VFD Harmonics

A six-pulse VFD rectifier doesn't sip current as a sine — it gulps it in pulses, injecting harmonic currents at h = 6k±1 back into the bus. Each harmonic ampere drops voltage across the source impedance, so the drive distorts the current and the system turns it into voltage distortion that every neighboring load sees. Step through the mitigation ladder, soften or stiffen the bus, and watch both waveforms react.

Mitigation
Front-end / filter

Operating Point
Drive load
Short-circuit ratio Isc/IL

SCR is bus stiffness: available fault current ÷ drive full-load current. Low SCR = soft bus, Xs = 1/SCR pu on the drive base.

Waveforms — Two Cycles at the Bus
Bus voltage (pu) Drive current (÷ I₁ peak) Fundamental current only
Harmonic Spectrum
Current Ih/I₁ (%) Voltage Vh/V₁ (%) — current × h/SCR
Readout

The Physics

Why pulses, not sines. A diode bridge only conducts when the incoming line voltage exceeds the DC-link capacitor voltage — a narrow window near each line-to-line voltage crest. The drive therefore draws two short gulps of current per half-cycle. A periodic non-sine current is a fundamental plus harmonics; the gulps are just how h = 5, 7, 11, 13… look when stacked.

Why h = 6k±1. Three-phase symmetry (three identical currents 120° apart) cancels every triplen (3rd, 9th, 15th…), and half-wave symmetry kills the even orders. What survives a p-pulse rectifier is h = pk±1: a 6-pulse bridge injects 5, 7, 11, 13, 17, 19…, each roughly shrinking as 1/h. A 12-pulse connection — two bridges fed 30° apart — makes the 5th and 7th of one bridge arrive in antiphase with the other's, cancelling them and leaving h = 12k±1.

THDi = √(Σh≥2 Ih²) / I₁  ·  Irms = I₁·√(1 + THDi²)

Current becomes voltage distortion. Source reactance rises with frequency, so each harmonic ampere drops h times more voltage than a fundamental ampere. On the drive's own base, Xs = 1/SCR pu — a stiff bus (high SCR) shrugs the harmonics off, a soft one doesn't:

Vh = Ih · h·Xs = (Ih/I₁) · (h / SCR) pu  ·  THDv = √(Σ Vh²)

Model limit: the spectra here are fixed per mitigation option, but on a genuinely soft bus the source reactance itself acts like a line reactor and tames the drawn current — so the no-choke + low-SCR corner of this model overstates THDv. Treat anything past ~10% as "this combination needs a real study", not as a prediction.

Power factor without power. Harmonic currents carry no average power (they have no matching voltage component to multiply against in a stiff source) — but they do carry rms amperes, so they heat cables and transformers and eat into true power factor even when the displacement angle is nearly zero:

PF = DPF / √(1 + THDi²)   (DPF ≈ 0.98 for a diode bridge)
🪜 The mitigation ladder. Each rung buys lower THDi for more money and panel space: nothing (THDi ≈ 95%) → 3% line reactor (≈ 44%, one cheap part — longer, gentler conduction) → 5% reactor / DC choke (≈ 35%, diminishing returns) → 12-pulse (≈ 12%, phase-shifting transformer cancels 5th & 7th) → active front end (≈ 4%, IGBTs synthesize a near-sine at the price of cost, losses, and EMI filtering). Walk the segmented control up the ladder and watch the spectrum drain.