MLCC DC Bias: Why Your 100 µF Capacitor Isn’t 100 µF

Engineers and buyers usually choose a ceramic capacitor by matching capacitance, case size, voltage rating and dielectric code. However, MLCC DC bias means that the capacitance printed on the reel is not what the circuit gets once voltage is applied.

That gap matters most in high-value X5R parts in small cases, such as a 100 microfarads (µF) capacitor in a 1206 case. Their thin ceramic layers can lose a large share of capacitance at the rail voltage. It’s critical to note that two parts with identical catalog values can behave differently on the board.

This guide explains why that happens, which parts hold capacitance better and how to check a substitute before you approve it.

What Is MLCC DC Bias?

MLCC DC bias is the drop in capacitance that a ceramic capacitor shows when a DC voltage is applied across it. The effect comes from the dielectric itself. Class II ceramics such as X5R and X7R use barium titanate, which gives them high capacitance in a small case. But that same material responds to an electric field by losing capacitance.

The thinner the ceramic layers, the stronger the field across each layer at a given voltage. That is why high-value parts in small cases are hit hardest. Class I dielectrics such as C0G barely change with voltage, but they are only practical at much lower capacitance values.

The nominal value on a datasheet is measured with a small AC signal and no DC voltage. For example, Kyocera AVX lists 1 kHz at 1 root-mean-square voltage (Vrms) for parts up to 10 µF and 0.5 Vrms above that. So the headline number describes the part at rest, not at 5 V or 12 V.

Next step: Compare 10 µF X5R options across manufacturers on Findchips

Why Doesn’t the Dielectric Code Warn You?

The three-character code on an MLCC describes temperature behavior, not voltage behavior. Each code sets a temperature range and a maximum capacitance change across that range.

CodeTemperature RangeCapacitance Change Over Temperature
X5RMinus 55 C to 85 C±15 percent
X7RMinus 55 C to 125 C±15 percent
X7SMinus 55 C to 125 C±22 percent
C0G (Class I)Minus 55 C to 125 C±30 ppm per degree C

KEMET’s X7R datasheet, for instance, specifies that ±15 percent with reference to 25 C and 0 V DC applied. In other words, the code tells you nothing about what happens once the rail is powered. An X7R part is not automatically safer under bias than an X5R part. The case size and rated voltage matter more than the letter.

This is a common blind spot in substitution. A buyer who upgrades from X5R to X7R for temperature margin can assume the swap also improves bias behavior. It does not, unless the new part also has thicker layers.

Next step: Check X7R 10 µF 1206 availability on Findchips

How Much Capacitance Can a Small MLCC Lose?

The loss depends on the part, but it is often large. Electronic Design reported that a 1 µF, 6.3 V X5R capacitor can fall to 0.36 µF at 3 V DC, less than half its rated voltage. That is a 64 percent loss before temperature or aging is counted.

Three factors drive how much a given part loses:

  • Capacitance density. More microfarads in the same case means thinner layers and a stronger field.
  • Case size. A larger case can hold the same capacitance with thicker layers.
  • Rated voltage. Higher-rated parts use thicker layers, so they lose less at the same operating voltage.

These factors combine. A 10 µF, 6.3 V part in an 0402 case running at 5 V can retain only a fraction of its value. Meanwhile, a 10 µF part in a 1206 case rated for 50 V keeps far more at the same 5 V.

Manufacturers publish the curve for each part. TDK, for example, plots a DC bias characteristic on its MLCC characterization sheets. Read that curve at your actual rail voltage, not at zero.

Next step: Look up 10 µF 0402 X5R sources on Findchips

Which MLCCs Hold Their Capacitance Better?

Distributor Rutronik, drawing on TDK lab tests, lists three ways to reduce DC bias loss. All three work the same way: They put thicker dielectric layers behind each microfarad.

OptionWhy It HelpsSourcing Trade-Off
Higher rated voltageThicker layers at the same value and caseCan cost more and narrow the manufacturer pool
Larger case sizeMore room for thicker layersUses more board area and can require a layout change
Several smaller values in parallelLower-value parts keep more capacitanceMore placements and more BOM lines to manage

Each option changes what you buy. A higher-voltage part can come from fewer manufacturers. A larger case can force a footprint change that locks out drop-in alternates. Parallel parts multiply the number of BOM lines that can go into shortage.

So the best choice is the one that delivers the effective capacitance you need with the widest set of qualified sources. Settle that during the design phase, while the footprint can still change.

Next step: Find 10 µF 50 V 1206 X5R options on Findchips

Why Can Same-Spec Substitutes Behave Differently?

Matching capacitance, case, voltage and dielectric code is the standard way to cross-reference an MLCC. But those four values do not include the DC bias curve. Two parts that match on all four can use different layer counts and ceramic formulations. As a result, their effective capacitance at the same voltage can differ.

The manufacturers acknowledge this. TDK’s characterization sheet for the C3216X7R1C106M160AC, a 10 µF X7R part listed as not recommended for new designs (NRND), names a recommended alternate. It then adds that interchangeability is not guaranteed.

For buyers, the fix is to treat effective capacitance as part of the specification:

  • Record the operating voltage next to each bulk or decoupling capacitor on the BOM.
  • Note the minimum effective capacitance the design needs at that voltage.
  • Check each proposed alternate’s DC bias curve against that minimum before it goes on the approved vendor list.

That extra column turns a cross-reference from a guess into a check. It also gives the engineer a clear way to approve a substitute quickly when the original part is short. When you send quotes out, list the approved alternates by MPN, as covered in the RFQ process guide.

Next step: Compare the original and alternate TDK part on Findchips

What Does Current Demand Data Show?

Supplyframe Intelligence data for September shows capacitors holding 51 of the top 100 sourcing slots, down from 61 in July. Within that cluster, activity shifted toward high-value parts: 47 µF and 100 µF X5R in 1206 and 0805 cases, plus 10 to 22 µF X7R and X7S parts rated for 100 V in 1210 and 2220 cases.

One example stands out. The C3216X5R1A107M160AC from TDK – a 100 µF, 10 V X5R part in a 1206 case – climbed from rank 98 in July to rank 38 in September. Within that cluster, activity shifted toward two groups of high-value parts:

The high-value X5R parts and the 100 V parts sit at opposite ends of the DC bias question. A 100 µF rating in a 1206 case is the kind of high-density part that loses the most capacitance under voltage. The 100 V parts in larger cases have thicker layers and more headroom. Activity data shows where buyers are active, not why. But both groups need a DC bias check before they replace anything already on the board.

For the market side of this shift, see The 2026 High-capacitance MLCC Crunch.

Next step: Search MLCC availability across distributors on Findchips

What Should You Check Before Ordering an MLCC?

A short check prevents most effective-capacitance surprises:

  1. Write down the DC voltage the capacitor will see in operation.
  2. Find the minimum capacitance the circuit needs at that voltage.
  3. Read the manufacturer’s DC bias curve for the exact MPN at that voltage.
  4. If the part falls short, step up the voltage rating or case size, or split the value across parallel parts.
  5. Repeat the curve check for every alternate before you approve it.
  6. Approve more than one manufacturer at the final value, case and voltage.

Running these checks during the design phase keeps sourcing options open. For the broader case for that habit, see Sourcing as a Design Constraint.

Next step: Check sources for a 100 µF 1206 X5R MLCC on Findchips

Frequently Asked Questions

What Is DC Bias in an MLCC?

MLCC DC bias is the drop in capacitance that a ceramic capacitor shows when a DC voltage is applied across it. Class II dielectrics such as X5R and X7R are affected most. Within those, high-value parts in small cases lose the most capacitance.

Does X7R Have Less DC Bias Than X5R?

Not by definition. The X7R and X5R codes describe capacitance change over temperature, measured with no DC voltage applied. DC bias depends mainly on layer thickness, which is set by case size, capacitance value and rated voltage.

How Do I Reduce DC Bias Loss in a Ceramic Capacitor?

Choose a part with a higher voltage rating, move to a larger case or split the value across several smaller capacitors in parallel. Each option uses thicker dielectric layers, which keep more capacitance under voltage.

Can I Substitute an MLCC With the Same Capacitance, Case and Voltage?

Only after checking the DC bias curve. Parts that match on capacitance, case, voltage and dielectric code can still have different effective capacitance at the same operating voltage.