Capacitors And Capacitance Changes

In circuit applications, the capacitor can be subjected to numerous electrical, mechanical, and environmental stresses. One of the most noticeable effects of these stresses is the phenomena of capacitance variation.

The fact that the capacitance does vary will come as no surprise to most design engineers. The purpose of this bulletin is to describe capacitance changes, and compare the extent of the variation for the common capacitor dielectrics.

The basic formula for capacitance is:

C= KA

Where C = Capacitance

K = Dielectric Constant

A = Effective Area of electrodes

d = Distance between electrodes

"C" varies directly with "A" and "K" and inversely with "d". Any change in "C" must come as a result of some change or combination of changes in "A", "K", or "d".

"A" is set by design and, once a capacitor is made, it is almost impossible for "C" to change due to a change in "A". This is not a normal factor in capacitance variation.

"d" is also set by design. Some small changes in "d" can occur on completed units due to external or internal pressure changes resulting in mechanical movement of the electrodes. This is not usually critical nor does it result in any large variations.

"K" is also initially set by design in the choice of dielectric material used to make the capacitor. Many factors will cause the "K" to change, and this change in "K" will vary for different materials.

The "K" in the basic formula is the effective dielectric constant of the total "space" between the electrodes. This "space" will consist of the dielectric material (or materials if a multiple dielectric design), air, impregnant (if an impregnated unit), and even moisture (if present). All of the dielectrics are effectively in series and therefore the resultant Kr would be:

The major factors that will cause a change in "K" are moisture, voltage, frequency, and temperature.

Moisture

Whenever moisture vapor penetrates into the dielectric of a capacitor, the capacitance will increase somewhat depending on the amount and effectiveness of the penetration, the percent of the total distance between the electrodes that is represented by air, and the percent of the air that is saturated or, in effect, replaced by the moisture.

The following example illustrates the results of extreme moisture absorption:

An increase of capacitance of approximately 27% was seen due to moisture absorption. Increases up to 5% are not uncommon on non-hermetically sealed commercial units when tested to the MIL-STD accelerated moisture tests.

Voltage

With the exception of the General Purpose (high K) ceramics, voltage stress has only a very minor effect on the "K" of the standard dielectric materials. In the case of the high "K" ceramics and AC voltage will cause the "K" to increase while a DC voltage will cause a decrease in the "K". The amount of change will depend upon the original value of K for that particular ceramic mixture.

As an example, for a K = 1200 mix, it is not uncommon to see changes amounting to approximately +20% with 20 VAC (RMS) applied and -30% with 200 VDC applied.

Frequency

The primary area of interest for most applications is from DC to 30 KHz. In this area, the "K" of mica, glass, TFE, polystyrene, and NPO type ceramic dielectrics does not exhibit any measurable change. Polycarbonate film will show a slight decrease in K of about .4% at 30 KHz. Polyester will drop about 1.5 to 2.0%, high K (1200) ceramics approximately 2.0 to 2.5%, and paper-impregnated units between 3.0 and 6.0%, depending on the impregnant.

Temperature

Temperature will have an effect on the "K" of all standard dielectric materials. (The effect can be slight on some dielectrics and significant on others.)

The following charts compare the average curves of various dielectrics relative to the capacitance variations with temperature. Special processing and other factors can be used to alter these curves.

Figure 1 is based on "dry" type film units; that is, no impregnants have been used. Actually, the films will not impregnate, but the use of impregnants as a "filler" is quite common to achieve certain results. When this is done, some alterations in the curves will result.

Figure 2 illustrates quite vividly the impact that the various impregnating materials have on the resultant K of the material between the electrodes. And, to further complicate the picture, the use of various additives to the impregnating material can alter even these curves considerably.

Figure 3 shows the relative curves for other common dielectrics.

In the case of ceramic capacitors, a plot of a "typical" capacitance vs. temperature curve is not feasible since these units can be made to exhibit almost any characteristic desired depending on the dielectric mixture used, processing, method of assembly, and stabilization techniques used following manufacture.