Pink Ring and Wedge Voids: 6 Causes in Drilled PCB Holes

Pink ring is the halo that appears around a drilled hole in a microsection where the resin has separated from the inner layer copper, leaving an area that etches differently from the surrounding laminate. It is named for the colour seen under the microscope, and it is a bond failure rather than a stain.

The defect matters because the same separation that produces the halo also weakens the connection the drill has to make. A board with extensive pink ring can pass electrical test and then fail in thermal cycling, because the copper to laminate interface that should resist the stress has already opened.

What Pink Ring Looks Like and What It Means

In a polished section, the affected area is a lighter ring between the hole wall and the undisturbed laminate, following the circumference of the hole at the inner layer that was drilled through. Under magnification the resin has pulled back from the copper, and the gap exposes material that polishes and etches differently.

Three severities are usually described: a ring confined to one layer, a ring that extends into adjacent layers and a ring that reaches the hole wall. Only the last is normally a rejection, which is why the extent of the separation is measured rather than its presence alone. A ring limited to a single layer is usually treated as acceptable, while a ring that reaches the hole wall is a reject, and that limit belongs in the shop specification rather than in the judgement of the inspector.

Wedge Voids and the Difference Between Them

A wedge void is a gap in the same general location that does not extend around the full circumference. It appears as a triangular or crescent shaped separation between the copper and the resin, usually on one side of the hole, and it is often associated with the drill entry or exit direction.

The two share most of their causes, so a process change that removes one usually improves the other. The distinction matters for diagnosis: a ring that follows the whole circumference points at the bonding chemistry, while a local wedge points at a mechanical or thermal effect during drilling.

<img src="https://www.gopcba.com/wp-content/uploads/2026/06/以太网交换机PCBA.jpg" alt="Wedge void visible at the inner layer copper after PCB drilling and desmear” />

Why the Resin Separates From Copper

The bond at the inner layer is mechanical. The resin flows into the oxide structure on the copper during lamination and locks onto it when it cures. Anything that weakens that interlock, reduces the resin flow or damages the interface before or during drilling can produce a separation.

Stress then opens the weak point. Drilling generates heat and mechanical load at the hole wall, and the laminate expands faster than the copper. Where the interface is marginal, that difference is enough to pull the resin away, and the separation shows up as pink ring in the section.

Drilling Heat and Resin Damage

Heat is the most direct cause. A dull bit, too high a chip load, too many hits on one bit or a stack that is too tall all raise the temperature at the cutting edge. Once the resin reaches its glass transition it softens, and the softened material is easily displaced from the copper. Bit wear is usually the culprit, because a worn edge rubs instead of cutting and turns spindle power into heat at the hole wall, exactly where the laminate can least afford it.

Smear then follows. Softened resin is smeared across the hole wall and the inner layer connection, and the desmear step has to remove it before plating. Where the smear is heavy, desmear may remove the resin that has already separated, leaving a void rather than a bonded joint.

Oxide and Bonding Chemistry

The bonding treatment decides how much grip the resin has. A thin, uneven or contaminated oxide leaves areas where the interlock is weak, and those areas are where the separation starts. The oxide treatment and its rinse therefore belong in the investigation of any pink ring problem.

Residue is the second factor. Manganese from permanganate desmear, oil from handling and moisture absorbed by a core waiting too long before lay-up all reduce the bond. Each leaves the surface chemically changed in a way that a visual check cannot detect, which is why coupon and weight gain data are used instead.

Desmear and Rinse Effects

Desmear removes resin smear from the hole wall and prepares the copper for plating. If it is too aggressive it can remove resin that was still bonded, opening a gap that then plates over; if it is too weak it leaves smear that insulates the inner layer from the plating.

Rinse quality decides whether the desmear chemistry stops at the right moment. Carried-over permanganate continues to act on the resin after the tank, and neutraliser that is not fully rinsed leaves residue that shows up later as a plating adhesion fault around the same circumference.

Lamination Pressure and Resin Flow

Lamination has to push resin into the oxide texture and hold it there while it cures. Insufficient pressure or a short cycle leaves the resin short of full contact, while excessive pressure drives resin out of the bond line and leaves a resin-starved interface that separates easily.

Resin flow depends on the prepreg as well. A batch with lower flow than the qualification batch will not fill the same texture, and the difference is usually invisible until a section is cut. Where a supplier changes resin content or flow, the lamination parameters should be re-qualified rather than carried over. Vacuum in the press matters as well, since trapped air creates local pockets where resin cannot reach the oxide texture however hard the press pushes.

Detection: Microsection and Dye and Pry

Pink ring is confirmed by a microsection cut through the hole and examined for separation at each inner layer. Preparation quality matters here, because a poorly polished sample produces artefacts that look exactly like the defect being investigated.

Dye and pry, used more often on assembled boards, forces dye into any gap and then mechanically separates the joint to reveal where it penetrated. It shows whether a separation is connected to the surface, which a section alone cannot always establish. It is also the practical method for a finished assembly, because a section can only be taken where the board can be cut, and cutting a populated board usually destroys the evidence around the joint.

Fixing the Process in the Right Order

Start with the drill. Bit condition, hits per bit, chip load, stack height and entry and backup material set the temperature that the resin sees, and they are the cheapest variables to correct. Confirm the change with sections from the worst position on the panel rather than from a convenient one.

Then review the chemistry and the lamination. Bonding treatment, desmear control and rinse quality come next, followed by lamination pressure and resin flow. Because the causes stack, changing several at once makes it impossible to attribute the improvement, and the defect tends to return when the undocumented change is reversed.

FAQ

Is pink ring always a reject? No. A ring confined to a single layer is often accepted, while a ring that reaches the hole wall is normally a rejection because it can affect plating adhesion. The acceptance limits belong in the shop specification and should match the customer requirement.

Can pink ring be detected without cutting a section? Not reliably. The defect is internal, so it appears only in a section or through a destructive test such as dye and pry. Process data, including drill life and desmear control, is what prevents it rather than inspection catching it.

Does a different laminate help? Sometimes. Resins with better adhesion to copper and higher toughness resist separation more effectively, but a change of material is expensive and only makes sense after the drilling and bonding processes have been verified.