What Actually Determines PCB Quality
Ask five people what makes a good circuit board and you will get five answers, most of them about price or delivery. The properties that decide whether a board works in three years are measurable, and they are all set inside the fabrication process rather than at final inspection. Understanding them is what allows a buyer to judge a supplier on something other than the invoice.
Copper Thickness Control
The copper thickness printed on a drawing is a nominal figure with a tolerance, and the tolerance is what matters. A trace specified at one ounce that arrives at three quarters of an ounce carries substantially less current and has different impedance from the one the designer calculated.
Variation happens in two places. The foil itself arrives with a tolerance, and the plating process adds copper unevenly across the panel, so a trace in the centre and one at the edge can differ. A fabricator who measures thickness on coupons from each panel and adjusts the process accordingly is producing a different product from one who assumes the nominal value.
Drill Quality and Hole Wall Condition
A drilled hole is not a clean cylinder. The bit tears the resin, leaves debris in the barrel and generates heat that can smear resin across the inner layer copper. All of that has to be removed before plating.
Drill quality shows up in ways that are hard to see from outside: a smear that was not fully removed becomes a high resistance connection or an open circuit under thermal cycling, and a hole with a rough wall may plate thinly at the point of greatest roughness. Cross sectioning a sample hole is still the definitive check, and a fabricator who does it routinely has the data that a fabricator who does not simply lacks. The consequences for inner layer connection are described in our plasma desmear article.
Plating Uniformity
Plating thickness must be sufficient on the hole wall as well as on the surface, and the two are not the same. Current density is higher at the surface and at the outer edges of a panel, so the barrel of a deep hole is always the hardest place to plate adequately.
This is the constraint that limits aspect ratio, and it is why a thick board with small holes demands more process control than a thin one. Uniformity is verified on coupons and by cross section, and the figures should be available as part of the build record rather than only on request.
Solder Mask Adhesion and Registration
The mask protects the copper and defines where solder may go. Its adhesion depends on surface preparation before application, and its registration decides whether the pads are fully exposed or partially covered.
A mask web that is too narrow breaks down during assembly and allows solder to bridge between pads, which is a manufacturing defect that appears at the assembly stage rather than at fabrication. A mask that lifts after thermal cycling exposes copper to corrosion. Both are consequences of process control rather than of the mask material alone.
Surface Finish Consistency
The finish decides how the pads behave during soldering and how long the board can be stored. Thickness and composition both matter, and both vary with bath condition. Our notes on OSP surface finish describe the storage consequences of one common choice.
Soldering a sample pad and measuring the wetting behaviour is a practical check, but the process record is what proves the finish was applied within specification on the panel that actually shipped.
Documentation and Traceability
Quality is also a question of whether the supplier can tell you what they did. Coupon measurements, plating thickness, surface finish thickness, cross section photographs and the results of electrical test all form part of the record.
A supplier who can produce those figures for a specific panel is operating a controlled process. One who can produce a certificate but not the measurements behind it is offering paperwork rather than evidence, and the difference becomes visible the first time a fault has to be investigated. The verification methods involved are described in our PCB electrical test coverage article.
Why Price Alone Misleads
Two quotes for the same bare board can differ by a factor of two, and the difference is usually in the process steps that the buyer cannot see. Thinner plating, a cheaper finish, skipped cross sectioning and reduced inspection all lower the cost and none of them appear on the drawing.
Where the product is a prototype for a bench test, that difference may not matter. Where the board will be assembled into a product with a service life, the cheapest quote transfers the risk to the buyer without transferring the discount. Judging the supplier on measurable process evidence rather than on price alone is the only reliable way to choose.
How Quality Is Verified in Practice
Verification in a fabrication shop is built around coupons rather than around the product boards. Coupons are small structures on the panel border that carry the same copper, the same plating and the same stackup as the boards beside them, and they can be measured or destroyed without affecting the customer shipment. Plating thickness, dielectric thickness and impedance are all checked on coupons, and the result belongs to the panel, not to a general batch average.
Cross sectioning is the other half of the method. A sample hole is mounted, ground and polished until the barrel and its connection to the inner layers are visible under magnification, then photographed and measured. This is the only direct evidence that the plating is thick enough, that the barrel is connected and that no smear is left between the copper and the plating. A fabricator who sections samples routinely has a process that is being watched; one who sections only on request does not really know what the process is doing.
What a Buyer Can Realistically Check
Most buyers do not have the equipment to verify plating thickness, and it would be unreasonable to expect them to. What a buyer can do is ask for the measurements that already exist, and to notice whether the answer arrives as a number or as an assurance. A figure with a method beside it indicates a controlled process. A general statement of compliance indicates a certificate.
Visual checks have some value. Board dimensions, hole positions, the appearance of the finish and the condition of the solder mask can all be inspected with a ruler and a light. They will not reveal a thin barrel, but they do reveal the careless process that would also have produced one. Treating those visible signals as a proxy for the invisible ones is not rigorous, and it is better than ignoring the question entirely.
Cost, Quality and the Trade a Buyer Is Making
Cheaper boards are not made by working faster. They are made by reducing process steps, using thinner plating, skipping cross sectioning, using a less controlled finish bath or inspecting less. All of those reduce cost, and none of them appears on the drawing or on the certificate.
Understanding this turns a price comparison into a decision about risk. For a board that will be tested on a bench for a week, the cheapest supplier is a reasonable choice. For a board that will be installed in equipment with a service life, the saving is a transfer of risk to the buyer, and the transfer is not accompanied by a corresponding reduction in the buyer’s exposure. Our notes on component tolerance and reliability describe how those risks develop over the life of a product.
FAQ
Is a thicker copper layer always better? It carries more current and spreads more heat, but it makes fine features harder to etch and costs more. The right figure follows from the current and thermal requirement.
What is the most useful quality document to request? Coupon measurements and cross section results for the specific panel, along with the surface finish thickness. Those three cover most of the process risk.
Can quality be judged from the finished board? Only partly. Appearance and dimensions can be checked, but plating thickness and hole wall condition require the process record or destructive analysis.
