I got a call the other day from an engineer who said their ABB ACS580 VFD kept tripping on overvoltage. They were already looking at 5 HP VFDs online and wanted my opinion on a replacement. I asked what was on the output side. Pause. A pump, maybe.
That's when I knew this was going to be a lesson in something other than drives.
The Surface Problem: Everyone Blames the Drive
Over six years of managing procurement for a mid-sized industrial services company, I've signed off on roughly 200 VFD repairs and replacements. ACS355s, ACS580s, ACS800s, ACH580s—you name it. I don't have hard data on industry-wide misdiagnosis rates, but based on our own history, my sense is that at least 40% of the drives we pulled and replaced had a correctable problem outside the drive.
Here's the thing: a VFD is not a light bulb. It doesn't just burn out on its own. It fails because something else made it fail—and if you don't fix that something, the replacement will fail too.
I wish I had tracked the split between drive-level faults and system-level faults more carefully. What I can say anecdotally is that the most common ABB VFD problem in our shop was not the VFD at all. It's the machine around it.
Deep Causes: What's Actually Making the Drive Fault
When a drive trips, the panel gives you a code. That code is the end of a conversation, not the beginning. You have to work backward.
In my experience, the real culprits fall into a few ugly familiar categories:
- The load is telling the drive lies. A pump with cavitation, a fan with a tensioning issue, a conveyor with a mechanical bind—these all look like overcurrent or overload to the drive. The drive is protecting itself. It's not the problem.
- The environment is hostile. Heatsink dust, a blocked cooling fan, high panel temperature, humidity. These are the easiest to miss because the drive may trip only after hours of running.
- The installation is electrically noisy. Poor grounding, missing line reactor, long unterminated motor cables. The drive sees voltage spikes and throws fault codes that make no sense.
- The control circuit is the weak link. A noisy sensor, a corroded terminal, a floating analog signal. A $200 sensor once created a $4,000 drive replacement.
To be fair, some drives genuinely die. Lightning, flood, a catastrophic short. But in my experience—and I'm using that word intentionally—a large portion of ABB VFD failures are really the drive sending us a warning we didn't understand.
One example: In Q2 2024, we replaced an ACS580 that had tripped on ground fault three times. The repair shop found the motor insulation was breaking down, not the drive. The motor was 30 years old. Replacing the VFD first was like replacing a fuel filter because the engine won't start, when the actual problem is the fuel pump.
This is a good place to say the obvious: if you've never had to search for 'how to replace fuel filter' on your genset, you probably don't think much about upstream causes. But the mindset is exactly what VFD troubleshooting requires. The filter, like the drive, is usually the victim. Start upstream.
I should add that not every translation is obvious. A Dometic marine air conditioner control panel, for instance, is not a VFD at all. It's a control and display unit. I once had a marina operator call me about a bad VFD on a boat's air conditioner. It turned out the unit didn't have a VFD; the Dometic control panel was throwing a fault because of a bad relay on the control board. If he'd replaced the panel first without checking, he'd have been no better off—and $400 lighter. The lesson applies broadly: identify the actual control architecture before you buy parts.
The Cost of Getting It Wrong
Let's talk about what this costs. I don't mean list price. I mean the total cost of the wrong decision.
A 5 HP ACS580 drive is typically in the $800–$1,400 range depending on options and enclosure type (based on quotes I've reviewed from authorized distributors, March 2025; verify current pricing). That's not a shocking figure. But the real cost shows up in three places: the replacement drive itself, the labor to swap it, and—this is the one that hurts—the downtime while you wait for the replacement to arrive.
I keep a spreadsheet of every VFD-related invoice, going back to 2019. Total spend: about $180,000. When I audited our 2023 spending, I found that roughly $22,000—or 12% of our cumulative VFD budget—went toward replacements that could have been avoided with better diagnosis. In 2023 alone, that was $7,300.
I know $7,300 doesn't sound huge in the grand scheme of an industrial budget. But it buys a lot of training. It buys a lot of spare parts. It buys a lot of late-night service calls that shouldn't have been necessary.
The worst part is the decision pressure. I remember one case where we had a down line, a customer waiting, and a drive that kept tripping on overcurrent. The upside of swapping the drive fast was getting the line running. The risk was replacing a faulty drive and still having the line trip the new one. I kept asking myself: is $1,300 worth potentially losing a day of production? In that case, it wasn't—we swapped the drive to get running, then bench-tested the old one. The old drive was fine. The problem was a mechanical bind that showed up intermittently.
To be fair, that wasn't a wasted $1,300 in the long run; we now check mechanical loads before we authorize VFD replacements. But it was a memorable way to learn.
What I Actually Recommend Now
I'm not going to tell you never replace a VFD. That would be absurd. I'm going to tell you what I tell our own buyer:
- Get the fault code and the exact time of the fault. That's free.
- Check the motor and the mechanical system before you check the drive. A megohmmeter and a good look at the coupling are worth more than a new drive.
- Inspect the drive's environment: heatsink, cooling fan, panel filters, ambient temperature.
- Look at the last few events in the drive log. According to ABB's ACS580 documentation (library.abb.com, accessed March 2025), the basic control panel lets you review the fault history. It often tells you more than a screenshot of the current code.
- If the drive is under warranty, don't open it. Use the ABB VFD rep locator (via abb.com/drives, verified March 2025) to find an authorized service provider. I've used it twice; the directory lists local reps, service partners, and panel builders. My one caveat: call first and ask if they actually have a drive repair bench.
If the drive is physically damaged—burned terminals, visible board failure, flooded—then yes, replace it. If the repair estimate is more than half the price of a new unit, replace it. If the drive is still under warranty, definitely replace it through official channels.
I should also be honest about when a new ABB VFD isn't the right answer. I generally like the ACS580 for fan and pump applications; the built-in pump logic is genuinely useful, and the panel is easy for technicians to navigate. But if you have a simple fixed-speed motor that only needs start/stop and basic protection, you probably don't need an AC drive at all. And if you need a niche low-speed/high-torque drive, a general-purpose VFD may not be the best fit.
My experience is based on about 200 repairs and replacements across ACS355, ACS580, ACS800, and ACH580 units. If you're working with a completely different product line or a custom motion-control system, your experience might differ.
Bottom Line
The drive is not always the problem. Treat it as the messenger, not the message.
Before you buy a replacement, spend an hour verifying the system around it. That hour is cheaper than a same-day VFD order, cheaper than a repair bench diagnostic, and cheaper than explaining to your maintenance manager why the new 5 HP VFD is faulting on the same code as the old one.
And if someone hands you a part number without telling you what the drive's fault log says, ask to see the log. It's probably worth more than the quote.
