Mastering MLCC DC Bias Derating for Accurate Capacitance Selection

You pick a 10 µF X7R capacitor for a 5 V rail. In the circuit, it measures only a small part of that value. Then the regulator acts up. This same problem happens in labs all over. The cause is dc bias derating.

How do you figure out and compare effective capacitance under DC bias before you pick a part? This post gives you a step-by-step method. You will learn the science behind mlcc dc bias derating, see how package size and voltage rating change the curves, use simulation tools from different vendors, apply one shared derating formula, follow the effects on the whole system, and check results on the bench. That way lets you choose mlcc parts with confidence.

Key Takeaways

• DC bias lowers MLCC capacitance, so always look at the derating curve for your specific part.

• Larger package sizes and higher voltage ratings help keep more capacitance when DC bias is applied.

• Use vendor simulation tools to compare the effective capacitance at your actual operating voltage.

• Use one single formula to combine the effects of DC bias, temperature, aging, and tolerance.

• Check the effective capacitance on the bench and write it down in your BOM for future projects.

MLCC DC Bias Derating: Why Datasheets Mislead

The Physics of Capacitance Loss

Class II and Class III dielectrics lose effective capacitance when the DC voltage goes up. The loss gets bigger with higher dielectric constant. These materials use barium titanate, a ferroelectric ceramic. When you apply DC voltage, domain dipoles line up with the electric field. This lining up stops further dipole motion. The change in charge per voltage change drops, so capacitance falls.

The effect is huge. A 1-µF, 6.3-V-rated X5R MLCC can lose 50–80% of its capacitance at 3 V DC bias, a significant loss common to Class II dielectrics.

Designers must account for DC bias capacitance loss in MLCCs, though this source does not provide a specific misleading example.

Long-term effects make things worse. Domain wall motion causes extra capacitance loss over hundreds to thousands of hours. This aging adds 10–20% loss. The first capacitance drop is bigger, but at lower voltages aging becomes more important.

Reading Derating Curves Correctly

Generic curves are not accurate. You must get the manufacturer's DC bias curve for the exact part number. Follow these steps:

1. Find the operating voltage on the X-axis.

2. Read the capacitance percentage on the Y-axis.

3. Multiply by the nominal value.

Compare parts at the DC bias voltage where the circuit will actually run them, not at zero bias. Use Murata SimSurfing or TDK's DC Bias Model to read effective capacitance at the operating voltage. When AC ripple is very small, the dielectric stays controlled by the DC field. The effective capacitance then matches the lower edge of the DC bias curve.

Package Size and Voltage Rating in DC Bias Derating

The Package Size Trade-Off

A bigger case size keeps more usable capacitance when dc bias derating is the same. Inside a larger package, the stack of dielectric layers is thicker. More ceramic material means the electric field puts less stress on each layer. The dipoles hold out against lining up longer. Your capacitance stays nearer to the nominal value.

Think about two mlcc parts that both have a 10 µF rating. One comes in an 0402 case. The other comes in an 0805 case. At 5 V, the 0402 part may lose most of its capacitance. The 0805 part keeps much more. You pay for that gain with board space. A bigger footprint may not fit your layout. You must weigh the room you have against the capacitance you need.

Voltage Rating and Dielectric Effects

A higher voltage rating moves the derating curve in your favor. A 25 V mlcc holds its capacitance better at 5 V than a 6.3 V part does. The manufacturer builds more dielectric layers or thicker layers into the higher-rated part. That construction stands up to dc bias derating. The trade-off is real. Higher voltage parts cost more and take up more room.

Dielectric type also changes how steep the curve is. X5R, X7R, and X6S each lose capacitance at different rates. A higher voltage rating alone does not tell you the whole story. You must check the specific curve for your exact part number. Frame the choice as a design decision. Balance footprint, cost, and effective capacitance for your target rail. No single factor wins every time.

Comparing Parts for Component Selection

Using Multi-Vendor Simulation Tools

You cannot judge a part only by its label. Multi-vendor selection platforms let you overlay DC bias curves for several parts at the voltage your rail uses. The overlay matters most at your real rail voltage, not at zero bias. That single view shows which parts hold their value and which ones lose it under bias.

Tool	Vendor	DC Bias Curve Support	Overlaid Comparison Feature
EMCompass	Samsung Electro-Mechanics	Yes (thermal derating modeling)	Yes — multi-part comparative overlays
SimSurfing	Murata	Yes — combined C-V-T derating models	Not explicitly stated
Component Characteristics Viewer (CCV) & SEAT	TDK	Yes — dynamic DC superimposition curves	Not explicitly stated
K-SIM	KEMET	Yes — in-circuit capacitance under user-defined rail voltages	Not explicitly stated

Each tool answers a different question. KEMET K-SIM gives you absolute capacitance vs. voltage and temperature for worst-case checks. It models capacitance, ESR, and bias-temperature effects. It also has an energy-density calculator and aging predictor, and exports SPICE models. Murata SimSurfing also plots absolute capacitance against voltage and temperature. It goes further with S-parameters and full impedance curves up to gigahertz. You can also combine multiple capacitors inside the tool to make the impedance response flat. TDK's Component Characteristics Viewer and SEAT tools give dynamic DC superimposition curves. They show how the part acts when the bias voltage changes. Samsung's EMCompass is best for overlays. You can stack several parts on one chart and see where their curves separate.

Building a Side-by-Side Comparison Workflow

For every value you look at, compare at least two vendors and two case sizes. This habit shows derating differences that one datasheet cannot show. Match the effective capacitance at the real operating voltage, not the nominal value. Class II mlcc parts can lose 50–80% of their capacitance under DC bias. Check ESR and ripple current margins at your switching frequency, like 100 kHz or 300 kHz, not at 120 Hz. Use the derating rule that fits each capacitor family, because there is no single rule for all. Estimate how long the part lasts with Arrhenius scaling. Life doubles for every 10°C drop in temperature. Check safety certifications like X/Y class or IEC 60252-1 for motor capacitors when needed. Read the full part number before you decide. It includes capacitance, tolerance, voltage, dielectric, case size, termination, and packaging. A wrong case size or finish breaks the build even if the electrical value is right.

Some parts come without vendor curves. Second-source and unmodeled mlcc parts are in this group. Test them on the bench with a bias sweep, then record the results. Sweep the bias in steps and log capacitance and dissipation factor at each point. A simple table keeps every candidate equal:

Part Number	Nominal Value	Effective Capacitance at Operating Voltage	Package	Voltage Rating	Cost

This structure makes component selection a repeatable decision. You see the real effective capacitance next to footprint and price. So dc bias derating is no longer an afterthought. When a shortage happens, the same table lets you compare a tantalum polymer part with matching effective capacitance to the original mlcc. Write the measured value in the BOM so the next engineer can use your work.

Unified Derating Formula and System Impact

You now know how to read DC bias curves and compare package sizes. But bias voltage is not the only thing that lowers capacitance. Temperature changes, aging over time, and manufacturing tolerance all shrink the value more. You need one formula that mixes every effect together. Then you can figure out the real capacitance your circuit will see.

You can combine all these effects by multiplying the nominal capacitance by the derating factors for DC bias, temperature, aging, and tolerance. Each factor comes from a manufacturer's curve or datasheet. For example, k_DC is the capacitance percentage from the DC bias curve, k_temp from the temperature characteristic curve, k_age from the aging curve, and k_tol is the tolerance factor (e.g., 0.80 for ±20%). Multiply them together to get the effective capacitance your circuit really uses.

Combining DC Bias, Temperature, and Aging

DC bias is usually the biggest factor, but you cannot ignore the others. Temperature and aging add to the loss over the product's life. Picture a 10 µF X7R MLCC running at 5 V DC bias at 85°C after 10,000 hours of use. The DC bias curve might give you only 60% of nominal. The temperature curve might give you 85% at 85°C. The aging curve matters a lot here.

Capacitance loss from aging follows a logarithmic pattern. It is steepest in the first 1000 hours, with a typical drop of 10–20% over hundreds to thousands of hours.

After 1000 hours, the effective capacitance basically levels off because further drops become tiny. But in those first 1000 hours, you lose about 9% of your capacitance just from aging. Multiply that by the DC bias loss and the temperature loss, and a 10 µF part can drop to 4 µF or less. You have to pick a nominal value that covers all these factors together, not just DC bias.

Power Integrity and Loop Stability Consequences

Not enough effective capacitance hurts power integrity. The output ripple voltage in a switch-mode power supply goes up, and the regulator control loop can become unstable. Low capacitance means the reactive ripple from the inductor current charging and discharging the capacitor gets bigger. The voltage swings at the load go past your design budget.

You can see the effect clearly in ceramic capacitor stacks. MLCC parts have very low ESR and ESL, so the ripple you see is almost all from the capacitance value. If your effective capacitance drops below the regulator design target, the ripple voltage rises right away. The regulator compensator, tuned for a certain output capacitor bank, loses its phase margin. You risk oscillation or poor transient response.

The output ripple voltage comes from three sources: capacitance, ESR, and ESL. The capacitance part comes from the inductor current ripple charging and discharging the capacitor. ESR makes a ripple proportional to the inductor current ripple. ESL makes a square-wave voltage spike. In MLCCs, ESR and ESL are very low, so the ripple is almost all capacitance-driven. To lower it, you need higher effective capacitance, higher switching frequency, and minimized inductor current ripple. But none of that works if you pick parts based on nominal values alone.

The formula output must guide your component selection. Write the effective capacitance value in the BOM documentation. Do not list the nominal value without the derating note. Your circuit needs the real capacitance at operating conditions. That is the value you should compare, simulate, and verify. It stops regulator misbehavior and keeps your power rail within tolerance over the product lifetime.

When you document the BOM, include k_DC, k_temp, k_age, and k_tol next to the effective capacitance. That gives the next engineer a complete derating trail. They can check your assumptions, update factors for their conditions, and avoid repeating your measurements. Discipline around dc bias derating becomes more important as rails drop in voltage and rise in current. Start now.

Bench Verification of Effective Capacitance

Measurement Techniques and Fixtures

A precision LCR meter gives you the real answer. Set the meter to sweep DC bias in steps and log capacitance and dissipation factor at each point. This direct method shows you exactly how much capacitance your mlcc loses at your rail voltage. An indirect method works too. You can measure the impedance curve and derive capacitance from it, but the direct bias sweep is faster for most labs.

Fixture setup decides whether your numbers mean anything. Keep leads short to cut stray inductance. Use a proper DC bias adapter or bias tee so the DC source does not load the meter. Allow settling time before each reading. The dielectric needs a moment to respond to the new field. Skip this step and your readings drift.

Community Benchmarks and Field Insights

Field data confirms how far measured values fall from nominal. In ferroelectric ceramics like barium titanate, the dielectric constant can drop significantly under DC bias, leading to a massive loss in effective capacitance. For example, Class II MLCCs can lose 50–80% of their nominal capacitance under bias. A 10 µF mlcc rated 6.3V on a 5V rail may lose 50% to 70% of its capacitance, acting like 3 µF or 4 µF. This deficit causes bulk decoupling problems, high-frequency noise, and voltage spikes. In DC-DC converters, the reduced effective capacitance can push the control loop into oscillation.

Dielectric type changes the picture. C0G/NP0 stays close to nominal with minimal DC bias effect. X7R, X8R, X7S, and X5R all lose capacitance under bias, and the loss depends on the operating voltage. Document measured values for second-source and unmodeled parts. Build an internal derating library so the next project starts with real data, not guesses. This discipline around dc bias derating saves repeated bench time.

Nominal capacitance is just a starting point, not the final spec. Figure out and compare effective capacitance before you lock in a part.

Follow this workflow: check derating curves for your exact part number, weigh package and voltage trade-offs, simulate across vendors, apply the unified formula, and verify on the bench.

Before you buy, run this checklist:

1. Confirm the bias voltage for your capacitance.

2. Apply every derating factor to your capacitance.

3. Compare at least two vendors for effective capacitance.

4. Validate capacitance with measurement.

5. Document the effective capacitance in the BOM.

Rails keep dropping in voltage and rising in current. That trend makes dc bias derating discipline more important, not less. Your mlcc capacitance budget depends on capacitance. Capacitance matters.

 

 

 

 

 

Written by Jack Elliott from AIChipLink.

 

AIChipLink, one of the fastest-growing global independent electronic   components distributors in the world, offers millions of products from thousands of manufacturers, and many of our in-stock parts is available to ship same day.

 

We mainly source and distribute integrated circuit (IC) products of brands such as Broadcom, Microchip, Texas Instruments, Infineon, NXP, Analog Devices, Qualcomm, Intel, etc., which are widely used in communication & network, telecom, industrial control, new energy and automotive electronics. 

 

Empowered by AI, Linked to the Future. Get started on AIChipLink and submit your RFQ online today!