Capacitive touch sensors report finger position by centroid-weighted averaging across multiple activated pads

Capacitive sensing measures how a conductive object alters the electric field around a charged plate, changing the plate’s capacitance. When a finger touches or approaches a pad it couples additional capacitance to ground, changing the measured value. Multi-pad sensors (like Trill) enable continuous position tracking: a finger simultaneously activates several adjacent pads, and a centroid computation (weighted average of pad positions by activation level) gives a sub-pad-resolution position. A zig-zag pad layout ensures fingers always contact multiple pads, keeping centroid detection reliable. Sensitivity must be tuned per material: conductive thread, graphite, water, and metal disturb the field differently and need different baseline calibration.

Examples

Trill Craft: 30 capacitive channels, each an independent plate. A finger at position 0.5 activates pads 14 and 15 at slightly different strengths; centroid = (14·s14 + 15·s15)/(s14 + s15) ≈ 14.5.

Assessment

Why does a zig-zag pad pattern give more accurate centroid position than straight parallel strips? What must change in software when switching from a metal touch surface to conductive thread?