Spectrum Analyzer Controls
Knowing that a spectrum analyzer exists and what it does in theory is only the beginning. The real skill is in setting the instrument's controls correctly for the measurement you need. Every spectrum analyzer — whether a $50 TinySA or a $50,000 Rohde & Schwarz — has the same fundamental set of controls, and understanding them deeply will make your measurements fast, accurate, and meaningful. Set them incorrectly and you will get results that are wrong or misleading in subtle ways you may not even notice.
This lesson explains each control in detail: what it does physically inside the instrument, how it affects the display, what the right setting is for different types of measurements, and what mistakes to watch for. Working through a specific example — measuring the harmonic output of a 40m transmitter — ties all the controls together into a coherent workflow.
Center Frequency and Span
These two controls define what portion of the spectrum is displayed. The center frequency sets the frequency at the horizontal center of the display. The span sets how wide a frequency range is shown — the total frequency range from the left edge to the right edge of the display.
For example, if you set the center frequency to 14.200 MHz and the span to 10 MHz, the left edge of the display will be at 14.200 − 5 = 9.2 MHz and the right edge at 14.200 + 5 = 19.2 MHz. The horizontal axis covers a 10 MHz window centered on your transmit frequency.
Many instruments also let you specify the measurement in terms of start frequency and stop frequency rather than center and span. These are equivalent: start = center − span/2, stop = center + span/2.
Choosing the Right Span
The right span depends entirely on what you are trying to measure. Two general principles apply:
Use a wide span first, then zoom in. Starting with a wide span (for example, 0 to 50 MHz to cover the HF spectrum) gives you the big picture and reveals signals you might not have expected. Once you have identified the signals of interest, narrow the span to see them more clearly.
Make the span wide enough to see all relevant signals simultaneously. For a harmonic measurement on a 40m (7 MHz) transmitter, you want to see the fundamental (7 MHz), second harmonic (14 MHz), and third harmonic (21 MHz) all on screen at once. A span of 0 to 30 MHz would accomplish this. Once you have confirmed where the harmonics are and their approximate levels, you can narrow the span to measure each one precisely.
Transmitter frequency: 7.150 MHz. You need to find and measure harmonics up to the fifth (35.75 MHz).
Initial setup: Center = 18 MHz, Span = 40 MHz (showing 0 to 40 MHz). This puts the fundamental, second, third, fourth, and fifth harmonics all on screen.
After identifying harmonic positions: Zoom to center = 14.3 MHz, span = 2 MHz to measure the second harmonic precisely. Then center = 21.45 MHz, span = 2 MHz for the third harmonic, etc.
Resolution Bandwidth (RBW)
The resolution bandwidth is the bandwidth of the IF filter that selects each frequency slice during the spectrum analyzer's sweep. It is the most important control affecting measurement accuracy and quality, and it has two competing effects that must be balanced: frequency resolution (ability to distinguish nearby signals) and noise floor (minimum measurable signal level).
How RBW Affects Frequency Resolution
If two signals are separated by less than the RBW, the analyzer cannot distinguish them — it sees a single broadened peak rather than two separate peaks. To resolve two signals 5 kHz apart, you need an RBW of 3 kHz or less. Two signals 100 kHz apart can be easily resolved even with a 30 kHz RBW.
For most amateur radio harmonic measurements, this is not a concern — harmonics are separated by the fundamental frequency (7 MHz, 14 MHz, 21 MHz...), which is far wider than any practical RBW. However, when looking for spurious products close to the fundamental, or when measuring the spectral occupancy of an SSB signal, narrow RBW becomes important.
How RBW Affects the Noise Floor
The noise floor decreases by 10 dB for every factor-of-10 reduction in RBW. At a 3 kHz RBW, the noise floor is 10 dB lower than at a 30 kHz RBW, and 20 dB lower than at a 300 kHz RBW. This is because a narrower filter passes less noise power from the analyzer's own thermal and electronics noise.
The relationship is: Noise Floor (dBm) = DANL (dBm/Hz) + 10·log₁₀(RBW in Hz)
A TinySA has a DANL of approximately −150 dBm/Hz. What is the displayed noise floor at different RBW settings?
RBW = 300 kHz (300,000 Hz): −150 + 10·log₁₀(300,000) = −150 + 54.8 = −95.2 dBm
RBW = 30 kHz (30,000 Hz): −150 + 10·log₁₀(30,000) = −150 + 44.8 = −105.2 dBm
RBW = 3 kHz (3,000 Hz): −150 + 10·log₁₀(3,000) = −150 + 34.8 = −115.2 dBm
Using the 3 kHz RBW gives 20 dB better noise floor than the 300 kHz RBW, allowing you to see signals 20 dB weaker. For harmonic measurements on a transmitter where harmonics might be 50–60 dBc below the fundamental, the narrower RBW is essential.
RBW Trade-off: Sweep Time
A narrower RBW filter takes longer to respond to a signal — the filter must be allowed to "ring down" before the analyzer moves to the next frequency. Professional analyzers automatically enforce a minimum sweep time based on the RBW and span: sweep time ≥ k × span / RBW², where k is a constant around 2–3. If you force a sweep faster than this, the displayed signal levels will be inaccurate — typically reading lower than the true level.
In practice: for harmonic measurements, a 3 kHz or 10 kHz RBW is usually appropriate. For quick "what's on the band" surveys, 100 kHz or 300 kHz RBW gives fast sweeps at the cost of poorer noise floor, which is acceptable when you only need to locate strong signals.
Video Bandwidth (VBW)
After the detector measures the amplitude at each frequency, the result is further processed by a video bandwidth filter — a low-pass filter applied to the detected amplitude signal. The VBW smooths the display by averaging out noise fluctuations.
With a wide VBW (VBW >> RBW), the display is "noisy" — you see the actual noise fluctuations of the spectrum. Each noise peak and dip is visible, and the noise floor appears jagged. This is the most accurate representation of the signal: each pixel shows the actual power in that RBW-wide slice, noise included.
With a narrow VBW (VBW << RBW), the display is smoothed. Noise fluctuations average out, and the noise floor appears as a smooth, flat line. Weak signals that would be hidden in the noise fluctuations become visible above this smoothed floor. A VBW of 1/10 of the RBW is a common choice for revealing weak signals, and a VBW of 1/100 of the RBW provides even more smoothing at the cost of slower sweeps.
The tradeoff is identical to RBW: narrowing VBW improves the ability to see weak signals but increases sweep time. A general rule: if you are looking for weak signals near the noise floor, set VBW = RBW/10. If you are making quick surveys looking for strong signals, set VBW = RBW.
Reference Level and Scale
The reference level is the power level displayed at the top of the screen. It sets the measurement range. If your reference level is +10 dBm and the scale is 10 dB/division, the display spans from +10 dBm (top) down to +10 − 10×10 = −90 dBm (bottom), covering a 100 dB range on screen.
The reference level should be set just above the level of the strongest signal you expect to see. Setting it too high wastes display space and compresses weak signals near the noise floor into a small region at the bottom. Setting it too low will cause strong signals to go off the top of the screen and may even cause overload in the analyzer's input stages (visible as compression or distortion of the displayed signal).
For a 40m harmonic measurement: if the fundamental is +43 dBm (20 watts), set the reference level to +50 dBm. Scale: 10 dB/division. The fundamental will appear near the top, and you will have 100 dB of dynamic range below it to see harmonics as weak as −50 dBm — 50 dBc below the fundamental. For a 50 dBc harmonic requirement, this is right at the edge; a 2 dB/division scale with a narrower range would be used to measure a specific harmonic precisely once you have located it.
Input Attenuator
Separate from the reference level, most spectrum analyzers have a front-panel or electronically controlled input attenuator that reduces the signal level before it reaches the sensitive mixer stage. This protects the input from overload and prevents intermodulation products generated inside the instrument from appearing as false signals on the display.
A good practice: always start with maximum attenuation (20–30 dB) when connecting an unknown signal source. Reduce attenuation only until the signals of interest are comfortably above the noise floor. If adding or removing attenuation changes the apparent signal levels by more than the attenuator change, something is wrong — the input may be overloaded or the signal may be at the noise floor.
Detector Modes
The detector converts the amplitude-modulated IF signal into a DC voltage that drives the display. Different detector modes process this conversion differently, and the choice significantly affects what you are actually measuring.
| Detector Mode | What It Measures | Best For | Notes |
|---|---|---|---|
| Peak (Positive Peak) | Highest amplitude in the RBW window | Measuring CW signals, finding signal peaks | Never misses a signal; overestimates noise floor; default for most measurements |
| Negative Peak | Lowest amplitude in the RBW window | Measuring dips, checking null depth | Useful for verifying filter notch depth |
| Sample | Amplitude at a single instant during each RBW window | Displaying noise level accurately | May miss signals; noise floor appears at correct level |
| Average (Power Average) | RMS (power-averaged) amplitude in the RBW window | Measuring modulated signals, noise power | Reads 2.5 dB below peak for noise; correct for average power of noise-like signals |
| RMS | True RMS of the envelope | Accurate power measurements of any signal type | Best choice for measuring modulated signals such as SSB or digital |
| Quasi-Peak | Weighted amplitude per CISPR standard | EMC compliance testing | Required by CISPR 16 for emissions testing |
For the vast majority of amateur radio measurements — measuring CW or SSB transmitters, checking harmonics, characterizing filters — the peak detector is the right choice. It never misses a signal and provides conservative (worst-case) harmonic level readings, which is exactly what you want for compliance testing.
Sweep Time
The sweep time is how long the analyzer takes to sweep from the start frequency to the stop frequency. It is determined by the span, RBW, and VBW settings. Professional analyzers compute the minimum valid sweep time automatically and enforce it; the user can increase sweep time beyond the minimum but cannot reduce it without getting inaccurate results.
The approximate relationship is: Sweep Time ≥ k × Span / (RBW × VBW), where k ≈ 2–3 for most architectures. For a span of 50 MHz, RBW of 10 kHz, and VBW of 1 kHz:
Sweep Time ≥ 2 × 50×10⁶ / (10×10³ × 1×10³) = 100×10⁶ / 10×10⁶ = 10 seconds
This means a careful measurement with narrow RBW and VBW over a wide span can take 10–30 seconds per sweep. This is normal and expected. If the instrument allows a 1-second sweep at these settings, the results will be meaningless.
Markers and Delta Markers
Markers are the most practical feature for reading accurate measurements from a spectrum analyzer display. Without markers, you must estimate signal levels by counting divisions from the reference level — inaccurate and tedious. With markers, you place a cursor on a signal peak and read the exact frequency and amplitude from a readout.
Single Marker
A single marker placed on a signal peak reads the frequency and amplitude. On a TinySA, pressing the marker button and rotating the tuning knob moves the marker. When placed on the 7 MHz fundamental of a transmitter, it might read: Freq = 7.150 MHz, Level = +43.2 dBm. This reading is reliable and accurate to the instrument's specification, unlike estimating from the grid scale.
Delta Marker
A delta marker is placed relative to a reference marker. You first place Marker 1 on the fundamental (the reference), then place the delta marker on a harmonic. The delta marker reads not the absolute level, but the difference in level between the two marker positions — the harmonic level in dBc.
Example: Marker 1 on 7.150 MHz fundamental reads +43 dBm. Delta marker on 14.300 MHz second harmonic reads −9 dBm. The delta marker displays Δ = −52 dB (the harmonic is 52 dBc below the fundamental). FCC Part 97 requires ≥43 dBc for transmitters above 5 watts — 52 dBc passes the test with 9 dB of margin. This entire measurement is done in seconds with the delta marker.
An annotated spectrum analyzer display showing a 7 MHz transmitter with its second harmonic at 14 MHz. Key controls are labeled: center frequency (7 MHz), span (30 MHz), reference level (+50 dBm), RBW (3 kHz), VBW (300 Hz). Marker 1 is on the fundamental and the delta marker is on the harmonic, reading the harmonic level directly in dBc.
View LargerPutting It Together: 40m Harmonic Measurement Workflow
Here is a complete workflow for measuring the harmonic output of a 40m CW transmitter running 20 W into a 50 Ω dummy load, using a TinySA with a 30 dB 100 W attenuator between the transmitter and analyzer input.
- Transmitter → 30 dB / 100 W attenuator → TinySA input (50 Ω)
- Transmitter output: +43 dBm (20 W). After attenuator: +43 − 30 = +13 dBm. Safe for analyzer.
- Start: 1 MHz, Stop: 50 MHz
- Reference level: +20 dBm, Scale: 10 dB/div
- RBW: 30 kHz, VBW: 3 kHz, Detector: Peak
- Identify fundamental at 7.0 MHz and visible harmonics at 14, 21, 28, 35 MHz
- Zoom to center = 14.0 MHz, span = 2 MHz
- Reference level: +20 dBm, Scale: 10 dB/div
- RBW: 3 kHz, VBW: 300 Hz (for better noise floor)
- Place Marker 1 on fundamental peak
- Place Delta Marker on harmonic peak
- Read delta in dBc: e.g. −55 dBc → passes FCC ≥43 dBc requirement
- The analyzer sees signals 30 dB lower than the actual transmitter output. The relative measurement (dBc) is unaffected by a flat attenuator. So the delta marker reading is already correct.
- To find absolute power at the transmitter output: analyzer reading + 30 dB. If analyzer reads −12 dBm at the harmonic, actual harmonic power = −12 + 30 = +18 dBm = 63 mW.
Frequently Asked Questions
What RBW should I use for most amateur radio measurements?
For harmonic and spurious emission measurements, 3 kHz to 10 kHz RBW works well. It gives a noise floor low enough to see harmonics 50–60 dBc below a typical transmitter. For a quick spectrum survey of an amateur band, 10–30 kHz is sufficient and sweeps quickly. For EMC compliance testing under CISPR standards, specific RBW values (200 Hz, 9 kHz, 120 kHz) are required by the standard. Always match RBW to the measurement objective — there is no single "best" setting.
My spectrum analyzer shows a noisy, jagged noise floor. How do I smooth it?
Narrow the VBW. Try setting VBW = RBW/10 — for a 10 kHz RBW, set VBW = 1 kHz. The display will smooth significantly at the cost of slower sweeps. You can also enable trace averaging (if available), which averages multiple sweeps together. Both approaches reveal weak signals by averaging out the random noise fluctuations that mask them.
Why does my signal level change when I change the input attenuator?
On a properly functioning analyzer with a properly scaled display, the displayed signal level should not change when you change the attenuator — the instrument compensates for the attenuation in the reference level readout. If the level does change, something is wrong: either the analyzer is overloaded at low attenuation (in which case adding attenuation reduces the displayed level to the correct value), or the signal is near the noise floor at high attenuation (in which case the signal rises above the noise as attenuation is reduced). Both conditions tell you something useful about the measurement setup.
Test Your Knowledge
Answer the questions below to check your understanding. Every answer can be found in the lesson above.