Copper Thickness Target: 4 Checks Before Plating Starts

A copper thickness target is the number the plating line is asked to hit, and it is rarely the number the designer had in mind. What is specified is usually a minimum, what is measured is an average, and what the bath produces is a distribution.

Confusing those three is the source of most arguments about copper thickness. A panel can meet the minimum in the specification and still be rejected by a measurement that was taken in the wrong place or compared with the wrong figure.

What the Target Has to Represent

The target is the value the process aims for, and it has to be set so that the whole distribution stays inside the specification. Aiming at the minimum is not a target; it is a guarantee of failure somewhere on the panel.

That means the nominal figure has to be the minimum plus the spread of the process plus a margin for the variation between measurement methods. Where the spread is wide, the aim point has to move up.

The aim point should be reviewed whenever the plating window, the panel size or the product mix changes, because the distribution moves with all of them.

Minimum, Nominal and the Tolerance Stack

Minimum thickness is a requirement, expressed as a value that no point on the panel may fall below. It is a worst-case statement, and it says nothing about where the process should be centred.

Nominal thickness is what the drawing shows for a finished trace or plane, and it usually includes the foil that the laminate arrived with. It is the number used for impedance and current calculations.

The tolerance is the band around the nominal, and it has to accommodate both the plating distribution and the measurement uncertainty. Where the band is tight, the plating process has to be tighter still.

Foil Plus Plating

Finished copper is the sum of the base foil and the copper added by plating, and both contributions carry their own variation. The foil has its own tolerance, and a panel that starts thin has less room for plating error than one that starts thick.

For that reason the foil specification belongs in the thickness calculation rather than in the purchasing file alone. A change of foil supplier can shift the whole distribution even when the plating bath is unchanged.

Where the panel is pattern plated, the copper in the holes and the copper on the surface come from the same bath but from different current densities, so the two targets have to be set separately.

Where the Target Is Hardest to Hold

The hardest points are the ones with the least current: the centre of a large panel, the middle of a sparse pattern and the barrel of a deep, small hole. Those are the points where the minimum has to be met.

The easiest points are the edges, the isolated pads and the areas around the contacts. Setting a target on those measurements flatters the process without proving anything about the panel.

Mapping the panel before the target is set is therefore worth the effort, and the coupon practice behind it is described in copper thickness coupons. The thinnest point is usually found at the centre of the panel and in the deepest, smallest holes, and those are the points the aim point has to cover.

Class Requirements and Customer Calls

Standards define classes of copper thickness for different products, and those classes are minimum figures rather than nominal ones. A product qualified to a class has to meet it at the thinnest point, not on average.

Customer requirements can be tighter than the class, and they often specify the measurement method as well. Where the method is specified, the target has to be set for that method rather than for a method the shop prefers.

Where the requirement is unclear, the question is worth asking before the first panel is plated. A target set on an assumption is discovered at final inspection, when the copper can no longer be changed.

Measurement Method and Its Bias

Different methods give different numbers on the same panel. X-ray fluorescence measures the total copper it sees, which includes the foil, while a microsection measures the plated layer in cross-section at the point where it was cut.

Each method has a bias, and the bias has to be known before the two are compared. Treating an XRF reading and a section measurement as interchangeable will produce a permanent disagreement between the line and the laboratory.

The measurement point matters as much as the method. A reading taken at the panel edge and one taken at the centre describe different parts of the distribution, and the earlier discussion of coupon placement applies directly.

Setting Targets for a New Product

A new product starts from the minimum the design requires, plus the process spread measured on a similar product, plus the measurement uncertainty. That gives an aim point that can be reviewed after the first article.

The first article is where the assumption is tested. If the distribution is wider than expected, the aim point moves or the process is tightened, and the decision should be made with data rather than with opinion.

Subsequent lots confirm the aim point rather than re-establish it. A target that is adjusted on every lot is not a target but a reaction.

Symptoms of a Wrong Target

A target set too low produces panels that fail on minimum thickness at the centre of the panel while looking healthy at the edge. Because the failure is local, it is often missed by a measurement taken in the wrong place.

A target set too high wastes copper, adds plating time and increases the risk of burning at high-current areas. It also loads the bath more quickly and increases the amount of copper that has to be etched away.

A target set without regard to the measurement method produces a permanent argument between two sets of correct numbers. The same target can look comfortable on one instrument and marginal on another, and both are reporting honestly.

Records and Review

The record should carry the aim point, the minimum it was derived from, the method used and the measured distribution for the lot. That is what allows a target to be reviewed on evidence rather than on impression.

The plating parameters that deliver the target are described in plating current density control, and the distribution that has to be held inside the band is covered in plating thickness distribution.

Where a product runs for years, the aim point should be reviewed at intervals rather than left as it was first set, and reference methods are published by IPC.

FAQ

Should the target be the minimum on the drawing? No. The minimum is a limit. The target has to be high enough that the whole panel distribution stays above that limit.

Why do two measurements of the same panel disagree? Because methods measure different things. XRF can include the foil, while a section measures the plated copper at one point, so each has its own bias.

How often should an aim point be changed? Rarely, and only on evidence. A target that is adjusted from lot to lot removes the reference that makes a trend visible.