Madhav Kavuru
This design is unpatented and will stay that way
Here is the paper trail so you don't have to take my word for it.
I filed US provisional patent application 63/819,460 on June 6, 2025, titled "Virtual Scroll Wheel Input Device with Adaptive Touch Sensing and Modular Haptic Feedback." I let it lapse — no non-provisional was filed, and the twelve-month window closed on June 6, 2026. In August 2025 I released the full specification and figures as open source, posting them at several online locations including LinkedIn, together with the one-page sell sheet in the gallery above. That publication is now more than a year old, which under US law bars me from patenting this and makes it prior art against anyone who files after that date. Patent searches turned up no prior art that would block implementation. A defensive publication is on file at Technical Disclosure Commons so the record is indexed and searchable by patent examiners.
So: if you make laptops, mice, touch controllers, or haptic drivers, you can build this. You owe me nothing. I would like attribution and I would like to see it exist.
And build the small version. Costed line by line below, the whole thing — arc, haptic motor, tilt, zoom — comes to roughly $4.80–8.70 of BOM at 100k units a year, and Core alone is roughly $3.70–6.50. But the reason to start with Core is not the dollar or two. Core is three components and a firmware loop: a curved capacitive arc, a smartwatch-class LRA, and centroid-based vertical scrolling with synthetic detents. It delivers what a mouse wheel delivers, in a 9.5 mm stack, without putting a vibration-isolation problem on your critical path. Tilt-to-horizontal is there when you want a second generation, and pressure-zoom adds nothing to the BOM at all.
If you are a hobbyist, the interesting unsolved parts are the arc electrode layout, tuning the contact-width gate against real finger data so press-down never registers as a stroke, and getting convincing detent feel out of a cheap LRA. Fork it and tell me what breaks.
— Madhav Kavuru, mkavuru2@gmail.com
The problem, stated precisely
Every laptop maker has been asked for a scroll wheel. None ship one. The reason is not taste, it is 15 millimeters.
Scroll wheels on current mice run roughly 15–22 mm in diameter. A thin laptop's keyboard section gives you about 10 mm total. Shrink the wheel below 15 mm and it becomes unpleasant to use, and you still have to redesign and miniaturize the optical encoder, the bearing, and the ratchet. So instead we get scroll strips painted onto a corner of the touchpad, which are too sensitive to be precise, break after driver updates, eat usable touchpad area, and force you to steer around them to avoid scrolling by accident. They feel like dragging a scroll-bar thumb, not like a wheel.
Arrow keys, meanwhile, remain terrible. In a word processor they move line by line and then jump a page at a time.
The idea
You never touch most of a scroll wheel. You touch an arc at the top. So build only that arc.
Take a disk segment — a circular segment no bigger than a semicircle — and stand it up through an opening in the chassis. Put a mutual-capacitance touch sensor along its curved outer face. Now your fingertip has the same curved reference geometry it gets from a real wheel, which is exactly what a flat strip fails to provide. With a 22 mm implied wheel diameter and a 17 mm chord, the segment is 4.02 mm tall — about the same as the exposed part of a mouse wheel — and about 2 mm of it pokes above the keycaps.
Then give it detents that don't exist. Bond a linear resonant actuator to part of the flat chord face of the segment. Smartwatch LRAs (12 × 5 × 3 mm, the class of part used in Apple Watch) are mature, cheap in volume, and already qualified for sub-20 ms haptic latency. Dimensions that are a problem in a watch are nothing in a laptop. Drive it with PWM waveform profiles and you can synthesize click detents, free-spin...
Read more »