If the detent and the electrical make point are out of sync, feel and effect misalign

What it is

The snap dome has collapsed, but the circuit has not made; or the lamp is already on, and the resistance break is still ahead. Detent–actuation desynchrony splits “done in the hand” from “done in the system.” People who end the action on feel may never have been received; people who wait for the effect keep pressing after feel already said pass. A real detent and enough travel can still void the receipt at this layer.

Why it happens

Mechanical over-center and electrical make are two tolerance stacks that can slide: collapse displacement of the dome, contact gap, debounce time, scan period. If the tactile point sits before make, people release at collapse, the key rebounds, contact never sits closed—a miss. If the tactile point sits after make, the system has already fired while the hand is still adding force waiting for the break—“it already happened and I am still pressing.” Repeats and long-press timers stretch, and the extra travel after make can take a neighbor with it. Firmware debounce that postpones electrical recognition lets the offset drift with temperature and scan rate. The hand treats the detent as a receipt because it is usually bound to make; desynchrony unbinds that learning, and people revert to bottoming or watching the screen.

Studying it

Sample force–displacement, electrical make, and the moment of release together, and deliberately shift the tactile point before or after make.

Independent variables: displacement gap between tactile point and make, debounce time, scan period, rebound speed. Dependent measures: missed actuation, double actuation, release time relative to make, neighbor coupling, whether people switch to a bottoming strategy.

Datasheet millimeters for “tactile point” and “make point” are not enough; include the human termination. Run a temperature chamber, because lubrication and plastic gaps move the two points relative to each other.

Where it stops holding

Hall and optical keys put “make” at a firmware threshold that can be aligned with a motor pulse; they lack a snap-dome mechanical split, but they can show a different split in which the pulse lags the threshold. Long-press and auto-repeat depend on hold time; an early tactile point makes people release too soon, and the function looks broken. Low-latency game and music keys would rather make slightly before feel than miss; form entry fears early make more. Wear can open two points that started aligned; a new-key pass does not mean mid-life sync.

Applying it

• Put the tactile drop and electrical make inside the same short displacement; allowed offset must be smaller than the time of a release action.
• Charge debounce and scan period against that budget; do not add tens of milliseconds of recognition after the mechanics already line up.
• Verify: record force, make, and release at high speed. Release often before make means the tactile event is early; force still rising after make, waiting for a drop, means it is late. Retest the same gap after hot/cold and after ten thousand presses. On long-press keys, score failures to hold the full time as their own case.

Related

• Same group: C10.08.1 A tactile detent (a sudden drop in resistance at the fire point) marks that the action was received · C10.08.2 Keys with too little travel cannot give clear mechanical feedback at the moment of press · C10.08.4 Thin devices often sacrifice travel and simulate a detent with a haptic actuator
• Adjacent: C10.16 Mechanical Reliability and Lifetime · C2.22 Bounce Clicks and Repeat Triggers
• Search: make point · tactile point · switch debounce