How a Piezo Cell Raises a Dot

The Actuator Underneath

A refreshable braille display raises and lowers each of its pins independently, responding to a new character in roughly 200 milliseconds. That speed, and the silence, and the low power draw, and the ability to pack eight actuators into a cell roughly five millimetres wide — none of it was possible with the technology that came first.

Early refreshable displays used solenoids: electromagnetic coils that pulled a ferromagnetic plunger to raise a pin. The principle is simple and robust. The problem is size. A solenoid powerful enough to raise a pin against the resistance of a fingertip — typically around 15 to 20 grams of force — requires a coil of significant diameter, and packing six or eight of them into the footprint of a braille cell demands engineering compromises that quickly hit a physical wall. Solenoid displays were large, heavy, power-hungry, and expensive to manufacture reliably. Some early units were also audible — a soft but persistent clicking that many users found fatiguing over long reading sessions.

Piezoelectric actuators changed the equation. They have no moving coil, no ferromagnetic core, no electromagnetic field that must be isolated from a neighbour. Their active element is a ceramic wafer. Voltage goes in; mechanical deflection comes out. The mechanism is quiet, fast, and — critically — small enough to be stacked in a matrix dense enough for practical braille.

The Bender, the Rod, and the Pin

The specific geometry used in almost every modern refreshable display is called a bimorph bender, sometimes a bender disc. Two thin layers of piezoelectric ceramic — most commonly a lead zirconate titanate formulation, abbreviated PZT — are bonded together with their polarisation axes oriented in opposite directions. A conductive layer between them and electrodes on the outer faces allow a voltage to be applied across the stack.

When voltage is applied, one layer tries to expand in the planar direction and the other tries to contract. Because they are bonded and cannot slip against each other, the wafer cannot expand or contract uniformly — it bends. The degree of bend, measured at the free end of a cantilever configuration, is proportional to the applied voltage and proportional to the square of the beam's length. In practice, a wafer roughly fifteen millimetres long, driven at somewhere between 100 and 200 volts, will deflect its tip by around 0.3 to 0.6 millimetres — enough, once the geometry is right, to raise a pin by the 0.48 millimetres that the standard braille dot height specifies.

That voltage figure surprises people. Refreshable braille displays run on batteries and connect to low-voltage electronics, yet the actuator demands a hundred volts or more. Every display therefore contains a boost converter — a small switching power supply that steps the battery voltage up to the drive voltage required by the ceramic. The conversion happens continuously and efficiently; the power consumed per actuator firing is small because the ceramic is a capacitive load, not a resistive one. Energy goes into bending the crystal lattice; most of it can be recovered when the voltage is removed.

The bender's deflection is transmitted to the pin through a mechanical push-rod, usually a thin stainless steel or hardened polymer shaft. The push-rod sits between the tip of the bender and the base of the pin, often with a small pivot or lever arrangement that both amplifies the mechanical displacement and translates it from the horizontal plane of the bender into the vertical motion the pin requires. The design of this lever is where manufacturers differentiate. A modest mechanical advantage — two or three to one — lets the designer use a slightly shorter, stiffer bender (which responds faster and is less prone to resonance) while still hitting the required pin travel. The trade-off is force: mechanical advantage that multiplies displacement divides force, so the lever ratio is constrained by the minimum push-force the pin must exert under a reading finger.

The pin itself is typically 1.4 to 1.6 millimetres in diameter at the tip, rounded to avoid cutting the skin, and it passes through a guide plate that keeps it aligned and prevents lateral wobble. The guide plate's hole is sized to a close tolerance — loose enough to allow free vertical travel at the speed required, tight enough to prevent the pin from tilting under finger pressure. When the actuator is not energised, a small return spring or the mechanical preload of the bender's resting position pushes the pin back down.

Why the Geometry Holds

Eight actuators — the extra two accommodate the cursor-tracking dot pair that some displays add to each cell, though most standard cells use six — must fit into a cell whose centre-to-centre spacing is 2.34 millimetres in the horizontal direction and 2.34 millimetres in the vertical, as specified by the ISO standard governing braille dimensions. The bender wafers themselves are longer than that spacing, which means they cannot be arranged in a simple grid directly below their pins.

Manufacturers solve this with a fanned or layered layout. The benders for the dots in a single column are staggered in depth — physically arranged at different heights beneath the face plate, angled slightly so their tips all converge on a narrow vertical stack of push-rods. Looking at a cross-section of a modern display cell, the actuators for a single column of three dots resemble a spread hand of cards, each bender at a slightly different elevation, each connected by its own push-rod to the pin above it. The depth required for this layering is the main reason braille display modules are as thick as they are: typically 15 to 25 millimetres from the face plate to the base of the actuator module.

The ceramic itself has one significant vulnerability: it is brittle. PZT, like other ceramics, will fracture under a sharp impact or if the pin encounters a load far beyond its design limit — an object placed on the display, for instance, or a user pressing hard on a pin that is already at full travel and mechanically stopped. Good mechanical design includes end-stops in the guide plate that take the load before it reaches the ceramic, so the bender never sees the force of a direct impact on the pin. The operating temperature range also matters: piezoelectric properties are temperature-dependent, and significant cold or heat will alter the effective drive voltage needed to reach standard dot height. Field devices used in outdoor or mobile contexts must compensate for this, either through temperature-sensing and voltage adjustment, or by designing the actuator to meet spec at the worst-case temperature within its rated range.

What Replaced What

Solenoid displays never entirely disappeared — a small number of specialised applications, particularly where very high pin force is required, continue to use electromagnetic actuation. But for the compact, battery-powered, multi-cell displays that have become the standard interface for accessing digital text through touch, PZT benders are now the dominant mechanism, and have been for several decades.

The reason is ultimately about energy density and packaging. A piezoelectric actuator delivers its mechanical work through the distortion of a crystal lattice, a process that requires no magnetic field, no coil winding, and no ferromagnetic mass. The energy per actuation cycle is small, the mechanism is repeatable to millions of cycles without wear in the conventional sense (fatigue in the ceramic bonding layer is the eventual failure mode), and the whole assembly can be produced with the dimensional consistency that precision pin geometry demands.

Understanding the bender also clarifies a constraint that refreshable braille displays have never escaped: each pin requires its own discrete actuator, its own push-rod, its own guide hole. A forty-cell display contains at minimum 240 independent mechanical systems, all of them built to micrometre tolerances, all of them driven by a high-voltage circuit that must switch each one individually. Elegance at the scale of a single bender becomes formidable complexity at the scale of a full line display, and that complexity is where the cost lives.