How to Test a VFD: Step-by-Step

How do you test a VFD? Testing a variable frequency drive means verifying input line voltage and phase balance, insulation resistance to ground on the motor and cable, DC bus voltage against the expected 1.35x-line ratio, and output waveform quality, following the lockout and capacitor-discharge procedure in IEC 61800-5-1 before any meter touches a terminal. Skip a step and the drive can pass a quick power-up only to trip on the first loaded run, or worse, put line voltage on a cabinet a technician assumes is dead. This sequence covers pre-test safety and discharge, input power and voltage unbalance, insulation resistance on motor and cable, DC bus and rectifier verification, output waveform and motor-side effects, and a functional test under load.

Before You Test: Lockout, Discharge, and What Still Has Charge

Isolate the incoming disconnect and apply lockout-tagout before opening the drive enclosure. The DC bus capacitors keep a lethal charge after the disconnect opens; discharge time depends on drive size and bleed-resistor design, typically 5 to 15 minutes, and the manufacturer's manual states the exact figure for that frame size. Do not trust the charge indicator LED alone.

What we see in the field: the LED discharge indicator runs off a separate low-power tap, and on some frames that tap fails open before the main bus is actually safe. Measure DC+ to DC- directly with a meter rated for the expected voltage class, and confirm a reading under roughly 42 V DC before touching bus terminals or output transistors.

Testing Input Power: Line Voltage and the Unbalance Formula

With the drive still isolated at the incoming disconnect, measure line-to-line voltage on the supply side: L1-L2, L2-L3, L1-L3. Compare each reading to the drive's nameplate voltage range, usually within about 10% of nominal. Then check each phase to ground for stray potential, which points to a wiring fault upstream rather than anything inside the drive.

A small difference between the three line-to-line readings is normal. What matters is how much that difference deviates from the average, because an unbalanced three-phase supply drives disproportionately higher current through one leg of the diode bridge and one phase of the motor winding.

The consequence compounds fast: a few percent of voltage unbalance can translate into a much larger current unbalance in the connected motor, since winding impedance is low and non-linear near rated current. That extra current shows up as heat on one phase of the rectifier and one motor winding, both of which fail from cumulative thermal stress, not from a single event.

See VFD voltage and current ratings for how nameplate tolerances are set, and IEC and NEMA standards for VFDs for the underlying reference documents.

Insulation Resistance: Testing the Motor and Cable, Not the Drive

Disconnect the motor cable from the drive's output terminals before running an insulation resistance test. A megger applies a DC test voltage, commonly 500 V DC for motors rated under 600 V, and that voltage will damage the drive's output IGBTs if it reaches them through the cable.

IEEE 43 gives a widely used rule of thumb: minimum acceptable resistance is roughly 1 MOhm plus 1 MOhm per kV of rated motor voltage, corrected to 40 degrees C. New windings read far above this floor. What matters more than the absolute number is the trend: a reading that drops toward the minimum across successive tests flags moisture ingress or insulation breakdown, well before the winding actually grounds out.

This depends heavily on motor age, ambient humidity, and how long the motor sat idle before the test, so a single low reading on an old, damp motor is not automatically a failure; a downward trend over months is the real signal.

Testing the DC Bus and Rectifier Diodes

With the drive fully discharged and isolated, a multimeter's diode-test function across each rectifier diode pair identifies a shorted or open device; a healthy diode shows a forward voltage drop in one direction and an open circuit in the other. One failed diode does not always stop the drive from running, but it raises ripple current on the DC bus and shortens capacitor life.

Under power, DC bus voltage should sit close to 1.35 times the RMS line-to-line input voltage; a 480 V three-phase supply produces a bus near 648 V DC. A reading noticeably below that, with input voltage confirmed normal, points to a rectifier problem rather than a supply problem.

Testing Output Waveform and Motor-Side Effects

An oscilloscope on an output phase should show a PWM pulse train, not a clean sine wave; that is normal drive behavior, not a fault. What matters for motor health is peak voltage at the motor terminals, not at the drive output, because fast IGBT switching creates a reflected-wave spike that grows with cable length and can approach twice the DC bus voltage on cable runs past roughly 15-50 m.

ABB ACS580, Schneider Altivar ATV340, and Siemens SINAMICS G120C drives all show this same reflected-wave behavior; the mitigation is a dV/dt reactor or sine-wave filter, not a brand-specific fix. Browse the full range of variable frequency drives when comparing filter-equipped models.

If the installation has a long motor cable and no output filter, checking terminal voltage at the motor end, not just at the drive, is the only way to catch this before it pits bearings or breaks down winding insulation. See motor cable length limits and output filters for the specific distance thresholds by voltage class.

Functional Test Under Load: Confirming Duty Rating and Trip Behavior

Run the motor at its intended speed setpoint under actual or simulated load and compare drawn current to nameplate full-load amps. A drive sized for normal duty (110% overload for 60 s) that nuisance-trips under a heavy duty load profile is not defective; it is undersized for the application. See normal duty and heavy duty overload ratings before concluding a trip is a hardware fault.

Record the final parameter set: motor nameplate data as entered, carrier frequency, ramp times, and any overload event. That record is the baseline the next technician compares against when something changes six months later.

Not always obvious. Some integrators run the functional test at no load and call it done, but a no-load test only proves the drive starts and turns the shaft; it says nothing about whether the DC bus sags or the drive derates under the actual current the application draws.

Conclusion

Testing a VFD in sequence, safety and discharge, input voltage and balance, insulation resistance, DC bus and rectifier, output waveform, then a loaded functional run, catches faults that a simple power-up check misses. Each step targets a different failure mode: supply-side imbalance, degraded winding insulation, a failed rectifier diode, cable-length-driven voltage spikes, or a drive undersized for its duty cycle. For the broader picture of how these subsystems interact, see the VFD engineering guide.