When a variable frequency drive switches its output transistors, the voltage rises extremely fast, a steep dV/dt, meaning a large change in voltage over a very short time. On short cables that is harmless, but on long motor leads the steep edge behaves like a wave traveling down a transmission line, reflecting off the motor and adding to the incoming pulse so the voltage at the motor terminals can spike to roughly twice the intended value. That repeated overvoltage hammers the motor's winding insulation and, over time, punches through it. It is a failure mechanism born from the combination of fast switching and long cable runs, which makes it a recurring headache on remote well pads where the drive and motor sit far apart.
dV/dt & Reflected Wave in one line: dV/dt is the rate at which a VFD's output voltage rises during each switching event, and the reflected wave is what happens when that fast-rising pulse travels down a long motor cable, reflects at the motor, and superimposes on the incoming pulse to nearly double the peak voltage at the motor terminals. This repeated voltage overshoot stresses and eventually breaks down winding insulation, which is why long-lead installations use dV/dt filters, terminators, or output reactors to tame it.
Modern drives use fast switching devices, commonly IGBTs, that turn on in a fraction of a microsecond. That gives a very high dV/dt: the output voltage jumps from near zero to full bus voltage almost instantly on each pulse. A fast edge is not a problem in itself, but it contains high-frequency content, and high-frequency signals on a long cable no longer behave like simple wiring. The cable starts to act like a transmission line, where a fast voltage step travels along it as a wave at a finite speed rather than appearing everywhere at once.
When that traveling wave reaches the motor, it meets a large impedance mismatch, because the motor's surge impedance is much higher than the cable's. At a mismatch, part of the wave reflects back toward the drive. The reflected wave adds on top of the still-arriving incoming wave, and where they overlap at the motor terminals, the voltages sum. In the worst case the reflected pulse nearly equals the incoming one, so the terminal voltage briefly reaches close to double the DC bus voltage. This is why the effect is often called voltage doubling: the motor sees peaks far higher than the drive's nominal output.
The reason cable length is the key variable is timing. On a short cable, the reflected wave returns before the incoming edge has finished rising, so it partly cancels rather than fully adding, and the overshoot stays modest. Once the cable is long enough that the wave's round trip takes longer than the rise time of the edge, the incoming pulse has fully arrived before the reflection returns, and the two add at full strength. That threshold length, which depends on the drive's rise time and the cable's propagation speed, is why every drive and cable combination has a length beyond which reflected-wave overshoot becomes severe.
The victim of reflected-wave overshoot is the motor's winding insulation, particularly the insulation on the first few turns of the winding closest to the terminals. Those turns absorb the brunt of the fast-rising, doubled-voltage pulse because the steep edge does not distribute evenly across the winding; it stresses the entrance turns hardest. Repeated thousands of times per second, every switching event, these overvoltage spikes gradually degrade the insulation through partial discharge and heating until it fails and the motor faults.
What makes this insidious is that the damage is cumulative and often silent until failure. A motor can run for months looking fine while its insulation quietly erodes under the constant hammering of overshoot pulses, then fail seemingly without warning. Because the mechanism depends on the specific combination of drive rise time, cable length, and motor insulation rating, two similar installations can have very different lifespans: the one with the longer cable run may cook its motor insulation while the one with a short lead never has trouble. This is why inverter-duty motors, built with reinforced insulation designed to survive these pulses, exist as a countermeasure.
The severity also interacts with drive settings and cable choices. A higher carrier frequency means more pulses per second, so more overshoot events battering the insulation, and certain cable constructions propagate and reflect the waves differently. None of these individually cause the failure, but together they set how hard and how often the motor terminals get hit. Understanding the mechanism is what lets an engineer diagnose a string of premature motor failures on long-lead wells as a reflected-wave problem rather than blaming the motors.
There are several standard fixes, and they attack the problem at different points. An output or load reactor placed at the drive slows the rise time of the voltage edge, softening the dV/dt so the reflection is less abrupt. A dedicated dV/dt filter does the same more aggressively, rounding the edges specifically to limit overshoot. A more complete sine-wave filter reconstructs a near-sinusoidal output so the motor barely sees switching edges at all. At the far end, a terminator, an impedance-matching network at the motor, can absorb the wave so it reflects less. Choosing among these is a balance of cost, size, and how severe the overshoot is for a given cable length.
Manufacturers publish cable-length limits precisely because of this mechanism. Beyond a stated lead length, they specify that a filter or reactor must be added, or that an inverter-duty motor must be used, because the terminal overshoot would otherwise exceed what standard insulation can tolerate. Respecting those limits during design is far cheaper than replacing burned-out motors, and ignoring them is a classic cause of repeat failures. The right question at design time is not just whether the drive can reach the motor electrically, but whether the cable is long enough to invite reflected-wave overshoot.
Remote oilfield sites make this a routine concern rather than an edge case, because drives and motors are often deliberately placed far apart, a surface drive feeding a downhole motor, or a control building set back from the equipment it powers. Those long runs sit squarely in the range where reflected-wave overshoot bites. When remote monitoring surfaces recurring motor faults or insulation-related trips across a fleet, correlating them with cable lengths and drive configurations can reveal a reflected-wave pattern that would be hard to spot one well at a time. A cloud SCADA view of motor faults across many sites turns a scattering of individual failures into a diagnosable trend, pointing the engineer toward filters or cable changes rather than repeated motor swaps.
The drive's fast switching edge travels down a long cable like a wave on a transmission line. When it reaches the motor, the impedance mismatch reflects part of it back, and the reflected wave adds to the still-arriving incoming pulse. Where they overlap at the terminals, the voltages sum, so the peak can reach nearly twice the drive's bus voltage. On short cables the reflection returns too soon to fully add, so the overshoot stays small.
There is no single number; it depends on the drive's voltage rise time and the cable's propagation speed. The threshold is roughly the length at which the wave's round trip takes longer than the rise time of the switching edge, because past that point the incoming pulse and its reflection add at full strength. Drive manufacturers publish specific cable-length limits for their products and require a filter, reactor, or inverter-duty motor beyond those lengths.
A dV/dt filter is a device at the drive output that slows and rounds the fast voltage edges so the reflected-wave overshoot at the motor stays within safe limits. You do not always need one: on short cable runs the overshoot is harmless. You need mitigation, whether a dV/dt filter, an output reactor, a sine-wave filter, or an inverter-duty motor, when the cable is long enough to exceed the manufacturer's stated limit, which is common on remote sites where the drive and motor sit far apart.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.