A motor can run for years on a drive and then fail in the windings, and the cable between the two is often the reason nobody looks at. The drive tested fine. The motor tested fine when it went in. What sat between them was a long cable carrying pulses with edges steep enough to double their own voltage by the time they reached the motor terminals. That is reflected wave, and it is one of the quietest motor killers in the plant.
This is a cable and motor protection problem, and it starts at installation and selection, not after the winding lets go. It lives in the VFD Installation Guide. The short version is that a VFD does not feed a motor the smooth sine wave the line did, and the difference shows up as voltage stress that scales with how far the cable runs.
The Cable Is a Transmission Line
A VFD makes its output by switching the DC bus on and off thousands of times a second, and the edges of those pulses are steep, rising in something like 50 to 100 nanoseconds on a modern IGBT drive. That steep edge is the whole trouble. Over a short cable it does not matter, but a motor cable is not a simple wire. At those switching speeds it behaves like a transmission line, and a transmission line reflects.
When a fast voltage pulse travels down the cable and reaches the motor, it meets an impedance much higher than the cable’s own, and part of the pulse bounces back the way it came. If the cable is long enough that the pulse’s travel time is a meaningful fraction of its rise time, the reflected pulse adds on top of the incoming one, and the voltage at the motor terminals climbs toward twice the DC bus voltage. On a 480 volt system, where the DC bus sits around 650 volts, that doubling puts spikes of 1200 to 1600 volts on the motor terminals, over and over, thousands of times a second.
The Critical Length Is Shorter Than You Think
The number that surprises people is how short the cable has to be for this to fully develop. On an IGBT drive with those fast edges, the reflection reaches its full doubling at a cable length of roughly 7 to 17 meters, which is about 25 to 50 feet. That is not a long run. It is a drive in one cabinet and a motor across the room.
This is where the common rule of thumb fails. People assume cable type only matters on long runs and reach for whatever wire is handy under 100 feet. The physics does not cooperate. The reflected wave is fully developed well before 50 feet, and the high frequency common mode current that drives bearing damage flows at any length, because it does not care how long the cable is. Cable length is not a footnote in the sizing math. In ORCA-Size, the framework behind the VFD Selection and Sizing Guide, it is one of eight multiplicative derating factors, sitting alongside ambient temperature, altitude, carrier frequency, and load class, each one shrinking the usable continuous rating. Run the Monte Carlo and the penalty does not wait for a long cable to show up. A 50 foot standard wire run carries most of the derating that a 300 foot run does. Length amplifies the problem. It does not create it, and it does not excuse it on a short run.
What the Overvoltage Does to the Motor
Those repeated spikes do their damage at the very front of the winding. The steep edge does not distribute evenly across the coil. It concentrates on the first few turns of the first coil, so the turn to turn insulation right at the winding entrance sees the brunt of every pulse. Over time, the voltage stress ionizes the tiny air gaps in the insulation, a process called partial discharge or corona, and that discharge slowly erodes the insulation until a turn shorts and the winding fails.
This is not a fringe mechanism dug up for an article. In the ORCA-VFD reliability work, electrical degradation appears as insulation breakdown from voltage stress and partial discharge, accelerated by the high frequency switching and the common mode voltages that come built into VFD operation. The reflected wave is that same stress with a cable length attached to it. Give the pulse a long enough run to double, point it at a standard winding, and the degradation the framework treats as a slow electrical wear-out arrives early and local, right at the first coil.
A standard motor built for line power has thinner insulation margins and no particular defense against this, so it fails faster, sometimes in a year or two, with a burned spot near the winding entrance and no other explanation. The motor did not wear out. It was being hit with voltage it was never designed to see, and the cable length decided how hard. The same long cable and common mode voltage also drive the shaft and bearing currents that flute bearings, which is why premature motor failures on drives so often trace back to the cable, a pattern the VFD Troubleshooting Guide treats as a wiring problem wearing a motor’s clothes.
There is a reason these failures look like bad luck. Winding failures sit at a Weibull shape near one, the signature of random failure, while bearings fail at a shape near three, the signature of wear-out. IEEE 493, the Gold Book, publishes the winding failure rates that make that curve. A reflected wave kill hides inside that random band. It is not random at all. It is a cable length and an insulation class deciding an outcome, dressed up as chance.
What Actually Protects the Motor
There are a handful of defenses, and they stack. The first and cheapest is the right motor. An inverter duty motor built to NEMA MG1 Part 31 is designed to survive the spikes, with insulation rated for roughly 1600 volt peaks and fast rise times. On any VFD driven motor this is the baseline, and on a long lead it is not optional.
When the motor alone is not enough, the fix moves to the drive output. A load reactor adds series impedance that softens the edge and knocks the peak down modestly, and it is the least expensive filter. A dV/dt filter does more, slowing the voltage rise from tens of nanoseconds to something like 800 to 2000 nanoseconds and pulling the peak at the motor down from around 1400 volts toward 1000, which takes most of the sting out of the reflection. A sine wave filter does the most, delivering a nearly sinusoidal voltage to the motor so the terminals never see a fast edge at all, at the highest cost and the most loss. The choice among them comes down to cable length, motor rating, and how critical the motor is, and it is a selection decision, not an afterthought.
And underneath all of it, use proper shielded VFD cable with symmetrical ground conductors, and keep the run as short as the layout allows. The shielded cable gives the high frequency common mode current a low impedance path back to the drive instead of through the motor bearings, so the right cable protects the windings and the bearings at once.
A field case makes the stack concrete. The ORCA-VFD dataset came off a seven year old ABB ACS880 driving a 15 horsepower, 1800 rpm, four pole inverter duty motor on a screw conveyor at 480 volts, with VFD-rated cable and less than 200 feet between the drive and the motor. Every defense was in place before the motor turned. The inverter duty insulation took the spikes, the VFD-rated cable gave the common mode current a path home instead of through the bearings, and the run was kept short. Seven years of telemetry off that machine show temperature and current drift, the slow honest stuff, not a burned turn at the winding entrance. The transmission line problem was engineered out at install, so the motor got to age instead of getting killed.
Setting It Up Right
Reflected wave is not a rare, long-cable curiosity. It is present on almost every VFD driven motor, fully developed by 50 feet, and it is the reason a healthy motor fails in the windings with the drive showing no fault at all. Specify the inverter duty motor, choose the output filter to match the cable length, run proper shielded cable, and keep it short, and you turn a quiet motor killer into a non event. Skip those at installation and you are betting the motor’s life on a cable being shorter than the physics requires.
Cable selection, output filters, and inverter duty motor specification get their full treatment in the VFD training program and in Before the First Fault.
Author: Dr. Carl Lee Tolbert, PhD, CMRP, Wayward Leaders LLC, waywardleaders.com
Frequently Asked Questions
What is reflected wave on a VFD?
Reflected wave is the voltage overshoot that happens when the fast switching pulses from a VFD travel down a motor cable and bounce back at the motor terminals. The reflected pulse adds to the incoming one, and the voltage at the motor can climb toward twice the DC bus voltage, roughly 1200 to 1600 volts on a 480 volt system. Those repeated spikes stress the motor winding insulation.
How long can a motor cable be before reflected wave is a problem?
Shorter than most people expect. On a modern IGBT drive the reflection reaches full voltage doubling at about 25 to 50 feet, and the common mode currents that damage bearings flow at any cable length. Treat cable type and motor protection as necessary even on short runs, not just long ones.
Why do long cables damage the motor?
The doubled voltage spikes concentrate on the first few turns of the winding, and over time the electrical stress erodes the turn to turn insulation through partial discharge until a turn shorts. A standard motor has thin insulation margins and can fail in the windings within a year or two, usually with a burned spot near the winding entrance and no drive fault to explain it.
Do I need an inverter duty motor on a VFD?
Yes. An inverter duty motor built to NEMA MG1 Part 31 has insulation rated for the voltage spikes a drive produces, around 1600 volt peaks, and it is the baseline defense on any VFD driven motor. On a long cable run it is not optional, because a standard motor will not survive the reflected wave stress.
What is the difference between a dV/dt filter and a sine wave filter?
A dV/dt filter slows the voltage rise time and lowers the peak the motor sees, taking most of the reflection stress out while keeping cost and loss moderate. A sine wave filter goes further and delivers a nearly sinusoidal voltage to the motor, so the terminals never see a fast edge, at higher cost and higher loss. The dV/dt filter is the common choice for cable length problems, and the sine filter for the most demanding or longest runs.
Does a short cable run mean the motor is safe?
No. The reflected wave is fully developed by about 50 feet, and the common mode current that drives bearing damage is present at any length because it does not depend on the cable being long. A short run reduces some of the risk but does not remove it, which is why inverter duty motors and proper shielded cable matter even on drives sitting close to their motors.