Building a repeatable EMC pre-scan process

A pre-scan is only useful for design comparison if the test setup changes less than the product. Move a cable, alter an operating mode or replace an uncharacterised pre-amplifier, and an apparent improvement may have nothing to do with the new PCB revision.

What makes an EMC pre-scan repeatable?

A repeatable EMC pre-scan is not intended to reproduce every feature of a formal compliance measurement. Its purpose is to provide a stable engineering reference against which hardware, firmware, enclosure and cable changes can be compared. The process must control the variables that materially affect coupling, detection and EUT behaviour.

Repeatability does not mean obtaining an identical trace every time. Ambient signals, connector repeatability, thermal conditions and normal instrument uncertainty will introduce variation. The practical goal is to make that variation small enough that a genuine design change remains visible.

Start by defining what the pre-scan is intended to answer. A conducted emissions investigation on a mains power port requires a suitable LISN, receiver or spectrum analyser, controlled cabling and a defined load condition. A radiated emissions pre-scan uses an antenna or appropriate near-field probe, a controlled geometry and a documented measurement chain. Neither setup demonstrates immunity performance. Conducted RF immunity instead uses devices such as CDNs or applicable current-injection methods, while radiated RF immunity requires a calibrated field-generation arrangement.

The applicable product or product-family standard should guide the pre-scan even where the setup is deliberately simplified. CISPR 11, CISPR 32 and other CISPR publications may define emissions requirements for products within their respective scopes. IEC 61000-4-x publications are basic immunity test methods, not emissions limit standards. Check the latest active edition, product scope, ports, frequency ranges, detectors, bandwidths, limits and operating conditions rather than applying a familiar setup by habit.

Freeze the measurement setup before comparing revisions

Photograph and measure the setup before collecting a baseline. Record antenna height and polarisation, EUT orientation, separation distances, table construction, ground-plane arrangement, cable lengths, support equipment positions and bonding details. For bench-level debugging, mark positions with tape or use a simple non-conductive fixture. A photograph alone is rarely enough because perspective makes distances difficult to recover.

Cable geometry deserves particular attention. At radiated emissions frequencies, an attached cable may behave as a significant common-mode radiator. Moving it closer to a ground plane or changing the length hanging from a table can alter the measured field even though the source current has not changed. Fix cable routes, bend radii, excess-length treatment and termination arrangements. Do not tightly coil surplus cable unless that is representative of the intended installation and permitted by the applicable method.

For conducted emissions, use the same LISN port, EUT supply lead, protective earth arrangement and auxiliary equipment configuration for every revision. The LISN provides a defined impedance and measurement port for relevant power-port disturbance voltage measurements. It is not a conducted immunity device. Where two LISNs or artificial networks are required by the test arrangement, record their locations and terminations clearly.

Near-field probes are excellent diagnostic tools, but probe pressure, angle and distance have a large effect on amplitude. A probe held by hand cannot usually support close numerical comparisons between revisions. A small fixture or scanner reference grid turns a qualitative sniffing exercise into a much more repeatable diagnostic process.

Keep the receiver chain unchanged

Record the analyser or receiver model, input attenuation, reference level, detector, resolution bandwidth, video bandwidth where applicable, sweep or dwell settings, trace mode and any frequency segmentation. Detector and bandwidth selections must follow the intended measurement purpose and applicable standard. Peak detection is useful for fast exploratory scanning, while quasi-peak and average detectors may be required for final emissions assessment in specified frequency ranges. They are not interchangeable measures of disturbance severity.

If a pre-amplifier is used, document its gain, frequency range, power supply and position in the chain. Excess gain can overload the analyser or the pre-amplifier itself, producing harmonics and misleading broadband responses. Insufficient gain may leave the system noise floor too close to the signal of interest. Guidance on choosing and applying EMC pre-amplifiers can help avoid both conditions.

Include cables, attenuators, transient protection and adapters in the configuration record. Replacing a coaxial cable with one of different loss changes the displayed amplitude. For higher-confidence work, apply current transducer factors and cable-loss corrections consistently rather than comparing uncorrected traces from different measurement chains.

Define operating modes that expose the real emitters

An idle product often produces an attractive but commercially meaningless emissions plot. Exercise the functions that create the highest clock activity, switching current, data traffic, motor loading or external cable current. For a networked product, that may require continuous bidirectional traffic. A power converter may need representative input voltage and output load combinations. Display products should use controlled image patterns and brightness settings.

Firmware must also be identified. A change in spread-spectrum clocking, task scheduling, pulse-width modulation or communications timing can move or broaden spectral components. Record the firmware build, configuration files and any diagnostic commands used to force operating states. If an operating mode cannot be reproduced automatically, write a concise operator sequence and include expected indicators so another engineer can confirm that the EUT entered the intended state.

Use a mode matrix rather than attempting every possible combination. Select modes by coupling risk and intended use, then record why each was chosen. The matrix should identify:

• Hardware revision, populated options and enclosure state.
• Firmware build and configuration.
• Power source, input voltage and load condition.
• Active interfaces, cable types and terminations.
• Peripheral equipment and software versions.
• Expected activity, duty cycle and test duration.
• Any safety or functional monitoring required.

A mode described only as “worst case” is not reproducible. State what was active and why it was expected to produce the highest disturbance.

Build baselines that survive product iteration

The first baseline should include more than a saved screenshot. Retain native trace data where the instrument permits it, together with plots, setup photographs, configuration files and a short test log. Native data allows later overlays, marker analysis and revised correction factors. Screenshots flatten useful information and may conceal detector or bandwidth settings.

Measure the ambient environment with the EUT off but the support equipment operating. This identifies broadcast services, local digital systems and intermittent workshop sources. An ambient signal can hide an EUT emission or create a false failure impression. If the ambient is unstable, repeat the measurement, use time-domain observation where technically appropriate, or move to a more controlled facility.

Set an internal comparison threshold based on observed setup variation, not an arbitrary fraction of a decibel. Run the same baseline unit through repeated disconnect, reposition and power-cycle sequences. The resulting spread indicates what the setup can genuinely resolve. If repeatability is poor, tightening the pass margin in a spreadsheet will not improve the physics.

Control units are useful when several revisions are tested over weeks or months. Re-running a retained reference unit can reveal drift in the environment or measurement chain. A changed control trace should trigger an investigation before conclusions are drawn about the latest design.

Use logs that explain why a trace changed

A useful log links each plot to a defined EUT, operating mode and setup revision. Use unique identifiers rather than filenames such as “final_scan_2”. Include the engineer, date, instrument configuration, calibration status, environmental observations and deviations from the fixed setup.

Change one significant variable at a time where practical. If the PCB, firmware, cable and enclosure all change together, the resulting trace may show that the product improved, but not why. That weakens design learning and makes regression diagnosis slower.

Annotate known emission families. Clock harmonics, switch-mode power supply components and data-interface activity often have recognisable spacing or modulation. Tracking source hypotheses alongside frequency and amplitude helps engineers distinguish a shifted source from a genuinely removed one.

EMC Hire uses test equipment with calibration traceable through an appropriate ISO/IEC 17025 accredited calibration provider where calibration is relevant to the measurement. Suitable traceable calibration supports repeatability, confidence in recorded data and comparison between development work and later formal testing. It does not compensate for uncontrolled cable geometry or an unrecorded operating mode.

Typical scenario

Consider an illustrative development programme with three revisions of a networked industrial controller. The team has changed the DC-DC converter layout, added an enclosure bonding feature and updated Ethernet firmware. A repeatable EMC pre-scan must separate those effects rather than simply produce three unrelated plots.

The engineers first establish a conducted emissions setup on the relevant power port using a suitable LISN and a radiated pre-scan with fixed antenna geometry. They define high network traffic, maximum processor load and representative I/O activity as controlled operating modes. Cable positions are marked, the load simulator remains in one location, and ambient traces are recorded before each session.

Revision A becomes the baseline. Revision B is measured with the original firmware before the firmware update is introduced, allowing the hardware effect to be assessed separately. Revision C is then compared using both firmware builds where compatibility permits. Without that staged approach, a reduced spectral line might incorrectly be attributed to enclosure bonding when it was actually shifted by firmware timing.

Early investigation provides time to alter return paths, filtering or cable termination before tooling and documentation are frozen. EMC Hire can support this work through EMC pre-compliance testing, equipment selection, test facility access and on-site measurements. Where a formal programme is appropriate, the evidence can contribute to the technical file, EMC risk assessment, mitigation records and self-certification decisions. Pre-compliance results do not by themselves prove compliance.

When to Hire EMC Equipment

Hiring is often appropriate when a repeatable EMC pre-scan is needed for a defined design sprint rather than as a permanent laboratory function. It avoids committing capital to an analyser, LISN, probes, antennas or accessories that may not suit the next product family or frequency range.

Rental also covers irregular demand. Several project teams may need equipment during a short verification peak, followed by months with little use. Ownership during those gaps still brings storage, servicing, damage-control and calibration overheads. Hiring allows the measurement chain to be selected for the actual port, frequency range, detector requirement and expected signal level.

Selection still needs engineering judgement. An analyser with insufficient frequency coverage, a LISN unsuitable for the supply arrangement, or a pre-amplifier with inadequate linearity can undermine the comparison. EMC Hire's equipment selection guides and engineering team can help define a suitable chain for the intended pre-scan window.

For teams without a controlled space, booking a test facility or arranging on-site support may be more efficient than trying to force repeatability from a noisy workshop. On-site testing can preserve the real installation where moving the EUT would change cable lengths, bonding or operating conditions.

Common EMC Testing Mistakes to Avoid

Changing cable routing between revisions

Cable common-mode current is often the dominant radiated emissions mechanism. A route change can produce a larger difference than the PCB modification under investigation, leading to false confidence or an unnecessary redesign.

Comparing different detector or bandwidth settings

A peak trace cannot be compared numerically with a quasi-peak or average trace as though only the label changed. Likewise, changing resolution bandwidth alters captured noise power and the response to narrowband or pulsed signals. Record and lock settings by frequency segment.

Ignoring measurement-chain overload

A pre-amplifier may lower the displayed noise floor but reduce available headroom. Strong ambient or EUT signals can create compression and intermodulation products. Check levels with attenuation changes and, where needed, remove the pre-amplifier to confirm suspicious responses.

Using unrepresentative operating modes

Testing an idle EUT may miss emissions associated with data transfer, maximum power conversion or actuator activity. The resulting baseline remains repeatable but answers the wrong engineering question.

Failing to preserve raw data and setup records

A plot without trace data, cable photographs, equipment settings and revision identifiers cannot be reliably reproduced. It also provides weak evidence for later technical documentation or formal test preparation.

Frequently Asked Questions (FAQs)

How close should a pre-scan be to a formal compliance setup?

It should reproduce the coupling paths, EUT configuration and measurement settings that matter to the applicable requirement as closely as the development environment permits. Simplifications must be documented. Formal testing may require prescribed sites, distances, detectors and configurations that a bench setup cannot reproduce.

Can near-field probe results be used to predict a radiated emissions result?

Not reliably as a direct field-strength prediction. Near-field probes are effective for locating sources and comparing local changes, but probe coupling does not represent the complete enclosure, cable and far-field radiation behaviour. Use a fixture and fixed grid when comparing revisions.

How often should the baseline unit be re-tested?

Re-test it after a material setup change, instrument substitution, facility move or unexplained shift in ambient conditions. For long programmes, periodic control measurements also help expose gradual drift.

Should every emissions line be investigated?

Prioritise signals near the applicable limit, emissions that vary unexpectedly between revisions, and families linked to high-energy clocks or switching nodes. Also retain margin for measurement uncertainty, unit variation and differences between the pre-scan and formal site.

Can repeatable pre-scan data support CE or UKCA self-certification?

Calibrated engineering data may support the technical file, risk assessment and mitigation evidence where the applicable conformity route permits self-certification. The manufacturer remains responsible for identifying the relevant legislation, standards, test requirements and documentation. Testing alone does not complete every conformity obligation.

Where can I find more practical setup guidance?

EMC Hire's EMC test guides cover measurement methods and setup considerations. Always verify the latest active edition of the relevant CISPR, IEC, EN or product-specific publication, including frequency ranges, test levels, limits, configurations and contractual requirements.

Turn the process into defensible engineering evidence

A fixed setup, controlled operating modes, retained baselines and disciplined logs turn EMC pre-scanning into a design tool rather than a collection of attractive plots. The payoff is faster fault isolation, better comparison between revisions and fewer surprises when the product enters a formal test programme.

To discuss a repeatable EMC pre-scan, request an equipment hire quotation, arrange on-site testing, book space at the EMC Hire test facility or plan pre-compliance and formal compliance work, contact the engineering team on +44 (0)1462 817111 or email sales@emchire.co.uk.