Differential-mode: sizes the choke against the load-current ripple target on each phase.
1Core geometry
Closed ring core. Ae/le/Wa from a rectangular (square-cut) annulus — exact for ferrite and tape-wound/nanocrystalline toroids; rounded-cross-section powder toroids (MPP/Kool Mµ) typically have ~15-30% less real Ae than this estimate for the same OD/ID/height, so treat as an upper bound for powder cores.
Ae
120mm²
le
100.531mm
Wa
452.389mm²
2Core material
3Turns & topology
Pass-through fixes N=1 per busbar (a straight busbar through a core is a single-turn winding).
Required for target: ≈ 74.54 turns
Informational/report only — CM flux from all phases adds; DM flux uses the per-conductor N above.
4Electrical operating point
f1 = 400 Hz (context only — ripple sizing uses switching frequency)
5Differential-mode sizing
Worst-case ripple occurs near D≈0.5, regardless of absolute motor speed.
Core cross-section
OD 40 mmID 24 mmToroidal · height (stack) 15 mm · 30 turns wound
Results
✗ Fails target, saturation margin, or window fit guidance
Required turns (ripple target)
74.54turns
using N = 30
Achieved inductance
81µH
Achieved ripple
123.46A pk-pk
peak current 361.7 A
Peak flux density
8.139T
678.2% of Bsat (1.2 T)
Core loss (estimate)
151.622W
12,568.44 mW/cm³
Window fill factor
39.8%
typical guidance ≤ 40%
Reference & assumptions

New to CM / DM Chokes? Read the guide: Common-Mode & Differential-Mode Choke Sizing, Explained — a plain-language explainer of the standard behind this calculator.

Toroidal and Oval/racetrack core geometry is exact closed-form from the entered dimensions. U-core and E-core geometry is approximated as a simple rectangular magnetic loop — real datasheets publish part-specific Ae/le/Wa, and there is no single universal formula for those families the way there is for a toroid, so cross-check against manufacturer data for a final design. Toroidal Ae assumes a rectangular (square-cut) annulus — exact for ferrite/tape-wound/nanocrystalline toroids; rounded-cross-section powder toroids (MPP/Kool Mµ) typically have ~15-30% less real Ae than this estimate for the same OD/ID/height. Core loss uses a Steinmetz-style estimate with representative, editable coefficients, not manufacturer-verified loss curves. The CISPR 25 class selector sets a suggested starting target common-mode impedance — a filter-design rule of thumb, not a direct CISPR 25 clause value. CM saturation is driven by a user-editable worst-case DC imbalance/common-mode current, since a correctly wound balanced multi-phase CM choke has its load-current flux cancel. Treat this as a screening/estimation tool, not a substitute for prototype/EMC test-house qualification.

Validated: every closed-form formula (reluctance-model inductance L=µ₀µᵣAₑN²/lₑ, turns count, peak flux density, DM ripple sizing, CM impedance sizing, and Steinmetz core loss) was checked against an independent hand re-derivation and matched exactly, and the inverse pairs (turns↔inductance, ripple↔inductance, impedance↔inductance) all round-trip back to their inputs exactly. Toroidal core geometry (Ae, lₑ, window area) matches the standard published toroid formulas by hand for a 30/15/10 mm OD/ID/height example exactly.

Calculation steps
1. Core geometry (simplified 2D profile)
Ae = h·(OD−ID)/2, le = π·(OD+ID)/2, Wa = π·(ID/2)²
Toroidal profile as entered
Ae = 120 mm², le = 100.53 mm, Wa = 452.4 mm²
2. Reluctance-model inductance
L = µ0·µr·Ae·N²/le
µr = 60 (Nanocrystalline (µi ≈ 60)), N = 30, Ae = 120 mm², le = 100.53 mm
L = 81 µH
3. DM ripple sizing (inverter-leg volt-second balance)
L_required = Vdc·D·(1−D) / (ΔI_target·fsw)
Vdc = 400 V, D = 0.5, ΔI_target = 20 A, fsw = 10 kHz
L_required = 500 µH → N_required ≈ 74.54 turns (using N = 30)
4. Achieved ripple current with N turns
ΔI = Vdc·D·(1−D) / (L·fsw)
L = 81 µH
ΔI = 123.46 A pk-pk, peak current = Idc + ΔI/2 = 361.7 A
5. Saturation check
B_peak = L·I_peak/(N·Ae), margin = B_peak/Bsat
I_peak = 361.73 A (Idc + ripple/2), Bsat = 1.2 T
B_peak = 8.139 T (678.2% of Bsat) — FAIL: exceeds 85% margin
6. Core loss (Steinmetz estimate, representative coefficients)
Pv [mW/cm³] = k·f[kHz]^a·B[T]^b, P_core = Pv·Ve
k=6, a=1.5, b=2, f=10 kHz, B=8.139 T, Ve=12,064 mm³
P_core ≈ 151.622 W (12,568.44 mW/cm³)
7. Winding/window fit check
A_cu = N × conductor cross-section
N=30, conductor = 6 mm², Wa = 452.4 mm²
Fill factor = 39.8% — fits (typical guidance ≤ 40%)