Humidity and Condensation Testing of Assemblies

An assembly that passes every electrical test in a dry laboratory can fail in a humid climate, and the failure will be attributed to the environment rather than to the design. Humidity testing exists to make that failure appear in the laboratory, where it can be analysed.

What Humidity Does to an Assembly

Water vapour in the air is absorbed by the laminate, the soldermask and the residues on the surface. The absorbed water increases the dielectric loss, lowers the insulation resistance and allows ionic species to move.

Where the surface is contaminated, the water dissolves the contamination and forms an electrolyte. With a voltage present, that electrolyte supports electrochemical migration between adjacent conductors.

The third effect is corrosion of the metal surfaces, which proceeds slowly at ambient and quickly when a combination of humidity, contamination and bias exists. The metals most at risk are those in the finish and in the component terminations.

Condensation and the Dew Point

The damage from condensation is different from the damage from high humidity, because liquid water on the surface is a much better conductor and a much better solvent than vapour.

Condensation occurs when a surface is colder than the dew point of the surrounding air, which happens when a product is brought from a cold environment into a warm one or when a surface cools at night.

The thermal mass of the assembly determines how long it stays below the dew point. A heavy assembly with a metal enclosure is more at risk than a light board, because it takes longer to warm up. Our industrial assembly notes describe the environments where this occurs in practice.

Test Types

A steady state damp heat test holds a constant temperature and humidity, typically forty degrees and ninety three per cent relative humidity, for a defined period. It assesses absorption and corrosion rather than condensation.

A cyclic damp heat test varies the temperature to force condensation and to drive the moisture in and out of the materials. It is more severe than the steady state test and closer to a real environment.

A highly accelerated stress test uses a high humidity at a temperature above the boiling point of water in a pressurised chamber. It finds defects quickly and its results cannot be translated directly into service life.

Measurements That Matter

Insulation resistance is measured during the test with a bias applied, and a falling resistance indicates that a conductive path is forming. Continuous monitoring is more informative than a measurement at the end, because it shows when the change began.

The measurement must distinguish a genuine degradation from the normal drop caused by moisture absorption. The two look similar at a single point in time and differ in their recovery when the assembly is dried.

Functional testing before and after shows whether the assembly still works, and the difference between the two states is what the product will experience in service.

Recovery and Its Interpretation

A critical part of the test is what happens afterwards. An assembly that recovers completely when dried has an absorbed moisture problem, which may be acceptable in a product that runs warm and dry.

An assembly that does not recover has a permanent change, such as corrosion or a permanent conductive path, which is a failure regardless of the final measurement.

The recovery measurement should be taken at a defined time after the test, because the value changes as the assembly dries. Reporting only the measurement at the end of the test loses this information. Our quality notes describe how the results are recorded with the lot.

The Role of Cleaning and Coating

Residue is the fuel for electrochemical migration, so cleaning before the test removes the failure mechanism rather than the moisture. An assembly that fails a humidity test is often contaminated rather than exposed.

Conformal coating protects the surface, so the humidity test on a coated assembly measures the quality of the coating as much as the assembly. A coating with a pin hole or a missing area is where the failure appears.

Our coating notes describe the coverage requirements and the verification methods, which are what makes the coated assembly’s result meaningful.

Design Measures That Change the Outcome

Increasing the spacing between conductors reduces the electric field that drives migration and lengthens the path that a filament must grow along. This is the same principle as the creepage rules and it applies to the same features.

Removing exposed copper between conductors, covering the surface with a soldermask that adheres properly, and avoiding unnecessary via holes in the sensitive area all reduce the risk.

The bias matters as well, and a sensitive node that can be left unbiased while the product is dormant has a lower risk than one that is always powered. Our solder defects notes describe the assembly conditions that make a humidity failure more likely.

Designing the Test

The test should represent the environment the product will see, in temperature, humidity, duration, bias and power state. A test that the product does not resemble provides a number that cannot be used.

The sample should include the worst case assembly configuration, including the coating and the cleaning state that production will deliver. Testing a clean, uncoated board and shipping a coated one tests a different product.

The result should be reported with the conditions and the sample configuration, and the failure, when it occurs, should be analysed to identify the mechanism rather than only recording that the resistance fell.

Process Control and Verification

On a design of this kind, bias is the item that decides how the rest of the board is arranged. Running a first article through the same checks as the production panel confirms that the two agree, and that comparison is the cheapest form of process control available at prototype stage. Keeping a sample from the panel turns a dispute into a measurement, because the same coupon can be re-examined by both parties without rebuilding the batch.

Reviewing the design before the data is released is cheaper than correcting it after the panel is in the tank, because every step downstream inherits the decision made at the front end.

Process Control and Verification

On a design of this kind, bias is the item that decides how the rest of the board is arranged. Running a first article through the same checks as the production panel confirms that the two agree, and that comparison is the cheapest form of process control available at prototype stage. Keeping a sample from the panel turns a dispute into a measurement, because the same coupon can be re-examined by both parties without rebuilding the batch.

FAQ

Does passing a humidity test guarantee the product? No. It demonstrates that the mechanism did not appear under the tested conditions, which is evidence and not a guarantee for a longer period or a more severe environment.

Is a coated board always better in humidity? Where the coating is complete, yes. Where it is incomplete, it can trap contamination against the surface and make the result worse.

What does gopcb provide for humidity testing? We provide steady state and cyclic damp heat testing with continuous insulation resistance monitoring, bias and power applied as required, cleaning and coating verification before the test, recovery measurements after it, and sectioning or analysis of any failure to identify the mechanism.