Haptic feedback design depends on which receptor class it is meant to engage

What it is

The four mechanoreceptor classes on a patch of skin — Merkel cells, Meissner corpuscles, Pacinian corpuscles, Ruffini endings — peak in sensitivity to pressure, light touch/slip, high-frequency vibration, and skin stretch respectively. That means a haptic actuator's waveform, frequency, and contact area determine which receptor class it is mainly "talking to." No single waveform activates all channels equally well.

The common misreading: this is not a matter of "bigger amplitude, stronger feeling." The same amplitude at different frequencies engages different receptor combinations, producing sensations that differ in kind, not just in intensity.

Why it happens

The Pacinian corpuscle's lamellar capsule makes it sensitive to roughly 40–500 Hz vibration and lets that signal travel through solid media — tools, casings — over some distance, so most of the buzz produced by eccentric rotating mass (ERM) motors and linear resonant actuators (LRA) primarily drives this channel. Merkel cells respond to slow, sustained indentation, driven by changing force magnitude and contact area rather than vibration frequency, so pin arrays and pneumatic bladders that can hold a static deformation suit them better. Ruffini endings respond to local skin stretch and need an actuator that produces shear or tangential stretch rather than perpendicular indentation or vibration — a waveform that is engineering-harder to produce, which is why skin-stretch haptic devices remain far less common than vibration motors.

Picking the wrong waveform is therefore not "a weaker effect" — it can mean the target channel was never engaged at all. Driving a 150 Hz-only motor to simulate "a finger being slowly pressed in" lands outside the band Merkel cells are tuned to, and the intended percept simply does not emerge.

Studying it

Confirming that a given feel is driven by a target channel typically uses selective adaptation/masking paradigms: fatigue one channel with vibration in a specific band, then check whether the detection threshold for a target stimulus rises accordingly — a rise indicates the two share a channel. This method underlies Bolanowski and colleagues' four-channel psychophysical model, which maps subjective reports of "vibration," "pressure," and "stretch" onto specific receptor types.

In device and interface evaluation, this kind of paradigm explains why two products rated at the same nominal vibration intensity can feel different to users — the difference usually lies in spectral content, not total energy.

Where it stops holding

The channel correspondences were established under controlled lab conditions (fixed probe, fixed contact force, single-frequency sinusoidal vibration). Real devices produce composite waveforms (motor resonance, harmonics conducted through the casing), so isolating a single channel is hard to guarantee; individual sensitivity to each channel also varies, so the same device can activate a different weighting of channels across users.

Applying it

• Want a pressure feel? Don't just turn up amplitude. For a sustained-pressure percept, prefer mechanisms that hold a static deformation (pneumatic bladders, servo pins) over raising a vibration motor's drive voltage.
• Want fine texture? Use high-frequency, short pulses. To simulate fine texture or a crisp click, use short, high-frequency (near 200–250 Hz) LRA pulses to engage the Pacinian channel precisely, rather than a sustained low-frequency buzz.
• Want a posture/flexion feel? Avoid vibration motors. Feedback involving skin stretch (such as a finger-flexion signal in virtual grasping) calls for a compliant stretch actuator; vibrotactile hardware has limited reach into this percept.
• How to check: run forced-choice paired comparisons across waveforms with the same participants (which feels more like "pressing," which more like "buzzing") rather than only measuring detection thresholds — this directly reveals which channel is engaged, not just whether the stimulus is strong enough.

Related

• Same group: A4.01.1 Different receptors respond to pressure, vibration, stretch, and sustained contact · A4.01.5 Pacinian corpuscles sit deep in the subcutaneous layer, dedicated to high-frequency vibration, including signals conducted through tools
• Nearby: A4.03 The frequency band of peak vibrotactile sensitivity · A4.10 Fast and slow adaptation
• Search terms: vibrotactile actuator · ERM motor · LRA · selective adaptation · haptic rendering