THE INFLUENCE OF ELECTRICAL PULSES ON THICK FILM ( DU PONT 1421 BIROX ) RESISTORS

This paper presents data on the effect of electric pulses on thick film resistors made using Du Pont 1421 Birox resistor pastes. Resistance changes during the application of the electric pulses were investigated. Two types of change were observed: reversible and irreversible (i.e. catastrophic). In order to illustrate the causes of these changes, observations of the film on a scanning electron microscope were made. Microcracks were observed in the film, which were mostly responsible for the permanent resistance changes.


INTRODUCTION
The application of electronic systems to pulse work requires a knowledge of the processes taking place in the circuit elements, particularly the resistors, which can be exposed to high temporary stresses.For this reason pulse method investigations were carried out on thick film resistors, t-a based on Du Pont technology and paste.
A bridge method for detecting resistance change was used during the application of the electric impulses. 4This method enabled an analysis of the processes occurring in the film to be made and gave the possibility of the observation of the phenomena for a wide range of stresses and pulse durations.
In this paper an interpretation of the causes of the observed changes is discussed.

EXPERIMENT
Thick Film resistors of initial resistance Rxo(Rxo 902 -+5%) 1 mm.long by 1.5 mm.wide, produced on the basis of Du Pont paste (1421 BIROX) were used.The test pattern is shown in Figure 1.96% A1203 ceramic, 0.55 mm.thick was used as a substrate.
The resistance changes of the sample were studied using the measurement system shown in Figure 2(a). 4  The bridge was activated by rectangular pulses gene- rated from a stabilized voltage supply (U s 140 V).A change of up to 3% of the power dissipated in the test resistor, R x, was obtained by the use of a resistor of the value RN 1.4Rxomtn in series with the investi- gated resistor.
The power dissipated Px in the test resistors is obtained from the relation, Eq. 1, the internal resis- tance of the stabilized supplier being assumed to be approximately equal to zero.Rx ex Us (Rx + RN) (1) This relation is shown in Figure 2(b).The resistance RN was chosen in order to obtain minimum changes of the power dissipated in the test resistors during the investigations.The relation, Eq. 2, was used to obtain the value of RN RN 1 Pxmax Px Rx (2) resistor, Px, as a function of resistance, R x, and defines the terms used.The power dissipated in the test resistor is about 80 times greater than the recommended maxi- mum power.Using a constant power pulse, it was FIGURE 1 Test pattern of resistors.Resistor dimensions: length 1.0 mm.; width 1.5 mm.possible to keep the thermal energy constant which simplified the analysis of the restdting effects.The time measurement of the observed voltage deviations of the bridge were carried out using a Tektronix Memory oscilloscope, the results on which were redrawn to determine the relation: where R xo is the value of the test resistor at t 0.

RESULTS
Figure 3a presents a typical graph obtained on the oscilloscope, and Figure 3b presents the redrawn graph.On Figure 3b three stages of resistance change can be distinguished.In the first stage, up to time t l, reversible changes produced by the thermal behaviour of the film, (Temperature coefficient of resistance), were observed.Irreversible resistance changes occurred after a certain energy supplied to the resistor was reached, i.e. after time tl.This irreversible change was observed in terms of a sudden increase of resistance in the time range tl to t2.Finally, after reaching a maximum in resistance at a time t2, the resistance decreased down to a time t3, when catastrophic des- truction of the resistor occurred.By interrupting the applied pulse at different points along the time axis, scanning microscope observations were obtained.
In the first stage, during which the reversible tem- perature changes were observed, no macroscopic changes were detected.(Figure 5a, b, c.) The second stage, characterized by an irreversible resistance increase, was accompanied by the appearance of microcracks.(Figure 6a, b, c.).The width of the cracks was noted as being comparable with the diameter of the metallic particles in the resistor body.
As time (t) was increased the resistance, and also the density of the microcracks, increased.(Figure 7a, b).The next stage of resistance change was character- ized by a sudden decrease in resistance.The causes of this phenomenon might be due to the improvement of the contacts between the boundaries of the cracks that had been created as well as the effects of local temperature breakdowns.(Figure 8a, b, c.).The sur- faces along microcracks are shown in Figure 8c.These surfaces were obtained by breaking the resistor.Finally exceeding the glass softening temperature (t t3), led to the destruction of the resistor (Figure 9a, b, c.).

CONCLUSIONS
The following conclusions can be drawn from the investigations'-1) Studies conducted on a great number of samples confirmed the reproducibility of the observed effects.i l i ; i i !!J : : i i i i i i !i i !i i i i i i i i i i i i i i i !i !i i i i : : i i i i # l / i ::iiiN %. " ::"i:::.%.N :....::..:...%. . . ..:-"::.. FIGURES g to 9 Out of balance bridge voRage d related SEN micrographs of film at successive stages of the application of electrical pulses to thick film resistors.2) Both reversible (TCR) and irreversible resistance changes occurred.The irreversible changes occurred after a certain energy had been supplied to the film, i.e. after time tl.

FIGUREg t<t
3) The irreversible changes were associated with macroscopic changes in the films, namely the appear- ance of microcracks.These appeared to be due to local stresses caused by the temperature gradient.
4) The improvement of the contacts between the boundaries of the cracks that have been created as well as local temperature breakdowns caused a decreasing of the resistance in the time range from t2 to 5) After the glass softening temperature was excee- ded, total destruction of the films occurred.

FIGURE 2 (
FIGURE 2(a) Measurement system.R x is the test resistor.

FIGURE 2 (
FIGURE 2(b) Power dissipated in the test-resistors as a function of resistance.

FIGURE 3
FIGURE 3 Out of balance bridge voltages obtained during pulse testing (a) Oscilloscope trace.(b) Relative resistance change.

FIGURE 4
FIGURE 4 SEM micrographs showing the film before investigation.(a) Film surface. (b) Film cross section.