Abstract

The Direct-Current (DC) testing method in discrete products such as Bipolar Junction Transistor (BJT) and diodes is inadequate for detecting low contact defects. Defects such as disturbed wire or conductive particles with marginal contact on wrong terminals are common in the semiconductor assembly. This is a challenge in today's environment, where quality standards and expectations, especially for automotive consumers, are increasing each day. The defects typically have intermittent behavior, with the understanding that the capability of leakage test on detection depends on the selection of voltage bias. Suppose the voltage bias level is varied to a certain level. In that case, it will help add more resolution on the measurement to differentiate between good and defective parts, on top of the normal DC testing where the manufacturer will measure and remove products with gross short failure and typical leakage failure due to intrinsic properties. This project aims to find the suitability of varying the voltage bias level as an additional DC test coverage. The different voltage biases were tested to analyze the effectiveness of using an actual device with an actual defect. A voltage bias of 80% of device actual breakdown values was found to detect the defect without inducing over rejection. It was found that 9 out of 10 purposely induced wire defects were successfully detected through this test method. Varying the voltage is useful for detecting the low contact defect by the standard test. This is aligned with the quality standard where assembly-induced defects are captured in-house, thus protecting the end customer.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call