VFD Troubleshooting Guide: Fault Codes, Field Tests, and When to Replace the Drive

The drive on your pump tripped. You pressed reset, it ran, and twenty minutes later it tripped again. By the fourth reset someone is already asking whether the unit is defective.

Most of the time it is not. The drive detected an abnormal condition and protected itself. The fault code is information, and clearing it without reading it throws that away.

This VFD troubleshooting guide follows four steps: read the code, test in the field, judge repair or replace, and source the right AC drive if replacement is the answer.

1. Before You Touch Anything

Discharge the DC bus

The DC bus capacitors hold a lethal charge after the supply is switched off. Waiting time varies by brand and frame size, so check the label on the drive. Waiting is not proof: measure across the DC+ and DC- terminals with your meter on DC volts and confirm the voltage has fallen below the manufacturer's safe value before working inside. Lock out the incoming supply first.

A dark keypad does not mean the bus is discharged. The control supply and the bus are two different things.

Record this before you press reset

Reset often overwrites the fault log, and the log is your evidence.

• Full fault code, not a rough description
• The last three faults with timestamps
• Output frequency and current at the trip
• Heatsink or panel temperature, plus any smell, noise or moisture in the cabinet
• The operating state at the trip: starting, accelerating, running steady, or decelerating

That last point decides where you look. A drive that trips during acceleration and one that trips after forty minutes of steady running have almost nothing in common.

When to stop resetting

A single fault that does not return may have been transient. Record it and watch it. If the same fault appears more than twice in one shift, stop resetting and start investigating. Every reset puts the IGBTs and the motor insulation through the same abnormal condition again, and the damage accumulates.

2. Read the Code First: VFD Fault Codes by Brand

Each manufacturer uses its own code system, but the fault categories underneath are the same. Translate the code into a category, then go to that section. Always confirm the meaning in the manual for your exact series; the table below is for fast orientation, not a replacement for documentation.

Note whether your drive separates alarms from faults, and read the fault history rather than the live display: the code on screen is the last event, not the first.

Stock by brand: Siemens, ABB, Mitsubishi, Allen-Bradley, Omron, Schneider.

The sections below are ordered by how often each category appears.

3. Overtemperature Faults

Overtemperature is the largest single category of VFD trips across every major brand.

The drive is not measuring cabinet air. It measures the heatsink carrying the power semiconductors and trips when that passes the manufacturer limit, typically 85 °C to 105 °C (185 °F to 221 °F) depending on the model. A cabinet that feels cool tells you nothing about the heatsink.

Only two things can be wrong: heat going in faster than expected, or heat coming out slower. One question narrows it fast. Did this drive ever run cool? A new installation points to design and sizing. Years of good service followed by recent trips points to cleaning, fans and whatever changed nearby.

The carrier frequency trap

A motor whines, someone raises the carrier frequency, the whine stops, and everything looks fine for a month. Then the weather turns warm and the drive trips every afternoon. A higher carrier frequency means more switching events, and every switching event makes heat. Manufacturers publish a derating curve for exactly this: above the default setting, usable output current drops. If you raise it, look up the derating figure, recheck your current margin, and record the change.

4. Overcurrent and Overload: Two Faults, Two Fixes

These two words get used as if they mean the same thing. They do not, and confusing them sends you looking in the wrong place.

Overcurrent is instantaneous protection for the IGBTs, tripping within milliseconds when output current spikes to two or three times the drive rating. Overload is thermal protection for the motor, based on I²t, meaning the higher the current the shorter the time allowed; it reacts to sustained current slightly above the motor full load amps you entered.

A useful shortcut: if it trips the instant you press start, look at mechanics and parameters. If it trips after minutes of running, look at the load and the motor data.

Check the load before you blame the drive

With the supply isolated, turn the shaft by hand. It should move freely with even resistance. Listen to the bearings. Look for hardened material in a mixer, a jammed conveyor roller, a fouled impeller, or a pump started against a closed valve. Five minutes here saves hours at the panel.

Parameters that cause overcurrent

Motor FLA entered incorrectly. A value below the nameplate makes the overload protection trip while the motor works normally.

Wrong load type. This one costs whole afternoons.

The rule: fans and centrifugal pumps are variable torque; conveyors, screws, augers and compressors are constant torque. Factory defaults are often set for fans and pumps, so a conveyor needs the setting changed. Our article on Schneider VFD overload capacity explains how these ratings are specified.

Torque boost set too high. Excess boost raises voltage at low speed and pushes starting current up. If the motor ID run was skipped at commissioning, run it now.

Spinning loads and short ramps

A fan already turning backwards in a draft forces the drive to arrest that rotation before accelerating forward. It trips on an immediate restart but starts cleanly ten minutes later. Use a coordinated start sequence or the drive's flying start function, which detects the motor's actual speed and direction and synchronises to it.

A ramp that is too short demands more current than the drive can supply. Lengthen it. Fans and high inertia loads need far more time than the factory default.

Do not fix this by raising the overcurrent trip level or the torque boost. Both hide the symptom and the motor pays for it. If the fault survives a ramp already stretched to an unreasonable length, the drive is probably undersized for the inertia, which is a sizing question rather than a parameter question.

When it really is the IGBT

With the supply off and the bus confirmed discharged, use the diode range and test each output phase against the others and against the DC bus terminals.

A shorted reading ends the diagnosis. Stop testing and move to the replacement decision.

5. Overvoltage and Undervoltage: Reading the DC Bus

Most engineers assume overvoltage arrives from the utility. Usually it comes from your own load. When the drive slows the motor, inertia keeps the shaft turning faster than the new frequency, the motor becomes a generator, and energy flows back into the DC bus. If the bus cannot absorb it fast enough, voltage climbs until the drive trips. The first fix is almost always a longer deceleration time.

If a longer ramp is unacceptable, or if the load regenerates continuously (downhill conveyors, winders, hoists lowering loads), you need braking hardware. Occasional regeneration calls for a dynamic braking resistor; continuous regeneration calls for a regenerative front end. Sizing a resistor comes down to three values: resistance, never below the drive's stated minimum, peak braking power, and duty cycle.

The sag your meter cannot see

A dip lasting a few milliseconds never shows on a handheld meter, but the drive detects it within one supply cycle. If undervoltage is intermittent and the voltage measures fine whenever you check, log it with a power quality recorder. Two other causes are easy to miss: a loose input connection reads normal with no load and drops badly under load, and a line reactor drops voltage in proportion to current, so always measure at the drive input terminals at full load.

The DC bus sits at roughly the line voltage multiplied by 1.41: about 325 V DC on a 230 V supply, 565 V DC on 400 V, 680 V DC on 480 V and 810 V DC on 575 V. Confirm the trip thresholds in your manual.

6. Ground Faults: A Three-Step Isolation Method

Ground faults are rarely inside the drive. They are in the cable or the motor, and these three steps tell you which.

Step 1. Disconnect the motor cable from the drive output and megger each phase to ground. A low reading means you have already found the problem. Select the test voltage by motor rating, and remember acceptable values shift with motor size and temperature.

Step 2. Reconnect and run at low speed. A trip at low speed points to cable damage or moisture at a termination. Walk the cable route for sharp conduit edges, pinch points and rodent damage, then open the motor terminal box and look for water. A fault that appears on the first cold start and clears as things warm up is moisture.

Step 3. Raise to full speed. A trip only at full speed points to reflected wave stress: steep pulses reflect at the end of a long motor cable and add up, so motor terminal voltage runs much higher than the drive output. Marginal insulation passes the megger test and fails under running voltage. The remedy is an output filter, dv/dt or sine wave, sized for the cable length.

7. Communication Faults Are Usually Wiring Faults

One test settles most cases in half a minute. Watch the communication status while someone starts and stops the drive three times. If the faults line up with drive starts and stops, the cause is noise from the output cables coupling into the communication cable. Separate the runs, cross at ninety degrees where they must meet, and ground the shield at one end only.

Then check the cable suits the protocol: shielded twisted pair terminated at both ends for Modbus RTU, the specified cable with end termination for PROFIBUS, industrial shielded Cat5e or better for PROFINET and EtherNet/IP. Office patch cable in a drive panel causes intermittent faults nobody can trace later. If cabling is correct, compare baud rate, node address, data format and timeout at both ends.

One trap remains: a hot drive can produce an unstable communication interface that looks like a network problem. If the faults clear once the drive cools, fix the temperature first.

8. Repair or Replace? A Decision Framework

Most faults above have a fix that costs only time. A small number do not.

Signs the hardware has actually failed: no display with the control supply confirmed present; burn marks, swollen capacitors or cracked boards; input fuses blowing the moment power is applied; a shorted reading on the output diode test. One caution: blown input fuses usually point to something outside the drive, so check the supply, the contactor and the cabling first.

A third option gets overlooked: replace now and keep the repaired unit as a shelf spare. For a discontinued model this is often the best value available, because it restores production immediately and secures a spare that cannot easily be bought again.

Most plants compare the repair quote against the new price and stop there. Add one line: hourly cost of the stopped line, multiplied by production hours per day, multiplied by days waiting for the repair. Compare that against the price difference. It changes a lot of decisions.

One last check before ordering. If the root cause sits in the environment, the parameters or the installation, a new drive fails the same way. Is the root cause confirmed? Has the condition been corrected? Have the parameters been checked rather than assumed? Replacing a drive is easy. Replacing the condition that killed it is the actual repair.

9. Sourcing the Right Replacement Drive

Match by part number, verify by specification

Ordering the same part number is fastest, but only when the model is current and the configuration identical. The same base model often ships in variants covering filters, braking units, communication options and enclosure ratings, and one different character in the suffix can mean a unit that does not fit. Check these against the old drive and the motor:

The most frequent purchasing error is selecting by kW or HP alone. Rated current at the same power rating differs between manufacturers and between series. Compare current. Use our model library to check a part number before you order.

When your model is discontinued

Three routes exist. The manufacturer's successor series usually has an official migration reference, so start there. Stock channels can still supply the original model. Cross brand substitution is possible but carries the most risk.

Four traps catch people out: parameter numbers and meanings change between series, so a parameter list cannot be copied across; mounting dimensions change; control terminal assignments change; and communication option cards are rarely compatible between generations. If the old drive still powers up, export and save its complete parameter set first.

Original, refurbished, or aftermarket

These three are not the same thing, and refurbished units have a legitimate place as shelf spares. What matters is knowing which one you are buying.

• Serial number consistent across the drive label, the carton and the warranty document
• Date code readable and consistent with the seller's claim
• No tool marks on the case screws, no reworked solder joints inside
• Factory traceability and a warranty you can actually claim

If a supplier cannot show traceability, ask for it before you pay. CHENTUO supplies new original units across all the VFD brands we stock, with traceable documentation and a one year warranty.

Before you power up the replacement

Load and verify the parameter set, check motor nameplate data (FLA, voltage, frequency, poles), run the motor ID routine, and set carrier frequency for the actual conditions with any derating applied. Then record the baseline: DC bus voltage, no load and full load current, heatsink temperature. Ten minutes today is the reference that saves hours at the next trip. For brand specific setup, see our Siemens VFD guide and our guide to ABB VFD control functions.

FAQ

 

 

How do I know if the fault is in the VFD or the motor?

Disconnect the motor cable and test motor insulation to ground. If that passes, run the drive with no motor connected and see whether it still faults. If it runs clean, try a known good motor. Those three tests isolate it.

Can I test a VFD with a multimeter?

Yes, for three things: incoming supply voltage, DC bus voltage, and the output devices on the diode range. It cannot catch a millisecond voltage sag or measure insulation resistance, so a power quality recorder and a megger cover the rest.

How long should I wait before opening a VFD after switching it off?

Check the label, since the time varies by brand and frame size. Waiting alone is not enough: measure across the DC bus terminals and confirm the voltage has dropped below the manufacturer's safe value before working inside.

Why does my VFD keep tripping on overcurrent after every reset?

Work in this order: mechanical binding in the driven equipment, parameters (load type, motor FLA, torque boost), acceleration ramp too short, then hardware. The first two account for most repeat overcurrent trips.

How do I find a replacement for a discontinued VFD model?

Check the manufacturer's migration reference for the successor series, look for the original through stock channels, or consider a cross brand equivalent. Verify rated current, overload rating, control mode and communication option, and save the old parameter set first.

Can I use a different brand of VFD to replace mine?

Technically yes, provided voltage, rated current, overload rating, control mode and communication interface all match the application. Expect to reconfigure parameters and rewire control terminals, since neither transfers between brands. Send us the part number and we can confirm what fits.

The Short Version

The fault code is where diagnosis starts, not where it ends. Most repeat faults trace back to installation, parameters or environment rather than to the drive. And once a hardware failure is confirmed, choosing the right replacement matters more than ordering the fastest one.

Have the part number? Send it to us. We stock new original Siemens, ABB, Mitsubishi, Allen-Bradley, Omron and Schneider VFDs, ship within 48 hours by DHL, FedEx or UPS, back every unit with a one year warranty and accept returns within 30 days.

Request a quote with your model number, quantity and email, or message us on WhatsApp for a same day reply.

Free consultation