A spectrum analyzer displays signal strength across a range of frequencies. The horizontal axis shows frequency (in Hz, MHz, or GHz), and the vertical axis shows amplitude, which is the power or strength of each signal (typically in dBm). Once you understand these two axes and a handful of key controls, you can interpret nearly any measurement the instrument displays.
The Two Axes: Frequency and Amplitude
The horizontal axis represents the range of frequencies being measured. It might span from 100 MHz to 1 GHz, or zoom into a narrow slice like 840 MHz to 860 MHz. Frequency tells you how fast a signal oscillates, measured in hertz. You’ll see labels in kHz, MHz, or GHz depending on what you’re looking at.
The vertical axis represents signal level, or amplitude. This is almost always displayed in decibels relative to one milliwatt (dBm), though you may also encounter dBV or dBmV. The scale is logarithmic, meaning each division represents a multiplication of power rather than a simple addition. A signal at -10 dBm is ten times more powerful than one at -20 dBm. The top of the display is the strongest signal level, and the bottom is the weakest. By default, many analyzers set the top line at 0 dBm.
Center Frequency and Span
These two controls define your viewing window. Center frequency places a specific frequency in the middle of the screen, and span sets how wide the display is on either side. If you set a center of 850 MHz and a span of 20 MHz, you’re looking at everything between 840 MHz and 860 MHz. You can also enter start and stop frequencies directly, but center and span are used more often because they make it easy to zoom in or out on a signal of interest. To zoom in, decrease the span. To see a wider picture, increase it.
Reference Level and Input Attenuation
The reference level defines what power level the top line of the display corresponds to. If it’s set to 0 dBm, then a signal reaching the top of the screen is at 0 dBm. If your signal of interest is weak and buried near the bottom, lowering the reference level brings it into better view. If your signal is clipping at the top, raise the reference level.
Input attenuation is a protective setting that reduces the power reaching the analyzer’s sensitive internal components. Strong signals can overload or even damage the front-end mixer. If the analyzer warns you that the input signal is too high, increase the attenuation (30 dB is a common safe choice). Most spectrum analyzers have a maximum safe input level of +30 dBm (1 watt) with at least 10 dB of attenuation engaged. As a rule, never reduce attenuation below 10 dB unless you are completely certain the input power is well within safe limits.
Resolution Bandwidth (RBW)
Resolution bandwidth controls how finely the analyzer can distinguish between two signals that are close together in frequency. Think of RBW as the width of the lens the analyzer uses to scan across the spectrum. A narrow RBW can separate two signals that are only a few kHz apart, while a wide RBW lumps them together into a single blob.
RBW also directly affects the noise floor, which is the baseline level of internal noise the analyzer generates even when no signal is present. Narrowing the RBW reduces the displayed noise, making it easier to spot weak signals. The tradeoff is speed: a narrower RBW requires a slower sweep across the frequency range. If you force a fast sweep with a narrow RBW, the analyzer may display an “UNCAL” warning, meaning the amplitude readings are no longer trustworthy. Most analyzers can automatically couple the sweep time to your RBW and span settings to avoid this.
Video Bandwidth (VBW)
Video bandwidth smooths the displayed trace by averaging out rapid fluctuations. It acts as a filter on the final displayed signal rather than on the measurement itself. A low VBW relative to RBW produces a clean, stable trace that’s easier to read. A high VBW shows every jitter and fluctuation in the signal.
The ratio of VBW to RBW is what matters. When VBW is much smaller than RBW (a ratio of 0.01 or less), noisy fluctuations are almost entirely suppressed, revealing the true shape of the spectrum. Setting VBW equal to or greater than RBW leaves the trace noisy and harder to interpret. For most measurements, leaving VBW coupled to RBW in a 1:1 ratio is a reasonable starting point, and you can reduce VBW from there when you need a cleaner display.
The Noise Floor and Sensitivity
Every spectrum analyzer has a noise floor: the level of internally generated noise that appears on the display even with nothing connected to the input (just a 50-ohm termination). This is formally called the Displayed Average Noise Level, or DANL. It sets the lower limit on what you can measure. Any signal weaker than the noise floor will be invisible.
DANL is specified at a given RBW, such as -135 dBm at 1 kHz RBW. Narrowing the RBW lowers the displayed noise floor and improves your ability to detect weak signals. Some analyzers also include built-in preamplifiers that reduce the overall noise figure and push the noise floor even lower. If you’re hunting for a faint signal, try reducing the RBW and enabling the preamp if your instrument has one.
What Common Signals Look Like
A pure sine wave (called a continuous wave, or CW signal) appears as a single sharp peak rising above the noise floor. The narrower your RBW, the thinner and more defined that peak becomes. This is the simplest pattern you’ll encounter.
Harmonic distortion shows up as additional peaks at exact multiples of the original frequency. If your fundamental signal is at 100 MHz, a second harmonic appears at 200 MHz, a third at 300 MHz, and so on. These secondary peaks are typically lower in amplitude than the fundamental. As you increase the input power, harmonics become more prominent. Seeing unexpected peaks at integer multiples of your signal is a clear sign of distortion, either in the device under test or, if you’re not careful with attenuation, in the analyzer itself.
Amplitude modulation (AM) produces a carrier peak in the center with symmetrical sidebands on either side. The spacing between the carrier and each sideband equals the modulation rate. If you modulate a 500 MHz carrier at 10 kHz, you’ll see peaks at 500 MHz, 499.99 MHz, and 500.01 MHz. With deeper modulation, additional harmonic pairs appear further out.
Frequency modulation (FM) looks different depending on the deviation. Narrow-deviation FM resembles AM, with a strong carrier and modest sidebands. Wide-deviation FM spreads energy across many peaks, with no single dominant carrier. In both cases, the spacing between adjacent peaks still equals the modulation rate.
Using Markers to Measure
Markers are on-screen cursors you place on the trace to read exact frequency and amplitude values. Most analyzers have a “peak search” function that automatically snaps a marker to the highest point on the display. This is the fastest way to identify the frequency and power of the strongest signal.
Delta markers let you measure the difference between two points. You place a reference marker on one peak, then a delta marker on another. The readout shows the frequency offset and power difference between them. This is useful for measuring harmonic levels relative to the fundamental, checking sideband spacing, or quantifying the gap between a signal and the noise floor. If your second harmonic is 35 dB below the fundamental, the delta marker will read -35 dB directly.
Putting It All Together
A practical workflow for reading an unfamiliar signal starts with setting a wide span to survey the spectrum. Use the peak search marker to find the signal of interest, then center it on screen and narrow the span to see detail. Adjust the reference level so the peak sits near the top of the display without clipping. Narrow the RBW if you need to separate closely spaced signals or see below a noisy floor. Lower the VBW if the trace is too jagged to read clearly. Place markers to record specific frequency and power values.
Each control involves a tradeoff. Narrow RBW gives better resolution but slows the sweep. Low VBW gives a smoother trace but also slows things down. Lower reference levels reveal weak signals but can cause strong signals to overdrive the display. Once you develop an intuition for these tradeoffs, reading a spectrum analyzer becomes a straightforward process of zooming in, cleaning up, and measuring.