Variable Speed Drive

What is a Variable Speed Drive

A variable speed drive is a device, or a complete system, that controls the speed (and usually the torque) of a driven load by regulating the flow of energy from the power source to the machine. For an electric motor, the drive sits between the supply and the motor, regulating the power delivered so the shaft turns at the speed the process needs rather than the single speed fixed by line frequency and pole count. The same goal can be reached without electronics: a mechanical variator changes a transmission ratio, and a hydraulic drive meters fluid to a hydraulic motor. All three are correctly called variable speed drives.

The reason VSDs exist is that most rotating machines, pumps, fans, compressors, conveyors, mixers, and extruders, do not need full speed all the time. A pump throttled by a valve, or a fan choked by a damper, wastes the energy it produces and then destroys across the restriction. A VSD instead slows the prime mover to match demand, which on centrifugal loads saves energy in proportion to the cube of the speed reduction. Beyond energy, a VSD provides a controlled, ramped start that avoids the mechanical shock and the inrush current (typically 6 to 8 times rated) of a direct-on-line motor start, extending the life of belts, couplings, and bearings.

Historically, variable speed was first achieved mechanically and hydraulically. Reeves-type adjustable pulley drives and PIV (positively infinitely variable) chain variators gave stepless ratio change for machine tools and process lines long before power electronics existed; the Reeves drive on a Bridgeport milling head, for example, spans roughly 60 to 4,200 rpm by hand crank. DC motor drives then offered smooth electronic speed control but needed commutators and brushes. The decisive shift came with the AC variable frequency drive: the insulated-gate bipolar transistor (IGBT), introduced in 1983, came to dominate VFD output stages over the following two decades, making rugged, brushless, high-efficiency AC speed control affordable across the full power range.

Modern electronic VSDs span an enormous range of power. Low-voltage drives cover fractional-kilowatt OEM machinery up to roughly 5 to 6 MW at 230 V, 460 V, 575 V, and 690 V classes. Medium-voltage drives serve large motors at 2.75 kV, 3.3 kV, 4.16 kV, and 6.6 kV, reaching tens of megawatts for ID fans, large compressors, ship propulsion, and mine hoists. A single technology family therefore stretches from a 0.37 kW conveyor to a 100 MW class installation, which is why selection is always application-specific rather than generic.

Four engineering attributes determine whether a variable speed drive fits its duty: the power and current rating versus the load profile, the control method and resulting torque performance, the electrical environment it creates (harmonics, EMC, motor stress), and its serviceability over a 10 to 20 year plant life. The chapters below treat each in turn, beginning with how the broad category breaks down into electronic, mechanical, and hydraulic types.

Drive Types and Classification

Variable speed drives divide first by the physical method used to vary speed: electronic, mechanical, or hydraulic. Electronic drives dominate new installations because of efficiency and control flexibility, but mechanical and hydraulic drives remain in service and in catalogues where simplicity, very high power density, or freedom from electronics is valued. The table below summarises the three families and their typical envelopes.

Electronic drives are built from power electronics and a control processor, with no moving speed-varying parts. The AC variable frequency drive is by far the most common form; it controls a standard squirrel-cage induction motor or a permanent-magnet motor by synthesising a variable-frequency output. Most low-voltage 6-pulse IGBT VFDs are about 96 to 97 percent efficient at the drive stage, and because the same drive can ramp, reverse, brake, and communicate over fieldbus, it has displaced older DC drives and eddy-current couplings in the majority of fixed-installation duties.

Mechanical variators change the effective ratio between input and output shafts with no electronics. A Reeves (variable-pulley) drive uses two pairs of spring-loaded conical sheaves whose effective diameters move in opposition while a wide V-belt rides between them, keeping belt tension constant as ratio changes. A PIV chain drive uses a toothed link chain meshing into radially slotted cones, giving positive, no-slip transmission up to 100 kW and ratio spans around 6:1. Mechanical efficiency is highest near the 1:1 region (around 92 to 94 percent) and falls toward 82 percent at extreme ratios, where belt or chain wear also rises. These drives still suit dosing pumps, mixers, and feed sections where a robust, electronics-free stepless ratio is wanted.

Hydraulic drives include variable-fill fluid couplings and full hydrostatic transmissions (HST). A fluid coupling transmits torque through oil between an impeller and a runner; varying the oil fill varies slip and therefore output speed, which suits very large fans and conveyor starters where soft, controlled run-up of high inertia is the priority. A hydrostatic transmission uses a variable-displacement pump feeding a hydraulic motor, giving stepless speed and high torque density for mobile and marine machinery. Hydraulic drives tolerate harsh, hot, or hazardous environments but carry slip and pumping losses, so they are less efficient than a modern VFD at steady state.

A second classification cuts across all three families: by load characteristic. Variable-torque loads (centrifugal pumps and fans) need torque that rises with the square of speed and benefit most from speed reduction. Constant-torque loads (conveyors, positive-displacement pumps, extruders, mixers) need roughly the same torque at any speed. Constant-power loads (winders, spindles, some traction) need torque that falls as speed rises. Drives are rated and overload-sized differently for each, and naming the load type is the first input to any selection.

VFD Architecture and Control Methods

The electronic variable frequency drive deserves its own chapter because it dominates the category. Every low-voltage VFD shares the same three-stage power architecture, and differs mainly in the control algorithm that drives the output stage. Understanding both lets a buyer read past marketing names to what the drive actually does.

The three power stages are: (1) the rectifier, which converts the incoming AC line into DC, most commonly a 6-pulse diode bridge but optionally 12-pulse, 18-pulse, or an active front end; (2) the DC bus (DC link), where capacitors and a reactor smooth the ripple and store energy, typically holding around 540 V DC for a 400 V class drive and around 930 V DC for a 690 V class; and (3) the inverter, which switches the DC bus back into a variable-frequency, variable-voltage output. The inverter uses IGBT transistors driven by pulse width modulation (PWM), chopping the DC bus into pulses whose average follows a sine wave. The PWM carrier (switching) frequency is typically adjustable from about 2 kHz to 16 kHz on low-voltage drives, with 4 kHz a common default; higher carrier frequencies give smoother current and quieter motors but increase switching losses and motor stress.

AC motor speed is set by supply frequency, so the inverter varies frequency to vary speed. To keep the motor's magnetic flux constant as frequency falls, the drive raises voltage in proportion, the V/f relationship, until it reaches full voltage at base frequency (50 or 60 Hz); above that, frequency can keep rising in the field-weakening region to roughly 130 to 150 percent of nameplate speed at reduced torque. How precisely the drive manages flux and torque defines its control method. The table below compares the four mainstream methods.

V/f (scalar) control holds a fixed voltage-to-frequency ratio so flux stays roughly constant regardless of speed. It is the simplest and cheapest method, runs several motors from one drive, and is ideal where torque precision is not critical, fans, pumps, and conveyors. Its weakness is poor torque and stability at very low frequency, below roughly 5 Hz, where stator resistance voltage drop becomes significant; a manual or automatic torque-boost partly compensates.

Sensorless vector control builds a mathematical model of the motor from measured output currents to separate the flux-producing and torque-producing current components, then controls them independently. This delivers high starting torque, tight speed holding, and good low-speed behaviour without a physical encoder, which is why it is the default on most modern general-purpose drives. Direct torque control (DTC), pioneered by ABB, takes a different route: it regulates stator flux and electromagnetic torque directly through fast hysteresis comparators that select the optimum IGBT switching state every few microseconds, giving the quickest torque step response for cranes, elevators, and winders. Adding an encoder upgrades either method to closed-loop operation, achieving full rated torque at zero speed and the highest speed accuracy needed for positioning and tension control.

Standards, Harmonics, and Motor Effects

A variable speed drive is not a passive component: it both draws distorted current from the line and feeds a steep-edged pulse waveform to the motor. Three families of standard govern these effects, and a buyer who ignores them risks failing a power-quality audit, an EMC test, or a premature motor failure. The applicable standards are summarised below.

Line-side harmonics. A standard 6-pulse diode rectifier draws non-sinusoidal current rich in the 5th and 7th harmonics, with current total harmonic distortion (THD) approaching 100 percent at the drive terminals. IEEE 519 sets the distortion that is acceptable at the point of common coupling, where the facility meets the utility, so mitigation is frequently mandatory. A simple AC line reactor or DC choke cuts current THD to roughly 30 to 35 percent. A 12-pulse front end cancels the 5th and 7th to reach about 9 percent THD at full load; an 18-pulse front end reaches about 3.5 percent. An active front end (AFE), which replaces the diode bridge with switched IGBTs, achieves 2 to 3 percent THD at any load, corrects power factor toward unity, and allows regeneration of braking energy back to the line, at the highest cost.

EMC. The fast IGBT switching that makes drives efficient also radiates and conducts electromagnetic interference. IEC 61800-3 classifies installations into categories C1 (residential) through C4 (industrial, dedicated transformer) and specifies the EMC filtering needed for each. Most industrial drives ship with an integral or optional Category C2/C3 filter; meeting C1 for a sensitive environment usually requires an external filter and careful shielded-cable practice.

Motor stress. The inverter output is a train of PWM pulses, not a sine wave, and each switching edge has a high rate of voltage rise (dv/dt). On long motor cables the pulse reflects and can nearly double at the motor terminals, so a 460 V drive can impose peaks around 1,600 V with sub-microsecond rise times. Inverter-duty motors built to NEMA MG-1 Part 31 use reinforced insulation to survive this. On long cable runs or high carrier frequency (above roughly 5 kHz), a dv/dt filter or a full sine-wave filter is fitted to soften the edges. PWM common-mode voltage can also drive bearing currents that cause electrical-discharge-machining pitting, mitigated with insulated bearings, a shaft grounding ring, or symmetrical shielded cable.

Efficiency classes. IEC 61800-9-2 lets buyers compare drives and systems on a consistent basis. The drive alone (the complete drive module, CDM) is rated IE0, IE1, or IE2, measured at 100 percent current and 90 percent output frequency, with IE2 the most efficient. The full power drive system, motor plus drive, is rated IES0, IES1, or IES2, measured at 100 percent stator frequency and 100 percent torque; the reference system pairs an IE2 reference motor with an IE1 reference CDM. These classes underpin EU ecodesign requirements and make a documented efficiency claim auditable rather than promotional.

Key Specification Parameters

Drive data sheets list dozens of figures, but a focused set drives the selection. The most decision-relevant parameters are: power and current rating, voltage class, overload rating, output frequency range, control method, efficiency class, protection rating, and communication and safety features. Each is explained below; the comparison table that follows shows how representative low-voltage families stack up.

Power and current rating. The decisive number is rated continuous output current in amperes, not nominal kilowatts, because the same kW drive carries different current at 230 V, 400 V, 480 V, and 690 V. Manufacturers publish two ratings: a higher current for variable-torque (ND, normal duty) loads such as pumps and fans, and a lower current for constant-torque (HD, heavy duty) loads such as conveyors and crushers. Always size the drive against the worst-case continuous motor current at the chosen duty, then check the overload curve.

Overload rating. Expressed as a percentage of rated current for a defined time, typically 110 percent for 60 s on normal-duty and 150 percent for 60 s on heavy-duty, with short 180 to 200 percent peaks for a few seconds. Loads with high breakaway torque (mixers, extruders, reciprocating compressors) need the heavy-duty rating; undersizing here causes nuisance trips on every start.

Voltage class and frequency. Low-voltage classes are 200 to 240 V, 380 to 480 V (the global workhorse), 500 to 600 V, and 660 to 690 V; medium-voltage drives serve 2.3 to 6.6 kV. Input tolerance is typically +10 / -15 percent of rated voltage. Output frequency commonly ranges from 0 to 400 Hz (some drives reach 500 to 1,000 Hz for high-speed spindles), with the field-weakening region above base frequency trading torque for speed.

Other selection-critical specs:

• Control method: V/f, sensorless vector, DTC, or closed-loop, as covered in Chapter 3, dictates torque and speed accuracy.
• Efficiency class: IE2 (drive) and IES2 (system) per IEC 61800-9-2 for documented losses.
• Enclosure rating: IP20/IP21 for cabinet mounting, IP54/IP55/IP66 for wall mount in dusty or wet areas; UL Type 1/12 in North America.
• Carrier frequency: 2 to 16 kHz adjustable; higher reduces motor noise but raises losses and may require derating.
• Communication: PROFINET, EtherNet/IP, Modbus RTU/TCP, PROFIBUS DP, EtherCAT, and BACnet for HVAC.
• Functional safety: integrated Safe Torque Off (STO) to SIL 3 / PL e is now standard; safe stop and safe speed are optional.
• Braking: dynamic braking (resistor), regenerative (active front end), or DC injection, set by the load's inertia and duty cycle.

The table below compares representative general-purpose low-voltage VFD families on the headline parameters. Values are indicative of the series envelope; always confirm the exact frame rating for a specific motor current and duty against the manufacturer datasheet.

Selection Decision Factors

To turn the knowledge above into a specific model, follow the ordered sequence below. Most selection errors are not a single wrong figure but a decision taken before an earlier one is settled, so work the steps in order. These eight points double as a fixed RFQ template.

1. Load type and torque profile: First classify the load as variable-torque (pump, fan), constant-torque (conveyor, extruder, mixer), or constant-power (winder, spindle). This sets the duty rating (ND vs HD) and the control method before anything else.
2. Motor data and current: Take the rated current, voltage, power, and pole count from the motor nameplate, then size the drive to the worst-case continuous current at the chosen duty, never to nominal kW alone. Confirm motor type: induction, permanent-magnet, or synchronous reluctance.
3. Speed range and dynamics: Define minimum and maximum speed, whether full torque is needed near zero speed (which forces vector or closed-loop), and required acceleration and braking times, which size any braking resistor or active front end.
4. Overload and starting torque: Match the drive overload curve (110% / 150% for 60 s, peak 180 to 200%) to the load's breakaway and peak torque. High-inertia or high-breakaway loads need the heavy-duty rating.
5. Electrical environment: Check the supply voltage class and tolerance, then decide harmonic mitigation (line reactor, 12/18-pulse, or AFE) to meet IEEE 519 at the point of common coupling, and the EMC category (IEC 61800-3 C1 to C4) for the installation.
6. Motor protection and cable: For long cables or sensitive motors, add a dv/dt or sine-wave filter and confirm the motor meets NEMA MG-1 Part 31; plan bearing protection (insulated bearings, grounding ring) where bearing currents are a risk.
7. Enclosure, environment, and integration: Choose IP/UL rating for ambient dust, moisture, and temperature (derate above 40 to 50 degrees C and at altitude above 1,000 m). Select the fieldbus (PROFINET, EtherNet/IP, Modbus) and confirm Safe Torque Off and any further safety functions.
8. Total cost of ownership (TCO): Add purchase price, harmonic and EMC filtering, installation, commissioning, energy saved over the duty cycle, and spare-parts and service availability. On a centrifugal load the energy saving usually dominates a multi-year TCO and can justify a premium drive on payback alone.

One last dimension that is easy to overlook is serviceability over plant life: local spare-parts stock, field commissioning and repair support, firmware and parameter backup, and the availability of a compatible replacement frame in 10 to 20 years. ABB, Siemens, Danfoss, Schneider, Yaskawa, and Rockwell maintain global service and spares networks, which lowers downtime risk on critical loops; regional suppliers such as Inovance, INVT, and Delta offer strong value on non-critical duty where a fast local replacement is acceptable. For medium-voltage motors above roughly 375 kW at 3.3 to 6.6 kV, a dedicated MV drive family and a vendor with commissioning experience on that class are essential.