When a sound level meter fails its final field check, pause release of the affected results and preserve the evidence. A later satisfactory reading does not, by itself, establish that the earlier measurements were valid.
The immediate task is to identify what changed and which results may be affected. The following workflow is a recommended investigation structure. Apply the decision criteria and reporting obligations of your own measurement procedure.
Preserve the first failed observation
Record the indicated value, expected indication, time, settings and equipment identities. Save the measurement files and the initial check record. Note any adjustment already made, along with its timing.
Avoid repeatedly adjusting the meter until it displays the expected value. That can remove evidence about the original condition. If another check is appropriate, document it as a separate observation and record what changed before repeating it.
Check the setup systematically
| Question | Evidence to collect |
|---|---|
| Was the same reference signal selected? | Calibrator level and frequency settings for both checks |
| Was the microphone coupled correctly? | Adaptor identity, seating and visible condition |
| Did meter settings change? | Range, weighting and check configuration |
| Was there a power or connection issue? | Battery indications, connectors and recorded interruptions |
| Did handling or conditions change? | Impact, moisture, transport or environmental observations |
| Is the reference source itself questionable? | Calibrator certificate and any independent comparison |
These are investigation questions, not diagnoses. Change one relevant factor at a time where practical. Follow the manufacturer's instructions for handling the microphone and coupler; an inspection should not introduce further damage.
Define the affected period
Start with the last documented acceptable check made with a comparable setup. Identify subsequent measurement sessions, equipment changes and any unusual events. Extend the review further back if there is evidence that the problem began earlier.
Do not assume the fault appeared immediately before the final check. The check identifies a problem at that time, but may not establish when it started. Missing timestamps or equipment IDs can widen the period that needs review.
Separate current performance from historical validity
An investigation needs two decisions. First, is the equipment suitable for further use? Second, can the earlier results support their intended purpose? Evidence for one decision may be insufficient for the other.
The distinction between calibration and adjustment in JCGM 200:2012, entries 2.39 and 3.11, helps explain why changing an indication is not the same as establishing previous performance.
A subsequent laboratory result can help identify a fault or quantify current performance. It cannot automatically reconstruct every earlier field condition. Request observations before repair or adjustment when they are needed for the historical review.
Illustrative investigation scenario
Hypothetical example: A survey has an acceptable initial check and an unacceptable final check. The operator finds that the calibrator was not seated as specified. After reseating it, the next check meets the criterion.
The new result supports the possibility of a coupling problem during the final check. It does not prove that coupling was the only issue. The reviewer should examine the recorded sequence, equipment condition and any independent observations before deciding what to do with the survey.
If the cause remains unresolved, repeating affected measurements may provide a stronger basis than applying an unsupported correction. The appropriate action depends on whether repetition is possible and on the intended use of the results.
Document the disposition
Use an investigation record with these headings: event, original evidence, affected files, checks performed, changes made, findings, remaining uncertainty, decision and reviewer. Link each retained or repeated result to that record.
Where a report has already been issued, follow your organisation's process for assessing whether it needs correction or withdrawal. Keep the technical basis and communication record together.
Return the equipment to use
Record the work completed, supporting checks or laboratory results, authorised configuration and reviewer. Update the equipment status so another operator does not unknowingly use an unresolved setup.
Use the before-and-after field-check log to improve the next session's evidence. If similar incidents recur, review the calibration interval and intermediate-check programme. The main calibration guide explains how these controls fit together.
Questions about this topic
Can I subtract the final check error from every survey reading?
Do not apply a blanket correction without evidence that the correction is valid for the affected measurements and permitted by the measurement method.
Does a passing repeat check clear the original failure?
It is additional evidence. Preserve the original failure and document why the investigation supports the final decision.
Comments
Post a Comment