MLCC Dielectric Strength Test for Turkey: Five Spellings of 250 Percent Across 24 Walsin Sheets, and a 400 V Part Proved at 480 V

The Walsin MLCC dielectric strength test is written five ways across 24 series datasheets, and its voltage ladder dips: a 400 V part is proved at 480 V, below the 500 V applied to a 250 V part. We read all 24 sheets. Write your inspection limit from the sheet, never from the export.

Venatech Limited, an authorised Walsin distributor in Hong Kong, holds 28 datasheets from Walsin Technology (Walsin), 451 pages, and the 24,026-row capacitor export. This post is for the engineer in Turkey who has to write a proof-voltage line into an incoming-inspection or supplier-approval specification for a Walsin multilayer ceramic capacitor. It is written from the “Dielectric Strength” block of each of the 24 series approval sheets and from the Rated Voltage column of 00_ALL_MLCC_parts_combined.xlsx. Every part number is real. The question is simple: what voltage, as a multiple of rating, does the sheet say each part is proved at?

What does a Walsin MLCC datasheet call its dielectric strength test?

Four different things, across 24 sheets. We opened every one. “Dielectric Strength” heads the test on 20 sheets, “Voltage proof (Dielectric Strength)” on the two safety sheets, “Withstanding Voltage” on the feed-through sheet, and the automotive safety sheet has no heading for it at all. On the ceramic-body sheets the duration is 1 to 5 seconds and the charge and discharge current is limited to 50 mA. The safety sheets use 60 seconds and 0.05 A for the X-class test. The pass criterion is no damage, and on 23 of the 24 sheets also no flash over.

A search for the phrase misses 2 of 24 sheets. A reader who concludes “no proof test” from that search is wrong both times.

How many ways do the 24 Walsin sheets write the 250 percent proof test?

Five, on 17 sheets. The low-voltage rule is 2.5 times rated voltage, and we counted it written as 250%, 250% of rated voltage, 250% rated voltage, ≥2.5 times VDC and 2.5 times of UR. Three more sheets start their ladder above 100 V, one is 300 percent, and the three safety sheets use a fixed voltage. That accounts for all 24.

Four sheets mix two notations in one cell. 03, 10, 14 and 20 write the low-voltage band as a percentage and the next bands as ≥2 times VDC. The Microwave RF sheet also carries an exception we counted in the export: RF02 parts are proved at 300 percent. There are 291 RF02 rows in the combined file, all at 16 Vdc or 25 Vdc, so for them 48 V and 75 V replace 40 V and 62.5 V.

Why is a 400 V Walsin MLCC proved at a lower voltage than a 250 V one?

Because the multiple falls from 2.0 to 1.2 at 400 V and does not climb back until 500 V. Sheet 04_Middle_and_High_Voltage.pdf prints 2 times for 200 to 300 V, 1.2 times for 400 to 450 V, 1.5 times for 500 to 999 V, 1.2 times for 1000 to 3000 V and 1.1 times at 4000 V. Multiply it out and the absolute voltage is not monotonic. A 250 V part is proved at 500 V; a 400 V part at 480 V.

Seven sheets print the 400 to 450 V band: 04, 11, 12, 13, 14, 19 and 20. We multiplied every rated voltage in the Middle to High Voltage series by the sheet’s own ladder.

We have not seen this counted anywhere. 37 rows in the combined export sit in the 400 to 450 V band: 29 in the Middle to High Voltage series (10 at 400 V, 19 at 450 V) and 8 in the Ag-polymer soft termination series (1 at 400 V, 7 at 450 V), whose sheet 19 carries the same ladder. Eleven of them are rated 400 V and are proved at 480 V, while the 867 Middle to High Voltage rows rated 250 V are proved at 500 V. We cannot tell from the sheet whether the 1.2 is deliberate. We would not assume it is a typo either.

Which rated voltages fall between the bands?

One row, and it is a 3.5 kV part. Sheet 04 lists 3500 VDC in its ordering code (352=3500 VDC) and in its high-temperature load (endurance) test, but its dielectric-strength ladder has no band between 1000 to 3000 V and 4000 V. 1812B103K352 is rated 3500 Vdc and has no stated multiple on its own sheet. A 1.2 multiple gives 4200 V; a 1.1 multiple gives 3850 V.

Elsewhere the ladders cover every voltage in the export. We checked each series against its own sheet. General Purpose is the exception worth knowing: sheet 01_General_Purpose.pdf states the test for “≤100V” only, and 14 General Purpose rows are rated 200, 250 or 630 V. We covered that voltage-range gap in the Walsin MLCC rated-voltage audit. The feed-through series (8 rows, for example FT21B473M500 at 50 Vdc) is proved at 300 percent, so 150 V.

Does the Walsin MLCC export carry the proof voltage?

No, and nothing near it. The export has 16 columns; none states a dielectric-strength multiple, a proof voltage, a test duration or a current limit. The Rated Voltage column is the only input, and on the 173 S2 safety rows it holds the impulse voltage, 5000 or 6000 Vdc, not a working rating. The 1075 Vdc that sheet 16 applies to X capacitors, printed there as 4.3UR, appears nowhere in the file.

So a proof-voltage line cannot be built by formula from the export. You need the sheet for your series, the ladder for your voltage, and, from the table above, a check that your voltage is on it. For what the sheets say about the other tests, see MLCC DC bias and voltage derating.

What should a Turkey incoming-inspection specification quote?

The sheet’s own wording, by series, with the arithmetic shown. Three rules come out of 24 sheets: quote the series sheet and its month, state the proof voltage in volts and not as a percentage, and check your rated voltage against the sheet’s bands before you rely on them. A percentage hides the 400 V dip. A volt figure exposes it at once.

1. Write the voltage, not the multiple. “480 V for 1 to 5 seconds, 50 mA limit” is auditable; “120 percent” is not.
2. Name the sheet. 04 and 19 print a 400 to 450 V band; 01 prints none above 100 V.
3. Do not search for the phrase. Two sheets use another heading.

We hold 28 Walsin datasheets, 451 pages, and send the sheet for your part number on request. We will not tell you a part is equivalent to the one you use today. If you are replacing another maker’s capacitor, compare the proof-voltage line field by field, confirm against both datasheets and qualify in your own circuit. The range sits on the MLCC page, you can cross a part with the free cross-reference, and what is in Hong Kong is on ready stock: 1,125 part numbers as at 23 September 2026, a dated snapshot and not a live feed.

The short version

Walsin writes its MLCC proof test under four headings and five spellings of 250 percent on 17 sheets. Four sheets mix percentages and “times VDC” in a single cell. The 400 to 450 V band on seven sheets uses 1.2 times, so 11 rows rated 400 V are proved at 480 V, below the 500 V applied to 867 rows rated 250 V in the same series. One 3500 V row has no band. The export carries none of it. Send us the proof-voltage line from your specification and the rated voltage of the part, and we will return the sheet’s ladder for it. And a question back: does your current limit say 480 V or 120 percent, and would your inspector notice the difference?

Frequently asked questions