Using a Servo Stabilizer with Variable Frequency Drives (VFDs): Engineering Notes

Variable frequency drives are the workhorse of modern motor control — pumps, fans, machine tools, elevators, and HVAC all rely on them. But VFDs have their own vulnerability: the DC bus at the heart of the drive holds voltage roughly equal to the peak of the incoming AC line. Chronic grid over-voltage stresses the DC bus capacitors and IGBTs, and grid sags cause under-voltage trips that halt production. This guide explains how servo stabilizers and VFDs interact, and how to design the two together correctly.

The four VFD damage modes from grid voltage variation

How grid voltage variation damages VFDs

Failure mode	Grid voltage trigger
DC bus over-voltage trip	Line V > 110% causes bus > 800 V on 415 V VFD → fault F32 / OU
DC bus under-voltage trip	Line V < 85% causes bus dip → fault F13 / UV
Electrolytic capacitor life halving	Chronic bus V > 720 V (nominal 700) — Arrhenius + voltage stress
IGBT junction over-temperature	Cyclic voltage swings cause continuous ramp of switching loss

The last two are the silent killers — the VFD keeps running but its calendar life drops from 12 years to 5–6 years, and the failure often looks like "component random failure" until multiple VFDs on the same feeder fail on similar timelines.

Voltage tolerance published by the top VFD OEMs

VFD voltage tolerance — as published by the top OEMs

OEM	Steady voltage tolerance	Sag ride-through
ABB ACS580	-15% to +10%	300 ms typical at 70% V
Danfoss VLT AutomationDrive	-15% to +10%	200 ms at 60% V (with kinetic backup)
Yaskawa GA700	-15% to +10%	2 s with kinetic energy backup option
Schneider Altivar 630	-15% to +10%	500 ms typical

Note the asymmetry: every major OEM allows only +10% on the high side. Indian LT feeders violate this several times a week. Cumulative capacitor stress at chronic +8% to +12% is the single most common cause of "unexplained" VFD failure in Indian industry.

Where the servo stabilizer fits

Correct sequence: grid → SPD → stabilizer → distribution panel → VFDs → motors.

View data table

Source: Ozmist field measurement, food-plant pump station, 2025

With the servo stabilizer upstream, DC bus voltage sits within ±1% of nominal 705 V — well inside the safe operating window. Without protection, the bus swings between 690 V and 770 V, spending significant time in the "life-shortening" region above 720 V.

Sizing considerations specific to VFD loads

VFDs draw non-linear current, which affects stabilizer sizing:

1. Power factor — VFDs have PF near unity at full load but drop to 0.6–0.7 at low load. Size stabilizer on worst-case kVA, not on kW
2. Harmonic content — 6-pulse VFD rectifier front-end produces THD_i of 30–45%. This raises stabilizer transformer heating; add 10% derating
3. Inrush — VFDs draw large charging current at power-on. Not usually a problem if stabilizer has 25% headroom
4. Simultaneity — for VFD-heavy shops, assume 90–95% diversity because most VFDs run continuously

Stabilizer rating = Sum of VFD nameplate kVA × 1.25 headroom × 1.10 harmonic derating

For a shop with 8× 30 kVA VFDs = 240 kVA connected:

Stabilizer = 240 × 1.25 × 1.10 = 330 kVA → next standard 400 kVA

When to add a harmonic filter alongside

If the VFD front end is 6-pulse and total VFD load exceeds 30% of the transformer or stabilizer rating, THD_i can exceed IEEE 519 recommendations (5% at the PCC). Options:

Harmonic mitigation options for VFD-heavy installations

Option	Typical THD_i reduction	Notes
Line reactor 3–5% at each VFD input	45% → 32%	Cheapest; some drives ship with built-in
DC choke (built into VFD)	45% → 30%	OEM option on premium VFDs
Passive tuned filter at panel	45% → 8%	Effective, tuned to fixed frequency
Active harmonic filter	45% → 5%	Best but 3–4× cost of passive
12-pulse rectifier VFD	45% → 12%	OEM option; larger drive size

Ozmist typical recommendation: line reactors on each VFD + passive tuned filter at the main panel, all downstream of the servo stabilizer. This gives clean voltage into the VFD and clean current back into the grid.

Do VFDs and stabilizers ever conflict?

Rarely. Two edge cases:

1. Rapid load transients — a VFD accelerating a large motor causes a step change in current. If the stabilizer's servo motor cannot correct fast enough, brief voltage sag reaches the VFD. Solution: size stabilizer bigger, or use digital servo topology with faster correction
2. Regenerative braking — decelerating VFDs pump energy back into the DC bus. If the DC bus voltage rises above trip threshold, VFD faults regardless of upstream stabilizer. Solution: add brake resistor at the VFD, not related to stabilizer

Case: 240 kVA pump station in Ahmedabad

A municipal wastewater pump station in Ahmedabad ran 8 × 30 kVA VFDs driving 25 kW submersible pumps. Chronic VFD board failures: 6 boards in 2 years, ₹1.2 lakh replacement plus 4-day downtime each event.

Diagnosis: feeder voltage swung 380–460 V; DC bus voltage on 415 V drives touched 780 V under grid swells.

Retrofit: Ozmist 400 kVA oil-cooled servo stabilizer with automatic bypass, plus line reactors on each VFD input. Post-install DC bus voltage held 704–708 V steady across 12 months of operation. Zero VFD failures in the following 18 months. Payback: 14 months on avoided board replacement and pump downtime.

Frequently asked questions

Can I put the stabilizer downstream of the VFD?

No — never. VFD output is switched PWM at high frequency; the stabilizer's variable transformer wiper cannot handle non-sinusoidal input and will burn out. Stabilizer always upstream of VFD.

Do I need a stabilizer if my VFDs have active front-end (AFE)?

AFE VFDs handle wider input voltage internally (typically ±20%) and correct harmonics on their own. They still benefit from stable input, but the ROI is lower — often not worth the CAPEX unless combined with non-AFE loads.

What is DC bus voltage and why does it matter?

The DC bus is the internal capacitor bank inside the VFD holding the rectified line voltage. Its voltage is √2 × line V (approximately). Bus voltage above rating stresses the capacitors and IGBTs; too far below trips the drive.

Does a stabilizer help with grid frequency drift?

No. Servo stabilizers correct voltage only. For frequency, you need an online double-conversion UPS. Most VFDs handle frequency drift of ±3% without issue.

What about the harmonics VFDs put back on the grid?

Harmonics are current-side, not voltage-side. The stabilizer transformer sees these harmonics and heats up ~10% more than sinusoidal load. Size accordingly (10% harmonic derating factor).

Does the stabilizer need to handle motor inrush?

VFDs use soft-start ramping and rarely draw motor inrush from the line side. So no — stabilizer sizing does not need to include motor LRA if all motors are on VFDs.

Can I use one stabilizer for VFDs and non-VFD loads?

Yes, and it's common. Just ensure total connected kVA drives the sizing, and add harmonic mitigation at the VFD side to keep the stabilizer's own current cleaner.

References

1. IEEE 519. Recommended Practices and Requirements for Harmonic Control in Electric Power Systems.
2. IEC 61000-3-12. Limits for harmonic currents produced by equipment.
3. ABB. ACS580 Firmware Manual and Voltage Tolerance Table.
4. Danfoss. VLT AutomationDrive Design Guide.
5. NEMA. Motor and Drive Application Guide — Voltage Effects.
6. BIS IS 9815. Auto-transformer type servo controlled voltage stabilizers.
7. Ozmist Field Case Study. Pump station VFD reliability post-stabilization. 2025.