How Does a Touchscreen Work? Why Glass Can Feel a Finger

Beneath the outer glass of a capacitive touchscreen lies a pattern of electrodes made from indium tin oxide, a material that conducts electricity and is almost perfectly transparent. The electrodes run in rows and columns on two layers, or on the two faces of one layer, forming a grid of crossings. Each crossing acts as a tiny capacitor, a pair of conductors separated by an insulator that stores a small electric charge between them.

A controller chip drives the rows one at a time with a rapid electrical signal and measures how much of it reaches each column. With nothing touching the screen, every crossing returns the same value, and that baseline is measured continuously to allow for temperature and humidity.

The human body conducts electricity well enough to act as a third plate. When a fingertip comes close to a crossing, some of the electric field that was passing between the row and column electrodes is drawn into the finger instead, so the signal reaching the column drops. The controller notices which crossings dropped and by how much, and from the pattern of drops across several neighbouring crossings it calculates the centre of the touch more finely than the grid spacing.

Because each crossing is measured separately, several fingers produce several separate dips, and the controller reports each one as its own touch. That is what makes pinching, rotating and two-thumbed typing possible, and it is the difference between this design, called projected capacitive or mutual capacitance, and the simpler early capacitive screens that could only find one point.

Anything that does not conduct will not register. A woollen glove holds the finger too far away to disturb the field, which is why touchscreens ignore it; gloves sold as touchscreen-compatible have conductive thread woven into the fingertips to bridge the gap. A plastic pen does nothing, but a stylus with a conductive rubber tip, or a metal one, works because it carries the field to the hand holding it.

Water is the nuisance. A drop on the screen conducts and disturbs the field much as a finger would, which is why a wet phone registers ghost touches, and rain on a screen is worse than a single drop. Modern controllers deal with it partly by measuring each electrode's capacitance to ground as well as between rows and columns, a second mode that responds to fingers but much less to water, and switching between the two.

Before 2007 most touchscreens, on cash machines, ticket kiosks and early handhelds, were resistive. Two flexible conductive layers were held apart by tiny spacers, and pressing the surface pushed them together at one point, closing a circuit whose position could be measured. They worked with any object, gloves included, and were cheap, but they needed real pressure, could only register one touch, and the flexible layer dulled the display beneath it.

Capacitive screens had existed since the 1960s but had been single-touch and expensive. The change came when projected capacitive sensing was combined with a glass front, a fast controller and an operating system built around gestures. The first iPhone was not the first capacitive phone, but it was the one that made the technology the default, and within five years resistive screens had almost disappeared from consumer devices. The chain of events in one tap:

• •The controller pulses each row and measures every column
• •A fingertip draws field away from the crossings beneath it
• •Those crossings return a smaller signal than the baseline
• •The pattern of reduced signals is turned into a coordinate
• •Coordinates are tracked frame by frame to recognise a gesture

A touchscreen works by measuring a grid of tiny capacitors under the glass; a conductive fingertip steals a little of the field at the crossings it touches, and the controller turns the pattern of disturbance into a position. That is why it senses several fingers at once, ignores plastic and ordinary gloves, and misbehaves in the rain.