Keep Detector and Bandwidth Consistent When Comparing EMI Traces

Two EMI traces can look inconsistent because the instruments evaluated the same signal differently. Detector type, bandwidth and observation time are part of the measurement definition. Compare those settings before concluding that the product’s emissions changed between a development bench and a test laboratory.

Identify the detector actually used

Rohde & Schwarz’s ESU operating manual describes different detector responses, including quasi-peak evaluation of pulsed disturbances. A peak, average or quasi-peak result is not simply another label on the same amplitude value. Use the required measurement method for the product’s actual evaluation.

Record instrument mode, detector, resolution bandwidth, relevant filter characteristics, dwell or sweep time and trace processing. Distinguish an average detector from averaging several displayed traces. Similar-looking menu options can implement different operations.

Retain the signal’s time behavior

Document whether the source is continuous, periodic, bursty or dependent on an occasional operating event. A short sweep can miss infrequent activity even when its bandwidth appears correct. Use an observation strategy that covers the relevant product behavior.

For example, a display interface active only during a screen update may produce a different result from continuous transfer. Keep the update sequence and repetition interval with the test record. Do not compare that trace with an idle-screen measurement and attribute the difference solely to a board revision.

Reconcile the setups systematically

Start by matching frequency range, detector and bandwidth, then inspect observation time and operating mode. Also compare transducers, cables, correction factors and reference levels. Avoid changing all settings simultaneously, because that makes the cause of the discrepancy difficult to identify.

If the bench instrument does not implement the laboratory’s detector or filter requirements, describe its result as diagnostic evidence with that limitation. It can still be valuable for controlled comparisons. It should not be presented as an equivalent formal result simply because the display uses the same amplitude units.

Use limits with their required method

Associate each limit line with the applicable measurement procedure and product configuration. Do not move a peak trace against an unrelated average limit and report a pass or failure without explaining the mismatch. Resolve the correct evaluation with the responsible test laboratory.

The final comparison should include the settings and operating state beside the relevant traces, plus the remaining differences between setups. Preserve exported configuration files where available. This gives another engineer enough information to reproduce the conclusion and prevents a detector change from being reported as an electrical improvement or a new hardware defect.