How to interpret radiated susceptibility failures

A radiated susceptibility failure is rarely explained by frequency alone. The type of upset, recovery behaviour, antenna polarisation, EUT orientation and test threshold usually reveal far more about the coupling path.

What the failure is telling you

Radiated susceptibility testing exposes equipment to a controlled RF electromagnetic field while its safety, performance and interfaces are monitored. Defence programmes commonly reference methods within MIL-STD-461, while commercial product standards may call up IEC 61000-4-3 for radiated RF immunity. These methods are not interchangeable. Test levels, modulation, frequency ranges, dwell arrangements, field generation and acceptance criteria must come from the applicable contract, product standard and approved test plan.

The first task after a failure is classification, not redesign. A narrow-band communications error at one frequency suggests a different mechanism from a processor reset across a broad frequency range. Likewise, a transient analogue deviation that self-recovers when the RF field is removed should not be recorded in the same way as corrupted non-volatile memory or loss of a safety-related function.

Record the observed effect in engineering terms. Words such as “upset” and “unstable” are too vague on their own. State which function changed, when it changed, whether the effect tracked modulation, whether operation recovered automatically, and whether any data or configuration was lost.

Separate temporary degradation from destructive or persistent effects

A temporary display disturbance may be acceptable under one product-specific functional criterion and unacceptable under another. Defence acceptance criteria are often contractual and may define permitted degradation by operating mode, mission function or safety consequence. Do not assume that commercial immunity performance criteria apply to a military programme.

Useful failure categories include:

• Observable degradation only while the field is applied.
• Automatic recovery after the disturbance is removed.
• Recovery requiring operator intervention.
• Processor reset, watchdog action or power cycling.
• Latched fault requiring maintenance action.
• Loss or corruption of stored data.
• Permanent damage or degraded performance after exposure.

The recovery category affects both compliance interpretation and debug priority. A self-clearing telemetry error can indicate demodulation within an interface. A full reset points towards supply disturbance, reset-line injection, clock disruption, watchdog behaviour or firmware exception handling. Permanent damage requires the test to stop and the applied conditions, monitoring evidence and EUT state to be preserved before further exposure.

Map the failure against every controlled variable

A single notation such as “failed at 180 MHz” is not enough for a defensible investigation. At minimum, correlate the response with applied field level, frequency, modulation, dwell time, antenna polarisation, antenna position, EUT orientation, cable arrangement and operating mode. Where the method requires substitution or field calibration, retain the relevant forward-power and field-generation records.

Frequency can suggest the physical scale of the coupling structure, but avoid treating a cable length as a simple resonant antenna in isolation. Harness terminations, common-mode impedance, bonding, enclosure apertures, cable shields and nearby conductive structures all alter the response. A failure that moves when a harness is repositioned strongly implicates cable coupling. One that remains fixed despite controlled cable changes may be dominated by an enclosure aperture, PCB trace, internal interconnect or direct coupling into an unshielded circuit.

Polarisation is equally informative. A marked difference between horizontal and vertical fields can identify a preferred coupling geometry. If rotating the EUT or changing the exposed face causes a large threshold shift, inspect seams, displays, ventilation openings and connector panels on the sensitive face.

Find the threshold without overstressing the EUT

Once a repeatable failure has been observed, reduce the field and determine the approximate onset threshold using a controlled method agreed with the test authority. Repeatedly applying the maximum level tells you little about design margin and may conceal a nonlinear response.

Threshold information helps distinguish a marginal design from a gross susceptibility. It also allows modifications to be compared quantitatively. A mitigation that merely shifts the failing frequency by a few megahertz may have changed the resonance without improving immunity. A repeatable increase in failure threshold across the affected band is stronger evidence that the coupling path has been reduced.

Watch for hysteresis. Some faults remain latched after the field falls below the original onset level, while thermal protection, automatic gain control or firmware recovery can introduce time-dependent behaviour. Record upward and downward sweeps separately rather than assuming they are equivalent.

From symptom to likely coupling path

RF energy usually reaches the victim circuit through cables, apertures, conductive enclosure paths or direct field coupling. The observed symptom narrows the search.

Audio-frequency or low-rate analogue modulation appearing on a sensor channel suggests rectification at a semiconductor junction or protection device. Test whether the disturbance follows the applied modulation. If it does, investigate filtering, source impedance, cable common-mode current and nonlinear input structures before changing firmware.

A digital reset deserves a disciplined sequence. Monitor the regulated rails, reset line, clock, watchdog output and relevant fault flags with probes that do not materially alter the RF coupling path. Long oscilloscope ground leads can act as receiving antennas and create a new susceptibility mechanism. Fibre-optic monitoring, isolated instrumentation or properly configured high-bandwidth differential probing is often more representative.

Interface errors may result from common-mode current entering through an external harness and converting to differential voltage at an impedance imbalance. Check connector bonding, shield termination around the connector circumference, filter placement and the current return path. A filter mounted several centimetres from the enclosure entry can be ineffective because the intervening conductor radiates internally and bypasses the intended boundary.

Near-field probes can help identify sensitive internal regions, but they do not reproduce the calibrated far-field test. Use them as comparative diagnostic tools. A local injection that recreates the same functional signature is useful evidence, not proof that the compliance coupling mechanism has been duplicated.

Typical scenario

Consider an illustrative defence controller undergoing radiated susceptibility testing. During vertical polarisation, one communications channel reports intermittent errors over a limited band. At a higher field level, the processor resets. No response occurs in horizontal polarisation, and the fault disappears when the external harness is moved closer to the ground plane.

The team should first preserve the original configuration with photographs, cable measurements, software version, operating mode and monitoring logs. Repositioning the harness without recording its initial route destroys the most useful evidence.

Next, reproduce the communications error at a reduced field and establish its threshold. Controlled changes can then be introduced one at a time: harness position, connector shield termination, temporary common-mode suppression, filtered feedthrough configuration and EUT orientation. If the error threshold follows harness geometry, cable coupling becomes the leading hypothesis. Measuring common-mode cable current with a suitable current probe may provide additional comparative evidence, although that measurement is not a substitute for the specified radiated test.

The reset requires separate investigation. The communications disturbance may trigger an unhandled firmware condition, or the same RF current may be entering the power or reset circuitry. Monitoring supply rails and reset logic helps separate these mechanisms. Combining both symptoms under one generic “RF upset” label risks applying shielding changes when robust interface handling is also required.

Early work in an accessible pre-compliance setup allows the team to investigate without consuming a formal test slot for every design iteration. EMC Hire can support equipment selection, test facility access, pre-compliance engineering and on-site testing. Where appropriate, formal compliance testing can then generate evidence for the programme technical file, EMC risk assessment, mitigation records and contractual review. Pre-compliance data improves confidence but does not automatically demonstrate compliance.

Common EMC Testing Mistakes to Avoid

Changing several variables together

Adding ferrites, rerouting cables and changing firmware in one step may remove the failure, but it does not identify the mechanism. The resulting fix can fail when cable length or installation geometry changes. Alter one controlled variable at a time and retain threshold data for comparison.

Ignoring cable geometry

Harness height, length, branching and termination affect induced common-mode current. An undocumented cable change can shift a resonance or reduce coupling enough to create false confidence. Photograph cable routes and record dimensions at every meaningful test stage.

Treating radiated emissions as susceptibility data

A product can exhibit low radiated emissions and still be susceptible to an external field. Emissions measurements may identify resonant structures or shielding weaknesses, but they assess energy leaving the EUT rather than immunity to an applied field. EMC Hire's radiated emissions test system information describes a separate measurement discipline.

Using the wrong injection method

A BCI probe used for an applicable current-injection immunity method does not replace a radiated susceptibility antenna test. Likewise, a CDN is associated with conducted RF immunity methods such as IEC 61000-4-6, not radiated field exposure. The conducted immunity overview and conducted immunity system information explain those distinct coupling methods.

Monitoring the wrong operating mode

An idle processor or static interface may pass while the mission-representative mode fails. Exercise maximum data rates, sensor states, loads and control transitions where required by the test plan. Otherwise, the most sensitive circuitry may never be active during exposure.

Recording only pass or fail

A binary result omits onset threshold, recovery, polarisation, dwell and functional evidence. That weakens repeatability and leaves future engineers unable to distinguish a genuine design regression from a setup difference.

When to Hire EMC Equipment

Radiated susceptibility systems involve RF sources, amplifiers, antennas, directional couplers, power monitoring, field probes, control equipment and suitable test environments. The required frequency coverage and power depend on the test method, field level, antenna efficiency, chamber losses and test distance. Buying an amplifier because its headline output power appears adequate can be an expensive mistake if it cannot generate the required calibrated field across the relevant band.

Hiring is technically sensible when demand is short-term, programme-specific or irregular. It gives a team access to a suitable configuration for a defined debug window without committing capital to equipment that may not suit the next programme. Storage, servicing and calibration overheads also remain proportionate to actual use.

Rental can cover project peaks when internal equipment is already committed, or support on-site investigation where moving the EUT is impractical. EMC Hire can help select equipment based on frequency coverage, modulation, monitoring requirements and the intended test method. Equipment used for measurements is supplied with calibration traceable through an appropriate ISO/IEC 17025 accredited calibration provider where calibration is relevant. Suitable traceability supports repeatability, confidence in recorded data and comparison between development and formal testing.

For defence work, pre-compliance support can reduce late redesign risk, but final testing may need an appropriately accredited laboratory where the contract or programme requires it. Review the latest active standard edition, test levels, frequency ranges, equipment configuration, acceptance criteria and customer-specific test plan before fixing the setup. Further sector information is available through EMC Hire's defence EMC testing support.

Frequently Asked Questions (FAQs)

Does a reset automatically mean the power supply is susceptible?

No. RF disturbance may affect a supply rail, but it can also couple into reset circuitry, clocks, communications, watchdog logic or firmware error handling. Instrument the likely signals carefully and compare their timing with the reset event.

Why does the failure frequency move after adding a ferrite?

The ferrite changes common-mode impedance and therefore alters the cable resonance and current distribution. A shifted failure is not necessarily an improvement. Compare susceptibility thresholds across the full affected band.

Can near-field probing replace a repeat radiated test?

No. Near-field injection is useful for localisation and comparative debugging, but it does not reproduce the calibrated field, illumination geometry or coupling conditions of the specified radiated method.

Should we test only at the original failing frequency?

Use the original frequency for rapid comparisons, then investigate around it and repeat the required band after modification. A mitigation can move the response rather than remove it.

How should functional criteria be agreed?

Derive them from the applicable standard, contractual requirements, equipment functions and safety or mission consequences. Define permitted degradation and recovery before testing. The manufacturer and programme authority remain responsible for confirming the applicable requirements.

Can pre-compliance results support formal documentation?

Calibrated pre-compliance data can support engineering decisions, mitigation evidence, EMC risk assessments and the technical file. It does not automatically replace testing required by a contract, product standard or conformity assessment route.

Plan the next diagnostic step

A useful failure record connects the symptom to field level, frequency, polarisation, orientation, cable configuration, operating state and recovery behaviour. That evidence points towards the next controlled experiment and prevents repeated testing based on guesswork.

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