VFD care and fault prevention: cooling, parameters and harmonics

VFDs have internal fans, and fans are wear parts. Replace them at 20,000–30,000 running hours (about three years of continuous duty); do not wait for one to seize. Once a fan stops, capacitor and IGBT temperatures climb fast — first over-temperature alarms, then a blown drive.

Clean dust off the heatsinks regularly. In dusty locations clean every three months, using low-pressure compressed air with a vacuum running alongside. Wash cabinet filters monthly.

Keep the cabinet below 40 °C. In hot shops during summer, consider a cabinet air conditioner or heat exchanger. Every 10 °C higher inside roughly halves electrolytic capacitor life.

Electrolytic capacitors are the VFD's life-limiting component. Judge them by: bulging or leaking cases, DC bus voltage ripple, or the drive's built-in capacitor life monitor (many brands have one). Once capacitance has fallen more than 20%, plan the replacement.

Record every parameter — especially motor nameplate data, ramp times, overload protection settings, and V/F or vector control parameters. When replacing a drive, a parameter sheet means the job takes half an hour; without one you commission from scratch.

Motor nameplate parameters must be entered exactly. Errors throw vector control off, heat the motor up, and make protection unreliable. After any motor change, re-run parameter identification (auto-tuning).

Do not set ramp times too short. Short ramps swing the DC bus voltage harder, age the capacitors faster, and trip on overcurrent more readily.

Never loosen protection settings casually. Overcurrent, overload and over-temperature protection exist to protect the equipment; widening them is the same as removing them. Frequent trips call for a root cause, not a parameter change.

Supply voltage must stay within tolerance (generally ±10%); beyond 3% phase imbalance, fit a line reactor. An unbalanced supply enlarges DC bus ripple, hurting both capacitors and the rectifier bridge.

Where several drives share a busbar, fit line or DC reactors — they suppress harmonics and protect the rectifier bridges. In heavily distorted supplies (many drives on one transformer), consider active filtering.

Never add capacitors or power-factor correction on a drive output — it destroys the drive outright. For long motor cables (generally over 50–100 m) fit an output reactor or dv/dt filter, or overvoltages will break down the motor insulation.

Use shielded cable between drive and motor, with the shield earthed at both ends (360° termination at the drive) — a safety requirement and a must for interference suppression.

Drive faults sort by root cause into three families, and with the family clear, troubleshooting has direction. Family one, environment: poor cooling (blocked air paths, dirty filters, aged fans) causing over-temperature trips; dust and moisture causing module breakdown and spurious operation; poor supply quality (voltage swings, harmonics, lightning) damaging the rectifier and inverter sections. Environmental faults are the majority - and almost all are preventable. Family two, parameters: accel/decel times set wrong causing overcurrent or overvoltage; V/F curve mismatched to the motor causing heat and weak torque; protection parameters too loose or too tight causing failure to act or nuisance trips; parameters lost or never backed up, making a replaced drive impossible to restore. Family three, load: mechanical jamming causing stall overcurrent; load transients exceeding drive capacity; current anomalies from transmission faults (slipping belts, damaged couplings).

The recommended troubleshooting order is "classify the symptom first, then check the matching stage": over-temperature trips point to environment and cooling; overcurrent trips start with the load (uncouple and test no-load), then parameters; overvoltage trips point to the supply and decel time; undervoltage to supply and rectifier; communication faults to wiring and interference; a dead display to supply input and the precharge circuit. This order covers the great majority of field cases and is far faster than testing modules one by one.

One more rule of thumb worth remembering: after a drive fault, do not immediately reset and restart three times. Repeated reset-and-energize can escalate a repairable fault into a blown module. The right response is to power down first, check the externals (load, supply, environment), eliminate the obvious causes, then energize once more as a test. If it still trips, hand it to a professional - do not gamble with the equipment.

Drive parameters are managed in three classes by importance. Class one - must be recorded and reproducible: motor nameplate data (power, voltage, current, frequency, speed), control mode (V/F or vector), accel/decel times, protection settings (overcurrent, overload, overheat), and communication parameters (station address, baud rate, protocol). Lose these and the drive essentially cannot be restored to its function. Class two - process parameters worth recording: multi-speed settings, PID parameters, frequency limits, terminal function definitions. Class three - factory defaults: generally no need to record, but note whether they were modified.

The backup method depends on the drive model: those with keypad storage can be copied panel-to-panel; those with communication ports can be uploaded via software; for neither, at minimum photograph or transcribe the key parameters and file them. Name backup files with equipment number and date, linked to the equipment file. On our retrofit and maintenance projects, parameter backup is a standard deliverable, and the delivery list states where the backup files are kept and how to restore them.

Parameter changes require discipline: record the original values before the change, verify the effect after it, and only on successful verification confirm the change and update the backup. Ad-hoc parameter edits under production pressure are a major hidden hazard - changed with no record, verified by no one, and at the next failure nobody knows the parameters were touched. Writing this rule into the equipment management system is far less effort than after-the-fact investigation.