Insulation resistance testing on a VFD-fed motor — why the drive must first be disconnected
Insulation resistance testing on a VFD-fed motor — why the drive must first be disconnected
The guide on insulation resistance testing covers the standard procedure and the test voltages per circuit type (250/500/1000 V DC), and the guide on DC-bus residual voltage in variable frequency drives covers why the capacitors in the DC intermediate circuit can retain dangerous residual voltage after switch-off. This article covers a third concern, specific to VFD-fed motors: why a regular insulation resistance test must not simply be performed at the motor terminals while the motor is still connected to the drive.
The problem: the test voltage damages the drive
An insulation tester (megger) typically applies 500 V or 1000 V DC for a motor circuit (see the insulation resistance testing guide for the full table). That voltage is appropriate for the motor's own winding insulation, but far above what the semiconductor circuits (IGBTs, freewheeling diodes) and the electrolytic capacitors of the DC intermediate circuit in a variable frequency drive are designed to withstand. If the insulation test is performed with the motor still connected to the drive, the test voltage is also applied to these components via the motor cable — with a real risk of irreparable damage to the drive.
The solution: fully disconnect the motor
For an insulation resistance test of the motor winding, the motor cable must therefore be fully disconnected from the variable frequency drive — not merely stopped electronically (see the Safe Torque Off (STO) guide for why STO does not interrupt the supply), but physically disconnected at the drive's output terminals, so that the test voltage is applied only to the motor winding itself.
Sequencing with the residual-voltage check
Before working on the drive side of the cable (for example to disconnect the motor cable at the output terminals), it must first be established that the DC-bus residual voltage has decayed to a safe value — see the guide on DC-bus residual voltage and capacitor discharge time. Only after that check, and with the motor cable physically disconnected from the drive, is it safe and meaningful to perform the regular insulation resistance test on the motor winding itself.
Practical relevance
During a periodic NEN 3140 inspection or a fault investigation on a VFD-fed motor (pump, fan, conveyor), the insulation resistance test must always be performed after physically disconnecting the motor cable from the drive — never directly through the drive, even if the drive itself is switched off or in STO. When in doubt about the correct test sequence: check residual voltage first, then disconnect, then perform the insulation test.
Common mistakes
- Performing an insulation resistance test with the motor still connected to the drive — the test voltage (500-1000 V DC) can damage the drive's semiconductors and DC-bus capacitors.
- Assuming STO (Safe Torque Off) is sufficient to make the drive "safe" for an insulation test — STO interrupts motor control, not the physical connection between the motor cable and the drive.
- Not checking the DC-bus residual voltage before working on the drive side of the motor cable to disconnect it.
- Not distinguishing between an insulation test of the motor winding itself and a test of the entire cable including the drive — these require a different approach, and for the drive itself typically the manufacturer's specific test procedure rather than a generic megger test.
Related
Further reading
- InspectieInsulation resistance measurement — method and limit values
- IEC 62446-1Insulation resistance testing on the DC side of PV strings
- LOTO / IEC 60204-1DC-bus residual voltage — capacitor discharge time on VFDs and PV inverters
- InspectieLocating cable faults — TDR versus insulation resistance testing
- Praktijk / IEC TS 60034-25Bearing currents in variable-speed motors — why a motor bearing can fail prematurely with no mechanical cause
- InspectieTest sequence for initial verification — why dead tests come before live tests