Hipot Testing of PCB Assemblies: Process and Standards

A hipot test applies a high voltage across an insulation barrier and checks that no significant current flows. It is the standard way to verify that the isolation built into a product is actually present after assembly, and it is one of the few tests that can damage the object it is measuring if it is applied wrongly.

What the Test Verifies

Hipot testing verifies the integrity of an isolation barrier: the spacing and the material between a hazardous circuit and a touchable or low voltage circuit. It detects contamination bridging the barrier, a component that has failed short, a creepage distance that has been reduced by a manufacturing defect and, sometimes, a mistake in the assembly.

The test is not a measurement of the design. It is a check that the unit in front of the operator matches the design, which is why it is applied to every production unit in products where the isolation is a safety function.

Choosing the Test Voltage

The test voltage is derived from the working voltage of the circuit and from the applicable standard, which specifies a test level that exceeds the working voltage by a defined factor for a defined duration. The figure is not chosen by the manufacturer; it is taken from the standard that governs the product.

Applying more voltage than the standard requires is a common mistake. It appears to add margin, but it stresses the insulation, and repeated over-testing degrades the barrier that the test is supposed to verify. The correct level is the specified one, applied consistently.

Ramp, Dwell and Discharge

A hipot test has three phases. The voltage is ramped up over a controlled period, typically a second or two, held at the test level for a specified dwell, and then ramped down. The ramp rate matters because a fast rise produces a capacitive charging current that can be mistaken for leakage.

The discharge is as important as the test. The capacitance of the barrier and of the assembly holds charge after the test, and a board left charged is a hazard to the operator and to the next process. A defined discharge path and a dwell at zero before the fixture is opened are part of the procedure rather than an optional step.

Leakage Limits and Interpretation

The pass criterion is a limit on the leakage current that flows at the test voltage, and the limit has to distinguish between a real insulation defect and the charging current of the assembly. A large assembly with a large barrier capacitance draws a measurable current even when the insulation is perfect.

The limit is therefore set from a measurement on a known good unit rather than from a general figure, and it is set high enough to accommodate the capacitance but low enough to catch a defect. An oscillating or unstable reading usually indicates a charging effect rather than a fault. Where a current sensing shunt is used, the measurement should be made on the low side of the barrier, with the test equipment referenced correctly.

Effects on Components

The test voltage is applied across the barrier, not across individual components, but a component that bridges the barrier sees the full test voltage. Optocouplers, digital isolators, isolation amplifiers and Y class capacitors are all rated for a withstand voltage, and applying more than the rating shortens their life or destroys them.

Components that are not rated for the test voltage should be fitted after testing, or the test should be arranged so that they are not stressed. That is a production sequence decision, and it needs to be made with the circuit design rather than discovered during the first production run.

Safety Practice

A hipot test involves a lethal voltage, so the fixture must be enclosed and interlocked, the operator must be trained and the equipment must be in calibration. Manual probing of a barrier at test voltage is not acceptable practice under any circumstances.

The test station should also define what happens to a unit that fails. A failed hipot test means the barrier has been compromised, and the unit must be quarantined and investigated rather than retested until it passes. A unit that passes on the second attempt is evidence of a marginal condition, not of a repair.

Testability and Design

The design has to make the test possible. That means accessible test points on both sides of the barrier, defined isolation boundaries and no component that prevents the voltage from being applied where it is needed.

Where the barrier runs between the primary and secondary side of a power supply, the test is usually applied between the mains terminals and the output terminals. Where a signal path crosses the barrier, the test may need to be applied to each circuit separately. The layout decisions that make this practical are part of the isolation design, and the protective layers that keep the barrier effective are described in conformal coating and board protection.

Routine Test and Type Test

A type test establishes that the design meets the standard and is performed on a sample, usually together with the other safety tests. A routine test is performed on every unit in production and is applied at the same or a slightly reduced level, for a shorter duration.

The distinction matters because the type test can be destructive and is not repeated, while the routine test must not damage good units. The station should also be laid out so that the test does not depend on the operator remembering which probe goes where, since a reversed connection measures something different from what was intended. The design rules that make a fixture reliable are the same ones described in PCB design guidelines for manufacturability. A production station that applies the type test level to every unit will eventually destroy units that would have passed, and the damage appears later as a barrier that no longer holds its rated voltage.

Records and Traceability

Each unit should have a record of the test: the voltage applied, the dwell, the measured leakage and the pass or fail result. The record is the evidence that the safety function was verified, and it is what an auditor will ask for.

The record also makes trends visible. A gradual rise in leakage across a production run points to a process drift, such as contamination or a change in a coating, and catching it from the data is far cheaper than catching it from a field return. The wider set of characteristics that a good board should demonstrate is described in PCB design quality characteristics.

FAQ

Can a hipot test damage a good board? It can if the voltage or the dwell exceeds the specification, or if components across the barrier are not rated for it. Applied correctly, a routine test does not degrade the insulation.

Should the test be done before or after coating? Usually after coating for the final check, and often before coating as well to identify defects early. The sequence depends on whether the coating is part of the insulation.

What if a unit fails and then passes? It should be treated as a failure and investigated. Intermittent pass results indicate a marginal barrier, which is exactly the condition the test exists to find.