How an LCD Display PCB Is Built and Bonded

An LCD display is a glass sandwich that needs a driver, a backlight and a mechanical frame, and the board behind it has to serve all three. Unlike a board that only carries electronics, an LCD display PCB is part of a mechanical assembly with tight tolerances in one direction and almost none in another, and it carries signals that are fast enough to matter and currents that are large enough to warm the panel. Designing one means balancing those constraints against a thickness budget that is often measured in millimetres.

What the Board Has to Deliver

The panel itself is a grid of thin film transistor cells, each of which must be addressed with a voltage that sets the brightness of one sub-pixel and then held until the next frame. The driver integrated circuits that perform the addressing are usually bonded directly to the glass rather than mounted on the board, and the board’s job is to supply them with power, timing and data through a flexible connection.

The backlight is the second function. Whether it uses edge-lit light guides or a direct array of emitters, the backlight needs a driver that regulates current rather than voltage, a dimming input that may operate at high frequency, and a layout that keeps its switching noise away from the data lines. A third function, less obvious but always present, is mechanical: the board often carries the connectors, the touch controller and the mounting features that locate the whole module.

Bonding and the Flexible Connection

The connection between the board and the glass is where most of the difficulty lives. Chip on glass places the driver directly on the panel, and the board reaches it through a flexible printed circuit that is bonded with an anisotropic conductive film. That film conducts only in the direction of compression, so the bond is formed by applying heat and pressure across the connection area, and the pad geometry, the adhesive thickness and the bonding parameters all have to be controlled together.

Chip on film is the alternative, where the driver sits on the flexible circuit instead of the glass. It allows finer pitch and a smaller glass border, which is why it dominates high resolution displays, but it moves the assembly difficulty onto the flex. In both cases the board’s contribution is a set of pads with the correct pitch and the correct surface finish, and the finish matters more than usual because the bond is formed at a relatively low temperature. Our component tolerance and reliability notes describe how that joint is assessed for thermal and mechanical stress.

Layout for a Thin Assembly

The board is usually a small number of layers, because the mechanical envelope is thin and the routing is not dense. Two or four layers is typical, with the ground plane placed to give the data lines a reference and to spread heat from the backlight driver. Component height is the binding constraint: parts are chosen for their low profile, and anything tall is placed away from the area that must sit against the panel.

Connector placement is what determines whether the module assembles at all. The flexible tail has to reach the panel bonding area without being stretched or folded sharply, and the connector that joins it to the main board must be positioned so that the tail lies flat. Where the board carries both the display connector and a touch connector, they should be arranged so that the two tails do not cross or press on each other.

Backlight Driver and Thermal Considerations

The backlight driver is a switching regulator, and its layout follows the usual rule: keep the loop formed by the input capacitor, the switch and the return path as small as possible. On a display board this matters twice over, because the switching node is physically close to the data lines and to the panel itself, and a poorly contained field couples into the display and appears as a faint ripple in the image.

Thermal behaviour is dominated by the backlight. The emitters convert most of their input into heat, and in a thin assembly that heat has nowhere to go but into the board and the frame. Metal-backed laminates are used where the dissipation is high enough to justify them, and thermal vias beneath the driver and the emitter pads carry heat into the spreader. Our thermal management article describes how the copper areas are estimated for that case.

Signal Integrity and EMI

The data interface to the panel runs at a rate that depends on resolution and refresh, and on a high resolution display it can be fast enough to require controlled impedance and length matching. The traces should be routed over a continuous reference plane, kept short, and kept away from the switching node of the backlight driver.

Emission control is also about the flexible tail, which acts as an antenna for anything that couples onto it. Grounding the tail properly, keeping its signal and ground returns adjacent, and if necessary adding a shielding layer are the measures that keep the module within the limits of the applicable standard. The same care applies to the touch interface, which is a high impedance sensor network that is easily disturbed by nearby switching circuits.

Reliability in the Field

Display assemblies fail for mechanical reasons more often than electrical ones. The bonded connection between the flex and the glass is sensitive to moisture and to thermal cycling, and the failure appears first as a line of dead pixels or as an intermittent column. Design measures that help include keeping the bond area away from heat sources, avoiding sharp bends in the tail and specifying a coverlay that does not stiffen the tail exactly where it must bend.

The board itself should be designed so that the tail is not the only thing holding the assembly together. Mechanical fixing of the board to the frame, with the flexible connection left free of load, is what turns a fragile assembly into a product that survives a drop. Our design release checklist places these checks in the review sequence.

COG Bonding and What It Asks of the Board

COG bonding is the process that attaches a driver integrated circuit directly to the glass of the panel using an anisotropic conductive film, and it changes what the board has to provide. Because the driver sits on the glass, the board supplies it through the flexible tail rather than through a package, and the number of signals that must cross the tail is reduced to power, timing, data and a handful of control lines. The pads that mate with the tail must be flat, clean and finished with a surface that forms a reliable bond at the temperature the film requires.

The practical consequence is that the board and the panel are no longer independent. The tail length, the pad pitch and the position of the bonding area are fixed by the panel, so the board layout has to accommodate them rather than the other way round. When a panel changes supplier, the tail often changes with it, and a board designed with a generous pad area and a defined keep-out around the bonding region is far easier to adapt than one that was laid out to the minimum possible size.

FAQ

Why is the driver bonded to the glass instead of mounted on the board? Because the number of connections to a high resolution panel is enormous and the pitch is very fine. Bonding directly to the glass removes thousands of solder joints and the space they would occupy along the panel edge.

Can an LCD display PCB be a single layer? Occasionally, in small low resolution modules where the routing is trivial. Most designs use two or four layers because a reference plane is needed for the data interface and for heat spreading.

What causes a vertical line of dead pixels? Usually a broken connection in the bonded area or in the flexible tail rather than a driver failure, because a whole column shares one output. Moisture ingress and thermal cycling are the common causes.