Why Your Danfoss VFD Keeps Tripping on Alarm 4 – And What the Manual Doesn't Tell You

The Call at 4:47 PM on a Friday

I got the call in late February 2024. A system integrator in Ohio had a line down—an HVAC chiller plant serving a data center. The Danfoss VLT FC 102, a 50 HP unit, was flashing Alarm 4 (Mains phase loss). This particular unit powers the primary chilled water pump for a server room cooling loop. The room was already climbing toward 80°F (26.7°C). Their on-site tech had replaced the input fuses, checked the incoming voltage at the disconnect (which was fine—480V balanced), and swapped the control card from a spare unit. The alarm persisted. Normal troubleshooting time for my team is 2-3 hours for a remote diagnostic, but this client needed an answer within 60 minutes, or they were facing a $12,000 penalty clause for the data center downtime. I didn't have time to re-read the manual. I had to triage based on what usually breaks, not what the documentation says can break.

The Surface Problem: What the Manual Says About Alarm 4

Every Danfoss VFD technician knows Alarm 4: "Mains phase loss." The standard troubleshooting flowchart in the FC 102 manual tells you to check the input supply—verify voltage, check fuses, look for loose connections. That's usually correct. For about 70% of calls, it is a blown fuse or a bad breaker. But there's a second, less common root cause that the manual buries in a footnote (note to self: I need to start reading footnotes more carefully). That secondary cause is a failed input rectifier, specifically one of the six diodes in the three-phase bridge. If a single diode fails open, the drive sees a missing phase on the DC bus even though the incoming AC power is balanced. The manual will tell you this can happen. It won't tell you when it happens most often.

The Hidden Pattern: It's Not Random

In my experience coordinating rush repairs for about 200+ VFD failures over the past four years, a rectifier diode failure almost always follows one of two events:

1. A recent mains surge or brownout: The input rectifier diodes take the brunt of a voltage spike. If the surge suppressors (MOVs) on the input failed first (which they usually do, silently), the next hit goes straight to the diodes. This is the most common scenario.
2. A catastrophic pre-charge circuit failure: This is rarer, but when it happens, it usually takes out two or three diodes at once. The manual tells you to check the DC bus voltage. What it doesn't tell you is that a zero DC bus voltage reading with good AC input is the definitive sign of a failed rectifier bridge.

The Deep Cause: Why the Diagnostic Took 45 Minutes Longer Than It Should Have

The integrator's tech had checked the AC voltage at the drive's input terminals. Three phases: 478V, 481V, 479V. Perfectly balanced. He had documented this in his log (I really should have asked for that data earlier). Based on that, he ruled out a power supply issue. But here's where the logic breaks down—and this is the part the Danfoss training doesn't drill into you: You cannot confirm a healthy DC bus by measuring AC input voltage alone. The drive's internal DC bus is what the motor control logic actually uses. I had to ask him to put his multimeter on the DC bus terminals (88/+ and 89/-). The reading: 0V. Dead short. That 0V reading confirmed my suspicion: the input rectifier was gone. The alarm was real—the DC bus had no power—but the root cause wasn't upstream of the drive; it was inside the drive.

The Cost of Ignoring This (A Cautionary Tale)

Everything I'd read about Alarm 4 said it's an external supply issue. In practice, for drives operating in environments with known power quality issues (like data centers with big UPS systems switching over, or industrial sites with welding loads), internal rectifier failure is the culprit in roughly 15-20% of 'phase loss' alarms. That's based on our internal data from 200+ rush jobs. The conventional wisdom is to swap the drive first. My experience suggests you should measure the DC bus first—it takes 30 seconds and saves the $800 cost of a rush replacement that you might not need.

The Consequences of Skipping This Step

Let me be very specific about what happens if you ignore this. The integrator in Ohio had a spare drive, a Danfoss VLT FC 102, sitting on a shelf. He was about to authorize a rushed replacement—overnight shipping, plus a service call for the install, which would have run about $1,200 total (the $800 rush drive premium plus $400 for the tech). The alternative was a $200 rectifier diode kit from a local electronics supply house and an hour of labor to install it. The downside risk of the repair path was that it might not fix it. The worst case: the diode fails again because of an underlying surge issue, and we're back to square one plus $200. The best case: the drive is back online in two hours for $200. The expected value said try the repair first. His gut—based on years of 'swap the drive' culture—said swap it. I had to talk him off that ledge.

We paid $12 for the part, had it delivered by 6 PM, and the drive was back online by 7:30 PM. The data center's temperature had peaked at 84°F (28.9°C), but the equipment was safe. The client's $12,000 penalty clause was avoided. If we had swapped the drive, we would have lost 24 hours and spent $1,200. The lesson: never trust an Alarm 4 diagnosis without a DC bus voltage check. The manual tells you to check the input. The job tells you to check the DC bus.

What You Should Actually Do (Short Version)

If you're getting a persistent Alarm 4 on your Danfoss VFD—especially an FC 102 or a VLT Micro Drive that's been in service for more than three years—don't start by ordering a replacement drive. Do this first:

1. Measure AC voltage at the drive's input terminals (L1, L2, L3-N). Verify it's within spec and balanced.
2. If AC looks good, measure DC bus voltage between terminals 88 and 89. You should see approximately 1.414 x the input RMS voltage (e.g., ~675 VDC for a 480V input).
3. If DC bus voltage is zero or very low (below 300V), the rectifier bridge is almost certainly failed. Isolate the drive and test the diodes with a multimeter.
4. If you identify a shorted or open diode, a standard 35A, 1600V rectifier diode pack will cost you $15-40. Replacing it is a 45-minute job for a competent technician.

Prices as of January 2025; verify current rates and part availability. This is not a replacement for full system diagnostics, but it will save you from making an expensive, time-wasting mistake (augh, I've seen it happen too many times).