EMC troubleshooting with a portable near-field kit

A near-field probe can find the clock harmonic behind an emissions failure in minutes, but poor probe choice and inconsistent positioning can just as easily send a debugging team towards the wrong component.

A portable near-field kit is most useful when treated as a repeatable diagnostic system rather than an RF divining rod. The objective is not to recreate a compliant radiated emissions measurement on the bench. It is to identify sources, coupling paths and design changes that alter unwanted RF energy before laboratory time becomes expensive.

What a portable near-field kit can actually tell you

Near-field source hunting measures local electric or magnetic fields close to a PCB, cable, enclosure seam or component. At these short distances, the relationship between the measured signal and far-field emissions is not straightforward. Probe geometry, orientation, spacing and loading all affect the displayed amplitude.

The strength lies in comparison. A repeatable scan can show that a processor clock harmonic is concentrated around a particular power island, that common-mode current is developing on an external cable, or that fitting a common-mode choke has reduced energy at the frequencies under investigation.

It cannot, by itself, establish compliance with a radiated emissions limit expressed in dBµV/m. A small loop probe responds mainly to local magnetic field, while formal radiated measurements use a specified antenna, test distance, site and detector arrangement. Treating the loop-probe trace as equivalent to an antenna measurement creates false confidence.

Selecting probes for PCB debug and source hunting

Magnetic loops reveal current paths

Magnetic-field probes are commonly formed as shielded loops. They respond to changing current and are particularly effective around switching loops, DC-DC converters, clock traces, processor supply networks and connector return paths. Smaller loops provide better spatial discrimination but usually lower sensitivity. Larger loops detect weaker activity more readily, although the response covers a broader area and makes individual sources harder to separate.

Start with a larger loop to identify the active region, then move to a smaller loop for localisation. Beginning with the smallest probe can waste time because a weak source may disappear below the receiver noise floor. Conversely, stopping with a large loop may wrongly identify an entire processor area when the dominant mechanism is a single return-path discontinuity beneath one trace.

Loop orientation matters. Maximum response generally occurs when magnetic flux passes through the loop. Rotating the probe can therefore help infer current direction. If orientation changes are not controlled, a 10 dB change may be caused by probe geometry rather than the PCB modification being assessed.

Electric-field probes identify high-impedance nodes

Electric-field probes respond primarily to local voltage fields. They are useful around oscillator pins, high-impedance switching nodes, display interfaces, unterminated tracks and enclosure apertures driven by internal conductors. A strong electric-field indication does not automatically mean that the node is the dominant radiated source. Radiation also requires an effective structure through which energy can couple.

For example, a clock node may produce a strong local indication but have a compact return path. A lower-level signal reaching a long I/O cable can become the larger far-field problem because the cable acts as an efficient common-mode radiator. Combining electric and magnetic field probes helps distinguish source activity from the current paths that can carry it elsewhere.

Probe bandwidth must cover the mechanism

Check the characterised frequency range and transfer information in the probe manufacturer's documentation. A probe being physically connected to a 3 GHz spectrum analyser does not make it useful to 3 GHz. Probe response, amplifier bandwidth, cable loss and connector performance must all support the frequencies being examined.

Equally, the highest clock frequency is not the upper limit of interest. Fast digital edge rates generate harmonics well above the fundamental. The applicable product or product-family standard may require radiated emissions investigation above 1 GHz, depending on product scope and internal frequencies. Confirm the latest active standard, limits, frequency range and measurement method rather than relying on a generic rule.

Receiver, preamplifier and cable selection

A portable kit normally needs more than probes. The measurement chain may include a spectrum analyser or EMI receiver, a preamplifier, coaxial cable, adaptors, an attenuator and a probe positioning aid.

Sensitivity must be balanced against overload margin. A broadband preamplifier can expose low-level harmonics, but a strong nearby clock or radio transmitter may compress the amplifier or analyser front end. Compression produces misleading amplitude comparisons and can generate intermodulation products that do not exist on the PCB. Check signals first without gain, then add a suitable amplifier if the noise floor prevents useful observation.

Use low-loss, well-screened RF cables with connectors suited to the measurement bandwidth. A damaged connector or poorly screened lead can respond directly to the EUT field, so moving the cable changes the plot even when the probe remains stationary. Suitable RF and microwave cables and connectors reduce this uncertainty, but the lead should still be routed consistently and kept away from the strongest sources.

For comparative PCB debug, a spectrum analyser's peak detector and maximum-hold function are often useful for finding intermittent activity. Formal CISPR emissions work may require peak, quasi-peak and average detection with resolution bandwidths specified for the relevant frequency range. Those settings should not be copied blindly into near-field debugging, nor should a development scan be presented as a formal emissions result.

Building a repeatable bench method

Repeatability turns a promising trace into useful engineering evidence. Record the analyser centre frequency or span, resolution bandwidth, detector, preamplifier state, attenuation, probe type, orientation and distance. Save the reference trace before changing the design.

A simple non-conductive spacer can hold the probe at a consistent height. Without it, moving from 2 mm to 5 mm above a track may produce a larger amplitude change than replacing the component under investigation. Probe fixtures or marked scan grids are worthwhile when several engineers must reproduce the work.

Exercise representative operating modes. A prototype sitting at an idle menu may not activate its display traffic, motor drive, Ethernet link or highest processor load. Triggering only one mode can hide the source that appears during formal testing. Where operation is cyclic, use maximum hold cautiously and correlate the trace with known firmware events.

Keep configuration records, photographs and firmware identifiers. Document attached cables, cable lengths, grounding, loads, peripherals and power source. These details determine the return paths and cable coupling behaviour. An undocumented cable movement can make a shielding change appear successful when the real difference is altered common-mode impedance.

Following energy from source to coupling path

Efficient source hunting works from frequency correlation. First identify frequencies of concern from a chamber result, conducted emissions scan or bench antenna observation. Then inspect the PCB using a broad-area magnetic probe and progressively smaller probes. Finally, examine connectors, seams and cables at those same frequencies.

If a harmonic is strong at a processor, weaker along the track and strong again at an external connector, investigate the return path between those points. A split reference plane, badly placed stitching capacitor or connector shield bonded through a long inductive path may be converting differential energy into common-mode current.

Cable coupling deserves separate attention. Clamp-on RF current probes can measure common-mode cable current, although they require appropriate characterisation and should not be confused with small near-field loops. A near-field probe can still compare activity along a cable, but its response depends heavily on conductor position and probe orientation.

Testing candidate fixes close to the source usually gives clearer diagnosis. Ferrite sleeves fitted randomly to every external lead may reduce one symptom without revealing the conversion mechanism. When the ferrite is later removed for cost or mechanical reasons, the failure returns.

Typical scenario

Consider an illustrative prototype controller that shows narrowband radiated emissions around harmonics of a high-speed digital interface. The formal test antenna identifies the frequencies, but not the responsible track or coupling path.

On the bench, the engineering team uses a portable near-field kit comprising magnetic and electric probes, a spectrum analyser, suitable RF cable and an optional preamplifier. A larger magnetic loop finds activity around the processor and interface connector. A smaller loop shows concentrated current near a break in the reference-plane return path. Scanning the external cable at the same frequencies suggests that energy is reaching a structure capable of common-mode radiation.

The team must decide whether to alter routing, improve return continuity, change edge-rate control, revise connector bonding or add filtering. Choosing only an electric-field probe could emphasise the driven signal pin while missing the problematic current loop. Using excessive preamplifier gain could also make small design changes meaningless because the measurement chain is already compressed.

Early investigation allows several PCB modifications to be compared before another formal test slot. It does not prove that the revised unit complies, but it reduces the chance of entering the next programme with the original mechanism intact.

EMC Hire can support this work through near-field probe hire, suitable receiving equipment, test setup guidance, pre-compliance investigation, on-site testing and access to EMC test facilities. Where appropriate, the revised design can then proceed into formal compliance testing to generate evidence for the technical file, Declaration of Conformity or other project documentation.

When to Hire EMC Equipment

A portable near-field kit is often needed intensively for one prototype phase and then not again for months. Hiring avoids tying up capital in probes, analysers and accessories that may not suit the bandwidth or sensitivity needs of the next programme.

Rental is also useful when an internal analyser is available but the right probes, low-noise amplifier or RF cables are not. Accessing a matched set for a defined test window reduces the risk of assembling a measurement chain with an unknown weak point.

Short-term hire can cover project peaks, parallel debugging activity or an unexpected formal-test failure. It also avoids long-term storage, servicing and calibration overheads. Buying under time pressure carries another risk: the selected probe set may be optimised for fine PCB localisation but too insensitive for enclosure work, or may lack the characterised bandwidth needed for future products.

Where measurement traceability matters, EMC Hire uses test equipment with calibration traceable through an appropriate ISO/IEC 17025 accredited calibration provider. Suitable traceable calibration supports repeatability, confidence in recorded data and better comparison between development and formal testing. Calibration does not remove uncertainties caused by hand positioning or an uncontrolled EUT configuration, so the method still needs discipline.

Common EMC Testing Mistakes to Avoid

Treating probe amplitude as a compliance result

A near-field display is not directly equivalent to a specified far-field measurement in dBµV/m. Using it to declare a margin against a radiated limit creates an unsuitable evidence trail and may conceal an efficient cable or enclosure radiator.

Changing probe position between comparisons

Small changes in height, angle and lateral position can dominate the result. Without a fixture, spacer or marked grid, apparent improvement may only be a positioning error.

Ignoring the measurement cable

An RF lead draped across the PCB can disturb the field or pick up energy directly. If cable routing changes after every component modification, the traces are not comparable.

Overloading the analyser or preamplifier

Too much gain can cause compression and internally generated products. A clean-looking harmonic may then be an artefact of the test chain rather than an EUT emission.

Scanning an unrepresentative operating mode

Idle firmware may disable the interface responsible for the laboratory failure. Record loads, data rates, peripherals and duty cycles so the source is active during each comparison.

Fixing the hotspot but not the coupling path

Reducing the field immediately above an IC may not reduce radiation if common-mode current still reaches a cable. Recheck connectors, enclosure seams and external leads after every source-level change.

Frequently Asked Questions (FAQs)

Should I choose an electric or magnetic near-field probe?

For general PCB debug, use both. Magnetic loops are usually more informative for current loops and return-path problems. Electric probes help locate high-voltage or high-impedance nodes. Comparing their responses gives more information than either probe alone.

Can a portable near-field kit replace chamber testing?

No. It is a diagnostic and comparative tool. Formal radiated emissions testing requires the antenna, site, distance, EUT arrangement, detectors and limits defined by the applicable standard or test plan.

How small should the magnetic loop be?

Use the largest loop that provides enough localisation for the current task, then reduce loop size as the suspect area narrows. Very small loops improve spatial resolution but may require more receiver sensitivity and stricter positioning.

Do I need a calibrated spectrum analyser?

Calibration is strongly preferable when measurements will be recorded, compared over time or used to support engineering decisions. Relative debugging can still be undermined by front-end compression, cable loss, configuration changes or poor probe repeatability, none of which calibration alone corrects.

Can near-field results support CE or UKCA self-certification?

They can support design investigation and mitigation records, but do not automatically demonstrate conformity. The manufacturer or responsible economic operator must confirm the applicable legislation, standards, conformity assessment route and documentation requirements. Formal compliance data may be needed to create a defensible technical file.

What should be checked before hiring a kit?

Confirm the frequencies of interest, required probe types, analyser bandwidth, sensitivity, overload environment, connector types and whether a preamplifier is appropriate. Also identify whether the task is fine PCB localisation, enclosure scanning or cable coupling investigation. EMC Hire's range of antennas and near-field probes covers different measurement roles, so selection should begin with the suspected mechanism rather than connector compatibility alone.

Planning the next investigation

A well-selected portable near-field kit shortens the route from an emissions frequency to a physical mechanism. The useful result is not merely a lower trace. It is a repeatable explanation of where the energy originates, how it reaches a radiating structure and why the proposed change interrupts that path.

For help selecting equipment, arranging on-site troubleshooting, booking pre-compliance or formal compliance testing, or reserving space at the EMC Hire test facility, speak with the EMC Hire engineering team on +44 (0)1462 817111 or email sales@emchire.co.uk. The discussion can begin with your frequencies of concern, PCB architecture, available test evidence and intended test window.