When an ABB ACH580 on a supply fan or a chilled-water pump goes to a solid red LED and drops the load, the plain-text line on the Hand-Off-Auto panel is only half the story. The four-digit code beside it points at a specific protection function, and knowing which one saves you from the two failure modes we see most in Central Florida mechanical rooms: guessing at a bad drive and swapping it (expensive, usually wrong), or repeatedly resetting a fault that is protecting the motor from a real problem (cheap, and it burns up a fan motor). This is a code-by-code field guide for the faults that actually show up on HVAC fans and pumps, with every meaning traced to ABB's ACH580 Installation, Operation and Maintenance Manual and the shared ACS580/ACH580 firmware fault list. Nothing here is invented; if a code is not in ABB's table, it is not in this guide.

For a refresher on why these drives are on the fans and pumps in the first place, the motor-control and energy fundamentals, see our pillar explainer, Variable Frequency Drives in the HVAC Market. If you also run other drive lines, the companion pieces walk the same first-checks discipline: Danfoss FC-102 Fault Codes on HVAC Fans and Pumps and Yaskawa Z1000 Fault Codes on HVAC Fans and Pumps. This post assumes you already know what a VFD does and need to get one running again.

How the ACH580 reports a fault

The ACH580 separates warnings from faults. A warning (an A-code, for example A780) means the drive detected something off but keeps modulating, the fan or pump stays running. A fault stops modulation immediately to protect the drive, motor, or cable, and the panel LED goes solid red. On a BACnet or Modbus-integrated job, that same code lands in the BAS as a numeric alarm, so the number you read on the panel is the number your controls contractor sees. Faults latch until reset, and (this is the part people skip) a fault that re-trips on reset is almost always reporting a real condition, not a nuisance. Reset once to confirm; if it comes back, troubleshoot before you reset again.

The overcurrent family: 2310, 2340, 3381

2310 Overcurrent is the single most common fault we clear in the field. Per the ABB manual, it means output current has exceeded the internal fault limit, and the manual explicitly lists an earth fault or an output phase loss as possible root causes, so 2310 is not always a "too much load" problem. First checks, in order:

• Mechanical load first. On a fan, a seized bearing, a bird-screen or coil packed with cottonwood, or a backdraft damper that failed closed will spike current on acceleration. On a pump, a closed isolation valve or an air-bound casing does the same. Bar the shaft over by hand with the drive locked out.
• Acceleration time. A ramp that is too aggressive for the connected inertia (a big plenum fan wheel is a lot of WR²) trips 2310 on start. Lengthen parameter 23.12 (Acceleration time 1) and re-test.
• Motor cable. Megger the motor leads to ground and phase-to-phase. A nicked cable in a wet Florida roof curb reads fine until it rains, then trips 2310 or 2330.
• Sizing. Confirm the drive frame is rated for the motor FLA. A drive dropped a size to save cost will trip on any real load.

2340 Short circuit is the harder-fault cousin: ABB defines it as a short circuit in the motor cable(s) or motor. This is not a reset-and-run condition. Lock out, disconnect the motor leads at the drive, and megger both the cable and the motor windings before you re-energize. The manual also notes to verify there are no power-factor-correction capacitors in the motor cable. Those belong on the line side of a VFD, never between drive and motor. 3381 Output phase loss means the drive sees a missing motor connection; check that all three output leads are landed and tight, a classic after a contactor or bypass swap.

Ground faults: 2330 and 3181

2330 Earth leakage is ABB's ground-fault trip: the drive detected load unbalance typically due to an earth fault in the motor or motor cable. In roof-mounted AHU and cooling-tower fan applications, this is the fault that finds moisture intrusion and insulation breakdown before the motor fails outright. First checks: measure insulation resistance of the motor and cable to ground (below ~1 MΩ is suspect on a running motor), look for the same PFC-capacitor problem noted above, and confirm long motor cables are not driving capacitive leakage. The manual notes scalar control mode as a diagnostic step to isolate whether the trip is genuine load unbalance versus a measurement artifact. Do not defeat 2330 to keep a fan running: a ground fault on a wet motor is a shock and fire path.

3181 Wiring or earth fault is different and worth knowing cold: ABB defines it as incorrect input power and motor cable connection, i.e., line and load leads crossed, or a motor cable landed on the drive's input terminals. It shows up on new installs and after panel rework, not on a system that has been running for years. If you get 3181 on a drive that was fine yesterday, suspect that someone moved wiring.

DC-bus faults: 3210, 3220, 3130

These three tell you about power quality upstream of the drive, which matters in Florida where summer afternoon thunderstorms and utility switching hit the grid hard.

• 3210 DC link overvoltage: excessive intermediate-circuit DC voltage. On a high-inertia fan that is decelerating, the motor regenerates into the bus and pushes it over the limit. Lengthen the deceleration time (parameter 23.13), confirm overvoltage control is enabled, and verify supply voltage matches the drive rating. If the fan must stop fast, the fix is a brake chopper and resistor, not a longer ramp.
• 3220 DC link undervoltage: intermediate-circuit DC voltage is not sufficient, per ABB due to a missing phase, blown fuse, or rectifier fault. Check the incoming three-phase, the input fuses, and the disconnect. A single dropped phase from a corroded lug is the usual Florida culprit.
• 3130 Input phase loss, ABB: DC voltage is oscillating due to a missing input line phase or a blown fuse. Same hunt as 3220: input fuses, loose power connections, and supply imbalance. A drive that trips 3130 under load but not at idle is telling you one input phase is intermittent.

Overtemperature: drive power unit (4210, 4290, 4310, 5081) vs. motor (4981)

Florida runs these drives hard, and ABB splits over-temperature into two distinct groups that are easy to confuse. The first is the drive's own power unit running hot; the second is a motor temperature sensor. Reading the code tells you which enclosure to open. In an unconditioned rooftop AHU section or a mechanical mezzanine that hits 105 °F in August, the drive-side faults are a seasonal event, not a defect.

• Drive power-unit over-temperature. 4210 IGBT overtemperature means the estimated drive IGBT temperature is excessive; 4290 Cooling means the drive module temperature is excessive; 4310 Excess temperature means the power-unit module temperature is excessive. All three are the drive, not the motor. The checks are the same: verify air flow and cooling-fan operation, check the heatsink fins for the fine dust that coats every Florida mechanical room (the number-one cause, blow it out), confirm ambient temperature is within the drive rating, and check motor power against drive power. If the drive's cooling fan itself has failed, you will typically see 5081 Auxiliary fan broken, an auxiliary cooling fan is stuck or disconnected, which is the code to act on for a dead fan.
• Motor over-temperature. 4981 External temperature 1 is your motor-temperature protection when a motor thermistor/PTC or sensor is wired to the drive's external-temperature input: measured temperature 1 has exceeded the fault limit. ABB's guidance is to check parameter 35.02 (the measured-temperature 1 source) and the cooling of the motor. On a belt-drive fan held below ~30 Hz for long stretches, a TEFC motor loses much of its own shaft-fan cooling, so 4981 is often reporting a real motor condition, not a sensor glitch. This is the fault that ties directly to how the drive is run, which is worth understanding before you widen a limit (see the affinity-law note below).

Motor protection: 7122 Motor overload and 7121 Motor stall

7122 Motor overload means motor current is too high, the drive's thermal model has integrated enough over-FLA current (the I²t curve) to protect the motor. This is the one people are most tempted to reset away, and it is the one most likely to be saving a motor. Before you touch parameters 35.51 to 35.56, find out why the motor is drawing more than nameplate: a fan pushing against a dirty filter bank, a pump with a failed check valve short-circuiting flow, or a motor that is simply undersized for the static it now sees after a duct change. 7121 Motor stall, motor is operating in the stall region because of, for example, excessive load or insufficient motor power, is the low-speed, high-torque version of the same story: a fan or pump that cannot break away. Fix the mechanical cause; don't just widen the stall window.

Comms and safety: 6681, 7510, 7081, 5091

These don't mean the drive is broken: they mean it lost a signal it needs.

• 6681 EFB comm loss: communication break in the embedded fieldbus. On a BACnet MS/TP or Modbus RTU job, check the fieldbus master and the wiring to terminals 29, 30, 31. A daisy chain with a missing terminating resistor drops out under electrical noise.
• 7510 FBA A communication: cyclical communication between the drive and the fieldbus adapter module A or PLC is lost. Check the adapter module seating and parameter groups 50 to 53.
• 7081 Control panel loss: the panel (or PC tool) set as the active control location stopped communicating. If the drive is in Hand and the panel cable is loose, the drive faults rather than run blind. Reseat the panel.
• 5091 Safe torque off: the STO safety circuit opened during start or run. This is a safety function doing its job: check the STO wiring and the device that opened it (a firestat, a smoke-detector relay, an E-stop) before assuming a drive fault.

Why the low-speed fault checks matter: the affinity laws

Several of the checks above (long low-speed running that overheats a TEFC motor, a fan that stalls at low Hz) trace back to how these drives save energy in the first place. Fan and pump power follows the cube of speed: the U.S. Department of Energy's guide Variable Speed Pumping: A Guide to Successful Applications states that the energy a fan or pump needs varies with the cube of motor speed, which is why modulating speed to match demand produces such large savings. Slow a fan to 80% speed and shaft power falls to roughly 0.8³ ≈ 51%, about half the draw for four-fifths of the airflow. That cube law is the entire economic case for putting an ACH580 on a constant-volume AHU in a 4,000-plus cooling-hour Florida climate. But the same physics means a drive spends much of its life at reduced speed, where a self-cooled motor gets less airflow over its own frame, which is exactly why the motor-side faults 4981 and 7122 are worth taking seriously instead of resetting.

A reset discipline for the field

One rule keeps you out of trouble: read the code, decide whether it is protecting something, and only then reset. Overcurrent (2310), overload (7122), stall (7121), ground fault (2330), and short circuit (2340) all point at a physical condition, find it first. Overvoltage (3210) and the DC-bus and comms faults are usually a settings or wiring hunt you can work methodically. Log the code, the date, and what you found; on a fleet of rooftop units, the pattern in those logs (every 2310 on the same fan every August) tells you more than any single trip. If a fault re-trips after one confirming reset and the mechanical and electrical checks above come back clean, that is the point to escalate to a drive-level diagnosis, not before.

Serving Orlando and Central Florida, we supply, program, and service ACH580 drives on commercial AHUs, cooling towers, and chilled- and condenser-water pumps. If a code in this list is re-tripping on your equipment, we can walk the first checks with your team or come turn a wrench.