PowerFlex 525 fault codes: what the drive is measuring before it trips

I get some version of this call a few times a year. A 525 trips F004 twice a shift, and the guy on the phone has already metered the incoming line at the disconnect. 478 volts, rock steady. All true, and all beside the point. The drive never looked at his line. It watched its own DC bus, and by the time that bus dropped under 390 volts the line had done something a handheld meter averaged straight out of existence.

Most PowerFlex 525 fault codes work like that. The word on the display names a symptom. What the drive actually measured to get there sits somewhere else, and that is the part that leads anywhere useful.

Five PowerFlex 525 fault codes below, what each one watches, and the numbers it trips at.

With PowerFlex 525 fault codes, the type matters more than the number

Rockwell splits the 520-series fault list into types, and the type decides what happens in the second after the trip. Type 1 can auto-reset and run again. Type 2 cannot, so the drive sits there until somebody walks out and clears it.

F004, F005 and F021 are Type 1. F003 and F012 are Type 2.

That one line reframes half the complaints I hear. "It keeps restarting itself" and "it will not come back" are not two moods of the same problem. Different families, different hunt.

Code	Name	Type	What the drive measures	Where to look first
F003	Power Loss	2	DC bus ripple while running, or undervoltage already present at power-up	Input phases, fuses, upstream terminations
F004	UnderVoltage	1	DC bus below the minimum for its voltage class	Bus voltage in b005 at the moment it trips
F005	OverVoltage	1	DC bus above the maximum for its voltage class	Decel time, regeneration, motor insulation
F012	HW OverCurrent	2	Output current past a fixed hardware limit	Load, A530 boost setting, DC brake level
F021	Output Ph Loss	1	Output phase current against A576, and only when A557 is on	Motor leads and the motor itself

 

The five PowerFlex 525 fault codes that generate the most calls, and what sits behind each one.

Where each of these PowerFlex 525 fault codes is actually measured. Three of the five never look at the motor at all.

 

F004 and F005 are one measurement with two limits

Undervoltage and overvoltage come off the same sensor. The drive rectifies your incoming line into a DC bus, everything downstream runs off that bus, and these two PowerFlex 525 fault codes just describe which side of the window it left. F004 means the bus fell under the floor. F005 means it went through the ceiling. Neither is a statement about your line voltage, which is exactly why metering the disconnect proves so little.

Input voltage class	F004 UnderVoltage trips at	F005 OverVoltage trips at
100...120V AC	190V DC bus (about 75V AC line)	405V DC bus (about 150V AC line)
200...240V AC	190V DC bus (about 150V AC line)	405V DC bus (about 290V AC line)
380...480V AC	390V DC bus (about 275V AC line)	810V DC bus (about 575V AC line)
525...600V AC	390V or 487V DC bus, depending on the P038 setting	1005V DC bus (about 711V AC line)

 

Rockwell's published DC bus trip points for the 520-series, by input voltage class. Two PowerFlex 525 fault codes, one window.

The operating window on the DC bus. The two PowerFlex 525 fault codes that come off it sit at opposite edges of one measurement.

 

Those are the numbers worth arguing with, not the ones your meter shows at the disconnect. Put b005 [DC Bus Voltage] on the display and watch it while the machine does whatever it does right before the trip. On a 480 volt drive you are looking for 390. If the bus never comes close and you are still getting F004, you are chasing the wrong fault.

Two parameters exist so a 525 can survive a sag instead of tripping on it. A548 [Power Loss Mode] set to 1 [Decel] tells the drive to decelerate the load on a power loss and use the inertia of that load to regenerate voltage back into the bus, which buys real time on anything with a flywheel on it. A549 [Half Bus Enable] keeps the motor powered at half the normal input voltage through short sags. Neither is on out of the box. A factory-default 525 does exactly one thing about undervoltage, and that is trip.

What actually trips F005 on a machine that has been running fine

Fast decel and an overhauling load are the textbook answers and they are usually right on a new install. Shorten the stopping time on a loaded conveyor, the motor turns into a generator, the energy has nowhere to go, and the bus climbs until it hits the ceiling. The fixes are boring: lengthen decel, add a dynamic brake resistor sized for the job, or stop with a coast.

The one that catches people is the drive that ran clean for two years and then starts throwing F005 at random up at speed. Rockwell's guidance for the 4-class drives, which take this same measurement, puts intermittent shorting inside the motor near the top of that list. Arcing in the windings happens at low current and high voltage, and it comes back down the leads as a bus event. I have opened more than one motor terminal box after a run of random overvoltage trips and found insulation cooked where nobody thought to look.

So on a machine that used to be fine, megger the motor before you touch a decel time. Check rotation while you are in there. A pump running backwards throws overvoltage on acceleration more often than it has any right to.

F012 is not a parameter problem

Of all the PowerFlex 525 fault codes, F012 HW OverCurrent is the one where reading through the parameter list is wasted effort. It is a hardware limit. Current sensing shuts the output stage down before the processor gets a vote, and Rockwell is direct about it across the PowerFlex family: the hardware overcurrent trip is fixed and cannot be defeated or mitigated. No threshold to raise, no timer to lengthen. It is Type 2 as well, so it will not auto-reset.

What you can change is what the drive is being asked to do. Rockwell's corrective action for F012 on the 520-series comes down to excess load, an A530 [Boost Select] setting pushing too much voltage at low speed, and DC brake volts set too high. Boost is the first thing I look at when a drive trips F012 the instant it starts. Somebody cranks it to get a sticky load moving, and the drive shoves too much current into a motor that has not started turning yet.

If it trips F012 with the motor leads off the drive, stop diagnosing. That is the power stage.

F021 does nothing until you switch it on

Here is where the PowerFlex 525 fault codes stop being a reading problem and turn into a configuration problem, and it costs people real time. Output phase loss detection on a PowerFlex 525 is off by default. A557 [Out Phas Loss En] has to be set to 1 before the drive will ever raise F021 at all. So "it never gave me a phase loss fault" is not evidence that all three output phases are there. Nine times out of ten it is evidence that nobody turned the check on.

From firmware 5.001 up there is a second parameter, A576 [Phase Loss Level], which sets the threshold each motor phase has to exceed. Drop the percentage and you make it less sensitive. On a lightly loaded motor that is sometimes the difference between a real detection and a nuisance trip.

Then the part that gets mixed up constantly. The 520-series has output phase loss protection only. There is no input phase loss protection in it, none at all. Lose a phase on the line side and this drive will not tell you with F021.

A lost input phase reaches the drive as bus ripple and comes back out as F003. F021 only ever watches the output side, and only once A557 turns it on.

 

F003 and F004 are not the same fault

These two PowerFlex 525 fault codes both mean low bus voltage, so the names get used interchangeably out on the floor. The drive does not treat them that way.

F003 Power Loss fires in two situations: an undervoltage condition already present as the drive powers up, or excessive bus ripple while it is running. Rockwell's description for the 520-series is blunt about the usual cause, single phase operation detected with excessive load, and ripple is how the drive infers it. On a 460 volt system, d380 [DC Bus Ripple] sitting above 60 volts for any stretch of time is what pushes it over the line.

F004 UnderVoltage is the other story entirely. The drive powered up clean, ran clean, and the bus dropped later.

The practical difference is the type again. F003 is Type 2 and stays down. F004 is Type 1 and can be set to restart on its own with A541 [Auto Rstrt Tries] and A542 [Auto Rstrt Delay]. A drive that comes back by itself after a brownout and a drive that has to be reset by hand are telling you two different things about your incoming power, and the display already said which.

FAQ

Which PowerFlex 525 fault codes reset themselves, and which ones stay down?

Type 1 can auto-reset, Type 2 cannot. That puts F004, F005 and F021 in the group that will restart on its own once A541 [Auto Rstrt Tries] is set above zero, and it leaves F003 and F012 in the group that stays down no matter what you program.

The line voltage checks out and the PowerFlex 525 still trips F004. Where do I look?

Stop metering the disconnect and read b005 [DC Bus Voltage] off the keypad while the event is happening. A sag short enough for a handheld meter to average away is still long enough to pull a 480 volt bus under 390. If you want the drive to ride it out rather than trip, that is what A548 [Power Loss Mode] and A549 [Half Bus Enable] are there for.

Can any of the PowerFlex 525 fault codes be turned off or moved?

F021 works the other way round. It is off until you enable it with A557 [Out Phas Loss En], and its threshold moves with A576 [Phase Loss Level]. F012 cannot be adjusted or defeated at all. The undervoltage and overvoltage trip points are fixed by voltage class.

Does the PowerFlex 525 detect a lost input phase?

Not directly. The protection it carries sits on the output side only. A missing input phase normally shows up as F003 Power Loss, driven by the bus ripple that single-phase operation creates once there is load on the motor.

Before you order a replacement drive

Most of the PowerFlex 525 fault codes above point at something outside the drive: a feeder that sags, a motor with cooked insulation, a decel time nobody has touched since commissioning. Rule those out first, because swapping a drive that was correctly reporting a real problem just moves the same fault onto newer hardware. When the power stage genuinely is gone, we keep PowerFlex 525 drives across the voltage classes on the shelf, tested and ready to ship.