VFD Troubleshooting Guide

---

Before you look at individual fault codes, there is a principle that applies to every troubleshooting call. The person who called you is describing what they see, hear, or smell. Not the root cause.

"The drive keeps tripping." "The motor will not start." "The panel smells hot." "There is a noise that was not there yesterday." These are symptoms. They are observable evidence that something is wrong. They are not diagnoses.

A drive that keeps tripping on overcurrent might have an overloaded motor, incorrect motor data entered in the parameters, a mechanical binding in the driven equipment, the wrong control mode selected, a torque boost set too high, or a variable torque setting applied to a constant torque load. The symptom is the same (overcurrent fault). The root causes are completely different, and each requires a different fix.

The fault code narrows the field. The baseline measurement from commissioning provides the context. The correct process works from symptom to fault code to measurement to root cause, in that order. Resist the urge to jump to conclusions or to fix the first thing that looks wrong. The most common troubleshooting mistake is fixing a symptom instead of the cause, which guarantees the fault will return.

If you took the time to record an operational baseline during commissioning (covered in our VFD Commissioning Guide), this is where it earns its keep. A technician looking at a 62-degree inverter temperature has no idea whether that is normal for the installation. A technician with the baseline that says the inverter was 44 degrees at commissioning immediately sees an 18-degree rise, checks current, finds the closed valve, and resolves the problem in twenty minutes instead of twenty hours.

A fault that appears once and does not return after a reset may have been a transient event. Document it and monitor for recurrence. A fault that returns repeatedly is telling you the underlying condition has not been corrected. Each reset exposes the drive and motor to the same abnormal condition, and repeated exposure accelerates component degradation.

The general rule: if a fault occurs more than twice in a shift, stop resetting and start investigating. Read the fault log (most drives store the last several faults with timestamps and operating conditions). Look for patterns. The information is there. The drive is talking. Listen.

Overtemperature is the single most common fault category, accounting for 50% to 60% of all reported VFD issues across every major manufacturer. The drive monitors the temperature of its power semiconductor heatsink and trips when that temperature exceeds the manufacturer's limit, typically 85 to 105 degrees Celsius depending on the drive.

An overtemperature fault means one thing: heat in is exceeding heat out. Either the drive is generating more heat than expected, or its cooling system cannot remove heat fast enough. Both paths lead back to installation decisions.

Blocked airflow or dirty filters. This is the number one cause of overtemperature faults in existing installations and a frequent cause in new installations where construction debris has entered the panel. Check the drive's cooling fan (is it spinning?), the air intake filter (is it clean?), and the airflow path through the panel (are cables or components blocking it?). Enclosure clearances above and below the drive are not suggestions. If they were compromised during installation, the drive will overheat.

Ambient temperature above design. The drive was sized and the panel was designed for a specific maximum ambient temperature. If the actual ambient exceeds the design temperature, the cooling system cannot remove heat fast enough. Common scenarios: a drive in a mechanical room where a new heat source was added, a rooftop panel where summer temperatures exceed expectations, or a panel door that was left closed when it was designed to operate with the door open.

Altitude derating not applied. Above 1,000 meters elevation, air density decreases and convective cooling drops by approximately 1% per 100 meters. A drive installed at 2,000 meters has roughly 10% less cooling capacity than the same drive at sea level. If the altitude derating was not applied during sizing, the drive will hit its thermal limit sooner than expected.

Carrier frequency too high. This is a classic "fixed the noise, caused the overheating" scenario. A drive was running at 4 kHz carrier frequency with the default settings. A maintenance tech heard a high-pitched whine from the motor, opened the parameters, and increased the carrier frequency to 12 kHz. The whine went away. Six weeks later, the drive started tripping on overtemperature every afternoon. Tripling the switching frequency increased the drive's internal switching losses by roughly 30%, pushing it past its thermal limit on warm afternoons.

Each switching event generates heat, and more switching events per second means more heat. A drive running at 8 kHz generates roughly 20% to 30% more heat than the same drive at 4 kHz. If carrier frequency was raised during commissioning without applying the corresponding derating, the first hot day will find the limit.

Drive undersized for actual load. If the drive was sized by matching its HP rating to the motor HP rating without accounting for actual operating conditions (ambient, altitude, carrier frequency, load duty cycle), it may not have enough thermal margin. A drive operating above 90% utilization has almost no headroom for additional thermal stress.

Before you reach for a replacement drive, work through the environment. Most overtemperature faults are environmental or configuration problems, not hardware failures. Enclosure design, panel layout, cooling method, and component spacing are covered in depth in our VFD Installation Guide. If the physical install was done right, the drive almost never runs hot.

Overcurrent and overload faults account for approximately 30% of first-week issues. Both involve output current exceeding a threshold, but they are not the same fault.

An overcurrent fault is an instantaneous protection that trips when output current exceeds a high threshold (typically 200% to 300% of the drive's rated current) for even a fraction of a second. This protects the IGBTs against short circuits, severe mechanical binding, or sudden load events.

An overload fault is a time-integrated protection (I-squared-t) that trips when output current exceeds a lower threshold (typically 105% to 110% of the configured motor FLA) for a sustained period. This protects the motor from thermal damage.

Read the fault code carefully. Most drives distinguish between the two types and provide context such as the current at the time of the trip and the output frequency at the time of the trip. Then work through the causes.

Mechanical binding in the driven equipment. A pump with a seized bearing, a conveyor with a jammed roller, or a mixer with hardened product demands far more current than normal operation. If the drive trips immediately on startup (overcurrent) or within seconds (overload), check the mechanical equipment before you touch the drive.

Incorrect motor data. If the motor FLA entered during commissioning is wrong, the drive's overload protection trips prematurely. If the entered FLA is lower than the actual motor FLA, the drive thinks the motor is overloaded when it is operating normally.

Wrong load type setting. This is one of the most common commissioning errors, and it has cost more than one technician a long afternoon. A grain handling facility installed a VFD on a 75 HP inclined belt conveyor. The drive was sized correctly. The installation was clean. The commissioning technician set the speed reference to 30 Hz, pressed Run, and the drive briefly spiked to its current limit and tripped on overcurrent. Reset. Same result. He increased the acceleration ramp from 10 seconds to 30 seconds. The belt moved a few inches before tripping again.

After two hours of parameter adjustments, the distributor's application engineer asked one question: what is the load type parameter set to?

Variable Torque.

The drive had shipped with a default load profile for fans and pumps. In Variable Torque mode, the drive limits output current to 110% for one minute, which is fine for a centrifugal load that needs almost no starting torque. But a loaded conveyor is a constant torque application. It needs 150% to 160% of rated current just to break the belt loose against static friction, material weight, and the inertia of the belt and rollers. The drive was never going to move that conveyor in VT mode, no matter how long the acceleration ramp got. Changing the load type to Constant Torque took thirty seconds. The conveyor started on the next attempt.

Torque boost set too high. Torque boost increases output voltage at low speeds to compensate for motor stator resistance. Too much boost causes excessive current at low speed, leading to overcurrent trips during starting. If you ran the motor ID Run during commissioning, the drive calculated the appropriate boost automatically. If the ID Run was skipped or if boost was adjusted manually, it may need recalibration.

Starting into a spinning load. Two fans in close proximity, one fan's draft causing the other to spin backward. Starting the free-wheeling fan causes an overcurrent because the drive must first stop the reverse rotation before accelerating forward. Solutions include coordinated start sequences, partitions to block cross-draft, or enabling the drive's flying start function, which detects a spinning motor and synchronizes to it before commanding direction.

Ramp time too short. An OL2 fault (overcurrent during acceleration) deserves special attention because it is the most common first-week fault on fan applications. The drive is trying to accelerate the motor and its load faster than the available current allows. The fix is almost always to lengthen the acceleration ramp. For fan applications, start with 60 seconds. For large fans (100 HP and above) or heavy wheels, 90 to 120 seconds is not uncommon. Do not fix this fault by raising the overcurrent trip level or the torque boost. Both of those approaches mask the symptom while the motor pays the price.

Drive undersized. If the fault persists even with ramp times above 120 seconds, the drive may be undersized for the application's inertia. Our VFD Selection and Sizing Guide covers how to apply the correct derating factors so the drive has enough margin for actual operating conditions, not just nameplate horsepower.

Voltage-related faults account for 15% to 20% of first-week issues. Overvoltage trips when the DC bus exceeds the drive's maximum limit (typically 810 to 850 VDC on a 480V drive). Undervoltage trips when the DC bus drops below the minimum (typically 350 to 400 VDC).

Deceleration too fast. This is the most common cause of overvoltage faults and one of the most common first-week issues overall. When the drive decelerates the motor, the motor's mechanical inertia keeps the shaft spinning faster than the new electrical frequency. The motor becomes a generator, feeding energy back through the inverter's flyback diodes into the DC bus. If the deceleration is faster than the bus can absorb this energy, the bus voltage rises until the drive trips.

The fix is almost always to increase the deceleration time. For fans and high-inertia loads, a default decel of 5 to 15 seconds is usually too aggressive. Try 30 seconds first. For high-inertia loads, 60 to 90 seconds may be needed. The longer ramp lets the motor's friction and windage absorb the kinetic energy during deceleration.

If increasing the deceleration time to an acceptable value does not eliminate the fault, or if the application requires fast deceleration, install a properly sized dynamic braking resistor. Do not rely on the drive's auto-deceleration or stall prevention feature as the primary solution. It is useful as a safety net, but it produces variable, unpredictable stopping times.

Overhauling loads. Some loads push energy back into the drive during normal operation, not just during deceleration. An inclined conveyor moving material downhill, a winder maintaining back-tension, or a crane lowering a load all generate regenerative energy continuously. Without a braking resistor or a regenerative front end, the drive trips on overvoltage. This is a specification issue, not a troubleshooting issue.

Input voltage too high. If the utility supply is running high (say, 510 VAC on a 480V system), the DC bus starts proportionally higher, which means less headroom before the overvoltage threshold. Check the input voltage against the commissioning baseline.

Voltage sags on the utility. This is the most common cause of undervoltage faults and often the most frustrating because the fault may not be reproducible on demand. A voltage sag caused by a large motor starting elsewhere in the facility, or a switching event on the utility grid, can momentarily drop the DC bus below the drive's trip threshold. Modern drives detect sags within a single power frequency cycle.

Here is the hard part: a sag that lasts a few milliseconds will not show up on a standard multimeter. If the drive is tripping on undervoltage intermittently and the voltage looks normal when you measure it, the cause is almost certainly a brief sag that your meter cannot catch. A power quality analyzer that logs voltage events over time is the right tool.

Loose input connections. A loose connection on one of the three input phases creates a high-resistance joint that drops voltage under load. With no load, voltage looks normal. Under full load, voltage drops significantly because current is now flowing through the high-resistance joint. Check and retorque every input connection. Thermal imaging during loaded operation identifies loose connections as hot spots in seconds.

Line reactor voltage drop at full load. A line reactor creates a voltage drop proportional to current. At full load, a 5% reactor drops roughly 5% of the supply voltage. If the post-reactor voltage at full load is below the drive's minimum operating threshold, the drive trips. Measure at the drive's input terminals under full load, not at the reactor's upstream side.

Ground fault detection senses current flowing between the output phases and ground. Under normal operation, there should be no significant current to ground. Small capacitive leakage is normal with long cable runs, but well below the trip threshold.

The systematic way to investigate a ground fault is a three-step sequence, and most faults are caught in the first step.

Step 1: Megger test with the drive disconnected. Disconnect the motor cable from the drive's output terminals. With a 1,000 VDC megger (500 VDC for smaller motors), test T1 to ground, T2 to ground, and T3 to ground. Readings should be above 1 megohm. Anything lower means you have either a cable problem or a motor insulation problem, and you have found the fault before doing anything else.

Step 2: Reconnect and run at low speed. If the megger test passed, reconnect the motor and run at 5 Hz. If the ground fault occurs at low speed, the cable is probably damaged at a specific point, or there is moisture at a termination. Inspect the cable path, open the motor junction box, and check for physical damage or water intrusion.

Step 3: Increase to full speed. If the fault does not occur at low speed but does at full speed, the most likely cause is reflected wave stress on marginal motor insulation. Long cable runs between the drive and motor can double the voltage at the motor terminals due to reflected waves, and a motor with borderline insulation may pass the megger test but fail under running voltage stress. The solution is an output filter (dv/dt filter or sine wave filter) sized for the cable run.

The common causes behind what you find:

Cable damage. Physical damage to the output cable insulation from installation, sharp edges in conduit, being pinched by a panel component, or rodent damage creates a path between the conductor and the cable shield or conduit. If the fault occurs immediately on startup or consistently at the same operating point, inspect the output cable.

Moisture ingress. Water or condensation in the motor junction box, in the conduit, or at cable terminations creates conductive paths to ground. Common in installations that sat idle before commissioning, or in outdoor and washdown environments. If the fault clears after the motor and cables warm up and dry out, moisture is the cause.

Motor insulation breakdown. Degraded winding insulation (from age, from reflected wave stress, or from a manufacturing defect) allows current leakage from windings to motor frame. This is why the megger test during commissioning matters so much.

Shield termination problems. A strand of shield braid that shifted during termination and is now touching a phase conductor inside the motor junction box will register as a ground fault. This is the same failure mode that has caused many a Monday morning restart problem. Our VFD Installation Guide. covers cable installation and 360-degree shield termination in depth, because getting this right during installation prevents a category of problems that is otherwise very difficult to diagnose.

Communication faults indicate a loss of communication between the drive and its control source (PLC, BMS, SCADA, or other supervisory system). They are especially common in installations where the drive receives its start/stop commands and speed reference from a network rather than hardwired signals.

Communication cable routed with power cables. This is the most common cause of communication faults in VFD installations, full stop. Communication cables running parallel to VFD output cables pick up electromagnetic noise from the PWM switching. The noise corrupts communication packets, triggering timeout faults.

The diagnostic clue is almost too easy. If communication faults correlate with drive operation, especially with drive starts and stops, the cause is EMI from the output cables coupling into the communication cable. Rerouting or re-shielding usually fixes it. Cable separation, 90-degree crossings, shielded cable matched to the protocol, and single-end shield termination are covered in the cable separation section of our VFD Installation Guide. 

Heat-related processor instability. If the drive's internal temperature is elevated, the microprocessor and communication interface can become unstable. Communication faults that appear during hot weather or heavy load and clear when the drive cools down are a thermal problem dressed up as a communication problem. Fix the overtemperature condition first.

Incorrect parameter configuration. Baud rate, node address, data format, timeout values. If any of these do not match between the drive and the controller, the link fails. This should have been caught during commissioning, but it sometimes surfaces as a first-week fault if the communication system was not fully tested.

Termination resistors missing or incorrect. Serial networks like Modbus RTU and PROFIBUS need termination resistors at both ends of the segment. A missing terminator causes signal reflections that corrupt data, especially as network traffic increases.

Wrong cable for the protocol. EtherNet/IP drives need shielded Cat5e or Cat6, not unshielded. PROFINET needs industrial Cat5e STP with M12 D-coded connectors. Modbus RTU uses shielded twisted pair. Each protocol has a specification, and generic substitutions create intermittent problems that are nearly impossible to diagnose after the fact.

If the drive shows no display at all after you apply power, if you can see visible damage to internal components (burn marks, swollen capacitors, cracked circuit boards), if fuses blow immediately on power-up, then you may actually have a hardware failure.

But check your installation first. A soft-charge resistor that burned because a contactor failed to close is not a defective drive; it is a damaged drive that reveals a failed support component. Blown input fuses are almost always a symptom of something external, not the drive itself.

The honest truth from talking to manufacturer support lines for years: most "defective drive" calls turn out to be installation problems, configuration errors, or environmental issues that the caller had not worked through. Before you pull the drive off the wall and box it up for warranty, walk through the fault category one more time, compare against the commissioning baseline, and confirm that the installation and parameter configuration match what the application actually requires. Nine times out of ten, the drive is fine. It is reporting what it found.

The drive does not trip for fun. Every fault connects back to a decision made during selection, installation, or commissioning, and reading the fault in that context turns a mystery into a diagnosis. Overtemperature points back to environment and sizing. Overcurrent points back to sizing, motor data, and load type. Overvoltage points back to deceleration and braking specification. Undervoltage points back to supply adequacy and connection integrity. Ground faults point back to cable installation and motor insulation. Communication faults point back to cable routing and protocol cable specification.

The path forward is not faster reset. The path forward is better diagnosis. Compare to the baseline. Read the fault log. Trace the symptom to the cause. Fix the cause, not the symptom.

Before the First Fault: A Field Guide to VFD Installation and Reliability covers the complete fault diagnostic framework in Chapter 18, with explicit cross-references back to every installation chapter where the root cause originated. The book includes diagnostic flowcharts for each fault category, a quick-reference card designed to be laminated and taped to the drive panel door, and the reset decision framework that tells you when to stop resetting and start investigating. The companion online training program walks through real-world fault scenarios with video demonstrations and downloadable reference forms.

For ongoing fault prevention and condition monitoring after commissioning, we cover trending and early warning indicators in our Maintenance and Reliability Guide.

Get the book: Before the First Fault

Enroll in the course: Before the First Fault training program

Contact the author: Dr. Carl Lee Tolbert, PhD, CMRP Wayward Leaders LLC www.waywardleaders.com

The drive is the messenger. The installation is where the story lives. Learn to read the faults, and you stop spending your nights at the panel.