Phase rotation and motor direction field guide

Phase rotation, and what it decides

Phase rotation is the order in which the three phases of a three-phase supply reach their peak voltage. The three voltages are a third of a cycle apart, and they crest in a fixed order, either A-B-C or A-C-B. That order is the whole story, because it sets the direction the rotating magnetic field sweeps around the inside of a motor, and the rotor follows the field. Change the order and the motor turns the other way.

People use phase rotation and phase sequence to mean the same thing, and you can treat them as one. The phases get called A-B-C, L1-L2-L3, R-S-T, or U-V-W depending on the drawing and the country, but the idea does not change. The supply has a sequence, the motor follows it, and the driven equipment, the pump or the fan or the compressor, only does its job when the shaft turns the way the maker intended.

This sits on top of two ideas worth keeping straight. Three-phase power itself, the wye and delta connections and the voltages, is covered in the three-phase guide. How that motor gets up to speed, across the line or through a soft starter or a drive, is covered in the motor-starting guide. This guide is about one thing on top of both: which way the shaft turns, and how you make sure it turns the right way before anything is coupled to it.

What the wrong rotation actually costs

Backward rotation is not a cosmetic problem. On the wrong equipment it does damage in the time it takes to walk back to the disconnect.

A centrifugal pump run backward still throws some liquid out by sheer centrifugal force, so it looks like it is pumping, but the flow and pressure fall off hard, commonly to a fraction of rated, and on a threaded impeller the nut can back off and the impeller can spin loose in seconds. A fan run backward keeps moving air in roughly the right direction at a fraction of the volume, which is the most deceptive case of all, because it looks and sounds like it is running. The scroll or screw compressor is the one that bites hardest. Run a scroll backward and it does not compress, it overheats, and sustained or repeated reverse running ruins it, sometimes within minutes. A conveyor or a machine tool driven the wrong way can wreck product, jam, or hurt someone before anybody reaches the stop.

Notice the pattern. The pump and the fan fool you by half-working. The compressor punishes you fast. None of them announce the problem in a way a quick glance catches, which is exactly why the check has to happen before the equipment is coupled and started, not after the call comes in that something runs wrong.

What is phase sequence, ABC vs ACB?

Phase sequence is the order the three phases peak, and there are only two possibilities that matter: A-B-C or A-C-B. A-B-C is the standard, the one most systems are set up to, and it is sometimes called positive sequence. A-C-B is the reverse, sometimes called negative sequence. Everything else is the same order written from a different starting point.

That last part trips people. A-B-C, B-C-A, and C-A-B are all the same sequence, because the order is a loop and it does not matter which phase you name first. The opposite loop, A-C-B, is the same as C-B-A and B-A-C. So when a rotation meter reads CBA, that is the reverse of ABC, not some third option. There is the standard direction and the reverse direction, full stop.

The convention on a job is to keep A-B-C consistent from the service through the gear to the equipment, so that a motor landed the same way every time turns the same way every time. The trouble starts when that consistency breaks somewhere upstream and nobody knows, which is why you verify the sequence rather than trust that it was kept.

How motor direction is set by the connection

The direction a three-phase motor turns is set by which line lands on which motor terminal. A standard three-lead motor has terminals marked T1, T2, and T3, and the supply lines are L1, L2, and L3. Land L1-L2-L3 onto T1-T2-T3 and the motor turns one way. Change which line goes to which terminal and you can change the direction.

The key fact, the one the whole reversal trick rests on, is that swapping any two of the three connections reverses the sequence the motor sees, and so reverses the rotation. Swap all three in a rotated pattern and nothing changes, because that is the same loop. Swap exactly two and you flip it. That is covered in its own section because it is the fix you will reach for most.

Worth saying plainly: NEMA marks the leads and standardizes the terminal nomenclature, but it does not hand you a universal rule that A-B-C onto T1-T2-T3 always means clockwise for every three-phase motor. The actual direction depends on the motor's internal connection, and the maker marks the expected direction on the nameplate where it matters. So the connection determines rotation, the nameplate tells you what to expect, and the meter or the bump test confirms it. You do not assume.

This is why wire color and labeling do not set rotation. Keeping A-B-C the same color through the system is good practice that speeds the next person's work, but rotation depends on which physical conductor lands on which terminal and the sequence it actually carries, not on the color of the tape. A conductor mislabeled upstream, a feed brought in on a different color convention, or a re-pull where the marks did not follow the wire can all put the wrong phase under the right-colored tape. The motor follows the sequence, not the label. Keep the identification consistent, and still confirm rotation with a meter or a bump, because the consequence of trusting the color is a reversed compressor.

How do you reverse a three-phase motor?

To reverse a three-phase motor, swap any two of the three line leads feeding it. That is the entire method. Pick any two of L1, L2, L3 at the motor connection or at the load side of the starter, trade their positions, and the rotation flips. It does not matter which two you pick, because every two-lead swap turns the sequence into its reverse.

The common shop habit is to swap L1 and L3 and leave L2 alone, mostly so everyone does it the same way and the next person can read the change. That is a convention, not a requirement. Swapping L1 and L2, or L2 and L3, reverses the motor exactly the same.

Do the swap at a clear, documented point. Doing it at the motor terminal box is clean and local to the one motor. Doing it at the starter output affects only that motor too. What you do not want is to swap phases somewhere upstream of other equipment, because then you have reversed everything fed from that point, not just the motor you were chasing. Reverse the one motor that is wrong, label what you changed, and leave the rest of the system on its original sequence.

The phase-rotation meter

A phase-rotation meter is the tool that reads sequence directly. You connect its three leads to the three phases, A to one, B to the next, C to the last, energize, and it tells you whether the sequence is ABC or CBA, usually with a display or a pair of arrows or rotating lights. It answers one question fast: is this supply on the standard sequence or the reverse.

The everyday use is on the line side. Before you land motor leads, clip the meter to the three phases at the disconnect or the starter and confirm the supply sequence is what the job expects. After a service change or a new feed, the same check tells you in seconds whether the source still rotates the way it used to. Many of these meters also flag a missing or open phase, which is a useful second job from one tool. Typical units read sequence on systems up to a few hundred volts AC, so confirm the meter's rating against the system you are on.

Reading the line sequence is not the same as knowing which way a given motor will spin, because that also depends on how the motor is wound and landed. The line-side meter confirms the supply. To confirm the machine, you want the motor-rotation function or a bump test, covered next.

The motor-rotation tester, no power needed

A motor-rotation tester checks which way a motor is set to turn without ever energizing it. You connect its leads to the motor's T1, T2, T3 terminals on a dead, disconnected motor, then turn the shaft by hand in the direction you want. The tester reads the small voltage the spinning rotor generates and tells you whether that hand-turn matches an ABC connection or a reversed one. You learn the motor's direction before a single line is landed and before anything is energized.

This is the tool that lets you permanently connect and tape up the motor leads, then verify, instead of making a temporary hookup just to bump it and see. That matters most where a temporary energized hookup is slow, costly, or hazardous, which is most medium and large motors. Combined with a line-side sequence check, you can match a known-good supply sequence to a verified motor and land it once, correct.

There is a ceiling. Above 600 V the manual no-power method runs out of road, and the bump test becomes the practical way to confirm a large machine. Match the tool to the voltage and read the maker's limits before you trust the readout.

How do you check motor rotation before starting?

You check motor rotation before starting by confirming the direction with the motor uncoupled or with a meter, never by coupling the load and hitting start to see what happens. The order that keeps equipment alive is verify first, couple second, run third.

There are three practical ways, and you pick by the equipment. Use a motor-rotation tester to read direction with no power at all, the safest and the one that needs nothing energized. Use a phase-rotation meter on the supply to confirm the line sequence, then match it to a motor you know is wound standard. Or uncouple the load and do a quick bump test, watching the bare shaft. On a direct-coupled pump or compressor where uncoupling is real work, the no-power tester earns its keep, because you confirm rotation without ever spinning the load the wrong way.

The rule that gets ignored and then regretted: do not energize a coupled pump or compressor just to check rotation. The few seconds of reverse running while you watch the shaft is exactly the few seconds that can spin an impeller loose or start cooking a scroll. Confirm the direction before the load is in the path.

The bump test

A bump test is a momentary jog of the motor to see which way the shaft starts to turn. You energize it for a fraction of a second, just long enough to watch the shaft break and pick a direction, then drop it. It is the oldest field method and it works, with one condition attached.

Do it uncoupled, or do it knowing the load can take a brief reverse without harm. On a motor decoupled from its pump or compressor, a bump is clean: the shaft moves, you read the direction off the fan or a paint mark, and nothing downstream is at risk. On equipment that is already coupled, a bump is a gamble, because even a short reverse spin can loosen an impeller or stress a compressor. That is the whole reason the no-power tester exists.

Read the direction against the rotation arrow on the equipment, not against your gut. Stand where you can see the shaft end or the fan, give it the briefest jog, and watch the first motion. If it is wrong, kill power, lock out, swap two leads, and bump again to confirm before anything is coupled back up.

Pump rotation

A centrifugal pump is built to turn one way, and the maker casts or prints a rotation arrow on the casing or the nameplate. Match the shaft to that arrow. Run it backward and the pump does not simply stop, which is what makes it dangerous to assume it is fine.

Backward, the impeller still slings liquid outward by centrifugal action, so you get some discharge and the gauge shows some pressure. It just falls well short of rated, often landing at a fraction of normal flow and head. The system looks like it is working and underperforms quietly, and the operator chases a flow problem for weeks. Worse, on a pump with a threaded impeller the reverse torque can back the retaining nut off, and the impeller can spin loose and chew up the pump in seconds. Reverse running also drives cavitation and runs the seals and bearings hot for lack of proper flow.

So the pump is the half-working trap. Check the arrow, confirm rotation uncoupled or with the no-power tester, and only then couple and run. A pump that has been spun backward even briefly deserves a look at the impeller and the nut before it goes back in service.

Fan rotation, the deceptive one

A fan running backward is the easiest wrong rotation to miss, because it still moves air. A reversed centrifugal fan or blower keeps pushing air in roughly the intended direction, since the housing and the outlet do not change, but the blade angles are now working against the design, so the volume and the pressure drop off hard. It is running, it is moving air, and it is moving far less than it should.

That is the trap. Nobody walks up to a spinning fan that is blowing air and suspects the rotation. They suspect a dirty filter, a closed damper, a duct problem, anything but the motor turning the wrong way. On a propeller fan the reversed unit moves much less air; on a squirrel-cage blower the reversal will not flip the airflow direction but it guts the performance.

Check the rotation arrow on the fan housing or scroll and confirm the wheel turns that way. On an air-handling unit the rotation can be hard to see through the housing, so confirm it at the motor or with a meter rather than guessing from the airflow you feel at a register. A fan that moves some air is not proof of correct rotation.

What happens if a compressor runs backward?

If a scroll or screw compressor runs backward it does not compress, it heats up, and sustained or repeated reverse running can ruin it, sometimes in minutes. This is the most damaging wrong-rotation case in the trade, and it is why refrigeration and HVAC startups treat rotation as a hard gate, not a nicety.

On a scroll, forward rotation walks the gas inward and compresses it. Backward, the scrolls cannot build pressure, the motor draws hard, the unit gets loud, and the internal protector eventually trips on overload. A brief reverse from a quick power blip is usually survivable, the makers say as much, but a compressor that is repeatedly restarted and run in reverse, or that runs reversed for a stretch on a miswired three-phase hookup, takes permanent damage. Screw compressors are similar in spirit: built for one direction, harmed by the other.

So on any three-phase compressor, confirm rotation before the first start, and protect it from the source changing later. The makers of larger scrolls build phase monitoring and a timed lockout into the protection module for exactly this reason. If your compressor does not have that built in, that is the strongest argument for a phase-monitor relay ahead of it, which is the next section.

The phase-monitor relay

A phase-monitor relay, also called a phase-sequence relay or phase-failure relay, watches the three phases and drops the control circuit when something is wrong before the wrong condition reaches the motor. It is the standing guard that catches a rotation problem you did not see coming, especially one introduced after commissioning.

A typical unit trips on reverse phase sequence, on the loss of a phase, and on voltage imbalance between phases, and many add undervoltage and overvoltage. Wire its contact into the motor control circuit so that any of those faults stops the starter from pulling in. The value is most obvious on a compressor: if the utility or a service change flips the sequence months after startup, the relay catches the reversal and locks the compressor out instead of letting it run backward and self-destruct.

This is the cheap insurance on the expensive machine. A compressor or a critical pump that has no built-in phase protection should have a phase-monitor relay ahead of it, set per the equipment, so that a source change cannot quietly reverse the rotation and the equipment never starts wrong. Verify the trip thresholds against the equipment's requirements; the device protects only as well as it is set.

Where the rotation really comes from: the source

The rotation of everything in a building starts at the source. The utility or the upstream supply sets the phase sequence, and your gear, your feeders, your motors all inherit it. As long as that source sequence holds, a motor landed the same way turns the same way for years. The problem is that the source can change without anyone telling the people downstream.

Utility work, a transformer swap, a re-termination at the service, a new feed brought in for an expansion: any of these can land the phases in a different order than before. When that happens, every three-phase motor on the system reverses at once. Crews chase it as a dozen separate equipment faults when it is one upstream change. The lesson the field teaches is to re-verify rotation after any work on the service or the source, not to assume the sequence survived.

On a new service, a temporary power setup, or a generator, the sequence is not guaranteed to match anything until you check it. Land it, meter the sequence, and confirm it against what the building expects before the first motor runs. A new feed that comes in reverse is easy to correct at the service while the gear is dead; it is expensive to discover after a compressor has run backward.

Generators deserve their own line on this. A standby or temporary genset has to present the same rotation as the utility it backs up, or equipment runs one way on utility power and the other on the generator. With an automatic transfer switch the load swings between the two sources, and if their sequences do not match, motors reverse on every transfer, a problem that hides until the first switchover. Meter both sources, make them agree before the gear is put under load, and where generators are paralleled, matching rotation is one of the conditions for tying together, alongside voltage, frequency, and phase angle. How the motor itself gets started off either source is in the motor-starting guide.

Single-phase motors reverse differently

Single-phase motors do not have a three-phase sequence to swap, so the swap-any-two rule does not apply to them. Their direction is set by the relationship between the run winding and the start winding, and you reverse them by swapping the leads of the start winding relative to the run winding, per the motor's connection diagram. Not every single-phase motor is field-reversible, and the nameplate or terminal diagram tells you whether and how.

This matters mainly so nobody tries to fix a backward single-phase motor by swapping line and neutral, which does nothing useful. Read the connection plate. If it shows a reversible arrangement, follow it; if it does not, the motor turns the way it turns. For NEMA single-phase machines there is a defined standard direction tied to the connection, which is the opposite situation from three-phase, where the direction is not fixed by sequence alone.

The VFD sets rotation by programming

A variable frequency drive controls the rotation of the motor it runs, and you can set the direction two ways. You can swap two of the three output leads from the drive to the motor, the same physical trick as a line-fed motor, or you can change it in the drive's parameters, a direction setting that flips the rotation without touching a wire. On a drive, the parameter is usually the cleaner move, because nothing in the power wiring changes and the setting travels with the program.

The input rotation to the drive does not set the motor's direction. A VFD rectifies the incoming three phases to DC and synthesizes its own three-phase output, so the line sequence feeding the drive is largely irrelevant to which way the motor turns. The drive's output is what spins the motor, and the program governs that output. This is different from a contactor-fed motor, where the line sequence is everything.

Still verify the actual shaft direction at commissioning. Set the parameter or land the leads for the direction you want, then confirm the shaft turns that way against the equipment arrow before the load is coupled. The drive does what it is told; what it was told is what you check. How drives compare to other starting methods is in the motor-starting guide.

Safety: de-energize and verify dead before you swap

Swapping leads to reverse a motor is work on energized-capable conductors, so you de-energize and lock out before you touch them. Open the disconnect, apply your lock and tag, and verify the conductors are dead with a meter you have proven on a known live source first. Swap the two leads, torque them to the terminal's value, then restore power under controlled conditions and bump to confirm.

The hazard people underrate is the live rotation check. A phase-rotation meter on the line side is used energized by design, so treat it as energized work, with the right PPE and the right approach, and follow your site's electrical safety practices for working on or near live parts. The no-power motor-rotation tester avoids that exposure entirely on the motor side, which is part of why it is worth carrying.

Verify dead. Then work. That order does not bend for a job that looks like a two-minute lead swap.

Rotation in commissioning and startup

Rotation verification is a line item on any real startup, not an afterthought. The commissioning sequence for rotating equipment is: confirm the supply, confirm the motor direction uncoupled or with a meter, correct it if wrong while the gear is dead, then couple and run. NETA acceptance practices and equipment startup procedures both put a rotation check before the first loaded run for exactly the reasons in this guide.

On critical and data-center equipment, where a chilled-water pump or a CRAC compressor failing at startup has consequences beyond the one machine, the rotation check is non-negotiable and it is documented. Add the standing protection too: phase-monitor relays on the equipment that cannot tolerate reverse running, so a later source change is caught before it reaches the compressor. The commissioning record should show that rotation was verified, by whom, and how.

Build the check into the schedule, not into the panic after a backward start. A reversed motor caught on the commissioning checklist is a five-minute lead swap. The same reversal caught after a coupled compressor has run is a replacement.

Common mistakes

• Energizing a coupled pump or compressor just to see which way it turns, instead of verifying uncoupled or with a no-power tester first.
• Trusting the wire color or the label to prove sequence instead of metering the rotation.
• Not re-checking rotation after utility work, a transformer swap, a new feed, or any change to the service.
• Assuming a fan is fine because it is moving air, when a reversed fan still moves air at a fraction of the volume.
• Leaving a critical compressor with no phase-monitor relay and no built-in phase protection.
• Swapping all three leads in a rotated pattern and expecting a reversal, when only a two-lead swap reverses rotation.
• Swapping phases upstream of other equipment to fix one motor, reversing everything else fed from that point.
• Forgetting the generator and utility must match, so motors reverse on every transfer.

Field checklist

Standards and references

NEMA MG 1 standardizes motor terminal markings and the way rotation direction is described, including the convention of viewing the drive end and the defined standard direction for single-phase machines. For three-phase machines NEMA does not fix a universal direction from the phase sequence alone, so the motor nameplate's marked direction and the connection govern, and you verify rather than assume. Confirm the marking and the expected direction on the actual nameplate.

The equipment manufacturer is the authority on the direction the driven machine needs. The rotation arrow cast or printed on a pump casing, a fan housing, or a compressor, and the startup instructions in the manual, are what control the call, and they override any general rule of thumb. For scroll and screw compressors in particular, the maker's literature is explicit about reverse-rotation damage and the built-in phase protection on larger units.

NETA acceptance testing practice places a rotation check in the commissioning of rotating equipment before the first loaded run. The phase-rotation and motor-rotation meter makers publish the voltage limits and the correct connection for their tools; read those before you trust a readout, and respect the ceiling where the no-power method gives way to a bump test. Standards and editions are adopted and amended by jurisdiction, so confirm the version that applies and follow your site's electrical safety practices for any energized work.

Terms and conventions

Phase rotation goes by several names across drawings, nameplates, and meters, and the same idea reads differently depending on the source. Keeping the vocabulary straight saves an argument on the gear.

Phase rotation and phase sequence mean the same thing. ABC is the standard sequence, also called positive sequence; ACB or CBA is the reverse, also called negative sequence. The phases get labeled A-B-C, L1-L2-L3, R-S-T, or U-V-W depending on the convention. Motor terminals are marked T1, T2, T3 on a standard three-lead machine. Rotation direction is read as clockwise or counterclockwise, and which end you view from matters, so the convention is to state the viewing end, commonly the drive end or the end opposite the drive.

Phase rotation / phase sequence

The order the three phases reach peak voltage, ABC or ACB, which sets motor direction

ABC / positive sequence

The standard sequence; ABC, BCA, and CAB are all the same order

ACB / CBA / negative sequence

The reverse sequence; ACB, CBA, and BAC are all the same reversed order

L1, L2, L3

The three supply line conductors landed on the motor terminals

T1, T2, T3

Standard three-lead motor terminal markings that receive L1, L2, L3

Phase-monitor relay

A relay that trips the control circuit on reverse sequence, phase loss, or imbalance

Bump test

A momentary jog of the motor to observe shaft direction, done uncoupled

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is the difference between wye and delta?
• What is across-the-line starting?
• What is on a motor nameplate?
• What color are the wires in a 480 volt system?
• What is a motor overload relay?