EMC pre-compliance testing: a practical checklist before the test house

Formal electromagnetic compatibility testing often happens late, after the electronics, mechanics and cable set have converged. Discovering a problem at that point can trigger a board revision, enclosure change and another laboratory booking.

Pre-compliance testing moves useful measurements into development. It aims to find likely failures, compare design changes and build margin before the accredited or competent test programme.

It is not a substitute for the final test. A normal office or engineering laboratory does not reproduce a calibrated chamber, compliant site, defined antenna distance or every detector and bandwidth requirement. Results should therefore be treated as engineering evidence with known uncertainty.

1. Identify the standards and product configuration

Do not begin by waving a near-field probe over the board.

Identify the markets, product family and likely emissions and immunity standards. Establish which ports, cable lengths, operating modes and accessories are relevant. If the product contains a radio, determine how radio and EMC requirements interact.

Create a configuration list covering:

• hardware and firmware revision;
• power supply;
• enclosure and shields;
• connected cables and peripherals;
• maximum data traffic and processing load;
• motors, converters, displays or radios active during the test;
• intended grounding or earthing arrangement.

The worst emissions mode may not be the normal user mode. Firmware may need a repeatable test routine that exercises clocks, interfaces and loads continuously.

2. Define what the bench can and cannot prove

A pre-compliance setup may provide strong correlation for conducted emissions and useful comparative evidence for radiated emissions. It is usually less able to reproduce absolute chamber measurements without careful calibration and a controlled environment.

Write down the objective of each test:

• locate a source;
• compare two layouts or filters;
• measure margin to an approximate limit;
• reproduce a laboratory failure;
• check that a fix remains effective across operating modes.

This prevents an approximate measurement being presented later as formal compliance.

3. Check ambient noise first

Measure the environment with the equipment under test switched off. Broadcast transmitters, mobile phones, lighting, nearby power supplies and computers can all appear in the trace.

Save the ambient result, then repeat it at intervals. A signal that does not change when the product is disconnected is unlikely to be fixed by redesigning the PCB.

For radiated work, a screened room or suitable cell improves repeatability. Where that is not available, use close-range comparative measurements and be cautious about absolute field strength.

4. Set up conducted emissions consistently

For mains or DC power-port measurements, a line impedance stabilisation network, or LISN, presents a defined impedance and provides a measurement port. The correct network and safety arrangement depend on the port and applicable method.

Rohde & Schwarz identifies LISNs, current or voltage probes and suitable analysers as core pre-compliance tools. The LISN must be installed and earthed correctly, with the required ground reference and cable arrangement.

Use appropriate attenuation and protect the analyser input. Record resolution bandwidth, detector, frequency range and any preamplifier or transient limiter.

Run line and neutral or positive and negative configurations where required. Keep the equipment placement and cable routing consistent between comparisons.

5. Use near-field probes to locate sources

Near-field probes are excellent debugging tools. Magnetic-field probes can identify high-current loops and switching nodes. Electric-field probes can respond strongly to high-voltage nodes and structures.

They do not directly predict the far-field compliance result. Probe response changes with distance, angle and position.

Create a repeatable scan method:

• mark a fixed probe height;
• keep orientation consistent;
• use a defined grid or photograph;
• record analyser settings;
• scan the PCB, connectors, seams and cables;
• repeat the same locations after a change.

A strong local field is a clue, not automatically the dominant far-field radiator. The cable or enclosure can provide the effective antenna.

6. Investigate common-mode cable current

Cables are frequent radiators because common-mode current can turn them into efficient antennas.

Use a current probe where appropriate to compare cable current across frequency. Move or reconfigure the cable in a controlled way and observe which peaks change. Check shields, connector termination, chassis bonding and the return path for common-mode current.

Filtering a signal conductor while leaving the return path uncontrolled may not solve the problem. Cable shields should be treated as part of the enclosure and grounding design, not as an afterthought in the harness.

7. Search by harmonic pattern

Regularly spaced peaks often point towards a clock, switching frequency or data pattern. Use the spacing to identify likely sources, then change one operating parameter at a time.

Temporarily alter a clock, PWM frequency, converter mode or data rate where safe. If the spectral family moves with it, the source has been narrowed.

Do not stop at the source. Identify the coupling path and radiator. A processor clock may appear on a cable because of reference-plane disruption or connector return-path design.

8. Test fixes as controlled experiments

Ferrites, copper tape, absorbers and temporary capacitors are useful diagnostic tools. They should answer a question rather than become an undocumented production fix.

For each experiment, record:

• the change made;
• the predicted mechanism;
• before-and-after traces;
• side effects on function, temperature and signal quality;
• whether the change can be manufactured consistently.

A capacitor that suppresses a peak may also overload a driver or degrade signal integrity. A slower edge may reduce emissions but increase switching loss. The production solution needs a full engineering check.

9. Include immunity pre-checks carefully

Immunity testing can involve electrostatic discharge, electrical fast transients, surge, conducted RF and radiated fields. Some tests involve hazardous voltages and specialised coupling networks.

Use competent personnel and appropriate equipment. Do not improvise high-energy surge testing.

Lower-risk development checks can still be planned around reset recovery, brownouts, cable disturbances and controlled ESD methods, provided the team understands the safety requirements and limitations.

Define pass criteria before the test. A product that recovers after a reset may be acceptable in one standard or performance class and unacceptable in another.

10. Preserve margin, not merely a narrow pass

Bench correlation is imperfect and production units vary. A trace sitting just below an approximate limit offers little confidence.

Set an internal margin that reflects measurement uncertainty and the difference between the pre-compliance setup and formal site. The appropriate figure depends on the method and correlation history.

Build correlation over time by comparing pre-compliance results with final laboratory reports. The organisation can then identify which bench measurements are predictive and which are only diagnostic.

What to take to the test house

Prepare the intended production configuration, support equipment, cables, firmware and operating instructions. Include a clear method for exercising worst-case modes.

Take design information that helps diagnosis if a failure occurs: schematics, PCB plots, clock list, converter frequencies and photographs. Bring safe temporary components and tools only with the laboratory's agreement.

Record every configuration change during testing. A pass obtained after moving a cable or applying tape is not a production result until that arrangement is defined and implemented.

The final pre-compliance checklist

• Applicable standards and ports identified.
• Worst-case operating modes defined and repeatable.
• Hardware, firmware, enclosure and cable configuration frozen.
• Ambient emissions recorded.
• Measurement path protected and settings documented.
• Conducted emissions checked with the correct network.
• PCB, seams, connectors and cables scanned consistently.
• Common-mode cable current investigated.
• Diagnostic fixes tested for functional side effects.
• Adequate margin established against an appropriate reference.
• Formal test plan and pass criteria agreed.

The purpose of pre-compliance work is not to create a cheaper imitation of an accredited laboratory. It is to make the product less surprising when it reaches one.

Technical sources

• Rohde & Schwarz, EMI pre-compliance testing.
• Rohde & Schwarz, EMI debugging and analysis.
• Rohde & Schwarz, EMC accessories guidance covering LISNs, probes and antennas.