Measure the mechanism, not just the symptom

Extra-class practice questions often present several technically familiar terms and ask which one belongs to the measurement or failure in front of you. Start at the physical boundary: what quantity enters the instrument, what reference sets its accuracy or limit, and where an unwanted signal is first created.

Build a measurement chain

An oscilloscope shows amplitude against time; a frequency-domain instrument shows amplitude against frequency; a vector network analyzer applies a stimulus and measures reflection or transmission at its ports. Once the domain is fixed, locate the controlling reference. A digitizer's sampling rate limits faithfully represented frequency and undersampling folds energy into a false lower-frequency trace. Its reference voltage instead sets amplitude headroom. A frequency counter compares counted cycles with a gate interval, so its time base transfers its fractional error directly to the reading.

For a margin calculation, first find received level from the signed ledger. Then form the required level from receiver threshold plus the required signal-to-noise ratio. Margin is received level minus required level; preserve both minus signs until the final subtraction. For conducted or radiated interference, classify the path before choosing a remedy. Differential current leaves on one conductor and returns on another; a common-mode component moves in the same direction on the conductors and returns through the surroundings, so shielding material alone does not guarantee containment.

The traps

• Swapping instrument domains. Amplitude-versus-time belongs to the oscilloscope, while spurious transmitter outputs and swept filter response require frequency-domain measurements; SWR is not an axis label.
• Using the wrong internal reference. Converter reference voltage governs full-scale amplitude, sampling rate governs captured bandwidth, and a counter's time base governs its frequency accuracy. Confusing them also makes aliasing look like a vertical-calibration fault.
• Calling every nearby-signal problem intermodulation. Oscillator phase noise produces reciprocal mixing, a nonlinear device creates new combination frequencies, and blocking is tied to gain compression rather than the third-order intercept definition.
• Losing the logarithmic reference. Thermal noise is stated per unit bandwidth, bandwidth changes use 10 log₁₀ of the ratio, and dBm remains referenced to one milliwatt when converting to watts.
• Assigning one noise tool every job. An automatic notch can remove a wanted steady carrier, a blanker targets impulses and can distort strong signals, while digital reduction can help several broader noise classes.
• Treating separate grounds as extra protection. Entry protectors belong at one bonded panel so a surge raises their references together; an outlet, service panel or isolated rod leaves a damaging difference between paths.

Try it