Measuring Capacitance Against Voltage | Hackaday

Ceramic capacitors are almost naughty in the electronic world. If you sprinkle enough things on the circuit, everything will work. However, these ceramic capacitors are not the latest technology: you can find them in radios from the 1930s, and they have an annoying characteristic: their capacitance changes with voltage.

If you rely on RC filters or ceramic covers in the power supply, this will be a problem. What you need is a device that can map capacitance and voltage [limpkin]

.

Of course, capacitance is usually measured by timing the time required to charge and discharge the capacitance through an RC oscillator. This requires at least one known value (in this case a 0.1% resistor), and the unknown capacitance can be calculated by measuring the time required for the circuit to oscillate.

Everything is fine, but how to measure the capacitance under the bias voltage?

 Build a simple circuit around an operational amplifier. The operational amplifier is just a comparator, and the voltage provided by the remaining circuits is proportional to the percentage of charge in the capacitor.

This small project is [limpkin] becomes

,this is

. That is to say, measuring capacitance relative to voltage is not something that any'ol meter can do. We are very happy that [limpkin] can use an easy-to-use tool to measure this phenomenon.

By the way, the circuit suggested by EDN will not work due to the capacitor ESR!

How high must the ESR be to significantly interfere with the measurement?

On the edn circuit? Only a few ohms...

Can you clarify? The charging current of the EDN circuit to it is less than 1mA, which makes its offset per mOhm ESR less than 1uV.

The voltage dependence depends on the type of capacitor used. Type I dielectrics have nothing to do with voltage, or at least I did not see it on any type I capacitors used. Hell, I have some capacitors that have been charged to close to 2KV without changing their capacitance.

If you are using a voltage-dependent capacitor, you can always overestimate it, eliminating most voltage-dependent effects. Obtaining a 50V or 100V rated capacitor is not expensive, nor is it bigger. Designers need to pay attention to this, but as we all know, this is reflected in the data sheet or type specification of the cap.

"It's not expensive to get 50V or 100V rated capacitors, and it's not bigger."

The DC bias effect has not improved much either.

TL; DR-If a 6.3V capacitor is used on the 5V rail, the effective capacitance will be much lower than the factory rating (many of us know this part). The interesting caveat here is that replacing it with a 16V capacitor does not necessarily help, because the voltage coefficient of the capacitor may vary greatly and will not be as advertised as tempco, such as X5R, X7S, etc... .

As you pointed out, the 80% value only applies to the more exotic dielectric materials they must use to get enough capacitance in a miniature SMD package. I assume that you do not get a voltage of 2KV on a 1mm SMD capacitor. :)

Dave talked about this last year.

For most digital circuit decoupling, the important thing is high-frequency impedance (determined by parasitic effects) rather than capacitance. Even with better materials, when you increase the electric field strength (such as shrinking the package size, applying a DC bias equivalent to its rated voltage, etc.), you will get some nonlinear effects.

If you are worried about analog circuits, you can use better capacitors than ceramic capacitors (depending on the application).

agree! I will not put any of those 10uF ceramic covers where strict tolerances are required. For example, bandpass/lowpass/highpass active filtering.

Doesn't the graph obtained from the constant current source charging the capacitor tell you everything?

You mean to see the slope of the charge? of course. The first step will tell you the ESR. Collecting data points fast enough will tell you the slope at any given voltage, which will tell you the capacitance at that voltage.

It cannot tell you all the information, because it can only tell you what happens to DC under a variable voltage. It will not tell you the dissipation factor or frequency-related losses/parasitics, piezoelectric effects, temperature effects, etc.

Of course, this is all the circuit tells you. You are talking about laboratory quality testing.

Go to the Kemet website and try the K-sim tool. This tool allows you to plot the capacitance, ESR, impedance, inductance and frequency of the kemet capacitor product line. Eye-opening!

The super low quality dummies version of these things we saw in the flea market sold for $2

Please be kind and respectful to help make the comment section great. (

)

The site uses Akismet to reduce spam.

By using our website and services, you expressly agree to our placement of performance, functionality and advertising cookies.