A partial first article is justified only when retained evidence still represents the unchanged configuration and new verification covers every affected product and process question.
A partial PCBA first article can be justified when a valid earlier baseline exists, the effect of a change is understood, and the applicable requirements permit retaining the unaffected evidence. The reasoning must explain both what needs new verification and why the remaining evidence still represents the current product and route.
Calling a change small does not supply that reasoning. A single component substitution can affect several characteristics, while a larger administrative change may leave a particular result applicable. The important boundary is the effect on the evidence, not the number of edited lines in a document.
This guide presents an engineering framework for a partial first article rationale. It does not replace a contractual first article requirement or reproduce a licensed standard. It also does not describe a GNSEMS approval record. Use the actual project requirements to decide the verification and approval scope.
Decide Whether Partial First Article Rules Apply
Before selecting a partial approach, identify the governing customer, purchase-order and product requirements. Some projects use an internally agreed first article process. Others invoke a specific standard or customer form. The applicable requirement determines what the record must contain and who can approve the approach.
The IAQG 9102 first article standard page describes a formal standard developed for aviation, space and defense organizations. Its existence does not make that standard automatically applicable to every commercial PCBA project.
IAQG’s published 9102 revision-change presentation addresses partial first article scope in relation to affected or no longer represented characteristics. This supports the importance of an impact-based rationale; consult the applicable standard and customer requirements for the actual obligations.
For the broader sequence of preparation, inspection and review, see the SMT first article inspection guide . Here the narrower question is whether earlier evidence can still support the changed build.
Verify the earlier baseline before reusing its results
Locate the previous first article record and establish what it actually represents. Identify the assembly, revision, variant, relevant route and sample identities. Check its review status and any unresolved findings. A report that was prepared but never accepted may not provide the baseline the new rationale assumes.
Review the evidence references rather than relying on the report’s cover page. A result may belong to a different configuration, an earlier fixture or an article that received special processing. Those conditions matter when deciding whether the result remains applicable.
Record any gaps in the earlier package. Missing result details, unclear program identities and incomplete sample history limit the ability to justify retention. Do not fill those gaps with assumptions simply because repeating work would delay the order.
If the earlier baseline cannot be established, reconsider whether a partial approach is appropriate. The team may need additional investigation or broader verification. The decision should acknowledge the evidence limitation instead of treating the most recent report date as proof of a valid baseline.
Describe the change as an effective configuration difference
Identify the old and new conditions clearly. Include the change reference, effective point and affected variants or orders. A marked-up drawing can help, but the analysis should also consider changes in material, manufacturing route, software, equipment or verification conditions where relevant.
Separate intended changes from incidental differences. A component change might be the approved design action, while a different feeder setup or test program update occurs during implementation. The first article rationale should address the actual proposed build, not only the headline change.
Connect the comparison to PCBA engineering change control . The same effective baseline should appear in the change record, production inputs and first article plan. Conflicting revision references make it difficult to determine which evidence belongs to the new condition.
State which work has already occurred under the old condition. A partial first article plan for the new route does not automatically settle the disposition of existing work in process. Keep those population decisions linked but distinct.
Trace direct and indirect effects on characteristics
The affected feature list should include relevant indirect effects. Direct effects are the obvious changed features: a different component, footprint, location, connection or configured behavior. List the requirements associated with those features, including identity, fit and function as applicable. Avoid reducing the scope to a visual check of the edited item.
Indirect effects arise through connected conditions. A changed package can affect assembly access or inspection visibility. A revised test fixture can alter how a result is obtained. A firmware change can modify behavior while leaving the physical assembly unchanged.
These are questions for the responsible technical owners, not automatic conclusions that every nearby feature has changed. Ask what mechanisms could invalidate the earlier evidence and what records support excluding them. A bounded analysis is stronger than either ignoring indirect effects or declaring every characteristic affected without explanation.
Use the PCBA validation checklist to identify relevant verification interfaces. The first article scope may need to reference other validation evidence without pretending that first article inspection alone replaces the entire validation program.
Classify Evidence for a Partial First Article
Every excluded check needs a basis that a reviewer can follow. Use a small number of explicit outcomes. Retain means the earlier evidence remains applicable with a stated reason. Repeat means new verification is required. Not applicable means the characteristic genuinely does not apply to the new configuration, with its basis recorded. Unresolved means the team has not yet established the right outcome.
Evidence disposition in a partial PCBA first article review
Evidence outcome
Required explanation
Weak shortcut to avoid
Retain
Why the old result still represents the current characteristic and conditions
No drawing change, so all evidence is valid
Repeat
Which effect invalidates or leaves a gap in the old evidence
Repeat only the visibly changed component
Not applicable
Why the requirement is outside the new configuration
Leave the cell blank
Unresolved
What information or technical decision is still needed
Mark complete to avoid delaying the review
The rationale is strongest when a reviewer can follow each retained result back to its source and forward to the current requirement. A broad statement that “all other features are unchanged” may be a conclusion, but it is not the supporting comparison.
Choose verification methods that answer the changed question
For every repeated check, identify the requirement, method, sample condition and acceptance criteria. The method must be capable of evaluating the concern. Repeating an inspection that could not detect the relevant difference does not strengthen the rationale.
For example, a marking check can help establish component identity but may not demonstrate required functional behavior. A functional test can demonstrate selected behavior under its conditions but may not establish every material or assembly requirement. Use complementary evidence where the question requires it.
Do not broaden the claim beyond the actual verification. A passing first article result is evidence about the represented product and conditions. It should not be described as proof of long-term process capability or lifetime reliability unless the separate evidence and analysis support those conclusions.
Identify who reviews the method and result. The person completing a measurement may not own the design requirement or the decision to retain earlier evidence. Keep those roles clear in the review package.
Apply the rationale to variants and shared processes
A partial first article should preserve the earlier result as a traceable reference, not silently relabel it as evidence for a changed configuration.
A change can affect only part of a product family, but that boundary needs evidence. Map the variants that use the changed component, firmware, fixture or route. Shared resources can extend the relevant scope beyond the order where the change was first raised.
Conversely, a product family name does not prove every variant is affected in the same way. Some may not contain the changed feature. Others may use a different configuration that needs a separate analysis. Record the inclusion and exclusion logic instead of attaching one unexplained family-level status.
The high-mix PCBA production guide provides useful context for managing those differences. In the partial first article record, use explicit variant identities and evidence references rather than general descriptions of manufacturing flexibility.
If the same inspection or programming resource serves several products, assess the effective implementation carefully. A new version intended for one variant should not silently alter the verification basis for another. Keep deployment and applicability aligned.
Use a worked change example to challenge the proposed scope
Consider a hypothetical connector replacement on a controller assembly. The engineering change names a new approved part. The team initially proposes checking only the connector marking because the PCB outline has not changed. That proposal is too narrow unless a technical comparison supports it.
The reviewer examines the affected interface, orientation, fit and applicable assembly requirements. The review also asks whether the change affects insertion, soldering, inspection access or the test connection. These are candidate effects to assess, not a statement that all connector changes create the same problems.
Some previous evidence may remain applicable. For instance, an unrelated characteristic with unchanged relevant design and manufacturing conditions may have a defensible retention rationale. The record should explain that basis and link the original result rather than repeating a measurement solely to make the report longer.
Meanwhile, affected characteristics receive suitable new verification. If the new connector requires a changed test adapter, the review includes the evidence needed to establish the relevant test condition. A result obtained through the new adapter needs a known relationship to the requirement it evaluates.
The final plan is partial only because the impact boundary is justified. It is not partial because the supplier or customer prefers fewer checks. If the comparison leaves important effects unresolved, the team should resolve them or select a broader verification scope.
Recognize when the evidence does not support a partial approach
Reject a partial first article when the change affects an unverified characteristic, invalidates the earlier route, or removes the evidence needed to reconstruct the decision.
The contract and evidence determine whether a partial approach is appropriate. A partial approach becomes difficult to defend when the earlier baseline is unknown, the actual change is not bounded, or the affected characteristics cannot be identified reliably. The appropriate next step can be investigation, additional evidence or broader first article verification, depending on the project requirements.
Signals that the proposed partial scope needs reconsideration
Condition
Why the rationale is weak
Decision to assess
Previous record lacks configuration identity
The retained evidence has no reliable baseline
Recover the baseline or plan broader verification
Several undocumented changes occurred
The actual difference is not established
Reconstruct the effective configuration before scoping
Changed route has uncertain effects
Unchanged drawings do not establish unchanged evidence
Assess process effects and expand checks as needed
Customer requirement does not permit the proposal
The plan lacks contractual acceptance
Follow the applicable requirement or obtain authorized agreement
Important evidence remains unresolved
The release would rely on assumptions
Keep the gap visible and complete the needed decision
Broader verification is not automatically better if it repeats unsuitable checks while leaving the core question unanswered. Choose scope and methods together. The objective is adequate, traceable evidence for the current build, not a predetermined amount of inspection paperwork.
Keep planned retention separate from accepted results
At planning stage, the matrix describes a proposed evidence strategy. Mark it accordingly. A check listed for repetition is not complete merely because its method has been selected. Similarly, a proposed retention rationale may still need review by the responsible technical authority. Keeping those statuses explicit prevents a draft plan from being used as the final release package when production moves quickly.
When results arrive, compare them with the planned conditions as well as the acceptance criteria. Confirm the sample identity, configuration, method and relevant equipment or program references. A numerical pass obtained under a different condition may not close the planned check. Record the discrepancy and decide whether the result remains useful, needs an applicability review or must be replaced with suitable evidence.
Account for cumulative changes
Review the total difference from the baseline you intend to retain. Several individually limited changes can accumulate into a configuration that the original evidence no longer represents. If earlier partial records exist, establish how they connect to the original baseline and to each other. Do not compare only the latest change while overlooking a previous alteration that affects the same characteristic or verification method.
A practical record can identify the controlling result for each characteristic. Some results may come from the original first article and others from later approved verification. The record should make that combination understandable without implying that all results came from one physical article. This is especially important when configuration history spans multiple revisions and different sample populations.
Close the rationale against the actual build
Before approval, compare the as-built configuration with the planned new baseline. Check whether production introduced an approved substitute, a changed route or another relevant difference during execution. An otherwise sound impact analysis can become incomplete if it still describes the pre-build plan rather than the condition actually inspected. Resolve the difference before extending the result to future work.
Record the final outcome for every unresolved item. If the decision limits the release to a defined configuration or route, carry that limit into the operational instruction. If additional verification remains necessary, keep its owner and affected population visible. Closing the document should reflect completion of the required evidence and review, not simply completion of the scheduled meeting.
What Should the Partial First Article Decision Contain?
The record should show why previous evidence still represents unchanged features. The review package should contain the old baseline, new baseline, change reference, impact comparison, evidence disposition and planned or completed repeated checks. Include the approval scope and any unresolved items. Keep the document concise enough that the decision can be understood without reconstructing the entire project.
Use references that remain accessible after release. A local filename without a controlled location can break the evidence chain when a laptop or shared folder changes. The project should decide how to preserve the required records and their revision context.
Before the review, prepare the existing first article record, proposed change and affected variant list. Discuss a partial first article rationale with GNSEMS using that evidence. The question to resolve is which earlier results remain valid and what the changed build must demonstrate.
Retain the reviewed comparison alongside the result references so future revisions can start from an established evidence map. This makes later decisions traceable without treating the present approval as unlimited permission for unrelated changes.
Questions about a partial PCBA first article
Does a one-component change always justify a partial first article?
No. The number of changed components does not establish the effect on the evidence. Review the associated characteristics, relevant process effects and applicable requirements before deciding the scope.
Can unchanged drawings justify retaining every old result?
Not by themselves. Manufacturing, software or verification conditions can change even when a drawing does not. Retention needs a basis showing that the earlier evidence still represents the current condition being claimed.
What if the previous first article report is incomplete?
Identify the missing evidence and its effect on the proposed baseline. Recover suitable records where possible. If the baseline cannot support retention, reconsider the scope instead of treating an incomplete report as a completed approval.
Who should approve the partial verification rationale?
Use the authority required by the actual contract and first article process. The review may involve design, quality, manufacturing and customer roles according to their delegated responsibilities. A generic job title does not establish approval authority.
