DC Fast-Charging Module Passive Components for Norway: 32 of 5,410 Walsin Inductors Clear 10 A, and None Publishes a Saturation Current

DC fast-charging module passive components run out at the filter stage, not at the bus. We corrected the unit defects in Walsin_Inductor_ALL.xlsx and counted: 32 of 5,410 rows clear 10 A, they sit in two series, and not one publishes a saturation current.

---

Venatech Limited, an authorised Walsin distributor in Hong Kong, wrote this from two Walsin Technology (Walsin) parametric exports — Walsin_Inductor_ALL.xlsx and 00_ALL_MLCC_parts_combined.xlsx — re-counted for this post. Norway appears nowhere in our data and we have no Norwegian entity. What we can do is tell a charger-module designer in Oslo or Trondheim where this catalogue stops. We hold 26 MLCC datasheets and zero inductor datasheets, so every inductor figure is export-sourced and flagged. Every part number is real.

What does a DC fast-charging module filter stage actually find on Walsin’s inductor shelf?

A very short list. An off-board charger module — the power block inside a fast-charging post, not the on-board charger in the car — puts tens of amps through its input and output filter stages. We counted the rows that can carry it. 32 of 5,410 clear 10 A. Two series hold all 32: AIP with 20 rows and VEFL with 12. 108 further rows sit between 5 A and 10 A, across 14 series. Every one of the 32 is ±20 percent, and none publishes a saturation current.

Rated current threshold	Rows of 5,410	Series holding them	Rows publishing a saturation current
3 A and above	273	18	8
5 A and above	140	14	6
8 A and above	53	4	0
10 A and above	32	2 — AIP, VEFL	0
15 A and above	11	2	0
20 A and above	3	1 — AIP	0

The highest rated current that survives correction is 30 A, on AIP0630MB-R15M at 0.15 µH, 1.8 mΩ, in a 7.0 × 6.6 mm footprint 2.8 mm tall. We have not seen this counted anywhere.

Why does the rated-current column need correcting before you filter on it?

Because a raw filter returns an 0402 rated at 50 A. We computed the implied self-heating for every row — rated current squared, times the published DC resistance — and discarded any row implying more than 2 W, on the grounds that no chip part of these dimensions dissipates that. That discards 173 of 5,410 rows, across exactly four series: CSME 73, CSMEA 53, WLFI 30 and WQCW 17.

The two mechanisms behind those 173 are both defects we have already published on this line: the CSME and CSMEA rows write DC resistance in a unit that cannot sit beside their current rating, and a handful of WLFI and WQCW rows carry a rated current that is physically implausible for the package. Without the correction, a naive read of this column puts an 0603 at 80 A against 5.5 Ω — about 35 kW in a 1.8 × 1.12 mm part.

What inductance can a charger-module filter get above 10 A?

0.10 µH to 2.2 µH, and 13 values in between. We listed every distinct inductance among the 32 surviving rows: 0.10, 0.15, 0.22, 0.33, 0.47, 0.56, 0.68, 0.82, 1.0, 1.2, 1.5, 1.8 and 2.2 µH. There is nothing above 2.2 µH at 10 A or more anywhere in the 5,410. Thickness across the 32 runs from 1.3 mm to 5.4 mm, and footprints from 4.4 × 4.2 mm up to 7.2 × 6.5 mm.

Part number	Series	Footprint L × W	Thickness	Inductance	Tolerance	RDC	Rated current	Match / mismatch for a 10 A charger-module filter leg
AIP0630MB-R15M	AIP	7.0 × 6.6 mm	2.8 mm	0.15 µH	±20%	1.8mΩ	30A	Match on current, mismatch on inductance — the lowest value in the set
AIP0630MB-R22M	AIP	7.0 × 6.6 mm	2.8 mm	0.22 µH	±20%	2.3mΩ	25A	Match on current, mismatch on saturation — not published
VEFL2C0630-1R0MPA	VEFL	7.1 × 6.5 mm	3.0 mm	1.0 µH	±20%	5.3mΩ	18A	Match on both, mismatch on height — 0.2 mm taller than the AIP rows
AIP0630GX-1R8MPB	AIP	7.2 × 6.5 mm	3.0 mm	1.8 µH	±20%	8.5mΩ	13A	Match, mismatch on DCR — 4.7x the 30 A row
AIP0420MB-R10M	AIP	4.4 × 4.2 mm	1.8 mm	0.10 µH	±20%	4mΩ	13A	Match on current and height, mismatch on inductance and on board area — 2.5x smaller
VEFL5C0654-1R5MPA	VEFL	7.1 × 6.5 mm	5.4 mm	1.5 µH	±20%	6.1mΩ	11.4A	Mismatch on height — 5.4 mm, the tallest inductor in the whole set
AIP0615MB-R47M	AIP	7.0 × 6.6 mm	1.3 mm	0.47 µH	±20%	8.5mΩ	10A	Match on current, and the flattest row above 10 A at 1.3 mm

We will not tell you any of those is equivalent to the part on your incumbent drawing. We will show you where each one differs — and on this shelf the differences are height and DCR, because the inductance choice is already down to 13 values.

What does the high-voltage MLCC shelf above 1,000 Vdc offer an off-board charger?

1,840 rows across eight voltage steps, with a capacitance ceiling that collapses as the voltage rises. Walsin’s Middle to High Voltage series alone holds 4,095 parts spanning 200 V to 4 kV — a figure we have published before. For this post we re-cut the whole 24,026-row MLCC export at 1,000 Vdc and above, setting aside the 173 S2 safety rows because that series’ Rated Voltage field is the impulse voltage, not the working voltage.

Rated voltage	Rows	Case sizes present	Highest capacitance	Thinnest part
1,000 Vdc	861	0805 to 2225	220 nF	0.80 mm
1,250 Vdc	2	1210 only	10 nF	2.50 mm
1,500 Vdc	61	1111 to 2225	100 nF	0.80 mm
2,000 Vdc	511	1206 to 2225	56 nF	0.80 mm
2,500 Vdc	29	1206 to 2220	10 nF	1.25 mm
3,000 Vdc	316	1808 to 2225	18 nF	1.25 mm
3,500 Vdc	1	1812 only	10 nF	2.50 mm
4,000 Vdc	59	1808 to 2225	1.8 nF	1.25 mm

Two things follow. 0805 exists at 1,000 Vdc and nowhere above it — take the bus rating up one step and the smallest footprint becomes 1206, or 1111 in the RF series. And the single 3,500 Vdc part, 1812B103K352, is 2.50 mm thick, which we have published before and which is still the entire 3,500 Vdc shelf.

Will those high-voltage MLCC fit a charger-module stack-up?

Only if seven heights are enough. We read the T column across all 1,840 rows and found exactly seven discrete published thicknesses — 0.80, 0.95, 1.25, 1.60, 2.00, 2.50 and 2.80 mm — and 47 rows that publish no length, width or thickness at all. 283 of the 1,787 rows that do publish a height are 2.50 mm or taller. Above 2,500 Vdc, nothing is thinner than 1.25 mm.

Published thickness	Rows of 1,840	Note
0.80 mm	183	the only sub-millimetre option, and it stops at 2,000 Vdc
0.95 mm	46	a single step, easy to miss in a height sort
1.25 mm	684	the most common height on the shelf
1.60 mm	209	1206 and 1210 only
2.00 mm	382	mostly 1808
2.50 mm	239	includes the only 3,500 Vdc part
2.80 mm	44	the ceiling, all 2220 and 2225
no figure published	47	no stack-up is possible until we ask Walsin

We have not seen the height ladder above 1,000 Vdc counted anywhere. The 47 blank rows sit in three places: 10 Soft Termination 1206 rows, 12 Middle to High Voltage 1206 rows and 9 Microwave 1111 rows at 1,500 Vdc, including RF22N100J152. Two are AEC-Q200 parts — MT31N560J102 and ST31B152J102 — the worst place for a dimension to be missing.

Does the inductor and MLCC data pair up on one charger-module filter stage?

Not symmetrically. We checked both exports against the same stage and the gaps fall in different fields. The inductor side publishes geometry on all 5,410 rows and saturation current on 146. The MLCC side publishes capacitance and voltage on all 1,840 and geometry on 1,793.

Field the filter stage is decided on	Inductor side	MLCC side above 1,000 Vdc	Match / mismatch between the two exports
Value	published on 4,769 of 5,410	published on all 1,840	Mismatch — 641 inductor rows are beads and carry an impedance instead
Tolerance	published on 5,387	published on all 1,840	Match
Footprint and height	published on all 5,410	published on 1,793 of 1,840	Mismatch — 47 MLCC rows publish no dimension
Voltage rating	no column	published on all 1,840	Mismatch — the inductor export has no voltage field anywhere
Current rating	published on all 5,410, 173 needing correction	no column	Mismatch — neither export carries both
Saturation or derating basis	146 of 5,410	DC bias not in the export	Mismatch — the non-linear behaviour is absent on both sides
Example real part	AIP0630GX-1R8MPB	1812B102K202	both real, neither settled by its export alone

The one place the two exports agree is the one that matters least: both publish a tolerance on nearly every row. On the non-linear behaviour — saturation on the inductor, DC bias on the capacitor — both are silent, and both need a curve from Walsin before a 1,000 V stage is signed off. For the capacitor side we hold the sheet: WTC_MLCC_Middle_and_High_Voltage.pdf, and our note on MLCC DC bias and voltage derating.

What should a Norwegian charger-module manufacturer do next?

Treat the two shelves as two different problems. The MLCC shelf is 1,840 rows wide above 1,000 Vdc and we hold 26 datasheets for it; the inductor shelf above 10 A is 32 rows wide and we hold no datasheet at all. Venatech has no Norwegian entity and no stock in Norway — our locations are Hong Kong, Shenzhen, Singapore and Jiangmen, with Chengdu in preparation, and shipment is from Hong Kong with lot and date-code traceability and a certificate of conformity per shipment on request.

1. Fix your current budget before you choose a value. Above 10 A this catalogue offers 13 inductance values in two series, all at ±20 percent. If your filter calculation needs 4.7 µH at 12 A, this line does not have it.
2. Ask for the two curves in writing. The inductance-versus-current curve for any inductor above 10 A, and the DC bias curve for any X7R above 1,000 Vdc. Neither is in either export.
3. Cross the whole module bill of materials in one pass. Paste up to 500 part numbers or drop your Excel or CSV into the free cross-reference tool, and read every stated difference. Then check ready stock — a dated snapshot as at 23 September 2026, not a live feed, with inductors at 0.04% of it.

For the voltage side, see middle and high voltage MLCC, 200 V to 4 kV. The inductor line is on the Walsin inductor page.

The short version

For DC fast-charging module passive components, the constraint is the filter inductor, not the bus capacitor. We corrected 173 rows of unit and plausibility defects in Walsin_Inductor_ALL.xlsx and counted 32 of 5,410 rows clearing 10 A, in two series, all ±20 percent, none publishing a saturation current, with nothing above 2.2 µH at that current. On the capacitor side, 1,840 rows sit at 1,000 Vdc and above once the S2 safety series is set aside, on seven discrete published heights, with 47 rows publishing no dimension at all and 0805 available at 1,000 Vdc and nowhere above it. Saturation and DC bias are missing from both exports. We hold 26 MLCC datasheets and zero inductor datasheets, and we say which figure came from which. Qualification stays with you.

Send us the filter stage, not the part list — the current, the ripple target, the bus voltage and the height you have — and we will come back with what the 32 rows and the 1,840 rows can actually serve, what they cannot, and the two curves we will request from Walsin in writing. One question back: on your last charger module, did the inductor choice come from a datasheet curve or from a rated-current column?

---

Frequently asked questions

---