Controlling resistance tolerance is a major obstacle in the process development of low cost direct-deposit resistors on organic substrates. State of the art resistance tolerance of direct deposited resistors is typically in the range of 10 - 20% after deposition. Tolerance in processing is frequently either regarded as randomly distributed and is ignored, or is controlled using modern empirical methods such as statistical processing control (SPC) methodologies. The tolerance can be further reduced if the origins of the tolerance have been identified. Unfortunately, neither the origins nor the physics have been identified and understood, such that additional laser trimming process is used to obtain tight tolerance. In this study, the origins of the resistance tolerance in two resistor s...[ Read more ]

Controlling resistance tolerance is a major obstacle in the process development of low cost direct-deposit resistors on organic substrates. State of the art resistance tolerance of direct deposited resistors is typically in the range of 10 - 20% after deposition. Tolerance in processing is frequently either regarded as randomly distributed and is ignored, or is controlled using modern empirical methods such as statistical processing control (SPC) methodologies. The tolerance can be further reduced if the origins of the tolerance have been identified. Unfortunately, neither the origins nor the physics have been identified and understood, such that additional laser trimming process is used to obtain tight tolerance. In this study, the origins of the resistance tolerance in two resistor systems, screen-printing of polymer thick film and electroless nickelphosphorous (Nip) thin film, were investigated with the objective to develop a physically based methodology to reduce resistance tolerance in direct-deposit resistors on organic substrates.

The resistance of a printed thick film resistor is dependent on the resistivities of the material and its printed geometries. In this study, two patterning methodologies, polyester mesh-screen printing and stainless steel stencil printing were investigated. The effect of the methodologies on the printed geometry was investigated, and the as-deposited resistance tolerances from over 600 resistors were evaluated. Experimental results indicated that resistance variations of screen-printed resistors were size dependent. The size dependent behavior of screen-printed resistance tolerances has been identified to be associated with the edge roughness of the printed resistors, and with the mesh size of the screen used. Elimination of the edge roughness using stencil printing gave considerably smaller tolerance than those printed by mesh-screen.

In comparison with broadly ranged polymer thick film, thin film Nip resistors are suited for making resistors for higher frequency applications where the resistance under 1k Ohm. The resistance tolerances of electroless-plated Nip resistors had been reported to be size dependent but the origins of the size dependent behaviors were undetermined. In Nip resistors, geometric variation is determined by the patterning quality and the sheet resistivity variation is determined by the plating quality. In this investigation, both geometric variations and sheet resistivity variations were analyzed as a function of size on different substrates. The plating conditions were identified to have a notable effect on the overall resistance tolerance. Analyses of the cavity depth of the resistor pattern and the corresponding sheet resistivity variation revealed that the plating quality is associated with normalized cavity depth. With this understanding, the study experimentally demonstrated that low resistance tolerances can be achieved using thinner electrode together with smooth substrate. Based on these experimental results, recommendations on process design are given on both patterning methodologies to reduce the as-deposited resistance tolerances.