How to Insulation-Resistance Test a Motor Circuit

Back to Blog

Isolate First: What You Need

The test applies hundreds of volts DC to conductors, so everything begins with isolation: the circuit de-energized, verified dead with a rated tester, and locked out under the site's control-of-hazardous-energy procedure. That verification and the lockout are the domain of qualified personnel; the megohmmeter comes out only after the circuit is proven dead. Isolation has a second, less obvious dimension - isolating the things the test would destroy. Drive and soft-starter power electronics, surge protection devices, capacitors, and some instrumentation cannot tolerate the test voltage, so the circuit is opened at the points that leave only the cable and winding under test. Megging into a connected VFD is one of the classic ways to convert an insulation question into a drive repair.

The equipment list is short: a megohmmeter with the appropriate test-voltage ranges, test leads in good condition - a leaky lead reads as a leaky winding - and the machine's test history if any exists, because insulation resistance is a trend far more than it is a number. Note the winding temperature and the weather while you are at it: insulation resistance falls steeply as temperature rises, roughly halving for a modest rise, which is why IEEE 43 corrects readings to a common base temperature before comparing them, and why a winter reading and a summer reading are not directly comparable.

Running the Test

The core measurement is winding-to-ground: the megohmmeter's line lead on the phase conductors - commonly bonded together for the test - and its earth lead on the motor frame or ground conductor. Choose the test voltage per the equipment rating; IEEE 43's test-voltage table puts low-voltage machines commonly at 500 or 1000 VDC. Apply the voltage and watch the reading climb as the insulation absorbs charge; the convention is to record the value at one minute, which puts every test on the same footing. Erratic or steadily falling readings during the test are themselves findings, suggesting moisture or tracking rather than stable, dry insulation.

For larger machines, the polarization index extends the same measurement: the ten-minute reading divided by the one-minute reading. Dry, healthy insulation keeps absorbing charge and the ratio climbs; wet or contaminated insulation conducts steadily and the ratio stays near one. IEEE 43 describes both the method and its limits, including the fact that very high one-minute readings make the index uninformative. When the test ends, discharge the winding through the megohmmeter's discharge function or per the site practice, for at least as long as the test voltage was applied - the tested circuit is a charged capacitor until proven otherwise, and it belongs in the same mental category as any other stored energy.

Reading the Numbers, Splitting Cable from Winding

IEEE 43 provides the recommended minimum values by machine class and era, including the long-standing rule of thumb of one megohm per kilovolt of rating plus one for older insulation systems and higher minimums for modern windings. But the minimum is a floor, not a health certificate: the diagnostic power is in the trend. A machine whose corrected readings have slid steadily downward over successive tests is telling a story that a single above-minimum reading does not refute, and the interpretation of that story - and every decision about returning the circuit to service - belongs to qualified personnel applying the site's engineering judgment, not to a table alone.

A low reading taken from the starter end condemns the circuit, not the motor: the cable, its terminations, and the winding are all in parallel under the test. Split them by opening the motor connection box and testing the cable and the winding separately. Wet cable ends, water in a conduit run, or a compromised termination read low just as convincingly as a failing winding, and they are far cheaper to fix; motors have been pulled for readings that lived entirely in a flooded junction box. Where a facility's maintenance records live alongside its operating data - in a platform such as Merobix - filing each test's corrected reading against the asset turns the next low reading into a trend point rather than an isolated alarm, and the common mistakes become visible: testing with the drive still connected, skipping the discharge, comparing readings taken at wildly different temperatures, and condemning a winding for a cable's sins.

Frequently Asked Questions