SDG #095 MLCC Capacitors Values Not What You Think?! Tests on MLCC

Multilayer ceramic capacitors are the default choice for decoupling, filtering and bulk storage in modern designs, yet the number printed on the reel is often the most optimistic figure you will ever see from the part. This episode puts a spread of 0805 MLCCs on the bench and measures what they genuinely deliver once real operating conditions are applied, rather than trusting the datasheet headline.

The core issue examined is DC bias derating. Class 2 dielectrics such as X5R and X7R rely on ferroelectric ceramics whose permittivity falls as the applied field increases, so a capacitor that measures close to its marked value at zero volts can lose a large fraction of its capacitance by the time it sees a working voltage well inside its rating. The tests show how severe that loss becomes on physically small packages and high capacitance-per-volume parts, and how a higher voltage rating or a larger case size buys back stability. Class 1 dielectrics such as C0G/NP0 are used as the reference point, since they stay effectively flat across bias but cannot reach the same capacitance densities.

For anyone specifying parts, the practical takeaways are worth more than the individual numbers: assume derating rather than hoping for it, treat tolerance and temperature coefficient as separate problems from bias loss, and be sceptical of cheap unbranded capacitor assortment books where the actual dielectric grade is unknown. Bulk decoupling, DC-DC output filters and timing or filter networks each demand a different level of caution, and this testing makes clear where a nominally correct BOM value can quietly leave a design short of the capacitance it needs.