Common PCB Design Mistakes and How to Avoid Them
Most board problems that reach production are not exotic. They are the same short list of mistakes, repeated across companies and across products, and almost all of them are avoidable with a review process that looks for them deliberately. A layout that has never been checked against a checklist is a layout that contains at least one of the following.
The mistakes fall into a few groups: things that cannot be manufactured, things that cannot be assembled, things that cannot be tested, and things that work on the bench but not in the product. Each group has its own set of symptoms, and each is cheaper to fix before the artwork is released than after the first boards arrive.
Footprints That Do Not Match the Component
The single most common cause of a board that will not assemble is a footprint that does not match the datasheet. Pad lengths that are too short produce joints without a fillet, pad widths that are too narrow leave the component floating, and a courtyard that is too small makes placement impossible next to a neighbour. The error is usually inherited from a library created years earlier and copied from project to project without ever being checked.
A second version of the same mistake is a footprint that matches the package but not the pin numbering or the orientation convention. Pin one placed on the wrong corner, a mirrored footprint for a bottom mounted part, or a numbering scheme that differs from the schematic symbol all produce a board that has to be reworked or scrapped. Reviewing a new footprint, pad by pad, against the datasheet and against the same pad design standards the library already follows is the only reliable prevention.
Clearance, Annular Ring and Drill Rules
Copper to copper clearance below the fabricator capability, annular rings that are too small for the drill tolerance, and via drill sizes outside the standard tool list all raise the price or cause a rejection at the panel level. These are rarely deliberate decisions; they appear when a design is tightened to fit a dense area and the fabricator rules were not loaded into the tool. The fix is to load the rules before the layout begins rather than to check them afterwards.
The same applies to solder mask and silkscreen. A mask dam narrower than the process can hold will not survive, so adjacent pads lose their separation and the assembly suffers bridging. Silkscreen printed over a pad interferes with soldering and is usually removed by the fabricator, which leaves the board without the marking the assembly house expected. Both are checked in a design rule review against the manufacturable design guidelines, and both are invisible to a visual inspection of a plot at normal zoom.
Thermal and Current Problems
A trace that is adequate for the nominal current may not be adequate for the inrush current of the load it feeds, and a copper pour connected to a small pad without thermal relief is difficult to solder. Copper area is often added for electrical reasons and then forgotten in the thermal calculation, so a board that works on the bench fails in an enclosure at elevated ambient.
The related mistake is placing a heat generating component where its heat has nowhere to go. A regulator placed in the middle of a dense area with no copper to spread into will run hotter than the same part placed near a plane or a thermal via. Thermal vias under a power device, connected to an internal or bottom layer plane, are a cheap and effective measure, and their absence is one of the most frequent findings in a design review. Checking the current path and the heat path together is the practical approach, and the calculation methods are set out in the guidance on trace width and current.
Return Paths and Reference Planes
A high speed net that crosses a split in its reference plane has no continuous return path, and the resulting loop radiates and couples into neighbouring traces. The error is easy to make because the split is often on an internal layer that is not visible while routing the outer layers, and it is very hard to find afterwards without a check that looks for it specifically.
The related problem is a connector or a memory interface where the return path is routed as a signal. Return currents need as much attention as the signals they accompany, and a single ground pin in a connector that carries a high speed bus is a common finding. Assigning ground pins deliberately, and routing the reference plane under the whole bus, removes the problem before it exists. A dedicated check for plane splits under critical nets catches most of what remains.
Testability and Assembly Access
Test points that are too close together, or that are placed under a component, make in circuit testing impossible. A board whose only test access is a connector forces a functional test instead of a structural one, which is slower and less diagnostic. Designing the test access at layout time, with the test engineer involved, is much cheaper than adding it later.
Assembly access is the same problem seen from a different angle. Components placed too close to a connector can prevent the mating part from seating. Tall parts next to a low profile area interfere with a nozzle or a shield can. A fiducial placed where the machine cannot see it because a component shadows it defeats the purpose of having one. These issues are all visible in a three dimensional review, which is why a mechanical check with the enclosure and the mating parts is worth the time it takes.
Documentation and Handover
The final group of mistakes is not in the board at all. A stack-up that is not documented, an impedance target that is stated as fifty ohms without saying on which layer, a drill table that does not match the artwork, and an assembly drawing that omits the polarity of a polarised part all cause problems downstream. The fabrication house will guess, and the guess may be reasonable but not what was intended.
Producing a short package of documentation with the layout, including the stack-up, the impedance requirements, the drill table, the surface finish and the assembly drawing, takes an hour and prevents most of these errors. It also makes the design review faster, because the reviewer can check the intentions rather than reverse engineering them from the artwork. A board that ships with complete documentation is one that a second engineer can modify safely, which is a benefit that appears long after the project is finished.
Additional Considerations for This Build
Practical attention to PCB design mistakes pays for itself here, because it decides whether the finished board behaves as the drawing intended. Where the requirement is not stated on the fabrication drawing or in the assembly notes, the shop has to assume a default, and that default is rarely the value the design was simulated with. Stating PCB design mistakes explicitly, together with the tolerance that applies, removes the assumption and keeps the result predictable from batch to batch.
Deliberate attention to footprint review pays for itself here, because it decides whether the finished board behaves as the drawing intended. Where the requirement is not stated on the fabrication drawing or in the assembly notes, the shop has to assume a default, and that default is rarely the value the design was simulated with. Stating footprint review explicitly, together with the tolerance that applies, removes the assumption and keeps the result predictable from batch to batch.
FAQ
What is the most common mistake overall? A footprint that does not match the component datasheet. It is also the cheapest to prevent, since the check takes a few minutes per part.
Can a design rule check catch everything? No. It catches geometry that violates the rules, but it cannot see that the wrong rule was used or that the topology is wrong. A review with a person who understands the circuit is still required.
How much review is enough? Enough to cover the power path, the return paths, the fine pitch components, the test access and the documentation. Those five areas account for the large majority of problems that reach production.
