On this page
A calibration fixture can measure a resistor card repeatably and still predict the installed system poorly. The fixture may reference different datums, press at a different radius, use a different footprint, travel without housing backlash or sample after a different settling time. Correlation begins with the contact trajectory rather than the fixture drawing. Position, normal force, skew, speed, reversal, excitation and acquisition timing must represent the intended assembly closely enough for the calibration decision being made.
System boundary
A calibration fixture, representative contact, resistor card, force and motion controls, electrical acquisition and the installed mechanism used for correlation. The fixture owner controls metrology; ChipSimple supplies drawing-defined card features; complete sensor calibration remains with the system integrator.
System integration decisions
- Reference fixture travel to the same functional origin used in the assembly.
- Match the contact footprint and force window, not merely endpoint stroke.
- Quantify correlation error separately from card transfer error.
Define trajectory as position, orientation and footprint
Represent the moving contact by its center path, angular orientation and full footprint at every calibration coordinate. A nominal line or arc omits skew and wiper width. Include vertical compliance because it influences which part of a crowned or tilted contact touches. Overlay fixture and assembly paths relative to card holes or edges with identical datum rules. Mark transition, active and prohibited zones.
Separate fixture indication from installed position
A local correlation model allocates offset, scale, angular and repeatability terms.
x_inst=a+b x_fix+r Delta_theta+e_seat+e_rev
- a and b are fitted offset and scale between coordinate systems.
- r Delta_theta is cross-track or arc displacement caused by angular mismatch.
- e_seat is card seating contribution and e_rev is direction-dependent reversal contribution.
The fitted relation applies only over the tested travel and assembly configuration.
Calculate a calibration consequence from scale mismatch
If fixture coordinate scale is 0.4 percent high, an indicated 50.0 mm point corresponds to 49.8 mm before other errors. If local card slope is 3 ohms per millimeter, the position difference represents about 0.6 ohm. The values illustrate propagation only. Actual correction requires traceable coordinate measurements and cannot be inferred from electrical slope alone.
Reproduce normal force and contact mechanics
Use the intended contact material, tip geometry, footprint, spring direction and bounded force wherever the fixture result depends on contact resistance or track traversal. Measure force through travel rather than setting one endpoint. A rigid probe may be appropriate for direct film characterization but is not automatically representative of a production wiper. Keep characterization and system-correlation modes clearly labeled.
Match speed, dwell and reversal states to the decision
Static point measurement can isolate curve geometry; continuous travel can expose bounce, friction and acquisition latency. Specify approach direction, acceleration, dwell, sample window and number of reversals. Do not average fill and return sweeps if hysteresis is itself an acceptance characteristic. Record environmental conditions and surface preparation because contact behavior may change during repeated cycling.
Use paired fixture and installed records
Measure the same identified cards in both configurations without changing their surface condition unnecessarily. Retain raw position, resistance or voltage, force where available, direction and time.
| Attribute | Fixture evidence | Installed comparison |
|---|---|---|
| Datum origin | Traceable coordinate acquisition | Housing or mechanism reference |
| Contact path | Mapped footprint over card | Witness or measured product trajectory |
| Force | Force-versus-position record | Assembly contact window |
| Signal | Raw synchronized output | Receiver-boundary output |
Build a correlation uncertainty rather than hiding bias
Separate fixture position calibration, card seating, contact repeatability, electrical instrument error, timing alignment and fitted-model residual. Correct stable bias only when its physical cause and configuration are controlled. A small average difference can coexist with large endpoint or direction errors. Report residuals over position and by direction. The customer decides how correlation uncertainty enters calibration acceptance.
Diagnose divergence by its shape
Constant fixture-to-assembly difference suggests origin. Growing difference suggests scale or angular mismatch. A loop between forward and reverse directions indicates backlash, friction or contact effects. Isolated spikes can result from surface events or sampling skew. A difference that follows card identity points toward card geometry; one that follows the fixture setup points toward seating or instrumentation. Preserve raw evidence before adjustment.
Release the fixture for a bounded calibration purpose
State whether the fixture screens resistance, assigns curve calibration, verifies endpoints or studies contact behavior. Qualify it only for that purpose and the tested card, contact and assembly family. ChipSimple can review drawing-defined card datums and track geometry; fixture design and installed correlation remain customer-controlled. Reopen after probe, spring, datum, software, acquisition, card mounting or mechanism revisions.
Fixture maintenance must preserve the qualified trajectory, not merely machine operability. Control locator wear, probe replacement, spring adjustment, software coordinate resets and cable routing. Establish checks that independently reveal origin, scale, force and acquisition changes. A golden card may provide continuity of comparison only when its storage, surface condition and stability are understood; it is not a substitute for traceable mechanical and electrical references. Before and after maintenance, compare mapped residuals rather than one midpoint value. If the fixture serves multiple card variants, use keyed setups and configuration validation to prevent a correct program from running with the wrong locator or contact. Retain the fixture identifier and setup revision on every calibration record.
Environmental correlation deserves its own boundary. Fixture temperature, humidity and contamination can change contact behavior or card resistance differently from the installed assembly. If calibration occurs at a controlled laboratory condition, define the correction or allowable transfer to operating conditions. Record warm-up, stabilization and self-heating. Repeat selected installed points after fixture calibration to confirm that environmental compensation has not merely improved agreement at one reference state.
Before routine use, confirm software limits, emergency stops and operator access do not alter the intended measurement path. Document the card cleaning rule and maximum permitted contact cycles between checks.
Provide the installed and fixture trajectories
Correlation needs both coordinate systems and the measurement decision.
- Card datums, track drawing and required transfer coordinates.
- Installed wiper path, footprint, force, speed and reversal behavior.
- Fixture locators, contact, motion profile and acquisition architecture.
- Correlation acceptance, uncertainty treatment and responsible owners.
The drawing-upload form loads as you reach this section.

