A variable-frequency drive may look like a single electronic box mounted in a control cabinet. Electrically, however, it is a carefully coordinated power-conversion system: a rectifier receives the incoming supply, a DC link stores and stabilizes energy, and an inverter produces the controlled output needed by the motor.
Understanding those stages is useful even when an engineer never designs an inverter from scratch. It explains why a drive can report undervoltage during a motor start, overvoltage during deceleration, excessive current during acceleration, or overtemperature after months of apparently normal operation.
The essential question is not simply whether the drive receives electrical power. It is where the energy flows, which components control that flow, and which conditions cause a protection circuit to intervene.
1. The three power stages inside a typical industrial VFD
The common architecture of a two-level, voltage-source variable-frequency drive can be summarized as follows:
| Stage | Main components | Function |
|---|---|---|
| Input rectifier | Power diodes, or an active semiconductor front end | Converts incoming AC power into DC |
| DC link | Capacitors, busbars and sometimes inductors | Buffers energy and limits bus-voltage ripple |
| Output inverter | IGBTs or other controlled power switches | Synthesizes variable-frequency, variable-voltage output for the motor |
| Supporting systems | Gate drivers, control electronics, sensors, protection and cooling | Command and protect the power path |
The control board supervises the stages, but it does not itself carry the main motor current. The main power path passes through the rectifier, DC link and inverter.
Not every drive follows exactly this arrangement. Regenerative drives, matrix converters and specialized high-power systems can use different topologies. Nevertheless, the rectifier–DC-link–inverter arrangement is an effective starting point for understanding many factory drives.
2. What the rectifier actually does
A conventional three-phase drive often begins with a six-diode bridge rectifier. The bridge directs portions of the incoming AC waveforms into a unidirectional DC supply.
The output is not perfectly smooth. Its instantaneous value changes as different diode pairs conduct. Under balanced three-phase, 50 Hz operation, the characteristic six-pulse ripple component occurs at 300 Hz, although the actual bus waveform depends on the source impedance, loading and DC-link components.
A diode rectifier is simple and robust, but its power flow is essentially one-way. That limitation matters when a motor is being driven by its load and starts returning energy toward the DC link.
Some industrial converters use a controlled rectifier or an active front end instead. An active front end can control input current and, when designed for it, return regenerated energy to the supply network.
Engineering distinction: A standard diode-front-end drive and a regenerative drive may control the same motor but behave very differently when energy flows back from the machine.
3. Why the DC link is more than a large capacitor
The DC link is the electrical connection between the rectifier and inverter stages. Its capacitors absorb short-term differences between the power entering the drive and the power leaving it.
For a typical 400 V three-phase supply, the peak line-to-line voltage is approximately 400 multiplied by the square root of two, or about 566 V. This provides a useful conceptual reference for an unloaded diode-rectified DC link, not an acceptance value for any particular drive. Real voltages vary with supply conditions, load, rectification topology and the manufacturer's design.
The DC-link capacitors cannot be treated as unlimited energy storage. When the inverter draws a sudden burst of power, the bus voltage can fall. When a moving load sends energy back toward the drive, the bus voltage can rise.
Capacitor ageing, elevated temperature and high ripple-current stress can gradually reduce the link's ability to perform its buffering function. Some drives also contain DC chokes or input reactors that influence current ripple and harmonic performance.
Diagnostic principle: A DC-bus alarm describes the energy balance in the converter; it does not automatically identify a defective capacitor.
4. How an IGBT inverter creates a three-phase output
In a conventional two-level three-phase inverter, six controlled semiconductor switches are arranged in three legs. Each leg connects one motor phase alternately toward the positive or negative DC bus under controlled switching.
The switching devices are commonly IGBTs in many established industrial designs, although MOSFETs and silicon-carbide devices are important in other voltage, power and efficiency ranges.
An IGBT is an insulated-gate bipolar transistor. Its gate circuit controls the switching state while the power terminals conduct the load current. Gate-driver circuits provide the correct control signals and incorporate protection functions appropriate to the device and converter.
The motor does not receive a pure sinusoidal voltage directly from the individual switching events. Instead, the converter produces high-frequency voltage pulses. The motor winding inductance helps smooth the resulting current, and the drive's modulation strategy determines the fundamental output voltage and frequency.
The physical IGBT module pictured with this article contains semiconductor dies, internal interconnections and associated power paths. It is a building block, not a complete variable-frequency drive.
5. Pulse-width modulation: voltage pulses, controlled motor current
Pulse-width modulation (PWM) changes the timing of the semiconductor switching events so that the fundamental component of the motor voltage follows the required command.
A motor drive may vary both fundamental output frequency and voltage according to its control mode. In simple scalar control, the voltage-to-frequency relationship is central. Vector-control methods also use motor models and current information to manage flux and torque more precisely.
Higher switching frequency can improve aspects of the output waveform, but it normally increases switching losses and may affect electromagnetic interference. Lower switching frequency reduces some switching losses but can increase motor-current ripple and audible effects.
| PWM-related factor | Potential benefit | Engineering trade-off |
|---|---|---|
| Higher switching frequency | Smoother motor current in some applications | Greater semiconductor switching loss and possible thermal derating |
| Lower switching frequency | Reduced switching loss | More current ripple, motor noise or torque ripple |
| Faster voltage edges | Efficient switching transition | Higher common-mode interference and motor-insulation stress |
| Output filtering | Reduced high-frequency components reaching the motor | Added cost, losses, size and application-specific constraints |
This is why selecting a drive is more complicated than matching motor power alone.
6. Where the heat comes from
An inverter may be highly efficient and still dissipate substantial power. Semiconductor losses are usually divided into two broad families.
Conduction losses occur while a device carries current. They depend on the device's on-state characteristics and operating conditions.
Switching losses occur during transitions between conducting and blocking states. Their magnitude depends on voltage, current, switching speed, frequency and the particular device.
The capacitors, inductors, rectifier and auxiliary circuits also dissipate energy. Thermal design therefore involves more than fitting a heat sink to an IGBT module. Airflow, cooling fan operation, ambient temperature, mounting conditions and cabinet ventilation influence the equipment's operating margin.
A cooling restriction that develops slowly can make a drive trip under a load that previously caused no difficulty. Conversely, an overtemperature alarm does not prove that the IGBT module itself has failed.
Diagnostic principle: Separate the heat-producing condition from the component that detects or reports it.
7. Why a motor can cause a DC-bus overvoltage alarm
An induction motor or permanent-magnet motor can return energy toward the drive when mechanical energy from the load exceeds the energy required to maintain the commanded motion.
Examples include an overhauling conveyor, a lowering mechanism, a high-inertia load during rapid deceleration, or a machine whose process momentarily drives the motor.
With an ordinary diode input bridge, this returning energy cannot simply flow backward into the AC supply. It accumulates in the DC link and raises the bus voltage unless the system can absorb or export it.
Depending on the equipment, a braking chopper and resistor may dissipate excess energy, or a regenerative input stage may return it to the mains. Drives also protect themselves by limiting operation or reporting a DC-bus overvoltage fault.
This is fundamentally different from a DC-bus undervoltage event caused by inadequate incoming energy.
Engineering distinction: Overcurrent, undervoltage and overvoltage are separate protective responses. One is not a generic substitute for the others.
8. Reading drive faults as evidence about the power path
A drive's recorded fault history is usually more useful than the simple statement that the motor stopped. The exact fault message, event sequence and stored values should be interpreted in the context of the machine.
| Indication | What it may suggest | Important alternative explanation |
|---|---|---|
| DC-bus undervoltage | Input-supply disturbance or high power demand relative to supply capability | A supply event elsewhere in the plant |
| DC-bus overvoltage | Regenerative load energy exceeding the available absorption path | A deceleration or operating-sequence condition |
| Output overcurrent | Excessive demanded torque, an abnormal motor/load condition or converter fault | An applied mechanical brake or incorrect motor data |
| Heatsink or module overtemperature | Insufficient thermal margin, high load or cooling problem | Blocked airflow or elevated cabinet temperature |
| Input phase or supply fault | Abnormal upstream power conditions | Fuse, connection or distribution problem |
| Ground-fault indication | Insulation or cable-related problem, or drive-specific leakage detection | Cable characteristics and filtering interactions |
These are diagnostic directions, not proofs of failure. Fault definitions and detection algorithms vary by manufacturer.
For safety, the analysis should begin with documented fault records, operating history and permitted non-invasive information. Electrical access and measurements inside power equipment are work for appropriately qualified personnel under the applicable procedures.
9. Reconstructed factory case: a conveyor drive that tripped only with product
Consider an illustrative conveyor installation with a three-phase induction motor and a variable-frequency drive.
The conveyor accelerated normally when empty, but periodically tripped during startup after the production line was changed. Operators described the event as an "inverter fault" and assumed that a power semiconductor was beginning to fail.
The drive's event history instead showed an overcurrent indication during the acceleration ramp. The condition correlated with loaded starts. The machine had begun accumulating more product before the conveyor was commanded to move.
The increased starting torque and inertia changed the current demand. The drive protection was responding to an operating condition that had changed outside the electronics enclosure.
The engineering lesson is not that every overcurrent alarm comes from the machine. It is that converter behaviour must be interpreted together with motor demand, mechanical loading and the production sequence. Replacing an inverter without establishing the cause may leave the original fault untouched.
10. Electromagnetic compatibility and motor-cable effects
PWM drive outputs contain fast voltage transitions. Those transitions interact with motor cables, parasitic capacitance, grounding, cable shielding and the motor's insulation system.
In a plant, the consequences may include unwanted currents, interference with analog measurement circuits or disturbance of nearby communication equipment. Long motor cables can also alter the voltage stress seen at motor terminals, depending on cable and drive characteristics.
Appropriate cable selection, equipment layout, bonding and filtering are therefore part of the complete engineering design. They must be assessed against the specific drive and motor manufacturers' installation requirements.
A converter can be electrically healthy while the installation produces an EMC problem elsewhere in the machine. For a related example, see Why Does a 4–20 mA Signal Jump When a Motor Starts?.
11. Five common misconceptions about industrial power electronics
- "A DC link always has a fixed voltage." The bus changes with supply conditions and energy flow.
- "An IGBT produces a sine wave by itself." A coordinated switching strategy produces a PWM waveform whose fundamental component drives the motor.
- "An overcurrent trip means the IGBT has failed." Motor, load, wiring, control and semiconductor faults are different possibilities.
- "A larger cooling fan always solves an overheating fault." The root cause may involve load, environment, thermal interface or the overall thermal design.
- "Any frequency inverter can regenerate into the mains." Regeneration requires an appropriate input and energy-handling architecture.
Technical conclusion
Industrial power electronics becomes much easier to understand when the machine is viewed as a controlled flow of energy rather than as an isolated "drive box."
The rectifier establishes a DC supply, the DC link buffers changing power demand, and the inverter uses semiconductor switches to deliver a controlled motor output. Their interaction explains many common factory symptoms: acceleration overcurrent, DC-bus undervoltage, braking overvoltage, heat-related trips and electromagnetic interference.
For maintenance and engineering work, the best starting point is to identify which part of that energy path is involved and correlate the converter's documented behaviour with the mechanical process.
Further technical reading
- Infineon: IGBT modules and industrial applications
- Related article: Why Does a VFD Trip During Acceleration?
- Related article: Common Causes of Overheating in Industrial Electrical Control Cabinets
Cover photograph: inside an IGBT power module, by DanielD, via Wikimedia Commons, licensed under CC BY-SA 3.0. The photograph has been resized, cropped and compressed for this article.*



