Humidity Testing and Damp Heat for Electronic Assemblies

Water is the most common cause of long term electronic failure, and it rarely arrives as an obvious splash. A thin adsorbed film, a slow condensation cycle, or a residue that becomes conductive when the air turns humid are all enough to destroy a circuit that passed every dry test. Humidity testing exists to expose those conditions before the product does.

What Humidity Testing Is For

The purpose of humidity testing is to accelerate the chemical and electrochemical processes that moisture enables. Contamination dissolves, insulation resistance falls, metal corrodes and metal ions migrate. None of these requires liquid water; a film a few molecules thick is sufficient given enough time and a voltage difference.

The test therefore targets the interaction between the assembly and its environment rather than the assembly alone. Two identical boards with different cleaning quality will produce very different results, which is exactly the information a qualification programme needs. The mechanisms involved overlap with those described in this overview of short circuit investigation.

Steady State Versus Cyclic Damp Heat

Steady state damp heat holds the chamber at a constant temperature and humidity, commonly eighty-five degrees Celsius and eighty-five percent relative humidity. It is simple, repeatable and effective at accelerating corrosion and contamination-related leakage, but it never produces condensation because the sample stays at the chamber temperature.

Cyclic damp heat alternates between a hot humid phase and a cooler phase, allowing the sample to fall below the dew point and collect liquid water. The wetting and drying cycles concentrate contaminants and deliver fresh electrolyte repeatedly, which makes cyclic testing considerably more aggressive for many failure mechanisms. Choosing between them depends on whether the product will experience condensation in service.

Condensation and Dew Point

Condensation occurs when a surface is cooler than the dew point of the surrounding air. In a chamber this happens during the transition from hot to cold, when the sample cools faster than the air around it, or when the humidity is raised while the sample is still cool. Controlling the ramp rates is therefore part of defining the test severity.

In service the same principle applies. Equipment that cools overnight in a humid environment will collect water on its surfaces, and equipment that is brought from a cold store into a warm room will condense on every surface. Both conditions are worth reproducing in a qualification programme if the product will genuinely encounter them.

Bias and Its Effect

Applying a voltage during the test changes the picture substantially. An unbiased assembly may suffer corrosion but no migration, because migration requires an electric field to drive ion transport. A biased assembly can develop dendrites and leakage paths that would never form without the field, and the effect is strongest with direct current at a fixed polarity.

Bias also affects the failure location. Conductive paths tend to form between adjacent conductors of opposite polarity, so the biasing scheme determines which features are at risk. A test that applies bias to the wrong nets can therefore pass while the real product fails, which is why the bias scheme deserves as much attention as the temperature and humidity settings. Documenting which nets were biased and at what voltage is therefore as important as recording the chamber profile.

Insulation Resistance Measurement

Insulation resistance is the standard electrical measurement during and after humidity exposure. It is typically measured on a comb pattern or between selected nets, and the acceptance criterion is usually a minimum value at the end of the test plus a limit on how far the value is allowed to fall from the initial reading.

Measuring inside the chamber avoids the recovery that occurs when a board dries out, but it requires feedthroughs and careful fixturing that does not itself become a leakage path. Measuring after removal is simpler but can miss a failure that recovers within minutes. Whichever method is used, the measurement voltage and the dwell time before reading should be stated, because both affect the result.

Corrosion and Electrochemical Failure

Corrosion appears as discoloured metal, as a change in surface finish appearance, or as a rise in contact resistance. Silver finishes darken as they form sulphide and oxide layers, copper develops green or brown products, and aluminium corrodes rapidly in the presence of chloride. Each of these has a different threshold and a different sensitivity to the specific contaminants present.

Electrochemical failure is the more dangerous outcome because it can be invisible until it becomes a short. A corrosion product that bridges two conductors has the same effect as a component failure, and it may appear only after the assembly has been in the field for months. Post-test inspection therefore needs to look for the early signs rather than only for gross damage. The interpretation of such findings is part of PCB quality judgement.

Chamber Control and Verification

The chamber’s own performance must be verified, because a reading taken from a drifting sensor produces a test that is less severe than reported. Calibration of the temperature and humidity sensors, plus spot checks with an independent instrument, keeps the reported conditions meaningful.

Humidity sensors are particularly prone to drift and to contamination, and a chamber that has been used extensively may hold a different humidity near the sample than at the sensor. Placing an independent logger beside the samples for at least one run confirms that the intended conditions are actually being delivered where it matters.

Interpreting Results

A failure at the end of a damp heat test is not automatically a product defect. Contamination from handling during the test, an unsuitable fixture, or a board that was never intended to survive the applied bias can all produce a failure that says nothing about the product. Establishing a baseline with a known-good sample is essential before drawing conclusions.

Equally, a pass is only meaningful if the test was genuinely representative. A test performed on a clean board with no bias and a benign profile has limited predictive value for a product that will operate outdoors with a permanent voltage difference between adjacent nets. Matching the test to the service environment is the only way to obtain useful assurance. Where the service environment is not well characterised, running two profiles — one benign and one condensing — brackets the risk far better than a single test at a guessed severity level.

Specification and Test Durations

Durations for damp heat tests commonly range from one hundred to one thousand hours, with longer tests reserved for products where the reliability requirement is high and the exposure is continuous. The duration should be justified by the expected service life and by the acceleration factor of the chosen conditions, rather than selected from habit.

The specification should also state the bias scheme, the monitoring method, the recovery period before measurement and the inspection criteria after the test. Those details determine whether two laboratories testing the same product will reach the same conclusion. Documenting them alongside the rest of the build records described in this guide to the PCB production flow prevents an avoidable disagreement later.

FAQ

What is the difference between damp heat and humidity testing? The terms overlap, but damp heat usually refers to a controlled combination of temperature and humidity applied in a chamber, often at eighty-five degrees Celsius and eighty-five percent relative humidity. Humidity testing on its own may be run at a lower temperature where the emphasis is on moisture rather than thermal acceleration.

Why does bias make the test more severe? Because electrochemical migration requires an electric field. Without bias, contamination can dissolve and corrode metal, but it cannot form a conductive filament between two conductors. Applying voltage at a fixed polarity creates the conditions for dendrite growth and leakage paths.

How long should the test run? It depends on the product and the exposure it will see, but durations of several hundred hours are typical for qualification work. The relevant question is whether the duration, combined with the chosen temperature and humidity, represents enough of an acceleration to predict service behaviour with confidence.