Three components that look similar on a datasheet and solve entirely different problems. Putting one in the wrong position is a common and expensive error.
All three are three-phase iron-core inductors. On a specification sheet they can look nearly identical — same kVA class, same current rating, similar impedance. Their position in the circuit is what determines what they do, and a component installed in the wrong position does not merely underperform. It can cause the failure it was bought to prevent.
A line reactor sits on the AC input, ahead of the drive's rectifier. It addresses problems flowing in both directions:
A load reactor sits on the drive output. Its job is entirely different: it slows the rate of rise of the PWM voltage waveform reaching the motor.
Modern IGBT drives switch with rise times in the range of 0.05–0.1 µs. That fast edge, travelling down a cable whose surge impedance does not match the motor's, reflects at the motor terminals and can produce a standing wave approaching twice the DC bus voltage. On a 480 V system the bus is roughly 650 V DC, so motor terminal peaks approaching 1,300 V are entirely possible — which is what destroys the first turns of the stator winding.
A load reactor increases the effective rise time, reducing both the peak and the dv/dt. It also reduces motor heating from PWM harmonic content and can reduce audible noise.
A DC link choke is installed in series with the positive DC bus, between the rectifier and the bus capacitors. Many drives either include one or provide terminals for an external unit.
Its harmonic benefit is comparable to a 3% line reactor — it smooths DC bus current ripple, which widens the rectifier conduction angle in much the same way. Its advantage is that it produces no fundamental voltage drop on the AC input, so no motor torque is lost.
Its limitation is equally clear: because it sits behind the rectifier, it provides no protection at all against line-side voltage transients. Installations that need both harmonic reduction and transient protection often use a DC link choke together with input transient protection, or simply use a line reactor.
| 3% impedance | 5% impedance | |
|---|---|---|
| Input current distortion | ~40% | ~35% |
| Fundamental voltage drop | ~3% at full load | ~5% at full load |
| Transient attenuation | Moderate | Better |
| Typical use | Default where torque margin is tight | Stiff supply, transient-prone site, harmonic headroom needed |
The trade-off is real: impedance reduces the voltage available to the motor. On an application already operating near full load at rated speed, 5% impedance can mean the motor cannot reach rated torque at rated speed. On a variable-torque load such as a fan or centrifugal pump, that rarely matters. On a constant-torque application at full load, it can.
Reflected-wave severity depends on cable length relative to the critical length for the drive's rise time. As a widely used field guideline for 480 V systems:
These thresholds shorten at 600 V and lengthen at 240 V, and they tighten further if the motor predates NEMA MG1 Part 31. If the motor is not rated for inverter duty, treat every threshold as conservative.
| Symptom | Component |
|---|---|
| Drive rectifier or DC capacitor failures | Line reactor |
| Nuisance DC bus overvoltage trips | Line reactor |
| Input harmonic current too high | Line reactor or DC link choke |
| Motor winding failures, long cable run | Load reactor or dv/dt filter |
| Harmonics high but no torque margin to give up | DC link choke |
| Motor audible noise and heating on PWM | Load reactor |
No. Line reactors are rated for line-frequency input duty. On the output they see PWM switching content they were not designed for, and they will not solve the reflected-wave problem a load reactor addresses. Use a component rated for the position it occupies.
Frequently yes. They solve unrelated problems — one protects the drive from the supply and reduces input harmonics, the other protects the motor from reflected voltage waves. A long-cable installation on a stiff service commonly needs both.
For harmonic reduction it is roughly equivalent to a 3% line reactor, with the advantage of no fundamental voltage drop. It provides no protection against line-side voltage transients, because it sits behind the rectifier. Choose based on whether transient protection is also needed.
On 480 V systems, reflected wave effects generally become a concern beyond roughly 50 feet with a standard motor, and a load reactor or dv/dt filter is commonly recommended past 100 to 200 feet. Beyond about 600 feet a sine wave filter is often required. Thresholds tighten at 600 V and with motors not rated to NEMA MG1 Part 31.
It reduces the voltage available to the motor by roughly 5% at full load, which can prevent rated torque at rated speed on a constant-torque application already running near full load. On variable-torque loads such as fans and centrifugal pumps it is rarely a practical limitation.
Fast IGBT switching plus a long cable equals voltage doubling at the motor terminals. Understanding transmission line behaviour in a motor feeder explains a whole category of otherwise inexplicable winding failures.
Data center distribution and grid-scale battery storage both present harmonic and thermal problems that traditional plant experience handles badly. The differences are specific and they change the component selection.
NEMA 1, 3R, 4X and 12 are not a simple ladder of increasing protection. Each addresses different hazards, and the wrong choice either wastes money or fails in service — often through heat rather than ingress.