Derate ceramic capacitors for DC bias

[ChrisSmolinski]

An FYI for those building circuits... the capacitance of ceramic capacitors varies dramatically as a function of the applied vs maximum voltage rating. Obviously critical in tuned RF circuits, but also a potential (pun intended) issue even just for decoupling caps.

So using a higher rated component will typically give you more capacitance at a given voltage, ie: if you have 5 volts on the cap, a 50 volt part is "better" than 6.3 volt part.

muRata's explanation: https://www.murata.com/en-us/support/faqs/products/capacitor/mlcc/char/0005

Stackoverflow post:
https://electronics.stackexchange.com/questions/280719/how-to-derate-a-ceramic-capacitor-for-dc-bias

[redhat]

This is dependent on the type of dielectric being used.  X7R and other non-NPO types being particularly vulnerable.  Also as I found out, bypass caps can act as piezoelectric speakers.  There are some thing that can be done to avoid this.  For critical applications, RF type NPO's such as ATC parts are recommended, or better yet metal clad mica types.

+-RH

[ChrisSmolinski]

Ahh yes, I am well acquainted with the piezoelectric effect, having been a victim in the past.

In a previous life, I worked in industrial controls. Some of our radiation sensors had a very low output current, in the nA and pA range. And by that I mean the maximum current.  In one application, we were measuring current by accumulating charge over a 100 millisecond period, read by a special 20 bit A/D converter. I recall doing the math and determined each bit on the A/D was about a thousand electrons.  We had to take great care with component selection, and circuit fabrication. Teflon insulated posts were your friend, with lots of dead bug construction.