.
- PDF Lab Assignment 1 - Spectrum Analyzers - Department of Electrical and ... — 3 Demonstrate use of the spectrum analyzer demodulation function to recover audio from a few stations. 4 Configure the spectrum analyzer to measure signals over the range from 88 to 108 MHz. Observe how spectrum analyzer settings, including the RF attenuation resolution bandwidth, video bandwidth, and sweep time, affect the display.
- PDF RSA306B Real-Time Spectrum Analyzer Installation and Safety ... - Tektronix — The RSA306B is a portable, 40 MHz Real-Time Spectrum Analyzer that contains an acquisition system inside a small module. The user interface and display resides on a user-provided PC running SignalVu-PC software or a user-developed application. The host PC provides all power, control, and data signals over the USB 3.0 cable included with the ...
- TEKTRONIX SPECTRUM ANALYZER OWNER'S MANUAL Pdf Download — Page 5 These advanced triggers provide the ability to capture a seamless record of RF signals into memory and perform time-correlated, multi-domain analysis. Unlike traditional swept spectrum analyzers, the Real-Time Spectrum Analyzer offers a Power Trigger with user-settable span, enabling spectrum capture whenever the power of any signal ...
- RF Education and Teaching - Tektronix — Add Vector analysis, frequency and phase trends, Wi-Fi demodulation and more. Our smallest USB Spectrum Analyzer is full featured ( real-time 6.2 GHz) yet smaller than a paperback book! It is perfect for in-class or field demonstrations, lecture hall teaching as well as lab work.
- Spectrum Analysis Measurements - Related Documentation - HP Memory Project — The AN150 series of application notes became almost a generic title for engineers. One by one they recap all the basic knowledge required to achieve the best measurements possible using the 1970-era spectrum analyzers. Today's spectrum analyzers integrate a lot of processor power to improve the ease of use and to gain in measurement accuracy.
- Review: Tektronix RSA306 Spectrum Analyzer, Part 1 — The spectrum analyzer has always been a vital tool for the EMC engineer. Until the last few years, these instruments have been rather large and heavy desktop instruments, weighing up to 60 pounds, or more. With the breakthroughs in components used for wireless technology, the size and weight of these instruments has decreased dramatically.
- RSA306B USB Spectrum Analyzer | Testforce — Advanced vector signal analysis at your fingertips. The RSA Series is powered by SignalVu-PC Vector Analysis software - the same software used on the full-line of Tektronix Real-time Spectrum Analyzers. Delivering unlimited access to advanced measurement and analysis capabilities that you can save, customize or share with peers.
6.3 Industry Standards and Guidelines
- US EMC Published Standards | National Standards Institute — Purchase: Search IEEE Standards - Enter C63 Standard number then Search (Enter) - Click on the version you want - Click on Purchase. C63.4: C63.4-2014 American National Standard for Methods of Measurement of Radio- Noise Emissions from Low-Voltage Electrical and Electronic Equipment in the Range of 9 kHz to 40 GHz.
- IEC 61000-6-3 - Electromagnetic compatibility (EMC) - EMC Directory — This part of IEC 61000 also applies to electrical and electronic equipment intended for use at other locations that do not fall within the scope of IEC 61000-6-8 or IEC 61000-6-4. The intention is that all equipment used in the residential, commercial and light-industrial environments are covered by IEC 61000-6-3 or IEC 61000-6-8.
- Iec 61326-1:2020 — Track status of your standards library; Create an account. Products ... Electrotechnical Commission) is the world's leading organization for the preparation and publication of international standards for all electrical, electronic and related technologies. International Standards facilitate technical innovation, efficient and sustainable ...
- PDF EN 300 386 - V1.6.1 - Electromagnetic compatibility and Radio spectrum ... — ETSI 2 ETSI EN 300 386 V1.6.1 (2012-09) Reference REN/ERM-EMC-312 Keywords EMC, network, testing ETSI 650 Route des Lucioles F-06921 Sophia Antipolis Cedex - FRANCE
- PDF Fundamentals of Spectrum Analysis - UVa — 6.3.4.2 Spectral power weighting with modulation filter (IS-136, TETRA, WCDMA) 198 6.3.4.3 Channel power measurement in time domain 200 6.3.4.4 Spectral measurements on TDMA systems 201 References 204 The current spectrum analyzer models from Rohde & Schwarz 207 Block diagram of spectrum analyzer described in this book 220 Table of ConTenTs
- PDF How to use spectrum analyzers for EMC testing - Tekbox — Typically, the resolution bandwidth filters on spectrum analyzers use Gaussian shaped intermediate frequency (IF) filters with adjustable bandwidths that follow a 1 -3 -10 sequence, e.g. 100 Hz, 300Hz, 1 kHz, ïkHz, ì kHz, ì kHz… In order to be compliant with CISPR standards, the spectrum analyzer must additionally provide so called
- PDF RSA603A and RSA607A Series Real-Time Spectrum Analyzers ... - Tektronix — Observe the following guidelines when recycling an instrument or component: Equipment recycling Production of this equipment required the extraction and use of natural resources. The equipment may contain substances that could be harmful to the environment or human health if improperly handled at the product's end of life.
- PDF A Guide to United States Electrical and Electronic Equipment ... - NIST — This guide addresses electrical and electronic consumer products, including those that will . In addition, it includes electrical and electronic products used in the workplace as well as electrical and electronic medical devices. The scope does not include vehicles or components of vehicles, electric or electronic toys, or recycling ...
- List of common EMC test standards - Wikipedia — The following list outlines a number of electromagnetic compatibility (EMC) standards which are known at the time of writing to be either available or have been made available for public comment. These standards attempt to standardize product EMC performance, with respect to conducted or radiated radio interference from electrical or electronic equipment, imposition of other types of ...
- How to correctly use spectrum analyzers for EMC pre-compliance tests — Spectrum analyzers sweep the frequency range in discrete steps. Typically, the number of frequency steps per sweep is identical with the number of display pixels in X-direction. The Siglent SSA3021X, as an example, has a resolution of 751 equidistant frequency points per sweep. Other common spectrum analyzers have 601 measurement points per sweep.








