VFD Monitoring

VFD Fault Modes and Early Detection

The common internal failure modes of variable frequency drives, how each develops, what early indicators appear before a drive fails, and which monitoring parameters detect them.

Key Principle

Most VFD failures do not happen suddenly. IGBT degradation progresses through measurable increases in on-state voltage and switching losses. DC bus capacitors lose capacitance over thousands of hours. Cooling fans slow before they stop. Each of these progressions is detectable before the drive trips.

IGBT and Power Module Failure

IGBTs (Insulated Gate Bipolar Transistors) are the switching elements that convert DC bus voltage into the variable-frequency output a VFD delivers to the motor. They are also the component most likely to fail in a drive that has been in service for several years.

How the Failure Develops

The primary mechanism is thermal cycling stress on the solder joints between the IGBT die and the copper substrate beneath it. Every load cycle creates a temperature swing at the junction. The IGBT and the substrate expand and contract at slightly different rates, putting shear stress on the solder layer with each cycle. Over tens of thousands of cycles, microscopic cracks propagate through the solder until the joint fails mechanically.

A second pathway is overvoltage transient puncture of the gate oxide layer. High-inertia regenerative loads returning energy to the DC bus, or upstream supply surges, can briefly exceed the IGBT's voltage rating and permanently damage the gate insulation.

Measurable Progression

As the solder joint degrades, two parameters change before the device fails outright. First, on-state voltage drop increases: the junction-to-substrate thermal path becomes more resistive, and this shows up in some drives' internal diagnostic data as rising on-state voltage measurements. Second, thermal resistance increases: the junction temperature rises faster per unit of dissipated power, so the drive's thermal monitoring sees junction temperatures trending upward under the same load conditions that previously produced lower readings.

Eventually the degradation progresses to gate drive failure, or the damaged IGBT conducts simultaneously with its complement in the same phase leg, creating a shoot-through short across the DC bus.

Early Detection Methods

DC Bus Capacitor Degradation

The DC bus capacitor bank filters the rectified supply voltage and stores the energy that buffers load transients. In most VFDs this means a bank of aluminum electrolytic capacitors, and electrolytics are the component with the shortest defined service life in the drive.

Degradation Mechanisms

Two mechanisms drive capacitor aging. The first is electrolyte evaporation: the liquid electrolyte inside each capacitor gradually escapes through the seal, reducing capacitance and increasing the capacitor's equivalent series resistance (ESR). Heat accelerates this process significantly. A capacitor running 10 degrees Celsius above its rated temperature loses roughly half its service life by the Arrhenius relationship.

The second mechanism is electrochemical stress from charge-discharge cycling. Each time the load draws a large transient from the DC bus, the capacitor discharges and recharges at the line frequency ripple rate. Over millions of cycles, this stresses the oxide layer inside each cell.

What Degradation Looks Like in Operation

Increased ESR means the capacitor generates more heat internally, adding to the thermal environment inside the drive enclosure. More visibly, DC bus voltage ripple at line frequency harmonics increases as capacitance falls. Drives that report DC bus voltage in their diagnostic data will show this as a higher peak-to-peak variation around the nominal DC bus value.

Reduced total capacitance also reduces braking performance: the drive stores less energy during regenerative events, so the DC bus voltage rises further during each deceleration cycle. In severe cases this triggers overvoltage trips that were not occurring previously.

Detection Options

Drives that expose DC bus ripple voltage or ripple current in their diagnostic registers allow continuous trend monitoring without physical access to the drive. This is the preferred approach for critical applications.

For drives without diagnostic DC bus data, periodic capacitance and ESR measurement requires de-energizing the drive, discharging the bus completely, and measuring each capacitor or the bank as a whole with an LCR meter. Manufacturer guidelines typically specify measurement intervals based on operating hours and ambient temperature class.

Cooling System Failure

VFDs dissipate a percentage of the power they handle as heat, typically 2 to 3 percent of rated power at full load. That heat exits through a combination of forced air convection over internal heat sinks and, in larger drives, through liquid cooling circuits. The cooling system is a maintenance item that degrades on a defined schedule.

Failure Pathways

Cooling fans accumulate dust and fibrous debris on blade surfaces and in the bearing housings. Increased blade mass and bearing drag raise motor current draw and reduce airflow volume. Fan bearings eventually wear to the point of increased vibration and noise, then seizure.

Filter media in filtered enclosures blocks progressively as debris accumulates. A fully blocked filter can reduce airflow to a fraction of design flow, causing heat sink temperatures to rise under loads that previously produced no warning.

Heat sink fins accumulate contamination between fin channels. The reduced surface area directly reduces convective heat transfer, raising operating temperatures even when fan airflow is adequate.

Detection Before Trip

The most reliable indicator is a trend in drive-reported heat sink temperature under consistent load conditions. A heat sink temperature that has risen 10 degrees Celsius from its baseline under the same ambient and load conditions is a maintenance trigger, not a fault code.

Audibly, a slowing or irregularly running fan is a direct indicator. Monthly visual inspection of filter media is the minimum required for fan-cooled enclosures in industrial environments. Most drive manufacturers publish inspection intervals for their specific enclosure designs.

Drives entering thermal derating (automatic reduction of output current to reduce internal heat generation) before a thermal trip is a late-stage warning. By the time derating is active, the cooling system is already significantly compromised and the drive is protecting itself at the cost of process output.

Gate Drive Circuit Faults

The gate drive circuit sits between the drive's control electronics and the IGBT gate terminals. It translates low-current control signals into the voltage pulses that turn each IGBT on and off at the correct moment. Degradation in this circuit produces output asymmetry without necessarily tripping the drive immediately.

Mechanism and Symptoms

Gate drive components include optocouplers, gate resistors, and bootstrap supply capacitors. Any of these degrading affects the timing or voltage of the gate pulses delivered to individual IGBTs. When one phase leg's gate drive differs from the other two, the switching behavior of that phase leg changes: rise time, fall time, and peak voltage all shift.

The result is phase-to-phase asymmetry in the output current waveform. At the same load and speed, one phase carries a different current magnitude from the other two. This asymmetry also generates increased common-mode voltage because the three phases no longer switch symmetrically around the DC bus midpoint.

Detection

Three-phase current monitoring at the drive output is the primary detection method. A consistent current imbalance between output phases under stable load conditions, not attributable to motor winding resistance imbalance, points to gate drive circuitry. Reviewing drive fault history for phase-loss or overcurrent codes associated with a single phase provides supporting evidence.

Common-mode current measurements on the motor cable can also reveal gate drive asymmetry, though this requires appropriate measurement equipment and a baseline for comparison.

Ground Fault Development

VFDs produce output voltages with very fast rise times: dv/dt values of 1,000 to 10,000 volts per microsecond are typical. Over years of operation, this voltage stress degrades motor winding insulation and cable jacket insulation. The result is a progressive increase in leakage current from one output phase to ground.

Failure Progression

Insulation degradation is not uniform. One phase, or one section of cable, tends to degrade faster depending on cable routing, junction box conditions, and the motor's thermal environment. As insulation resistance falls, leakage current increases. This current flows from the output phase conductor through the insulation to the motor frame or cable shield, then back through the ground path.

Most drives detect ground faults by monitoring the sum of the three output phase currents. Under balanced, fault-free conditions this sum is zero. A non-zero sum indicates leakage to ground. Drives trip on ground fault when this sum exceeds a threshold, typically expressed in the drive's I2t monitoring parameters.

Early Indicators

Intermittent ground fault trips are the characteristic early pattern. The drive resets and runs normally, then trips again under similar load or temperature conditions. As insulation continues to degrade, trip frequency increases. Eventually the drive trips consistently at startup or shortly into operation.

Insulation resistance (IR) testing on the motor and cable, measured with the drive isolated, provides a quantitative trend. A declining IR reading measured at consistent intervals (quarterly or annually depending on criticality) will show the degradation trajectory before the drive starts tripping. Values below 1 megohm on a 480V system motor are generally considered problematic; the trend matters as much as the absolute value.

Drive Fault Code Interpretation

VFD fault codes identify trip conditions, not root causes. The same fault code can result from several different underlying problems, and the same problem can produce different fault codes on different drives or under different operating conditions. Interpreting fault codes requires context: the load history, the fault frequency, and the conditions under which the trip occurred.

Common Fault Codes and What They Indicate

Overvoltage: DC bus voltage exceeded the protection threshold. Causes include a regenerative load with high rotational inertia returning energy faster than the braking resistor can absorb, a supply-side voltage surge, or a deceleration ramp time too short for the connected inertia. One overvoltage trip per year on a stiff supply is within normal operating experience. Weekly overvoltage trips require investigation of the braking circuit and supply quality.

Undervoltage: DC bus voltage dropped below the low-voltage protection threshold. Causes include supply voltage sag, a loose or corroded input power connection, or a blown fuse on one input phase. A single-phase loss will produce undervoltage on some drive models before a phase-loss code appears.

Overcurrent: Output current exceeded the drive's trip threshold. This is a broad fault category covering motor overload (the load required more torque than the drive could supply at the set current limit), a ground fault or phase-to-phase short, a failing IGBT module conducting excess current, or a locked rotor condition.

Overtemperature: Heat sink or internal ambient temperature exceeded the protection threshold. Primary cause is cooling system degradation as described above. Secondary cause is sustained overload operation above the drive's thermal rating for the ambient temperature class.

Ground fault: Leakage current from output phase to ground exceeded the detection threshold. The source is motor winding or output cable insulation degradation, as described above. Confirm by measuring motor and cable insulation resistance with the drive isolated.

Using Fault Frequency as a Diagnostic Tool

A fault code's significance depends heavily on how often it occurs. A single overcurrent trip during an unusual process event carries different diagnostic weight than 12 overcurrent trips in 30 days under normal load conditions. Drives that log fault history with timestamps allow this analysis directly. For drives with limited fault log capacity, maintaining a manual fault log at the panel is worth the effort for critical equipment.

The pattern that warrants immediate attention: any fault code that was previously absent and now appears with increasing frequency under unchanged process conditions. This is the signature of a developing hardware problem rather than a process anomaly.

Maintenance Actions by Fault Mode

Fault Mode Early Indicator Monitoring Method Maintenance Action
IGBT / Power Module Degradation Rising heat sink temperature under constant load; output phase current asymmetry; intermittent overcurrent or ground fault codes Drive thermal trend logging; three-phase current monitoring at drive output; fault history review Schedule module replacement during next planned outage; reduce thermal stress by verifying cooling system is clean and functional; review load cycle severity
DC Bus Capacitor Degradation Increased DC bus voltage ripple; overvoltage trips during deceleration that were not previously occurring; elevated internal drive temperature DC bus ripple monitoring via drive diagnostics (where available); periodic capacitance and ESR measurement with drive de-energized Replace capacitor bank per manufacturer's hour-based or calendar-based replacement schedule; prioritize replacement if ESR has increased more than 50% from nameplate specification
Cooling Fan / Filter Failure Rising heat sink temperature trend; audible fan speed reduction; drive entering thermal derating under normal load Drive-reported heat sink temperature trending; monthly visual and tactile filter inspection; thermal imaging of drive enclosure exterior Clean or replace filter media on schedule; replace fan if bearing noise or speed reduction detected; clean heat sink fins with compressed air during planned shutdown
Gate Drive Circuit Fault Consistent current imbalance between output phases under stable load; increased common-mode voltage; intermittent phase-related fault codes Three-phase current monitoring at drive output; drive fault history for phase-specific codes Circuit board replacement or component-level repair if gate drive PCB is serviceable; confirm motor winding balance first to rule out motor-side asymmetry
Ground Fault (Motor / Cable Insulation) Intermittent ground fault trips with increasing frequency; declining motor insulation resistance trend Drive fault log review for ground fault code frequency; motor and cable IR testing at scheduled intervals Replace motor cable if IR is declining and cable routing exposes it to mechanical or chemical damage; motor rewind or replacement if IR below acceptable threshold; consider output reactor to reduce dv/dt stress if replacement motor is same insulation class
Overvoltage (Braking / Supply) Overvoltage trips during deceleration; fault frequency increasing; braking resistor temperature higher than baseline Drive fault log for overvoltage code frequency; braking resistor thermal check; supply voltage monitoring Check braking resistor continuity and thermal rating against load inertia calculation; extend deceleration ramp time if inertia has increased; investigate supply voltage quality if fault occurs outside deceleration events

Frequently Asked Questions

How do IGBT failures develop inside a VFD?

IGBT failures develop through thermal cycling stress on solder joints between the IGBT die and the substrate. Each load cycle creates a temperature swing at the junction, causing microscopic cracks to propagate over thousands of cycles. The measurable precursors are rising on-state voltage drop and increasing thermal resistance before an actual gate drive failure or shoot-through fault occurs.

What causes DC bus capacitor degradation in a VFD?

Electrolytic capacitors degrade through two mechanisms: electrolyte evaporation (accelerated by heat) and charge-discharge cycling stress. Both reduce capacitance and increase equivalent series resistance (ESR). The visible effects are increased DC bus voltage ripple at line frequency harmonics and reduced braking energy storage. Drives that expose DC bus ripple in their diagnostic data allow trend monitoring without removing the drive.

What does an overvoltage fault code mean on a VFD?

An overvoltage fault indicates DC bus voltage exceeded the drive's protection threshold. The cause is typically a regenerative load with high rotational inertia returning energy faster than the braking resistor can absorb it, or a supply-side voltage surge. A single annual trip on a stiff supply is within normal range. Weekly overvoltage trips indicate a deteriorating braking resistor, a supply problem, or a deceleration ramp time too short for the connected load inertia.

How can cooling fan failure be detected before the drive overheats?

The earliest indicator is a rising trend in drive-reported heat sink temperature under the same load conditions. Audible changes in fan speed are a secondary indicator. Drives entering thermal derating (automatic output current reduction) before tripping is a late-stage warning. Monthly filter inspection and comparing current heat sink temperature readings against baseline under similar load conditions will catch degradation before a thermal trip.

What is the difference between a ground fault trip and an overcurrent trip on a VFD?

A ground fault trip occurs when the vectorial sum of the three output phase currents is non-zero, indicating leakage current flowing to ground through degraded motor winding or cable insulation. An overcurrent trip occurs when any individual phase current exceeds the drive's set threshold, from motor overload, a locked rotor, a short circuit, or a failing IGBT module. Both can trigger the same trip event on some drives, so checking insulation resistance on the motor and cable helps distinguish the root cause.

Monitor VFD Health Continuously

Tractian's condition monitoring platform connects to your electrical systems and tracks the parameters that predict VFD failures before drives trip. Thermal trends, current asymmetry, and fault code history, all in one view.

See Electrical Monitoring
This content is produced by electricalmonitoring.com, an educational resource operated by Tractian. Information is provided for general reference. Always follow equipment manufacturer guidelines, applicable electrical codes, and site safety procedures. Measurements and intervals described are illustrative; consult your drive manufacturer's documentation for equipment-specific requirements.