I've been wanting to investigate the actual capacitance of class-II MLCC, operating with the actual workingbias. And yes I know that MLCC manufacturers publish typical data.
Conceptually, the idea is simple: connect a LCR meter to the cap under test while adding in parallel a variable DC power supply.
But even a dimwit like myself knows that this would be a VERY STUPID IDEA. At best, the supply will add a parallel impedance which will load the reading. At worst, the LCR will let out the magic smoke.

What I have struggled to figure out, is how would one isolate the LCR from the supply and viceversa. I am aware that some higher end LCR meters incorporate a DC bias option.
Any ideas how that could be done?

Edit: I also have posted this same question in another forum.

 

Use a signal generator to output pulses at the desired voltage into an RC circuit and use a scope to measure the time constant.

 

BOB; You have given me an idea:
A capacitor will charge linearly (constant rate of voltage increase) when charged from a constant current source.
If the capacitance decreases as the voltage's value increases, the ramp's slope should change.

Hmmmmm....something to experiment with, as soona s I clean my bench from the thing that I am repairing now.

 

Using a differentiator circuit with a triangle source yields a square wave. Then a slow ramp or triangle DC offset generator can vay the DC bias and be used for X sweep on a scope, so you can view the change in amplitude of C = V/R * dt/dV
from Ic = C dV/dt