Nondestructive Testing of Multilayer Ceramic Capacitors

1. Field of the Invention

This invention pertains to the field of nondestructive testing and in particular to the nondestructive testing of multilayer ceramic capacitors. Specifically, this method of nondestructive testing is directed to discovery of delaminations and nonvisible cracks in a ceramic capacitor body.

2. Description of the Prior Art

In the prior art, the testing of multilayer ceramic capacitors for either delaminations or nonvisible cracks has been primarily destructive testing. As the name implies, this testing is restricted to samples of a lot and consequently is not effective with respect to individual units within the lot. Though nondestructive testing has been suggested, however, few economically successful techniques have been developed to date. Among the proposed or attempted means of nondestructive testing, there have been at least four alternatives. Ultrasonic scanning is a form of bulk testing which to date has been unable to isolate individual defective units within a lot and consequently has not been very successful. Neutron radiography is a very complex and expensive means of testing but has not yet provided an adequate resolution. The scanning laser acoustic microscope provides a very expensive means of testing but it does rely on human judgment and is relatively slow. Testing by acoustic emission is a very slow process.

The delamination of the layers in a multilayer ceramic capacitor shortens the life of the device, significantly decreases its reliability and causes problems in soldering related to thermal expansion. The delamination of layers in a multilayer ceramic capacitor is not detected by the usual capacitance and resistance tests.

The present invention provides a means to determine the quality of multilayer ceramic capacitors, and in particular the absence of delaminations or other nonvisible cracks in the ceramic material. Certain classes of multilayer ceramic capacitors, when poled by a bias voltage, show piezoelectric effects. Standing acoustic waves are set up corresponding to the dimensions of the ceramic chip capacitor. Where delaminations are present, the resonance pattern is disturbed and the resulting changes can be used to distinguish delaminated capacitors from good or acceptable capacitors. The resonances occur generally between 0.25 MHz and 6 MHz. At these high frequencies testing can be very rapid with a potential for testing 6,000 capacitors per hour.

FIGS. 1 and 2 are micrographs of good (acceptable) and bad (defective) Z5U ceramic capacitors taken in cross-section.

FIGS. 3 and 4 are impedance versus frequency curves for the good and bad Z5U capacitors depicted in FIGS. 1 and 2.

FIGS. 5a and 5b are micrographs of good X7R ceramic capacitors, the pictures being taken in cross-section.

FIG. 6 is a graph of impedance versus frequency for the capacitors of FIGS. 5a and 5b.

FIGS. 7a and 7b are micrographs of bad X7R capacitors taken in cross-section.

FIG. 8 is a graph of the impedance versus frequency curves for the capacitors of FIGS. 7a and 7b.

The present invention is directed to a method of nondestructive testing to determine the presence or absence of delaminations and concealed cracks within multilayer ceramic capacitors manufactured from materials having piezoelectric properties. In terms of the EIA standard designations and codes, the present method of nondestructive testing applies to what are known as class II and class III capacitors. These capacitors basically include materials which exhibit piezoelectric effects. Such capacitors would include capacitors of the types known in the trade as Z5U and X7R. It is possible with capacitors made from these materials to excite mechanical waves through electrical means. The frequency to match the mechanical resonance is determined by the length, width and height of the device made from these materials. When an electric field of one to two times rated voltage is applied, a mechanical wave feeds back a measure of the impedance. When this DC field is applied to the devices individually, standing acoustic waves are set up corresponding to the dimensions of the ceramic chip capacitor. If delaminations or hidden cracks are present, the resonance pattern is disturbed and the resulting changes can be used to distinguish the delaminated capacitors from good (acceptable) capacitors. Resonances occur generally between 0.25 MHz and 5 MHz.

The method of the present invention involves the following steps. The first step is to determine the resonance frequency and the impedance at resonance for a good (acceptable) ceramic capacitor of a specific size, materials and capacitance. A typical multilayer ceramic capacitor of that same size, capacitance and materials specified above is inserted into an impedance measurement bridge. A bias voltage is applied, this voltage being between 1 and 2.5 times the rated voltage of the capacitor. The impedance is then measured as a function of frequency. In the determination of what is a good or acceptable versus a bad or defective ceramic capacitor a high impedance at the resonance frequency indicates a good capacitor. In the case of a bad or defective ceramic capacitor, the impedance will be low.

The figures present some examples which illustrate the method of the present invention. In FIGS. 1 and 2 there are presented illustrations of good and bad capacitors of the type Z5U with a size of 0.10 inches.times.0.05 inches.times.0.05 inches and having a capacitance of 100 nanofarads. It is evident from the micrograph of FIG. 2 that there is a delamination in the lower third of the capacitor. The measured resonance frequency for the capacitor of FIG. 1 yields a f.sub.L of approximately 1 MHz and a f.sub.W,H of approximately 1.7 MHz with the velocity v=5.times.10.sup.3 meters/second. The terms f.sub.L, f.sub.W and f.sub.H specify the resonance frequency along the length, width, and height respectfully of the capacitor. The term "velocity v" means sound velocity in the capacitor material. The resonance f.sub.L occurs for example when the length L times the frequency f.sub.L is equal to the sound velocity V. Turning now to the graphs of FIGS. 3 and 4, in FIG. 3 one can determine that at 1.7 MHz, the impedance of the tested device of FIG. 1 is relatively high and consequently the capacitor is a good capacitor. In FIG. 4 the impedance of the tested device of FIG. 2 at 1.7 MHz is substantially less than that shown for a good capacitor in FIG. 3 and consequently there are delaminations or defects in that capacitor, as shown in FIG. 2.

FIGS. 5a and 5b are micrographs of good (acceptable) multilayer ceramic capacitors of the type X7R having the size of 0.125 inches.times.0.095 inches.times.0.055 inches and a capacitance of 67 nanofarads. The measured resonance frequency for these capacitors of FIGS. 5a and 5b is f.sub.L =0.625 MHz, and f.sub.W =0.875 MHz. Turning to the graph of FIG. 6, one notices that at 0.625 MHz the impedance is very high, and at 0.875 MHz, the resonance frequency corresponding to this width, the impedance is also high. This is true for both specimens. In FIGS. 7a and 7b, there are micrographs of two bad (defective) multilayer ceramic capacitors of the type X7R having the same measurements as the capacitors of FIGS. 5a and 5b. The delaminations in these capacitors are quite obvious from the photographs. Turning now to the graphs measuring the impedance versus the frequency for these bad capacitors, one notes that at the resonance frequencies of 0.625 MHz and 0.875 MHz, the measured impedance is significantly less than that for the good capacitors.

Consequently, a good (acceptable) capacitor can be recognized by a large impedance value at the measured resonance frequencies. A bad (defective) capacitor can be recognized by the change in impedance which appears at resonance frequencies between a good and a bad capacitor. The impedance values at resonance are significantly lowered for a bad capacitor.