Stoklink Technical Articles

VFD Braking: Dynamic, Regenerative and DC Injection

What is VFD braking? VFD braking is how a drive removes the energy a decelerating motor pumps back onto the DC bus, using one of three methods — dynamic braking (an IGBT chopper switching a resistor across the bus), regenerative braking (an active front end returning the energy to the line), or DC injection (a low-frequency current that holds the rotor, not true braking) — each judged against the bus overvoltage trip point, typically near 800-830 VDC on a 400 V supply. Pick the wrong method and the drive faults on overvoltage mid-ramp-down, or a resistor rated for occasional stops overheats on a crane duty cycle it was never sized for. This article covers why the bus voltage rises, chopper-and-resistor sizing, when an active front end earns its cost, DC injection's real role, and how load type (overhauling or not) decides the answer.

Why a Decelerating Motor Raises DC Bus Voltage

Cut the output frequency and the motor's synchronous speed drops faster than the rotor's mechanical inertia allows it to follow. For an instant the rotor spins faster than the field and acts as a generator: slip goes negative, and current flows back through the inverter's IGBTs into the DC bus capacitors instead of out to the load.

A standard 6-pulse diode rectifier front end is a one-way valve. It rectifies AC to DC without trouble, but a diode bridge cannot conduct in reverse to push that returning energy back onto the line. With nowhere else to go, the energy stacks up on the bus capacitors and the bus voltage climbs — on a nominal 400 V line, the bus normally sits around 540-560 VDC (1.35x line voltage), and an unmanaged deceleration can push it past 800 VDC within a few hundred milliseconds.

Every drive has an overvoltage trip threshold protecting the bus capacitors and IGBTs from that rise. Cross it and the drive faults — deceleration stops, the load coasts, and on a hoist or a centrifuge that fault is not a nuisance, it is a safety event. The three braking methods below exist to keep the bus under that threshold during any deceleration where the load, not the motor, is doing the driving — see how a VFD works for the rectifier-DC bus-inverter path these methods act on.

Key takeaway: A drive with no braking provision will trip on overvoltage the moment a load decelerates faster than its own friction and windage can absorb — check the load's overhauling behavior before specifying a bare drive with no chopper or active-front-end option.

Dynamic Braking: Chopper and Resistor

Dynamic braking adds an IGBT (the brake chopper) across the DC bus in series with an external resistor. When bus voltage crosses a set threshold — commonly a few percent above the normal bus level — the chopper turns on, connects the resistor across the bus, and burns the excess energy as heat. It turns back off once voltage drops below the threshold, holding the bus near its trip point through the deceleration ramp.

Most general-purpose drives up to a few tens of kW have the chopper built into the power module — ABB's ACS580 and ACS880, Schneider's ATV340 and ATV630, and Siemens' G120C all offer an internal or plug-in chopper option, available through Stoklink's variable frequency drives collection. The resistor itself is almost always external, mounted in free air or a separate enclosure, because the heat it dissipates has to go somewhere the drive's own heatsink was not sized for.

Sizing the Brake Resistor

Two numbers matter: peak power during the braking pulse, and average power over the full duty cycle. Peak power sets the minimum resistance the chopper's IGBT can switch without exceeding its current rating; average power sets the resistor's continuous wattage rating and, in turn, whether it needs a fan.

Formula: Brake Resistor Peak Power — Source: brake resistor sizing method (drive vendor application guides)

Pbr = Vdc² / Rbr

Symbol Description Unit
Pbr Peak power dissipated in the brake resistor while the chopper conducts W
Vdc DC bus voltage clamped at the chopper's turn-on threshold V
Rbr Brake resistor resistance Ω

Average power follows from the duty cycle: Pavg = Pbr x ED%, where ED% is the fraction of a repeating cycle the chopper actually conducts. A resistor sized only for peak power and dropped into an application with frequent stops — a packaging line indexing every few seconds — runs continuously hot and fails early. Vendors publish ED% derating curves precisely because the same resistor behaves differently on a once-an-hour stop versus a six-times-a-minute index cycle.

Duty cycle (ED%) is the ratio of braking (on) time to total cycle time, expressed as a percentage, used to derate a brake resistor's continuous power rating relative to its peak rating.

What we see in the field: undersized resistors rarely fail on the first hot day. They fail six months in, once ambient temperature and a slightly shortened cycle time stack on top of a marginal sizing decision. Size for the worst realistic duty cycle, not the nameplate one.

Regenerative Braking: Active Front End

An active front end (AFE) replaces the diode rectifier with a second IGBT bridge, controlled the same way as the output inverter — see VFD control methods for how vector and DTC algorithms apply to both bridges in an AFE-equipped drive. Instead of dissipating braking energy as heat, the AFE pushes it back through a line reactor onto the incoming AC supply at close to unity power factor. ABB's ACS880 with an active supply unit, and larger Schneider and Siemens process drives, offer this as a module or cabinet option rather than a built-in feature on compact general-purpose drives.

Active front end (AFE) is a PWM-controlled IGBT rectifier bridge that replaces a diode front end, enabling bidirectional power flow between the DC bus and the AC line, per IEC 61800-4 classification of drive topologies.

The case for an AFE is duty cycle and energy volume, not occasional stops. A crane lowering a load, a downhill conveyor, a test bench cycling a motor back and forth, or a centrifuge braking repeatedly all return meaningful energy on a near-continuous basis. Dissipating that as resistor heat wastes more than energy — it wastes cooling capacity, and on a high-duty machine the resistor enclosure can end up larger than the drive itself. An AFE recovers that energy instead, at the cost of a second power bridge, a line reactor, and a materially higher drive price.

Not every overhauling load justifies it. A single occasional-use hoist rarely returns enough energy over its life to pay back an AFE's added cost; a multi-shift production crane or a test stand running continuous regenerative cycles usually does. Run the energy-return estimate first — it's a payback calculation, not a default.

Key takeaway: Specify an active front end by duty cycle and energy volume, not by "the load overhauls" alone — a resistor and chopper handle occasional overhauling fine; frequent, sustained regeneration is where AFE cost pays back.

DC Injection Braking

DC injection is not braking in the sense of removing bus energy. It is a stopping method. Once the drive ramps the output frequency down to zero, or on command, it injects a DC current into two motor phases. That current sets up a stationary magnetic field in the stator, and the rotor's residual kinetic energy dissipates as heat in the rotor bars and windings rather than on the DC bus.

Because the dissipation happens inside the motor, not in an external resistor, DC injection has a duty limit set by motor thermal capacity, not drive or resistor rating. Vendors specify a maximum injection current (as a percentage of motor rated current) and a maximum injection time, typically single-digit seconds, precisely to avoid overheating stator windings on a motor that was not designed as a brake.

Its practical role is holding a load at zero speed against a small opposing torque, or providing a quick, low-cost stop where positioning accuracy at standstill matters more than deceleration time — an index table, a light conveyor, a fan coasting to a stop faster than natural friction allows. It does nothing for an overhauling load still generating during the ramp. That energy still has to go through a chopper resistor or an AFE before the motor ever reaches the frequency where DC injection engages.

Key takeaway: DC injection stops a motor faster than coasting but does not replace dynamic or regenerative braking on an overhauling load — it only handles the last step, holding or stopping at or near zero speed.

Choosing a Braking Method by Load Type

Start with whether the load overhauls the motor during any part of its cycle. Pumps and fans almost never do. Friction and windage absorb the kinetic energy well before the drive's own coast-to-stop ramp finishes, so many installations run with no brake resistor at all. Cranes, hoists, centrifuges, downhill conveyors, and dynamometer test stands overhaul routinely, and need dynamic braking at minimum.

Frequency of the overhauling event separates a chopper-and-resistor solution from an AFE. Some integrators default to the highest-duty resistor available "for margin," but that just moves the heat problem into an oversized enclosure with its own ventilation requirement — the real fix for high-duty regeneration is recovering the energy, not dissipating more of it.

Criteria Dynamic (Chopper + Resistor) Regenerative (AFE) DC Injection
Energy path Dissipated as heat in external resistor Returned to AC line at near-unity PF Dissipated as heat in motor windings
Typical duty Occasional to moderate overhauling Frequent, sustained overhauling Zero-speed hold or quick stop only
Typical loads Occasional-use hoist, single-shift crane Multi-shift crane, test stand, downhill conveyor Index table, light conveyor, fan coast-down
Relative cost Low (chopper + resistor) High (second bridge, reactor, cabinet space) Lowest (software function, no extra hardware)

Motor thermal rating matters on the DC injection side of that table too. A totally-enclosed fan-cooled motor standing still has no forced airflow, so repeated injection cycles on a stopped motor can overheat it even within the drive's rated injection current and time.

Frequently Asked Questions

Do I need a brake resistor on every VFD?

No. Pumps, fans, and most non-overhauling loads decelerate on their own friction and windage well within the drive's normal bus voltage range, and run with no resistor fitted. A resistor becomes necessary once the load overhauls the motor during any part of the deceleration or duty cycle — cranes, hoists, centrifuges, and downhill conveyors are the common cases.

What happens if a drive overvoltage-trips during braking?

The drive faults out and the output stage stops switching, so the motor coasts uncontrolled from whatever speed it was at. On a hoist or crane that is a dropped-load risk, which is why overhauling applications get a chopper, resistor, or AFE sized for the actual deceleration profile.

Can I use DC injection instead of a brake resistor on a crane?

No. DC injection only dissipates the small residual energy left once the motor is already near zero speed; it lacks the thermal capacity to absorb the sustained regenerative energy a crane or hoist produces while lowering a load at speed. That energy has to go through a chopper resistor or an active front end before the motor ever reaches the injection stage.

Why does my brake resistor run hot on light loads?

Resistor temperature tracks duty cycle (ED%), not just peak power. A resistor sized for occasional stops but installed on an application with frequent indexing or short cycle times runs at a higher average power than it was rated for, even though each individual braking event looks modest.

Is an active front end always better than a resistor for regenerative loads?

Not for occasional use. An AFE recovers energy and eliminates resistor heat, but it adds a second IGBT bridge, a line reactor, and a materially higher price. On a hoist that overhauls a few times a shift, a resistor is usually the lower lifetime cost; on a multi-shift crane or a continuously cycling test stand, the energy recovered typically pays back the added AFE cost.

Conclusion

Braking method follows directly from one question: does the load overhaul the motor, and how often? Non-overhauling loads like pumps and fans rarely need any brake hardware. Occasional overhauling (a single-shift hoist, a light crane) calls for a chopper and a correctly sized resistor, matched to the actual duty cycle rather than just peak power. Frequent, sustained regeneration on multi-shift cranes, downhill conveyors, or test stands is where an active front end's energy recovery pays back its added cost and cabinet space. DC injection sits underneath all of that, handling only the final hold or quick stop near zero speed — a stopping tool, not a substitute for the bus-voltage management the other two methods provide. Get the duty cycle and load behavior right before the drive is ordered, and the braking method falls out of the same numbers used to size the drive itself. See the VFD engineering guide for how braking fits into the rest of the sizing process, and normal vs heavy duty overload ratings for how the same overhauling-load logic affects overload sizing.

Comments (0)

    Leave a comment