Insulation Resistance Test for PCB Assemblies: 6 Rules

Insulation resistance is the resistance measured between two conductors that are supposed to be separated by an insulator. The measurement is simple in principle and difficult in practice, because the value depends on the test voltage, the dwell time, the temperature, the humidity and the cleanliness of the surface as much as on the material itself.

On a PCB assembly the insulator is rarely just the laminate. It is the laminate, the solder mask, any coating, the flux residue and the adsorbed water film, all in parallel. That is why the same board can read differently on two days, and why the test has to be controlled rather than merely performed.

What the Test Measures

A known voltage is applied between two points and the current that flows is measured. The resistance is calculated from the two. What flows is not a single current but a sum: a small bulk current through the material, a surface current along any film of moisture or contamination, and a charging current that decays as the capacitance of the structure fills.

Because of that charging current, a reading taken immediately after the voltage is applied is always lower than the steady value. The dwell time is therefore part of the measurement, and a specification that names a resistance without naming a dwell cannot be compared with another.

Where the board carries a coating, the coating is in series with the surface path. A coating that is thin, damaged or incomplete will allow the surface path to dominate, and that is exactly the condition an insulation resistance test is good at finding.

Test Voltage and Dwell

Higher voltage stresses the insulation more and can reveal a weakness that a low voltage would not, but it can also damage a marginal structure. Most product specifications use a voltage related to the working voltage of the circuit, and the value is chosen once and then held constant.

Dwell should be long enough for the reading to stabilise. A polarisation index, which compares the reading at one minute with the reading at ten minutes, is a useful way to see whether the material is still charging when the result is recorded.

The instrument itself needs to be capable of the range being measured. A meter with a resolution of a megohm cannot meaningfully report a teraohm result, and the leads and fixture leakage must be far below the value being tested, or the fixture becomes part of the measurement.

Humidity and Conditioning

Surface resistance falls sharply as relative humidity rises, because a thinner or more continuous water film covers more of the surface. A board measured in a dry room and the same board measured on a humid day can differ by orders of magnitude without anything having changed in the material.

Conditioning is the answer. The sample is held at a defined temperature and humidity for a defined time before the measurement, and the reading is reported with those conditions. Without them the number has little meaning.

Where the product will see condensation, a wet insulation test is used and the sample is measured while wet rather than after drying. That is a harsher and more realistic test for a board that will operate outdoors or in a condensing environment.

Surface Resistance Versus Bulk

The two paths behave differently and the distinction matters. Bulk resistance is a property of the laminate and the coating, and it changes slowly with moisture. Surface resistance depends on whatever is on the surface, and it can change within minutes.

Guard patterns are used to separate the two. A guard conductor held at the same potential as the measuring point intercepts the surface current so that the meter reads the bulk path instead of the combined one.

In practice, most assembly-level failures are surface failures, caused by residue rather than by a defective laminate. That is why the measurement is so closely tied to cleaning and to handling. The contamination control notes describe the handling rules that keep a surface from becoming conductive, and the no clean residue review explains what residue does under bias.

Fixtures, Probing and Guarding

A test fixture has to insulate better than the board it is measuring. Fixture material that has absorbed moisture, or a cable that runs across a contaminated surface, will dominate the result. Clean, dry, low-leakage materials are used, and the fixture is verified with a reference board of known resistance.

Probing pressure and contact area also matter. A sharp probe that penetrates the coating gives a different reading from one that rests on the surface, and the difference is the coating rather than the board.

For a coated assembly, the test should be performed at the same points before and after coating, so that the effect of the coating is visible rather than inferred. The adhesion test notes describe the related checks that are made on the same samples.

Interpreting Results and Records

A pass and a fail are of limited value on their own. The trend across lots tells far more, because insulation resistance drifts in one direction when a cleaning bath degrades or a flux changes, and the drift appears long before the limit is crossed.

Records should include the test voltage, dwell, conditioning, temperature and humidity, together with the instrument and the fixture used. A result without those details cannot be reproduced.

Where a board has been exposed before test, the moisture history has to be considered as well, since a package that has absorbed water can release it into the assembly during a reflow and change the surface. The floor life guidance covers that exposure.

Common Test Mistakes

The most common mistake is to read leakage current too early. The charging current decays over seconds to minutes, and a reading taken as soon as the voltage is applied describes the capacitance of the board rather than its insulation.

The second mistake is to test a board that has just been handled. Salts from a fingerprint raise the surface conductivity for a while, and the reading then reflects the operator rather than the product.

The third is to compare results taken at different humidities. A specification that gives a limit without a conditioning regime invites a comparison between two numbers that were never measured under the same conditions.

Testing the same board repeatedly is also misleading, because the voltage itself can drive ionic material into the surface and lower the reading on the second test. Fresh samples, or a defined recovery time, keep the comparison valid.

FAQ

Why does the reading fall while the test is running? Because the structure is still charging. The current that flows at the start includes a capacitive component that decays, and the steady value is lower. That is why a dwell time is specified.

Should insulation resistance be tested before or after conformal coating? Both. The before-reading shows whether the surface is clean, and the after-reading shows whether the coating has covered the board properly. Comparing them is more informative than either alone.

Can a board pass in the factory and fail in service? Yes, if the failure is caused by moisture or contamination that arrives later. The factory test describes the board at the moment of test, and the environment decides what happens afterwards.