Skip to content

How to create an immunity retest plan

How to create an immunity retest plan
11 min read

Repeating a failed immunity test without controlling the setup may reproduce the symptom, but it rarely proves that the corrective action works.

A useful immunity retest plan must separate three questions: can the original failure be reproduced, does the proposed change remove the failure mechanism, and does the product remain acceptable across the full applicable test conditions?

Start with the failure evidence, not the proposed fix

Before booking equipment or laboratory time, reconstruct the original event. Record the test method, applied level, frequency or transient condition, modulation, dwell time, coupling path, EUT operating state and observed degradation. The relevant performance criterion from the applicable product or product-family standard should also be identified.

Descriptions such as “unit reset during RF test” are too weak for retest planning. A reset at one conducted RF frequency may indicate demodulation in an analogue input, disturbance of a supply rail, common-mode current on an external cable or corruption of a communications interface. Each mechanism points towards different diagnostics and corrective actions.

Gather the original test report, laboratory notes, photographs, software logs, oscilloscope captures and configuration records. Check whether the failure threshold was explored. If testing stopped immediately at the specified level, the team may know only that the EUT failed, not whether its immunity margin was slightly inadequate or substantially deficient.

Do not infer more than the evidence supports. Suspected mechanisms should be recorded as hypotheses until measurements confirm them.

Define the purpose and acceptance criteria

An immunity retest plan should state what each test stage is intended to prove. Otherwise, teams tend to rerun the complete campaign after every modification, consuming time while learning very little.

A practical retest normally has three layers:

  • Reproduction testing confirms that the controlled setup can recreate the original failure.
  • Diagnostic verification examines whether the corrective action changes the susceptibility threshold or removes the identified coupling path.
  • Regression testing checks the applicable test range, ports, polarities, orientations and operating modes for new or displaced weaknesses.

Acceptance criteria must come from the applicable product standard, contractual test plan or justified engineering specification. IEC 61000-4-x documents are basic immunity test methods. They do not, by themselves, establish the complete set of product requirements. The applicable product or product-family standard may define test levels, port applicability, operating modes and performance criteria.

For every retest, verify the latest active editions, test levels, frequency ranges, configurations, methods, performance criteria and documentation requirements. Customer-specific plans may add conditions not present in a generic product standard.

Create a setup lock before changing the EUT

A setup lock is the controlled description of the test arrangement used to compare the failed and corrected configurations. Without it, a cable moved by 200 mm, a different auxiliary supply or a changed earth bond can alter the coupling path enough to create false confidence.

The setup lock should capture:

  • EUT hardware revision, serial number, firmware, software settings and populated options.
  • Operating mode, data traffic, loads, duty cycle and monitored functions.
  • Cable type, length, screening, termination, routing, height and orientation.
  • Bonding method, ground-plane arrangement and earth connections.
  • Auxiliary equipment, support equipment and isolation arrangements.
  • Coupling device, injection point, calibration data and test equipment identity.
  • Photographs from repeatable viewpoints, supported by a dimensioned sketch.
  • Environmental conditions where they could influence EUT behaviour.

Record intentional deviations rather than quietly absorbing them into the setup. For conducted RF immunity under a method such as IEC 61000-4-6, changing the CDN type or replacing a CDN with an injection probe is not a minor substitution. The coupling impedance and current distribution differ. A BCI probe may be correct for an applicable automotive, military, aerospace or product-specific current-injection method, but it is not automatically interchangeable with a CDN.

The same discipline applies to transient testing. Capacitive coupling clamps, direct injection networks and ESD simulators represent different disturbance mechanisms. Using the wrong coupling device may exercise a different path from the one that produced the original failure.

Turn corrective actions into controlled variables

Changing a filter, enclosure bond, PCB layout, cable screen termination and firmware watchdog simultaneously makes attribution impossible. If the retest passes, the team cannot identify which change worked. If it fails, there is no clear next step.

Where practical, test one controlled change at a time. Assign each corrective action an identifier, engineering rationale and expected observable effect. For example, a common-mode choke added to an external interface should be linked to an expected reduction in injected cable current or disturbance at a sensitive circuit node. A firmware recovery change may improve functional behaviour without reducing RF coupling, so it should not be described as an electrical immunity improvement unless the evidence supports that conclusion.

Temporary modifications can be highly informative, but their construction must be recorded. A short copper bond applied during debugging may perform far better than the production bracket eventually proposed. Replacing it with a long wire adds inductance and can remove the high-frequency benefit that the trial appeared to demonstrate.

Choose verification measurements that expose the mechanism

Pass or fail is not enough during engineering investigation. Select monitoring that shows what happens immediately before the functional degradation.

Useful measurements might include supply-rail disturbance, reset-line activity, communications error counts, watchdog events, analogue output deviation or injected cable current. Monitoring equipment must not materially change the susceptibility. A long oscilloscope ground lead or unfiltered probe cable can create an additional RF path and produce a failure that disappears when the instrumentation is removed.

For swept conducted RF immunity, targeted testing around the original failure frequency is often efficient during debugging. Explore a controlled frequency window, repeat the relevant modulation and dwell conditions, and establish a repeatable threshold where the method permits. The formal regression stage should then cover the full applicable range and required conditions rather than only the known weak point.

ESD, electrical fast transient, surge and radiated RF failures need their own method-specific verification. A fix that resolves a conducted RF problem does not demonstrate resistance to transient or radiated phenomena.

Build the retest sequence around decision gates

A disciplined immunity retest plan prevents a full campaign being spent on an EUT that still fails at the first known condition.

  1. Configuration check: confirm the setup lock, EUT revision, operating mode, monitoring and equipment status.
  2. Baseline reproduction: use the failed EUT configuration, where safe and practical, to reproduce the original symptom.
  3. Corrective-action check: install the controlled modification and repeat the known failure condition.
  4. Margin exploration: assess the surrounding conditions or threshold where permitted by the method and test plan.
  5. Regression sweep: test the full applicable range, ports, orientations, polarities and operating states.
  6. Production-representative confirmation: replace temporary fixes with the intended implementation and repeat the relevant testing.

Define stop conditions. Unexpected overheating, unsafe operation, loss of monitoring or damage to protection components should trigger review rather than automatic continuation. Also state who can authorise deviations and whether a failed diagnostic step requires redesign, further investigation or escalation to formal testing.

Typical scenario

Consider an illustrative industrial controller that resets during conducted RF immunity testing on a screened communications cable. The failure appears within a narrow part of the sweep, but the first report records neither cable-screen termination details nor the controller traffic mode.

The engineering team first reconstructs the setup using photographs and laboratory notes. A setup lock is created, including CDN selection, cable routing, ground-plane arrangement, software load and communications traffic. Reproduction testing then confirms whether the reset remains repeatable.

Diagnostic work may compare a revised screen bond, interface filtering and a controlled firmware recovery change. These are tested separately. Current monitoring and local rail measurements can help distinguish reduced RF coupling from improved software recovery, provided the probes do not disturb the setup.

Once the preferred corrective actions are transferred into a production-representative build, the applicable frequency range, ports and operating modes are retested. This staged approach reduces the chance of paying for a formal campaign that merely repeats the original failure.

EMC Hire can support this work through EMC pre-compliance testing, equipment selection, test-facility access, on-site investigation and formal compliance testing where appropriate. Hiring the required generator, amplifier, coupling devices and monitoring equipment can also cover a defined debugging window without creating a long-term ownership burden.

When to Hire EMC Equipment

Hiring is particularly useful when immunity testing demand is irregular or tied to a short corrective-action programme. A complete system must be selected as a compatible chain. Frequency coverage, amplifier power, directional coupling, monitoring, modulation capability and coupling-device ratings all need to support the required method and level.

For a conducted RF project, review the available conducted immunity equipment and the associated conducted immunity system information. The appropriate CDN, clamp or injection probe depends on the specified method and EUT port. EMC Hire’s equipment selection guides can help define the wider system, although the current standard and project test plan remain controlling documents.

Rental avoids capital expenditure for equipment that may have little use after the retest. It also reduces storage, servicing and calibration overheads, supports project peaks and limits the risk of buying a system unsuitable for future programmes.

Where calibration affects the measurement or applied test level, EMC Hire uses test equipment with calibration traceable through an appropriate ISO/IEC 17025 accredited calibration provider. Suitable traceability supports repeatability, comparison between development and formal testing, and a stronger engineering evidence trail. UKAS information on accreditation provides further context on the role of accredited conformity assessment and calibration activities.

Common EMC Testing Mistakes to Avoid

Retesting only the original failure point

A narrow check can verify a local improvement, but it cannot reveal a shifted resonance or a new weakness on another port. Use targeted testing for diagnosis, followed by an appropriate regression programme.

Changing the setup along with the fix

Moving cables, changing auxiliary equipment or improving bonding while fitting the corrective action destroys the comparison. The pass may result from the altered setup rather than the product modification.

Using an unrepresentative operating mode

An idle communications interface or lightly loaded power stage may not exercise the vulnerable circuitry. A passing result under that condition can create false confidence and an unsuitable compliance evidence trail.

Substituting coupling devices without justification

A CDN and BCI probe impose disturbances through different arrangements. An unjustified substitution can change common-mode impedance, injected current and cable behaviour, making comparison with the original test invalid.

Failing to document temporary fixes

Conductive tape, ferrites and improvised bonds often move between trials. Without part details, position, orientation and photographs, the result cannot be reproduced or transferred reliably into production.

Confusing recovery with immunity

A watchdog may restore operation after disruption while the sensitive circuit remains affected. Whether that behaviour satisfies the required performance criterion depends on the applicable product standard and test plan.

Evidence for formal compliance and technical documentation

The final record should include the approved plan, equipment identities, calibration status, setup photographs, EUT configuration, raw observations, deviations, corrective-action revisions and signed engineering conclusions. This material can support the technical file, EMC risk assessment, mitigation evidence, Declaration of Conformity and self-certification process where the applicable route allows it.

Testing alone does not complete every conformity obligation. The manufacturer or responsible economic operator remains responsible for identifying applicable legislation, standards, conformity assessment routes and documentation requirements. Defence, automotive and aerospace programmes may also require final testing by an appropriately accredited laboratory under contractual or programme rules, although pre-compliance work remains useful for reducing redesign risk beforehand.

Frequently Asked Questions (FAQs)

Must the original failure be reproduced before testing the fix?

Preferably, yes. Reproduction confirms that the setup still exercises the relevant weakness. If reproduction is unsafe or impossible, document why and identify the resulting limitation in the evidence.

Should a retest cover the full immunity range?

Diagnostic testing can concentrate on the known failure region. Final verification should cover the applicable range, ports, operating modes and conditions needed by the product standard or agreed test plan.

How many corrective actions should be tested at once?

Use one controlled variable at a time where practical. Combined changes may be necessary for schedule or safety reasons, but they reduce the ability to attribute the result to a specific mechanism.

Can pre-compliance results prove conformity?

No. They provide engineering evidence, identify weaknesses and improve confidence before formal testing. Their role in self-certification depends on the applicable legislation, product requirements and adequacy of the documented evidence.

What makes a retest result defensible?

A controlled setup, suitable calibrated equipment, representative EUT operation, clear acceptance criteria, complete records and justified deviations. Repeatability matters as much as the final pass indication.

Plan the retest before booking the test window

EMC Hire’s engineering team can help review an immunity retest plan, select suitable equipment, arrange on-site testing, support pre-compliance investigation or discuss formal compliance testing. Test-facility space can also be booked for controlled debugging and verification.

For an equipment hire quotation or a discussion about the failed test, contact +44 (0)1462 817111 or email sales@emchire.co.uk. Providing the applicable standard, failed condition, EUT port details and available test records will make the initial technical review more productive.

Disclaimer: Content is for informational purposes only and does not constitute formal engineering or regulatory advice. Always verify testing procedures against current official standards (e.g., ISO, MIL-STD, DEF STAN). EMC Hire Limited accepts no liability for outcomes resulting from the use of this information.