ABB VFD Earth Fault: Three Scenarios, Three Fixes (From a Quality Manager Who's Seen All Three)

If you've ever seen "EARTH FAULT" flash on an ABB VFD keypad and felt that knot in your stomach, you're not alone. It's one of the most common alarms, and also one of the most misleading.

The problem is that "earth fault" on an ABB drive isn't a diagnosis—it's a symptom. It can mean a fried motor cable, a damp motor, or a misconfigured parameter. Treating them all the same way is how you end up replacing a perfectly good drive.

After reviewing about 200 fault-related service reports in Q1 2024 alone—and walking through this exact issue with a client who was about to scrap a brand-new ACS580—here's how I break it down. The question isn't "is there an earth fault?". The question is "which kind of earth fault do you have?".

The Three Scenarios

In my experience, 90% of ABB VFD earth faults fall into one of three categories. The fix is completely different for each, so getting the category wrong means wasting time (or money).

Scenario A: The Hardware Fault (Motor Cable or Motor)

What it looks like: The fault appears immediately when you start the drive, or within seconds. It's repeatable—every time you try to run, you get the same alarm. The drive might trip immediately on start.

What's happening: There is a genuine, measurable short circuit between a phase conductor and ground. This is usually in the motor cable (damaged insulation, pinched wire) or inside the motor itself (wet windings, failed insulation). On an ABB drive, the hardware is sensitive enough to detect leakage currents as low as a few milliamps that older drives would have ignored.

The fix:

• First, isolate the problem. Disconnect the motor cable from the drive output. Try to start the drive without a load. If the fault goes away, the problem is downstream (cable or motor). If it stays, you might have a damaged drive output stage.
• Use a megohmmeter, not a multimeter. A standard multimeter won't reliably detect insulation breakdown that only appears at high voltage. Test motor winding resistance to ground. Anything below 1 megaohm on a 480V system is suspect. I've seen motors read 0.4 megaohms and trip the drive every time.
• Check the cable routing. I once spent 2 hours troubleshooting a fault on an ACH580 only to find the motor cable was running through a conduit that had filled with water. The insulation was fine, but the water created a path to ground.

Scenario B: The Configuration Fault (Parameter Sensitivity)

What it looks like: The drive runs fine for a few minutes, or even hours. Then it faults. It's inconsistent—sometimes it happens, sometimes it doesn't. Maybe the weather is humid, or the motor is running hot.

What's happening: This is the one that catches most people off guard. ABB drives (especially ACS880 and ACS580 series) have configurable earth fault sensitivity. If the parameter is set too sensitively for your specific installation, the drive will trip on leakage current that is electrically real but operationally harmless. Long motor cables—anything over 100 meters—naturally have higher capacitive leakage current. Standard drives set to "low leakage" mode can trip on this.

The fix:

• Check parameter 31.02 (ACS580) or the equivalent fault settings. This allows you to adjust the earth fault response. Some applications need to be set to "High leakage" or "Reduced sensitivity" specifically because of long cable runs.
• Log the fault with a timestamp. I keep telling field techs: write down what the drive was doing when it faulted. "It was running at 45Hz for 3 hours on a hot day" is useful information. "It just faulted" is not.
• Consider putting a reactor at the output. In installations with very long cables (over 200m), a dV/dt filter or output reactor reduces the peak voltage stress and capacitive leakage current. I've seen this fix intermittent faults that no amount of parameter tweaking could solve.

Scenario C: The Grounding Problem (Installation Issue)

What it looks like: The fault happens randomly. It might correlate with other equipment starting/stopping (compressors, large motors). It might happen when you touch certain parts of the cabinet.

What's happening: The drive's ground reference is not stable. This is surprisingly common when the drive is installed on a system that uses non-linear power supplies (VFDs drive other VFDs, for example). The grounding system has noise or potential differences between the drive's ground and the motor's ground.

The fix:

• Verify a single-point ground. All equipment—drive, motor, and any filters—should terminate to the same ground reference point. Ground loops are a common source of nuisance earth faults.
• Check the incoming power quality. I once troubleshooted a fleet of ABB drives that intermittently showed earth faults. The issue wasn't the drives—it was a failing transformer on the utility feed that was creating harmonics. The drives were doing exactly what they were supposed to do: detecting leakage that shouldn't have been there.
• Look at the shielded cable termination. If you're using shielded motor cable (which ABB recommends for EMC compliance), make sure the shield is grounded at one end only—preferably the drive end. Grounding at both ends creates ground loops that can trigger earth fault detection.

How to Know Which Scenario You're In

Here's a quick decision tree I use when I get a call about an earth fault:

1. Does it happen every time you try to start? → Scenario A (Hardware). Go get the multimeter.
2. Does it happen after running for a while, or in specific conditions? → Scenario B (Configuration). Check your cable length and parameter settings.
3. Is it completely random, or tied to something else in the plant? → Scenario C (Grounding). Check your installation before blaming the drive.

What I've learned after 5 years of dealing with these faults is that the drive is usually telling the truth—the leak path is real. The question is whether that leak path is a problem you need to fix in the cable, a sensitivity you need to adjust in the configuration, or a ghost you need to exorcise from the grounding system. Treating the symptom instead of the scenario is how perfectly good motors get replaced and perfectly good drives get sent back for warranty. And trust me—I've seen both.

Prices as of January 2025; verify current rates. Megohmmeters typically range from $150-$600 for a quality unit adequate for 480V motor testing.