MLCC DC Bias in South Korea: How Much Capacitance an X7R Loses, and the 179 Walsin 1 µF X7R Parts Venatech Can Cross Across 12 Voltage Ratings

MLCC DC bias can take a large share of an X7R’s value, and the X7R code’s ±15% is measured with no voltage applied. For South Korean designs the fix is headroom: Walsin Technology (Walsin) lists 179 X7R parts at 1 µF across 9 cases and 12 voltage ratings, crossed and supplied by Venatech, the authorised Walsin distributor in Hong Kong.

---

Venatech Limited, an authorised Walsin distributor in Hong Kong, wrote this for a power or hardware engineer in South Korea who has watched a 1 µF X7R behave like something much smaller on the bench. We parsed Walsin’s MLCC export, 00_ALL_MLCC_parts_combined.xlsx (24,026 rows, 23,601 distinct part numbers), read Walsin’s own General Purpose datasheet and MLCC application guide, and checked our stock list as at 23 September 2026. The market statement names its source and the day we fetched it, 11 November 2026.

Why is MLCC DC bias a live question in South Korea this quarter?

Because Korea’s own MLCC maker says the demand is for high-capacitance, high-reliability parts. According to Money Today on 28 October 2026, Samsung Electro-Mechanics posted record third-quarter revenue of 3.7633 trillion won, up 30.3%, with its component division, which makes MLCCs, at 1.883 trillion won, and it credited demand from AI servers.

The same report says the company is expanding supply of high-capacitance, high-reliability MLCCs for AI servers. Those are the paper’s figures, not ours. The engineering point we draw is ours: every power rail that leans on high-value Class II capacitors runs into MLCC DC bias, and a Korean board team that second-sources those parts needs to know what the second part does at the rail voltage, not at zero volts. We say nothing about any company’s purchasing.

How much capacitance does an X7R MLCC lose under DC bias?

More than most first-time users expect, and the amount depends on the part. Samsung Electro-Mechanics’ own MLCC catalogue (Part I, December 2025) states that the capacitance of a high dielectric constant Class II MLCC “is changed by applied DC & AC voltage”, and plots an X5R 10 µF 6.3 V example on an axis that runs down to minus 100%.

The physics is the same for X7R. The barium titanate dielectric that gives Class II parts their high capacitance loses permittivity as the DC field across each layer rises. So the loss grows with the rail voltage as a fraction of the rated voltage, and it grows as the dielectric layers get thinner, which is why a high value squeezed into a small case is the worst case. The same catalogue says designs with narrow tolerance should account for DC voltage, temperature and aging together, and that heat treatment at 150 °C for 1 hour resets aging. Walsin’s General Purpose datasheet applies the same de-aging step, 150 °C for 1 hour, before every Class II measurement. Read the bias curve for the exact part number; never carry one maker’s curve over to another maker’s part.

What does the X7R code actually guarantee, and what does it leave to the curve?

Only temperature. Walsin’s General Purpose datasheet, 01_General_Purpose.pdf, specifies X7R as within ±15% capacitance change from -55 to +125 °C, measured “with no electrical load”, and Samsung’s catalogue likewise specifies its X7R temperature limit without DC bias.

So a part marked X7R K (±10%) can meet every line of its datasheet and still deliver far less than 1 µF on a 12 V rail. Samsung’s catalogue even lists an “X7R(S)” code, X7R with the temperature characteristic specified under 0.5 times rated voltage, which tells you the plain X7R code is a zero-bias promise. Walsin’s MLCC application guide adds a second reason to keep headroom: it says failure rate follows a cube law of voltage, so running a part at working voltage Uw on rated voltage UR scales the failure rate by (Uw/UR) cubed. Halving the voltage stress cuts that factor to one-eighth.

Which Walsin 1 µF X7R parts give you DC bias headroom?

A wide grid. We counted 179 distinct Walsin X7R part numbers at 1 µF in 00_ALL_MLCC_parts_combined.xlsx, spread over 9 case sizes from 0402 to 2225 and 12 voltage ratings from 6.3 V to 630 V, and 11 of them were on our stock list as at 23 September 2026.

We checked each one against our snapshot. The table shows the 11 stocked parts, ordered by the headroom they give a 12 V rail. The ratio column is ours: rail voltage over rated voltage, the number the bias loss and the cube law both depend on.

Walsin part	Case	Rated V	Tol.	Thickness (mm)	12 V rail as % of rating	Match / mismatch for a 12 V, 1 µF brief	On the shelf, 23 Sep 2026
0402B105K6R3	0402	6.3 V	±10%	0.50	190%	Mismatch: rating below the rail	yes
0603B105K100	0603	10 V	±10%	0.80	120%	Mismatch: rating below the rail	yes
0805B105K100	0805	10 V	±10%	1.25	120%	Mismatch: rating below the rail	yes
0603B105K160	0603	16 V	±10%	0.80	75%	Match on rating, little bias headroom	yes
0603B105K250	0603	25 V	±10%	0.80	48%	Match	yes
0805B105K250	0805	25 V	±10%	1.25	48%	Match, larger case	yes
1206B105K250	1206	25 V	±10%	1.15	48%	Match, larger case	yes
1206B105K500	1206	50 V	±10%	1.60	24%	Match, strong headroom	yes
1206B105J500	1206	50 V	±5%	1.60	24%	Match, tighter tolerance	yes
1210B105K500	1210	50 V	±10%	1.25	24%	Match, strong headroom	yes
1210B105K101	1210	100 V	±10%	2.00	12%	Match, most headroom	yes

Beyond the shelf, the same grid holds AEC-Q200 MT parts such as MT21B105K500 and MT31B105K101, and soft-termination SH parts such as SH21B105K500. A match here is a match on rating and value from the export; the delivered capacitance at 12 V still comes from the bias curve. Our guide to MLCC DC bias and voltage derating explains how to read one.

How should a Korean design team size an X7R for a 12 V rail?

Size for the capacitance you need at 12 V, not at zero. Our working rule from the 179-part grid: pick a rating at least twice the rail, so 25 V or more for 12 V, then step up a case size before you step up the value, and confirm on the exact part’s bias curve.

1. Write the required effective capacitance at the rail voltage and temperature, not the nominal value.
2. Read the bias curve for the candidate part number from the maker’s own data, and request Walsin’s characteristic data through us for a Walsin part.
3. If the effective value falls short, move up a case: at 1 µF, Walsin lists 25 V parts in 0603, 0805, 1206, 1210, 1812 and 2220.
4. Then consider a higher rating: 50 V parts exist at 1 µF in 0603, 0805, 1206, 1210, 1812, 2220 and 2225 in the export.
5. Measure it on your own board. A bench LCR meter with DC bias applied settles the question faster than any catalogue.

The NP0 vs X7R vs X5R guide covers when to leave Class II altogether.

How can Venatech help a Korean design team pick the Walsin part?

How Venatech helps: we cross your current part and show the bias-relevant fields side by side. Venatech Limited is Walsin’s authorised distributor in Hong Kong; our free VenaMatch tool decodes 18 manufacturers’ part numbers and crosses 800,000+ parts to Walsin, field by field, with every difference stated.

Paste up to 500 part numbers into the cross-reference tool, including Samsung, Murata or TDK numbers, or upload a whole BOM. The ready-stock list is a Hong Kong snapshot dated 23 September 2026 and holds the 11 parts in the table above; for anything else in the 179-part grid, request a quote and we source it through our authorised Walsin channel. We serve small lots from Hong Kong and give no lead-time or stock promise. The Walsin MLCC range lists the families, and our note on crossing Samsung MLCC families to Walsin covers the part-number side. We will not tell you a part is equivalent. We will show you where it differs.

The short version

An X7R’s ±15% is a zero-bias, temperature-only promise; DC bias is extra and part-specific. For a 1 µF part on a 12 V rail in South Korea, Walsin’s grid of 179 X7R part numbers lets you buy headroom by voltage and case, and 11 sat on our Hong Kong shelf on 23 September 2026.

Send us the rail voltage and the effective capacitance you need, and we will return Walsin candidates with rating, case and thickness side by side. What does your 1 µF read at 12 V?

Related reading

• MLCC DC bias and voltage derating
• NP0 vs X7R vs X5R MLCC dielectric
• Crossing Samsung MLCC families to Walsin
• Free cross-reference tool and ready-stock snapshot

Last reviewed 11 Nov 2026.

Sources

Fetched 11 November 2026 unless stated.

• Money Today, “Samsung Electro-Mechanics posts record Q3 results as AI boom drives operating profit up 2.7-fold”, 28 October 2026: https://www.mt.co.kr/en/industry/2026/10/28/2026102811552838741
• Samsung Electro-Mechanics, Multilayer Ceramic Capacitors, Part I, Commercial/Industrial catalogue, December 2025: https://product.samsungsem.com/resources/file/product-catalog/MLCC_2512.pdf
• Walsin General Purpose MLCC datasheet, held as 01_General_Purpose.pdf: https://www.passivecomponent.com/wp-content/uploads/datasheet/WTC_MLCC_General_Purpose.pdf
• Walsin MLCC application guide, January 2022, held as 00_MLCC_Application_Guide.pdf: https://www.passivecomponent.com/wp-content/uploads/datasheet/MLCC_application_guide.pdf
• Walsin MLCC export 00_ALL_MLCC_parts_combined.xlsx and the Venatech stock list as at 23 September 2026, counted by us.

---

Frequently asked questions

---