VFD faults for maintenance techs: read the code, not the panic
A drive fault stops the line and everyone looks at you. Good news: the VFD just told you more about the problem than any other component on the machine ever will. The fault code sorts the whole problem into one of four buckets — before you open a single panel.
Open the drive and watch which part complains.
Power comes in on the left, gets rectified to a DC bus, and is switched back out to the motor on the right. Each fault family lights up a different section — that's why the code tells you where to go.
Running — nothing to protect against
Supply comes in, the rectifier fills the DC bus, the caps hold it steady, and the inverter switches it back out to the motor at the frequency you asked for. The fan moves air across the heatsink the whole time.Back up your parameters while it looks like this. A baseline turns tomorrow's "it's acting weird" into a five-minute comparison.
Illustration only — codes vary by manufacturer. The DC bus holds a lethal charge after power-off: wait the stated discharge time and verify zero energy before opening a drive.
Get the free PM checklistHere's the mental shift that separates a controls-minded tech from a reset-and-pray tech: the drive is a measuring instrument that happens to also turn a motor. It watches current, bus voltage, temperature, and ground leakage every millisecond. When it faults, it's handing you a measurement. Most drive "problems" aren't in the drive at all — the usual suspects are the supply feeding it, the motor and cable it feeds, the load on that motor, or the environment it lives in.
First move: reset once — but know why you're resetting
When a drive faults on my floor, my first move is to reset it and see what happens.
That sounds like the opposite of the advice I give on overload relays, where I inspect before touching anything. It isn't. The difference is how confident I am about the cause before I act.
In my building, a VFD fault is a loaded belt roughly 99% of the time. The conveyor is full of boxes and the drive is protecting itself. I'm not resetting blind — I'm resetting a fault I've already diagnosed from experience, and using the reset to confirm it. If it comes back, I go clear the belt.
That's the actual rule underneath both articles: confident about the cause → reset and confirm. Not confident → look first. Experience doesn't skip steps; it changes which step is worth doing first.
01First read: which fault family?
Every manufacturer names them differently, but nearly every trip lands in one of four families — and each family points somewhere different:
Overcurrent
The most common field fault. Output current exceeded the limit — think mechanical overload or jam, a shorted motor/cable, or accel set too aggressive. The drive is protecting its own transistors; it reacts faster than any breaker.
Over/Undervoltage
DC bus out of range. Overvoltage usually = decelerating too fast — the motor acts as a generator and pumps energy back. Undervoltage = supply sag or a starving connection. Both point upstream or at the ramp settings, not at the motor.
Overtemperature
The drive is cooking. Dead or dying cooling fan, clogged heatsink or filters, blocked cabinet airflow, or ambient heat. Environment kills more drives than electronics do.
Ground fault
Current leaking to ground. Motor insulation breaking down or cable damage — and if it correlates with rain or washdown, moisture is getting in somewhere. This one is a megger job, not a reset job.
How you read the fault depends on what your system gives you
Two eras, two workflows — and plenty of facilities are still living in the first one.
Without a display at the drive: you get the fault code off the unit, look it up in the manual, and work from there. Slower, and it means the manual has to actually be findable when you need it at 2 a.m.
With an operator interface or graphical display: the fault shows up stated plainly, and you skip the lookup entirely.
If your operation doesn't have the display, the practical fix isn't buying one — it's making sure the fault-code reference is where the tech is standing, not filed in an office. A laminated sheet in the panel beats a manual on a shelf.
02The reset discipline: one and done
Every tech resets a drive once — sometimes a trip really is a power blip or a one-off jam. The discipline is what happens next. If the same code comes back, stop resetting. Each repeat trip is the drive repeating the same measurement, and repeatedly resetting into a real fault can finish off whatever's failing — reset into a ground fault enough times, and you can turn a marginal motor winding into a dead one.
Before that first reset, do the ten-second data grab: log the code, and check the drive's fault history screen — most drives store the last several trips with current, bus voltage, and frequency at the moment of failure. That history tells you whether this is trip #1 or trip #14, and whether it always happens at the same frequency, the same time of day, or the same point in the cycle. Patterns like that solve faults.
03Working each family
Overcurrent → start at the load
Same logic as an overload relay: lock out and turn the load by hand first — jam, seized bearing, dead gearbox. If the load turns free, megger the motor and cable (disconnected from the drive) for a short or insulation fault. Only after those check clean do you look at accel time — a ramp set too aggressive for the inertia will trip on every loaded start.
Overvoltage → look at the decel, not the drive
A decelerating load back-feeds the DC bus. If OV trips happen when stopping — especially with heavy or high-inertia loads — extend the decel ramp or add/check a braking resistor. If OV happens while running steady, suspect supply transients instead.
Undervoltage → chase the supply
Sagging incoming power, an undersized feed, or a loose/corroded supply connection starving the bus. If UV trips cluster at the same time of day, something big is starting on the same feed. Check terminations before blaming the utility.
Overtemperature → it's almost always airflow
Cooling fans are wear items — typical life is a few years, not the life of the drive. Check the fan spins, the heatsink fins aren't packed with dust, cabinet filters are clean, and nothing is stacked against the enclosure. A drive in a hot, dirty cabinet lives half as long as the same drive in a clean one.
Ground fault → megger the motor circuit
Disconnect the motor leads at the drive (never megger INTO a drive — the test voltage kills electronics), then megger phase-to-ground on the motor and cable. Low readings name your culprit. Intermittent GF that tracks weather or washdown = moisture ingress at a connection box, conduit, or the motor itself.
Parameter drift → "someone fixed something"
A drive that ran fine for years and suddenly acts different without a fault often didn't change itself. Compare parameters against your baseline backup — and if you don't have a baseline backup, today is the day. One drift pattern worth knowing: repeated bearing failures on a VFD-fed motor can be electrical fluting from shaft current, not a mechanical problem at all.
The overvoltage-on-decel pattern is worth burning into memory because it fools people into replacing drives. The bus voltage sits flat all day, then spikes over the trip threshold in the two seconds after a stop command — the drive is fine; the ramp is wrong for the load.
The fix costs nothing: lengthen the decel time. If the process can't tolerate a slower stop, that's what braking resistors are for — somewhere for the regenerated energy to go.
04Drives fail less than people expect
Worth setting expectations honestly: I've rarely had to replace a VFD.
In years of running this operation, one comes to mind — a drive on a multi-stage extendable conveyor that we eventually had to swap out. That's it. Everything else has been the drive doing exactly what it's supposed to do: faulting to protect itself and the motor from a condition that shouldn't be run through.
So when a drive faults, the odds heavily favor something the drive is protecting against, not the drive itself failing. Chase the cause first. Suspect the unit last.
05When the protection becomes the recovery problem
Two hours to clear one bad day
When a fault turns out to be a loaded belt, the recovery is simple in principle: clear the boxes. There's no way to start it in a mode that works through the load — the belt has to be cleared.
Where it gets painful is a major jam. Our system starts in a defined sequence, deliberately, to keep conveyors from restarting into each other and causing the exact jams that fault the drives. That sequencing works — but when the jam is bad enough, you can't satisfy the sequence until each conveyor in the chain is clear. So you end up clearing conveyor after conveyor after conveyor just to get the system to come back up.
Add photo eyes detecting jams and stopping sections on top of the drive faults, and it becomes a genuinely bad day. The last serious one took us two hours to clear.
The honest tradeoff nobody mentions
Sequenced startup and jam detection exist to stop small problems from becoming big ones — and they do that job well.
But the same design that prevents small problems amplifies recovery time on big ones. Every protective interlock you have to satisfy on the way back up is another thing standing between you and running.
That's not an argument against the protection. It's an argument for catching jams early, because the cost of a bad one isn't just the jam — it's the whole sequenced recovery behind it. And it's worth knowing before you promise operations a restart time: a small jam is minutes; a major one is a chain reaction you clear one conveyor at a time.
Drives have wear items — treat them like it
A VFD isn't a solid-state forever-box. Cooling fans typically last a few years; DC bus capacitors age out over roughly 5–10 years, depending on heat. A well-cooled drive in a clean cabinet can run 10–15 years; the same drive cooking in a dusty enclosure won't see half that. Proactive fan replacement and clean filters are the cheapest drive-life insurance there is — which is why they're on the annual list in our PM checklist.
A VFD is not de-energized when you kill the disconnect. The DC bus capacitors hold a lethal charge after power is removed — sometimes for several minutes or more. Follow the manufacturer's stated discharge wait time, then verify zero energy with a meter at the bus terminals before touching anything inside the drive.
All of it — lockout, discharge wait, verification, PPE — per your facility's energy-control and electrical safe-work policies, performed by a qualified person. Drive internals are not a "quick look" component.
And the same applies on the mechanical side: clearing a loaded belt, working at a drive, or opening an enclosure means working on or near energized equipment and stored mechanical energy. Confirm the equipment is de-energized and follow your facility's energy-control (LOTO) policy before anyone reaches into a conveyor or a panel. Drive and panel work is qualified-person work; your facility's procedure is the authority.
06The troubleshooting sequence
- Log the code and pull the fault history before anything else. Current, bus volts, and frequency at trip — plus how many times this code has hit. Ten seconds, and it's half the diagnosis.
- Sort it into its family: OC = load/motor/cable, OV = decel/regen, UV = supply, OH = cooling, GF = insulation. The family picks your path.
- One reset, eyes open. If it holds, watch the conditions that preceded the trip. If the code returns — stop resetting and start diagnosing.
- LOTO Lock out, wait the manufacturer's capacitor discharge time, verify zero energy before hands go inside the drive or on motor leads. The bus bites after the power's off.
- OC path: turn the load by hand (locked out) → megger motor + cable, disconnected from the drive → then review accel settings.
- OV/UV path: trips on stop = lengthen decel or check the braking resistor. Trips at random or same-time-of-day = chase supply and terminations.
- OH path: fan spinning? Fins clean? Filters clear? Cabinet ventilated? Fix airflow before suspecting the drive's sensors.
- GF path: megger motor and cable phase-to-ground (never into the drive). Weather-correlated = hunt moisture at boxes, conduits, and the motor.
- Back up parameters when it's healthy. The baseline backup turns tomorrow's "it's acting weird" into a five-minute comparison instead of a guessing game.
07By the book vs. on the floor
By the book
Read the fault code, look it up, and work the diagnostic tree from the top before touching anything.
Never reset a fault until you've identified the cause. A repeated fault means the drive needs evaluation.
Once the fault is cleared, restart the system.
On the floor
You already know what it usually is. Reset once to confirm it, then go clear the belt.
Resetting is a diagnostic step when you're confident — as long as one reset is the limit. A repeated fault almost always means the load is still sitting there, not that the drive is bad.
Clearing the fault is the easy part. Satisfying the startup sequence across a jammed line is the two-hour part.
08The takeaway for your floor
The drive is a measuring instrument — use it:
Most "bad drives" are innocent. The fault code, the fault history, and the four-family sort will point you at the load, the ramp, the supply, the airflow, or the motor circuit — where the real problem almost always lives. Save the drive swap for when the drive has actually earned it.
Got a drive story from your floor?
Every building's drives fail a little differently. If you've chased a VFD fault worth teaching — especially one that fooled everybody, or a recovery that ate your whole shift — email info@maintenancevaultpro.com. The best "from the floor" accounts get added to this article, credited however you like — including not at all.
Catch the dying fan before the dead drive
Grab the free conveyor PM checklist — daily, monthly, and annual inspections.
Almost there — check your inbox
Click the confirmation link we just sent, and the checklist is yours. If it isn't there in a minute, check your junk folder.