Dielectric Withstand (Hipot) Testing: Ramp, Dwell Time and Reading Leakage Current

The dielectric withstand test, usually called a hipot test, is the routine check that the insulation between live parts and accessible surfaces will not break down in service. It is simple in principle and surprisingly easy to run badly. Three things decide whether the result is worth writing on the report: how you got to the test voltage, how long you held it, and what current you let pass.

AC or DC, and why it matters

AC hipot stresses the insulation the way the mains does, and it is the usual choice for mains-connected equipment. DC hipot is common for cables and for devices where the capacitive charging current of an AC test would be enormous. They are not interchangeable: a DC test at the AC-equivalent voltage does not exercise the insulation the same way, so pick the one the product standard calls for and state it on the record.

Choice	Typical use	Watch out for
AC (50/60 Hz)	Mains-powered instruments and appliances	Capacitive inrush at start
DC	Cables, large capacitive loads	Residual charge after test
Both, for a set time	Some standards specify a combined routine	Discharge safely before handling

Ramp and dwell

Do not slam the voltage on. A controlled ramp, often to the test value over a few seconds, avoids a transient that trips the tester without telling you anything about the insulation. Once at the set voltage, hold it for the specified dwell time, most commonly 60 seconds for type tests and shorter for routine production. Cutting the dwell to "save time" is the most common reason a borderline unit passes on the line and fails in the field.

• Ramp to the test voltage over a few seconds, not instantly.
• Hold for the full dwell; a short hold misses slow breakdowns.
• Record the actual applied voltage, not the dial setting.
• Discharge the unit before anyone touches it.

Reading the leakage current

The trip threshold is the limit, not the measurement. A healthy unit draws a small capacitive and resistive leakage current; a failing one draws far more, and the tester trips. But a trip is a binary event. Logging the actual leakage current tells you far more: a unit sitting near the limit today will exceed it next month. Set the trip at the standard value, then also note the current you measured.

Set-up errors that invalidate the test

A loose earth bond, a test lead with a poor connection, or a probe touching the wrong point will either fail a good unit or pass a bad one. The return path through the earthing terminal must be solid, because that is what the tester measures against. Cover or insulate exposed metal that is not part of the test, or the tester will trip on a harmless arc to a screw head.

Interpreting a trip

A trip at the very start, before the ramp finishes, usually points at set-up or a hard short. A trip near the end of the dwell points at a slow breakdown or a marginal design. Either way, the response is not "reset and try again"; it is to find where the current went. A hipot result that says only "pass" is not a result that survives an audit.