Conformal Coating Thickness Measurement: Methods, Limits and Sampling
A conformal coating specification is written as a thickness with a tolerance, and the only way to know whether a board complies is to measure the coating thickness at points that represent the worst case. Measuring where the coating is thickest proves nothing, and that is exactly what happens when the sampling plan is not defined.
What the Coating Specification Actually Requires
A typical specification calls for 25 to 75 microns of cured coating over the assembled surface, with a minimum that applies to every point on the board rather than to the average. Some products state a nominal figure with a tolerance, others state a minimum only, and a few state different figures for different areas such as over a solder joint, over a component body or on a connector keying feature.
Reading the specification correctly is the first step, because the three common forms lead to different sampling plans. A minimum only specification is verified by measuring the thinnest point, which is usually at a sharp edge, in a corner or beside a tall component where the spray shadows the surface. A nominal with a tolerance has to be verified at both ends, which means measuring the thickest point as well, typically where the spray concentrates on a flat area. The specified thickness applies to the cured film, not to the film as applied. Most coatings lose a significant fraction of their thickness during cure as solvent leaves, so a wet film measurement of 120 microns may correspond to 50 microns cured. State which measurement the specification refers to before designing the inspection.
The Three Measurement Methods and Where Each Works
Three techniques are used in production. Magnetic or eddy current gauges measure the lift off of a probe from a conductive or ferrous substrate and are fast and non destructive, but they read only over a conductive surface and are affected by the substrate material. Ultrasonic gauges measure the time of flight of a pulse through the film and work over most substrates, though they need a couplant and a flat enough surface. Optical methods use a microscope or an interferometer to measure a step or a cross section, and they are the reference when the other two disagree.
None of the three is universally correct. An eddy current gauge calibrated on a copper coupon reads high over a solder joint and low over a gold plated pad, because the calibration assumed a specific conductivity. An ultrasonic gauge needs a minimum film thickness to resolve the echo, and it struggles under about 20 microns. An optical step measurement needs an exposed edge of coating, which means it is normally done on a coupon or a scrap board rather than on a production unit. Use a fast method for routine control and verify it against a reference method at intervals, re establishing the correlation whenever the coating, the substrate or the gauge changes. For most conformal coating work, an eddy current gauge over the copper areas and an optical step measurement on a coupon cover the requirement. Record which method was used at each point.
Eddy Current and Magnetic Methods on Coated Boards
An eddy current gauge works by measuring the impedance change produced in a coil by the conductive substrate below the film. The reading depends on the film thickness and on the conductivity of the substrate, so the gauge has to be calibrated against foils of known thickness placed on the same substrate. Calibrating on a steel block and then measuring over copper introduces an error that can exceed the tolerance of the specification.
Probe size sets the spatial resolution. A small probe reads a small area and can measure between fine pitch components, but it is more sensitive to curvature and to the edge of the substrate. A large probe averages over a wider area and is more repeatable on flat surfaces. Keep both sizes and state in the procedure which is used where, because a reading taken with the wrong probe can differ enough to change a pass into a fail. The method is also sensitive to the film itself if the coating is filled. A heavily loaded coating with a high filler content changes the effective dielectric properties in the gap and shifts the reading, so the calibration has to be made with the same coating that will be measured. Verify the gauge against a reference foil of known thickness before each measurement session and record the deviation.
Ultrasonic and Optical Methods for Thin Films
Ultrasonic measurement sends a pulse through the film and reads the reflection from the coating to substrate interface. It works over any substrate and it can measure on a component body, which is a common requirement where the specification applies to the whole assembly. The limitations are the minimum thickness, the need for a couplant that does not attack the coating, and sensitivity to surface roughness.
Optical step measurement is the reference method and it is destructive by nature. A coupon or a scrap board is sectioned or masked during coating and the step measured under a microscope with a calibrated scale, or the coating is cut and the edge viewed under magnification. Because it exposes a true cross section it also shows whether the coating has flowed around an edge or bridged between two close features, which the non destructive methods cannot see. Use the optical method to qualify a process and to settle disputes, not for routine production control. A sectioned coupon from each coating batch is enough to establish that the process is producing the intended thickness and coverage, and the non destructive readings taken on production units are then interpreted against that evidence.
Wet Film Measurement During Application
Wet film measurements are made immediately after application with a comb gauge that has teeth of graduated depth, or with a wet film wheel. The reading is taken while the coating is still liquid, and it converts to a cured thickness by the solids content of the coating. A coating with 40 percent solids applied at 100 microns wet produces roughly 40 microns dry, though the relationship is not exactly linear because of solvent retention and shrinkage.
The value of a wet film measurement is that it catches an application problem before the boards are cured, when the coating can still be adjusted or removed. Its limitation is that the measurement is taken on a flat area where the coating is easy to comb, which is rarely the thinnest point on the board. Use it to confirm that the spray or the dip is delivering the intended amount, and use the cured measurement to confirm the coverage over the whole assembly. Wet film gauges have to be cleaned immediately after use and checked for wear. A comb with a bent tooth reads a thickness that does not exist, and the error is systematic, so every board coated that shift inherits it.
Sampling: Where to Measure and How Many Points
Sampling has to target the thinnest point rather than a convenient one. The thinnest coating is normally found at a sharp edge, in a corner or recess, in the shadow of a tall component relative to the spray direction, and on a surface that was not directly exposed to the spray. Identify those points from the board layout and the spray path, and make them the mandatory measurement points.
The number of points follows from the tolerance and from the process capability. Where the process is well centred with a narrow distribution, five points per board on a sample of boards per batch is adequate; where the process is at the edge of the window, more points are needed to detect a shift. The points should be marked on a drawing that is part of the inspection instruction, so that the same locations are measured every time and the trend is meaningful. Include at least one point over a solder joint and one over a component body, since the specification often applies to both and the coating behaves differently on each. A tin lead or lead free joint is a rough, oxidised surface that the coating wets differently from the laminate, and the thickness over it is usually lower.
Measuring Over Components, Edges and Tall Features
Coating flows and levels, so the thickness at the top of a component body is usually much lower than the thickness on the board surface beside it. Where the specification applies to the component surface, measure there, and expect to find readings at the bottom of the allowable range. Where the specification applies only to the board surface, the component reading is informative but not the acceptance criterion.
Edges are the other difficult case. Coating thins over a convex edge because surface tension pulls material away, and it thickens in a concave corner where material collects. A gauge probe placed over an edge reads an average that describes neither. Measure a fixed distance from the edge, typically at least 1 mm, and state that distance in the procedure so that two inspectors measure the same thing. Tall features cast shadows. A spray applied at a fixed angle will deposit less behind a connector, a shield or a large capacitor, which is why the coverage inspection and the thickness measurement should be read together. A thin reading in a shadowed area is evidence that the spray path or the board orientation needs changing, not that the coating material is at fault.
Cure and Its Effect on the Reading
Cured thickness is the specified quantity, so measurements taken before cure is complete are only indicative. A coating that is still releasing solvent is thicker than it will be, and the same point measured again after full cure will read lower. Measure after the cure specified by the coating manufacturer, and record the cure schedule with the measurement so that a change in schedule can be distinguished from a change in application.
Some coatings continue to change for days after the initial cure, particularly those that cure by moisture or by a secondary reaction. Where the specification is tight, measure after the full cure period rather than after the oven step, and if the product is measured at both times, record both figures so that the difference is understood. Cure also affects adhesion, and a thickness that passes while the coating is under cured will not stay compliant in service. Read the thickness result together with the coverage inspection and the coating coverage evidence, because a coating that is thick but poorly adhered to the surface below it is not protecting anything.
Records, Reference Standards and Disputes
Keep reference foils or shims of known thickness with each gauge, and verify the gauge against them at the start of every session and after any probe change. Record the verification result, the substrate used for calibration, the probe size, the measurement points and the operator. When two parties disagree about a thickness, these records decide which reading is credible.
Because different methods give different numbers, state the method in the report as well as the result. A 30 micron reading by eddy current and a 38 micron reading by cross section on the same board are not a contradiction; they are two methods with different sensitivities and different sampling volumes. Where the specification is contractual, agree the method and the sampling points in writing before the first delivery, drawing on the same conformal coating practice that governs the application.
FAQ
What thickness should a conformal coating be? Most specifications call for 25 to 75 microns of cured coating over the assembly, with a minimum that applies to every point. Read whether the requirement is a minimum only or a nominal with a tolerance, because that determines whether you sample for the thinnest point or for both extremes.
Can an eddy current gauge measure coating over a solder joint? It can, but the reading depends on the conductivity of the metal underneath, so a gauge calibrated on copper reads differently over a solder alloy and over gold. Calibrate on the same substrate you will measure, and verify against a reference foil at the start of each session.
Why does the coating thickness read lower after curing? Because the film loses solvent and shrinks during cure. A wet film measurement converts to a cured thickness only through the solids content, and even then the relationship is approximate. Measure the cured film for acceptance, and record the cure schedule with the result.
