STUDIES ON THE PERCENTAGE VARIATION OF RESISTANCE OF POLYMER THICK FILM ( PTF ) RESISTORS

The percentage variation of resistance of polymer thick film (PTF) resistors has been studied. Two different kinds of 
polymer resin, i.e., epoxy resin and polyimide resin, as well as four different kinds of carbon black are adopted in the 
preparations of PTF resistors. Among the four different kinds of carbon black, i.e., e45, MA-100, MA-600 and 
HS-500 (all trade names), HS-500 shows the least variation with temperature changes. This can be attributed to the 
high structure property of HS-500 carbon black. The variation will be lowered down with an increase in the loading 
fraction of carbon black in PTF resistors. Same tendency will be observed in both epoxy and polyimide resin PTF 
resistors. However, the variations are smaller in polyimide resin resistors than in epoxy resin resistors. The percentage 
variations of resistance of PTF resistors, after stored at 120℃ for 1000 hrs, or after 25 cycle thermal shock tests, or 
dipped in a 200℃ solder pot for three seconds, as well as the variations after the short term overload tests, have also 
been measured and studied. The variations are still smaller in polyimide resin PTF resistors. A possible explanation on 
this smaller variation in polyimide resin PTF resistor is also presented.


INTRODUCTION
From the previous studies of the authors, 1,2 it is known that the mixtures of carbon black and polymer resins can be utilized as thick film resistor.Their electrical properties: such as sheet resistivity, 3,4 TCR, and cur- rent noise, 6 have been obtained and studied.However, TCRs are only the measures of resistance changes be- tween two specific temperatures, i.e., 25C for Hot TC and a 5C to -55C for Cold TCR.No information of resistance variations, for the temperatures in between, can be known.So the percentage variations of resist- ance of PTF resistors prepared by different ingredients are measured and studied to try to get a better under- standing about the thermal effects on PTF resistors.
In order to find out the effects of different kinds of carbon black, as well as different kinds of polymer resins, on the stability of resistance values of PTF resis- tors, several experiments have been conducted.Besides the measurements of resistance variations with respect to temperature changes, the percentage variation of resistance after being stored at 120C for 100 hrs, or after thermal shock tests, or the resistance variations after short term overload (STOL) tests or Solder dip- ping tests, are also measured and studied.A possible explanation on the experimental results will also be given.

EXPERIMENTAL PROCEDURE
The experimental procedures will, in general, be the same as those in the previous studies.There are two kinds of polymer resin used, i.e., epoxy resin and polyimide resin.Four different kinds of carbon black, i.e., #45, MA-100, MA-600 and HS-500, are adopted in the preparations of PTF resistors.Their physical properties have also been stated in the previous studies.
The conductor terminations are prepared by the mix- tures of Ag/Cu powder with polymer resins.The resin used in the preparation of conductor termination will be the same with the resin used in the overlapped resis- tor.
The PTF resistor pastes are prepared by mixing of carbon blacks with polymer resins, and xylene or cyc- lohexaneon is adopted as the diluent.The substrates used are polyimide glass fibre laminates.After screen- printing on the substrate, the PTF resistors are dried and then cured.The curing conditions are at 200C/10 hrs for epoxy resin PTF resistors and 250C/10 hrs for polyimide resin PTF resistors.
The percentage variations of resistance are measured on the cured PTF resistors in a thermostat chamber.The variations are measured for every 25C tempera- ture change.Starting from room temperature, however, the lowest temperature is down to 55C and the highest is 125C.There are 15 pieces of each sample point.
Long term resistance drifts are measured and expre- ssed as percentage variations of resistance.The drift tests are carried out in the thermostat chamber by set- ting the temperature at 120C.The resistance values are measured after 1000 hrs.
The effect of solder dipping on the resistance varia- tion is measured at room temperature.By dipping the resistor samples into a 200C solar pot for three sec- onds, then take it out and measure the resistance values till its cooled down.The results are also expressed as percentage variations.
Thermal shock tests are practiced by placing the samples at 100C and 0C for 30 seconds each time, and the process is repeated 25 times.The resistance variations are measured.
The short term overload (STOL) test is carried out by applying voltage to the resistor samples, which are square in shape and the length of each side is 2mm, till the power reaches a value of 250 mW.After 20 sec- onds, remove the voltage source and record the resist- ance variation after it has been in room temperature for 30 minutes.The rating of the resistors are approxi- mately 100 mW/4 mm2.

EXPERIMENTAL RESULTS
Figure 1 and Figure 2 are the relationships between AR/R x 100% and the temperature of carbon black/ polyimide resin PTF resistors.In Figure 1 the contnet of carbon black is 10% (wt.), and in Figure 2 it is 20% Four different kinds of carbon black are used.In both Figure 1 and Figure 2, the HS-500 carbon black shows the smallest variation.The reason might be attributed to the high structure property of HS-500 carbon black.
It is smallest both from 25C to 125C to 55C.
No unique relationship can be obtained between car- bon blacks and the percentage variations.However, if we neglect the results of #45 carbon black temporarily, then a resistor with higher sheet resistivity (MA.100 in this case) will show a higher variation, both from 25C to 125C and from 25C to 55C.The result is co- incident with the works done by other authors, 7'8 and all variations are in negative direction when temperature goes higher.A decrease in the percentage variation of resistance has also been observed when the content of carbon black is increased.
In Figure 3(a) and Figure 3(b), there shows the resist- ance variations in epoxy resin and in polyimide resin PTF resistors.Here the temperature range is from 25C to 125C and the content of carbon black is 20% in weight.In both resistors, HS-500 carbon black still possess the smallest variation.The percentage vari- ations observed in Figure 3(a), show a direct relationshi with the oil absorption property can be taken as a measure of the structure of carbon black, lower struc- ture carbon black, which is higher in sheet resistivity value, will show higher variation.This result is coinci- dent with that discussed on above.The #45 carbon black in Figure 3(b), like it is in both Figure 1 and Figure 2, is still the only exception and will be discussed later.However, almost all other variations are negative in values, and are larger in epoxy resin PTF resistors than in polyimide resin PTF resistors.Figure 4 shows the long term resistance drift charac- teristics.After storing at 120C for 1000 hrs, the drift of polyimide resin PTF resistor is within + 2%, while for expoxy resin PTF resistor is within -6%.The variation of resistance is still larger in epoxy resin resistor than in polyimide resin resistor.
The PTF resistors, after dipped into a 200C solder pot for three seconds, the variations in sheet resistivity is within 1.5% for both types of PTF resistors.The samples shown in Figure 5 are of different loading frac- tions of carbon black.The variations are still larger in epoxy resin PTF resistors.
Figure 6 shows the measured variations of PTF resis- tors after thermal shock tests, i.e., placed at 100C and 0C for 30 seconds each time, and measured after 25 times test.The percentage variations are smaller than 1.5%.The STOL tests carried out on the 2 mm 2 mm resistors (rating 100 mW), by applying 250 mW on it for 20 seconds.The variations measured after 30 minutes power off are shown as Figure 7.The vari- ations are within 1% and it's still larger in epoxy resin PTF resistors.100C/30 sec + 0C/30 sec).
From the experimental results obtained in tile study of percentage variations of PTF resistors with tempera- ture changes, it can be seen that #45 low structure carbon black has shown many distinguishable effects.
In Figure 1, when the temperature is increased from 25C, #45 resistor first showed a negative variation, then it crossed the zero-variation line and toward positive.This might be a result of competition between the Brownian motion of carbon black 1 and the thermal expansion effect of matrix resin. 11At first, When tem- perature is increased from 25C, the Brownian motion of #45 carbon black is dominant, so a negative vari- ation is observed.However, by further increasing the ambient temperature, the effect of the thermal expansion of polyimide resin will exceed that of the Brownian motion, and a positive variation is expected.The effect of thermal expansion of matrix resin will be decreased with an increase in the amount of carbon black, like that shown in Figure 2.However, the only effect in PTF resistor, from -55C to 25C, seems to be the thermal motion of carbon blacks.
The percentage variations of resistance of PTF resis- tors, except those prepared by #45 carbon black, are negative in value.This can be taken as other evidence that the thermal motion of carbon black might be the dominant factor in the resistance variations of PTF resistors, when subjected to temperature changes.
The resistance variations of PTF resistors, after undergoing different tests, are expressed as percentage variations as shown above.The variations are larger than those of the commercial cermet resistors.However, one can see that the PTF resistors prepared by polyimide resin will have smaller variation than those prepared by epoxy resin.The reason that the variation is larger in epoxy PTF resistor than in polyimide resin PTF resistor can be understood by considering the strength of linkage in the polymer resins.In polyimide resin PTF resistors, the rr electrons 12 will resonate in the polymer and release the resonate in the polymer and release the resonant energy, so the linkage in polyimide resin will be more stable.While in epoxy resin, no such resonance occurs, so the linkage in epoxy resin, no such resonance occurs, so the linkage in epoxy resin is not so stable as in polyimide resin.This might be the reason why the percentage variation of resist- ance is larger in epoxy resin PTF resistor than in polyimide resin PTF resistor.
The reason that larger variations in resistance are observed in epoxy resin PTF resistors than in polyimide resin PTF resistors can be attributed to the resonance of rr electrons in polyimide resin.They will release the resonant energy and make the linkage in polyimide resin more stable.While in epoxy resin, no such resonance occurs, so the variations in resistance is larger in epoxy resin PTF resistor under different test conditions.REFERENCES 5. CONCLUSION Among different kinds of carbon black used in the prel: arations of PTF resistors in this study, HS-500 carbon black shows the smallest variation.This can be attri- buted to the high structure property of HS-500 carbon black.
Almost all variations (except those of #45 carbon black) show negative values, implying that the Brownia motion of carbon black are more effective than that of the thermal expansion of the matrix resins.
The percentage variations of resistance after long term drift, solder dipping, thermal shock and short term overload tests are, in general, larger than those of cer- met type thick film resistors, and implies that the stabil ity of PTF resistors is worse than that of cermet resis- tors. 25 FIGURE

FIGURE 2 FIGURE 3
FIGURE 2 AR/R versus temperature of carbon black/polyimide resin resistors (ID.

FIGURE 4
FIGURE4 Long term drift characteristics of PTF resistors at 120C.