Motor Starting Methods: Across the Line, Soft Start and VFD

There are three families of method for getting a three-phase induction motor from standstill to full speed, and the load picks the method, not the motor. Across the line is the cheapest, has the fewest parts to fail, and gives you every bit of torque the motor can make. A soft starter trades torque for a gentler current draw and a gentler mechanical ramp. A variable frequency drive gives you the ramp plus speed control, and brings three motor-side consequences that a contactor never had. The older electromechanical reduced-voltage starters (autotransformer, wye-delta, part winding, primary reactor) sit inside the middle family and are compared in the table further down. This page starts with the work practices that have to be in place before anybody opens a panel, then covers how to read the load, the voltage-squared torque relationship that constrains every reduced-voltage method ever built, overload and short-circuit protection as two separate jobs, NEMA design letters, and the enclosure rating the controller needs where it is actually going to hang.

Shop motor controls, starters and drives

On this page: Before you open the enclosure · Start with the load · Half the voltage, a quarter of the torque · Across the line · Two devices, two jobs · Soft start · VFD · NEMA design letters · Method comparison · Enclosure ratings · Decision path · Before you call · FAQ

Before anyone opens the enclosure

Everything past this point happens inside equipment that will kill a person who assumes it is off. That covers setting an overload, changing a starter parameter, reading what a motor is actually drawing, and deciding whether a bus is dead. Establish an electrically safe work condition first, every time. Treat it as a precondition for the whole page rather than as a note at the end of it.

Start with the load, not with the motor

Every argument about starting methods is really an argument about the shape of the load's torque curve. Get that right and the rest follows. There are three shapes worth naming.

Constant-torque loads demand torque from the instant the shaft moves. Variable-torque loads demand almost none at zero speed, which is why a centrifugal fan is easy to start and a loaded screw conveyor is not. Constant-horsepower work is almost always a drive application to begin with, since the whole point is holding power while the speed changes.

Which loads have high breakaway

Breakaway torque is what it takes to get the shaft moving from dead stop, and on a constant-torque machine it commonly runs 150 to 250 percent of running torque. It is the number that kills reduced-voltage starting when nobody measured it. The usual offenders:

• A screw conveyor or drag conveyor that sat full of settled material over a weekend.
• A long inclined belt conveyor, loaded and cold, where the belt has to stretch before the tail pulley knows anything happened.
• A positive-displacement pump on cold, thick fluid, or one that has been sitting long enough for the product to set up.
• An extruder with resin that cooled in the barrel.
• A ball mill, rotary kiln or any drum that has to lift its charge up the wall before it will roll.
• A reciprocating compressor restarted against system pressure with no unloader or bypass.
• A rotary airlock with product wedged past the vane tips.

Half the voltage is a quarter of the torque

Induction motor torque varies with the square of the applied voltage. Halve the voltage and you have a quarter of the torque. Apply 65 percent and you have 42 percent. This is not a characteristic of any particular starter. It is the machine itself, and it applies to every reduced-voltage method ever built: autotransformer, primary reactor, part winding, wye-delta and solid-state alike.

Motor current, by contrast, falls roughly in direct proportion to voltage. That asymmetry is the entire trade. At 50 percent voltage you have halved the motor's current and quartered its torque.

Autotransformer line currents ignore transformer magnetizing current, which adds a few percent. Part-winding start requires a motor wound and marked for it. Wye-delta requires a six- or twelve-lead motor.

The autotransformer is the exception worth knowing, because it is a transformer and trades voltage for current the way transformers do. At a 65 percent tap the motor sees 65 percent voltage and 65 percent of its locked-rotor current, but the line sees only about 42 percent, the same fraction as the torque. A solid-state soft starter set to the same 65 percent produces the same 42 percent torque while the line still carries 65 percent of locked-rotor current. If the binding constraint is a generator, a small transformer or a utility flicker limit rather than mechanical shock, that difference decides the job.

What the square rule actually rules out

The common misreading is that reduced-voltage starting is unsuitable for anything with a real load on it. That is wrong, and the example people reach for first is backwards.

The square rule does not say you cannot start a loaded conveyor. It says you cannot start it from a low pedestal. On a solid-state starter you set the initial torque or initial current just above what the belt actually breaks away at, then ramp from there. That is a measured setting, not a fixed 50 percent tap. The constraint is where the ramp begins, not whether you may ramp at all.

Loads that genuinely rule out reduced voltage are the ones whose torque demand is at or near maximum from the first degree of shaft rotation and cannot be relieved:

• A reciprocating compressor restarted against system pressure with no unloader, no bypass and no way to blow the head down. The pistons push against full discharge pressure from the start. Unload it or start it across the line.
• A crusher, hammermill or shredder that stopped loaded or jammed. It has to break the charge before it turns at all. Reduced voltage will not, and the motor will sit at locked rotor while you find out.
• A positive-displacement pump against a closed discharge or a plugged line. Full pressure differential at zero flow, and no amount of ramping changes it.
• Any load whose accelerating time under the ramp would exceed the motor's safe locked-rotor time. Winning the torque argument and losing the thermal one is still losing.

There is a fifth case that is subtler and catches more people. A NEMA Design B speed-torque curve is not flat. It dips between locked rotor and breakdown at the pull-up minimum, typically somewhere in the 20 to 50 percent speed range. Reduce the voltage and the whole curve scales by the square, dip included. The motor breaks the load away, accelerates into the pull-up region, and stalls there with the load curve sitting above it. It presents as a starter problem and it is not one. Plot the reduced-voltage motor curve against the load curve at every speed, not just at zero.

Across the line: full voltage, full torque, full inrush

Across the line (direct on line, or a full-voltage non-reversing starter) is a contactor, an overload relay and short-circuit protection ahead of it. It is the cheapest way to start a motor, and it delivers every bit of torque the motor is capable of producing. In most plants it remains the right answer for the majority of motors, and the discipline is to justify moving away from it rather than to justify staying.

What you pay for that is inrush. Locked-rotor current on a general-purpose motor runs roughly six to eight times full-load amps, and it is present for the whole of the acceleration, not just for an instant.

You do not have to guess at the number. The nameplate carries a locked-rotor indicating code letter, and the NEC code-letter table, Table 430.7(B), converts it to locked-rotor kVA per horsepower. From there:

Locked-rotor amps = (code kVA/HP x HP x 1,000) / (1.732 x volts)

Worked: a 25 HP, 460 V motor with code letter J, taken at the top of its band at 7.99 kVA per horsepower, gives 7.99 x 25 x 1,000 / (1.732 x 460) = about 251 A. The NEC full-load current table value for 25 HP at 460 V is 34 A, so the ratio is about 7.4 to 1. That is what the transformer, the conductors and the breaker see every time somebody presses start.

That current does three things. It dips the bus voltage, which is why the lights flicker and why the other motors on the same transformer notice. It heats the rotor. And it applies full-voltage torque to the driven machine as a step function, which is why gear teeth, chain, coupling elements, shear pins and conveyor splices are the parts that tell you a plant starts everything across the line.

Two protective devices, two completely different jobs

This is the most commonly muddled part of a motor circuit, and the muddle burns motors.

The overload relay protects the motor from sustained overcurrent: a load that crept up, a bearing going, a belt over-tensioned, a fan that picked up buildup, a lost phase. It is a thermal model of the motor, and it is sized off the motor nameplate full-load amps. NEC 430.6(A)(2) is explicit that overload protection is based on the nameplate current rating, not on a table value. NEC 430.32(A)(1) sets the ceiling for continuous-duty motors over 1 HP: 125 percent of nameplate full-load current for motors with a marked service factor of 1.15 or greater, or a marked temperature rise of 40 deg C or less, and 115 percent for all other motors. If the motor will not start or trips at that setting, 430.32(C) permits an increase to 140 percent and 130 percent respectively, and no further. Those are ceilings, not a troubleshooting sequence. A relay that still trips at the permitted maximum is reporting something real: a load that has changed, a bearing on its way out, a lost phase, a motor that is simply too small for the job. Raising the setting past what the Code allows only moves the failure out of the relay and into the winding.

The breaker or the fuses ahead of the starter protect the circuit from short circuits and ground faults. They are sized from the table in NEC 430.52 as a multiple of the motor full-load current taken from the NEC tables in 430.247 through 430.250, per 430.6(A)(1), deliberately not from the nameplate. For a three-phase squirrel-cage motor the permitted multiples are large: an inverse-time breaker up to 250 percent, a non-time-delay fuse up to 300 percent, a dual-element time-delay fuse up to 175 percent, all so the device rides through inrush without opening. Read the row for the motor type you actually have, out of the Code edition your jurisdiction enforces. That table has carried more than one number over the years.

One more item at the assembly level: NEC Article 409 covers industrial control panels and requires the panel to be marked with a short-circuit current rating, which has to equal or exceed the available fault current where it is installed. Substituting a component inside a listed combination starter can invalidate that marked rating even when the replacement is electrically equivalent.

Trip class 10, 20 and 30

Trip class names the maximum time an overload relay will permit at six times its current setting, starting from cold. That is the NEMA definition. IEC-rated relays use the same class numbers but define them at 7.2 times the setting, so a Class 10 on one scale is not the identical curve on the other. Check which standard the relay is built to before carrying a class number across a substitution.

Pick the lowest class the load will start on. Class 30 is not a forgiving setting to be generous with. It is thirty seconds of near-locked-rotor current the relay has agreed to permit, and on a motor whose rotor is finished at twelve seconds hot it is not protection at all. Get the safe locked-rotor time, hot and cold, from the motor data sheet, and confirm the relay curve sits under the motor thermal damage curve across the whole range rather than crossing it somewhere in the middle.

Electronic overload relays are worth the difference over bimetallic on anything that matters: selectable trip class, phase-loss and phase-imbalance detection, ground-fault and jam detection, and a current readout that tells you what the motor is drawing today versus what it drew at commissioning. Single-phasing is one of the fastest ways to lose a three-phase motor, and it does not always raise the remaining phase currents enough, fast enough, to move a bimetallic element in time.

Soft start: less current and less shock, no speed control

A solid-state soft starter is three pairs of back-to-back SCRs that phase-control the voltage to the motor, ramping it from a set initial pedestal up to full voltage over a set time, then closing a bypass contactor so the SCRs stop making heat for the rest of the run.

What it does

• Reduces line current during acceleration, which is what the utility, the generator or the upstream transformer cares about.
• Removes the torque step. The driven machine sees a ramp instead of a slam. This is the reason the device exists, and it is what saves conveyor splices, gear teeth, coupling elements, chain and shear pins.
• Gives a controlled stop on the way down. On a centrifugal pump this is often worth more than the soft start, because a coasting pump letting a check valve slam is what cracks pipe and breaks pump feet.
• Fits in far less panel space than an autotransformer starter, with no tap contactors and no transition to time.

What it does not do

• It does not give you speed control. When the ramp finishes, the motor is across the line at whatever speed the load and the slip settle on. If anyone in the conversation wants to run at 40 Hz, this is the wrong device and no amount of setup will change that.
• It does not reduce the torque the load demands. If the machine needs 90 percent of full-voltage locked-rotor torque to break away, the starter has to deliver essentially full voltage to break it away, and you have bought very little.
• It does not save energy while running. Once bypassed it is a contactor.
• It does not reduce motor heating during acceleration. It increases it.

Ramp time is bounded by the rotor, not by the ramp setting

This is the point most often missed. During acceleration the rotor bars heat essentially adiabatically, since the start is over long before that heat conducts anywhere. What the rotor spends is a current-squared-times-time budget. Stretching a six-second across-the-line start into a thirty-second soft-start ramp does not make the start gentler on the motor. It runs a lower current for five times as long, and since the current came down only in proportion to voltage while the time went up by a factor of five, the rotor almost always finishes hotter than it would have.

The bound is the motor's safe locked-rotor time, quoted on the data sheet as safe stall time, hot and cold, and commonly somewhere in the range of 8 to 20 seconds for a general-purpose TEFC motor. Total accelerating time under the ramp has to finish inside that with margin, and the hot number is the one that matters, because the second start of the shift is the one that kills the motor. An over-long ramp cooks the rotor: bars soften and distort, bar-to-end-ring joints crack, and the motor comes back from the rewind shop with a broken rotor bar nobody can explain from the load data.

Set it in this order:

• Set the initial torque or initial current just above what the load actually breaks away at. Watch the shaft, not the nameplate, from outside the guarding and with the enclosure closed.
• Use the shortest ramp that brings the mechanical shock down to where you wanted it. Longer is not safer.
• Use the current-limit setting as a ceiling, not as the primary control. A starter sitting in current limit is a motor sitting near locked rotor.
• Check total accelerating time against safe stall time, hot, with margin.
• Check the starter's own duty rating. SCR stacks and bypass contactors have thermal limits and a published starts-per-hour figure at a stated current and ramp time. A unit rated for two starts an hour is a different device from one rated for ten.

Open-chassis units for mounting in an existing panel are listed under chassis soft starters (52 available to order). If the starter needs its own rated box, the packaged versions are covered on enclosed VFD and motor control packages.

VFD: speed control, and the three things that come with it

A variable frequency drive rectifies incoming AC to a DC bus and inverts it back into a synthesized variable-voltage, variable-frequency output. Because it holds the volts-per-hertz relationship it keeps the motor's flux constant, so rated torque stays available all the way down to very low speed, and down to zero speed on a closed-loop or sensorless-vector drive. That is what makes it both a speed control and the most capable starting method available. Two limits belong in the same breath. Above base speed the drive runs out of voltage and can no longer hold volts per hertz, so flux falls and available torque falls roughly inversely with speed from there, through the constant-horsepower region. And the drive's own overload rating is the ceiling on starting torque, commonly 150 percent for 60 seconds on a heavy-duty rating and only 110 percent for 60 seconds on a normal-duty one, so a drive chosen on running amps alone may not start the machine at all. A correctly sized drive starts a load at 100 to 150 percent torque while drawing something close to that same percentage of full-load current from the line, not six to eight times it.

On a variable-torque load the energy case frequently carries the purchase on its own. Flow varies with speed, pressure with the square of speed, and shaft power with the cube. A fan at 80 percent speed is drawing roughly half the power. Throttling a damper or a control valve to reach the same flow throws that difference away as heat and noise.

Then the drive hands you three motor-side problems the contactor never had. All three are manageable. All three are ignored often enough to be worth stating plainly.

1. Reflected wave and dV/dt stress on the winding

The drive output is a train of fast-rising DC bus pulses, not a sine wave. The cable and the motor are a mismatched transmission line: the cable surge impedance is on the order of tens of ohms and the motor's is far higher, so each pulse partially reflects at the motor terminals. When the cable is long enough that the round-trip travel time exceeds the pulse rise time, the reflected pulse adds to the incident one and the terminal voltage approaches twice the DC bus. On a 480 V system the bus peak is about 679 V, so a full reflection puts roughly 1,300 to 1,400 V on the motor terminals, and bus overvoltage during deceleration pushes it higher still.

The stress is not distributed evenly. A fast edge does not divide equally across the winding, so the first turn of the first coil takes a disproportionate share of it, which is why reflected-wave failures show as turn-to-turn shorts near the lead end rather than as general insulation aging.

The critical cable length is set by the pulse rise time and the cable's propagation velocity, so faster switching devices produce full doubling over shorter runs. Do not use a remembered number. Use the drive manufacturer's published maximum lead length for your carrier frequency, cable type and voltage, and then:

• Specify a motor whose insulation system is rated for it. NEMA MG 1 Part 31, which covers definite-purpose inverter-fed motors, calls for withstand of a peak voltage of 3.1 times rated line-to-line voltage with a rise time no shorter than 0.1 microsecond (about 1,488 V peak on a 480 V motor). General-purpose motors are built to a materially lower withstand than that, which is exactly why a general-purpose motor on a long lead run is a short story.
• Shorten the leads if you can. Locating the drive near the motor beats every filter.
• Add an output reactor, a dV/dt filter or, for very long runs, a sine-wave filter. Lowering the carrier frequency helps drive heating but does not slow the edge.

2. Reduced cooling at low speed

A totally enclosed fan-cooled motor is cooled by a fan bolted to its own shaft. Slow the shaft and you slow the fan. On a constant-torque load the motor draws roughly the same current at 20 Hz as it does at 60 Hz, so the losses stay where they were while the cooling airflow falls away underneath them.

Variable-torque loads largely protect themselves here, because torque falls with the square of speed and the heating falls with it. The problem is a constant-torque problem: conveyors, positive-displacement pumps, extruders, mixers, hoists. The fixes:

• Specify the constant-torque speed range, in writing. Motors sold as inverter duty are commonly rated for a 20:1 constant-torque range, but that figure is a market convention rather than something a standard hands you. MG 1 Part 31 is an insulation and construction standard for inverter-fed motors, not a turndown rating. A general-purpose TEFC motor is usually good for far less. Inverter duty printed on a nameplate is not a specification by itself. Ask for the constant-torque and variable-torque turndown ratios and get them onto the order.
• Use a separately powered blower. A totally enclosed blower-cooled motor runs its own cooling fan from the line at constant speed regardless of shaft speed, which removes the problem entirely and is the standard answer on wide-turndown constant-torque drives.
• Oversize the frame, or restrict the low-speed operating range in the drive parameters so the operator cannot park the machine at 10 Hz all shift.

3. Shaft voltage and bearing currents

This is the most common VFD-attributable motor failure in the field, and the one that gets misdiagnosed as a bad bearing batch.

The three PWM phase voltages do not sum to zero at every instant the way three sine waves do, so the motor's electrical neutral swings against ground at the carrier frequency. That common-mode voltage couples capacitively across the air gap onto the rotor. When the resulting shaft voltage exceeds the dielectric strength of the grease film in the bearing (a few volts to a few tens of volts, and lower as the film thins with load, speed and temperature), it punches through and an electrical discharge melts a small crater in the race. The bearing has become a spark gap.

The progression is recognizable once you have seen it:

• Millions of discharges produce first a frosted, grey, matte appearance on the race where it should be mirror bright.
• Then fluting: evenly spaced transverse ridges running across the race, a washboard pattern that is unmistakable and that no mechanical cause produces.
• The grease darkens and carbonizes, going from tan to grey to near black, loaded with melted metal debris that then works as an abrasive.
• Audibly it starts as a whine or a growl that rises over weeks or months, and it is often blamed on alignment or on the bearing supplier before anybody puts a scope on the shaft.

Confirming it takes an oscilloscope and a shaft-riding brush with the machine running: the sawtooth charge-and-discharge waveform on the shaft is diagnostic. That is a live measurement on rotating equipment: qualified persons only, guarding in place, probe reference bonded to the motor frame. The fixes are well established and they are cheap relative to a rewind:

• Shaft grounding ring on the drive end, a conductive microfiber ring that gives the shaft current a path to frame with lower impedance than the bearing. This is the single highest-value item on the list.
• Insulated or hybrid ceramic bearing on the non-drive end on larger frames. Do not insulate both ends without also grounding the shaft: with nowhere else to go, the current leaves through the coupling and destroys the bearings in the gearbox or the pump instead, which is a far more expensive repair than the one you avoided.
• Symmetrical shielded VFD cable with the shield terminated 360 degrees at both the drive and the motor. A pigtail ground is not a high-frequency termination and the common-mode current does not care what it was called on the drawing.
• A low-impedance bond between motor frame and drive enclosure. Building steel and conduit threads are not a high-frequency return path.
• Common-mode chokes on the drive output where cable practice cannot be fixed.

One line-side note. The six-pulse rectifier front end on a standard drive draws non-sinusoidal current and injects harmonic current back at the supply. IEEE Std 519 is the standard covering harmonic control in electric power systems, and its limits apply at the point of common coupling with the utility, not at the drive terminals, which is where it is most often misapplied. A handful of drives on a stiff service rarely matters. Converting most of a plant's connected load to drives, or running drives on a generator, does, and the remedies are line reactors or DC bus chokes, multi-pulse or active front-end drives, or a harmonic filter at the service.

Open-chassis drives for panel mounting are listed under chassis variable frequency drives (77 available to order), and the full control line including starters, contactors and overloads is in motor controls (1,058).

NEMA design letters A, B, C and D

The design letter on the nameplate is a description of the motor's speed-torque curve, and it decides whether the motor will start the load at all. It is the field most often dropped when somebody crosses a motor to a replacement, and dropping it is how a machine that used to start reliably stops doing so.

How that interacts with the starting-method choice:

• High breakaway plus a nuisance voltage dip is not automatically a soft-start job. A Design C motor started across the line gives you 200 percent-plus locked-rotor torque at normal locked-rotor current, which is the opposite of what reduced voltage does to you. Sometimes the cheaper fix to a hard start is the motor, not the starter.
• Design A's higher locked-rotor current is why Design A motors so often sit on reduced-voltage starters. Size the starter, breaker and conductors against the motor's actual locked-rotor current, not against a Design B assumption.
• Design D exists so the motor can slip. On a flywheeled press the slip is the shock absorber. Substituting a Design B and expecting the machine to behave the same way does not work.
• Design E was a high-efficiency class whose inrush caused starter compatibility problems. It is effectively out of the field, and the efficiency levels it chased now come from ordinary premium-efficiency Design B motors.

On a VFD, the design letter stops governing the start. The drive sets voltage and frequency and holds the motor near its own torque-per-amp optimum, so a Design B motor on a correctly sized drive develops full-load torque or better from a standstill and does it without the inrush. The drive's overload rating, not the motor's design letter, sets how much is actually on offer. Design B is the right choice for drive service. The high-resistance rotor construction that gives C and D their starting torque also gives them higher rotor loss and poorer performance on a drive. When you are specifying a motor that will live on a drive, spend the specification on the insulation system, the cooling arrangement and bearing protection rather than on the design letter. Browse the motor line in electric motors (2,153 in the collection; the motor line runs 0.25 to 600 HP) or narrow by frame, RPM, voltage and enclosure with the electric motor finder.

Comparing the three methods

Enclosure ratings for the controller: 1, 12, 3R, 4 and 4X

The starter or drive is only as good as the box it lives in, and the rating question comes up after the starting-method question because a VFD in a sealed enclosure is a different thermal problem than a contactor in one.

NEC Table 110.28 lists which enclosure types are acceptable against each of these conditions in non-hazardous locations, and it is the fastest way to settle an argument about whether a 3R is good enough for a given spot. In classified areas the area classification governs the whole installation, and the equipment rating has to meet or exceed the Class, the Division or Zone, the gas or dust Group and the temperature code. Meet or exceed: a Division 1 rated device is acceptable in a Division 2 area, and the temperature codes run in the opposite direction from what most people expect, with a lower T-number meaning a higher permitted surface temperature.

Pre-engineered assemblies with the drive or starter already mounted, wired and rated are covered on enclosed VFD and motor control packages. If you are assembling your own panel, start from the chassis drives and chassis soft starters and specify the enclosure separately against the conditions above.

A decision path you can run in five minutes

Work down the list. The first one that fits is your answer.

• Does anyone need to vary speed, flow, pressure or rate during normal operation? VFD. Nothing else does this, and every workaround costs more over five years than the drive did.
• Is it a centrifugal fan or pump that is currently throttled by a damper or a control valve? VFD. Power falls with the cube of speed, and you are presently converting the difference into heat and noise at the damper.
• Is the complaint mechanical: splice failures, broken gear teeth, sheared pins, chain wear, water hammer, spilled material at a transfer, gearbox backlash slam? Soft start, unless you also want speed, in which case VFD. A loaded belt conveyor, a bucket elevator and a long screw conveyor are all textbook cases here.
• Is the complaint electrical: a voltage dip at every start, a generator that will not hold, a transformer at its limit, a utility flicker limit? Reduced voltage of some kind. Compare an autotransformer starter against a solid-state starter on line current at equal torque before choosing, then verify the reduced-voltage torque curve stays above the load curve at every speed, including the pull-up dip.
• Is the load high breakaway, started against real load, with no way to unload it? Across the line, and consider a Design C motor. Do not soft start a reciprocating compressor against pressure, a jammed crusher or a positive-displacement pump against a closed discharge.
• Does the machine start many times per hour? VFD. Both other methods are limited by accumulated inrush heating: the motor's rotor on across the line, the SCR stack and bypass on a soft starter.
• None of the above, and the motor is small relative to the service? Across the line. It is the cheapest and simplest option, and it has the fewest ways to fail. There is no prize for over-specifying a 5 HP fan motor.

One thing that is true regardless of which branch you land on: the drivetrain downstream deserves the same look. A starting method that is gentle on the motor and brutal on the reducer has not solved anything. Ratio, service factor and shock-load class are covered on speed reducers.

What to have in front of you before you call

Ninety percent of the back-and-forth on a starter or drive quote is missing nameplate data. Photograph the motor nameplate before anything comes apart, and have these ready:

• From the motor nameplate: horsepower, full-load amps, voltage and phase, RPM, frame, enclosure, service factor, insulation class, NEMA design letter, locked-rotor code letter, and duty. The full-load amps and the design letter are the two that get dropped most often and they are the two that matter most here.
• From the driven machine: what it is, whether it starts loaded or unloaded, breakaway behavior if anyone has watched it, and starts per hour.
• From the installation: supply voltage and available fault current, distance from the controller to the motor in feet, indoor or outdoor, wash-down or dust, ambient temperature, and whether the area is classified.
• From the existing panel: what is there now, what it is protected by, and what has been failing.

Texas Belting and Supply ships motors, starters, drives and reducers from our Houston warehouse and supplies plants, OEMs and repair shops across the Texas Gulf Coast and nationwide. Everything in the motor control line is available to order. Call (888) 203-2358 with a part number or a nameplate photo and we will confirm the fit and the lead time.

Frequently asked questions

Related guides and collections

Last updated: August 2026