Avoiding the voids: plating voids are usually blamed on the metalization process, yet the causes are plentiful. (The Plating Rack).

THE GERMAN AUTHOR Kurt Tucholski defined a hole as:

"A hole is where nothing exists."

"A hole is a permanent companion on the 'non-hole.'"

"If man sees a hole he intends to fill it. While doing so, he often falls into it."

Certainly, the objective of the metalization operation is to put something into the holes; namely, a continuous, void-free and tightly adherent deposit of copper. However, despite good intentions, one does not always accomplish this. The result is voiding--those areas on the resin or glass, or at times both, that do not amass the copper deposit.

Mention "voids" in a PCB facility and watch everyone scramble. The problem with voids in plated-through holes is that with so many potential causes, it is difficult to quickly identify the source or sources of the problem.

Voiding is most often blamed on the metalization process. Certainly, this process is not blameless. However, since PTH metalization and desmearing are so interdependent, it is easy to see that the upstream processes can influence a defect that manifests itself later in the sequence. We will discuss this in detail later.

IPC-A-600 provides excellent guidelines with respect to voids. Basically, if your company is building PCBs to Class 3 specifications, any voiding is considered a nonconforming defect. This includes glass voids, resin voids, circumferential voids and so on. Voids come in all shapes and sizes. Some are obvious (FIGURE 1). Others are much more subtle (FIGURE 2). The ideal is a continuous, void-free deposit, as shown in FIGURE 3.

[FIGURES 1-3 OMITTED]

Say that voiding has been discovered and analysis reveals that it is confined to the glass. The resin coverage looks fine. Backlight testing confirms voiding through the electroless copper process. The voids are on multilayer PCBs made with a tetrafunctional resin. This resin system is used regularly in PCB shops, so there are no new wrinkles here. Where would troubleshooting start? Let's list the possible causes of glass voiding.

* Insufficient cleaning and conditioning of the holes.

* Drill debris in holes.

* Poor hole wall quality overall.

* Insufficient or ineffective desmear.

* Permanganate or manganate residues remaining on resin.

* Insufficient catalyzation (in electroless copper preplate).

* Insufficient acceleration (leaving excess tin and not allowing palladium to initiate the deposition reaction).

* Air/gas bubbles lodging in small diameter vias.

* Inadequate rinsing.

* Panels racked too close together.

* Rack agitation.

These are just a few potential causes of glass voiding. One of the first places to look is the cleaner/conditioner step. This step is critical to the success of the electroless copper process in that the conditioner enhances the adsorption of the activator species (palladium) to both the resin and glass. Inadequate conditioning will reduce the charge density on the glass and resin, resulting in less than optimal palladium (activator adsorption). This condition will lead to a lack of a continuous, void-free copper deposit in the hole.

Sufficient activator is necessary to catalyze the electroless copper reaction. Essentially, the cleaner/conditioner can be described as an activator promoter. If a complete examination shows the problem is minor glass voiding (no evidence of resin voids), then the first place to look is the cleaner/conditioner. Often the concentration of the cleaner/conditioner, or the operating temperature, is below that recommended by the process supplier. Another cause: the useful life of the made-up solution has been exceeded. In this case the "spent" chemistry must be sent to waste treatment and a new solution readied. If this doesn't solve the problem, there are other areas to check.

Another test to perform is a copper deposition rate measurement using the proper test coupon. Sometimes this test will reveal that the deposition rate of the copper is less than previous measurements. Let's say the process capability is designed to deposit 60-80 microinches of copper in the normal process time, yet the latest measurements yield a measurement of 50 microinches or less. This can be a cause of the glass voiding. We will discuss these other causes in subsequent columns.

MICHAEL CARANO is vice president of sales and marketing at Electrochemicals (Minneapolis, MN). He can be reached at [e-mail].