Lamination Cycle Control In PCB Pressing

Lamination is the step in which the layers of a multilayer board are bonded into a single panel. The stack is placed in a press, heated and held under pressure, and the resin in the prepreg flows, fills the spaces around the copper and then cures. The way the cycle is controlled decides the thickness, the flatness and the bond quality of the finished panel.

This article covers what the cycle has to do, how the parameters are set, and how the result is measured.

What The Cycle Has To Do

The cycle has to accomplish three things in the right order. First the panel must be heated so that the resin softens and its viscosity falls. Then pressure must be applied while the resin is fluid, so that it flows into the spaces and the stack consolidates. Finally the temperature must be held long enough for the resin to cure, so that the bond is strong and the panel keeps its shape when it is cooled.

Getting the order wrong produces characteristic defects. Pressure applied too early, before the resin has softened, squeezes it out of the stack and leaves a dry bond and a thin board. Pressure applied too late, after the resin has begun to gel, leaves voids and a thick board, because the resin can no longer flow to fill the gaps. The relationship between the resin and the copper it has to fill is described under balanced stackup and odd layer count.

Parameters And Their Interaction

The cycle is defined by a temperature profile, a pressure profile and a time, and the three interact. A higher temperature lowers the resin viscosity, which allows a lower pressure to consolidate the stack. A slower ramp gives the heat time to reach the centre of a thick stack, but it also gives the resin more time to advance its cure before the pressure is applied. A stack with many layers needs a longer cycle than a simple one, because the heat has further to travel.

The prepreg itself is a variable. Its resin content, its flow behaviour and its gel time are properties of the material, and they change with age and with storage conditions. A batch of prepreg that has been stored warm has partially advanced, so it flows less and needs more pressure, or it may not fill the spaces at all. Keeping the prepreg in controlled storage and recording its batch with the lamination run is what allows a defect to be traced back to the material.

Resin Flow And Thickness Control

The final thickness of the panel is set by the amount of resin in the prepreg, the copper thickness, and how much resin flows out of the stack. The press controls the last of those through the pressure and the point at which it is applied. A panel that is thinner than intended has lost too much resin, which is a sign of pressure applied too early or of a press that is closing too fast.

The thickness also varies within a panel. A region with dense copper has less room for resin than a region with little copper, so the resin flows from the sparse area into the dense area and the dielectric is thinner over the copper. The variation is one of the reasons an impedance specification is difficult to hold on a board with uneven copper, and the countermeasure is the copper balance described elsewhere in the design rules. The measurement of the finished thickness is described under PCB dimensional stability and expansion.

Voids And Their Causes

A void is a region of the stack that the resin did not fill. It can come from a prepreg that did not flow enough, from a stack that was pressed unevenly, or from the copper pattern itself, where a deep pocket between two heavy traces traps air that the resin cannot displace. A void beneath an inner layer connection is a reliability problem, because it removes the mechanical support around the barrel.

The remedies are a longer or a hotter cycle, a slower press closure so that the resin has time to fill from the edge, and a design measure: avoiding deep, narrow pockets in the copper that the resin cannot reach. Voids are detected by microsection or by an ultrasonic scan of the panel, and the acceptance limit depends on their size and their position relative to a hole.

Controlling The Process

The press is instrumented, and the temperature and pressure are recorded against time for every cycle. That record is the primary evidence that the cycle met its specification, and it is the first thing examined when a lamination defect appears. The record also shows whether the press itself is behaving, because a slow thermocouple or a leaking ram appears as a deviation from the expected curve.

The other control is the incoming material. Prepreg is checked for resin content and flow, and the batch record links the material to the panel. Where a defect appears in a group of panels, the batch record narrows the search immediately to the material, the press cycle or the operator, rather than requiring an investigation of the whole process.

Measurement And Verification

The verification of a lamination cycle is partly dimensional and partly destructive. The thickness is measured on the finished panel, at defined points, and compared with the target. A microsection shows the bond line, the resin distribution and the presence of voids. A thermal stress test, in which a sample is floated on molten solder and then examined for delamination, shows whether the bond holds at temperature.

The peel strength of the copper to the laminate and the glass transition temperature of the cured resin are measured on a coupon for a demanding product. Together with the press record, these measurements form the evidence that the lamination is under control. The qualification programme that establishes the cycle for a new stack is described under multilayer prototype requirements.

Thickness control also depends on the tooling that defines the stack height. Where the press closes to a fixed stop, the thickness is set by the stop and the resin flow adjusts to it; where the press closes to a fixed force, the thickness is the result of the stack and of how much resin flows out. The two arrangements behave differently when a prepreg batch varies, and the arrangement should be recorded with the cycle so that a change in incoming material is interpreted correctly.

FAQ

Why does the panel come out thinner than intended? Usually because pressure was applied before the resin had softened, which squeezed resin out of the stack. Check the temperature at which the press closes and the rate of closure.

Can a thick stack use the same cycle as a thin one? No. The heat has further to travel, so the ramp has to be slower and the hold longer. Using a cycle that was developed for a thin stack on a thick one leaves the centre under-cured.

What causes a void under an inner layer pad? A resin that did not flow enough to fill the pocket around the pad, often because the copper pattern leaves a deep narrow gap. A slower press closure and a longer flow stage help, and the layout can avoid the geometry.