EV Traction Inverter Passives for Germany: 36 of Walsin’s 2,014 MLCC Above 1,000 Vdc Are AEC-Q200

We counted every row in Walsin’s MLCC export rated 1,000 Vdc or higher: 2,014. Exactly 36 sit in an AEC-Q200 series. EV traction inverter passives on an 800 V bus want 1,600 Vdc at a 2x derating. Walsin’s rungs are 1,500 and 2,000 Vdc, and neither carries one AEC-Q200 part.

---

Venatech Limited, an authorised Walsin distributor in Hong Kong, measured this from Walsin Technology (Walsin) data rather than from a selector. The file is 00_ALL_MLCC_parts_combined.xlsx, 24,026 rows, and the relevant family is the Middle to High Voltage series at 4,095 of them. Every part number below is real, every figure is a count of part numbers rather than pieces, and the ones we cannot confirm are flagged as such.

How much of Walsin’s high-voltage MLCC shelf sits above 1,000 Vdc?

Just under a third. We parsed all 4,095 rows of the Middle to High Voltage series and grouped them by the Rated Voltage string: 1,261 rows are 1,000 Vdc or above, resolving to 1,146 distinct part numbers. The series name says 200 V to 4 kV and the data agrees, but the rungs are not evenly loaded. 1,000 Vdc holds 612 rows and 1,500 Vdc holds 29.

Rated voltage	Rows in the series	Case sizes present	Highest capacitance
1,000 Vdc	612	0805, 1206, 1210, 1808, 1812	100 nF
1,500 Vdc	29	1206, 1210, 1808, 1812	22 nF
2,000 Vdc	356	1206, 1210, 1808, 1812	22 nF
2,500 Vdc	28	1206, 1808, 1812	2.2 nF
3,000 Vdc	212	1808, 1812	6.8 nF
3,500 Vdc	1	1812	10 nF
4,000 Vdc	22	1808, 1812	1.8 nF

The shape of that table is the whole design problem. The two rungs a traction inverter designer reaches for first — 1,500 and 2,000 Vdc — are a 29-part shelf and a 356-part shelf, and both stop at 22 nF. We have not seen the 1,500 Vdc rung counted anywhere, and at 29 part numbers across four case sizes it is thinner than most engineers assume when they write 1,500 V into a requirement. The series datasheet we hold is 04_Middle_and_High_Voltage.pdf, and it is sent on request.

What does a 2x derating on an 800 V bus cost you in case size?

The entire 0805 footprint, and 78% of your capacitance. An 800 V bus at a 2x derating wants 1,600 Vdc, and 1,600 Vdc does not exist in this series — the rungs either side are 1,500 and 2,000 Vdc. Round down and you are below your own rule. Round up to 2,000 Vdc and we measured what you give away.

Moving from 1,000 Vdc to 2,000 Vdc	1,000 Vdc	2,000 Vdc	What changes
Rows available	612	356	42% fewer part numbers
Smallest case size	0805 (111 rows)	1206	0805 disappears entirely above 1,000 Vdc
Capacitance ceiling	100 nF, at 1812	22 nF, at 1812	a 4.5x reduction
Ceiling at 1206	22 nF	3,300 pF	6.7x reduction at the same footprint
Ceiling at 1210	47 nF	8,200 pF	5.7x reduction

We checked the 0805 rows directly: all 111 of them sit at 1,000 Vdc and the highest capacitance among them is 10 nF, on 0805B103K102. Above 1,000 Vdc, 0805 is absent from this series. So on an 800 V bus the honest statement is that a 1.25x derating keeps 0805 on the table and a 2x derating does not, and if your layout depends on an 0805 pad the derating rule and the footprint are in direct conflict.

The 1206 column is where the cost bites hardest. 1206B222K152 at 2,200 pF and 1,500 Vdc and 1206B332K202 at 3,300 pF and 2,000 Vdc are the kind of part you end up paralleling, and parallelling high-voltage ceramics changes your current sharing and your failure mode, not just your line count. Both figures are read from the export; the derating curve behind them is in Walsin’s own application guide, not in any column.

How many MLCC above 1,000 Vdc are in an AEC-Q200 series?

Thirty-six of 2,014. We widened the measurement past the Middle to High Voltage series and read every one of the 24,026 MLCC rows: 2,014 rows are rated 1,000 Vdc or higher, spread across nine series. Exactly 36 of them sit in MT or ST, the two series Walsin qualifies to AEC-Q200. That is 1.8%. The rest are commercial, soft-termination SH and SG, Big Sizes, Microwave RF, the S2 safety series, or MG, which Walsin itself names Without AEC-Q200 Certification.

Series holding rows at 1,000 Vdc or above	Rows	AEC-Q200 qualified
Middle to High Voltage (200 V–4 kV)	1,261	no
Big Sizes 1825/2220/2225	258	no
Soft Termination SH (Ag-poly)	256	no
Safety Certificated X1/Y2 (S2)	173	no
Automotive MT	20	yes
Automotive Soft Term ST	16	yes
Microwave RF	15	no
Soft Termination SG (Cu-poly)	12	no
Automotive MG	3	no — Walsin’s own series title says so

That 36 is the number a German traction inverter team should carry into its first supplier conversation. It is not a shortage story and we are not implying any part is unobtainable — it is a catalogue-shape story. Above 1,000 Vdc, automotive-qualified ceramic is a short list at every manufacturer, and this is what Walsin’s published catalogue looks like when you count it instead of filtering a web selector. The MT and ST series datasheets, 11_Automotive_MT.pdf and 12_Soft_Term_Automotive_ST.pdf, are both in our 451 pages.

Is there anything AEC-Q200 between 1,250 and 3,000 Vdc?

No. We read all 36 AEC-Q200 rows above 1,000 Vdc and the voltages are 1,000 Vdc on 29 of them, 1,250 Vdc on two, and 3,000 Vdc on five. There is nothing in MT or ST at 1,500 Vdc, nothing at 2,000 Vdc and nothing at 2,500 Vdc. So an 800 V bus with a 2x derating rule and an AEC-Q200 requirement has no Walsin ceramic that satisfies both: you go to 1,250 Vdc at 1.56x, or you jump to 3,000 Vdc at 3.75x and accept a 2220 footprint.

Part number	Series	Size, dielectric, capacitance	Rated voltage	Thickness	Match / mismatch vs 800 V bus, 2x derating, AEC-Q200
MT32N103J12E	MT, AEC-Q200	1210 NPO 10 nF ±5%	1,250 Vdc	2.50 mm	Mismatch — 1.56x derating, short of a 2x rule
ST32N103J12E	ST, AEC-Q200, soft term	1210 NPO 10 nF ±5%	1,250 Vdc	2.50 mm	Mismatch — same voltage; adds soft termination for board flex
ST55B103K302	ST, AEC-Q200, soft term	2220 X7R 10 nF ±10%	3,000 Vdc	2.50 mm	Match on voltage at 3.75x; mismatch — 2220 footprint
ST55B222K302	ST, AEC-Q200, soft term	2220 X7R 2,200 pF ±10%	3,000 Vdc	2.50 mm	Match on voltage; mismatch — 2,200 pF is a snubber, not a link
ST43N101K302	ST, AEC-Q200, soft term	1812 NPO 100 pF ±10%	3,000 Vdc	not stated	Match on voltage, smaller footprint; mismatch — 100 pF, and the row has no thickness
MT32N153J102	MT, AEC-Q200	1210 NPO 15 nF ±5%	1,000 Vdc	2.50 mm	Mismatch — 1.25x derating only; highest MT capacitance above 630 Vdc
ST46B104K102	ST, AEC-Q200, soft term	1825 X7R 100 nF ±10%	1,000 Vdc	2.00 mm	Mismatch — 1.25x derating; the only AEC-Q200 part at 100 nF and 1,000 Vdc
1812B223K202	Middle to High Voltage, commercial	1812 X7R 22 nF ±10%	2,000 Vdc	2.50 mm	Match on a 2.5x derating; mismatch — not an AEC-Q200 series

We will not tell you a part is equivalent. We will show you where it differs, and here the difference is a choice between the derating rule and the qualification, made explicit.

Why will a Walsin high-voltage part number not fix your board height?

Because the code does not carry thickness. We counted duplicates in the series: 208 of its 3,871 distinct part numbers appear on more than one export row — 432 rows for 208 codes — and 202 of those 208 differ only in the T field. Inside the 1,000 Vdc band and above, 101 codes are duplicated and 100 of them differ on thickness. 1206B472K152 at 1,500 Vdc is published at both 1.25 mm and 1.60 mm under one part number.

For a traction inverter that is not a cosmetic defect. These parts sit under busbars, inside potted assemblies and beneath cold plates, where 0.35 mm of unexpected height is a mechanical interference rather than a tolerance. Across the whole above-1,000 Vdc set the published thickness runs from 0.80 mm to 2.50 mm. State a maximum height on the order line and ask us to confirm which row you are getting. That is a question the part number cannot answer and a human can.

What is the single 3,500 Vdc part, and how thick is it?

One part number, 2.50 mm. 1812B103K352 is the only row at 3,500 Vdc in all 4,095 — X7R, 10 nF, ±10%, 1812, with published dimensions of 4.50 mm by 3.20 mm and a thickness of 2.50 ±0.30 mm. That fact is ours and already published; we re-measured it today and it is unchanged. The next rung up, 4,000 Vdc, holds 22 rows and tops out at 1.8 nF.

So the extreme end of this series is a single part, and a designer who writes 3,500 Vdc into a specification has written a single-source line without knowing it. Say so early: that is the kind of line where a lead time and a second source have to be settled before layout. Ask us for the datasheet and we will send 04_Middle_and_High_Voltage.pdf with the row it came from marked.

What should a German traction inverter team do with this?

Fix the derating rule and the qualification requirement before you touch a part number. We measured the shelf so that you can see which of your three constraints — voltage margin, AEC-Q200, footprint — has to give, and then cross the whole list in one pass rather than one line at a time.

1. Decide which constraint moves. If AEC-Q200 is immovable, the Walsin ceramic answer on an 800 V bus is 1,250 Vdc at 1.56x or 3,000 Vdc at 2220. If the derating rule is immovable, the 2,000 Vdc shelf is 356 commercial rows and the qualification has to come from elsewhere in your stack.
2. Cross the whole BOM at once. Paste up to 500 part numbers or drop your Excel or CSV into the free cross-reference tool. It is matched against Walsin’s published catalogue, returns candidates field by field with every difference from your original stated, and lets you filter by automotive grade or for soft termination.
3. Check what is physically in Hong Kong on ready stock before you commit a build slot. That page is a dated snapshot as at 23 September 2026, not a live feed, and 149 of its 1,125 part numbers are in this Middle to High Voltage series.

For the series itself, see Middle and high voltage MLCC: 200 V to 4 kV. For why a rated voltage is not a working voltage, see MLCC DC bias and voltage derating. The full range is on the Walsin MLCC range page.

The short version

Walsin’s Middle to High Voltage MLCC series holds 4,095 rows from 200 V to 4 kV, and 1,261 of them are 1,000 Vdc or above. Across the whole 24,026-row export, 2,014 rows reach 1,000 Vdc and exactly 36 sit in an AEC-Q200 series — 29 at 1,000 Vdc, two at 1,250 Vdc, five at 3,000 Vdc, and nothing at 1,500, 2,000 or 2,500 Vdc. An 800 V bus at a 2x derating wants 1,600 Vdc, which is not a rung; rounding up to 2,000 Vdc costs you the entire 0805 footprint and drops the capacitance ceiling from 100 nF to 22 nF. 202 part numbers in the series are published at two thicknesses, so the code cannot fix your board height. One part exists at 3,500 Vdc and it is 2.50 mm thick. Qualification stays with you.

Send us the high-voltage ceramic lines off your traction inverter BOM with your derating rule written down, and we will come back with Walsin candidates field by field, the AEC-Q200 ones marked, the thickness rows separated, and the Hong Kong position as at 23 September 2026 beside each. If your rule says 2x on an 800 V bus, which do you break first — the margin, the qualification, or the footprint?

---

Frequently asked questions

<!– wp:rank-math/faq-block {"questions":[{"id":"faq-question-01","title":"How many Walsin MLCC are rated 1,000 Vdc or above?","content":"2,014 rows across the whole 24,026-row export, of which 1,261 are in the Middle to High Voltage series and resolve to 1,146 distinct part numbers. The remainder sit in Big Sizes, soft-termination SH and SG, the S2 safety series, Microwave RF and the three automotive series. We counted it from the export rather than a selector.","visible":true},{"id":"faq-question-02","title":"Which Walsin MLCC above 1,000 Vdc are AEC-Q200?","content":"Thirty-six rows, 20 in the MT series and 16 in the ST soft-termination series. Twenty-nine are at 1,000 Vdc, two at 1,250 Vdc — MT32N103J12E and ST32N103J12E — and five at 3,000 Vdc, all in ST. There is nothing AEC-Q200 at 1,500, 2,000 or 2,500 Vdc in the data we hold.”,”visible”:true},{“id”:”faq-question-03″,”title”:”What rated voltage should an 800 V EV bus use?”,”content”:”That is your decision and your derating rule, not ours. Arithmetically a 2x rule asks for 1,600 Vdc, and Walsin’s rungs are 1,500 and 2,000 Vdc, so there is no exact landing. Read the derating guidance in Walsin’s own application guide, apply your own margin, and ask us for the qualification report before design-in.”,”visible”:true},{“id”:”faq-question-04″,”title”:”Does moving from 1,000 Vdc to 2,000 Vdc cost case size?”,”content”:”Yes, and measurably. We checked all 111 of the 0805 rows in the series and every one is at 1,000 Vdc, so 0805 is absent above it. The smallest case at 2,000 Vdc is 1206. The capacitance ceiling falls from 100 nF to 22 nF, and at 1206 from 22 nF to 3,300 pF.”,”visible”:true},{“id”:”faq-question-05″,”title”:”Why does one Walsin part number show two thicknesses?”,”content”:”Because the ordering code does not encode thickness. We counted 208 of the 3,871 distinct part numbers in the series appearing on more than one export row, and 202 of those differ only in the thickness field. 1206B472K152 is published at both 1.25 mm and 1.60 mm. State a maximum height on the order line.”,”visible”:true},{“id”:”faq-question-06″,”title”:”What is the highest-voltage Walsin MLCC in the export?”,”content”:”4,000 Vdc, on 22 rows, with a capacitance ceiling of 1.8 nF at 1812. Below that, 3,500 Vdc is a single part number, 1812B103K352, at 10 nF, X7R, 1812 and 2.50 mm thick. The 3,000 Vdc rung is larger at 212 rows and reaches 6.8 nF.”,”visible”:true},{“id”:”faq-question-07″,”title”:”Do you hold the datasheet for these parts?”,”content”:”Yes. We hold 28 Walsin datasheets, 451 pages, and the Middle to High Voltage, MT and ST series are all covered. Ask for the sheet covering your part number and it will be in your inbox. What we do not hold is a per-part qualification report, PPAP or IMDS entry, and we say so rather than implying coverage.”,”visible”:true}],”listStyle”:””,”titleWrapper”:”h3″,”titleCssClasses”:””,”contentCssClasses”:””,”sizeSlug”:””} –>

---