Mechanical detents come from a physical pawl; electronic detents are motor-simulated, and they feel different

What it is

Mechanical steps come from a pawl or a ball dropping into a notch; the force curve is metal and spring, and it survives power-off. Electronic steps are a slotless encoder plus a motor: at the angle software thinks should stop, the motor bumps or drags. Mechanical and electronic detents feel different—not which is more advanced, but where the event comes from, how late it is, and how it fails. Step density is the previous layer. This one asks whether the step is a notch or a pulse.

Why it happens

A pawl’s drop is bound to geometric lock of the shaft; delay is material, near zero; reverse, and the notch is still at the same angle. A motor simulation samples angle, computes, then applies force; the pulse lags true angle, and a fast twist outruns it, so the step becomes “a buzz after you already passed.” A motor can also change step count and strength with mode, which a pawl cannot; but when the driver is off, in power-save, or saturated, the steps vanish and the knob is suddenly a smooth shaft—quantized semantics fall back to continuous. A worn mechanical notch gets shallow; the step is still there, only dull. Electronic steps can be rewritten overnight in software, resetting the hand’s counting ruler. Both can “have detents”; the receipt lives in geometry in one and in a servo loop in the other.

Studying it

Same diameter, same nominal steps per turn, compare a pawl part with a force-feedback encoder, sweeping rotation speed.

Independent variables: detent implementation (pawl / ball / motor), command delay, whether software may change step count, power-off or driver-off. Dependent measures: missed steps in a fast twist, whether reverse still finds the step at the same angle, behavior with the driver off, whether people still count by steps.

Separate “liked” from “can count.” Electronic detents can mimic well at slow speed; defects show in flicks, reverse, and power loss. Measuring driver-off is required, not a failure demo.

Where it stops holding

An incremental encoder with no motor can be electrically fine and completely smooth in the hand; do not call it an electronic detent. Automotive center knobs whose function changes almost require electronic steps so volume and lists can use different densities; a pawl cannot host two step sets in one knob. A safety-critical “must feel a step” that must survive power loss cannot ride a motor alone. Cold thickens grease and makes mechanical notches heavy; motors may weaken under current limits; the two failures are not the same “worse feel.”

Applying it

• Channels whose steps must remain with no power get a pawl or ball, not a servo impersonating them.
• Knobs that must change granularity with mode get a motor, and the change of step must be noticeable at mode switch so the same feel does not suddenly mean a different increment.
• A motor scheme must define semantics with the driver off: free, braked, or adjustment forbidden.
• Verify: slow turns, flicks, and fast reversals, tens of each; count missed and misplaced steps. Turn the motor off and twist; see whether people still believe steps are there. Retest in a temperature chamber. If flick misses exceed what you can accept, drop steps per turn or go back to a mechanical notch—do not only make the pulse louder.

Related

• Same group: C10.10.1 Detents per revolution cap the granularity a single turn can achieve · C10.10.3 Too little damping lets a knob turn by itself under vibration · C10.10.4 The resistance curve between detents decides whether people can count steps by feel alone
• Adjacent: C10.02 Knobs and Continuous Adjustment · C10.08 Key Travel, Tactile Detent, and Confirmation
• Search: haptic encoder · mechanical detent · virtual detent