PCB Copper Thickness Measurement Methods and Techniques

Copper thickness is one of the few PCB characteristics that directly determines current capacity, impedance, and the fatigue life of a plated barrel, yet it is also one of the easiest to report carelessly. Two labs can measure the same panel and disagree by twenty percent if they use different methods or sample different locations. This guide explains what each technique actually measures and how to write a requirement that a fabricator can verify.

Why Copper Thickness Measurement Matters

Copper carries current, spreads heat, and holds the mechanical load of every thermal cycle. If the surface foil is thinner than specified, trace resistance rises and a design that passed its current calculation in theory can run hotter than expected. If the hole wall is thin, the barrel may survive assembly but crack after a few hundred temperature excursions in the field.

The measurement also feeds commercial decisions. A board quoted as two ounces of copper but delivered at 1.6 ounces has been, in effect, substituted with a cheaper construction. Because the difference is invisible to visual inspection, the only protection is a documented measurement plan with defined methods, locations, and limits that both parties accept before the order is placed.

Standards, Units and Acceptance Limits

The industry states copper thickness in micrometres, in mils, or as copper weight in ounces per square foot, where one ounce corresponds to roughly 35 micrometres. Standards such as IPC-6012 define minimum requirements for both surface copper and plating in holes, with different limits for Classes 1, 2 and 3. Those minima are floors, not targets, and they apply at specified locations rather than as an average.

Designers should translate their electrical need into a specification that a laboratory can measure. Asking for minimum 25 micrometres in the barrel and minimum 35 micrometres on the surface is workable; asking for nominal 35 micrometres everywhere is not, because plating naturally varies with geometry, current distribution, and hole position on the panel.

Microsection: The Reference Method

A microsection is a physical cross-section. The sample is encapsulated in resin, ground and polished until the plane of interest is exposed, then examined under a calibrated microscope at 100 to 500 times magnification. Because the operator sees the actual copper, the technique reveals barrel thickness, corner thinning, resin smear, and inner-layer separation in a single view.

Preparation dominates the result. Aggressive grinding smears copper over the surface or rounds the edges of the barrel, and polishing with the wrong compound can etch the copper boundary so it appears thinner than it is. Good labs use a controlled sequence, measure at defined clock positions, and report the minimum rather than a flattering average. The method is destructive, so it is applied to coupons or to a scrap area.

Non-Destructive Eddy Current Gauging

An eddy current gauge induces a small current in the copper and reads the resulting impedance change to infer thickness. It is fast, non-destructive, and ideal for production screening on a coupon or a copper-clad test area. The catch is that it measures total conductive metal within its field, so it can be confused by nearby traces, ground planes, and finishes such as nickel or gold.

Calibration is everything. The instrument must be zeroed on a sample of the same alloy and finish as the production board, ideally with a certified reference of similar thickness. Expect readings to be less reliable over very thin foils, over dense circuitry, and wherever the probe cannot sit flat. Use the gauge for trend monitoring and the microsection for proof.

XRF and Other Surface Techniques

X-ray fluorescence identifies elements and estimates coating thickness by the fluorescence spectrum, and it is widely used to check finishes such as nickel, gold, and tin. When applied to copper it can indicate foil thickness on a bare laminate area, but the measurement volume and the surrounding material influence the result, so it is better suited to process control than to arbitration.

Other useful techniques include four-point probe resistance measurement, which is quick and repeatable on a flat coupon, and cross-sectioning combined with scanning electron microscopy when a failure needs detailed investigation. Whichever tool is used, the report should state the instrument, the calibration reference, the measurement location, and the uncertainty so that the number can be interpreted correctly.

Coupon Design and Where to Cut

Because microsectioning destroys what it measures, production boards are qualified through coupons attached to the panel. A well-designed coupon places test traces, a plated hole of the smallest diameter in the design, and a surface copper area where the probe can sit flat. Panel position matters too: edges plate differently from centres, and a coupon tucked in a corner will not represent the middle of the array.

Many fabricators place coupons on both sides of the panel so that current distribution can be judged, and some customers ask for coupons per panel lot rather than per order. The coupon must be processed through the same drilling and plating sequence as the product, which is why a coupon taken from a different panel batch proves very little about the boards you receive.

Surface Copper Versus Plated Hole Wall Copper

Surface copper and barrel copper come from different process steps and rarely match. Surface foil begins as the laminate’s copper, later thickened by pattern plating or covered by a solder mask; the hole wall is deposited by electroless seeding followed by electrolytic plating, and its thickness varies with hole diameter, depth, and position. A single number cannot describe both.

Reports should therefore separate the two and quote limits for each, with the minimum barrel thickness measured at the thinnest point, usually the bottom corner or the mid-barrel. When a customer specification mentions copper thickness without qualification, ask which surface is intended. Confusion here is a frequent cause of disputes that could have been avoided with one extra line in the purchase order.

Measurement Pitfalls: Smear, Etch Factor and Bias

Resin smear drags dielectric across the copper surface during drilling; if it is not removed before plating, the copper may bond poorly even while measuring acceptably thick. Etch factor works in the opposite direction, thinning the top of a trace so that a cross-section shows a trapezoid rather than a rectangle. Both effects mean the number on the report may not describe the copper that carries current.

Operator bias is subtler. Measuring where the barrel looks thickest, choosing the magnification that flatters the edge, or reporting an average instead of a minimum all produce a passing sheet on a marginal board. Counter it with a written procedure that fixes the clock positions, the magnification, and the statistic reported, then audit the results against photographs.

Sampling Plans, Records and Reporting

Sampling should be tied to lot size and to the criticality of the product. A prototype may need one coupon for information, while a medical or automotive order may require coupons per panel and per plating bath. Define how many, from where, and at what frequency before production starts, and write the plan into the plating thickness requirements so that it is quoted rather than negotiated after delivery.

Records complete the loop. Each report should identify the date, the panel or lot, the method, the calibration reference, the measurement locations, and the minimum and average values. Photographs of the sectioned barrel make the data verifiable years later if a field failure occurs. Storing results in a simple trend chart turns a series of separate inspections into a picture of process stability.

FAQ

Which method should I specify for copper thickness? Use microsectioning as the reference for dispute resolution and qualification, and eddy current gauging for routine production screening. Add XRF or four-point probing when you need fast surface data. Name the method in the specification, because a requirement without a method is an invitation to measure whichever way is most convenient.

Is one ounce of copper exactly 35 micrometres? It is approximately that value, and the exact figure depends on the density assumed in the conversion. More importantly, delivered foil has a tolerance, and plating adds thickness unevenly across the panel. Specify a minimum in micrometres at defined locations rather than relying on the ounce rating alone.

Why do my coupon readings disagree with the fabricator’s? Different methods, different locations, and different statistics usually explain the gap. Confirm that both parties measure the same feature, at the same clock positions, with the same magnification and the same rule about reporting the minimum rather than an average. Sharing a reference sample resolves most disagreements quickly.