Two test stations can both pass a familiar board and still disagree on other units or operating points. Before moving production, compare their measurement behavior using a defined plan. The goal is to identify relevant differences, not to force identical numbers by adjusting one station until it resembles the other.
NIST’s measurement-process characterization overview distinguishes repeatability, reproducibility, and stability. A station comparison should consider more than a single reading because short-term noise, fixture differences, and changes over time can affect the result.
Define the comparison boundary
Record the software recipe, limits, instruments, fixtures, cables, corrections, and environmental conditions for both stations. Decide which measurements are critical and what difference would matter to the product decision. Establish the analysis and acceptance approach before seeing the results.
Choose stable, representative units that exercise the relevant measurement range and variants. Include any known difficult condition that the new station must handle. Do not select only central, easy-passing samples and assume they represent behavior near a decision boundary.
Control sequence and repetition
- Identify every comparison unit independently of the test attempt.
- Measure units on both stations using the same approved procedure.
- Include repeated measurements to observe within-station variation.
- Vary or balance the order where time or unit drift could matter.
- Preserve raw results and all interventions.
If moving a unit changes its temperature or state, include the required preparation before each reading. Otherwise, a time-dependent product effect may appear to be a station bias. Record the preparation rather than leaving it as an informal operator habit.
Look beyond average agreement
| Pattern | Question to investigate |
|---|---|
| Nearly constant offset | Reference, correction, or connection difference? |
| Difference grows with signal level | Range or scale-factor difference? |
| One station has wider scatter | Contact, settling, acquisition, or noise issue? |
| Only one variant disagrees | Recipe or fixture mapping difference? |
These patterns guide investigation; they do not establish the cause by themselves. Confirm any proposed correction with independent evidence and repeat the affected comparison. Review paired differences for individual units, since opposing errors can cancel in an average and create apparent agreement while each station makes different decisions for particular assemblies at the same specification boundary.
Document which measurements and conditions are supported by the study. A limited correlation exercise is not proof of universal equivalence. Retain the study with the station release and revisit it after changes that affect the measurement path. The configuration manifest guide helps define which changes need that review.
Move from reading to a build-ready project
Use the fabrication and assembly checklists to prepare drawings, production files and test requirements.
Open the DFM checklist