DC Bus Overvoltage on a High-Inertia Fan Drive

Kinetic energy to be removed. ω = 2π × 1,480 / 60 = 154.99 rad/s E = ½ J ω² = 0.5 × 22 × 154.99² = 264,000 J ≈ 264 kJ

Average braking power over the 8 s ramp. P_avg = 264 kJ / 8 s = 33 kW, less motor and drive losses and fan windage, so the resistor will see roughly 230–250 kJ in practice. Use 264 kJ as the conservative design figure.

Why the bus capacitors cannot take it. The energy stored between the nominal and trip levels is: E_cap = ½ C (U_trip² − U_nom²) = 0.5 × 4,700 µF × (800² − 540²) = 0.00235 × 348,400 ≈ 820 J. That is about 0.3% of the 264 kJ being returned. The 6-pulse diode rectifier is a one-way bridge, so the energy cannot go back to the supply; it charges the bus until the 800 V trip operates.

Corroboration. The fault appears only on deceleration, the supply is stable, and no fault is found on the drive or motor. That is the signature of regeneration, not of a supply transient. Note also that 800 V is close to the practical limit for the series-connected DC bus capacitors in a 400 V drive, so the trip is a protective function and must not be raised or disabled - the correct fix is to remove the energy, not to raise the threshold.

One check before proceeding: confirm with a scope on the DC bus and a current log that the motor is actually regenerating rather than being driven by an overhauling load. If the fan can overhaul, the drive may see continuous regeneration, which changes the resistor duty completely.

Choose a chopper pulse power above the average braking power. A pulse power around 1.3× average allows the chopper to switch rather than conduct continuously: P_pulse ≈ 45 kW at 780 V R = U² / P = 780² / 45,000 = 13.5 Ω → specify 13.5 Ω (a 15 Ω standard value gives 40.6 kW and 52 A, also acceptable). With 13.5 Ω: I_peak = 780 / 13.5 = 57.8 A, which is inside the 60 A chopper rating but has little margin; with 15 Ω the peak is 52 A, which is the safer choice.

Chopper duty during the ramp. D ≈ P_avg / P_pulse = 33 / 40.6 ≈ 0.8. A high duty is normal for an 8 s ramp; what matters is that the resistor's thermal mass and the chopper's I²t rating accept it.

Why the rating is not 33 kW. The continuous rating of a brake resistor is set by its ability to dissipate heat to the surrounding air over a cycle, not by the instantaneous power: Energy per cycle / cycle time = 264 kJ / 120 s ≈ 2.2 kW. So a 3–5 kW continuous resistor with a specified short-time overload (for example 40 kW for 10 s) is the correct selection. It must be mounted outside the enclosure, on a heat-resistant surface, with a thermal switch wired to trip the drive, and cabled for the peak current (10 mm² minimum for 52 A in a short duty).

Chopper settings.

• Switch-on threshold about 760 V, switch-off about 740 V, both comfortably below the 800 V trip.
• Verify the drive's overvoltage trip is not raised as a workaround.
• Enable the drive's braking-resistor protection function (thermal I²t model) if it has one, and set it to the resistor's actual continuous rating.
• Check the chopper's I²t against 57.8 A for 8 s at 80% duty.

Verification after fitting: DC bus trace on a deceleration, resistor case temperature after ten cycles, and panel internal temperature with the drive at 42 °C ambient.

Options in order of cost:

1. Brake chopper plus resistor (selected). Lowest capex, all energy to heat; recovers nothing. Acceptable because 2.2 kW average is small.
2. Regenerative front-end or active front-end drive. Returns the 2.2 kW average to the supply and removes the resistor, but costs a large fraction of a new drive and requires supply-side harmonic compliance with the network operator.
3. Common DC bus with another drive on the same process that is accelerating while this one stops - only viable if the cycle times genuinely coincide.
4. Mechanical brake or a holding brake. Solves holding, not stopping; introduces wear and maintenance.
5. DC injection braking. Dissipates energy in the motor rotor, is ineffective at low speed, and risks motor overheating on a high-inertia stop. Not suitable here.

Checks that apply whichever option is chosen:

• Torque capability. Braking torque = 33 kW / 155 rad/s ≈ 213 N·m, about 44% of the motor's rated torque. That is within the motor's capability, but the available braking torque at low speed must be confirmed against the fan's load curve.
• Overhauling load. If the fan is driven by the process, the drive may not be able to follow the 8 s ramp at all, and the resistor becomes continuously rated. Confirm from the process data.
• Motor thermal duty. Frequent braking increases motor losses; check the motor's duty cycle rating.
• Panel ambient. At 42 °C, the chopper and drive are already derated; adding a resistor inside the enclosure is not acceptable.
• Control interaction. With overvoltage stall control disabled, the drive will trip rather than extend the ramp. Leaving it disabled is correct here because the customer needs a defined stop time, but it must be a documented decision.

Risk if the resistor is fitted without the checks: the resistor overheats on a cycle where the fan overhauls or where the ramp cannot be followed, and the resulting failure is a fire risk rather than a nuisance trip.