Executive answer
Choose the reporting rule before collecting the readings
- Keep the two force inputs paired to the same key and observation before calculating a percentage.
- A mean of ratios and a ratio of means answer different summary questions; write the chosen operation into the report.
- A passing panel summary cannot replace a per-key requirement, mounted evaluation or electrical evidence.
Current guidance context
What OEM engineers and sourcing teams should decide first
A quotation or sample sheet may say 'average tactile ratio: 50%.' Before comparing it with another supplier's number, ask for the calculation behind that field. Does every key get one ratio, or are force readings combined before the ratio is calculated? Does the requirement concern every key or the panel's summary?
This is a report-definition question, not a request to maximize a percentage. The worksheet below uses the same two fictional pairs in both calculations so the difference cannot be blamed on a different sample. The practical output is a paired-data record and an explicit acceptance field.
Name the force fields before using the formula
Snaptron describes tactile ratio as the percentage drop from peak force to a minimum after snap. In that convention, TR = 100 × (Fmax − Fmin) / Fmax. The values must use compatible force units and the same declared observation and point-selection rule. A peak-force reading by itself cannot supply the second input.
Its Sapphire tester field list distinguishes Fmin, the minimum after Fmax, from Fc at contact, Fb at break and Frmin/Frmax on return. Use those labels as a source-specific dictionary, not permission to substitute whichever low value is easiest to find. Another report's word 'release' needs its own definition and trace location before it is mapped to Fmin.
Keep the method identity separate from the equipment advertisement. ASTM's official F2592-16 record is withdrawn in 2023 with no replacement; its public scope discusses force-displacement characterization beyond a single ratio. This article does not reconstruct that procedure or make a percentage equivalent to a tactile slope or slope angle. The existing force-curve guide addresses the wider method and electrical-event record.
Work the same two force pairs two ways
Hypothetical HYR editorial arithmetic only: imagine one recorded observation for each of two keys, A and B, under one common method and specimen state. These are scalar inputs, not measured hardware, complete force curves or evidence that either key can be manufactured. A uses Fmax = 2.0 N and Fmin = 0.5 N. B uses Fmax = 6.0 N and Fmin = 4.5 N.
Calculate each row first. A gives 100 × (2.0 − 0.5) / 2.0 = 75%. B gives 100 × (6.0 − 4.5) / 6.0 = 25%. With equal weight for these two rows, the arithmetic mean of their ratios is (75% + 25%) / 2 = 50%.
Now average the forces first: mean Fmax = 4.0 N and mean Fmin = 2.5 N. Applying the same ratio formula to those means gives 100 × (4.0 − 2.5) / 4.0 = 37.5%. The data have not changed. The order of operations has changed the summary by 12.5 percentage points.
For positive peak forces and the same complete paired rows, the ratio of force means equals a peak-force-weighted mean of the row ratios. Here B receives three times A's weight because its peak is three times as large. That is an arithmetic identity, not a recommendation to weight heavier keys more strongly. Choose the intended statistic before interpreting the result.
Attach the limit to a field, not the word average
To expose the decision problem, introduce an explicitly hypothetical minimum of 40%. This is not an acceptable-feel recommendation or an ASTM limit. If it applies to each of the two row ratios, A satisfies it and B does not. The mean of row ratios, 50%, satisfies a 40% minimum on that summary. The ratio of mean forces, 37.5%, does not satisfy a 40% minimum on that different summary.
Those are three different rules, not competing verdicts on an approved product. A summary-only requirement leaves individual values uncontrolled unless it also names a per-key condition. Conversely, a per-key requirement must not be replaced after measurement by a favorable panel average. Retain any range, outlier or critical-key rule the project actually requires.
State the reporting hierarchy too. Repeated cycles of one key, different keys on one panel and different panels are distinct observation groups. If a row is already a group mean, do not give every such row equal weight while calling the result an observation-level mean unless that is the intended rule. Keep group sizes and weights with the calculation, and do not infer process capability from this two-row exercise.
Keep component and mounted records separate
Snaptron's overlay guidance separates a dome's own characteristics from the force-displacement response experienced at the finished button surface. The overlay, actuator, adhesive, spacer and support belong to that installed context. Do not average standalone dome readings with external-button readings and describe the mixture as one construction's tactile ratio.
Identify the physical stage on every row, alongside the key and specimen. If the purpose is to compare two panel constructions, use the agreed support and pressing method for both. If the purpose is to evaluate the installed interface, preserve the relevant housing and stack revision. A numerical summary cannot establish that those configurations are equivalent.
Keep electrical observations separate as well. The presence of a calculated force ratio does not establish a make or break event, correct mapping, return behavior or durability. Request the additional channels and checks that answer the project's actual question rather than adding unsupported meaning to the percentage.
Catch five summary failures before acceptance
These are hypothetical report-review failures, not observed HYR defects or customer cases. Resolve the missing definition rather than silently changing a number.
- An unexplained average: the sheet gives a percentage but not the order of operations. Request the row formula, aggregation and weighting.
- A substituted low force: electrical contact or a return-path minimum is used where the agreed formula requires a post-peak minimum. Recover the trace and correct the field mapping.
- Broken pairing: peaks and minima come from different keys, cycles or specimen states. Rebuild matched rows before calculating ratios.
- Approval by summary: a favorable panel mean is substituted for a stated per-key requirement. Apply the original field-level rule and disposition the exceptions.
- Hidden data handling: missing rows, unequal groups or rounding alter the summary without being recorded. Preserve the available precision, exclusion reason and intended weighting.
Release the paired-data record with the RFQ
Give the supplier a compact record layout: specimen and key ID, observation ID, physical state, method revision, force units, paired Fmax/Fmin, trace reference, calculated row ratio and row disposition. Add a separate summary record that names its included rows, operation, weighting, rounding and acceptance field. Open inputs can be labeled as quotation assumptions until the responsible team defines them.
Retain the unrounded values available from the agreed measurement system; do not manufacture extra precision. Recalculate the summary from the released rows and compare it with the reported result. If Fmax is zero, the ratio is undefined. If a selected minimum is inconsistent with the agreed post-peak definition, investigate the observation rather than clipping it into a convenient percentage.
Use HYR's secure RFQ to share the available panel drawing, tactile-element and mounting context, and a nonconfidential report format or question. The review can clarify the information needed for comparison. It does not promise a tactile target, measurement service, test result, quotation outcome or delivery schedule.