PCB vibration testing should start with a defined operating environment, a representative assembly, a mounting arrangement and measurable acceptance criteria. A shaker booking alone cannot establish whether an industrial controller will survive service. The useful result connects a specified mechanical exposure with recorded electrical behavior and a traceable hardware configuration.
For an OEM preparing a new assembly, this planning decision belongs before the test purchase order. Mechanical engineering defines mounting and excitation assumptions. Electronics engineering identifies functions that must remain stable. Quality agrees how failures, interruptions and deviations will be recorded. Procurement can then compare quotations against the same scope.
This guide covers assembled boards and their product-level interfaces. It does not prescribe a universal acceleration, frequency range or duration. Those values must come from the applicable product specification, measured environment or approved engineering justification.
Define what PCB vibration testing must demonstrate
Separate design qualification, manufacturing screening and failure investigation before choosing a method. Qualification asks whether a representative design meets defined requirements. Screening applies an approved process to production hardware. An investigation reproduces a suspected mechanism. The same equipment might support all three jobs, but specimen selection and disposition can differ.
A successful qualification run does not automatically create a production screening recipe. Screening can consume fatigue life or expose hardware to stresses outside normal operation. Its limits need a separate rationale. An investigation may intentionally progress toward failure, which would be inappropriate for a shipment acceptance test.
Write the decision in one sentence. For example, an engineering team might need to determine whether a revised mounting bracket changes intermittent connector behavior under a specified machine operating condition. That statement names a change, an observable response and a use condition. It gives the laboratory something more useful than a general request for a reliable board.
Connect this decision with the broader industrial application requirements . Include installation orientation, enclosure mass, cable restraint and expected operating modes. Where environmental data are missing, record the uncertainty and its owner. Approval of a convenient laboratory profile should not silently turn an unknown field environment into a verified requirement.
Select an excitation profile with a stated basis
Sinusoidal and random vibration describe different inputs. A sine test excites one frequency at a time or follows a defined sweep. A random test distributes energy across a frequency band. Shock introduces a transient event. The test plan must identify which physical condition each exposure represents.
IEC 60068-2-6 describes a sinusoidal vibration method. It supplies a method framework, while the relevant specification establishes the required severity and acceptance. Referencing the standard number without the selected conditions leaves the quotation and the result ambiguous.
For random vibration, a single overall acceleration value does not describe the complete profile. Frequency breakpoints, spectral density, exposure duration and axis definition also matter. Two profiles can have the same overall RMS value while placing different energy near an assembly resonance. Avoid comparing supplier quotations solely by their headline acceleration.
Check units and amplitude conventions explicitly. Specify displacement as peak or peak-to-peak. Acceleration may be RMS or peak, depending on the method. Ask the responsible engineer to confirm transition points, control tolerances and any approved notching or response limits. Treat any departure during execution as a recorded deviation.
Input
What the plan records
Decision it supports
Operating environment
Machine mode, location and installation assumptions
Whether the exposure represents intended use
Excitation
Method, units, frequency definition and duration
Whether quotations describe equivalent work
Axes and orientation
Product coordinates and mounting direction
Whether each required load path is covered
Limits and deviations
Control tolerance, response limits and approval route
When to continue, stop or repeat
Component mass, connectors and mounting points form part of the test configuration.
Freeze the sample configuration before testing
A bare PCB and a populated PCBA respond differently. Heavy components, connectors, heatsinks, shields and attached cables change loading. An enclosure or support bracket changes boundary conditions. PCB vibration testing must identify exactly which combination represents the product being assessed.
Record board revision, assembly variant, component substitutions, fastening hardware, firmware and any prior repair. A convenient prototype with a different capacitor package or temporary connector can answer a development question, but its limitations should appear in the report. Otherwise the team may extend a result to production hardware that was never represented.
Define sample selection before seeing outcomes. Consider build variation and the failure modes under investigation. A sample count cannot be chosen responsibly from a generic article. The quantity and confidence argument depend on the project, applicable requirements and consequences of failure. If only one unit is available, state what that evidence can establish.
Photograph each specimen and record its initial condition. Document damaged pads, cracked housings, repaired joints and loose hardware before installation. Agree whether the laboratory may tighten fasteners, reseat connectors or replace consumables. Such interventions can remove the very symptom the test is meant to investigate.
The PCB assembly review should carry the same controlled revision into purchasing, build instructions and specimen identification. Keep the tested configuration distinct from later engineering changes. A passing report should never be attached to an unspecified family of board variants.
Identify each specimen and preserve its configuration before and after exposure.
Review PCB vibration testing fixtures and cable restraint
The fixture determines how motion reaches the assembly. Mounting an installed product through different holes or clamping unsupported board edges can create a different load path. Review supports, bracket interfaces, fastener specifications and contact surfaces against the intended installation.
Fixture dynamics can also influence the result. A resonance in the fixture may amplify local response even when the controller follows its input target. Ask how the laboratory characterizes the setup and where it will place control and response sensors. The record should distinguish fixture behavior from specimen behavior.
NASA’s vibration-testing lessons discuss planning accelerometer positions around structural interfaces and predicted responses. These are useful planning concepts. Aerospace mission profiles should not be copied into an industrial PCBA requirement without an approved applicability assessment.
Cables deserve the same attention as fasteners. A cable hanging freely can apply loads absent from the finished product. Excessive restraint can suppress a real connector problem. Record routing, unsupported length, tie positions and external equipment interfaces. Check that instrumentation wiring will not short a circuit, disturb a sensitive signal or constrain a moving part.
Freeze photographs of the installed setup for each axis. A note saying “mounted normally” cannot explain a later disagreement about bracket orientation. Document remounting between axes, including any hardware replacement and the reason for it.
Monitor electrical behavior during PCB vibration testing
Functional checks before and after exposure can miss intermittent events. A connector might open briefly during excitation and recover when motion stops. A processor may reset and restart before the final inspection. Choose monitoring that can detect the failure mechanism named in the test objective.
Define powered and unpowered conditions separately. If the product must operate during exposure, record supply conditions, load, firmware mode and connected peripherals. Identify which signals or functions require continuous monitoring and which can be checked at agreed intervals. Measurement bandwidth and event detection settings must suit the expected disturbance.
Agree how an event will be timestamped and correlated with the mechanical exposure. Useful records may include resets, communication loss, unexpected current change, output excursions or continuity interruptions. Do not label every transient as a board failure before examining instrument behavior and the test setup.
The OEM should own functional limits or explicitly delegate their approval. A laboratory can execute a method accurately while testing against incomplete product criteria. “Still powers on” is insufficient when the product’s job depends on measurement accuracy, controlled outputs or persistent communication.
Use the quality evidence review to connect each acceptance requirement with a measurement and record. If a function cannot be monitored during exposure, record the resulting blind spot and the alternate evidence. Resolve that limitation before interpreting a passing test as proof of uninterrupted operation.
An inspection baseline makes later component movement or visible damage easier to investigate.
Set interruption and inspection rules in advance
PCB vibration testing can stop because of specimen behavior, an equipment limit, a loose fixture or a measurement fault. These situations require different responses. Record stop conditions and identify who may authorize continuation. An operator should not have to invent acceptance policy while an anomaly is developing.
Preserve evidence before disturbing a suspected failure. Save logs, note the active axis and exposure point, photograph the setup and identify affected functions. Power cycling or reseating a cable may help diagnosis, but it can erase information. Record every intervention and keep the original observation separate from later interpretation.
Define inspections between exposures where the approved method calls for them. Compare visible condition and selected electrical measurements with the initial baseline. A cosmetic mark and a cracked solder joint are different findings. Inspection criteria should name the features that matter, such as component movement, fastener displacement or damaged interconnects.
Any repeat needs a clear reason. Repeating until a unit passes introduces a selection bias and can consume further life. Decide whether the first run was invalid, whether the specimen changed or whether a design correction requires a new qualification scope. Keep earlier failed or interrupted runs in the record.
Observation
Immediate record
Disposition question
Communication interruption
Timestamp, operating mode and correlated channels
Product fault, connection fault or measurement artifact?
Unexpected mechanical response
Input and response traces with fixture condition
Did the setup or specimen depart from the approved basis?
Visible damage
Location, photographs and comparison with baseline
What evidence is needed before moving the hardware?
Stopped or repeated run
Reason, exposure completed and approving person
Which result remains valid and what must be repeated?
Select inspection methods for the suspected failure mechanism; an equipment photograph does not establish a test result.
Read the report before releasing a design
A useful report makes the exposure reproducible and its limits visible. Check specimen identity, fixture details, instrumentation, calibration references, input profiles, response data, operating states and inspection results. Review deviations alongside the final pass or fail statement.
Ask what the evidence excludes. Passing a board mounted in one enclosure does not prove performance in another mounting arrangement. A change in a heavy component or cable restraint can justify further review. The decision to retest should follow the changed load path and failure mechanism, rather than an automatic rule based only on a revision letter.
If an anomaly occurred, retain the connection between its observation, investigation and corrective action. The PCBA failure-analysis process can guide evidence preservation and fault isolation. A suspected solder defect should not become the recorded root cause merely because it was the first visible finding.
Keep test acceptance separate from production release. Test evidence may close one qualification requirement while drawings, supplier changes or manufacturing controls remain open. The release owner should identify which requirement the report closes and which unresolved conditions still restrict use.
Send a testable request with the manufacturing package
Before requesting PCB vibration testing, assemble the product requirement, controlled assembly identity, mechanical installation information and functional acceptance criteria. Add the purpose of the test, requested timing and any specimen-handling restrictions. Ask prospective providers to list assumptions and exclusions against that package.
Compare quotations by included engineering review, fixtures, instrumentation, axes, operating states, inspections and reporting. A lower price may describe less monitoring or a different sample scope. Resolve those differences before selecting a provider. Where a requirement is unresolved, request a separate planning activity instead of accepting an unspecified test.
For an assembly review, send the BOM, board and enclosure drawings, connector details and the approved test specification through the GNS contact page . Identify which responsibilities belong to the OEM, assembly provider and test laboratory so the requested evidence matches the intended release decision.
FAQ
Can an unpopulated board prove PCBA vibration performance?
No. Mounted components, connectors and mechanical interfaces affect response. A bare-board test supports only its stated specimen and objective; assess the assembled configuration separately.
Is an overall acceleration value enough for a test quote?
No. Specify the method, frequency definition, duration, axes, mounting and operating state. For random vibration, include the profile shape and approved control limits.
Must a board remain powered during the entire exposure?
That depends on the requirement. Define operating states and monitoring from the expected failure mechanism. Record any functions that cannot be observed during exposure.
Does a bracket change require another vibration test?
Review its effect on mounting stiffness, load paths and resonance. The responsible engineer should document whether existing evidence remains applicable or further testing is needed.